polish journal of microbiology · polish journal of microbiology 2007, vol. 56, no 3, 151Œ152...

P O L S K I E T O W A R Z Y S T W O M I K R O B I O L O G Ó WP O L I S H S O C I E T Y O F M I C R O B I O L O G I S T S

Polish Journal of Microbiology

formerly

Acta Microbiologica Polonica

2007

POLISH JOURNAL OF MICROBIOLOGY

(founded in 1953 as Acta Microbiologica Polonica)

www.microbiology.pl/pjm

EDITORIAL OFFICE

EDITOR IN CHIEF: Miros³awa W³odarczyk

EDITORS: Ryszard ChróstHanna DahmJaros³aw DziadekAnna Skorupska

EDITORIAL SECRETARY: Anna Kraczkiewicz-Dowjat

POSTAL ADDRESS: Polish Journal of MicrobiologyMiecznikowa 102-096 Warsaw, POLAND

CONTACT: Phone: (48) 22 554 1318Fax: (48) 22 554 1402E-mail: Editorial Office ([email protected])Editor in Chief ([email protected])

EDITORIAL BOARD President: Andrzej Piekarowicz (Warsaw, Poland)

Waleria Hryniewicz (Warsaw, Poland) Zdzis³aw Markiewicz (Warsaw, Poland)El¿bieta K. Jagusztyn-Krynicka (Warsaw, Poland) Gerhardt Pulverer (Cologne, Germany) Miros³aw Kañtoch (Warsaw, Poland) Geoffrey Schild (Potters, Bar, UK)Donovan P. Kelly (Coventry, UK) Wac³aw Szybalski (Madison, USA)Józef Kur (Gdañsk, Poland) Torkel Wadstrom (Lund, Sweden)Tadeusz Lachowicz (Wroc³aw, Poland) Jadwiga Wild (Madison, USA)Wanda Ma³ek (Lublin, Poland)

PUBLISHER: POLISH SOCIETY OF MICROBIOLOGISTS

Published quarterly with the financial support of the Ministry of Science and Higher Education

SUBSCRIPTION:For information for Polish subscribers contact Secretary of Polish Society of Microbiologists, Che³mska 30/34,02-725 Warsaw, Poland; phone: (48) 22 841 3367, fax: (48): 22 842 2949, e-mail: [email protected]

For information for foreign subscribes contact �Ars Polona� � e-mail: [email protected]; phone:+ 48 (22) 509 8665 also www.arspolona.com.pl

Cover illustration: Airborne bacterial and fungal colonies (Jerzy Pi¹tkowski, Uniwersity of Wroc³aw, Poland)

Typesetting and print: Publishing House Letter Quality, 01-216 Warsaw, Brylowska 35/38Circulation: 300

CONTENTS

IN MEMORIAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

ORIGINAL PAPERS

Multiplex-PCR assay for identification of Klebsiella pneumoniae isolates carrying the cps loci for K1 and K2 capsulebiosynthesisGIERCZYÑSKI R., JAGIELSKI M., RASTAWICKI W., KA£U¯EWSKI S. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

FAME profiles in Pseudomonas vesicularis during catechol and phenol degradation in the presence of glucoseas an additional carbon sourceMROZIK A., PIOTROWSKA-SEGET Z., £ABU¯EK S. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

Reliability of the Etest in light of the correlation between an antibiotic�s critical concentration (Cc) and MIC valuesBEDNAR M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

Antibiotic susceptibility and molecular characterisation of Proteus mirabilis isolates in hospitals fromthe west pomeranian area of PolandM¥CZYÑSKA I., GIEDRYS-KALEMBA S. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Prevalence of antibiotic resistance profile in relation to phylogenetic background among commensal Escherichia coliderived from various mammalsBALDY-CHUDZIK K., STOSIK M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

Human neutrophil peptides in vaginitis/cervicitis of different etiologyWIECHU£A B.E., FRIEDEK D.A., EKIEL A.M., ROMANIK M.K., MARTIROSIAN G. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

Biological activity of phenylpropionic acid isolated from a terrestrial StreptomycetesNARAYANA K.J.P., PRABHAKAR P., VIJAYALAKSHMI M., VENKATESWARLU Y., KRISHNA P.S.J. . . . . . . . . . . . . . . . . . 191

Ultra-structural studies on root nodules of Samanea saman (Jacq.) Merr. (Leguminosae)QADRI R., MAHMOOD A., ATHAR M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

Growth and antioxidant activity of Desulfotomaculum acetoxidans DSM 771 cultivated in acetate or lactatecontaining mediaPAW£OWSKA-ÆWIÊK L., PADO R. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Bioremediation of aflatoxins by some reference fungal strainsEL-SHIEKH H.H., MAHDY H.M., EL-AASER M.M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215

INSTURCTION TO AUTHORS AVAILABLE AT www.microbiology.pl/pjm

Polish Journal of Microbiologyformerly Acta Microbiologica Polonica

2007, Vol. 56, No 3

Polish Journal of Microbiology2007, Vol. 56, No 3, 151�152

Miros³aw Kañtoch(1928�2007)

On the 5th of May, 2007, passed away professordr hab. med. Miros³aw Kañtoch, full member of thePolish Academy of Sciences, eminent microbiologist,the creator of modern Polish virology, expert of theWHO for medical microbiology, initiator and mainexecutor of scientific programs in collaboration withthe Center for Disease Control in Atlanta, devoted toviral infections of social significance in Poland.

Miros³aw Kañtoch was born on the 1st of January,1928 in Sosnowiec. He studied at the Faculty of Medi-cine of the Medical Academy in Wroc³aw, where hereceived his medical doctor�s degree in 1951. He wasa student, and then assistant of professor LudwikHirszfeld. He defended his doctoral thesis, the advi-sor of which was professor Henryk Makower, at theMedical Academy in Wroc³aw in 1956 and obtainedthe position of assistant professor at the same time atthe Department of Medical Microbiology, MedicalAcademy and at the Institute of Immunology and Ex-perimental Therapy of the Polish Academy of Sciences.In 1956 he organized the Electron Microscopy labo-ratory which he was then the head of for close to10 years. In 1961, at the age of 33, he presented anddefended his habilitation thesis and then left ona Rockefeller Foundation stipend for Baltimore, wherehe worked under the guidance of professor F.B. Bang.In his recollections professor Kañtoch always mentio-ned the names of professors L. Hirszfeld, H. Makowerand F.B. Bang as his mentors and the photos of thosethree professors always hung in his office.

Three years after returning from the USA, in 1965,professor M. Kañtoch was appointed by the Ministerof Health as the head of the Department of Virologyat the National Institute of Hygiene in Warsaw, whichposition he held until the year 2000.

In 1970 he became an associate professor and in1978 full professor. In 1986 he was elected corre-sponding member of the Polish Academy of Sciences(PAN) and in 1994 full member of the PAN.

Prof. M. Kañtoch was a member of several Com-mittees of the PAS, including the Committee of Micro-

biology, Committee of Immunology, Committee of Hu-man Ecology Etiopathogenesis, Committee of Immu-nology and Human Disease Etiopathogenesis. He washonorary member of the Polish Society for Microbio-logy, the Society of Epidemiologists, I.I. MiecznikowCommittee of Microbiologists and Infectious DiseasePhysicians in Russia, full member of the WarsawScientific Society, the Polish Society of Epidemio-logists and Infectious Disease Physicians, member ofthe team of experts of the World Health Organization.

Prof. M. Kañtoch was on the Scientific Boardsof many research institutions, such as the LudwikHirszfeld Institute of Immunology and ExperimentalTherapy of the PAS, the National Institute of Hygiene,The Military Institute of Hygiene and Epidemiology.Professor M. Kañtoch was also a member of the edito-rial boards of several scientific journals, e.g. �Przegl¹dEpidemiologiczny� �Postêpy Mikrobiologii�, �Medy-cyna Do�wiadczalna i Mikrobiologia�, Postêpy Higie-ny i Medycyny Do�wiadczalnej�, �Polish Journal ofMicrobiology� and a member of the international edi-torial board of �Acta Virologica�.

Professor M. Kañtoch contributed in a majorway to the development of scientific research on the

IN MEMORIAM

152 In Memoriam 3

etiopathogenesis, immunology and immunoprophyla-xis of viral infections. Worth particular attention is hiscontribution in studies on the safety, effectiveness anddissemination of vaccinations against the polio virus,measles and rubella and well as in the pioneer at thetime research on the teratogenic action of the rubellavirus and the occurrence of congenital defects causedby the infection of mothers during pregnancy withsuch viruses as the cytomegaly virus, herpes simplex(cold sores) and varicella zoster (chickenpox andshingles). In continuing the work of professor FeliksPrzesmycki, professor M. Kañtoch became a co-origi-nator of Polish medical virology.

The research mentioned above was documented inover 200 original scientific papers several score con-gress and meeting reports, which professor Kañtochwas the author or co-author of. The Professor was alsothe author of several monographs and the textbook

�Medical Virology� intended mainly for medical stu-dents and physicians, which had several reprints.

In recollecting professor M. Kañtoch, we cannotoverlook his didactic achievements. We, the employ-ees of the Department of Virology, National Instituteof Hygiene, remember the Professor above all asa carer, doctoral thesis advisor and a person who in-spired us and demanded we work on our habilitationtheses. The professor was the thesis advisor of 20 doc-torates, 17 of which were from the Department ofVirology of the National Institute of Hygiene andthree from the time of his work in Wroc³aw. We re-member the Professor as a very demanding Person butalways knew we could in every case count on hishelp, aid and protection.

Employees of the Department of VirologyNational Institute of Hygiene


ORIGINAL PAPER

Introduction

Klebsiella pneumoniae, an important nosocomialpathogen, causes suppurative infection, pneumonia,urinary tract infection and septicaemia in humans,especially immunosuppressed (Podschun and Ulmann,1998) or suffering from underlying diseases like diabe-tes mellitus (Fung et al., 2002). Persons of low socialstatus and alcoholics constitute the main patientspopulation at risk, comprising up to 66% of those suf-fering from community acquired pneumonia, that isa very severe illness with a rapid onset and a highmortality rates (Podschun and Ulmann, 1998; Sahlyand Podschun, 1997). Despite the discovery of othervirulence factors such as fimbriae, siderophores andO-antigens, capsular antigens are considered to beultimate determinants of K. pneumoniae pathogenicity(Podschun and Ulmann, 1998; Sahly and Podschun,1997; Fang et al., 2004; Yu et al., 2007). Clinical iso-lates of this species produce capsular polysaccharide(CPS) (Fang et al., 2004; Ørskov and Ørskov, 1984).Among 77 capsular serotypes (K-types) of K. pneumo-niae (Ørskov and Ørskov, 1984), strains belonging toserotypes K1 and K2 are considered the most virulentto mice (Simoons-Smit et al., 1984) and humans (Fang

et al., 2004; Yu et al., 2007). Moreover, strains of K1and K2 are believed to escape the opsonin-independentlectin phagocytosis (Podschun and Ulmann, 1998;Kabha et al., 1995). Clinical studies on 134 patientswith K. pneumoniae liver abscess exhibited predomi-nation of serotypes K1 (63.4%) and K2 (14.2%)(Fung et al., 2002).

The capsular swelling (quellung) reaction andcounter-current immunoelectrophoresis are the mostcommonly used techniques for identification of K. pneu-moniae serotypes K1 and K2 (Janda and Abbott, 1998).The availability and costs of the antisera, which can beproduced in specialised laboratories, limit the practiceof serotyping. Therefore, novel molecular-serotypingtool was recently developed (Brisse et al., 2004).Although, this method is capable to identify all 77K-types of K. pneumoniae, it requires time consuminglong-range PCR followed by the endonuclease diges-tion and computer aided analysis of the electrophoreticpatterns. Thus, despite its indisputable advantages,molecular-serotyping is not optimal for rapid identifi-cation of K1 and K2 strains in routine diagnostic. Onthe other hand, recently described PCR-based assaysfor differentiation of the major serovars of Listeriamonocytogenes (Doumith et al., 2004), Streptococcus

Multiplex-PCR Assay for Identificationof Klebsiella pneumoniae Isolates Carrying the cps Loci for K1

and K2 Capsule Biosynthesis

RAFA£ GIERCZYÑSKI*, MAREK JAGIELSKI, WALDEMAR RASTAWICKIand STANIS£AW KA£U¯EWSKI

Department of Bacteriology, National Institute of Hygiene, Warsaw, Poland

Received 26 June 2007, revised 20 July, accepted 25 July 2007

A b s t r a c t

Multiplex-PCR assay for identification of Klebsiella pneumoniae isolates carrying gene clusters for biosynthesis of capsular polysaccha-ride (CPS) types K1 and K2 was developed. Genes wzc and orf10 of the cps cluster were applied as K1 and K2 specific markers respec-tively. The assay specificity was confirmed using 147 isolates of Klebsiella spp. including 77 K-antigen reference strains. The multiplex-PCR assay was found simple and cost-effective tool for identification of K. pneumoniae clinical isolates of K1 and K2 geno-serotypes.

K e y w o r d s: Klebsiella, K. pneumoniae K1, K. pneumoniae K2, genoserotyping, multiplex-PCR

* Corresponding author: R. Gierczyñski, Department of Bacteriology, National Institute of Hygiene, Chocimska Street 24,00-791 Warsaw, Poland; phone: (48) 22 5421244, fax: (48) 22 5421307; e-mail: [email protected]

154 Gierczyñski R. et al. 3

pneumoniae (Kong and Gilbert, 2003), Yersiniapseudotuberculosis and Y. pestis (Bogdanovich et al.,2003) have been found a rapid and practical alternativeto laborious classical serotyping. For these reasons,we aimed to develop multiplex-PCR assay for identi-fication of K. pneumoniae strains genetically compe-tent to produce K1 and K2 capsular polysaccharides.

Experimental

Materials and Methods

Bacterial strains. We examined 147 isolates(Table I) including complete set of 77 Klebsiella spp.K-antigen reference strains (Ørskov and Ørskov,1984) and ten reference K1 and K2 strains describedelsewhere. Prior to serotyping by the counter-currentimmunoelectrophoresis using K1 and K2 antisera(Statens Serum Institut, Denmark) all clinical isolateslisted in Table I were biochemically identified by clas-sical tube tests. Noncapsulated variants of K. pneu-moniae strains K1 (n = 1) and K2 (n = 6) were de-signed as described previously (Ka³u¿ewski, 1968).

PCR procedure. Template DNA was preparedfrom 0.5 ml of an overnight culture at 37°C in nutrient

broth as described previously (Gierczyñski et al.,2004) but the lysozyme treatment was omitted. Prim-ers listed in Table II were used for amplification offragments of wzc, orf10 and K. pneumoniae 16S rRNAgene in multiplex-PCR. PCRs were carried out in 20 mlreaction volumes in a thermalcycler (Mastercycler,Eppendorf, Germany), with 0.75 U of the recombi-nant Taq DNA Polymerase (Fermentas, Lithuania),1×Mg-free PCR buffer with (NH4)2SO4, each deoxy-nucleoside triphospate at a concentration of 0.2 mM,3.0 mM MgCl2, each primer at a concentration shownin Table II and 2.5 µl of the template DNA solution.A general program consisting of 35 cycles for 45 sof each denaturation at 94°C, annealing at 60°Cand elongation at 72oC was used for amplification.Finally, DNA synthesis was completed at 72°C for3 min. Prior to cycling, 5 min denaturation step at94°C was included. The 2% gel (MP Biomedicals,Germany) in TAE (40 mM Tris-acetate pH 8.0, 1 mMEDTA) was used for the multiplex-PCR productsseparation. Gels were run at a constant voltage of 80 Vfor 2 hours, stained in 2 µg/ml ethidium bromidefor 10 min and photographed under UV by Gel-Scanapparatus (Kucharczyk, Poland). Each strain was ana-lysed in triplicate.

A5054 (O1:K1)a 1 K1 + � (Ørskov and Ørskov, 1984)

A5054b 1 NTe + � This study

408 (SB3182)c 1 K1 + � (Brisse et al., 2004)

468 (SB3186)c 1 K1 + � (Brisse et al., 2004)

643 (SB3188)c 1 K1 + � (Brisse et al., 2004)

920 (SB3192)c 1 K1 + � (Brisse et al., 2004)

B5055 (O1:K2)a 1 K2 � + (Ørskov and Ørskov, 1984)

B5055b 1 NT � + This study

1584 (SB3201)c (C2b)d 1 K2 � + (Brisse et al., 2004)

777 (SB3202)c (C2c)d 1 K2 � + (Brisse et al., 2004)

34 (SB3203)c (C2d)d 1 K2 � + (Brisse et al., 2004)

778 (SB3199)c (C2e)d 1 K2 � + (Brisse et al., 2004)

B4631 (O2:K2) 1 K2 � + (Kauffmann, 1954)

B7380 (O2:K2) 1 K2 � + (Kauffmann, 1954)

K3-K82a 75 NT � � (Ørskov and Ørskov, 1984)

Clinical isolates K1 3 K1 + � This study

Clinical isolates K2 10 K2 � + This study

Clinical isolates K2b 5 NT � + This study

Clinical isolates 40 NT � � This study

Total: 147

Table IList of tested strains and results of the multiplex-PCR assay

Strain Numberof isolates

Capsular type(serotyping)

Geno-serotypeReference

K1 (wzc) K2 (orf10)

a Klebsiella K-antigen reference strains excluding false serotypes K73 and K75-78 (Ørskov and Ørskov, 1984),b noncapsulated variants,c genomic DNA template (capsular type cited from the reference),d C-patterns (subgenotypes) of K. pneumoniae K2 (Brisse et al., 2004),e NT, strains nontypeable by K1 and K2 antisera.

155K. pneumoniae K1 and K2 identification by multiplex-PCR3

Results and Discussion

In order to select marker loci specific for serotypesK1 and K2 we performed comparative analysis of thecps gene clusters for K1 and K2 capsule biosynthesisdeposited in GenBank (http://www.ncbi.nlm.nih.gov)under accession numbers: AY762939 and D21242 re-spectively. For multiplex-PCR assay we selected genewzc encoding tyrosine-protein kinase and the openreading frame 10 (orf10) encoding putative innermembrane protein for K1 and K2 serotypes respec-tively. Fragment of the K. pneumoniae 16S rRNAgene served as a positive multiplex-PCR control. Theexpected size of 16S rRNA gene amplicon was 508bp while PCR-product for wzc (K1) and orf10 (K2)was 356 bp and 663 bp respectively.

The multiplex-PCR assay result for K1 and K2was judged as a positive when two bands were present� one specific for the 16 RNA gene and the other spe-cific for wzc or orf10 respectively. The presence ofthe16S rRNA gene amplicon alone indicted that testedDNA sample contained neither wzc, orf10 and PCRinhibitors. In this case, the assay result was valid butnegative for K. pneumoniae K1 and K2 geno-sero-types. The optimal yield of PCR products was ob-served for concentrations of wzc and orf10 primers

ranging from 0.50 to 0.25 µM and the 16S rRNA geneprimers between 0.250 and 0.125 µM.

The multiplex PCR yielded DNA fragment ofabout 500 bp for all tested strains, whereas additionalfragments about 350 bp and 650 bp were detectedfor strains of capsular type K1 and K2 respectively(Fig. 1). No bands were observed for DNA-free nega-tive control (data not shown). Specificity of the PCRproducts was confirmed by DNA nucleotide sequenc-ing performed as described previously (Gierczyñskiet al., 2004). Notably, orf10 amplicons were obtainedfor strains of K. pneumoniae K2 belonging to differentsubgenotypes (C-patterns) (Brisse et al., 2004). Thisfinding proved usefulness of the developed multiplex-PCR assay for identification of genetically diversestrains of capsular type K2 (Table I). Moreover, devel-oped assay correctly identified rarely occurring strainsO2:K2. The wzc and orf10 were also detected in thenoncapsulated variants of strain A5054 and B5055respectively. Consequently, orf10 was also traced innoncapsulated derivatives of clinical K2 isolates. Thisis in agreement to previous findings (Brisse et al.,2004), that molecular serotyping was capable to deter-mine a potential serotype of capsule-deficient isolates.Except the 500 bp band, no PCR-products were gene-rated for Klebsiella spp. K-antigen reference strains

wzcf wzc AY762939 5'-GATACAGGTGTATTGTCGC-3' 8947�8966 0.4 µM

wzcr wzc AY762939 5'-GAGCTCTATATGTTGGATGC-3' 9283�9302 0.4 µM

or10f orf10 D21242 5'-CCAGAGTTAGACCCGATATTC-3' 14205�14225 0.4 µM

or10r orf10 D21242 5'-GAAGTCTATTACCCCTGAAG-3' 14848�14867 0.4 µM

K16Sf 16S rRNA AF453251 5'-AGGGTGCAAGCGTTAATCGG-3' 493�512 0.2 µM

K16Sr 16S rRNA AF453251 5'-TGTCTCACAGTTCCCGAAGG-3' 981�1000 0.2 µM

Table IIPrimers used in this study

Primername Target locus

GenBankaccesion number Primer sequence

PrimerPosition

Primerconcentration

Fig. 1. Result of the multiplex PCR for identification of K. pneumoniae capsular genoserotypes K1 and K2M � DNA size ladder 100 bp step (GeneRuler 100 bp, Fermetas, Lithuania). Lines: K1 � A5054 (K1 reference strain), K2 � B5055

(K2 reference strain), 1 � B5055 (noncapsulated), 2 � F5052 (K6), 3 � 889/50 (K20), 4 � 636/52 (K58), 5 � 438 (K66), 6 � B4631 (O2:K2),7 � Kp90 (non-K1 and non-K2 clinical isolate), 8 � Kp57 (clinical isolate K2), 9 � A5054 (noncapsulated), 10 � 408 (K1), 11 � Kp229

(clinical isolate K1), 12 � 778 (K2), 13 � 1584 (K2), 14 � 34 (K2).

156 Gierczyñski R. et al. 3

tested other than A5054 or B5055, as well clinical iso-lates nontypeable by K1 and K2 antisera (Table I). Itis noteworthy, strain K58 that was reported to cross-react with K1 in serotyping did not yield wzc specificamplicon. The lack of such cross-reactions may be anadvantage of developed assay when compared witha classical serotyping.

Taken together, obtained results show that devel-oped multiplex-PCR assay is potentially useful toolfor identification of K1 and K2 serotypes of K. pneu-moniae. Moreover, developed assay is capable to de-termine whether a capsule defective strain is K1 orK2 derivative. Thus, the assay detects K1 and K2geno-serotypes of K. pneumoniae in fact. However,due to reported horizontal transfer of the cps clusterto strains representing other species of Enterobacteri-aceae (Nelson and Selander, 1994; Rahn et al., 1999)the multiplex-PCR assay may not be used instead theclassical biochemical tests for K. pneumoniae identifi-cation (Janda and Abbott, 1998). We recommend thisassay as a relatively inexpensive and robust tool forscreening for K. pneumoniae K1 and K2 genosero-types. However, to diversify capsule producing andcapsule deficient isolates classical serotyping with K1and K2 antisera is recommended. The multiplex-PCRmay help to reduce total cost and workload of K. pneu-moniae K1 and K2 capsular types identification inepidemiological surveys and routine diagnostic.

AcknowledgementsThis work was supported by grants-in-aid for scientific re-

search (3P05D00225) from the Ministry of Science and HigherEducation of Poland.

Dr. A.A. Zasada assisted in the template DNA isolation. Weare thankful to Dr. S. Brisse for a support of a total DNA of se-lected strains of K. pneumoniae K1 and K2.

Literature

Bogdanovich T., E. Carniel, H. Fukushima and M. Skurnik.2003. Use of O-antigen gene cluster-specific PCRs for identifica-tion and O-genotyping of Yersinia pseudotuberculosis and Yersiniapestis. J. Clin. Microbiol. 41: 5103�5112.Brisse S., S. Issenhuth-Jeanjean and P.A.D. Grimont. 2004.Molecular serotyping of Klebsiella species isolates by restrictionof the amplified capsular antigen gene cluster. J. Clin. Microbiol.42: 3388�3398.Doumith M., C. Buchrieser, P. Glaser, Ch. Jacquet and P. Mar-tin. 2004. Differentiation of the major Listeria monocytogenesserovars by multiplex PCR. J. Clin. Microbiol. 42: 3819�3822.

Fang Ch.T., Y.P. Chuang, Ch.T. Shun, S.Ch. Chang and J.T.Wang. 2004. A novel virulence gene in Klebsiella pneumoniaestrains causing primary liver abscess and septic metastatic com-plications. J. Exp. Med. 199: 697�705.Fung C.-P., F.-Y. Chang, S.-C. Lee, B.-S. Hu, B. I.-T. Kuo,C.-Y. Liu, M. Ho and L.K. Siu. 2002. A global emerging diseaseof Klebsiella pneumoniae liver abscess: is serotype K1 an impor-tant factor for complicated endophthalmitis? Gut 50: 420�424.Gierczyñski R., S. Ka³u¿ewski, A. Rakin, M. Jagielski, A. Zasa-da, A. Jakubczak, B. Borkowska-Opacka and W. Rastawicki.2004. Intriguing diversity of Bacillus anthracis in eastern Poland� the molecular echoes of the past outbreaks. FEMS Microbiol.Lett. 239: 235�240.Janda J.M. and S.L. Abbott. 1998. The Enterobacteria, Lippin-cott-Raven, (ed.) Philadelphia, New York, pp 110�130.Kabha K., L. Nissimov, A. Athamna, Y. Keisari, H. Parolis,L.A. Parolis, R.M. Grue, J. Schlepper-Schafer, A.R. Ezekowitzand D.E. Ohman. 1995. Relationships among capsular structure,phagocytosis, and mouse virulence in Klebsiella pneumoniae.Infect. Immun. 63: 847�52.Ka³u¿ewski S. 1968. Some partial antigens of unencapsulatedvariants of group O2 Klebsiella: I. characteristics of strains and anti-gen O preparations. Exper. Med. Microbiol. 20: 16�32.Kauffmann F. 1954. Enterobacteriaceae., Second edition, EjnarMunksgaard Publisher, Copenhagen, pp. 223�247.Kong F. and G.L. Gilbert. 2003. Using cpsA-cpsB sequence poly-morphisms and serotype-/group-specific PCR to predict 51 Strepto-coccus pneumoniae capsular serotypes. J. Med. Microbiol. 52:1047�1058.Nelson K. and R.K. Selander. 1994. Intergeneric transfer andrecombination of the 6-phosphogluconate dehydrogenase gene(gnd) in enteric bacteria. Proc. Natl. Acad. Sci. USA 91:10227�10231.Ørskov I. and F. Ørskov. 1984. Serotyping of Klebsiella, In:T. Bergan (Ed.) Methods in Microbiology, Vol. 14, AcademicPress Inc. New York, NY, pp 143�164.Podschun R. and U. Ulmann. 1998. Klebsiella spp. as noso-comial pathogens: epidemiology, taxonomy, typing methods, andpathogenicity factors. Clin. Microbiol. Rev. 11: 589�603.Rahn A., J. Drummelsmith and C. Whitfield. 1999. Conservedorganization of the cps gene clusters for expression of Escherichiacoli group 1K antigens: relationship to colanic acid biosynthesislocus and the cps genes from Klebsiella pneumoniae. J. Bacteriol.181: 2307�13.Sahly H. and R. Podschun. 1997. Clinical, bacteriological, andserological aspects of Klebsiella infections and their spondylarthro-pathic sequelae. Clin. Diagn. Lab. Immunol. 4: 393�399.Simoons-Smit A.M., A.M. Verwey-van Vught, I.Y. Kanis andD.M. MacLaren. 1984. Virulence of Klebsiella strains in experi-mentally induced skin lesions in the mouse. J. Med. Microbiol.17: 67�77.Yu V.L., D.S. Hansen, W.Ch. Ko, A. Sagnimeni, K.P. Klugman,A. von Gottberg, H. Goossens, M.M. Wagener, V.J. Benedi,J.M. Casellas, G. Trenholme, J. McCormack, S. Mohapatraand L. Mulazimoglu. 2007. Virulence characteristics of Kleb-siella and clinical manifestations of K. pneumoniae bloodstreaminfections. Emerg. Infect. Dis. 13: 986�993.


ORIGINAL PAPER

Introduction

Fatty acids are essential structural components ofbacterial cell membranes that regulate their stabilityand fluidity. The membrane is the site of the primarycontact with the environment and has an importantrole in maintaining the viability and functionality ofbacterial cells. The main function of the membrane isto form permeability barriers regulating the passageof solutes between the cell and the external environ-ment. This function is mainly determined by mem-brane lipid composition (�ajbidor, 1997; Denich et al.,2003). The analysis of bacterial membrane fatty acidsis also of interest for studies on toxicity of many con-taminants that generate environmental stress. Manyfindings documented that xenobiotics, such as organicsolvents and aromatic hydrocarbons influence bacte-rial fatty acid composition (Heipieper et al., 1992;Sikkema et al., 1994; Kabelitz et al., 2003). Low-mo-

lecular weight aromatic hydrocarbons, such as catecholand phenol are the simplest structurally aromatic com-pounds and enter the environment as a consequenceof human activities. For example, they widely occurduring the production of dyes, pesticides, pharmaceu-ticals, wood processing chemicals, polymers andexplosives. Since aromatic compounds exhibit toxic,mutagenic and carcinogenic properties, there is a se-rious concern about their elimination from environ-ment. One of the most promising methods is theapplication of hydrocarbon degrading bacteria toclean-up contaminated sites. For this purpose, numer-ous bacteria mainly from the genera Pseudomonas,Acinetobacter, Klebsiella and Bacillus are used in thefield of phenols degradation (Ahamad and Kunhi,1996; Chang et al., 1997; Ali et al., 1998; Heesche-Wagner et al., 1999; Beendorf et al., 2001).

The rate of degradation processes of many aro-matic substrates can be improved by supplementing

FAME Profiles in Pseudomonas vesicularis during Catecholand Phenol Degradation in the Presence of Glucose

as an Additional Carbon Source

AGNIESZKA MROZIK*1, ZOFIA PIOTROWSKA-SEGET2 and SYLWIA £ABU¯EK1

1 Department of Biochemistry, 2 Department of Microbiology,University of Silesia, Katowice, Poland

Received 22 March 2007, revised 28 May 2007, accepted 1 June 2007

A b s t r a c t

The aim of this study was to evaluate the impact of catechol and phenol added to culture media separately and with glucose as anadditional, easily-degradable carbon source on fatty acid methyl ester (FAME) composition in Pseudomonas vesicularis. Simultaneously,the degradation rates of aromatic substrates used were investigated in single and binary substrate systems. Both catechol and phenoltreatments caused changes in the distribution of tested groups of fatty acids. The most noticeable changes included an increase in degreeof fatty acid saturation, the appearance of branched and disappearance of hydroxy fatty acids as compared to the control sample withglucose. Under catechol or phenol treatment sat/unsat ratio showed the values of 8.63 and 11.38, respectively, whereas in control cells itreached the value of 2.66. The high level of saturation comes from the high content of cyclopropane fatty acids in bacteria under exposureto aromatic substrates, regardless of the presence of glucose. In these treatments their content was more than 3-fold higher compared to thecontrol. It has been demonstrated that glucose supplementation of culture media containing single aromatic substrate extended the degra-dation rates of catechol and phenol by P. vesicularis, caused an increase in number of cells but did not significantly change the fatty acidprofiles in comparison with bacteria growing on catechol and phenol added to the media individually.

K e y w o r d s: Pseudomonas vesicularis, catechol and phenol degradation, fatty acid composition

* Corresponding author: A. Mrozik, University of Silesia, Department of Biochemistry, Jagielloñska 28, 40-032 Katowice, Poland;phone: (48) 32 200 94 42; fax: (48) 32 200 93 61; e-mail: [email protected]

158 Mrozik A. et al. 3

of culture media with additional carbon sources orother compounds such as nitrate, phosphate as well asmineral constituents. Glucose, sodium glutamate andyeast extracts are known as conventional carbonsources that influence the biotransformation and bio-degradation processes (Wang et al., 1996; Wang andLoh, 1999, 2001). For example, Yu and Ward (1994)have observed that the rate of pentachlorophenoldegradation by mixed bacteria cultures significantlyincreased by the addition of glucose and peptone toculture medium. However, some additional carbonsources may inhibit aromatic compound degradation.In studies conducted by Ampe et al. (1998) it has beenshown that Ralstonia eutropha degraded phenol lesseffectively in the presence of acetate as comparedto the culture with phenol alone. Moreover, glucoseenrichment repressed catechol degradation by Pseu-domonas sp. CF600 (Mrozik et al., 2006). In turn, theaddition of glucose and sodium glutamate did notaffect the dynamics of phenol degradation by Pseudo-monas putida ATCC49451 (Loh and Wang, 1998).Bacteria posses a regulatory mechanism that allowsthem to use a preferential carbon source over a mix-ture of several other substrates and this phenomenonis usually called catabolic repression. It has been de-scribed for various bacteria, however, the molecularmechanism of gene expression for peripheral cata-bolic enzymes in the presence of preferred substratediffer substantially between species (Saier, 1996;Stülke and Hillen, 1999; Petruschka et al., 2001).

In fact, many aromatic compounds partition intophospholipid bilayer and modify its fatty acids com-position and membrane properties. Accumulation ofthese compounds in the membrane disturbs many bio-logical processes such as respiration, growth, ions andnutrient transport and may even cause lysis of the cell(Sikkema et al., 1995; Weber and de Bont, 1996;Denich et al., 2003). As a response to phenols expo-sure bacteria modify their membrane lipid composi-tion by de novo synthesis of fatty acids, isomerizationof cis to trans unsaturated fatty acids, changing theproportion between iso and anteiso branched fattyacids, altering the average of chains length and pro-tein content (Keweloh et al., 1990; Heipieper et al.,1994; Sikkema et al., 1995). These mechanisms havebeen related to homeoviscous adaptation and havebeen investigated by several authors (Shinitzky, 1984;Heipieper et al., 1992; Härtig et al., 2005).

However, there is no available information on theinfluence of additional carbon sources on fatty acidprofiles of bacteria during the biodegradation ofphenols. The objective of this work was to establishchanges in cellular fatty acid patterns in Pseudomonasvesicularis during catechol and phenol degradation inculture media supplemented with glucose as an addi-tional source of carbon and energy.

Experimental


Bacterial strain. The experiments were performedusing Pseudomonas vesicularis strain isolated frommixed populations of activated sludge collected fromsewage-treatment plant in Czêstochowa, Poland. Toselect phenol-degrading bacteria the increasing dosesof phenol were added to sample of sludge for 30 days.To isolate phenol-degrading bacteria 10-fold dilutionsof sludge suspensions were plated onto mineral me-dium (Kojima et al., 1961) amended with 0.188 g/l ofphenol. Among isolated strains P. vesicularis wasdominant. It was identified on the basis of cellularfatty acids derivatized to methyl esters (FAMEs) andanalysed by gas chromatography using the MIDI Mi-crobial Identification System (Newark, USA).

Culture conditions. Cultures were grown in modi-fied minimal medium containing: 3.78 g of Na2HPO4×12H2O; 0.5 g of KH2PO4; 5.0 g of NH4Cl; 0.2 g ofMgSO4×7H2O and 0.1 g of yeast extract in 1.0 l ofdeionised water (Kojima et al., 1961). To study theeffect of glucose on FAME profiles bacteria were culti-vated in Kojima medium containing catechol or phe-nol, at the concentration of 0.440 g/l and 0.376 g/l,respectively and in binary mixtures containing singlearomatic substrate and 1.0 g/l of glucose. The final pHof the medium was 7.2�7.3. Liquid cultures were grownin 500 ml flask on rotary shaker (125 rpm) at 30°C.

Bacterial growth. Samples of the cultures werewithdrawn every two hours until 8 h of incubation, andthen at 16 and 24 h of the experiments. Cell density(OD) was measured spectrophotometrically as the ab-sorbance of the suspension at 600 nm, with referenceto a standard curve calibrated by plate enumeration.

Determination of catechol and phenol concentra-tions. Concentrations of tested aromatic compoundswere measured at the same sampling time when ODwas measured. Determination of catechol concentrationwas based on color reaction between catechol andsodium molybdate by measuring absorbance at 480 nm(Evans, 1946). Phenol concentration was estimatedusing spectrophotometry method with diazotisedp-nitroaniline by measuring the absorbance of colorsolution at 550 nm (Lurie and Rybnikova, 1968).

Determination of glucose concentration. Re-moval of glucose in media was calculated using testsGLUCOSE EO produced by Biochemtest, Poland.This test is based on glucose oxidation to gluconicacid by glucose oxidase with production of hydrogenperoxide in the presence of peroxidase andchromogene ABTS. Absorbance of solution was mea-sured with spectrophotometer at 675 nm.

Enzyme activity assay. The activities of catecholdioxygenases were measured spectrophotometrically

159FAME profiles in Pseudomonas vesicularis3

by monitoring the formation of the first product ofaromatic ring cleavage, cis, cis-muconate at 260 nmfor catechol 1,2-dioxygenase and 2-hydroxymuconicsemialdehyde at 375 nm for catechol 2,3-dioxygenase(Feist and Hegeman, 1969). Detailed procedure of en-zymes isolation was described in the previous paper(Mrozik et al., 2006). Enzymes activities were expres-sed as mmol of cis, cis-muconate and 2-hydroxy-muconic semialdehyde formed per mg of protein perminute for catechol 1,2-dioxygenase and catechol2,3-dioxygenase, respectively. The protein content ofcell-free extract was estimated by the method ofBradford (1976) with lysozyme as a standard.

Fatty acid extraction and analysis. Fatty acidcomposition of bacterial strain was determined at mid-exponential phase of culture growth. For analysis ofcellular fatty acids cells grown in single- and binary-substrates systems were used. Bacteria were harvestedby centrifugation (8000×g) at 4°C for 30 min. The cellpellets obtained from each culture were washed with10.0 ml of 0.85% NaCl to remove residue of culturemedium. To decrease the humidity of bacterial cell,pellets were left through 2 h at room temperature. Next40 mg of bacterial biomass was transferred in tripli-cate to reaction tubes (Pyrex). To each sample 1.0 mlof 3.27 M NaOH in MeOH:H2O (1:1) for saponifica-tion was added. Then the samples were vortexed andplaced in 100°C water bath for 30 min. Following thissaponification step, fatty acids were converted to fattyacid methyl esters (FAMEs) by adding 2.0 ml of 6.0 MHCl:MeOH (1:0.85) to each tube and were incubatedat 80°C in water bath for 10 min. FAMEs were ex-tracted from the aqueous phase by addition of 1.15 mlof hexane:methyl tert-butyl ether (MTBE) (1:1) to eachtube. Then samples were rotated end-over-end for10 min. After removing aqueous (lower) phase, 3.0 mlof 0.3 M NaOH in H2O was added and the tubes wereagain rotated for 5 min (Sasser, 1990). Finally, the or-ganic (upper) phases containing FAMEs were trans-ferred to gas chromatography vials. Fatty acids wereanalysed by gas chromatography (Hewlett-Packard6890, USA) using capillary column Ultra 2-HP (cross-linked 5% phenyl-methyl silicone 25 m, 0.22 mm ID,thickness 0.33 mm) and hydrogen as a carrier gas.FAMEs were detected by a flame ionisation detector(FID) and identified by MIS (Microbial IdentificationSystem) software, using the aerobe TSBA40 methodand TSBA40 library (MIDI, USA).


Cell growth and aromatic compounds degrada-tion. To estimate the effect of glucose on catechol andphenol degradation by P. vesicularis bacteria weregrown in media containing only single aromatic sub-

strate and in the same media supplemented with glu-cose. The strain was able to metabolize catechol com-pletely at the concentration of 0.440 g/l and phenol atthe concentration of 0.376 g/l served as a single sub-strate during 10 and 15 h, respectively. The highestcatechol removal was observed during the first 4 h ofculturing and in this time 65% of dose added to themedium was degraded. In comparison, in that time inphenol containing medium its concentration decreasedabout 42%. In both experiments significant differencesin growth as indicated OD value of P. vesicularis werenot found. The substrate removal profiles and growthcurves are presented in Figure 1A and B.

The addition of glucose to media with aromaticsubstrates resulted in the increase of culture OD andaltered the time of catechol and phenol degradation.The time necessary for complete removal of both aro-matic substrates by tested bacteria extended to 24 h(Fig. 1A and B). In comparison with single-substratesystem, OD of bacterial culture in binary mixtureswas markedly higher and reached the value of 0.8 and1.0 for catechol with glucose and phenol with glucose,respectively. In control medium with glucose servedas a sole source of carbon and energy P. vesicularismetabolized it during 6 h of culturing (data notshown). Time of glucose utilization did not change inthe medium containing phenol whereas in mediumwith catechol was 2 hours shorter as compared to thecontrol sample. Interesting changes were revealedwhen compared the dynamics of catechol and phenolbiodegradation. In mixture containing glucose and phe-nol both substrates started to be degraded at the sametime whereas in the mixture containing catechol andglucose were not degraded simultaneously (Fig. 1Aand B). Catechol biodegradation by P. vesicularisstarted when 90% of glucose added was metabolized.These results indicated that glucose was preferentiallyutilized by P. vesicularis and it might repress catecholdegradation. Similar phenomenon was observed dur-ing studies on catechol biodegradation rate in the pres-ence of glucose by strain Pseudomonas sp. CF600. Thetime of catechol degradation in binary system withglucose was longer than that when catechol served asa sole carbon source. In contrast to P. vesicularis itstarted to degrade both substrates immediately aftertheir addition to the culture medium (Mrozik et al.,2006). The effect of glucose and sodium acetate onaromatic compounds biodegradation by bacteria fromthe genus Pseudomonas was also observed by Kaoet al. (2005). They have revealed that addition of theseextra carbon sources did not enhance pentachlorophe-nol (PCB) degradation by Pseudomonas mendocinaNSYSU. They have explained this phenomenon bythe fact that this strain isolated from PCB-contaminatedsoil did not receive inputs of glucose and acetatefrom natural sources and the cometabolism is not the


dominant biodegradation mechanism of PCB by thisbacterium. In turn, impact of glucose on phenanthrene(PHE) degradation by Sphingomonas sp. strain LB126in chemostat cultures was studied by van Herwijnenet al. (2003). They found that PHE removal in thepresence of glucose was much higher as compared tophenanthrene and fluorene grown culture without thehigh impact on growth cells. Besides biodegradationexperiments many studies are being conductedon molecular mechanism of catabolic repression inbacteria grown on glucose or other easily-degradablecarbon sources and aromatic compounds as inducers.For example, Duetz et al. (1996) described catabolicrepression of the TOL pathway by succinate under

different conditions of inorganic-nutrient limitation.In other studies the role of Crc regulator in the repres-sion of several catabolic pathways for the assimila-tion of some sugars and aromatic compounds in Pseu-domonas putida was shown (Morales et al., 2004).Such studies are necessary for better understandingthe correlations among degradation ratio of variousorganic substrates by bacteria.

In parallel to biodegradation studies the activitiesof enzymes involved in aromatic ring cleavage werecalculated. As shown in Table I, P. vesicularis in eachexperiment treatment synthesized both catechol 1,2-and 2,3-dioxygenases. In bacterial cells growing oncatechol only the activity of catechol 1,2-dioxygenase

GC ODC

ODC+G

CG C

1.1

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

00 2 4 6 8 10 12 14 16 18 20 22 24

Time, hours

1.1

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

Car

bon

sour

ce, g

/l

OD

, 600

nm

PG

P

GP

ODP+G

ODP

1.1

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

1.1

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

Car

bon

sour

ce, g

/l

OD

, 600

nm

0 2 4 6 8 10 12 14 16 18 20 22 24Time, hours

Fig. 1. Degradation rate of catechol (0.440 g/l) (A) and phenol (3.376 g/l) (B) in the presence and absence of glucose (1.0 g/l)and growth curves of Pseudomonoas vesicularis.

C � degradation of catechol, GC � degradation of glucose in binary system with catechol, CG � degradation of catechol in binary system withglucose, ODC � optical density of culture growing on catechol, ODC + G � optical density of culture growing on catechol and glucose, P � degradationof phenol, GP � degradation of glucose in binary system with phenol, PG � degradation of phenol in binary system with glucose, ODP � optical

density of culture growing on phenol, ODP + G � optical density of culture growing on phenol and glucose.

A

B


was significantly higher as compared to activity ofcatechol 1,2-dioxygenase and reached the values of2.55 and 0.39 µmol/min/mg of protein, respectively.In contrast, in bacteria cultured on phenol the activityof catechol 2,3-dioxygenase was 2.5-fold higher thancatechol 1,2-dioxygenase. These results indicate thatcatechol and phenol degradation proceeded both viameta and ortho metabolic pathways. The addition ofglucose slightly decreased the activity of measuredenzymes (Table I). Similarly, Tian et al. (2003) stu-dying the impact of glucose added on phenanthrenedegradation by P. mendocina demonstrated that glu-cose supplementation decreased the activities of hydro-carbon dioxygenase and catechol 2,3-dioxygenase.

FAME analysis. To estimate the changes in fattyacid composition the profiles of whole-cell fatty acidsisolated from P. vesicularis cultured in media contain-ing catechol or phenol degradation with or withoutglucose were analyzed. For the detailed interpretationof results all fatty acids obtained were divided intotwo major groups: saturated and unsaturated. The firstgroup of fatty acids included four sub-groups: straight-chain, branched, hydroxy- and cyclopropane fattyacids. Percentages of these fatty acid groups in eachexperiment treatments are presented in Table II. Bothcatechol and phenol treatment caused crucial changesin the distribution of the tested groups of whole cell-derived fatty acids in P. vesicularis. Bacteria culturedon aromatic substrates characterized by the higherproportion of saturated fatty acids as compared tocontrol with glucose. The percentage of these fattyacids composed 89.41% and 91.92% of total fattyacids when bacteria were grown on catechol or phenol,respectively, whereas in control sample they repre-sented 72.65% of total fatty acids (Table II). Similartendency resulting in the increase of the membranesaturation in the presence of toxic aromatic compoundsand aliphatic alcohols was earlier observed in studiesusing P. putida (Heipieper et al., 1992; Mrozik et al.,2005), Rhodococcus sp. 33 (Gutierrez et al., 1999),

Ralstonia eutropha H850 (Kim et al., 2001) andAcinetobacter calcoaceticus (Kabelitz et al., 2003).The addition of glucose to culture medium withcatechol did not significant change the abundance ofsaturated fatty acids in P. vesicularis as compared tocells collected from medium containing catechol only.In contrast, in bacteria growing in medium containingphenol and glucose the proportion of saturated fattyacids was about 10% lower than in bacteria grown onphenol used separately. In bacterial cells, irrespectiveof medium content, among saturated fatty acids thedominant group was straight-chain fatty acids. Thisgroup included the following fatty acids: 10:0, 12:0,14:0, 15:0, 16:0, 18:0 and 19:0. However, their per-centages in fatty acid profiling obtained from bacteriacultured in single- and binary system was lower ascompared to control (55.40%) and ranged from 48.39to 52.58%. The increase of degree of membrane satura-tion is well known adaptive mechanism allowing bac-teria to survive under toxic substrates stress (Sikkemaet al., 1994, 1995; Weber and de Bont, 1996). Catecholand phenol exposure drastically changed the contentof terminally branched and hydroxy fatty acids. Inter-estingly, this observed effect was independent of thepresence of glucose in the culture medium. The resultsshowed that both aromatic substrates caused the dis-appearance of hydroxy fatty acid 12:0 2OH, whereasin control sample with glucose it composed 5.70% oftotal fatty acids. In contrast, catechol and phenol inall tested systems caused the appearance of branchedfatty acid 15:0 iso and additionally 15:0 anteiso wasdetected in cells growing on phenol as a single carbonsource. However, the percentages of branched fattyacids in FAME profiles were generally low and rangedfrom 0.84 to 1.62% of total fatty acids for bacteriagrown on catechol and/or phenol in the presence orabsence of glucose (Fig. 2). Tsitko et al. (1999) stu-dying the impact of different aromatic compounds on

glucose 0.07 ± 0.02 0.11 ± 0.03

catechol 2.55 ± 0.07 0.39 ± 0.07

catechol + glucose 2.11 ± 0.11 0.19 ± 0.05

phenol 0.70 ± 0.05 1.85 ± 0.17

phenol + glucose 0.57 ± 0.06 1.12 ± 0.09

Table ICatechol 1,2- and 2,3-dioxygenase activities in cell-free

extracts of Pseudomonas vesicularis growing on catecholor phenol with/without glucose

Carbon source

Catechol1,2-dioxygenase

µmol/min/mgof protein

Catechol2,3-dioxygenase

µmol/min/mgof protein

Number of replicates, n = 3

Saturated 72.65 89.41 87.00 91.92 84.27

Straight-chain 55.40 50.99 48.39 50.52 52.58

Hydroxy 5.70 0.00 0.00 0.00 0.00

Branched 0.00 0.93 0.84 1.62 1.12

Cyclopropane 11.56 37.49 37.77 39.78 30.57

Unsaturated 27.35 10.36 12.99 8.08 15.73

Sat/unsat ratio 2.66 8.63 6.70 11.38 5.36

Table IIPercentages of total saturated, unsaturated fatty acids and sat/unsat ratio of Pseudomonas vesicularis growing on catecholor phenol, or/and glucose in single- and binary systems

Groupof fatty acids

% of total fatty acids

Glucose CatecholCatechol+ glucose Phenol

Phenol +glucose

Values are the averages of three independently performed experiments(standard errors < 5%).


R. opacus FAMEs composition revealed that these sub-strates also increased content of branched fatty acids.

It seems that the response of bacterial cells tomembrane active substrates to a large extent dependson individual physiological and biochemical featuresof given bacteria. In contrast to tested strain, Pseudo-monas sp. CF600 reacted to catechol and phenolexposure in an opposite way. This strain growing oncatechol or phenol in single- and binary systems withglucose synthesized both hydroxy and branched fattyacids (Mrozik et al., 2006). With regards to the chroma-tographic profiles of saturated fatty acids, the changesin the abundance of cyclopropane fatty acids 17:0 cyand 19:0 cy T8c were the most visible. The highestincrease of these fatty acids content was detected inbacterial cells grown on phenol. Their percentagereached the value of 39.78% whereas in control samplewith glucose showed the value of 11.56%. Surprisingly,in bacteria cultured in binary system containing phenoland glucose 19:0 T8cy fatty acid was not detected thatresulted in decreasing of the total amount of cyclopro-pane fatty acids. Such phenomenon was not observedin the experiment with catechol served as a sole carbonand energy source and in mixture with glucose (Fig. 2).The presented results as well as results obtained byother researchers indicate that content of cyclopro-pane fatty acids depends not only on the chemicalstructure and properties of hydrocarbons but also onthe features of bacterial strains (Ramos et al., 1997;Kim et al., 2001; Fang et al., 2004; Mrozik et al.,2006). Cyclopropane fatty acids have been known ascompounds that stabilize membrane lipids, make itmore rigid and in this way improve bacteria survivalunder unfavorable conditions. However, the detailed

role of these fatty acids in the regulation of bacterialmembrane stability and fluidity in the presence of aro-matic compounds is not fully understood yet and re-quire further investigations and explanations.

It has been found that tested aromatic substratesused both in a single- and binary systems significantlydecreased the amount of unsaturated fatty acids suchas 16:1T7c and 18:1T9c. Their abundance declinedabout 5-fold in comparison to control. In contrast, thepercentage of fatty acid 18:1T7c/T9t/T12t, which iswell known as typical for bacteria from the genusPseudomonas, increased from 6.34% in control to8.41% and 10.25% in bacterial cells growing in mediasupplemented with catechol or phenol and glucose,respectively. However, under exposure of catecholand phenol used individually the amount of this fattyacid slightly decreased.

The impact of various toxic compounds and theirinteractions with easily degradable carbon sources onbacterial MIDI-FAME profiles might be examined byanalysis of saturated/unsaturated ratio (Table II). Inthis study it has been showed that under catechol orphenol exposure this ratio was about 3.5-fold higherthan that in control and reached the value of 8.63 and11.38 for catechol and phenol, respectively. The addi-tion of glucose to culture medium containing the aro-matic substrates changed the response of bacterialcells to these toxic compounds which resulted in de-creasing of sat/unsat ratio in comparison with culturesgrown in the presence of catechol or phenol separately(Table II). Additionally, the protective effect of glu-cose against the toxicity of aromatic substrates duringtheir biodegradation was confirmed by marked in-crease of bacterial culture density (Fig. 1A and B).

100%

80%

60%

40%

20%

0%G C C + G P P + G

Fig. 2. Percentages of distinct groups of fatty acids in Pseudomonas vesicularis growing on catecholor phenol only and in binary systems with glucose, G � glucose, C � catechol, P � phenol

straight branched hydroxy cyclopropanestraight branched hydroxy cyclopropane unsaturated


The action of catechol or phenol in the presence ofglucose as an additional carbon source measured asFAMEs patterns of bacterial cells was slightly differentas compared to the action of these aromatic substratesadded individually. Generally, the differences amongthe fatty acid composition in bacteria cultured in mediacontaining catechol or phenol and in mixtures withglucose were slight but significant in comparison topattern of FAMEs obtained for control samples. Themost noticeable difference was associated with cyclo-propane fatty acid abundance. Beside straight-chainfatty acids they constituted the second dominant groupin FAMEs profiles. The high proportion of cyclo-propane fatty acids resulted in the increase of sat/unsatratio. The data obtained from biodegradation studiesand analysis of FAME profiles of P. vesicularis indi-cated that the addition of glucose as easily-degrad-able carbon source to media containing aromatic sub-strates such as catechol and phenol stimulated thegrowth of bacteria while did not have distinct influ-ence of the whole cell-derived fatty acid composition.

Literature

Ahamad P.Y.A. and A.A.M. Kunhi. 1996. Degradation of phenolthrough ortho-cleavage pathway by Pseudomonas stutzeri strainSPC2. Lett. Appl. Microbiol. 22: 26�29.Ali S., R. Fernandez-Lafuente and D.A. Cowan. 1998. Meta-pathway degradation of phenolics by thermophilics bacilli. Enz.Microb. Technol. 23: 462�468.Ampe F., D. Leonard and N.D. Lindley. 1998. Repression ofphenol catabolism by organic acids in Ralstonia eutropha. Appl.Environ. Microbiol. 64: 1�6.Beendorf D., N. Loffhagen and W. Babel. 2001. Protein synthe-sis patterns in Acinetobacter calcoaceticus induced by phenol andcatechol show specificities of responses to chemostress. FEMSMicrobiol. Lett. 200: 247�252.Bradford M.M. 1976. A rapid and sensitive method for quanti-tation of microgram quantities of protein utilising the principle ofprotein-dye binding. Anal. Biochem. 72: 248�254.Chang B.V., W.B. Wu and S.Y. Yuan. 1997. Biodegradation ofbenzene, toluene, and other aromatic compounds by Pseudomo-nas sp. D8. Chemosphere 35: 2807�2815.Denich T.J., L.A. Beaudette, H. Lee and J.T. Trevors. 2003. Ef-fect of selected environmental and physico-chemical factors on bac-terial cytoplasmatic membranes. J. Microbiol. Meth. 52: 149�182.Duetz W., S. Marques, B. Wind, J.L. Ramos and J.G. vanAndel. 1996. Catabolite repression of the toluene degradationpathway in Pseudomonas putida harbouring pWW0 under variousconditions of nutrient limitation in chemostat culture. Appl.Environ. Microbiol. 62: 601�601.Evans W.C. 1946. Oxidation of phenol and benzoic acid by somesoil bacteria. Biochem. J. 41: 373�382.Fang J., N. Lovanh and P.J.J. Alvarez. 2004. The use of iso-topic and lipid analysis techniques linking toluene degradation tospecific microorganisms: applications and limitations. Water Res.38: 2529�2536.Feist C.F. and G.D. Hegeman. 1969. Regulation of the metacleavage pathways for benzoate oxidation by Pseudomonas putida.J. Bacteriol. 100: 1121�1128.Gutierrez J.A., P. Nichols and I. Couperwhite. 1999. Changesin whole cell-derived fatty acids induced by benzene and occur-

rence of the unusual 16:1T6c in Rhodococcus sp. 33. FEMSMicrobiol. Lett. 176: 213�218.Härtig C., N. Loffhagen and H. Harms. 2005. Formation of transfatty acids is not involved in growth-linked membrane adaptationof Pseudomonas putida. Appl. Environ. Microbiol. 71: 1915�1922.Heesche-Wagner K., T. Schwarz and M. Kaufmann. 1999.Phenol degradation by an enterobacterium: a Klebsiella strain car-ries a TOL-like plasmid and a gene encoding phenol hydroxylase.Can. J. Microbiol. 45: 162�171.Heipieper H.J., R. Dieffenbach and H. Keweloh. 1992. Conver-sion of cis unsaturated fatty acids to trans, a possible mechanismfor the protection of phenol-degrading Pseudomonas putida P8from substrate toxicity. Appl. Environ. Microbiol. 58: 1847�1852.Heipieper H.J., F.J. Weber, J. Sikkema, H. Keweloh andJ.A.M. de Bont. 1994. Mechanisms of resistance of whole cellsto toxic organic solvents. Trends Biotechnol. 12: 409�415.Kao C.M., J.K. Liu, Y.L.Chen, C.T. Chai and S.C. Chen. 2005.Factors affecting the biodegradation of PCP by Pseudomonasmendocina NSYSU. J. Hazard. Mat. B124: 68�73.Kabelitz N., P.M. Santos and H.J. Heipieper. 2003. Effect ofaliphatic alcohols on growth and degree of saturation of mem-brane lipids in Acinetobacter calcoaceticus. FEMS Microbiol.Lett. 220: 223�227.Keweloh H., G. Weyrauh and H.J. Rehm. 1990. Phenol-in-duced membrane changes in free and immobilized Escherichiacoli. Appl. Environ. Biotechnol. 33: 66�71.Kim I.S., H. Lee and J.T. Trevors. 2001. Effects of 2,2�,5,5�-tetrachlorobiphenyl and biphenyl on cell membranes of Ralstoniaeutropha H850. FEMS Microbiol. Lett. 200: 17�24.Kojima Y., N. Itada and O. Hayaishi. 1961. Merapyrocatechasea new catechol cleaving enzyme. J. Biol. Chem. 236: 2223�2231.Loh K.C. and S.J. Wang. 1998. Enhancement of biodegradationof phenol and a nongrowth substrate 4-chlorophenol by mediumaugmentation with conventional carbon sources. Biodegradation8: 329�338.Lurie J. and I. Rybnikova. 1968. Chemical Analysis of Indus-trial Sewages (in Russian). Gaschmizdat, Moskwa.Morales G., J.F. Linares, A. Belosso, J.P. Albar, J.L Martizezand F. Rojo. 2004. The Pseudomonas putida Crc global regula-tor controls the expression of genes from several chromosomalcatabolic pathways for aromatic compounds. J. Bacteriol. 186:1337�1343.Mrozik A., S. £abu¿ek and Z. Piotrowska-Seget. 2005. Changesin fatty acid composition in Pseudomonas putida and Pseudomo-nas stutzeri during naphthalene degradation. Microbiol. Res. 160:149�157.Mrozik A., Z. Piotrowska-Seget and S. £abu¿ek. 2006. Cellu-lar fatty acid patterns in Pseudomonas sp. CF600 during catecholand phenol degradation in media supplemented with glucose asan additional carbon source. Ann. Microbiol. 56: 57�64.Petruschka L., G. Burchhardt, C. Müller, C. Weihe and H.Herrmann. 2001. The cyo operon of Pseudomonas putida is in-volved in carbon catabolite repression of phenol degradation. Mol.Genet. Genomics 266: 199�206.Ramos J.L., E. Duque, J.J. Rodriquez-Herva, P. Godoy, A. Hai-dour, F. Reyes and A. Fernanadez-Barrero. 1997. Mechanismsfor solvent tolerance in bacteria. J. Biol. Chem. 272: 3887�3890.Saier M.H. Jr. 1996. Regulatory interactions controlling carbonmetabolism: an overview. Res. Microbiol. 147: 639�447.�ajbidor J. 1997. Effect of some environmental factors on thecontent and composition of microbial membrane lipids. Crit. Rev.Biotechnol. 17: 87�103.Sasser M. 1990. Identification of bacteria by gas chromatographyof cellular fatty acids. MIDI Technical Note 101. Microbial ID,Inc., Newark (USA).Shinitzky M. 1984. Physiology of Membrane Fluidity. Vol. 1,pp. 1�52, CRC Press, Boca Raton , USA.


Sikkema J., F.J. Weber, H.J. Heipieper and J.A.M. de Bont.1994. Cellular toxicity of lipophilic compounds: mechanisms,implications and adaptations. Biocatalysis 10: 113�122.Sikkema J., J.A.M. de Bont and B. Poolman. 1995. Mechanism ofmembrane toxicity of hydrocarbons. Microbiol. Rev. 59: 201�222.Stülke J. and W. Hillen. 1999. Carbon catabolite repression inbacteria. Curr. Opin. Microbiol. 2: 195�201.Tian L., P. Ma and J-J. Zhong. 2003. Impact of the presence ofsalicylate or glucose on enzyme activity and phenanthrene degrada-tion by Pseudomonas mendocina. Proc. Biochem. 38: 1125�1132.Tsitko I.V., G.M. Zaitsev, A.G. Lobanok and M.S. Salkinoja-Salonen. 1999. Effect of aromatic compounds on cellular fatty acidcomposition of Rhodococcus opacus. Appl. Environ. Microbiol.65: 853�855.Van Herwijnen R., B.F. van de Sande, F.W.M. van der Wielen,D. Springael, H.A.J. Govers and J.R. Parsons. 2003. Influenceof phenanthrene and fluoranthene on the degradation of fluorene

and glucose by Sphingomonas sp. strain LB126 in chemostat cul-tures. FEMS Microbiol. Ecol. 46: 105�11.Wang K.W., B.C. Baltzis and G.A. Lewandowski. 1996. Kine-tics of phenol biodegradation in the presence of glucose.Biotechnol. Bioeng. 51: 87�94.Wang S.J. and K.C. Loh. 1999. Facilitation of cometabolic de-gradation of 4-chlorophenol using glucose as an added growthsubstrate. Biodegradation 10: 261�269.Wang S.J. and K.C. Loh. 2001. Biotransformation kinetics of Pseu-domonas putida for cometabolism of phenol and 4-chlorophenol inthe presence of sodium glutamate. Biodegradation 12: 189�199.Weber F.J. and J.A.M. de Bont. 1996. Adaptation mechanismsof microorganisms to the toxic effects of organic solvents onmembranes. Biochem. Biophys. Acta 1286: 225�245.Yu J. and O.P. Ward. 1994. Studies on factors influencing thebiodegradation of pentachlorophenol by mixed bacteria culture.Int. Biodeter. Biodeg. 34: 209�221.


ORIGINAL PAPER

Introduction

Quantitative antibiotic sensitivity can be estimatedby either agar/broth dilution methods, or by agar dif-fusion methods. The minimum inhibitory concentra-tion (MIC) in dilution methods is not affected by thegrowth of standard bacterial inoculums because theantibiotic is completely active from the beginning ofthe incubation period. On the other hand, in diffusionmethods (Etest, multidisc method) the concentrationof the antibiotic at the edge of inhibition zone at thetime of its formation relates to the growing bacterialpopulation. This antibiotic concentration, the actualbacterial density and the time when the edge of inhi-bition zone is founded are called the critical concen-tration (Cc), the critical population and the criticaltime, respectively (Linton, 1961; Cooper, 1963; Barry,1980; Delignette-Muller and Flandrois, 1994). Thecritical time generally lasts for a period of severalhours after the start of incubation. In slow growingbacteria, the critical time is longer. The critical popu-lation is therefore higher than the standard inoculumand, consequently, the critical concentration can theo-retically differ from MIC. The question we hope toanswer is, does it really differ? If yes, then the Etest

wouldn�t work properly. However, the Etest has beenreported to be reliable by many authors, although ithas not been well studied (Kronvall, 2000). The pur-pose of this paper is to uncover the theoretical expla-nation for Etest accuracy.

How the antibiotic critical concentration can bemeasured? If two or more discrete discs with differentamounts of the same antibiotic are used, the criticalconcentration of the antibiotic can be calculatedfrom the content of the discs and from the diameter oftheir respective inhibition zones. The following rulesapply to the theory of inhibition zone formation in thedisc method:

● The critical time for the same bacteria strain,antibiotic, and cultivation conditions is indepen-dent of the amount of antibiotic in the disc (ofthe disc content) (Barry, 1980).

● As soon as the edge of a zone is formed, itsdiameter will not change (Linton, 1961; Barry,1980).

● The relation between the area of inhibition zoneand natural logarithm of antibiotic disc con-tent is linear (Barry, 1980; Kronvall, 1982;Delignette-Muller and Flandrois, 1994).

Reliability of the Etest in Light of the Correlation between an Antibiotic�s CriticalConcentration (Cc) and MIC Values

MAREK BEDNAR*

Department of Medical Microbiology, Charles University, 3rd Faculty of Medicine, Prague, Czech Republic

Received 24 February 2007, revised 20 July 2007, accepted 24 July 2007

A b s t r a c t

The study relates to the theory of diffusion methods for antibiotic sensitivity testing. The aim of the study was to show the relationshipbetween the antibiotic critical concentration (Cc) and its minimum inhibitory concentration (MIC). The results contribute to the explana-tion of the Etest�s reliability and support the scientific basis for MIC determination using agar diffusion methods. Susceptibility among 90clinical isolates of 12 common aerobic bacterial species to gentamicin, erythromycin, or oxacillin was assessed using the multidisc method(for Cc), by the agar dilution method (for MIC) and by the Etest. The results of all three methods were statistically compared and found tobe closely related. The regression equation for Cc values and MIC was log

2(MIC) = 0.99×log

2(Cc)�0.13; r = 0.99; p<0.05; the regression

equation for Cc values and Etest-MIC (Et) was log2(Et) = 0.86×log

2(Cc)+0.34; r = 0.96; p<0.05; the regression equation for Etest-MIC

values and MIC was log2(MIC) = 1.12×log

2(Et)�0.50; r = 0.96; p<0.05.

K e y w o r d s: Etest, multidisc method, antibiotic critical concentration (Cc), antibiotic sensitivity testing

* Corresponding author: M. Bednar, Department of Medical Microbiology, Charles University, 3rd Faculty of Medicine, Ruska 87,100 00 Prague, Czech Republic; phone: (42) 26 7162580; fax: (42) 26 7162516; e-mail: [email protected]

166 Bednar M. 3

● The antibiotic diffuses into the agar according toFick�s second law of diffusion (Vesterdal, 1947;Humphrey and Lightbown, 1952; Koch, 1999).

Note on rule 3): the relation between the radius ofthe inhibition zone and antibiotic disc content is ex-pressed by the equation

{1}

where rn is the radius of inhibition zone in n-th discand mn is the content of the disc. A graphic image ofthis dependency is a straight line. Its intercept a andslope b can be calculated from the measured data bycommon statistical procedures. Measurement of thezone radius from the center of discs provides a bettercorrelation with disc content and size of inhibitionzone, in terms of equation {1}, than does measure-ment of the zone radius starting from the edge of thedisc (Kronvall, 2000).

Note on rule 4): a derived radial law for two-dimensional cylindrical diffusion applies to the discmethod (Humphrey and Lightbown, 1952; Koch, 1999).

{2}

where Cc (µg/ml) is the critical concentration of anti-biotic, mn (µg) is the antibiotic disc content at thebeginning of diffusion, rn (cm) is the radius of therespective inhibition zone, measured from the disccenter D (cm2/s) is the diffusion constant of the anti-biotic, tc (s) is the critical time, d (cm) is the agardepth, and e is the base of natural logarithms. Thedenominator 4BDtcd reflects the drop in antibioticconcentration in the disc during the critical time andthe agar depth. Dimensional analysis of the formulayields µg/ml, which confirms the mathematical valid-ity of the formula.

It can be proven that b = 4Dtc. This is in agree-ment with the research done by others (Cooper, 1963;Drugeon et al., 1987; Delignette-Muller and Flandrois,1994; Koch, 1999). In formula {2}, you can substi-tute slope b for 4Dtc and we can substitute a + bln(mn)for r2

n. By consequent reduction, we arrive at resultingformula combining equations {1} and {2}

*{3}

where a is the intercept, b is the slope of regres-sion line according to equation {1}, and d is the agardepth. Formula {3} calculates the critical concentra-tion of an antibiotic without knowing its diffusionconstant or critical time (i.e., without calibration). Ifwe use at least three discs, the reliability of the resultcan be ascertained from the correlation of the loga-rithms of the discs� content and the sizes of the re-spective inhibition zones.

Experimental


Susceptibility among ninety clinical isolates ofcommon aerobic bacterial species to gentamicin,erythromycin, or oxacillin was assessed. We performed90 concurrent sensitivity measurements using themultidisc diffusion method, standard agar dilutionmethod and Etest. Sensitivity was measured 33 timesfor oxacillin, 27 times for gentamicin, and 30 timesfor erythromycin. These antibiotics were chosen due totheir diverse mode of action. They also provide sharp,clear zones of inhibition in the diffusion method.

Bacterial strains and culture media. Forty-fivestrains of Staphylococcus aureus, seventeen strains ofStaphylococcus epidermidis, six strains of Pseudomo-nas aeruginosa, six strains of Proteus mirabilis, sixstrains of Escherichia coli, four strains of Klebsiellaspp., three strains of Acinetobacter baumanii, twostrains of Enterobacter spp. and one strain of Entero-coccus faecalis were used. All the strains were iso-lated during routine investigations of various clinicalspecimens in a hospital laboratory. The strains werechosen according to the qualitative sensitivity testing(NCCLS, 1993a) so as to obtain three sets of strainsfor measurements of the sensitivity to given antibioticover a wide range of MIC values. Erythromycin oroxacillin sensitivity was checked in Gram-positiveswhereas gentamicin sensitivity in Gram-negatives. Theactual experiment involved a quantitative assessmentof the sensitivity to a particular antibiotic in selectedstrains using the multidisc method, agar dilutionmethod, and Etest. The tests were run concurrently foreach inoculum. We used Mueller-Hinton agar (Oxoid,Unipath Ltd., Basingstoke, Hampshire, England)poured to a depth of 4 mm in Petri dishes (diameter90 mm) for all three methods. The inocula for all threemethods came from colonies that were suspended inphysiological saline solution to a density of 0.5 onthe McFarland scale. Plates used for the multidiscmethod and the Etest were inoculated by swabbing.

Etest. Etest strips (Gentamicin low range; Erythro-mycin; Oxacillin) were used according to the manu-facturer�s instructions (AB Biodisk, Solna Sweden).Agar plates were incubated at 35°C. Results wereread after 24 hours according to the manufacturer�sreading guide.

Antibiotics used. Antibiotic solutions were pre-pared by dilution of injectable preparations: Gentami-cin LEK Pharmaceuticals and Chemical Co., Slovenia(gentamicin), Erythrocin Abbott Laboratories, USA(erythromycin), and Prostaphilin Bristol-Myers SquibbS.p.A., Italy (oxacillin).

Critical concentration (Cc) determinations(multidisc diffusion method). Four 6 mm � diameterblank paper discs (Oxoid, Unipath Ltd., Basingstoke,

r2n = a + b ln(m

n)

* Equation {3} used in this work was briefly described in theMedical Science Monitor (2000) 6: 168�170.

Cc = e�m

n

4BDt, d

r2n

4Dtc

Cc = e� ab

1Bbd

167Realibility of Etest, correlation between MIC and antibiotic Cc3

Hampshire, England) were placed onto the inoculatedagar. An appropriate amount of antibiotic, dissolvedin 20 µl distilled water, was dropped onto discs usinga pipette. The amount of antibiotic on the discs wasestimated on the basis of previous qualitative sensi-tivity testing of all strains (NCCLS, 1993a) so as toachieve a minimum of two measurable yet distinct in-hibition zones when using four discs (Table I). Agarplates were incubated at 35°C for 24 hours. Thediameters of inhibition zones were measured using anelectronic calliper. Non-rounded zone diameters anddata on antibiotic disc content were used to calculatethe slope b and intercept a according to equation {1}.These constants were subsequently employed for thecalculation of the critical concentration of the antibi-otics using formula {3}. Where at least three inhibi-tion zones formed, the percentage of variation in thezones explained by the disc content logarithms (coeffi-cient of determination R2 written as a percentage) wasalso calculated to show the reliability of the result.

MIC determination. The assessment of MIC us-ing the agar dilution method was performed in accor-dance with NCCLS guidelines (NCCLS, 1993b). Theantibiotics were the same as those used in themultidisc diffusion method. Final concentrations ofantibiotics in the agar ranged from 0.012 to 512 mg/l.The dilutions were based on a geometrical order (fac-tor 2) and were related to concentrations of 1 mg/l and1.5 mg/l (0.012, 0.016, 0.023, 0.031, 0.047, 0.063,0.094, 0.125, 0.19, 0.25, 0.38, 0.5, 0.75, 1.0, 1.5, 2, 3,4, 6, 8, 12, 16, 24, 32, 48, 64, 96, 128, 192, 256, 384,512 mg/l). The inoculum was applied to the agar sur-face by means of a pin replicator. Agar plates wereincubated at 35°C. Results were read after 24 hours.

Statistical evaluation of results. The results wereclassified by the applied method only, not by bacte-rial strain or the antibiotic used. The correlations ofEtest results with MIC, multidisc diffusion method re-sults with MIC and multidisc diffusion method resultswith Etest results, were expressed using the Spearmanrank correlation coefficient. After transforming the re-sults into base 2 logarithms, we expressed them bymeans of the Pearson correlation coefficient. Statisti-cal values were calculated by Statistica for Windows(StatSoft Inc.). Differences no greater than a twofolddilution factor between the MIC and the Etest or be-tween the MIC and the multidisc method, were usedto calculate agreement (Pfaller et al., 2000).

Results

Agreement of Cc with MIC was observed in 89out of 90 concurrent critical concentration measure-ments. Regression straight line is shown in Figure 1.We observed two disagreements between the MIC and

Fig. 1. Correlation of results of sensitivity measurements by multidisc method (Cc � critical concentration) and agar dilutionmethod (MIC � minimum inhibitory concentration). Broken lines indicate the 95% confidence band; some points overlap

Table IAntibiotic disc content in individual discs A, B, C, and D

in multidisc method rounded to three significant digits

In order that the number of discs in the multidisc method can be reducedto four, strains were initially subdivided into groups that were eithersensitive or resistant using a routine qualitative disc method. Because ofpractical reasons, the ratio between the neighboring discs� contents isusually (but not obligatory) regular.

Erythromycin 1000 200 40.0 8.00 200 40.0 8.00 1.60

Gentamicin 800 256 64.0 16.0 256 64.0 4.00 1.00

Oxacillin 5000 714 102 14.6 102 14.6 2.08 0.298

Antibiotic

Disc content (µg)

Resistants Sensitives

DCBADCBA

168 Bednar M. 3

the Etest. All disagreements occurred among resistantstrains. Both the multidisc method and the Etestcorrelated well with the dilution agar method: boththe Spearman and Pearson coefficients reached at least0.9. The relation between the results of the multi-disc method (Cc) and MIC was log2(MIC) = 0.99××log2(Cc)0.13; r = 0.99; p<0.05. The relation betweenthe Etest results (Et) and MIC was log2(MIC) =1.12×log2(Et)0.50; r = 0.96; p<0.05. The relation be-tween Cc and Et was log2(Et) = 0.86×log2(Cc) + 0.34;r = 0.96; p<0.05. The reliability of the multidiscmethod expressed as the average percentage of varia-tion in the zones explained by the disc content loga-rithms was 98.85% (92.36�100.00%). The b value informula {1} was not related to the sensitivity of thestrains. Thus, the critical time was independent fromstrain sensitivity. Neither species-dependent nor anti-biotic type-dependent irregularities in Cc-MIC rela-tionship were found.

Discussion

According to the classical theories (Cooper, 1963;Barry, 1980; Hedges, 1999), the bacterial growth rateimpacts the inhibitory zone diameter. In the Kirby-Bauer qualitative disc diffusion method, this pheno-menon is solved by the interpretative standards(NCCLS, 1993a). The crucial question is whethersuch �inaccuracy� substantially influences the Etestresult. The results of our work show that it does not,because the antibiotic (critical) concentration under-neath the edge of the forming zone practically equalsMIC. So in Etest the bacterial growth rate mayimpact the inhibitory zone shape but not its point ofintersection with the scale on the strip.

In the multidisc method, the concentration of anti-biotic on each additional discs is, optimally, four toseven times lower than on the preceding disc (depend-ing on the type of antibiotic). Because the experimen-tal design initially divided the strains into either sen-sitive or resistant, we were able to reduce the numberof discs used to four and still obtain at least two mea-surably distinct zones of inhibition for an accurate cal-culation of the critical concentration. Without this ini-tial categorization, five or six discs would be requiredto test over the full scale of an Etest strip. Such increas-ing disc number brings the multidisc method closer tothe Etest, which can be imagined as a chain of antibio-tic discs with exponentially growing antibiotic con-tent. The primary data in both methods are the criticalconcentrations of antibiotics. The Etest and multidiscmethod do have similarities � the zones (including thezero zones) are always formed after the critical timepasses. In the case of the Etest, the critical concentra-tion if r = 0 estimates MIC using a printed scale.

The correlation between the critical concentrationand MIC is not a new finding. Nevertheless, the extentof the correlation is surprising. It implies that in dif-

fusion quantitative methods, bacterial growth up untilthe critical time does not influence the result. Thisobservation contributes to an understanding of theaccuracy of the Etest on a wide variety of organisms,and indicates that the results obtained with quantita-tive diffusion methods (E-test, multidisc method) canbe expressed as MICs without any conversion.

AcknowledgementThe author would like to thank Marie Duskova and Zorka

Haasova for their technical laboratory assistance and Jiri Horacekfor the software production. The program for the critical concen-tration calculation is downloadable on http://www.lf3.cuni.cz/ustavy/mikrobiologie/download/atb_cc.zip; last accessed 7/07/07.

Literature

Barry A.L. 1980. Procedure for testing antibiotics in agar media:Theoretical consideration, pp: 1�23. In: V. Lorian (ed.) Antibioticsin Laboratory Medicine. The Williams and Wilkins Co., Balti-more/London.Cooper K.E. 1963. The theory of antibiotic inhibition zones,pp: 1�85. In F. Kavanagh (ed.) Analytical Microbiology. AcademicPress, New York and London.Delignette-Muller M.L. and J.P. Flandrois. 1994. An accuratediffusion method for determining bacterial sensitivity to antibio-tics. J. Antimicrob. Chemother. 34: 73�81.Drugeon H.B., M.E. Juvin, J.Caillon and A.L. Courtieu. 1987.Assessment of formulas for calculating critical concentration bythe agar diffusion method. Antimicrob. Agents Chemother. 31:870�875.Hedges A.J. 1999. The influence of factors affecting the �criticalpopulation� density of inocula on the determination of bacterialsusceptibility to antibiotics by disc diffusion methods. J. Anti-microb. Chemother. 43: 313.Humphrey J.H. and J.W.Lightbown. 1952. A general theory forplate assay of antibiotics with some practical applications. J. Gen.Microbiol. 7: 129�143.Koch A.L. 1999. Diffusion through agar blocks of finite dimen-sions: a theoretical analysis of three systems of practical signifi-cance in microbiology. Microbiology 145: 643�654.Kronvall G. 1982. Analysis of a single reference strain for deter-mination of gentamicin regression line constants and inhibitionzone diameter breakpoints in quality control of disk diffusion anti-biotic susceptibility testing. J. Clin. Microbiol. 16: 784�793.Kronvall G. 2000. MIC determination of fusidic acid and ofciprofloxacin using multidisk diffusion tests. Clin. Microbiol. In-fect. 6: 483�489.Linton A.H. 1961. Interpreting antibiotic sensitivity tests. J. Med.Lab. Technol. 18: 1�20.NCCLS. 1993a. National Committee for Clinical laboratory Stan-dards. Performance Standards for Antimicrobial Disk Suscepti-bility Tests � Fifth Edition: Approved Standard M2-A5. NCCLS,Villanova, PA, USA.NCCLS. 1993b. National Committee for Clinical laboratory Stan-dards. Methods for Dilution Antimicrobial Susceptibility Tests forbacteria that Grow Aerobically � Third Edition: Approved Stan-dard M7-A3. NCCLS, Villanova, PA, USA.Pfaller M.A., S.A. Messer, A. Houston, K. Mills, A. Bolmstromand R.N. Jones. 2000. Evaluation of the Etest method for deter-mining voriconazole susceptibilities of 312 clinical isolates of Can-dida species by using three different agar media. J. Clin. Microbiol.38: 3715�3717.Vesterdal J. 1947. Studies on the incubation zones observed inthe agar cup method for penicillin assay. Acta Pathol. Microbiol.Scand. 24: 272�282.


ORIGINAL PAPER

Introduction

Nosocomial infections are an important healthproblem worldwide and are closely related to the typeof diagnostic and therapeutic procedures performedon patients. Isolating the pathogen responsible for aninfection is one of the basic criteria for the identifyingthe type of infection, which considerably improvesa chance for therapy which should be preceded byantibiogram determination (Toltzis et al., 2001). Uri-nary tract infections caused by Proteus mirabilis arecommon and often severe, leading to acute pyelone-phritis, chronic inflammation, and bacteremia. Thefrequency of P. mirabilis infections in hospital pa-tients as well as in outpatients increases the risk ofendogenous infections, being infected by other pa-tients, hospital staff, or contaminated equipment. Itshows that the keeping the record of the exact charac-teristics of those microorganisms becomes necessary(Johnson et al., 1993).

Due to the increased antibiotic resistance, it beco-mes necessary to control the appearance of P. mirabilisstrains isolated from infections in hospital environment.

Wild-type strains of P. mirabilis are usually susceptibleto $-lactams. However, a progressive increase in $-lac-tam resistance, mediated by the production of $-lacta-mases, has occurred in this species (Perilli et al., 2002).

The most recent advances in molecular biology of-fer promising possibilities of examining epidemiolo-gical bacteria strains in a controlled hospital environ-ment. It is possible to determine the genetic profile ofthose microorganisms using pulsed field gel electro-phoresis (PFGE), which can be used for analysis ofchromosomal DNA restriction patterns, a gold stan-dard in hospital epidemiology. Demonstrating evidentrelationship among isolated strains from different hos-pital wards within a few years period indicates persis-tence of the population of microorganisms responsiblefor appearance of clonal outbreaks (Fernandez-Bacaet al., 2001; Hennekinne et al., 2003).

The aim of the present study was to characteriseP. mirabilis strains, isolated during 5 years period inthe West Pomeranian area of Poland, by moleculartyping using PFGE procedure. The results obtainedwith the application of PFGE were then compared toantimicrobial resistance patterns.

Antibiotic Susceptibility and Molecular Characterisationof Proteus mirabilis Isolates

in Hospitals from the West Pomeranian Area of Poland

IWONA M¥CZYÑSKA and STEFANIA GIEDRYS-KALEMBA

Department of Microbiology and Immunology, Pomeranian Medical University, Szczecin, Poland

Received 23 April 2007, revised 11 July 2007, accepted 31 July 2007

A b s t r a c t

Proteus mirabilis isolates (n = 177), collected between 1996 and 2000 in four hospitals in the West Pomeranian area of Poland, werecharacterized by antibiotype and pulsed-field gel electrophoresis (PFGE). The selected isolates were collected from different wards (inten-sive care unit, surgery, internal medicine, and urology). The strains were cultured from various specimen types, mostly from urine, woundsamples, bronchial exudates and sputa. The identification was done by biochemical test ID 32E ATB (bioMerieux). Analysis of PFGEpatterns was based on comparison of the banding patterns obtained by PFGE of chromosomal DNA digested with SfiI enzyme. Among allP. mirabilis isolates tested three major genotypes A (A1-A7), B (B1-B4), C (C1-C5) and 71 unique patterns were identified. The samegenotypes were obtained from different patients, treated in different wards and hospitals during a 5-year period. The strains w hichbelonged to the genotypes A and B were multiresistant and most of them produced ESBL; genotype C was more sensitive to antibiotics.

K e y w o r d s: Proteus mirabilis, antibiotic susceptibility, ESBL, nosocomial infection, PFGE

* Corresponding author: I. M¹czyñska, Department of Microbiology and Immunology, Pomeranian Medical University, PowstañcówWielkopolskich 72, 70-111 Szczecin, Poland; phone: (48) 91 4661654; e-mail: [email protected]

170 M¹czyñska I. and Giedrys-Kalemba S. 3

Experimental

Material and Methods

Bacterial isolates and clinical data. 177 P. mira-bilis clinical isolates were collected between 1996 and2000 from different patients in 4 hospitals in the WestPomeranian area: Clinical Hospital No 2 in Szczecin(SZC; n = 152 isolates � all isolates), 3 municipal hos-pitals (some resistant strains): Police (POL; 6 iso-lates), Choszczno (CHO; 2 isolates), Gryfice (GRY;17 isolates). The selected isolates were collected fromdifferent wards (intensive care unit, surgery, internalmedicine, and urology). The strains were cultured fromvarious specimen types, mostly from urine (124 iso-lates; 70%), wound samples (39 isolates; 22.1%),bronchial exudates (6 isolates; 3.4%) and sputa (8 iso-lates; 4.5%). Isolates were identified to the species bythe ID 32E ATB test (bioMerieux).

Antimicrobial susceptibility testing. The suscep-tibility to antibiotics was tested by the disk diffusionmethod on Mueller-Hinton agar according to the cri-teria of the Clinical and Laboratory Standards Insti-tute. The following amounts of antibiotics per discwere used: ampicillin � Amp (10 µg), amoxycillin/clavulanic acid � Amc (20/10 µg), piperacillin � Pip(100 µg), piperacillin/tazobactam � Tzp (100/10 µg),cephalotin � Cf (30 µg), cefuroxime � Cxm (30 µg),cefotaxime � Ctx (30 µg), ceftazidime � Caz (30 mg),imipenem � Imp (10 mg), gentamicin � Gn (10 µg),tobramycin � Tob (10 µg), netilmicin � Net (30 µg),amikacin � An (30 µg), pipemidic acid � Pi (30 µg),pefloxacin � Pef (5 µg), norfloxacin � Nor (10 µg),ciprofloxacin � Cip (5 µg) trimethoprim-sulphametho-xazole � Sxt (1.25/23.75 µg).

All isolates were recognized as ESBL producersby the double-disc test. Double-disk synergy test wasperformed on Mueller-Hinton agar with a centralamoxycillin-clavulanic acid disk and disks of the third

generation cephalosporins (cefotaxime, ceftazidime)placed 20 mm (centre to centre) from each other. Thetest was considered to be positive for ESBL produc-tion when the bacterial growth had a �champagne cork�appearance. For each strain the test was repeated twice.

Molecular typing. Isolates were typed by deter-mining PFGE SfiI DNA macrorestriction patternswith the GenePath Group 5 Reagent Kit (Bio-RadLaboratories) according to the manufacturer�s recom-mendation. Pulsed-field gel electrophoresis (PFGE)was performed using the GenePath System (Bio-Rad).Differences detected in band patterns analysed usingMolecular Analyst Fingerprinting software (Bio-Rad). The PFGE pattern was interpreted according toTenover et al., (1995) recommendations.

Results

SfiI PFGE patterns. Analysis of PFGE patternswas based on comparison of the banding patterns ob-tained by PFGE of chromosomal DNA digested withthe SfiI enzyme. Among all P. mirabilis isolates inthe collection, three major types A: n = 48 (subtypes:A1 � 14 strains, A2 � 16, A3 � 6, A4 � 5, A5 � 4, A6� 2, A7 � 1), B: n = 49 (subtypes: B1 � 36 strains, B2� 9, B3 � 2, B4 � 2), C: n = 9 (subtypes: C1 � 2 strains,C2 � 3, C3 � 1, C4 � 2, C5 � 1). The remaining strainshad some unrelated PFGE patterns (more than sixband differences), which were designated by romannumerals and letter P: P1 � P71 (Fig. 1).

The data for the three major types A (A1-A7), B(B1-B4), C (C1-C5) from hospitals in the West Po-meranian area are presented in Table I.

Molecular typing and antibiotic resistance. Ge-notype A (A1-A7) strains (n = 48) were resistant toampicillin, amoxycillin/clavulanic acid, piperacillin,cephalotin, cefuroxime, gentamicin, tobramycin, netil-micin, amikacin, pipemidic acid, pefloxacin, norflo-

Urology SZCa A1, A2, A3, A4, B1, C1, C2 A1, A2, A5, B1, C4 A1, A6, B1, B2, C4 A2, B1 A3, B1, B2, B4

Surgery II SZCa A1 A1 A1, A2

Surgery III SZCa C2 A5, A6, B2, B3 A1, A2, A7 A1

Intensive Care Unit SZCa A2, A3, C1 A3, A4, A5, B2 A2, B1

Internal Medicine SZCa A1, A2, A4, A5, C2, C3 A2, B1 B1, B4 A2, B1, C5 B1, B2

POLb B1

CHOc A4 A4

GRYd B1

Table IThe distribution of the three major types A (subtypes A1-A7), B (B1-B4), C (C1-C4) P. mirabilis strains isolated from four

hospitals in the West Pomerania between 1996 and 2000

Hospital (ward) 1996

a SZC � Clinical Hospital No 2 in Szczecin; b POL � Municipal Hospital in Police; c CHO � Municipal Hospital in Choszczno;d GRY � Municipal Hospital in Gryfice

1997 1998 1999 2000

171Antibiotic susceptibility of P. mirabilis hospital isolates3

xacin, ciprofloxacin, trimethoprim-sulphamethoxazole.Only 3 isolates were sensitive to cefotaxime, 5 toceftazidime, all strains were sensitive to imipenem.Only 5 (10.4%) isolates were resistant to piperacillin/tazobactam.

Genotype B (B1-B4) strains (n = 49), were resis-tant to ampicillin, amoxycillin/clavulanic acid, pipera-cillin, cephalotin, cefuroxime, gentamicin, tobramycin,netilmicin, amikacin, pipemidic acid, pefloxacin, nor-floxacin, ciprofloxacin, trimethoprim-sulphamethoxa-zole. Most of the genotype B strains were resistant tocefotaxime (96%) and ceftazidime (92%), 10 strains(20.4%) to piperacillin/tazobactam. One of the isolatesrepresenting genotype B sshowed very high antibioticresistance, including imipenem. This strain was isola-ted from a patient�s urine in the internal medicine SZC.

All genotype C (C1-C5) strains (n = 9) were resis-tant to ampicillin, cephalotin and trimethoprim-sulphamethoxazole. The 3 P. mirabilis isolates wereresistant to amoxycillin/clavulanic acid, 6 to pipera-cillin, 2 to cefuroxime, 1 to cefotaxime and ceftazidime,7 to pipemidic acid, 2 to pefloxacin, norfloxacin andciprofloxacin. All strains were susceptible to pipera-cillin/tazobactam, imipenem, netylmicin and amikacin.

Identification of $-lactamases. Thirty four iso-lates were found to be ESBL producers. The numberof strains producing ESBL isolated in particular yearsshowed an increasing trend. In 1996 only 2 (5.9%)strains showed the presence of ESBL, in 1997 � 4(11.8%), in 1998 � 6 (17.6%), in 1999 � 7 (20.6%),

and in 2000 � 15 (44.1%). All of the P. mirabilisESBL producers belonged to multiresistant strainsand to genotypes A1 (5 strains), A2 (7), A5 (2), A7 (1)and B1 (14), B2 (4), B3 (1).

The following antibiotic resistance patterns wereobserved:AmpAmcPipCflCxmCtxCazGnTobNetPiNorPefCipSxt

(20 isolates; 58.8%)AmpAmcPipCflCxmCtxCazGnTobNetAnPiNorPefCipSxt

(13 isolates; 38.3%)AmpAmcPipCflCxmCtxCazGnPiNorPefCipSxt

(1 isolates; 2.9%)

Discussion

Despite the progress of knowledge, improved pre-ventive and scrutiny procedures, hospital infectionsstill pose a serious clinical, therapeutic and epide-miological problem. The most common are urinarytract infections that comprise 35�45% of all hospitalinfections and often cause dangerous diseases suchas septicemia, pyelonephritis or wound infections.The most common pathogen isolated in urinary tractinfections is still Escherichia coli. However, otherbacteria including representatives of Proteus genera(especially P. mirabilis) begin to play ever greaterrole in pathogenesis of hospital infections, especiallyin urology wards (Chippendale et al., 1994; Claphamet al., 1990).

Fig. 1. Examples of pulsed-field gel electrophoresis profiles obtained for P. mirabilis isolates.Lanes: 1and 16, 8 ladder used as molecular size (MW) markers; 2�5, subtype C1 and C2; 6�12,

subtype A1 and A2, A3; 13�14, subtype B1; 15, unrelated pattern P4.

C 1 C 1 C 2 C 2 A 1 A 2 A 2 A 3 A 3 A 3 A 3 B 1 B 1 P 4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

8

48.5 kb

970 kb

8


P. mirabilis is the second most common cause ofurinary tract infections, and it is a frequent causeof nosocomial infections as well. It was confirmed inour own research, as much as 70% isolates isolatedfrom urine of patients suffering for urinary tract in-fections were classified as P. mirabilis, and only 30%from other materials. The existence of multiresistantP. mirabilis strains in hospital environment makesconstant monitoring for presence of those micro-organisms in specific hospital wards a necessity. It isalso necessary to monitor emerging new resistancemechanisms as well as transmission of strains be-tween patients and wards as it is commonly done incase of other bacteria (Bonnet et al., 2002; Mammeriet al., 2001).

In our research we used PFGE as a method foranalysing restriction patterns of chromosomal DNAof P. mirabilis strains in order to classify them intoparticular genotype. It has been shown that up to 60%of P. mirabilis strains isolated from hospital infectionsbelong to the three main genotypes A, B, C.

Type A and B showed high antibiotic resistance toall tested aminoglycosides and chinolons, and mostof $-lactam antibiotics. One strain belonging to geno-type B also showed resistance to imipenem. SubtypesC1-C5 differed from types A and B in regard to anti-biotics susceptibility. All those strains showed highersensitivity to antibiotics, but the diversity of antibi-otic profiles was observed among subtypes.

Demonstrating clear genetic relationship betweenP. mirabilis strains isolated from different environ-ments within a few years period indicates existenceof outbreak clones in wards of the clinical hospital aswell as the transmission of those strains to other hos-pitals in the region. $-lactamases producing P. mira-bilis strains with extended spectrum, ESBL, poseadditional therapeutic problems in treatment of infec-tions. The selection of ESBL producers occurs mostfrequently in surgery, urology and neonatal wards(Goering 1993; Saladin et al., 2002). In our researchseven-fold increase of ESBL producing strains wasobserved: from 5.9 % in 1996 to 44.1% in 2000 year.

ESBLs are functionally differentiated. Some ofthem efficiently hydrolyse cefotaxime and ceftriaxonbut not ceftazidime, while others clearly preferceftazidime and aztreonam. These differences maycause problems, particularly in interpretation of sen-sitivity to combinations of penicillins with inhibitors.ESBL are inhibited by $-lactamase inhibitors, butsome $-lactamases hydrolyse penicillin to a consider-able degree and so the inhibition effect may not suf-fice for the strain to be sensitive to a combination ofpenicillin with and inhibitor. Genes encoding forESBL are often located on plasmids which in a shorttime may cause a spread of resistance genes amongdifferent bacteria through conjugation and exchange

of plasmids (Neuwirth et al., 2001; Pa³ucha et al.,1999). The results concerning aminoglycosides resis-tance confirm that genes responsible for aminoglyco-side resistance are often located on the same plasmids,where genes encoding ESBL reside.

Our research has shown constant increase ESBLproducers among P. mirabilis strains, which indicatesproliferation of plasmid-coded resistance mechanism inthe given environment, a disconcerting tendency. Thephenomenon of proliferation of plasmid located resis-tance genes does not seem to be related to other gene-tic traits of the tested strains. ESBL producers belong-ing to multiresistant genotypes A and B, not to moresensitive to antibiotics genotype C. No unique patternamong ESBL producing strains has not been found.

Appearance of P. mirabilis strains with the samerestriction patterns in different wards of the same hos-pital as well as in different hospitals in the WestPomeranian area proves the spread of the strains inhospital environment. In addition to the strains classi-fied as particular genotypes, in our research 71 uniquestrains characterized by a singular restriction patternwere found. It indicates a relatively great varietyamong P. mirabilis strains present in our environment.The unique strains usually showed high sensitivity toantibiotics. They were probably �patients own strains�that caused endogenous infection. Appearance andremaining of the same clones of strains in differentwards, their transmission between wards and their ap-pearance in other hospitals may certainly cause theproblems. It is also an irrefutable proof for circula-tion of hospital strains in the environment and posesdanger to potential patients. Hospital strains are char-acterized by much higher resistance to antibiotics thanpatients� indogenous strains and frequently requiretreatment with expensive antibiotics of wide spectrumof antibacterial activity. It results in the prolongationof the time of hospitalisation and in increasing thecost of treatment. Scarce information appearing in theliterature concerning increasing number of clonalstrains among P. mirabilis is probably due to the factthat until recently these bacteria have not been conside-red to be particularly dangerous in hospital environ-ment, contrary to MRSA or Klebsiella pneumoniaestrains which are commonly known to be alert-patho-gens (Fiet et al., 2000; Traub et al., 1996). Our resultsindicate that also P. mirabilis may turn out be a danger-ous pathogen in hospital environment causing danger-ous clinical infections. This fact should make us awarethat greater attention must be paid to the situation.Periodic surveillance tests for P. mirabilis presenceshould be regularly performed as it is done with otherbacteria in a properly managed, modern hospital. Itwould certainly aid determining the frequency andtype of hospital infections in the relation to a givenward�s specific function and sanitary conditions.

173Antibiotic susceptibility of P. mirabilis hospital isolates3

Moreover, more intensive efforts of research teamsworking on hospital infections could contribute toa decrease in infection rates caused by Gram-negativerods of Proteus genus, especially hospitals isolates.

Literature

Bonnet R., H. Marchandin, C. Chanal, D. Sirot, R. Labia,C. De Champs, E. Jumas-Bilak and J. Sirot. 2002. Chromo-some-encoded class D $-lactamase OXA-23 in Proteus mirabilis.Antimicrob. Agents Chemother. 46: 2004�2006.Chippendale G.R., J.W. Warren and A.L.Trifillis. 1994. Inter-nalization of Proteus mirabilis by human renal epithelial cells.Infect. Immun. 62: 3115�3121.Clapham L., R.J.C. McLean and J.C. Nickel. 1990. The influ-ence of bacteria on struvite crystal habit and its importance in uri-nary stone formation. J. Crystal. Growth 104: 475�484.Fernandez-BacaV., F. Ballesteros, J.A. Hervas, P. Villalon,M.A. Dominguez, V.J. Benedi and S. Alberti. 2001. Molecularepidemiological typing of Enterobacter cloacae isolates froma neonatal intensive care unit: three-year prospective study. J. Hosp.Infect. 49: 173�182.Fiet J., A. Pa³ucha, B. Maczyñska, M. Stankiewicz, H. Przondo-Mordarska, W. Hryniewicz and M. Gniadkowski. 2000. A novelcomplex mutant $-lactamase, TEM-68, identified in a Klebsiellapneumoniae isolate from an outbreak of extended-spectrum $-lac-tamase-producing Klebsiellae. Antimicrob. Agents Chemother. 44:1499�1505.Goering R.V. 1993. Molecular epidemiology of nosocomial in-fection: analysis of chromosomal restriction fragment patterns bypulsed-field gel electrophoresis. Infect. Control Hosp. Epidemiol.14: 595�600.Hennekinne J.A., A. Kerouanton, A. Brisabois and M.L.De Buyser. 2003. Discrimination of Staphylococcus aureus bio-types by pulsed-field gel electrophoresis of DNA macro-restric-tion fragments. J. Appl. Microbiol. 94: 321�329.Johnson D.E., R.G. Russell, C.V. Lockatell, J.C. Zulty,J.W. Warren and H.L.T. Mobley. 1993. Contribution of Proteus

mirabilis urease to persistence, urolithiasis, and acute pyelone-phritis in a mause model of ascending urinary tract infection.Infect. Immun. 61: 2748�2754.Mammeri H., L. Gilly, G. Laurans, G. Vedel, F. Eb and G. Paul.2001. Catalytic and structural properties of IRT-21 $-lactamase(TEM-77) from a co-amoxiclav- resistant Proteus mirabilis isolate.FEMS Microbiol. Lett. 205: 185�189.Neuwirth C., S. Madec, E. Siebor, A. Pechinot, J.M. Duez,M. Pruneaux, M. Fouchereau-Peron, A. Kazmierczak andR. Labia. 2001. TEM-89 $-lactamase produced by a Proteusmirabilis clinical isolate: new complex mutant (CMT 3) withmutations in both TEM-59 (IRT-17) and TEM-3. Antimicrob.Agents Chemother. 45: 3591�3594.Pa³ucha A., B. Mikiewicz, W. Hryniewicz and M. Gniadkowski.1999. Concurrent outbreaks of extended-spectrum $-lactamase-producing organisms o the family Enterobacteriaceae in a Warsawhospital. J. Antimicrob. Chemother. 44: 489�499.Perilli M., B. Segatore, M.R. De Massis, N. Franceschini,C. Bianchi, G.M. Rossolini and G. Amicosante. 2002. Charac-terization of a new extended-spectrum $-lactamase (TEM-87) iso-lated in Proteus mirabilis during an Italian survey. Antimicrob.Agents Chemother. 46: 925�928.Saladin M., V.T.B. Cao, T. Lambert, J.L. Donay, J.L. Herr-mann, Z. Ould-Hocine, Ch. Verdet, F. Delisle, A. Philipponand G. Arlet. 2002. Diversity of CTX-M $-lactamases and theirpromoter regions from Enterobacteriaceae isolated in three Pari-sian hospitals. FEMS Microbiol. Lett. 209: 161�168.Tenover F.C., R.D. Arbeit, R.V. Goering, P.A. Mickelsen,B.E. Murray, D.H. Persing and B. Swaminathan. 1995. Inter-preting chromosomal DNA restriction patterns produced bypulsed-field gel electrophoresis: criteria for bacterial strain typing.J. Clin. Microbiol. 33: 2233�2239.Toltzis P., M.J. Dul, C. Hoyen, A. Salvator, M. Walsh, L. Zettsand H. Toltzis. 2001. Molecular epidemiology of antibiotic-resistant Gram-negative bacilli in a neonatal intensive care unitduring a nonoutbreak period. Pediatrics 108: 1143�1148.Traub W.H., B. Leonhard and D. Bauer. 1996. Gentamicin- andmethicillin-resistant Staphylococcus aureus: phenotypic and geno-typic characterization of three putative nosocomial outbreak strains.Chemotherapy 42: 21�36.


ORIGINAL PAPER

Introduction

Commensal bacteria inhabiting human and animalintestine, E. coli among others, are subjected to con-tact with various antibiotics applied at various con-centrations and with varied frequency. Antimicrobialagent resistance genes are situated in mobile geneticelements such as plasmids, transposons and integrons(Carattoli, 2001; Rowe-Magnus and Mazel, 1999).Once acquired resistance genes can be transferred be-tween bacteria. The host organism�s selective pressureselects the resistant bacteria that have specific patternsof resistance (Sayah et al., 2005). The observationsof the development of resistant bacteria have resultedin hypothesis that commensal bacteria serve as a res-ervoir of resistance genes (Wray and Gnanou, 2000).The research on the prevalence of the resistance be-tween natural E. coli in various mammals may pro-vide arguments supporting the hypothesis. The argu-ments confirming such a possibility are as follows:

1) common occurrence of E. coli as an element ofintestine microflora in mammals; 2) the occurrence,within E. coli species, of pathogens causing bothintestinal and extraintestinal diseases in humans andanimals; 3) the genetic structure of E. coli, which isof a clonal character and is composed of 4 main phy-logenetic groups A, B1, B2 and D. This is evidencethat pathogenic E. coli originate from commensalstrains revealing diversified preference to the acquisi-tion of certain virulence factors. E. coli of phylo-genetic group B2 accumulate extraintestinal virulencefactors. Enteropathogenic E. coli are assigned togroup D in prevailing number of cases. Groups A andB1 are determined as typical commensals (Duriezet al., 2001; Reid et al., 2000). Thus, the question ofthe prevalence of the resistance within the four mainphylogenetic groups is raised in the aspect of thediversified genetic structure of E. coli. It seems thatmammals, which are kept in zoological gardensand safari parks may serve as good model objects for

Prevalence of Antibiotic Resistance Profilein Relation to Phylogenetic Background among Commensal Escherichia coli

Derived from Various Mammals

KATARZYNA BALDY-CHUDZIK* and MICHA£ STOSIK

Department of Microbiology and Genetics, Institute of Biotechnology and Environmental Science,University of Zielona Góra, Zielona Góra, Poland

Received 9 January 2007, revised 26 June 2007, accepted 13 July 2007

A b s t r a c t

The paper describes the prevalence of resistant strains within the genetic structure of E. coli (phylogenetic group A, B1, B2 and D). A totalof 200 commensal E. coli strains have been derived from 10 species of healthy animals residing on ZOO Safari Park area, in �wierkocin,Poland. The phylogenetic structure of E. coli has been analysed with the use of a PCR-based method. The strains were tested in terms oftheir susceptibility to eight classes of antibiotics: aminoglycosides, penicillins, cephalosporins, tetracyclines, nitrofurans, sulphonamides,phinicols, and quinolones. The genetic structure of E. coli revealed a not uniform distribution of strains among the four phylogeneticgroups with significantly numerous representation of groups A and B1. Resistant E. coli were found within each of the phylogeneticgroups. Strains resistant to one class of antibiotics occurred significantly more frequently in phylogenetic groups B2 and D (potentialpathogens), whereas strains resistant to more than one class of antibiotics belonged to phylogenetic groups A and B1 (typical commensals)in a prevailing number of cases.

K e y w o r d s: commensal E. coli, phylogenetic groups of E. coli, resistance to antibiotics

* Corresponding author: K. Baldy-Chudzik, Department of Microbiology and Genetics, Institute of Biotechnology and Environ-mental Science, University of Zielona Góra, Monte Cassino 21B, 65-561 Zielona Góra, Poland; phone (48) 68 3287333; fax: (48)68 3287323; e-mail: K. [email protected]

176 Baldy-Chudzik K. and Stosik M. 3

such research, as they stay in a limited area for a longtime and are under continuous control.

Our research involved the analysis of commensalE. coli derived from carnivorous and herbivorousmammals staying in the grounds of ZOO Safari Parkin �wierkocin. The aim of the research was to analysethe prevalence of resistant strains within the four mainphylogenetic groups of E. coli.

Experimental


The source of strains. The material was derivedfrom adult, healthy individuals. The source organismswere five species of herbivorous animals from thetaxonomic order Artiodactyla: waterbuck (Kobusellipsiprymus), eland (Taurotragus oryx), yak (Bosmutus graniens), aurochs (Bos primigenius), buffalo(Bubalus bubalis) and five species from the taxo-nomic order Carnivora: lion (Panthera leo), lynx(Felis lynx), wildcat (Felis silvetris), racoon (Procyonlotor), dingo (Canis familiaris dingo). The researchincluded both herbivorous and carnivorous speciesbecause of the different diets and the different layoutof the walks in the Zoo�s area. The buffaloes and theyaks walks were separated with a walk common foraurochs, eland and waterbuck. The walks of the car-nivorous animals were isolated from both other car-nivorous as well as herbivorous animals.

Identification of E. coli. The samples were drawnonce. Each of the 10 tested animal species was repre-sented by samples drawn from three individuals.Thus, a total of 30 faecal samples were obtained.E. coli were isolated from each individual faecalsample. Inoculation was performed onto m-FC agarwith rosolic acid (Merck). After 24 h of incubation at44.5°C blue colonies (randomly 15 colonies fromeach sample) were passaged onto MacConkey agarplates (Difco). Lacto-positive isolates were verifiedwith a series of tests IMV and C (indole, methyl red,Voges-Proskauer, citrate) just as described earlier(Baldy-Chudzik and Stosik, 2003). All E. coli isolateswere stored at �70°C in Luria-Bertani (LB) broth con-taining 25% glycerol. For the experiments, the strainswere cultured in LB broth for 18 h at 37°C.

DNA template preparation. A single bacterialcolony was suspended in 25 µl sterile water thenheated to 99°C, 10 min, then cooled and centrifuged.

The obtained supernatant was the source of DNA tem-plate for PCR reaction.

BOX-PCR DNA fingerprinting. All the isolatesidentified in IMV and C tests as E. coli were used toDNA rep-PCR fingerprinting with BOX A1R primer(BOX-PCR). The primer sequence BOX A1R and theamplification conditions were used according to theones described earlier (Baldy-Chudzik and Stosik,2005). BOX-PCR products were analysed electro-phoretically in 0.8% (w/v) agarose in 1xTBE bufferand stained with ethidium bromide. Gels were docu-mented as TIFF files and analysed with Finger-printing II informatix software (BioRad). BOX-PCRgel lanes were normalized using 1 kb DNA Ladder(Fermentas), as external reference standards. The simi-larity matrices were calculated based on Pearson�ssimilarity coefficient with 1.5% tolerance for the po-sition of a band. Cluster analysis of similarity matrixwas performed by the unweighted pair group methodusing arithmetic averages (UPGMA). The correla-tions were expressed as the percentage of similarity.The comparative analysis of BOX-PCR patterns gene-rated by repeated analyses of strain E. coli K12 (CIP,Paris, France) (n = 50), revealed similarity >90%. Onthis basis, the similarity of BOX-PCR patterns of order85% was established as a cut-off value for determina-tion of unique strains. The isolates, which were derivedfrom hosts of the same species and which revealed thesimilarity of BOX-PCR pattern >85% were regardedto be the same and were eliminated from subsequentanalyses. Based on the similarity analysis of genomicpatterns, 200 unique E. coli strains were selected forthe subsequent studies (Fig. 1, Table I).

Determination of phylogenetic groups. Themethod used three pairs of primers of PCR reactionof sequences homologic to genetic markers specificfor phylogenetic groups of E. coli: gene chuA, geneyjaA, and an anonymous DNA sequence TspE4C2.The PCR primers, the amplification steps and theelectrophoretic analysis were all used according tothose given by Clermont et al. (2000). On the basisof a specific profile of PCR products, the determina-tion of the phylogenetic group was carried out in thefollowing way: chuA+, yjaA+, group B2; chuA+, yjaA�

, group D; chuA�, TspE4.C2+, group B1; chuA�,TspE4.C2�, group A.

Antimicrobial agents susceptibility. Antibiogramsand their interpretation were made using the disk diffu-sion method following the CLSI (formerly NCCLS)standards (National Committee for Clinical Laboratory

Fig. 1. Dendrogram of the similarity relation of BOX-PCR fingerprinting patterns of the 200 unique E. coli strains derivedfrom ten source species.

Each of the isolates is defined by: taxonomic species of the host / I, II or III refers to an individual of a given species, an Arabic numeral refersto a number of an isolate identified in an individual.

177Antibiotic resistance in commensal E. coli strains3


Standards, 2003) for 14 antimicrobial agents (Table II).These antimicrobial agents were selected on the basisof their importance in treating human or animal bac-terial infections and their use as feed additives to feedefficiency and/or disease prophylaxis in animals andon the basis of their ability to provide diversity forrepresentation of different antimicrobial agent classes.Each of E. coli isolate was grown overnight in LB

broth at 37°C and diluted with LB broth to an absor-bance at 600 nm of ~0.1. The diluted E. coli inocu-lum was swabbed onto a Mueller-Hinton agar (Difco)plate. Fourteen commercially prepared antimicrobialagent disks (Becton Dickinson) were placed on theinoculated plates. The plates were incubated at 35°Cfor 18 to 20 h. The diameters (in millimeters) of theclear zones of growth inhibition around the anti-microbial agent disks, including the 6-mm disk dia-meter, were measured by using precision callipers.The breakpoints used to categorize isolates as resis-tant or not resistant to each antimicrobial agent werethose recommended by manufacturer (BBL Sensi-Disc Antimicrobial Susceptibility Test Discs, BectonDickinson) for E. coli (Table II). Intermediate zonesof inhibition were counted as sensitive for purposesof this study. E. coli ATCC 25922 was used for qualitycontrol strain.

Statistical analysis. Chi-square was used for thecomparisons between groups (Sneath and Sokal, 1973).

Results

A total number of 200 E. coli strains were identi-fied, among which 100 were derived from carnivo-rous and 100 from herbivorous animals. The analysisof the genetic structure of E. coli isolates showedsignificant differences between strains derived fromcarnivorous and herbivorous animals (Table I). Strainsderived from carnivorous animals occurred in groupsA and D significantly more frequently than the onesderived from herbivorous animals (p<0.001, p<0.01respectively). E. coli from herbivorous species weresignificantly more frequently classified to group B1(p<0.001). The observed diversity of the geneticstructures of E. coli between carnivorous and her-

AMINOGLYCOSIDES:neomycin N 30 ≤12gentamicin GM 10 ≤12streptomycin S 10 ≤11amikacin AN 30 ≤14

PENICILLINS:ampicillin AM 25 ≤13amoxicillin/clavulanic acid AMC 20/10 ≤13

CEPHALOSPORINS:cephalothin CF 30 ≤14cefoperazone CFP 75 ≤15

TETRACYCLINES:tetracycline TE 30 ≤14doxycycline D 30 ≤12

NITROFURANS:nitrofurantoin FT 300 ≤14

SULPHONAMIDES:trimethoprim/sulfamethoxazole SXT 1.25/23.75 ≤10

PHINICOLS:chloramphenicol C 30 ≤12

QUINOLONES:nalidixic acid NA 30 ≤13

Table IIConcentrations and diffusion zone breakpoints for resis-

tance for antimicrobial agents in this study, sorted by classof antimicrobial agent

Antimicrobial agentDrugcode

Disk drugconcen-trations

(µg)

Diffusionzone

breakpoint(mm)

Carnivorous: lion 9 8 8 25 10/40.0 8/32.0 3/12.0 4/16.0

lynx 8 7 7 22 10/45.4 6/27.3 3/13.6 3/13.6

racoon 6 6 5 17 6/35.2 5/29.4 3/17.6 3/17.6

dingo 6 6 5 17 8/47.1 5/29.4 0/0 4/23.5

wildcat 7 6 6 19 6/31.6 6/31.6 4/21.0 3/15.8

Herbivorous: buffalo 6 6 5 17 5/29.4 7/41.8 3/17.6 2/11.8

waterbuck 8 8 8 24 7/29.2 9/37.5 5/20.8 3/12.5

aurochs 6 6 6 18 5/27.8 8/44.4 3/16.7 2/11.1

eland 8 8 8 24 8/33.3 10/41.7 3/12.5 3/12.5

yak 6 6 5 17 4/23.5 8/47.1 3/17.6 2/11.8

all 200 69/34.5 72/36.0 30/15.0 29/14.5

Table IGenetic structure of commensal E. coli strains derived from different animal species

Source species

Number of E. colifrom individuals:

Allnumber

n

Number/(%) of E. coli in phylogeneticgroups:

I AIIIII B1 B2 D


Aminoglycosides: neomycin 1.5 4 0 0 5.9 5.3 0 0 0 0 0

gentamicin 3 4 9.1 0 0 5.3 5.9 0 0 4.2 0

streptomycin 22.5 36 36.4 29.4 29.4 42.1 17.6 8.4 0 20.8 0

amikacin 3.5 8 9.1 0 11.8 0 0 0 0 4.2 0

Penicillins: ampicillin 13.5 20 9.1 11.8 29.4 42.1 5.9 4.2 0 12.5 0

amoxicillin/clavulanic acid

3 4 4.6 0 0 5.3 5.9 0 0 8.4 0

Cephalosporins: cephalothin 40 56 36.4 47.1 29.4 31.6 52.9 41.7 66.7 29.2 5.9

cefoperazone 22 24 22.7 41.2 29.4 26.3 11.8 20.8 27.8 12.5 5.9

Tetracyclines: tetracycline 38.5 44 50 47.2 11.8 42.1 52.9 41.7 16.8 50 17.6

doxycycline 15 20 13.6 17.6 11.8 21.1 11.8 29.2 5.6 12.5 0

Nitrofurans: nitrofurantoin 9 8 9.1 17.6 11.8 15.8 17.6 8.4 0 4.2 0

Sulphonamides: rimethoprim/sulfamethoxazole

49 20 50 58.8 47.1 47.4 52.9 50 83.3 62.5 23.5

Phenicols: chloramphenicol 3.5 20 0 0 5.9 5.3 0 0 0 0 0

Quinolones: nalidixic acid 9.5 20 0 17.6 23.5 26.3 11.8 0 0 0 0

Table IIIOccurrence of resistant E. coli in the examined animal species

The class of agent (antibiotic)

% of resistant E. coli for each animal species

Totaln = 200

Carnivorous: Herbivorous:

Lionn = 25

Lynxn = 22

Raccoonn = 17

Dingon = 17

Wildcatn = 19

Buffalon = 17

Waterbuckn = 24

Aurochsn = 18

Elandn = 24

Yakn = 17

bivorous species may be explained with different dietrequirements (resulting from their taxonomic posi-tion). Diet has been reported to be the key factor de-termining the relative abundance of E. coli phylo-genetic groups in mammals (Gordon and Crowling,2003). The revealed higher homogeneity in the ge-netic structure of E. coli in herbivorous species hasresulted from the fact that the examined species wereall ruminants rather than from the incidences of trans-mission of E. coli between them. The conclusion issupported by the results of BOX-PCR fingerprints,where the genomic similarity (>80%) proving thetransmission of strains, has been revealed in indi-vidual E. coli derived from: lynx and eland; wildcatand buffalo; yak, waterbuck, and eland as well as buf-falo and aurochs (Fig. 1).

Among the 200 examined strains, 74 were suscep-tible to all the antimicrobial agents tested. Resistancesto cephalosporins, tetracyclines, and sulphonamideswere generally most frequent (Table III). BetweenE. coli from herbivorous animals, 65% of strains fromyak were susceptible to all the antibiotics tested,whereas 38, 35, 29, and 17% of E. coli from eland,buffalo, waterbuck, and aurochs showed no resistanceat all, respectively. In carnivorous animals, 47% ofE. coli from dingo, 37% from wildcat, 36% from lionand lynx, and 35% from raccoon were susceptible toall the antibiotics tested. Strains from carnivorous ani-mals were more frequently resistant to aminoglycosides(p<0.001), penicillins (p<0.001), as well as nitro-furans (p<0.001) and quinolones (p<0.001) in com-

parison to E. coli from herbivorous animals. E. colimulti-resistant to antibiotics belonging both to thesame class as well as various classes were found incarnivorous animals more frequently than in herbivo-rous ones. One strain from wildcat presented resis-tance to 12 antimicrobial agents (N, GM, S, AM,AMC, CF, CFP, TE, FT, SXT, C, and NA). Amongstrains of all the herbivorous species, the most fre-quently identified multi-resistance pattern comprisedthe following antibiotics: CF, TE, and SXT. The sameresistance pattern was identified also in three strainsderived from lynx, what proves the transmission ofresistance factors between animals living in a neigh-borhood. Among multi-resistant E. coli, simultaneousresistance to cephalosporins (CF or/and CFP) and AM,characteristic for penicillins, was found in 26 strainsamong which 17 were derived from carnivorous.Simultaneous resistance to cephalosporin and AMwas shown in five E. coli whose multi-resistance pat-terns comprised five classes of the applied antibio-tics. Resistance to TE, D or SXT exclusively, occurredwith a comparable frequency in E. coli from herbivo-rous and carnivorous animals. The resistance patterncomprising SXT and D or SXT and TE occurred inboth herbivorous and carnivorous, whereas resistanceto SXT and CF was a characteristic feature for her-bivorous animals exclusively.

Strains from carnivorous animals differed fromE. coli of herbivorous both in the genetic structure andthe resistance patterns. The observed differences con-stituted the base for the following generalizations:


1) strains derived from carnivorous and herbivorousspecies were treated as two separate sets, withoutspecifying the source species. 2) the resistance waslater analyzed with regard to a class not to a particu-lar antibiotic since strains multi-resistant to antibio-tics of the same class occurred more frequently. Whatis more, because of a low percentage of E. coli resis-tant to chloramphenicol (phinicols), the class wasneglected in subsequent analyses. The analysis of thegenetic structure of resistant E. coli showed that a pre-vailing percentage of strains resistant to any classof the applied antibiotics belonged to phylogeneticgroups A and B1 (typical commensals), derived fromboth carnivorous and herbivorous animals (Fig. 2).Not numerous resistant E. coli from groups B2 and/orD usually co-occurred with the resistant strains fromgroups A and B1.

The susceptible strains constituted 36% of the setfrom herbivorous animals and 38% of set of E. coliderived from carnivorous animals (Fig. 3). A highpercentage of multi-resistant strains were revealedin each of the sets. Among E. coli from herbivorousanimals 24% strains revealed resistance to three

classes of antibiotics, whereas 38% strains revealedresistance to more than three classes of antibioticsin the set from carnivorous animals (from 4 up to7 classes) (Fig. 3). The genetic structure of multi-resistant strains revealed distinct features commonfor both analyzed sets of E. coli: 1) strains resistant toone class of antibiotics were represented by groupsB2 and D (potential pathogens) whereas the represen-tation of strains from group A and B1 did not occur(typical commensals); 2) strains resistant to two andthree classes of antibiotics were represented by all thefour phylogenetic groups; 3) strains resistant to morethan three classes of antibiotics in vast majoritybelonged to group A and B1 and the representationof group B2 was insignificant or did not occur, andgroup D did not occur.

Discussion

The highest levels of resistance were observed forsulfonamides, tetracyclines, and cephalosporins instrains derived from all animal species. Resistance to

Fig. 2. Genetic structure of E. coli resistant to the examined classes of antibiotics in comparisonto the genetic structure of E. coli obtained from carnivorous (A) and herbivorous (B) animals respectively.


sulphonamides and tetracyclines was shown in eachof the phylogenetic groups of E. coli, but strains re-sistant to a single antibacterial agent (sulphonamideor tetracycline) belonged exclusively to phylogeneticgroups B2 and D (potential pathogens). The preva-lence of resistance to these classes of antibiotics couldbe the result of widespread and lengthy use of theseantimicrobial agents in non-food producing animals(Klein and Bulte, 2003). Resistance to tetracycline isplasmid mediated. The large numbers of genetic de-terminants for tetracycline resistance make it morepossible for a susceptible bacterium to acquire resis-tance factors than if only a few determinants wereavailable (McEwen and Fedorka-Cray, 2002). Resis-tance to sulfonamides is plasmid mediated, but chro-mosomal mutations for sulfonamide resistance takeplace very seldom. Resistance to sulfonamides is

widespread in the environment and cross-resistancebetween sulfonamides is complete (Sköld, 2001).

A high level of resistance to cephalosporins wasobserved in both carnivorous and herbivorous animals,whereas resistance to penicillins occurred considerablymore frequently in carnivorous animals. The resis-tance to these classes of antibiotics as well as to theremaining examined classes co-occurred, formingpatterns of multi-resistance in strains from phylo-genetic groups A and B1. Cefoperazone (the thirdgeneration of cephalosporins) is used in veterinarymedicine for various species of animals. The first-generation cephalosporin�s � cephalothin are heavilyused for the treatment of bacterial infections and parti-cularly of urinary tract infections in cats and dogsbut rarely or never in ruminants (all herbivorous ani-mals in this study are ruminant) (Lanz et al., 2003;

Fig. 3. Genetic structure of susceptible and multi-resistant E. coli derived from carnivorous (A) and herbivorous (B) animals.Values above the bar correspond to the number of E. coli strains in each of phylogenetic group.


Donaldson et al., 2006). In herbivorous animals, theresistance against the third-generation of cephalo-sporins can develop resistance to cephalothin. The mostcommon mechanism of resistance to $-lactam anti-biotics (including the tested penicillins, and cephalo-sporins) is the production of various $-lactamases,which hydrolyze $-lactam ring. The E. coli resistancemay be caused by both mutations in ampC gene (con-ditioning constitutive $-lactamase) located on chro-mosomes, and a broad spectrum of $-lactamase geneslocated on plasmids. AmpC $-lactamases are not in-hibited by inhibitors such as clavulanic acid. The plas-mid resistance is the effect of a stable mutation andis easy to maintain by bacteria even at the absenceof the selective pressure of antibiotics. $-lactamaseslocated on a plasmid are sensitive to inhibitors suchas clavulanic acid (Livermore, 1995; Siu et al., 2003).The family of $-lactamases is numerous, and grow-ing. Among strains of a complex resistance patternsincluding: penicillins, penicillins with clavulanic acidand cephalosporins, a univocal identification of typesof developed resistance may be achieved by specificgene identification. It is the consequence of the factthat a simultaneous occurrence of both chromosomaland plasmid genes conditioning $-lactometers arefound more and more frequently in multi-resistantstrains (Tenover et al., 2003). The differences inresistance patterns between carnivorous and herbi-vorous animals may be caused by exposure to dif-ferent agents because of differences in the husbandryof these species or other factors that may have in-creased or decreased the likelihood of the develop-ment and conservation of resistant bacteria in the ani-mal species. For example, resistance to quinolonesoccurred considerably more frequently in carnivorousthan in herbivorous. The resistance to this class of anti-biotics is the result of chromosomal mutations, andnot the result of acquiring genes from other bacteriaof the same or other species, and the occurrence ofresistance is conditioned by the frequency of applyingthe medicine (Chen et al., 2001).

The research comprised 10 species of healthymammals with the aim to analyze the genetic struc-ture of antibiotic-resistant commensal E. coli. Suchdirect comparative analyses of E. coli in various hostspecies are rare in the literature and are usually con-cerned with clinical strains derived from a single hostspecies (Hill et al., 2003; Selander et al., 1986). Theobtained results are essential for the better recogni-tion of population biology of E. coli, because they in-dicate the fact that within a genetic structure of E. coliof various source species, phylogenetic groups A andB1, i.e. typical commensal strains, compose a basicreservoir of multi-resistant strains. They also suggestthat E. coli from groups A and B1 occupy similarniches in the organisms of the examined animals. In

such niches, gradual acquiring of resistance factorsmay result in increased surviving. It has been reportedthat multi-resistant E. coli survive better beyond thehost organism than the susceptible strains (Abu-Ghazaleh, 2001). The contamination of the habitat ofanimals with multi-resistant E. coli increases the realhazard of the prevalence of both such strains as wellas the resistance factors. The prevalence of multi-re-sistant E. coli from groups A and B1 in the environ-ment may be essential for the recently reported fail-ures in the treatment of extraintestinal infections ofE. coli in humans (particularly reoccurring infectionsof urinary tract in females) (Johnson et al., 2005;Moreno et al., 2006). The problems in the treatmentresulted from the occurrence of opportunistic, multi-resistant E. coli from groups A and B1 rather thanuropathogenic E. coli from group B2, while in mostcases uropathogenic E. coli from group B2 constituteda bacterial subpopulation sensitive to antibiotics.

Literature

Abu-Ghazaleh B. M.2001. Fecal coliforms of wastewater treat-ment plants: antibiotic resistance, survival on surfaces and inhibi-tion by sodium chloride and ascorbic acid. Microbiol. 24: 379�387.Baldy-Chudzik K. and M. Stosik. 2003. Heterogeneity of Esche-richia coli derived from artiodactyla animals analyzed with theuse of rep-PCR fingerprinting. Folia Microb. 48: 162�167.Baldy-Chudzik K.and M. Stosik. 2005. Specific genomic finger-prints of Escherichia coli strains with repetitive sequences andPCR as an effective tool for monitoring freshwater environments.Polish J. Environ. Stud. 14: 551�557.Carattoli A. 2001. Importance of integrons in the diffusion ofresistance. Vet. Res. 32: 243�259.Chen F.J., L.C. McDonald, M. Ho and H.J. Lo. 2001. Identifi-cation of reduced fluoroquinolone susceptibility in Escherichiacoli: a herald for emerging resistance. J. Antimicrob. Chemother.48: 936�938Clermont O., S. Bonacorsi and E. Bingen. 2000. Rapid andsimple determination of the Escherichia coli phylogenetic group.Appl. Environ. Microbiol. 66: 4555�4558.Donaldson S.C., B.A. Straley, N.V. Hegde, A.A. Sawant, C.DebRoy and B.M. Jayarao. 2006. Molecular epidemiology ofceftiofur-resistant Escherichia coli isolates from dairy calves.Appl. Environ. Microbiol. 72: 3940�3948.Duriez P., O. Clermont, S. Bonacorsi, E. Bingen, A. Chaventre,J. Elion, B. Picard and E. Denamur. 2001. Commensal Escheri-chia coli isolates are phylogenetically distributed among geographi-cally distinct human populations. Microbiology 147: 1671�1676.Gordon D.M. and A. Cowling. 2003. The distribution and geneticstructure of Escherichia coli in Australian vertebrates: host andgeographic effects. Microbiology 149: 3575�3586.Hill T.C.J., K.A. Walsh, J.A. Harris and B.F. Moffett. 2003.Using ecological diversity measures with bacterial communities.FEMS Microbiol. Ecol. 43:1�11.Johnson J.R., M.A. Kuskowski, T.T. O�Bryan, R. Colodnerand R. Raz. 2005. Virulence genotype and phylogenetic origin inrelation to antibiotic resistance profile among Escherichia coliurine sample isolates from Izraeli women with acute uncompli-cated cystitis. Antimicrob. Agents Chemother. 49: 26�31.Klein G. and M. Bulte. 2003. Antibiotic susceptibility patternof Escherichia coli strains with verocytotoxic E. coli associated


virulence factors from food and animal species. Food Microbiol.20: 27�33.Lanz R., P. Kuhnert and P. Boerlin. 2003. Antimicrobial resis-tance and resistance gene determinants in clinical Escherichia colifrom different animal species in Switzerland. Vet. Microbiol. 91:73�84Livermore D.M. 1995. $-Lactamases in laboratory and clinicalresistance. Clin. Microbiol. Rev. 8: 557�584.McEwan S.A. and P.J. Fedorka-Cray. 2002. Antimicrobial useand resistance in animals. Clin. Infect. Dis. 34 (Suppl 3): S93�106Moreno E., G. Prats, M. Sabaté, T. Pérez, J.R. Johnson andA. Andreu. 2006. Quinolone, fluoroquinolone and trimethoprim/sulfamethoxazole resistance in relation to virulence determinantsand phylogenetic background among uropathogenic Escherichiacoli. J. Antimicrob. Chemother. 57: 204�211.National Committee for Clinical Laboratory Standards. 2003.Approved standard M2-A8. Performance standards for antimicro-bial disk susceptibility tests, 8th. Ed. NCCLS, Wayne, PA.Reid S.D., A.C. Bumbaugh, R.K. Selander and T.S. Whittam.2000. Parallel evolution of virulence in pathogenic Escherichiacoli. Nature 406: 64�67.Rowe-Magnus D.A. and D. Mazel. 1999. Resistance gene cap-ture. Curr. Opin. Microbiol. 2: 483�488.Sayah R.S., J.B. Kaneene, Y. Johnson and R.A. Miller. 2005.Patterns of antimicrobial resistance observed in Escherichia coli

isolates obtained from domestic- and wild-animal fecal samples,human septage, and surface water. Appl. Environ. Microbiol. 71:1394�1404.Selander R.K., T.K. Korhonen, V. Vaisanen-Rhen, P.H. Wil-liams, P.E. Pattison and D.A. Caugant. 1986. Genetic relation-ships and clonal structure of strains of Escherichia coli causingneonatal septicemia and meningitis. Infect. Immun. 52: 213�222.Sköld O. 2001. Resistance to trimethoprim and sulfonamides. Vet.Res. 32: 261�273.Siu L K., P.L. Lu, J.Y. Chen, F.M. Lin and S.C. Chang. 2003.High-level expression ampC $-lactamase due to insertion ofnucleotides between �10 and �35 promoter sequences in Escheri-chia coli clinical isolates: cases not responsive to extended-spec-trum-cephalosporin treatment. Antimicrob. Agents Chemother. 47:2138�2144.Sneath P.H.A. and R.R. Sokal. 1973. Numerical Taxonomy.Freeman, San Francisco.Tenover F.C., P.M. Raney, P.P. Williams, J.K. Rasheed, J.W.Biddle, A. Oliver, S.K., Fridkin, L. Jevitt and J.E. McGowan Jr.2003. Evaluation of the NCCLS extended-spectrum $-lactamaseconfirmation methods for Escherichia coli with isolates collectedduring project ICARE. J. Clin. Microbiol. 41: 3142�3146.Wray C. and J.C. Gnanou. 2000. Antibiotic resistance monitor-ing in bacteria of animal origin: analysis of national monitoringprograms. Int. J. Antimicrob. Agents. 14: 291�294


ORIGINAL PAPER

Introduction

The female birth canal is rich in microorganisms.Infection starts when a pathogenic microorganism en-ters into the birth canal and dominates the physiologicmicroflora, breaking protective barriers and startingan inflammatory process. Vaginal environment (stabletemperature, humidity, presence of nutrients, and lowlevel of oxygen) is appropriate for the growth of micro-organisms. Development of infection depends on manyfactors, such as virulence of microbes, immune sys-tem activity, production of different factors inhibitinggrowth of microorganisms and others.

In Poland about 35% of infections of femalegenito-urinary tract are caused by yeasts (Go³¹b-Lipiñska and Kurnatowska, 2001), especially Can-dida albicans (80�95%), urogenital mycoplasmas;Ureaplasma urealyticum � 20% of cases of NGU(Denys, 2006), and Chlamydia trachomatis � 20�40%(Choroszy-Król et al., 2000; Zbroch et al., 2004).

Although group B streptococci (GBS) is a part ofthe physiologic microflora of vagina (colonization� 34%) the frequency of infection by this bacteriumincreases during inflammatory processes in thegenito-urinary tract (Dyba� et al., 2005).

Taking into account the possibility of relapses orsevere complications as infertility, and possibility oftransmission of infectious microorganisms to new-borns, it is very important to appropriate diagnose andtreat such infections. Because of recently observedincrease of microbial resistance to antibiotics, resear-chers are looking for alternative treatment methods.Many investigators worked on AMPs (antimicrobialpeptides) � small cationic peptides that have anti-microbial activity. Probably, AMPs can be used in thefuture as alternatives to antibiotics. For this reasonit is very important to study which peptides andin which concentrations are produced locally asa response to microbial inflammation. The purpose ofbasic protective mechanisms of the human body is the

Human Neutrophil Peptides in Vaginitis/Cervicitis of Different Etiology

BARBARA E. WIECHU£A1, DANIELA A. FRIEDEK1, ALICJA M. EKIEL1,MA£GORZATA K. ROMANIK1, and GAYANE MARTIROSIAN1,2*

1 Department of Medical Microbiology Medical University of Silesia, Katowice, Poland2 Department of Histology and Embryology Warsaw Medical University, Warsaw, Poland

Received 29 March 2007, revised 1 June 2007, accepted 18 June 2007

A b s t r a c t

Development of female genito-urinary infections depends on many factors, such as immune system activity, virulence of microorganismand production of factors inhibiting the growth of microorganisms. Taking into account the possibility of relapses or severe complications,it is very important to appropriately diagnose and treat infections. Because of recently observed increase of microbial resistance to antibi-otics, researchers are looking for alternatives. In our study we evaluated and compared the concentration of human neutrophil peptides(HNP 1�3) in cervico-vaginal lavages (CVL), obtained from women with vaginitis/cervicitis. Swabs from the posterior vaginal fornix andfrom the endocervical canal as well as CVL samples were obtained from 32 patients with vaginitis/cervicitis and 29 healthy women(control group). Supernatants of CVL were used for determination of concentration of HNP by ELISA. The difference between concen-trations of HNP 1�3 in studied and control groups was statistically significant (p = 0.018). The maximal concentration was determinedin patients with mixed infections (28.41 ng/ml), and Group B Streptococci, GBS, (28.06 ng/ml), the minimal concentrations in cases ofC. trachomatis (mean concentrations did not differ from those in the control group: 16.93 ng/ml and 16.39 ng/ml, respectively). Maximalcorrelation was determined for control-studied group with isolation of GBS (r = 0.79), and very high negative correlation for group ofGBS � C. trachomatis (r = �0.98).

K e y w o r d s: CVL, genito-urinary infections, HNP 1�3

1* Corresponding author: G. Martirosian, Department of Medical Microbiology Medical University of Silesia, Medyków18,40-752 Katowice, Poland; phone/fax: (48) 32 2526075; e-mail: [email protected]

186 Wiechu³a B.et al. 3

localization and elimination of pathogenic micro-organisms mainly by nonspecific mechanisms. Humanneutrophil peptides, HNPs, are "-defensins producedin female birth canal against infectious agents. HNPsare localized in the azurophilic granules of neutrophilsas the main proteins, participating in oxygen � inde-pendent phagocytosis of microorganisms (Ganz et al.,1985). "-Defensins are part of host�s natural anti-microbial immunity, responsible for the first line ofdefense against pathogenic microorganisms. Additio-nally they play an important role in acquired anti-microbial immunity through the production of spe-cific antigens and promote maturation of dendriticcells (Yang et al., 2002). In vitro HNPs demonstratea cytolytic effect against bacterial strains, yeasts andviruses. This causes increasing permeability of micro-bial membrane, pore formation and outflow of ionsand bigger molecules. HNPs can also competitivelysubstitute divalent cations, which form bridges be-tween lipopolysaccharide molecules (Hancock, 1997;Zasloff, 2002). It is possible that development of dif-ferent mechanisms of co-influence with bacterial cellwall plays an important role among the properties of"-defensins against different microorganisms. Mecha-nisms of "-defensin actions are similar, but effects aredifferent, depending on specific construction of targetcell wall (Lynn et al., 2004).

The aim of this study was to evaluate and comparethe concentration of HNP 1�3 in cervico-vaginallavages, obtained from women with vaginitis and cer-vicitis, caused by different etiological agents.

Experimental


Samples source. Samples taken from 61 non-preg-nant women aged 19�40 (mean age 28.5) attendingthe Department of Medical Microbiology at the Me-dical University of Silesia, Katowice for diagnosticpurposes were studied. The study group includes33 patients (mean age 28.6) with symptoms of vagi-nitis/cervicitis (redness of vaginal and cervical epi-thelium and/or mucopurulent endocervical dischargeand/or pain and contact bleeding) before antibiotictreatment. The control group includes 29 healthywomen (mean age 28.4). All patients gave informedconsent for this study.

Sampling procedure. Swabs from the posteriorvaginal fornix and from the endocervical canal andalso cervico-vaginal lavage samples were obtainedfrom each patient for this study (Fig. 1). Gram stainedmicroscopic slides were studied for bacterial vagi-

Fig. 1. Schematic presentation of materials and methodsused in this study

Vaginal swabs Cervical swab

Gram-staining Culturing Chlamydia direct IF Mycoplasma IST 2

Aerobic bacteria

CA CZ MC

Yeasts/fungi

Sab

Incubation at 37 ºC/24 h Incubation at 30 ºC/48 h

Agars:

CA � Columbia agar with 5 % sheep blood

CZ � Chapman agar for Staphylococcus aureu

MC � Mc Conkey agar for Enterobacteriaceae

Sab � Sabouraud chloramphenicol agar fo

yeasts/fungi

Cervico-vaginal lavage (CVL)

HNP 1-3 ELISA test kit

Isolation and identification

Agars:CA � Columbia agar with 5% sheep bloodCZ � Chapman agar for Staphylococcus aureusMC � McConkey agar for EnterobacteriaceaeSab � Sabouraud chloramphenicol agar for yeast/fungi

187Neutrophil peptides in vaginitis/cervicitis3

nosis (BV) using Amsel and Nugent criteria and pa-tients with BV were excluded from this study (Zbrochet al., 2004).

Microorganisms culturing and identification.Vaginal swabs were used for Gram staining andmicroorganisms culturing. Culturing was performedaccording to routine microbiological practice. Typingof $-hemolytic streptococci was performed by Micro-screen Strep (Microgen Bioproduct Ltd., UK). Yeastsand other fungi were identified by ID 32C test(bioMérieux, France). Identification of mycoplasmaswas performed by Mycoplasma IST 2 (bioMérieux,France), and cervical Dacron swabs were used forChlamydia Direct IF (bioMérieux, France) deter-mining antigens of Chlamydia trachomatis accordingto the manufacturer�s instructions.

HNP assay. Cervico-vaginal lavage samples wereobtained by introducing 5 ml of PBS by sterile syringefollowed by aspiration. Lavage-samples were centri-fuged for 10 min at 1000×g at 4°C. Supernatants wereused for determination of concentration of HNP 1�3by ELISA test kit (Cell Sciences, Inc., USA) accor-ding to the manufacturer�s instructions by usingspectrophotometer mQuant (Biotek Instruments Inc.,USA) at 8 450 nm.*

Results

No statistically significant difference was observedbetween mean ages of women in the studied and con-trol groups (28.6 and 28.4 respectively), also socio-economic conditions in both groups were comparable.

The studied group was divided into subgroups ac-cording to detected etiological agents: GBS (n = 5),C. albicans (n = 7), C. trachomatis (n = 6), U. urealy-ticum (n = 7) and mixed infections (n = 7). Etiologicagents of mixed infection are presented in Table I. In

control group only microorganisms of physiologicalmicroflora were detected. In the studied group themaximal concentration of HNP 1�3 was found in pa-tients with mixed infections (28.41 ng/ml) and in casesof GBS (28.06 ng/ml), but the minimal concentrationsof HNPs were determined in cases of C. trachomatis,mean concentrations did not differ from those in thecontrol group (16.93 ng/ml and 16.39 ng/ml, respec-tively) (Table II). The difference between concentra-tions of HNP 1�3 in studied and control groups wasstatistically significant (p = 0.018).

Maximal Pearson correlation index was deter-mined for control group-studied group with isolationof GBS (r = 0.79), and very high negative dependencewas determined for studied group with GBS and withC. trachomatis (r = �0.98). We demonstrated highcorrelation between number of neutrophils observedin microscopic smears and mean concentration ofHNP 1�3, especially in control group (no statisticallysignificant correlation) � Table III.

Discussion

In the beginning phase of infection neutrophilsadhere to the surface of epithelial cells and theirproteins determine the first defense against infection.

* Statistics. For statistical analysis the program Statistica 6.1(Statsoft, USA) was used. This study was approved by theBioethical Committee of Medical University of Silesia (NN-6501-113/04).

>5 n = 10 29.23 n = 13 19.98

4�5 n = 11 24.11 n = 9 16.52

0�3 n = 12 23.24 n = 7 9.55

Table III Correlation between number of neutrophils observed

in microscopic smear and mean concentrations of HNP 1�3in cervico-vaginal lavage

Neurophiles were counted at the magnification of 400 x

Numberof neutrophils

Studied group Control group

Number

Medianconcentr. ofHNP 1�3(ng/ml)

Number

Medianconcentr. of

HNP 1�3(ng/ml)

Studied group 33 25.35 ± 13.55*

GBS 5 28.06 ± 14.59

Candida albicans 7 25.96 ± 11.63

Chlamydia trachomatis 6 16.93 ± 15.12

Ureaplasma urealyticum 7 24.66 ± 16.56

Mixed infection 8 28.41 ± 10.33

Control group 29 16.39 ± 13.53*

Table II Concentrations of HNP 1�3 in cervico-vaginal lavage

of studied women

n Median ± SD (ng/ml)

* statistically significant difference (p = 0.018)

GBS + C. albicans 1 35.84

GBS + C. trachomatis+ U. urealyticum

1 31.24

C. albicans + U. urealyticum 1 28.71

C. albicans + C. trachomatis+ U. urealyticum

1 26.42

C. albicans + C. trachomatis 3 25.56

Table I Etiologic agents and concentrations of HNP 1�3 in mixed

infection subgroup

Etiologic agents n Concentrationof HNP 1�3 (ng/ml)


Infection process triggers morphological changes inepithelial cells (mainly in nucleus and cytoplasma).These changes usually depend on the type of infec-tious agent.

Our study demonstrated differences in expressionof HNPs depending on isolated etiological agent,although further genetic studies are required for con-firmation of these results. In the medical literature wedid not find any data regarding the expression of HNP1�3 in GBS infection, as well as mixed-infection. Incases of mixed infection concentrations of HNP 1�3determined in our study evidenced increasing immuneresponse, only in one case the level of HNP 1�3was low � 6.45 ng/ml (Table I). Such high immuneresponse to GBS infection may be connected with thefact, that GBS is also a part of vaginal physiologicalmicroflora. It was found that saprophytic microorgan-isms can induce defensin expression, but they alsodemonstrate relative tolerance against them. On theother hand, pathogenic microorganisms do not inducedefensin expression and also can evade innate immunemechanisms and cause disease (Yeaman and Yount,2003). Low concentration of defensins in women in-fected with C. trachomatis, observed in our study,confirms the results described by Wiesenfeld et al.(2002) and Porter et al. (2005), who described a lowerconcentration of HNP 1�3 in women with chlamy-diasis compared with women infected by Tricho-monas vaginalis or Neisseria gonorrhoeae. Studiesof Yasin et al. (1996) demonstrated minimal role of"-defensins against C. trachomatis. It is also a wellknown fact that HNP 1�3 with other antimicrobialpeptides, like LL-37, and HBD-1, acts as a synergisticbarrier and kills pathogenic microorganisms (Tollinet al., 2003). This mean that even low expression ofHNP 1�3 by epithelial cells (lower than active concen-tration) also promotes development of immune mecha-nisms. A very active response was detected duringC. albicans infection. Lehrer et al. (1988) and Rajet al. (2000) demonstrated that HNP-1 has a high anti-mycotic activity, but HNP-3 � very low. Structures ofHNP-1 and HNP-3 differ by one amino-acid in N-ter-minus of peptide, which determines defensin proper-ties (Oren and Shai, 1997). This was confirmed also infor other yeasts; Cryptococcus neoformans is inhibitedby HNP-1 and HNP-2 (Ganz et al., 1985), but intra-cellular growth of Histoplasma capsulatum in mureinof macrophages is inhibited by xenogenic expressionof HNP-1 (Couto et al., 1994; Salzman et al., 2003).

There are no reports on the effect of HNP 1�3 onurogenital mycoplasmas. In our study we demon-strated similar results for U. urealyticum as in the caseof yeasts. We didn�t observe correlation betweendefensins level and age of woman. Similar resultswere obtained by Wiesenfeld et al. (2002). They de-

monstrated lack of correlation between level of HNPand hormonal anticonception, phase of menstrualcycle and use of condoms. Increasing of NHP 1�3concentration is a response to inflammation by neutro-phil and epithelial cells. Valore et al. (2002) de-monstrated that because antimicrobial effect of vagi-nal proteins depends on concentrations, increasingof neutrophil defensisns concentration contributes tohost defense. It is also very important that thesepeptides can act synergistically with other factors,such as LL-37 (Nagaoka et al., 2000). On the otherhand many microorganisms developed mechanisms toevade bacteriocidal antimicrobial molecules (Ganz,2001). In the opinion of Lynn et al. (2004) probability,that these mechanisms stimulate adaptive evolutionof "-defensins is very high. For example, pathogensPseudomonas aeruginosa, Enterococcus faecalis andStreptococcus pyogenes produce sulfur compounds,which bind and neutralize HNP-1 (Schmidtchen et al.,2001), but protein SIC (streptococcal inhibitor of com-plement) inhibits antibacterial activity of several anti-microbial peptides, like lizozyme, SLPI, LL-37, HNP-1and $-defensins 1, 2 and 3 (Fernie-King et al., 2006).

We have shown that the ability of interactionbetween AMPs and microorganisms changes hostresponse to infection depending mainly on structureand properties of etiological agent. Further in vivo andin vitro studies of interactions of AMPs with differentmicroorganisms are required to shed light on the pos-sibility of using them against antibiotic-resistant mi-croorganisms.

AcknowledgementsThis work was supported by Grants No 2P05D 050 29 and

No 2P05D 060 27 from Ministry of Science and Higher Educa-tion in Poland.

Literature

Choroszy-Król I., J. Ruczkowska, A. Kowal and L. Pawlik.2000. Detection of Chlamydia trachomatis in urine specimensby using Ligase Chain Reaction (LCR). Adv. Clin. Exp. Med. 9:245�250.Couto M.A., L. Liu, R.I. Lehrer and T. Ganz. 1994. Inhibitionof intracellular Histoplasma capsulatum replication by murinemacrophages that produce human defensin. Infect. Immun. 62:2375�2378.Denys A. 2006. Macrolides in the treatment of genitourinary sys-tem infections (in Polish). Gin. Prakt. 3: 6�10.Dyba� I., A. Sidor-Wójtowicz and M. Kozio³-Montewka.2005. Bacterial flora and mycosis of the vagina in women withsymptoms of vaginal inflammation (in Polish). Ginekol. Pol. 76:385�390.Fernie-King B.A., D.J. Seilly and P.J. Lachmann. 2006. Inhi-bition of antimicrobial peptides by group A streptococci: SIC andDRS. Biochem. Soc. Trans. 34: 273�275.Ganz T. 2001. Fatal attraction evaded: how pathogenic bacteriaresist cationic polypeptides. J. Exp. Med. 193: F31�F34.

189Neutrophil peptides in vaginitis/cervicitis3

Ganz T., M.E. Selsted, D. Szklarek, S.S. Harwig, K. Daher,D.F. Bainton and R.I. Lehrer. 1985. Defensins. Natural peptideantibiotics of human neutrophils. J. Clin. Invest. 76: 1427�1435.Go³¹b-Lipiñska M. and A. Kurnatowska. 2001. Some aspectsof fungi multifocal invasions connected with the genital organs inwomen (in Polish). Wiad. Parazytol. 47: 137�142.Hancock R.E. 1997. Peptide antibiotics. Lancet 349: 418�422.Lehrer R.I., T. Ganz, D. Szklarek and M.E. Selsted. 1988.Modulation of the in vitro candidacidal activity of human neutro-phil defensins by target cell metabolism and divalent cations.J. Clin. Invest. 81: 1829�1835.Lynn D.J., A.T. Lloyd, M.A. Fares and C. O�Farrelly. 2004.Evidence of positively selected sites in mammalian alpha-defensins.Mol. Biol. Evol. 21: 819�827.Nagaoka I., S. Hirota, S. Yomogida, A. Ohwada and M. Hirata.2000. Synergistic actions of antibacterial neutrophil defensins andcathelicidins. Inflamm. Res. 49: 73�79.Oren Z. and Y. Shai. 1997. Selective lysis of bacteria but notmammalian cells by diastereomers of melittin: structure-functionstudy. Biochemistry 36: 1826�1835.Porter E., H. Yang, S. Yavagal, G.C. Preza, O. Murillo,H. Lima, S. Greene, L. Mahoozi, M. Klein-Patel, G. Diamondand others. 2005. Distinct defensin profiles in Neisseria go-norrhoeae and Chlamydia trachomatis urethritis reveal novel epi-thelial cell-neutrophil interactions. Infect. Immun. 73: 4823�4833.Raj P.A., K.J. Antonyraj and T. Karunakaran. 2000. Large-scale synthesis and functional elements for the antimicrobial acti-vity of defensins. Biochem. J. 347: 633�641.Salzman N.H., D. Ghosh, K.M. Huttner, Y. Paterson andC.L. Bevins. 2003. Protection against enteric salmonellosis in

transgenic mice expressing a human intestinal defensin. Nature422: 522�526.Schmidtchen A., I.M. Frick and L. Bjorck. 2001. Dermatansulphate is released by proteinases of common pathogenic bac-teria and inactivates antibacterial "-defensin. Mol. Microbiol. 39:708�713.Tollin M., P. Bergman, T. Svenberg, H. Jornvall, G.H. Gud-mundsson and B. Agerberth. 2003. Antimicrobial peptides in thefirst line defence of human colon mucosa. Peptides 24: 523�530.Valore E.V., C.H. Park, S.L. Igreti and T. Ganz. 2002. Antimi-crobial components of vaginal fluid. Am. J. Obstet. Gynecol. 187:561�568.Wiesenfeld H.C., R.P. Heine, M.A. Krohn, S.L. Hillier,A.A. Amortegui, M. Nicolazzo and R.L. Sweet. 2002. Associa-tion between elevated neutrophil defensin levels and endometritis.J. Infect. Dis. 186: 792�797.Yang D., A. Biragyn, L.W. Kwak and J.J. Oppenheim. 2002.Mammalian defensins in immunity: more than just microbicidal.Trends Immunol. 23: 291�296.Yasin B., S.S. Harwig, R.I. Lehrer and E.A. Wagar. 1996. Sus-ceptibility of Chlamydia trachomatis to protegrins and defensins.Infect. Immun. 64: 709�713.Yeaman M.R. and N.Y. Yount. 2003. Mechanisms of antimicro-bial peptide action and resistance. Pharmacol. Rev. 55: 27�55.Zasloff M. 2002. Antimicrobial peptides of multicellular orga-nisms. Nature 415: 389�95.Zbroch T., P. Knapp, E. B³oñska, M. Kobylec and P. Knapp.2004. Life style, Chlamydia trachomatis infection, bacterialvaginosis and their impact on the frequency of cervical lesions (inPolish). Ginekol. Pol. 75: 538�544.


ORIGINAL PAPER

Introduction

Actinomycetes are Gram-positive bacteria that arewide spread in nature and play a significant role inthe production of bioactive metabolites mainly anti-microbial compounds (Sanglier et al., 1993). At least90% actinomycetes isolated from soil have beenreported to be Streptomyces spp. (Anderson andWellington, 2001). Most of these are potentially usefulas pharmacologically and agriculturally active agents(Berdy, 2005). Pathogens found to be drug resistantrevealed the importance of novel bioactive com-pounds (Bonjan et al., 2004). Microbial metabolitescause increasing attention as potential plant protec-tion agents because they are expected to overcome thepollution problems caused by the synthetic chemicalpesticides. Several novel metabolites of actinomyceteswere widely useful for the control of plant diseases,insects and weeds (Li et al., 2003). Phenylpropionicacid is a member of the phenylpropanoid family, com-

prising a wide variety of C6-C3 compounds synthe-sized by plants from phenylalanine and important inplant physiology and defense mechanisms for the syn-thesis of flavonoids, insect repellents, UV protectantsand signal molecules (Hahlbrock and Scheel, 1989).Phenylpropionic acid is rarely encountered as micro-bial metabolite. Cremin et al. (1994) reported that3-phenylpropionic acid (3-PPA) is found in ruminalfluid as product of chemical reduction of dietary phe-nolic monomers by ruminal microorganisms. Duringthe screening of actinomycetes for bioactive com-pounds, an actinomycete strain was found to be pre-dominant in the random sampling of laterite soilspresent in different locations of Acharya NagarjunaUniversity (ANU) campus. The isolate was identifiedas Streptomyces and designated as strain ANU 6277.The strain was deposited at Microbial Type CultureCollection Centre (MTCC), IMTECH, Chandigarh(India) with accession number 6277. Very little isknown about the biological activity of 3-PPA and no

Biological Activity of Phenylpropionic Acid Isolatedfrom a Terrestrial Streptomycetes

KOLLA J.P. NARAYANA1, PEDDIKOTLA PRABHAKAR2, MUVVA VIJAYALAKSHMI1*,YENAMANDRA VENKATESWARLU2 and PALAKODETY S.J. KRISHNA3

1 Department of Microbiology, Acharya Nagarjuna University, Guntur, India2 Organic Chemistry Division-I, Indian Institute of Chemical Technology

3 Biotechnology Unit, Institute of Public Enterprise, Hyderabad, India

Received 28 May 2007, revised 1 August 2007, accepted 7 August 2007

A b s t r a c t

The strain ANU 6277 was isolated from laterite soil and identified as Streptomyces sp. closely related to Streptomyces albidoflavus clusterby 16S rRNA analysis. The cultural, morphological and physiological characters of the strain were recorded. The strain exhibited resist-ance to chloramphenicol, penicillin and streptomycin. It had the ability to produce enzymes such as amylase and chitinase. A bioactivecompound was isolated from the strain at stationary phase of culture and identified as 3-phenylpropionic acid (3-PPA) by FT-IR, EI-MS,1H NMR and 13C NMR spectral studies. It exhibited antimicrobial activity against different bacteria like Bacillus cereus, B. subtilis,Escherichia coli, Klebsiella pneumoniae, Proteus vulgaris, Pseudomonas aeruginosa, P. flourescens, Staphylococcus aureus and somefungi including Aspergillus flavus, A. niger, Candida albicans, Fusarium oxysporum, F. udum and Penicillium citrinum. The antifungalactivity of 3-PPA of the strain was evaluated in in vivo and in vitro conditions against Fusarium udum causing wilt disease in pigeon pea.The compound 3-PPA is an effective antifungal agent when compared to tricyclozole (fungicide) to control wilt caused by F. udum, but itexhibited less antifungal activity than carbendazim.

K e y w o r d s: Streptomyces strain ANU 6277, taxonomic studies, 3-phenylpropionic acid, biological activity of 3-PPA

* Corresponding author: M. Vijayalakshmi, Department of Microbiology, Acharya Nagarjuna University, Guntur-522 510, A.P.,India; e-mail: [email protected]

192 Narayana K.J.P. et al. 3

reports were found on its production by actinomycetes.Hence an attempt was made to evaluate biologicalproperties of the bioactive compound from the strain.The present paper define the taxonomy position of thenew isolated strain, presents the procedure of theextraction and elucidation of the biological activity ofa compound obtained from the strain.

Experimental


Strain used. Streptomyces strain ANU 6277 wasisolated from laterite soil sample collected at AcharyaNagarjuna University (ANU) campus by dilution platetechnique using asparagine-glycerol-salts agar mediumsupplemented with streptomycin sulphate (10 µg/ml)and amphotericin-B (50 µg/ml). The strain was main-tained on yeast extract-malt extract-dextrose (YMD)agar medium at 4°C (Williams and Cross, 1971).

Phenotypic studies. Morphological, cultural andphysiological characteristics of the strain were per-formed according to the methods described byShirling and Gottlieb, (1966). The strain was culti-vated on different media including those recommendedby International Streptomyces Project (ISP) mediaand non-ISP media. The media such as tryptone-yeastextract agar (ISP-1), YMD agar (ISP-2), oat meal agar(ISP-3), starch-casein-salts agar (ISP-4), glycerol-as-paragine-salts agar (ISP-5), peptone-yeast extract-iron agar (ISP-6), tyrosine agar (ISP-7), nutrient agarand Czapek-Dox agar media were employed for thestudy of growth characteristics of the strain (Dietz andThayer, 1980). Utilization of different carbon sourceswas studied in minimal medium containing those at1% concentration. The strain grown at 37°C for5 days on ISP medium 2 was used to study the micro-morphology with scanning electron microscope(SEM) (Yassin et al., 1997). The culture was fixedwith glutaraldehyde, and dehydrated with ethanol.The dehydrated samples were dried, mounted onaluminium stubs, and sputter coated with gold-palla-dium. Finally, they were observed with digital SEM(model JEOL JSM-5600).

Phylogenetic analysis. The chromosomal DNAof the strain ANU 6277 was isolated according tothe procedure described by Rainey et al. (1996).The 16S rRNA gene was amplified with primers8-27f (5�-AGAGTTTGATCCTGGCTCAG-3�) and1500r (5�AGAAAGGAGGTGATCCAGGC-3�). Theamplified DNA fragment was separated on 1% agarosegel, eluted from the gel and purified using Qiaquickgel extraction kit (Qiagen, Germany). The purifiedPCR product was sequenced with four forward andthree reverse primers namely 8-27f (5�AGAGTTT

GATCCTGGCTCAG-3�), 357f (5�-CTCCTACGGGAGGCAGCAG-�), 704f (5�-TAGCGGTGAAATGCGTAGA-3�), 1114f (5�-GCAACGAGCGCAACC-3�),685r (5�-TCTACGCATTTCACCGCTAC-3�), 1110r(5�-GGGTTGCGCTCGTTG-3�) and 1500r (5�-GAAAGGAGGTGATCCAGGC-3�), respectively (Esche-richia coli numbering system). The rDNA sequencewas determined by the dideoxy chain-terminationmethod using the Big-Dye terminator kit using ABI310 Genetic Analyzer (Applied Biosystems, USA).

The 16S rDNA sequence of the strain ANU 6277generated in this reaction (1478 bases) was alignedwith the 16S rDNA sequence of other closely relatedStreptomyces species retrieved from the GenBankdata base. A sequence similarity search was doneusing GenBank BLASTN (Altschul et al., 1997).Sequences of closely related taxa were retrieved,aligned using Cluster X programme (Thompson et al.,1997) and the alignment was manually corrected. Forthe neighbour-joining analysis (Saitou and Nei, 1987),the distances between the sequences were calculatedusing Kimura�s two-parameter model (Kimura, 1980).Bootstrap analysis was performed to assess the confi-dence limits of the branching (Felsenstein, 1985).

Cultivation of the strain for secondary metabo-lites production. Actively growing pure culture of thestrain was inoculated into 250 ml Erlenmeyer flasks,each containing 50 ml of seed medium consisting of0.4% dextrose, 0.4% yeast extract, 1% malt extractand 0.2% calcium carbonate (pH 7.2). The culturewas incubated on a rotary shaker (250 rpm) at 28°Cfor two days. The seed culture (10%) of the strainwas transferred into a culture medium (4% dextrose,0.9% proteose peptone, 0.1% yeast extract, 0.6% cal-cium carbonate, 0.1% K2HPO4, 0.1% MgSO4×7H2O,0.01% MnSO4×4H2O, 0.005% FeSO4×7H2O (pH 7.2)and incubated at 28°C for 5 days.

Extraction, purification and identification ofactive metabolite. The culture filtrate was collectedat the end of five day incubation period and extractedtwice with equal volume of ethyl acetate. The solventextract was evaporated in vacuo to dryness. The darkbrown residue was obtained and partially purified onsilica gel column chromatography (22×5 cm, Silicagel 60, Merck) and eluted with gradient solvent sys-tem consisting of ethylacetate: hexane. Active frac-tion was collected and concentrated. Further purifi-cation was carried out in HPLC preparative column(10 mm ×250 mm, 5 µ using hexane: 2-propanol(8:2 v/v). Structure elucidation of pure bioactive com-pound from the strain was carried out by FT-IR,EI-MS, 1H NMR and 13C NMR spectral studies.

Biological activity testing. Minimum inhibitoryconcentrations (MIC) of 3-PPA obtained from thestrain against different microorganisms including bac-teria and fungi were determined by conventional agar

193Phenylpropionic acid from terrestrial Streptomycetes3

dilution method (Cappuccino and Sherman, 1999)using nutrient agar for bacteria and Sabouroud�s agarmedium for fungi. Different concentrations of 3-PPA(0 to 1000 µg/ml) were prepared in dimethyl sulpho-xide (DMSO) and assayed against tested organisms.The organisms used in this assay are Bacillus cereusMTCC 430, B. subtilis MTCC 441, Escherichia coliMTCC 40, Klebsiella pneumonia MTCC 109, Pro-teus vulgaris MTCC 742, Pseudomonas aeruginosaMTCC 424, P. fluorescens MTCC 103, Staphylococ-cus aureus MTCC 96, Aspergillus flavus, A. niger,Candida albicans MTCC 183, Fusarium oxysporum,F. udum MTCC 2204 and Penicillium citrinum. Theantimicrobial activity was observed after 24�48 h in-cubation at 37°C for bacteria and 48�72 h incubationat 28°C for fungi. Each experiment was performedin triplicates and proper controls were done. Lowestconcentration of compound that showed antimicrobialactivity against test organisms was recorded as MICvalue (Hwang et al., 2001).

The cytotoxic activity of 3-PPA was tested onU-937 (Human leukemic monocyte lymphoma cellline) cells using MTT assay (Plumb et al., 1989).U-937 cells were obtained from National Centre forCell Science, Pune (India) and were cultured at 37°Cwith 5% CO2 using RPMI-1640 (Himedia®, India)media containing fetal bovine serum. U-937 (2×104

cells per well) were seeded in a 96-well plate contain-ing 100 µl of RPMI medium and incubated for 24 h.The cells were then treated with different concentra-tion of 3-PPA (0�150 µg/ml). After 48 h incubation,100 µl of MTT (3-(4,5-dimethylthiazol-2-yl)-2,5,-diphenyltetrazolium bromide) reagent (Sigma Chemi-cals, USA) was added to each well, and the plates wereincubated in a CO2 incubator at 37°C for 4 h. Thereaf-ter, the supernatant was removed from each well. Then100 µl DMSO was added to dissolve the coloredformazan crystals produced by the MTT. Subsequently,the optical density was measured at 570 nm using anELISA reader (Molecular Devices Corp., USA).

In vitro and in vivo antifungal activity of 3-PPAfrom the strain was studied against Fusarium udumMTCC 2204, the causal agent of Fusarium wilt inCajanus cajan L. The antifungal efficacy of 3-PPAwas also compared with the activity of commercialfungicides such as carbendazim and tricyclozole.Conidial suspension of F. udum was prepared usingthe culture grown on potato dextrose agar for 10 daysat 30°C (Hwang et al., 2001). The conidial suspen-sions were mixed with 3-PPA, carbendazim andtricyclozole to give the concentration of 0, 1, 10, 50,100, 500 and 1000 µg/ml.. After incubation for 4 h at28°C, conidial germination was microscopically exa-mined in three replicates.

In vivo antifungal activity of 3-PPA was evaluatedfor its ability to suppress Fusarium wilt on red gram

plants in a growth chamber. Antifungal substancesincluding 3-PPA, carbendazim and tricyclozole dis-solved in water + methanol (95:5) were diluted togive different concentration of 0, 10, 100, 500 and1000 µg/ml. Seeds of red gram (Cajanus cajan L.)were sown in glass beaker (18�14 cm) containingsteam sterilized soil drenched with antifungal solu-tion (30 ml). The soil was drenched with conidialsuspension (105 spores/ml) when the seedlings werethree day old (Hwang et al., 2001). Disease severityon plants was rated 15 days after inoculation basedon a scale from 0 to 5 as follows: 0 for no visibledisease symptoms, 1 for slightly wilted leaves, 2 for30 to 50% of the entire plant diseased, 3 for 50 to70% of the entire plant diseased, 4 for 70 to 90%of the entire plant diseased and 5 for a dead plant.Data are the mean of 10 plants per treatment and re-sult of two trials.


Cultural and physiological characteristics of thestrain are presented in Table I. The strain showedgood growth on ISP-1, ISP-2, ISP-4 and ISP-5 media.Moderate growth was observed on ISP-3, ISP-6, ISP-7and nutrient agar media. Pigment production by thestrain varied with the culture media employed. Darkbrown pigment was produced by the strain when grownon ISP-1,2 and 3, while yellowish brown to yellowpigments were found with ISP-4 and 5 and nutrientagar media. Diffused melanoid pigments were ob-served when grown on ISP-6 and ISP-7. Micromor-phology of the strain was examined by SEM (Fig. 1).The culture showed extensively branched aerial myce-lium and bear short chains of spores. As the sporo-genous hyphae (sporophores) were straight to flexuousin nature bearing the spores with smooth surface, thestrain may be placed in the rectus-flexibilis group ofStreptomyces species (Pridham et al., 1958).

Fig. 1. Scanning electron microscopic photographof strain ANU 6277 (magnification x 10,000, Bar 1 µm ----)


The strain had the ability to hydrolyze casein, es-culin, gelatin, starch, tyrosine and xanthine, but nothypoxanthine. The culture could tolerate NaCl levelsup to 7%. It produced enzymes such as amylase andchitinase. It exhibited resistance to chloramphenicol,penicillin-G and streptomycin and showed sensitivityto ampicillin, rifampicin and tetracycline. The utiliza-tion of various carbon sources by the strain indicatedits wide pattern of carbon assimilation potential.D-arabinose, D-fructose, D-glucose, D-galactose, gly-cerol, lactose, maltose, mannitol, mannose, raffinose,rhamnose, trehalose and xylose supported growth ofthe strain, whereas cellulose, dextrin and sucrose didnot support its growth.

Phylogenetic study of the strain was performed by16S rRNA analysis. An almost complete 16S rDNAgene nucleotide sequence (1478 bp) of the strain wasidentified by BLASTN programme and submittedto Genbank with accession number EF 142856. Thestrain showed high homology (96% identity) withStreptomyces albidoflavus. The evolutionary distancewas calculated by the Kimura�s two parameter modeland a phylogenetic tree was constructed using neigh-bour-joining method (Fig. 2). Based on observationssuch as white spore mass, yellow to brown pigmentproduction, rectus-flexibilis sporophore, spore withsmooth surface, antibiosis against fungi, resistance topenicillin-G and 7% NaCl, hydrolysis of starch and

xanthine and absence of growth at 45°C, the strainseemed to closely resemble Streptomyces albido-flavus. Theses findings are in confirmity with reportsof Williams et al. (1989), Gurtler et al. (1994) andAugustine et al. (2004).

The structure of white crystalline compound ob-tained from the crude extract after purification waselucidated by FT-IR, EI-MS, 1H NMR and 13C NMRstudies. In the FT-IR spectrum, Vmax was obtained at697.96, 931.96/cm (aromatic, C-H), 1218.19/cm (aro-matic, C = C), 1301.93/cm (C-O), 1699.90 (C = O),2928.47, 3030.43/cm (CH3-C-H) and 3390.13 (OH-group broad peak). The compound gave molecularions in positive mode at m/z are 150(50), 104(95),91(100), 78(40) and 51(35) suggested a molecularweight of 150 from EI-MS analysis. NMR data indi-cated a hydrogen count of 10 and a carbon count of9 in CD3OD at 300MHZ. 1H NMR showed protonsat 2.70* (t, 2H), 2.90* (t, 2H), 7.10 to 7.20* (dd,aromatic-protons) and 11.0* (broad, s, O-H,). The13C NMR spectrum of bioactive compound exhibitedpeaks at 30.0 (s, C-2), 36.0 (s, C-3), 126.0,128,129and 140 (aromatic carbons) and 180.0 (s, C-1). Basedon above data, the bioactive compound was charac-terized as 3-phenylpropionic acid with molecular for-mula C9H10O2 (Fig. 3).

The bioactive compound, 3-phenylpropionic acid(3-PPA) from strain showed antimicrobial activity

Hydrolysis ofcasein +esculin +gelatin +hypoxanthine �starch +tyrosine +xanthine +

Tolerance tolysozyme (0.05%) +NaCl (7%) +phenol (0.1%) �

Growth at 45°C �

Production ofmelanoid pigments +H

2S +

amylase +chitinase +

Resistance to antibiotics (µg/disc)ampicillin (50) �chlorampenicol (50) +penicillin- G (50) +rifampicin (50) �streptomycin (100) +tetracycline (100) �

Table ICharacteristics of strain ANU6277

Pigment production inISP-1 +, DBISP-2 +, DBISP-3 +, DBISP-4 +, YBISP-5 +, YBISP-6 +, MISP-7 +, MNutrient agar medium +, YCzapek-Dox �

Utilization of carbon sources (1%)D-arabinose +cellulose �D-glucose +D-fructose +dextrin �D-galactose +glycerol +lactose +mannitol +D-mannose +raffinose +rhamnose +sucrose �trehalose +xylose +

+, Positive result; �, Negative result; DB, Dark Brown; YB, Yellowish Brown; M, Melanin; Y, Yellow


aeruginosa and P. fluorescens are highly susceptibleto 3-PPA followed by B. subtilis, Escherichia coli andProteus vulgaris. Among fungi, F. udum exhibitedhigh sensitivity followed by Aspergillus flavus, Peni-cillium citrinum and A. niger. The bioactive com-pound (3-PPA) from S. albidoflavus strain ANU 6277did not exhibit significant cytotoxicity on U-937 cellsup to the concentration of 100 µg/ml..

The compound3-PPA showed inhibitory activity (IC50) on U-937 cellgrowth at 128.20 µg/ml, while the widely used anti-cancer drug, Etoposide (positive control) exhibitedcytotoxicity activity (IC50) on U-937 cells at 10.26µg/ml (Table III).

Fig. 2. Neighbour-joining tree based on 16S rDNA sequences showing the phylogenetic relationship between strainANU 6277 and other closely related species of the genus Streptomyces, Bootstrap values

(expressed as percentage of 1000 replications) greater than 50% are given at the nodes. The scale indicates 1% sequence variation.

Fig. 3. Molecular structure of 3-phenylpropionic acid

against different test microorganisms including bac-teria and fungi. The minimum inhibitory concentration(MIC) of 3-PPA ranged between 10 and 100 µg/ml(Table II). Among the test bacteria, Pseudomonas

3-PPA 128.20

* Etoposide 10.26

Table IIICytotoxic activity of 3-phenylpropionic acid on U-937

growth in vitro

* Positive control

Compound IC50

(µg/ml)

Table IIActivity of 3-phenylpropionic acid from strain ANU 6277

against different test microorganisms

Bacteria: Bacillus cereus 75

B. subtilis 50

Escherichia coli 50

Klebsiella pneumoniae 100

Proteus vulgaris 50

Pseudomonas aeruginosa 10

P. fluorescens 10

Staphylococcus aureus 100

Fungi: Aspergillus flavus 25

A. niger 50

Fusarium oxysporum 50

F. udum 10

Penicillium citrinum 25

Candida albicans 100

Tested microorganism MIC (µg/ml)

In in vitro conditions, conidial germination ofF. udum was totally inhibited with carbendazim at50 µg/ml, with 3-PPA at 100 µg/ml and tricyclozoleat 500 µg/ml (Fig. 4). In vivo efficacy of 3-PPA, carben-dazim and tricyclozole for the control of Fusariumwilt was evaluated (Fig. 5). The symptoms of wilt be-gan to appear on red gram plants one week after ino-culation. Initial symptoms of the disease consist of


wilting of individual branches. The foliage symptomsare characterized by drooping of the leaves followedby upland curling. Treatment with the antifungal sub-stances, 3-PPA and carbendazim greatly inhibitedthe wilt in red gram plants. The suppressive effect of3-PPA against Fusarium was observed at 500 µg/ml.In contrast, the commercial fungicide carbendazimcompletely inhibited the development of Fusariumwilt at the concentration of 100 µg/ml, while the treat-ment with tricyclozole showed maximum antifungalactivity at 1000 µg/ml. The efficacy of 3-PPA againstFusarium wilt was better than tricyclozole but lesseffective than carbendazim.

The bioactive compound, 3-PPA from the strainexhibited less cytotoxicity, while carbendazim thesynthetic fungicide showed high toxicity to humans,animals and plants (Mantovani et al., 1998). Soilbacteria such as Achromobacter, Nocardia and so-me Pseudomonas species could degrade 3-PPA (Fuand Oriel, 1999) indicating its suscepibility to themicrobial degradation in soil what results in the lackits accumulatin in nature like synthetic fungicides.Hence 3-PPA can be preferred over carbendazim tocontrol Fusarium wilt as an eco-friendly compound.

Among Streptomyces spp., S. albidoflavus is oneof the potential species that elaborate number ofindustrially and agriculturally important metabolites.Enzymes like chitinase and serine proteinases arereported from S. albidoflavus (Broadway et al., 1995;Bressollier et al., 1999). An odoriferous actinomycete,S. albidoflavus strain DSM 5415 was reported to pro-duce a new sesquiterpene, albaflavenone (Gurtler andPedersen, 1994). Antimicrobial properties of a non-polyene antibiotic (poly-hydroxy-poly ether com-pound) have been reported from S. albidoflavus strainPU 23 (Augustine et al., 2005). The bioactive com-pound dibutyl phthalate was reported from S. albido-flavus strain 321.2 (Roy et al., 2006). In the presentstudy, the strain ANU 6277 was found to elaboratea bioactive compound, 3-phenylpropionic acid (3-PPA).This is the first report of 3-PPA from actinomycetesespecially Streptomyces spp.

Phenyl acids like phenylacetic acid from Strepto-myces humidus are known to possess antimicrobialactivity against several bacteria and fungi (Hwanget al., 2001). Phenylpropionic acid was reported to bedetected in culture filtrates from media after inocula-tion with isolated rumen bacteria or rumen fluid inthe absence of added phenolic acids (Chesson et al.,1982). Clostridium bifermentans strain TYR-6 reportedfrom oil mill waste waters could convert cinnamicacid to 3-phenylpropionic acid (Chamkha et al., 2001).Phenylpropionic acid derivatives are pharmaceuti-cally important agents. Anti-inflammatory and anal-gesic drugs like isoprofen, ketoprofen, naproxenetc. are phenylpropionic acid based drugs (Saishoand Ishibashi, 1998). Nagano et al. (2001) reportedpyloricidin, a novel anti-Helicobacter pylori anti-biotic produced by Bacillus sp. Phenylpropionic acidmoiety of pyloricidin is essential for anti-H. pyloriactivity. The present paper investigated the extraction,physico-chemical properties and biological activitiesof 3-PPA from Streptomyces strain ANU 6277. Thebioactive compound, 3-PPA from strain ANU 6277 isa promising compound as it exhibited antimicrobialactivity against gram-positive as well as gram-nega-tive bacteria and fungi. It can also be useful as bio-control agent against Fusarium wilt.

Acknowledgement

The authors KJPN and PP are thankful to Andhra Pradesh-Netherlands Biotechnology Programme (A.P.N.L.B.P.), Hydera-bad, India and Department of Biotechnology, New Delhi, Indiafor the financial assistance.

Literature

Altschul SF., T.L. Madden, A.A. Schaffer, J.Zhang, Z. Zhang,W. Miller and D.J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.Nucleic Acids Res. 25: 3389�3402.

Fig. 4. In vitro antifungal activity of 3-PPA from strainANU 6277 against F. udum

Fig. 5. In vivo antifungal activity of 3-PPA from strainANU 6277 against F. udum causing wilt in pigeon pea plants


Anderson A.S. and E.M.H. Wellington. 2001. The taxonomyof Streptomyces and related genera. Int. J. Syst. Evol. Microbiol.51: 797�814.Augustine S.K., S.P. Bhavsar, M. Baserisalehi and B.P. Kapad-nis. 2004. Isolation, characterization and optimization of anti-fungal activity of an actinomycete of soil origin. Indian J. Exp.Biol. 42: 928�932.Augustine S.K., S.P. Bhavsar and B.P. Kapadnis. 2005. A non-polyene antifungal antibiotic from Streptomyces albidoflavusPU 23. J. Biosci. 30: 201�211.Berdy J. 2005. Bioactive microbial metabolites. J. Antibiot. 58:1�26.Bonjan G.H.S., M.H. Fooladi, M.J. Mahdavi and A. Shahghasi.2004. Broadspectrim, a novel antibacterial from Streptomyces sp.Biotechnol. 3: 126�130.Bressollier P., F. Letourneau, M. Urdaci and B. Verneuil. 1999.Purification and characterization of a keratonolytic serine protein-ase from Streptomyces albidoflavus. Appl. Environ. Microbiol. 65:2570�2576.Broadway R.M., D.L. Williams, W.C. Kain, G.E. Harman,M. Lorito and D.P. Labeda. 1995. Partial characterization ofchtinolytic enzymes from Streptomyces albidoflavus. Lett. Appl.Microbiol. 20: 271�276.Cappuccino J.G. and N. Sherman. 1999. Microbiology; a Labo-ratory Manual (Harlow: Benjamin) pp. 263�264.Chamkha M., B.K.C. Patel, G.J. Louis and M. Labat. 2001.Isolation of Clostridium bifermentans from oil mill waste watersconverting cinnamic acid to 3-phenylpropionic acid and emenda-tion of the species. Anaerobe 7: 189�197.Chesson A., C.S. Stewart and R.J. Wallace. 1982. Influence ofplant phenolic acids on growth and cellulolytic activity of rumenbacteria. Appl. Environ. Microbiol. 44: 597.Cremin J.D., J.R. James, K. Drackley, E. Grum, L.R. Hansenand G.C. Fahey. 1994. Effect of reduced phenolic acids on meta-bolism of propionate and plamitate in bovine liver tissue in vitro.J. Dairy Sci. 77: 3608�3617Dietz A. and D.W. Thayer. 1980. Actinomycete Taxonomy. So-ciety for Industrial Microbiology Special Publication, number 6,pp. 26�31.Felsenstein J. 1985. Confidence limits on phylogenies: an approachusing the boot strap. Evolution 39: 783�789.Fu W. and P. Oriel. 1999. Degradation of 3-phenylpropionic acidby Haloferax sp. D1227. Extremophiles 3: 45�53.Gurtler H. and R. Pedersen. 1994. Albaflavenone, a sesqui ter-pene ketone with a zizaene skeleton produced by a streptomycetewith a new rope morphology. J. Antibiot. 47: 434�439.Hahlbrock K. and D. Scheel. 1989. Physiology and molecularbiology of phenyl ropanoid metabolism. Annu. Rev. Plant Physiol.Plant Mol Biol. 40: 347�369.Hwang B.K., S.W. Lim, B.S. Kim, J.Y. Lee and S.S. Moon.2001. Isolation and in vivo and in vitro antifungal activity of phenylacetic acid and sodium phenyl acetate from Streptomyces humidus.Appl. Environ. Microbiol. 67: 3739�3745.Kimura M. 1980. A simple method for estimation of evolutionaryrate of base substitutions through comparative studies of nucleo-tide sequences. J. Mol. Evol. 16: 111�120.

Li Y., Z. Sun, X. Zhuang, L. Xu, S. Chen and M. Li. 2003.Research progress on microbial herbicides. Crop Protection 22:247�252.Mantovani A., F. Maranghi, C. Ricciardi, C. Marci, A.V. Stazi,L. Attias and G.A. Zapponi. 1998. Developmental toxicity ofcarbendazim: comparison of no observed-adversed affect leveland benchmark dose approach. Food Chem. Toxicol. 36: 37�45Nagano Y., K. Ikedo, A. Fujishima, M. Izawa, S. Tsuotani,O. Nishimura and M. Fugino. 2001. Pyloricidins, novel anti-Helicobacter pylori antibiotics produced by Bacillus sp. II. Isola-tion and structure elucidation. J. Antibiot. 54: 934�947.Pridham T.G., C.W. Hesseltine and R.G. Benedict. 1958.A guide for the classification of streptomycetes according toselected groups placement of strain in morphological sections.Appl. Microbiol. 6: 52�79.Plumb J.A., R. Milroy and S.B. Kaye. 1989. Effects of the pHdependence of 3-(4,5-dimethylthiazo-2-yl)-2,5-diphenyl-tetrazo-liumbromide-formazan absorption on chemosensitivity determinedby a novel tetrazolium-based assay. Cancer Res., 49: 4435�4440.Rainey F.A., N. Ward-Rainey, R.M. Kroppenstedt andE. Stackebrandt. 1996. The genus Nocardiopsis representsa phylogenetically coherent taxon and a distict actinomycete line-age: proposal of Nocardiopsiaceae fam.nov. Int. J. Syst. Bacteriol.46: 1088�1092.Roy R.N., S. Laskar and S.K. Sen. 2006. Dibutyl phthalate, thebioactive compound produced by Streptomyces albidoflavus321.2. Microbiol. Res. 161: 121�126.Saitou N. and M. Nei. 1987. The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Molecular Biol.Evol. 4: 406�425Saisho K. and N. Ishibashi. 1998. Rapid analysis of drugs,phenylpropionic acid-based anti-inflammatory and analgesic drugs.Gekkan Yakugi 40: 2883�2887.Sanglier J.J., E.M.H. Wellington, V. Behal, H.P. Feidler,R. Ellouz Ghorbel, C. Finance, M. Hacene, A. Kamoun, C.Kelly, D.K. Mercer, S. Prinzis and C. Tringo. 1993. Novel bioac-tive compounds from actinomycetes. Res. Microbiol. 144: 661�663.Shirling E.B. and D. Gottlieb. 1966. Method for characteriza-tion of Streptomyces species. Int. J. Syst. Bacteriol. 16: 313�340.Thompson J.D., T.J.Gibson, F. Plewniak, F. Jeanmougin andD.G Higgins. 1997. The Clustal X windows interface: flexiblestrategies for multiple sequence alignment aided by quality ana-lysis tools. Nucleic Acids Res. 24: 4876�4882.Williams S.T. and T. Cross. 1971. Isolation, purification, culti-vation and preservation of actinomycetes. Methods Microbiol. 4:295�334.Williams S.T., M. Goodfellow and X. Alderson. 1989. GenusStreptomycetes. In: Williams, S.T., Sharp, M.E., Holt, J.G. (eds)Bergey�s Manual of Systematic Bacteriology. The Williams andWilkins co., Baltimore, Vol.4. pp. 2452�2492Yassin A.F., F.A. Rainey, J. Burghardt, D. Gierth, J. Unge-rechts, I. Lux, P. Seifert, C. Bal and K.P. Schaal. 1997. De-scription of Nocardiopsis synnemetaformans sp.nov., elevation ofN. alba sub sp. prasina to N. prasina comb. nov., and designationof N. antarctica and N. alborubida as later subjective synonymsof N. dassonvillei. Int. J. Syst. Bacteriol. 47: 983�988.


ORIGINAL PAPER

Introduction

Samanea saman is a fast growing woody mimosoidlegume that is cultivated in many parts of Pakistan infarmlands and along roadsides. It provides shade andfuel wood. Its wood is also used for making bowls,trays, furniture etc. The tree forms nitrogen-fixingnodules with rhizobia (Allen and Allen, 1981). Theindigenous woody legumes and their root nodule bac-teria play an important role in the overall nitrogenincrement of Pakistani soils (Mahmood, 1999). A di-verse group of Gram-negative nodule forming bacterianamely Rhizobium, Bradyrhizobium, Allorhizobium,Azorhizobium, Mesorhizobium, Sinorhizobium andMethylobacterium have been recognized. They are themembers of the " and $ subgroup of the phylumProteobacteria (Chen et al., 2003), collectively knownas rhizobia (Amarger, 2001; Vessey et al., 2004). Thefixed nitrogen is used by S. saman for its growth andenrichment of the rhizosphere. The process of noduleformation is closely related to the infection of roots byappropriate rhizobia. Rhizobia enter the root via roothairs in majority of legumes (Iqbal and Mahmood,1992; Qadri and Mahmood 2003, 2004, 2005). Struc-

tural studies of tree legume nodules have been conduc-ted on Sesbania sesban (Mahmood and Jamal, 1977),Prosopis glandulosa (Baird et al., 1985), Andira sp.(Faria et al., 1986), Leucaena leucocephala (Iqbal andMahmood, 1992), Anadenanthera peregrina (Grosset al., 2002), Dalbergia sissoo (Qadri and Mahmood,2002, 2004), Albizia lebbeck (Qadri and Mahmood,2005) and Pithecellubium dulce (unpublished). Al-though S. saman has been cultivated in Pakistan fora long time, studies on structure of its nodules arelacking. This paper describes mode of infection anddevelopment and structure of S. saman nodules.

Experimental


Material collection and preparation for micro-scopy. Nodules of S. saman were collected from rootsof trees growing in the garden of the Department ofBotany, University of Karachi. For light microscopy,the nodules and roots were fixed in F.A.A. (formaline-acetic acid-ethyl alcohol) in the ratio of 5:5:90 for

Ultra-structural Studieson Root Nodules of Samanea saman (Jacq.) Merr. (Leguminosae)

RAIHA QADRI1, A. MAHMOOD1 and MOHAMMAD ATHAR2*

1 Department of Botany, University of Karachi, Karachi, Pakistan2 California Department of Food and Agriculture, Sacramento, CA, USA

Received 19 March 2007, revised 25 June 2007, accepted 12 July 2007

A b s t r a c t

Ultra-structural studies were conducted on root nodules of Samanea saman (Jacq.) Merr. collected from trees growing under naturalconditions. Nodules were distributed singly as well as in clusters on the main and lateral roots. Mature nodules were elongated, branchedand coralloid. Root hair curling was found but infection threads could not be observed. Rhizobia entered through the epidermis and movedintercellularly through the cortical region. Mature nodules of S. saman could be differentiated into meristem, cortex, vascular tissue andbacteroid tissue. The latter showed both infected and non-infected cells mixed together. Vascular bundles were inversely collateral anddistributed around the bacteroid tissue. The bacteroids were enclosed in peribacteroid membrane in groups and showed prominent gran-ules of polyhydroxybutyrate in their cytoplasm. Mycorrhizal hyphae were also observed along with rhizobia in the bacteroid tissue.S. saman with dual rhizobial and mycorrhizal infection is a potential tree for plantation in arid soils of Pakistan.

K e y w o r d s: Samanea saman, bacteroids, mycorrhiza, rhizobia, root nodules ultra-structure, woody legume

* Corresponding author: M. Athar, California Department of Food and Agriculture, 3288 Meadowview Road, Sacramento,CA 95832, USA; e-mail: [email protected]

200 Qadri R. et al. 3

18 hours. pieces of nodules (1�2 mm) were dehy-drated in ethanol series and infiltrated with L.R. (Lon-don resin) white at room temperature and polymeri-zed at 60°C for 24 hours. Serial sections (0.5�2 mm)were cut with a glass knife using a Sorrall J.B.-4 ultramicrotome and transferred to glass slides in a largedrop of water. The sections were dried on a hot plateat 40°C, stained with aqueous toluidine blue (in 1.0%borax, pH 4.4) and mounted in Canada balsam (Fariaet al., 1986). They were than examined under a Zeissstudent microscope.

Transmission electron microscopy assay. Fortransmission electron microscopy (TEM), small piecesof nodules (1�2 mm) were fixed in 2% gluteraldehydein 0.1 M phosphate buffer (pH 7) for 4 hours, washedwith three changes of buffer solution during three hoursand transferred to 1% aqueous osmium tetroxide for2�4 hours at room temperature. The fixed materialwas processed for transmission electron microscopyas described by Qadri and Mahmood (2003).

Scanning microscopy assay. Complete nodulesand free hand sections of nodules were fixed for scan-ning electron microscopy as for TEM. They weredehydrated in 100% ethanol followed by an ethanol/acetone mixtures up to 100% acetone (Faria et al.,1986). The specimens were then dried using a Polaroncritical point drier (BIO-RAD), coated with goldin coating unit (JFC-1100) and examined under (JeolT-20) scanning electron microscope.


Nodules of S. saman were distributed on the mainas well as lateral roots and occurred singly and inclusters (Fig. 1A). Although root hair curling was ob-served, infection threads could not be seen (Fig. 1B).Faria et al. (1987a, b) have surveyed the occurrenceof infection threads in the three sub-families of legumesnamely Caesalpinoideae, Mimosoideae and Papi-lionoideae. According to their survey, infection inmembers of Mimosoideae occurs by the movement ofrhizobia intercellularly rather than by infectionthreads. Similar observations have been reported byDart (1977), Chandler et al. (1982) and Calvert et al.(1984). The bacteria entered the ruptured epidermisof the root, from where they spread intercellularly intothe cortical region (Fig. 1C). Continuous proliferationof rhizobia in host cells resulted in the formation ofwell organized indeterminate nodules (Fig. 1D).

The general structure of the nodules of S. samanshared similarities with the majority of leguminousplants in having a nodule meristem (M), nodule cor-tex (NC), bacteroid region (B) and vascular supply(VS) (Fig. 1D). The nodule meristem was comprisedof numerous small compact cells. This is the region of

active nuclear division. Normally these cells containneither infection threads nor rhizobia. The meristem-atic region persisted throughout nodule development.The nodule cortex was comprised of 4�10 layersof non-infected parenchyma, isodiametric in shape.Cortical cells are derived by division of cells ofthe meristematic zone. Tannins were found scatteredthroughout the cortical region as idioblasts (Fig. 1D).Bacteroid region occupied the central part of thenodule. The bacteroid tissues of the nodules showedboth infected (IN) and uninfected (UN) cells mixedtogether (Fig. 1E). Similar observations have beenmade for Sesbania grandiflora (Harris et al., 1949),Cajanus indicus (Arora, 1956), Cyamopsis tetragono-loba (Narayana, 1963), Glycine max (Bergersen andGoodchild, 1973), Trifolium alexandrium (Naz andMahmood, 1976), Albizia spp. (Dart, 1977), Sesbaniasesban (Mahmood and Jamal, 1977), Parasponiaandersonii (Trinick, 1979), Phaseolus vulgaris (Bairdand Webster, 1982), Leucaena leucocephala (Iqbaland Mahmood, 1992), Dalbergia sissoo (Qadri andMahmood, 2002, 2004), Albizia lebbeck (Qadriand Mahmood, 2005) and Pithecellobium dulce(unpublished). A group of bacteria were enclosed ina common peribacteroid membrane (Fig. 2B). Thebacteria contained prominent granules of polyhydro-xybutyrate (PHB) (Fig. 2B). Both oval and rod shapedbacteria were observed (Fig. 2C). Enclosure of a groupof bacteria in a common peribacteroid membrane isa distinctive feature of infected cells in leguminousroot nodules as reported by a number of investigators(Newcomb, 1976; Lawrie, 1983; Chalifour and Ben-hamou, 1988; Qadri et al., 2006). The peribacteroidmembrane is derived initially from the host plasmamembrane and is a plant product. The peribacteroidmembrane become lost at certain points and bacteriaare released into the cytoplasm of the cell from thesesites (Fig. 2C). The liberated or free bacteria are al-ways surrounded by a peribacteroid membrane whichis derived from the bulges of the plasma membranesurrounding a group of bacteria (Fig. 2C) as describedby Newcomb (1976).

The vascular differentiation of the nodule is dis-cernible within a week after nodule initiation. Litera-ture on the subject has been reviewed by Bond (1948),Naz and Mahmood (1976) and Baird et al. (1985).The first indication of formation of conducting tissuebecomes evident in the form of a few cortical cellsthat start dividing parallel to the radius of the rootforming the procambial strands. Very soon thesestrands get connected with the protoxylem points ofthe vascular cylinder of the parent root. The vascularsupply may consist of one to four vascular strands(Bond, 1948). In S. saman two vascular strands wereseen making connection with the vascular supply of themain root (Fig. 1D). Two vascular strands have been

201Ultrastructure of root nodules of S. saman3

Fig. 1. A: Distribution of nodules on roots ofS. saman.B: Light photomicrograph of S. saman root,showing curled root hair (RH) (magnification× 400).C: Infection taking place through rupturedepidermis (E) Rhizobia move intercellularly.Patches of bacterial mass (BM) can be seen inthe root cortex (magnification × 200).D: Light photo micrograph of longitudinalsection of S. saman nodule showing deeplysituated meristem (M) central bacteriod region,B, and nodule cortex (NC). There is a heavydeposition of tannins (T) in the nodule tissue.Vascular tissue of the parent root VS(R), mak-ing connection with the vascular supply ofnodule VS (N) (magnification × 283).E: A scanning electron micrograph of rootnodule of S. saman showing infected (IN), anduninfected (UN) cells, and vascular bundles(VB) in the cortex (magnification × 68).


Fig. 2. A: An enlarged view of inversely col-lateral vascular bundle of S. saman showingxylem (X), phloem (P) and endodermis (EN)(magnification × 3000).B: Transmission electron micrograph ofa portion of bacteriod region (B) showingrhizobia enclosed in a common peribac-teriod membrane (PM). They show a highcontent of polyhydroxybutyrate (PHB) gra-nules (magnification × 6364).C: Transmission electron micrograph ofa portion of bacteroid region (B) of a rootnodule cell. Note that rhizobia (R) are en-closed in a common peribacteroid mem-brane (PM) and some of them are comingout of the membrane at certain points. Bothelongated (E) and oval (OV) forms arepresent (magnification × 15 600).D: Transmission electron micrograph ofa single bacteroid (B) cell of a root noduleshowing rhizobia (R) and vesicular-arbus-cular mycorrhizae � VAM (VA) (magnifica-tion × 2160).E: Transmission electron micrograph of anenlarged portion of bacteroid cell in Fig. 2-Dshowing a single hypha (H) along withrhizobia (R) (magnification × 15000).

203Ultrastructure of root nodules of S. saman3

reported in Vicia faba (Bieberdorf, 1938), Sesbaniagrandiflora (Harris et al., 1949), Pisum sativum(Bond, 1948), Melilotus alba, Trifolium alexandrinum(Naz and Mahmood, 1976) and Pithecellobium dulce.Once formed, the strands branched repeatedly encir-cling the central bacteroid region (Fig. 1E). The vas-cular strands never come in direct contact with thebacteroid tissue. A few layers of parenchyma alwaysseparate the vascular tissue from the bacteroid tissue(Fig. 1D). Vascular bundles were inversely collateral(Fig. 2A). The xylem elements faced away from whilephloem elements faced towards the center. Inverselycollateral bundles have been reported in pea nodulesby Bond (1948). The vascular bundles were enclosedby an endodermis (Fig. 2A).

The infected cells of S. saman along with rhizobiaalso contained mycorrhizal hyphae (Figs. 2D and 2E).Vesicular-arbuscular mycorrhizae (VAM) have beenreported in nodules of some leguminous trees such asSesbania grandflora (Habte and Aziz, 1985), Acaciamangium, Albizia falcata, (Dela Cruz et al., 1988) andLeucaena leucocephala (Young, 1990). The presenceof mycorrhizae is known to enhance nodulation andnitrogen fixation by legumes (Amora-Lazecano et al.,1998; Johansson et al., 2004). Mycorrhizal fungi andnitrogen fixing bacteria often act synergistically on in-fection rate, mineral nutrition and plant growth (Rabieand Almadini, 2005). The beneficial effects of nitro-gen-fixing bacteria in combination with mycorrhizalfungus on plant growth have been discussed by Patrezeand Cordeiro (2004) and Domenech et al. (2004).

In conclusion it may be said that most of the mycor-rhizal research with nitrogen fixing trees has revolvedaround only a few selected tree species (Aziz andSylvia, 1992). Studies on VAM interactions with nitro-gen-fixing tree species should be conducted on a largescale. Mahmood (1999) has analyzed the nitrogen-fix-ing potential of indigenous woody legumes and dis-cussed their role in the improvement of denuded andderelict lands of Pakistan. S. saman with dual rhizobialand mycorrhizal infection is a potential tree for planta-tion in Pakistani soils in future afforestation schemes.

AcknowledgementsThe first author is thankful to British Council for providing

the technical assistance under ODA linkage program to carry outresearch at the University of Reading, England.

Sincere thanks are due to Prof. Dr. John Barnett, Head ofSchool of Plant Sciences, University of Reading, England for pro-viding research facilities. The help of Miss Stacie Oswalt (Cali-fornia Department of Food and Agriculture, Sacramento, CA) inmaking linguistics corrections of the text is greatly appreciated.

Literature

Allen O.N. and E.K. Allen. 1981. The Leguminosae: a SourceBook of Characteristics, Uses and Nodulation. The University ofWisconsin Press, Madison.

Amarger N. 2001. Rhizobia in the field. Adv. Agron. 73: 109�168.Amora-Lazecano E., M.M. Vazquez and R. Azcon. 1998. Re-sponse of nitrogen-transforming micro-organism to arbuscularmycorrhizal fungi. Biol. Fertil. Soils 27: 1�13.Arora N. 1956. Morphological study of root nodules on Cajanusindicus. Proc. 43rd Indian Sci. Congr. (Agr.): 224�245.Aziz T. and D.M. Sylvia. 1992. Utilization of vesicular-arbuscularmycorrhizal fungi in the establishment of nitrogen-fixing trees.pp. 167�194. In: N.S. Suba Rao and G. Rodriguez Barrueco (eds).Symbiosis in Nitrogen Fixing Tree. Oxford & IBH Publishing Co.(Pvt.) Ltd., New Delhi.Baird L.M. and B.D. Webster. 1982. Morphogenesis of effectiveand ineffective root nodules in Phaseolus vulgaris L. Bot. Gaz.143: 41�51.Baird L.M., R.A. Virginia and B.D. Webster. 1985. Develop-ment of root nodules in a woody legume, Prosopis glandulosaTorr. Bot. Gaz. 146: 39�43.Bergersen F.J. and D.J. Goodchild. 1973. Aeration pathway insoybean root nodules. Australian J. Biol. Sci. 26: 729�740.Bieberdorf F.W. 1938. The cytology and histology of the rootnodules of some Leguminosae. J. Amer. Soc. Agron. 30: 375�389.Bond L. 1948. Origin and developmental morphology of rootnodules of Pisum sativum L. Bot. Gaz. 109: 411�434.Calvert H.E., M.K. Pence, M. Pierce, N.S.A. Malik and W.D.Bauer. 1984. Anatomical analysis of the development and distri-bution of Rhizobium infection in soybean roots. Can. J. Bot. 62:2375�2384.Chalifour F.P. and N. Benhamou. 1989. Indirect evidence forcellulase production by Rhizobium in pea root nodules duringbacteriod differentiation: cytochemical aspects of cellulose break-down in rhizobial droplets. Can. J. Microbiol. 35: 821�829.Chandler M.R., R.A. Date and R.J. Roughley. 1982. Infectionand root nodule development in Stylosanthes species by Rhizo-bium. J. Exp. Bot. 33: 47�57.Chen W.M., L. Moulin, C. Bontemps, P. Vandamme, G. Benaand C. Boivin-Masson. 2003. Legume symbiotic nitrogen fixa-tion by $-proteobacteria is widespread in nature. J. Bacteriol.185: 7266�7272.Dart P.J. 1977. Infection and development of leguminous nodules.pp. 367�472. In: R.W.F. Hardy and W.S. Silver (eds). A Treatiseon Dinitrogen Fixation. Section III. Biology. John Wiley and Sons,New York.Dela Cruz R.E., M.Q. Manalo, N.S. Aggangan and J.D.Tambalo. 1988. Growth of three legume trees inoculated with VAmycorrhizal fungi and Rhizobium. Plant Soil 108: 111�115.de Faria S.M., H.C. de Lima, A.A. Franco, E.S.F. Mucci andJ.I. Sprent. 1987a. Nodulation of legume trees from South EastBrazil. Plant Soil 99: 347�356.de Faria S.M., S.G. McInroy and J.I. Sprent. 1987b. The occur-rence of infected cells, with persistent infection threads in legumeroot nodules. Can. J. Bot. 65: 553�558.de Faria S.M., J.M. Sutherland and J.I. Sprent. 1986. A newtype of infected cells in root nodules of Andira spp. (Leguminosae).Plant Sci. 45: 143�147.Domenech J., B. Ramos-Solano, A. Probanza, J.A. Lucas-Garcia, J.J. Colon and F.J. Gutierrez-Manero. 2004. Bacillusspp. and Pisolithus tinctorius effects on Quercus ilex ssp. ballota:a study of tree growth, rhizosphere community structure andmycorrhizal infection. Forest Ecol. Manage. 19: 293�303.Gross E., L. Cordeiro and F.H. Caetano. 2002. Nodule ultra-structure and initial growth of Anadenanthera peregrina (L.)Speg. var. falcata (Benth.) Altschul plants infected with rhizobia.Ann. Bot. 90: 175�183.Habte M. and T. Aziz. 1985. Response of Sesbania grandiflorato inoculation of soil with vesicular-arbuscular mycorrhizal endo-phytes in an oxisol. Biol. Fertil. Soils 7: 164�167.


Harris J.D., E.K. Allen and O.N. Allen. 1949. Morpho-logical development of nodules on Sesbania grandiflora Poir.with reference to the origin of nodule rootlets. Amer. J. Bot. 36:651�661.Iqbal R. and A. Mahmood. 1992. Structural studies on root nod-ules of Leucaena leucocephala with particular reference to theinfection process. Pak. J. Bot. 24: 142�152.Johansson J.F., L.R. Paul and R.D. Finlay. 2004. Microbial in-teractions in the mycorrhizosphere and their significance for sus-tainable agriculture. FEMS Microbiol. Ecol. 48: 1�13.Lawrie A.C. 1983. Infection and nodule development in Aotusericoides (Vent.) G. Don., a woody native Australian legumes.J. Exp. Bot. 34: 1168�1180.Mahmood A. 1999. A comparison of nitrogen concentrationbetween wild and cultivated legumes of Sindh. Pak. J. Bot. 31:183�192.Mahmood A. and S. Jamal. 1977. A contribution to the histo-logy of root nodules of Sesbania sesban L. Pak. J. Bot. 9: 39�46.Narayana H.S. 1963. A contribution to the structure of rootnodules in Cyamopsis tetragonoloba Taub. J. Ind. Bot. Soc. 42:273�279.Naz S. and A. Mahmood. 1976. Histology of the root nodulesof Melilotus albus and Trifolium alexandrinum. Pak. J. Bot. 8:95�101.Newcomb W. 1976. A correlated light and electron microscopicstudy of symbiotic growth and differentiation in Pisum sativumroot nodules. Can. J. Bot. 54: 2163�2186.

Patreze C.M. and L. Cordeiro. 2004. Nitrogen-fixing and vesicu-lar-arbuscular mycorrhizal symbioses in some tropical legumetrees of tribe Mimoseae. Forest Ecol. Manage. 196: 275�285.Qadri R. and A. Mahmood. 2002. Occurrence of persistent in-fection threads in the root nodules of Dalbergia sissoo Roxb. Pak.J. Bot. 34: 397�403.Qadri R. and A. Mahmood. 2003. Presence of rhizobia in thexylary elements of root nodules in Samanea saman (Jacq.) Merr.Pak. J. Bot. 35: 819�823.Qadri R. and A. Mahmood. 2004. Structural study of theroot nodules of Dalbergia sissoo Roxb. Int. J. Biol. Biotech. 1:535�538.Qadri R. and A. Mahmood. 2005. Ultrastructural studies on theroot nodules of Albizia lebbeck (Roxb.) Benth. Pak. J. Bot. 37:815�821.Rabie G.H. and A.M. Alamdini. 2005. Role of bioinoculants indevelopment of salt tolerance Vicia faba plants under salinitystress. Afr. J. Biotech. 41: 210�222.Trinick M.J. 1979. Structure of nitrogen-fixing nodules formedby Rhizobium on roots of Parasponia andersonii Planch. Can.J. Microbiol. 25: 565�578.Vessey J.K., K. Pawlowski and B. Bergman. 2004. Root-basedN

2-fixing symbiosis: legumes, actinorhizal plants, Parasponia sp.

and Cycads. Plant Soil 266: 205�230.Young C.C. 1990. Effects of phosphorus solubilizing bacteria andvesicular-arbuscular mycorrhizal fungi on the growth of tree spe-cies in subtropical soils. Soil Sci. Plant Nutr. 36: 225�231.


ORIGINAL PAPER

Introduction

The sulfate-reducing bacteria (SRB � group 7) arecapable not only of assimilative sulfate reduction butalso of dissimilatory reduction of sulfate or sulfur. Thedissimilatory pathway is the source of energy forSRB. The reductive character of metabolism, espe-cially of the dissimilatory pathway requires strictlyanaerobic conditions for SRB growth. These path-ways are coupled to the utilization of hydrocarbonderivatives; lactate is a very good substrate for mostSRB. However, Widdel and Pfennig (1981) postulatedthat the Gram-positive strains of Desulfotomaculumacetoxidans never utilized lactate as an electron do-nor and sporulated only when acetate was the organicsubstrate. Consequently, Campbell and Singleton (inBergey�s Manual of Systematic Bacteriology) de-scribed this species as growing on media with acetate,but not with lactate (Campbell and Singleton, 1986).

In contrast, we found that D. acetoxidans DSM 771consumed lactate, too (Pado and Paw³owska-Æwiêk,2004). Because the ability of this species to grow onmedium with lactate remains controversial (Holt et al.,1994), we have attempted to thoroughly investigate thegrowth of D. acetoxidans DSM 771 on acetate and onlactate. We also determined the effects of these carbonsources on the antioxidative activity of this bacterium.

Experimental

Material and Methods

Material. Desulfotomaculum acetoxidans strainDSM 771 was grown at room temperature (19�23°C).The primary inoculum was 1 ml active culturefrom Deutsche Sammlung von Mikroorganismen.Half volume of the inoculum was used immediately

Growth and Antioxidant Activity of Desulfotomaculum acetoxidans DSM 771Cultivated in Acetate or Lactate Containing Media

LUCYNA PAW£OWSKA-ÆWIÊK* and RYSZARD PADO

Pedagogical University of Cracow, Departament of Microbiology, Kraków, Poland

Received 9 November 2006, revised 23 March 2007, accepted 20 May 2007

A b s t r a c t

Three independent 28 or 32-day stationary cultures of Desulfotomaculum acetoxidans DSM 771 strain were carried out under anoxicconditions in acetate or lactate-containing media. The acids were the sole carbon and energy sources in these media. During cultivation theturbidity (for calculation of cell division index) and hydrogen sulfide contents were determined in culture broth and reduced glutathioneand protein concentrations were assayed in culture broth supernatant. In these three successive cultures, the bacterium initially grew muchfaster on lactate than on acetate. However, after two weeks of culture this difference disappeared and in fact the growth rate was higher onacetate than on lactate. The level of H

2S formed (product of the dissimilatory pathway of sulfate reduction) demonstrated that this pathway

was more effective when lactate was a carbon source and the average H2S concentration was from over 3-fold to about 9-fold greater in

lactate than in acetate cultures. Also GSH (glutathione, product of the assimilatory sulfate reduction pathway) average level was about2-fold higher in lactate-grown cultures. The high negative values of the correlation coefficients between GSH and O

2 levels, especially

during the first 4 days of cultivation, indicate that GSH is a very important antioxidizing extracellular agent of D. acetoxidans. The rapidincrease in GSH level, preceding the release of H

2S, indicates the metabolic priority of the assimilation pathway of sulfate reduction. For

both carbon sources the highest coefficient of correlation was found between protein and H2S levels. These results suggest that hydrogen

sulfide is bound by proteins (which contain cysteinyl residues) secreted by D. acetoxidans cells. Indicated way of H2S bounding could

result in its acccumulation. This coefficient of correlation increased gradually in the successive cultures. The ratio of H2S concentration to

protein concentration increased gradually in the successive cultures, too.

K e y w o r d s: growth of D. acetoxidans on acetate and on lactate, antioxidant activity

* Corresponding author: L. Paw³owska-Æwiêk, Pedagogical University of Cracow Department of Microbiology, ul. Podbrzezie 3,31-054 Kraków, Poland; fax: (48) 12 6626709; e-mail; [email protected]

206 Paw³owska-Æwiêk L. and Pado R. 3

to inoculate 50 ml of medium with 42 mM acetateand the other half to inoculate 50 ml of medium with42 mM lactate. After 3 weeks both cultures weresupplemented with 50 ml of the respective fresh me-dium and further kept in the dark at room tempera-ture. After the next 3 weeks these cultures broths wereused as the inocula (45 ml) for the first 32-day culture(culture I). Each inoculum was supplemented with therespective fresh medium 3 weeks before the next cul-ture (culture II � strictly on completion of the firstcultures and culture III � six months after the firstculture was completed). All the cultures were con-ducted in parallel for acetate or lactate-supplementedmedia in 500 ml Erlenmayer flasks each containing450 ml of the culture medium. The lactate concentra-tion was determined on the basis of the results of ear-lier cultures (Pado and Paw³owska-Æwiêk, 2004). Inthe first series of cultures, also the culture in lactate-supplemented medium but inoculated with the acetate-containing inoculum was executed. After fixing oxy-gen detectors and inoculation, the media were imme-diately covered with a liquid paraffin layer (about5 mm thick), which was maintained throughout theculture. This paraffin layer made easier monitoring,particularly of oxygen level in the culture, without therisk of the culture being exposed to air.

Other medium components were as follows:21.12 mM Na2SO4, 1.15 mM KH2PO4, 4.02 mM KCl,5.61 mM NH4Cl, 1.13 mM CaCl2, 1.97 mM MgCl2,85.55 mM NaCl and trace elements (according toDSM-bank instruction) (Pado and Paw³owska-Æwiêk,2004, Paw³owska-Æwiêk and Pado, 2005).

Growth. The classic Monod�s method (Monod,1949) of graphical representation of bacterial growthin continuous cultures (bacterial growth curve) consistsin plotting the number of living cells in 1 ml of culturebroth as a function of cultivation time but this assay is

very time-consuming. Therefore it is frequently re-placed by a simpler method, based on culture turbiditymeasurements (nephelometry) used for cell numberevaluation (Gottschal, 1992). On the basis of the re-sults of our earlier study the high correlation coeffi-cient (0.6295) between the Monod�s and nephelometrymethods was found. The culture turbidity measure-ments facilitated much faster determination of the celldivision index (CDI). The turbidity of culture brothwas measured throughout the cultures (as shown inTable I) at 580 nm using a Specol 11 colorimeter witha TK attachment (Carl Zeiss Jena). Each result in thetable is an arithmetic mean from five measurementswith standard deviation ranging from 4 to 11%. Priorto sampling the flasks were gently manually agitatedfor 10 min. On the basis of turbidity values the celldivision index (CDI) was calculated:

CDI = Jx / J0

where: J0 � turbidity at J0 (CDI on the day of inocula-tion is 1.00), Jx � turbidity in successive days of theculture (Jx).

Chemical analysis. Proteins and reduced glu-tathione contents were assayed (without using any cellmembrane disrupting agents) in culture broth super-natant after centrifugation at 6000×g for 15 min. Theamount of proteins was estimated by the Lowry�smethod (Lowry et al., 1951). Reduced glutathioneconcentration was determined by the method de-scribed by Akerboom and Sies (1981). The concen-tration of hydrogen sulfide was measured in culturebroth by the methylene blue method of Fago andPopowsky (1949) but the samples with reagents wereleft overnight.

The colorimetric analyses and spectrum scanningwere performed using the CECIL 8020 spectrophoto-meter. The standard curves obtained for known concen-

acetate10.5 13.3 14.8 15.0 29.35 28.7 31.2 45.8 31.8 92.1 29.95 26.85

4.8/b 7.6/b 7.5/b 7.7/b 8.2/b 6.3/b 6.5/b 11.8/b 12.2/b 14.5/b 15.1/b 13.1/b

Ilactate

11.5 25.8 54.7 102.5 175.65 157.1 108.15 100.9 108.8 133.6 88.5 69.64.8/b 8.5/b 12.8/b 17.2/b 18.6/b 14.7/b 13.3/b 12.2/b 11.9/b 15.2/b 16.2/b 15.8/b

acetate21.3 26.1 23.0 26.2 26.0 10.8 26.0 23.9 16.7 36.7 46.1 29.7

II8.7/b 8.1/b 7.9/b 7.3/b 8.5/b 8.5/b 6.8/b 5.9/b 5.8/b 8.7/b 9.3/b 7.9/b

lactate38.5 49.0 54.8 67.8 59.7 53.2 43.4 59.7 61.1 61.5 92.4 66.9

9.1/b 9.9/b 10.2/b 13.1/b 9.2/b 11.3/b 10.5/b 9.9/b 10.3/b 11.5/b 12.9/b 9.6/b

acetate20.6 13.0 11.0 10.0 6.8 12.8 26.8 40.0 19.2 11.0 37.0 17.0 40.1

III8.4/b 5.2/b 5.6/b 5.0/b 6.9/b 7.2/b 6.4/b 6.6/b 6.5/b 5.0/b 10.1/b 10.3/b 12.3/b

lactate40.8 35.7 34.2 74.7 80.8 88.7 70.5 26.65 17.0 19.85 16.8 16.2 13.5

9.6/b 9.9/b 9.8/b 13.1/b 12.8/b 13.5/b 11.8/b 8.9/b 7.0/b 9.1/b 10.5/b 10.8/b 11.6/b

Table IAverage protein levels (mg/ml) in culture broth supernatant during successive cultures

Series 0/a

31

Culture day

2821 2414 171 2 3 7

/a � inoculation day; /b � standard deviation [%]

4 10

207Antioxidant activity of D. acetooxidans DSM7713

trations of respective standards. All the reagents (inclu-ding standards: albumin for protein, Na2S for hydro-gen sulfide and GSH) were of analytical grade fromMerck or Fluka. The results of triplicate assays are pre-sented as arithmetic mean ± the standard deviation (thelatter was similar for protein, GSH and H2S) (Table I).

The oxygen level in culture broth was measuredbefore mixing the cultures to avoid an error causedby air diffusion (the differences of oxygen concentra-tion before and after mixing were 2.6�4.3 µM O2).Because between the measurements, the cultureswere kept without mixing, these measurements weremade very gently for five different positions of theculture CTN-980 R oxygen detector (ELSENT Poland)coupled to a CX-315 microcomputer pH/oxygenmeter(ELMETRON Poland).


Cell division index. The obtained values of CDIrevealed that the bacterium grew faster on lactate(about 2-fold higher turbidity) than on acetate withinthe first two weeks of the culture (Fig. 1). After thistime, in the first and the third cultures, the CDI wasslightly higher on acetate. After 24 days CDI wasagain higher (about 3-fold) for the lactate containingculture medium (in culture III). However, the culture

on lactate medium but inoculated with acetate inocu-lum showed increase of neither cell division indexnor hydrogen sulfide level. On the other hand, thelack of an increase in CDI values when the lactatemedium was inoculated with acetate inoculum indi-cates that the adaptation process requires a relativelylong time for the changeover of metabolic path-ways, essential for switching on lactate catabolism.The necessity of a changeover of metabolic pathwayswas confirmed by an earlier observation regarding thesynthesis of different redox proteins in cultures ofthree Desulfovibrio strains grown on hydrogen or lac-tate (Steger et al., 2002).

The cell division is related to biosynthesis pro-cesses, particularly the production of proteins. Accor-ding to some authors (Hancock and Poxton, 1988;Russel, 1988) �free� wall-associated proteins willcontinue to be synthesized and will be released directlyinto the culture supernatant. The obtained averagecontent of determined protein in supernatant was atleast 2-fold higher in lactate cultures and in culture I(when the cultures were inoculated with the youngestinoculum) were even 3-fold higher than in acetate cul-tures (Table I and II). Interestingly, as the inoculumgrew older the average protein level was progres-sively reduced in both cultures, but more in the cul-tures on lactate. Thus, the above � mentioned decreasein the protein content in cultures inoculated with aged

Fig. 1. Cell division index during cultivation of cultures on acetate (thin lines) or lactate (thick lines): culture I � solid lines;culture III (six months after the first series was completed) � dashed lines; culture on a medium with lactate but inoculated

with an acetate inoculum � pointed line. For more clear illustration culture II (after the first series was completed) is notpresented, since it was similar to series I. Student�s t-test values for lactate to acetate culture in successive cultures: I � 1.967

(statistically insignificance where p>0.05); II � 3.969 (0.01 significance level); III � 2.588 (0.05 significance level).


inoculum suggests that the age of the inoculum com-promises the capability of cells to synthesis and secre-tion of proteins. The highest coefficient of correlationbetween CDI and protein level was found for culture I(Table III). However, this correlation coefficient inculture III (with 7-month-old inoculum) was muchhigher for the lactate than acetate culture.

Reports on D. acetoxidans are rather scarce. Thisspecies has not been grown earlier on lactate(Stackebrandt et al., 1997; Hristova et al., 2000;Scholten and Stams, 2000; Boschker et al., 2001;Londry and Des Marais, 2003; Londry et al., 2004),because it was commonly believed to be unable togrow on media containing lactate as a sole carbonsource (Widdel and Pfenning, 1981; Campbell andSingleton, 1986; Holt et al., 1994). Our experienceshowed that D. acetoxidans DSM 771 was also ca-pable of catabolic utilization of lactate (Pado andPaw³owska-Æwiêk, 2004; Pado and Paw³owska-Æwiêk, 2005; Paw³owska-Æwiêk and Pado, 2005).However, in agreement with earlier findings (Widdeland Pfenning, 1981; Campbell and Singleton, 1986),we did not observe sporulation, even after 80 daysof culturing in the presence of lactate (Pado andPaw³owska-Æwiêk, 2004).

The obtained relationship between the secretedprotein level and CDI is reflected in the correlationcoefficient but only in the first series, especially inthe acetate culture (Table III). These results are in ac-cordance with the data of Londry and Des Marais(2003) who used 13C acetate. Those authors provedthat D. acetoxidans (unlike three other species of

SRB) effectively incorporated acetate into biomass viaacetyl-CoA. Moreover, they observed that this specieswas capable of lithotrophic growth using carbonateand gaseous CO2. This lithotrophic growth capabilitycould explain the better growth of D. acetoxidans inacetate culture but only after two weeks of the cultiva-tion (in the first and second cultures), when carbonate(including dissolved CO2) accumulated as a conse-quence of acetate catabolism (Fig. 1).

Reduced glutathione. Also the GSH level washigher in cultures with lactate than in those with ace-tate (Fig. 2�4). As the cultures with lactate producedslightly higher levels of both GSH and H2S it suggeststhat lactate is more advantageous for the assimilatoryand dissimilatory sulfate reduction pathways too. It isknown that lactate contains more hydrogen atomsthan acetate and this is very important in sulfate re-duction processes. Since cell membranes were not dis-rupted prior to GSH determination, the measuredGSH was extracellular. The results obtained in allthree cultures (designated I, II and III) indicate thatGSH biosynthesis and secretion began immediatelyafter inoculation and in early cultures the GSH levelincreased more rapidly than that of H2S (Fig. 2�4).The early GSH domination over H2S suggests prior-ity of the assimilatory over the dissimilatory pathway.

The initial sulfate concentration in the media was21 mM. On the basis of the highest GSH levels (al-ways during the first four days of the cultures), sulfurincorporation from sulfate into GSH was counted: itranged from 0.35 to 2.55� in acetate cultures andfrom 0.68 to 3.74� in lactate cultures. Contrary to

Fig. 2. Hydrogen sulfide (solid lines) and glutathione (dotted lines) levels within culture I:acetate cultures (thin lines) or lactate cultures (thick lines).


our expectation we did not find any correlation be-tween GSH and H2S levels (Table III). The absenceof such correlation can explain different metabolic re-quirements of bacterial cells during of the cultivation.The observed regular, significant and negative valuesof correlation coefficients between GSH and oxygen(the more GSH the less oxygen) prove that the re-duced glutathione performs the role of an antioxidant,as expected (Table III, compare Fig. 2�4 and 5). Theantioxidant role of GSH was especially clear at thebeginning of cultivation (the first 4 days), becausewithin this period the greatest decreased in oxygenlevel was observed (by about 100 µM).

As it is well known, GSH is the major antioxidantagent (both extra- and intracellular) in all live organ-isms (Poot et al., 1995; Deneke 2000; Hand andHonek, 2005). The obtained results (the ratio of GSHand protein concentration) showed that lactate stimu-lated the production of GSH, so thus increasing theantioxidant activity of the examined strain. Theselevels of GSH were lower (Table II) as compared toFareleira et al. (2003) (1.8±0.6 nmol GSH per mgprotein of Desulfovibrio gigas cells). However, we de-termined the extracellular GSH, while those authorsdetermined the total GSH. In this work, a rapid in-crease in GSH level was observed at the beginning of

Fig. 3. Hydrogen sulfide (solid lines) and glutathione (dotted lines) levels within culture II:acetate cultures (thin lines) or lactate cultures (thick lines).

Fig. 4. Hydrogen sulfide (solid lines) and glutathione (dotted lines) levels within culture III:acetate cultures (thin lines) or lactate cultures (thick lines).


the culture (Fig. 2�4). However, Fareleira et al. (2003)did not observe any significant differences when oxy-gen concentration in the medium or the duration of theoxic period in D. gigas cells were increased. This phe-nomenon may be explained by GSH secretion ini-tiated by O2 and/or reactive oxygen species present inthe fresh medium and taxonomic differences betweenthese strains. Extracellular oxygen-utilizing processesare well known in eucaryotic Deuteromycotine (Odierand Artaud, 1992; Leonowicz et al., 2001). This stra-tegy should also apply to anaerobic bacteria becauseit protects the cells against the penetration of toxicradicals into the cytosol.

Hydrogen sulfide. The amounts of H2S (or sul-fide) were much greater in cultures grown on lactatethan on acetate (Table II): in the first culture about10-fold higher (during the first week of cultivation� Fig. 2); in the second culture about 30-fold higher(during the first week � Fig. 3); and in the third cul-ture over 45-fold higher (during the first week of cul-tivation � Fig. 4).

The absorption spectra of 2-days samples (in thefirst culture) after addition of methylene blue method�sreagents showed the presence of peaks at 411.6 and665 nm (which is characteristic for product formedin this assay) only for the lactate culture (Fig 5A).Our earlier research showed that the complex of4-hydroxybenzoate and ferrous ions (present in the re-agent used for this assay) exhibited absorption maxi-mum at 411.4 nm (Paw³owska-Æwiêk and Pado, 2005).In this culture (culture I) the absorption spectra of17-days sample from lactate culture showed peak at502.1 nm and from acetate culture only just 32-dayssample (Fig. 5B). This absorption maximum was cha-racteristic for complex ferrous ions and 4-hydroxy-3-

sulfobenzoate and this ligand was product of 4-hydro-xybenzoate sulfonation. The 4-hydroxy-3-sulfoben-zoate as extracellular metabolite was requisite for sul-fate transport processes in this strain (Paw³owska-Æwiêk and Pado, 2005). The obtained results suggestthat lactate was more efficiently for sulfate transportprocesses than acetate in D. acetoxidans. The moreefficiency of lactate requires further research. The in-corporated sulfur index was counted: in lactate cul-tures it ranged from 4.66 to 6.07�, while in acetateones from 0.83 to 1.45� (Table II). Kaplan andRittenberg (1964) observed sulfur isotope fractiona-tion for D. desulfuricans increasing in the order lac-tate, acetate and ethanol. A correlation between sulfatereduction rates and fractionation was also confirmedby continuous culture experiments (Chambers et al.,1975). Detmers et al. (2001) examined 32 species ofSRB and found that all incomplete-lactate-oxidizingsulfate reducers fractionated 2.0�17� of an isotopeof sulfur (34S), whereas all examined acetate-oxidizingspecies fractionated 18.0�22.0�.

The present results show that D. acetoxidans iscapable not only of complete acetate oxidation, butit can also utilize lactate. The incomplete oxidationof lactate to acetate by sulfate yields 3-fold moreenergy than the complete oxidation of acetate to CO2(Londry and Des Marais, 2003). So probably theexamined strain in the first stage metabolizes lactateto acetate but produces less hydrogen sulfide, and inthe next stage it oxidizes acetate generating more H2S(Fig. 2�4), as follows:

SO42� + 2 CH3CH(OH)COO�→→→ 2 acetylCoA +

+ 2 HCO3� + HS� (1)

SO42� + 2 acetylCoA →→→ 4 CO2 + H2S (2)

I 30.78 94.73 3.08 4.480 40.923 9.13 27.738 53.126 1.92

5.186 0.146/b 0.432/b 2.496 0.901/d 0.561/d 3.731(0.001) 0.83/c 6.07/c (0.05) 2.55/e 3.37/e (0.01)

II 26.04 58.97 2.26 9.579 34.622 3.61 7.719 13.249 1.72

15.40/a 37.75/a 11.123 0.368/b 0.587/b 3.352 0.296/d 0.225/d 3.069(0.001) 1.45/c 4.66/c (0.01) 0.35/e 0.68/e (0.02)

III 20.41 41.18 2.02 4.813 34.458 7.16 10.040 24.117 2.40

21.62/a 30,17/a 2.151 0.236/b 0.837/b 2.050 0.492/d 0.586/d 2.059nss 0.83/c 5.54/c nss 0.64/e 3.74/e nss

Table IIAverage levels of GSH (µM) and protein (mg/ml) in culture broth supernatant and H

2S levels

in culture broth and Student�s t-test values in individual series

SeriesProteinAcetate

ProteinLactate

AH

2S

AcetateH

2S

LactateA

GSHAcetate

GSHLactate

A

A � the ratio of protein or H2S or GSH average concentration in lactate culture to in acetate culture; bold values� Student�s t-test for lactate to acetate culture; ( ) � significance level; nss � statistically insignificance (p> 0.05);/a � decrease relative to previous series (%); /b � the ratio of H2S and protein concentration (nmol/mg protein);/c � sulfur incorporation from sulfate (initial concentration 21 mM) into hydrogen sulfide (on final cultivation day)(�); /d � the ratio of GSH and protein concentration [nmol/mg protein]; /e � sulfur incorporation from sulfate (21 mM)into glutathione (in early exponential phase) (�).


Although the average H2S level decreased in con-secutive cultures, this decrease was smaller in the cul-tures with lactate than acetate (Table II). The calcu-lated correlation coefficients show that H2S amountscorrespond to levels of determined proteins (Table III).Despite the levels of both proteins and hydrogen sul-fide were reduced in the next culture, but their corre-lation coefficients increased in the successive cul-tures, both with acetate and lactate. In lactate culture,

in the third culture, was found higher H2S level insamples from immediately inoculated culture than onthe next day (in contrast to other culture-compareFig. 2, 3 and 4). The higher H2S amount in samplesfrom freshly inoculated lactate culture proves extra-cellular accumulation of hydrogen sulfide. Moreover,the average amount of H2S per mg of protein in-creased in subsequent series, but only in lactate cul-tures (Table II). These results suggest that extracellular

Fig. 5. Absorption spectra of culture samples: 1 � acetate culture; 2 � lactate culture after 2 days of cultivation (A)and 17 days (lactate culture) or 32 days (acetate culture) of cultivation (B) with methylene blue method�s reagents.

CDI � cell division index; /a � whole period of cultivation; /b � from the fourth day to the end; /c � for the first 4 daysof cultivation; /d � from inoculation day to 21-th day; /e � from the second day to the end of cultivation.

I acetate 0.6772/a 0.2647 �0.2760 �0.5801 �0.4386/a �0.6049 0.1223/a �0.1394/a

0.7105/b �0.5426/c 0.1278/d 0.5968/e

I lactate 0.2033/a �0.3107 0.8451 0.3093 �0.5916/a 0.0382 0.1275/a �0.1326/a

0.3716/b �0.5856/c 0.1295/d 0.4821/e

II acetate �0.7571/a �0.7209 �0.3848 0.7549 �0.1341/a 0.5754 0.4226/a 0.0516/a

�0.7690/b �0.5587/c 0.4016/d 0.8278/e

II lactate �0.3301/a �0.6095 �0.6221 0.8162 �0.2531/a 0.7020 0.5886/a �0.2393/a

�0.7415/b �0.6826/c 0.6046/d �0.0346/e

III acetate 0.0407/a 0.2010 �0.1443 �0.1440 0.1863/a �0.0049 0.7449/a �0.5951/a

�0.0199/b �0.7210/c 0.7875/d �0.5105/e

III lactate 0.2181/a 0.0981 0.2761 0.0748 0.2286/a 0.4897 0.6760/a 0.4392/a

0.3623/b �0.5356/c 0.9357/d 0.2519/e

Table IIICorrelation coefficients

SeriesCulture

CDI vs.Protein

CDI vs.H

2S

GSH vs.CDI

GSH vs.Protein

GSHvs. O

2

GSHvs. H

2S

H2S vs.

ProteinH

2S

vs. O2


proteins bound dissimilated H2S. Our earlier researchshowed H2S accumulation by proteins linked to thecell wall this strain, because hydrogen sulfide levelswere much higher in the lysozyme-treated samplesthan in the untreated samples (Pado and Paw³owska-Æwiêk, 2004). Hydrogen sulfide can be bound bycysteinyl residues of proteins forming disulfides,which release the so-called labile sulfur in acidicenvironment (Ogasawara et al., 1994). However, theresults obtained in this work show a decrease of theamount of extracellular proteins with the age of usedinoculum, but may be, the copies number of proteincontaining of Cys residues increased (e.g. proteinsincluding in the transport processes). According toRussell (1988), these proteins could be associatedwith the cell wall, but not covalently linked.

Surprising was the fact that, contrary to expecta-tion, O2 did not decrease H2S level; in acetate cultures,in the first and the second series oxygen presence waseven advantageous, especially from the second dayuntil the end of cultivation (compare Fig. 2�4 and 6).Also the high values of correlation coefficients forthese cultures (Table III) show that oxygen could evenbe a positive factor for the dissimilatory sulfate reduc-tion pathway (e.g. through the influence on sulfatetransport processes) (Paw³owska-Æwiêk and Pado,2005). Also Johnson et al. (1997) found a positiveinfluence of oxygen (48 µM) on the growth of Desul-

fovibrio vulgaris if 250 µM hydrogen sulfide wasadded to the medium.

Conclusion. Although growth of D. acetoxidansDSM 771 on lactate requires at least two successivepassages on this medium the species grows better onlactate than on acetate, which is contrary to earlierobservations of other researches. D. acetoxidans lac-tate cultures produced higher levels of both GSH andH2S than cultures with acetate, so lactate is a bettersubstrate for metabolic processes, especially the sul-fate reduction pathways. The higher level of reducedglutathione in lactate cultures results in the increase ofD. acetoxidans antioxidant activity, which could bevery important for the survival in natural environment.

Literature

Akerboom T.P.M. and H. Sies. 1981. Assay of glutathione, glu-tathione disulfide and glutathione mixed disulfides in biologicalsamples. Meth. Enzymol. 77: 373�82.Boschker H.T., W. de Graaf, M. Koster, L. Meyer-Reil andT.E. Cappenberg. 2001. Bacterial populations and processes in-volved in acetate and propionate consumption in anoxic brackishsediment. FEMS Microbiol. Ecol. 35: 97�103.Campbell L.L. and R. Jr Singleton. 1986. Genus Desulfotoma-culum, p. 1200�1202. In: P.H.A. Sneath (ed.), Bergey�s Manual ofSystematic Bacteriology, Baltimore: The Williams and WilkinsCo., Baltimore.Chambers L.A., P.A. Trudinger, J.W. Smith and M.S. Burns.1975. Fractionation of sulfur isotopes by continuous culture of

Fig. 6. Oxygen levels during cultivation of cultures on acetate (thin lines) or lactate (thick lines): culture I� solid lines; culture III (six months after the first series was completed) � dashed lines. For more clearillustration culture II (after the first series was completed) is not presented, since it was similar to culture I.Student�s t-test values for lactate to acetate culture in successive cultures: I � 5.886 (0.001 significance

level); II � 2.795 (0.05 significance level); III � 1.424 (statistically insignificance where p>0.05).


Desulfovibrio desulfuricans. Can. J. Microbiol. 21: 1602�1607.Deneke S. 2000. Thiol-based antioxidants. Curr. Top. Cell. Reg.36: 151�179.Detmers J., V. Brüchert, K.S. Habicht and J. Kuever. 2001.Diversity of sulfur isotope fractionations by sulfate-reducingprokaryotes. Appl. Environ. Microbiol. 67: 888�894.Fago J.K. and M. Popowsky. 1949. Spectrophotometric deter-mination of hydrogen sulfide. Methylene blue method. Anal.Chem. 21: 732�734.Fareleira P., B.S. Santos, C. Antonio, P. Moradas-Ferreira,J. LeGall, A.V. Xavier and H. Santos. 2003. Response of a strictanaerobe to oxygen: survival strategies in Desulfovibrio gigas.Microbiology 149: 1513�1522.Gottschal J.C. 1992. Continuous culture, p. 559�572. In:J. Lederberg (ed.) Encyclopedia of Microbiology. Academic PressInc., San Diego.Hancock I.C. and I.R. Poxton. 1988. Isolation and purification ofcell walls, p. 55�67. In: I.C. Hancock and I.R. Poxton (ed.) Bac-terial Cell Surface Techniques. John Wiley & Sons, Chichester.Hand C.E. and J.F. Honek. 2005. Biological chemistry of natu-rally occurring thiols of microbial and marine origin. J. Nat. Prod.68: 293�308.Holt J.G., N.R. Krieg, P.H.A. Sneath, J.T. Staley and S.T. Wil-liams. 1994. Dissimilatory sulfate- or sulfur-reducing bacteria,p. 335�46. Bergey�s Manual of Determinative Bacteriology 9th ed.The Williams and Wilkins Co., Baltimore.Hristova K.R., M. Mau, D. Zheng, R.I. Aminov, R.I. Mackie,H.R. Gaskins and L. Raskin. 2000. Desulfotomaculum genus-and subgenus-specific 16S rRNA hybridization probes for envi-ronmental studies. Environ. Microbiol. 2: 143�159.Johnson M.S., I.B. Zhulin, M.E.R. Gapuzan and B.L. Taylor.1997. Oxygen-dependent growth of the obligate anaerobe Desul-fovibrio vulgaris Hildenborough. J. Bacteriol. 179: 5598�5601.Kaplan I.R. and S.C. Rittenberg. 1964. Microbiological frac-tionation of sulphur isotopes. J. Gen. Microbiol. 34: 195�212.Leonowicz A., N.S. Cho, J. Luterek, A. Wilko³azka, M. Wojta�-Wasilewska, A. Matuszewska, M. Hofrichter, D. Wesenbergand J. Rogalski. 2001. Fungal laccase: properties and activity onlignin. J. Basic Microbiol. 41: 185�227.Londry K.L. and D.J. Des Marais. 2003. Stable carbon isotopefractionation by sulfate-reducing bacteria. Appl. Environ. Micro-biol. 69: 2942�2949.Londry K.L., L.L. Jahnke and D.J. Des Marais. 2004. Stablecarbon isotope rations of lipid biomarkers of sulfate-reducing bac-teria. Appl. Environ. Microbiol. 70: 745�751.

Lowry O.H., N.J. Rosebrough, A.L. Farr and R.J. Randall.1951. Protein measurement with the Folin phenol reagent. J. Biol.Chem. 193: 265�275.Monod J. 1949. The growth of bacterial cultures. Ann. Rev.Microbiol. 3: 371�394.Odier E. and I. Artaud. 1992. Degradation of lignin, p. 162�185.In: G. Winkelmann (ed.), Microbial Degradation of Natural Pro-ducts. VCH, Wainheim.Ogasawara Y., S. Isoda and S. Tanabe. 1994. Tissue and sub-cellular distribution of bound and acid-labile sulfur, and the enzy-matic capacity for sulfide production in the rat. Biol. Pharm. Bull.17: 1535�1542.Pado R. and L. Paw³owska-Æwiêk. 2004. Changes during long-term growth of Desulfotomaculum acetoxidans DSM 771. ActaBiol. Crac. Ser. Bot. 46: 101�107.Pado R. and L. Paw³owska-Æwiêk. 2005. The uptake and accu-mulation of iron by the intestinal bacterium Desulfotomaculumacetoxidans DSM 771. Folia Biol. (Krakow) 53: 79�81.Paw³owska-Æwiêk L. and R. Pado. 2005. The role of benzoatescreted by Desulfotomaculum acetoxidans DSM 771 in sulfateuptake. Acta Biochim. Pol. 52: 797�802.Poot M., H. Teubert, P.S. Rabinovitch and T.J. Kavanagh.1995. De novo synthesis of glutathione is required for both entryinto and progression through the cell cycle. J. Cell. Physiol. 163:555�560.Russell R.R.B. 1988. Isolation and purification of proteins lin-ked to the cell wall in Gram-positive bacteria, p. 104�110. In:I.C. Hancock and I.R. Poxton (ed.) Bacterial Cell Surface Techni-ques. John Wiley & Sons, Chichester.Scholten J.C. and A.J. Stams. 2000. Isolation and characteriza-tion of acetate-utilizing anaerobes from a freshwater sediment.Microb. Ecol. 40: 292�299.Stackebrandt E., C. Sproer, F.A. Rainey, J. Burghardt,O. Pauker and H. Hippe. 1997. Phylogenetic analysis of thegenus Desulfotomaculum: evidence for the misclassification ofDesulfotomaculum guttoideum and description of Desulfotoma-culum orientis Desulfosporosinus orientis gen. nov. comb.nov.Int. J. Syst. Bacteriol. 47: 1134�1139.Steger J.L., C. Vincent, J.D. Ballard and L.R. Krumholz.2002. Desulfovibrio sp. genes involved in the respiration of sul-fate during metabolism of hydrogen and lactate. Appl. Environ.Microbiol. 68: 1932�1937.Widdel F. and N. Pfennig. 1981. Sporulation and further nutri-tional characteristics of Desusulfotomaculum acetoxidans. Arch.Microbiol. 129: 401�402.


ORIGINAL PAPER

Introduction

The aflatoxins are a biologically active polyketide-derived secondary metabolites (Bhatnagar et al.,1992). The aflatoxins are a group of closely relatedhighly oxygenated bisfurano-coumarin heterocycliccompounds (Buchi and Rae, 1969; Ellis et al., 1991).Aflatoxins are produced by some strains of Aspergil-lus flavus and most, if not all, Aspergillus parasiti-cus Speare (Smith and Moss, 1985) as well as theclosely related species Aspergillus nomius Kurtzman(Kurtzman et al., 1987; Cotty et al., 1994). Theseaflatoxin-producing species differ in their ability toproduce aflatoxins and some are entirely non-toxi-genic (Smith, 1997).

Chemically, aflatoxins are defined as a series of18 known bisfuran polycyclic compounds that fluores-cence strongly in ultraviolet light (Park et al., 2001).There are four naturally occurring aflatoxins B1, B2,G1 and G2 together with other aflatoxins which occurendogenously as metabolic products of microbial, ani-mal, or human metabolic systems (Smith, 1997).

Aflatoxins can be acutely toxic, carcinogenic,mutagenic, teratogenic, and immunosuppressive to

most mammalian species. The rank order of toxicity,carcinogenicity, etc., is AFB1 > AFG1 > AFB2 > AFG2implying that the unsaturated terminal furan of AFB1is critical for determining the level of biologicalactivity of the aflatoxins (Eaton and Gallagher, 1995;Smith, 1997).

The biological detoxification or the biotransforma-tion or degradation of aflatoxin by microbial systemsto a metabolite(s) that is either nontoxic when in-gested by animals or less toxic than the original toxinand readily excreted from the body is being studiedin several laboratories (Smith and Bol, 1989). As yet,such methods do not constitute a realistic practicalapproach (Smith, 1997) to the problem. Boller andSchroeder (1973, 1974) reported that A. cheralieriand A. candidus that dominated the mycoflora in ricealso showed marked inhibition in aflatoxin produc-tion by A. parasiticus Speare. Aspergillus oryzae andRhizopus nigricans (formerly Rhizopus stolonifer)and have also been reported to inhibit A. parasiticusand aflatoxin production (Christensen et al., 1973;Weckbach and Marth, 1977).

The capabilities of several fungal strains to resistaflatoxins as well as to biotransform and/or biode-

Bioremediation of Aflatoxins by Some Reference Fungal Strains

HUSSEIN H. EL-SHIEKH, HESHAM M. MAHDY* and MAHMOUD M. EL-AASER

Botany and Microbiology Department, Faculty of Science and Regional Center for Mycology and BiotechnologyAl-Azhar University, Nasr City, Cairo

Received 11 October 2005, resubmitted 20 April 2007, revised 12 July 2007, accepted 25 July 2007

A b s t r a c t

Aspergillus parasiticus RCMB 002001 (2) producing four types of aflatoxins B1, B

2, G

1 and G

2 was used in this study as an aflatoxin-

producer. Penicillium griseofulvum, P. urticae, Paecilomyces lilacinus, Trichoderma viride, Candida utilis, Saccharomyces cerevisiaeas well as a non-toxigenic strain of Aspergillus flavus were found to be able to exhibit growth on aflatoxin B

1-containing medium up to

a concentration of 500 ppb. It was also found that several fungal strains exhibited the growth in co-culture with A. parasiticus, naturalaflatoxins producer, and were able to decreased the total aflatoxin concentration, resulting in the highest inhibition percentage of 67.2% byT. viride, followed by P. lilacinus, P. griseofulvum, S. cerevisiae, C. utilis, P. urticae, Rhizopus nigricans and Mucor rouxii with totalaflatoxin inhibition percentage of 53.9, 52.4, 52, 51.7, 44, 38.2 and 35.4%, respectively. The separation of bioremediation products usingGC/MS revealed that the toxins were degraded into furan moieties.

K e y w o r d s: Aspergillus parasiticus, aflatoxins, bioremediation, GC/Mass determination

* Corresponding author: Hesham M. Mahdy; Botany and Microbiology Department, Faculty of Sciences, Al Azhar University,Madient Nasr City P.O. Box 11884, Cairo, Egypt; e-mail: [email protected]

216 El-Shiekh H.H. et al. 3

grade these compounds into several other metabolitesthat are either non- or less toxic than the original havebeen reported in this paper.

Experimental


Organisms used. Several strains of Aspergillusflavus group were examined for aflatoxin productioncapabilities. These cultures were provided by theRegional Center for Mycology and Biotechnology(RCMB) culture collection unit. A. flavus RCMB002002 (strains designated as 1, 2, 3, 4, 5, and 6),A. flavus var. columnaris RCMB 002003, A. parasiti-cus CMB 002001 (strains: 1 and 2), A. tamarii RCMB002004(2), A. oryzae RCMB 002015(1). Several otherfungal (provided by the same culture collection) werealso examined for their capabilities to grow in thepresence of aflatoxins. These strains were: P. griseo-fulvum RCMB 001007(2), P. urticae RCMB 001017,P. italicum RCMB 001018(1), P. roquefortii RCMB001009(6), P. nigricans RCMB 001013, P. citrinumRCMB 001011(1), P. notatum RCMB 001016, Circi-nella sp. RCMB 013001, Rhizopus nigricans RCMB014001, Mucor rouxii, RCMB 015002, Syncepha-lastrum racemosum RCMB 016001(2), Trichodermaviride RCMB 017002, Paecilomyces lilacinus RCMB018002, P. variotii RCMB 018003, Curvularia lunataRCMB 019002, C. clavata RCMB 019003, Rhizo-ctonia solani RCMB 031001, Torulomyces lagenaRCMB 030001, Acremonium rutilum RCMB 020002,Saccharomyces cerevisiae RCMB 006001 and Can-dida utilis RCMB 005002.

Preliminary detection of aflatoxin-producingfungi by fluorescent agar technique. Several strainsof A. flavus group were screened for their ability toproduce aflatoxin(s) on Sabouraud-dextrose yeast ex-tract agar plates, using the fluorescent agar techniqueof Hara et al. (1974).

Application of agar plug method. Seven-day oldcultures of Aspergillus spp. grown on MEA were exa-mined for aflatoxins production using an agar plugtechnique (Paterson and Bridge, 1994). An agar plugwas cut out with a flamed cork borer (inner diameter~0.4 cm) from the center of the colony. The plug wasremoved with the a sterile dissecting needle, wettedthe mycelial side of the plug with a drop of chloro-form/methanol (2:1, v:v) for a few seconds and touchedthe agar side to a thin layer chromatography (TLC)plate. The agar plug was placed at the origin of theTLC plate (20×20 cm Merck aluminum sheet, silicagel 60, layer thickness 0.2 mm). The diameter of theapplication spot should not be more than 0.6 cm. Afterdrying the spots, the TLC plate was developed in TEF

eluent (toluene/ethyl acetate/90% formic acid, 5:4:1,v/v/v) in a solvent saturated atmosphere using satura-tion pads. Griseofulvine was used as an external stan-dard. The dry developed TLC plates were viewed innormal white light, under long wave (366 nm), andshort wave (254 nm) UV light; then compared with thestandards and published data on colors and Rf valuesaccording to Paterson and Bridge (1994).

Production of aflatoxins. For the production ofaflatoxins yeast-extract sucrose (YES) liquid mediumwas used. For enhancement of aflatoxin production,1 ml trace element solution was added to 1 liter of YESmedium. The trace element solution is prepared by dis-solving 0.5 g magnesium sulphate, 0.5 g cupric sulpha-te and 0.5 g zinc sulphate in 100 ml of distilled water.

Preparation of spore suspension. Mould ino-culum was prepared by growing A. parasiticus onMEA slants for 7�10 days at 25°C until sporulation.The spores were harvested by adding 10 ml of steriledistilled water to the cultures on the surface of theagar slants and gently dislodging spores from conidio-phores with an inoculation loop. The spore suspen-sion was filtered through four layers of sterile cheese-cloth followed by filtration through Whatman No 1filter paper to remove mycelial debris. Spores werecounted using an Improved Neubauer bright linehemocytometer. Appropriate dilutions were made fromthe stock spore suspension in 0.1% peptone water asthe diluent to obtain the desired inoculum�s density of4×102 cells/ml (Ellis et al., 1991).

Extraction of aflatoxins. The broth filtrates weremixed with an equal volume of chloroform in a se-parating funnel. The residue was re-extracted twicefor complete extraction. The chloroform extract wasdefatted with hexane, concentrated in a rotary evapo-rator and purified using silica gel column chroma-tography. The column was washed with 3 ml eachof hexane, ethyl ether and methylene chloride. Theaflatoxins were then eluted with 6 ml chloroform:acetone (9:1, v/v) mixture. The solvent was removedby evaporation on a rotary evaporator. The residueswere reconstituted in 1 ml methanol for further chro-matographic analyses.

Analysis of aflatoxins derivative formation. 50 µlof trifluoroacetic acid (TFA) were added to 200 ml ofthe methanol solution of toxin extract. The mixturewas allowed to react at room temperature for 15 minand then was evaporated to dryness. The residue wasdissolved in 2 ml of water:acetonitrile (75:25, v/v) forHPLC (Frisvad and Thrane, 1993).

Determination of the fungus ability to grow onaflatoxin B1. Different fungal strains were examinedfor their ability to grow on mineral medium amendedwith three different concentrations of aflatoxin B1(100, 250 and 500 ppb). The broth mineral mediumas given by Atlas (1995) was used.

217Bioremediation of aflatoxins3

Inhibition of aflatoxin production by toxigenicstrain when co-cultivated with other fungal strains.Tested fungal strains were co-cultivated with anaflatoxigenic producing A. parasiticus strain on MEAbroth medium. Five Erlenmeyer flask of 250-ml ca-pacity were provided with 50 ml of the medium andinoculated with a spore suspension of A. parasiticusas well as the tested strain. The pH of the mediumwas adjusted to 6.5 with 0.1 N NaOH. The flasks wereincubated for two weeks at 25°C. The growth of thefungal strains as well as the aflatoxin production wasestimated by HPLC.

Separation of the degradation products. Thedegradation products of the aflatoxin bioremediationwere analyzed using SHIMADZU GC/MS-QP 5050Agas chromatograph-mass spectrometer using CLASS5000 software. WILEY Mass Spectral Database wasused in the identification of the separated peaks.

Results

Preliminary detection of aflatoxin-producingfungi. Four strains of A. flavus, two strains of A. para-siticus and one strain of A. flavus var. columnarisexhibited blue green or bluish-green fluorescence sur-rounding the colonies on the agar medium indicatingthe possible production of aflatoxins. While A. oryzae,A. tamarii and A. flavus RCMB 002002(6) were notshowing any fluorescence.

A confirmative test for aflatoxin production bythese strains was performed by TLC analysis. Fourstrains of A. flavus and the two strains of A. parasi-ticus and the strain of A. flavus var. columnaris thatshowed positive fluorescence were found to produceone or more spots on the TLC plate. Two strains ofA. parasiticus produced AFB1, AFB2 as well as AFG1and AFG2. Also, a marked increase in aflatoxin con-centrations in A. parasiticus, strain RCMB 002001(2)culture was observed (Table I). Thus this strain wasused for the further studies as aflatoxin-producingstrain. The chromatographic analysis of A. flavusRCMB 002002(6) indicated that this strain producedno or undetectable quantities of aflatoxins. Conse-quently, it was considered a non-aflatoxigenic fungus.

Investigation of the ability of selected fungalstrains to grow on medium emended with differentconcentrations of aflatoxin. Twenty three strainswere investigated for their ability to grow on mineralmedium containing different concentrations of afla-toxin B1 (Table II). Results indicate that several fungalspecies including P. griseofulvum, P. urticae, P. lilaci-nus, T. viride, C. utilis, S. cerevisiae as well as a non-toxigenic strain of A. flavus were able to grow ona medium containing different concentrations of afla-toxin B1; 500 ppb, 250 ppb and 100 ppb. While, P. ro-

quefortii, R. nigricans (= Rhizopus stolonifer), S. ra-cemosum, P. variotii, C. clavata, and A. fumigatus wereless tolerant to aflatoxin B1 and showed weak growth.Alternatively, two fungal species; P. notatum andR. solani were highly sensitive to the aflatoxin B1.A. rutilum showed absence of growth on the mediacontaining aflatoxin B1. Whereas, P. notatum, R. solanishowed only a weak growth on the lowest (100 ppb)concentration.

0 = not detected under the experimental conditions.

A. flavus RCMB 002002 (1) + + 0 0

A. flavus RCMB 002002 (2) + + 0 0

A. flavus RCMB 002002 (6) 0 0 0 0A. flavus RCMB 002002 (5) 0 + 0 0

A. parasiticus RCMB 002001 (1) + + + +

A. parasiticus RCMB 002001 (2) + + + +A. flavus var.columnaris

RCMB 002003 + 0 + 0

Table IAflatoxin production by different strains of the A. flavus group

StrainsAflatoxin

B1

B2

G1

G2

P. griseofulvum RCMB001007(2) ++ ++ ++ ++P. roquefortii RCMB001009(6) ++ + + 0

P. urticae RCMB001017 ++ ++ ++ ++

P. nigricans RCMB001013 ++ + + +P. notatum RCMB001016 ++ + 0 0

P. italicum RCMB001018 ++ ++ + +

P. citrinum RCMB001011 ++ + + +R. nigricans RCMB001014 ++ + + 0

Circinella sp. RCMB013001 ++ ++ ++ +

S. racemosum RCMB016001 ++ + + 0M. rouxii RCMB015002 ++ ++ + +

A. rutilum RCMB020002 ++ 0 0 0

P. variotii RCMB018003 ++ + + 0P. lilacinus RCMB018002 ++ ++ ++ ++

C. lunata RCMB019002 ++ + + +

C. clavata RCMB019003 ++ + + 0T. viride RCMB017002 ++ ++ ++ ++

T. lagena RCMB030001 ++ + + 0

C. utilis RCMB005002 ++ ++ ++ ++S. cerevisiae RCMB006001 ++ ++ ++ ++

R. solani RCMB031001 ++ + 0 0

A. fumigatus RCMB002008(2) ++ + + 0A. flavus(atoxigenic strain)

RCMB002002(6) ++ ++ ++ ++

Table IIInvestigation of the ability of fungal strains to grow

on aflatoxin B1 containing media

Organism name

Aflatoxin B1

concentration (ppb)

Control 100 250 500

0 = not detected underthe experimental conditions,+ = detectable growth, ++ = good growth.


Influence of antagonistic activities of A. parasi-ticus against some other fungal strains. Cultivationof A. parasiticus in mixed cultures with other fungalstrains demonstrated inhibition of the aflatoxins pro-duction. The growth of A. parasiticus was inhibited bythe presence of these fungal strains. T. viride was foundto be capable of inhibiting the growth of A. parasiticus.At thenhighest level. This was followed by P. chryso-genum, P. lilacinus and P. urticae. The aflatoxin con-centrations were decreased with all strains examined(Table III) and their aflatoxin inhibition percentagesare given hereafter in-between parenthesis.

The concentration of aflatoxin B1 was found to de-crease reaching a minimum value of 5.32 ppb (75.2%)by T. viride compared with 21.5ppb of the control.This is followed by 7.8 ppb (63.7%), 7.81 ppb (63.6%),8.1 ppb (62.3%), 8.2 ppb (61.8%), 9.83 ppb (54.3%)and 10 ppb (53.5%) for S. cerevisiae, P. griseoful-vum, C. utilis, P. lilacinus, M. rouxii and P. urticae,respectively.

The concentration of aflatoxin B2 was found to bedecreased and reaching a minimum value of 3.1 ppb(56.2%) by T. viride compared with 9.6 ppb of thecontrol. This is followed by 4.7 ppb (51.0%), 4.8 ppb(50.0%), 5.2 ppb (45.8%), 5.9 ppb (38.5%) and 6.1 ppb(36.4%), for S. cerevisiae, C. utilis, P. griseofulvum,P. lilacinus and P. urticae respectively.

The concentration of aflatoxin G1 was also de-creased reaching a minimum value of 3.1 ppb (65.1%)by T. viride compared with 8.9 ppb of the control.This is followed by 4.1 ppb (53.9%), 4.9 ppb (44.9%),5.1 ppb (42.6%), 5.4 ppb (39.3%), 5.8 ppb (34.8%),

and 6.2 ppb (30.3%) for P. lilacinus, S. cerevisiae,C. utilis, P. griseofulvum, P. urticae, and R. nigricans,respectively.

The concentration of aflatoxin G2 decreased aswell reaching a minimum value of 2.2 ppb (57.7%)by T. viride compared with 5.2 ppb of the control. Thetotal aflatoxins production were found to decreasereaching a minimum concentration of 14.8 ppb(67.2%) with T. viride when compared with 48.2 ppbof the control. This is followed by P. lilacinus, P. gri-seofulvum, S. cerevisiae, C. utilis, P. urticae, R. nigri-cans and M. rouxii, with total aflatoxin concentrationsof 20.8 ppb (53.9%), 21.5 ppb (52.4%), 21.3 ppb(52%), 21.8 ppb (51.7%), 25.3 ppb (44.0%), 27.9 ppb(38.2%) and then 29.2 ppb (35.4%), respectively.

Compounds produced after bioremediation ofaflatoxins by Mucor rouxii. M. rouxii has a uniquebehavior in the bioremediation of aflatoxins. In spiteGC analysis showed four characterized peaks (Fig. 1).The four main peaks have mass peaks; 189, 217, 216and 228 with molecular formula of C12H20N2O2,C10H18O, C22H46 and C11H18O2, respectively. The pres-ence of furan moiety in B and E as well as F indicatesthe degradation of aflatoxin molecules. The presenceof ketone molecule (D) indicated the degradation oftoxins by the fungus. Obviously, the other peaks werefragments of each principal peak. The aspergillic acid,which is a by-product of A. parasiticus shown in thecontrol, was also observed in the chromatogram.

Compounds produced after bioremediation ofaflatoxins by Rhizopus nigricans. The detected cyclo-pentane represents the cyclopentenone ring of aflatoxin

Control (A. parasiticus) 21.5 21.5 9.6 8.9 45.2

P. griseofulvum 7.81 5.2 5.4 3.1 21.5 63.6 45.8 39.3 40.3 52.4

P. roquefortii 12.8 7.4 6.9 4 31.1 40.3 22.9 22.4 23.0 31.2

P. urticae 10.0 6.1 5.8 3.4 25.3 53.5 36.4 34.8 34.6 44.0

P. nigricans 14.5 7.6 7.1 3.8 33 32.5 20.8 20.2 26.9 27.0

P. notatum 11.9 7.2 6.8 4.1 30 44.6 25.0 23.6 21.1 33.6

P. italicum 12.6 8.1 6.7 4.2 31.6 41.4 15.6 24.7 19.2 30.0

R. nigricans 10.8 6.9 6.2 4 27.9 49.7 28.1 30.3 23.0 38.2

S. racemosum 13.9 8.1 6.8 4.2 33.0 35.3 15.6 23.6 19.2 27.0

Cirinella sp. 15.2 8.3 7.2 4.1 34.8 29.3 13.5 19.1 21.1 23.0

P. lilacinus 8.2 5.9 4.1 2.6 20.8 61.8 38.5 53.9 50.0 53.9

C. lunata 13.1 7.8 7.8 4.7 33.4 39.0 18.7 12.3 9.6 26.1

T. viride 5.32 4.2 3.1 2.2 14.8 75.2 56.2 65.1 57.7 67.2

C. utilis 8.1 4.8 5.1 3.8 21.8 62.3 50.0 42.6 26.9 51.7

S. cerevisiae 7.8 4.7 4.9 3.9 21.3 63.7 51.0 44.9 25.0 52.0

M. rouxii 9.83 6.9 8.2 4.3 29.2 54.3 28.1 7.8 17.3 35.4

Table IIIInhibition of aflatoxin production by a toxigenic strain of A. parasiticus cultivated with certain fungal isolates

Organism nameAflatoxin production (µg/l) Inhibition (%)

G2

TotalB1

B2

G1

TotalB1

B2

G1

G2


B group after bioremediation by the R. nigricans.While the detected benzene and cyclopentane moietyindicates the cleavage of aflatoxin structure, the de-tection of furan moieties as well as dioctyl phthalateconfirming this degradation of aflatoxins. The chro-matogram (Fig. 1) also indicated the presence of se-

veral fatty acids such as oleic, palmitic, palmitolenic,linolelaidic and high amount of linoleic acid. Someterpines such as alpha-terpinol, farnesene, jasmoneand linalol were also observed.

Compounds produced after bioremediation ofaflatoxins by Penicillium griseofulvum. The gas

5 10 15 20 25 30minutes

Mucor rouxii

B

A C

D

E

1

2

3 45

AB C

7

6 8

9

D

5 10 15 20 25 30 35 40 45minutes

Rhizopus nigricans

5 10 15 20 25 30minutes

Penicillium griseofulvum

5 10 15 20 25 30 35minutes

Penicillum uricae

F

DC

AB

E

D

C

B

AGE

Fig. 1. Total ion chromatogram of GC/MS analysis of the compounds produced after remediation of aflatoxins by indicated fungi.Peaks for particular fungi are as follow:

Mucor rouxii: (A) aspergillic acid; (B) furan � 4,5diethyl-2,3-dihydro-2,3-dimethyl; (C) 2-docosane; (D) ketone � 2,2-dimethylcyclohexyl methyl;(E) mannofuranoside; (F) bi-furan, dicarboxylic acid dimethyl

Rhizopus nigricans: (1) linalool; (2) alpha-terpineol; (3) cis-jasmone; (4) farnesene; (5) palmitoleic acid; (6) palmitic acid; (7) oleic acid;(8) linoleic acid; (9) linolelaidic acid; (A) cyclodecafuranone; (B) benzyl benzoate; (C) cyclopentane � isopropenyl-2,3-dimethyl (D)

benzenedicarboxylic ethylhexyl esterPenicillum griseofulvum: (A) aspergillic acid; (B) oxyaspergillic acid; (C): furanone; (D) dioctyl phathalate; (E) ethanone, dihydrodimethyl-indol;

(F) benzofuran, cyclohexyl-dihydro-methyl; (G) griseofulvin.Penicillum uricae: (A) butyl hydroxy toluene; (B) aspergillic acid; (C) naphthofurandihydropropenyl; (D) naphthalene; (E) mannofuranoside.


chromatographic analysis of compounds producedafter the bioremediation process of aflatoxins by theP. griseofulvum was completely different of that of thecontrol. The position of peaks is shifted indicatingbioremediation of aflatoxins. Chromatogram showsfive characteristic peaks (A-D, F) with several peakfragments. The mass spectrum exhibited 2,6-di-tert-

butylphenol with a methyl group at the position 4, thisbutyl phenol absolutely not included in the chemicalstructure of the control, i.e. aflatoxins B1, B2, G1 andG2. The molecular weight of this compound is 220,with molecular formula C15H24O. On the other hand,aspergillic acid and oxyaspergillic acid (C12H20N2O2)were developed, respectively. The mass peaks are 248

5 10 15 20 25 30minutes

Paecilomyces lilacinus

5 10 15 20 25 30 35minutes

Trichoderma viride

A

B

C

D

1

2

4

A

BC

D

E

3

5

Fig. 2. Total ion chromatogram of GC/MS analysis of the compounds produced after bioremediation of aflatoxins by indicated fungi.

The peaks for particular fungi are as follow:Paecilomyces lilacinus: (A) phenol-bis � (1,1-dimethyl)-4-methyl; (B) methyl dimethoxyphenyl propanoate; (C) dioctyl phthalate;

(D): hexanone.Trichoderma viride: (1) aspergillic acid; (2) cyclopentanetione; (3) butabarbitol; (4) methyl jasmonate; (5) dioctyl phthalate;

(A) tinuvin; (B) limonene; (C): benzofuranone, hexahydrotrimethyl; (D) benzene; (E) androstanedione.Candida utilis: (A) aspergillic acid; (B) benzofuran; (C) tinuvin; (D) dioctyl phthalate.Saccharomyces cerevisiae: (A) aspergillic acid (B) furan; (C): Dimethyl-naphthalene; (D) 7-methoxy-coumarin; (E) dioctyl phthalate.

5 10 15 20 25 30minutes

Saccharomyces cerevisiae

5 10 15 20 25 30minutes

Candida utilis

B

C

E

D

B

C

A

D

A


and 416, respectively. The furan moiety was cleavedsince it appeared as furanon or within spiro structureas benzofuran. Presumably, the appearance of dioctylphthalate and the dihydro-dimethyl indole indicate thedegradation of aflatoxins.

Compounds produced after bioremediation ofaflatoxins by Penicillium urticae. The GC analysisof the compounds produced after the treatment ofaflatoxins by P. urticae is shown in Figure 1. Thereare several peaks present, but the most typical peaksare C and E. The last peak represents a manno-furanoside of molecular formula C11H18O2. The totalmolecular weight is approximately 1210 for bioreme-diated aflatoxins in contrast to 1292 of the controlsample. It seems that the P. urticae do not bioreme-diate aflatoxins very well as indicated from the num-ber of hydrogen atoms. The mass peak of the value of234 of naphthalene with phenyl- and methyl- sidechain as well as the butyl hydroxyl toluene also indi-cate this weak tendency to cleave aflatoxin molecule.

Compounds produced after bioremediation ofaflatoxins by Paecilomyces lilacinus. The gas chro-matographic analysis of the compounds produced afterbioremediation by P. lilacinus, indicate the presenceof four main characteristic peaks (Fig. 2). Meanwhilethe positions of these peaks are completely differentfrom peaks of control sample. The main difference isthat the compounds of P. lilacinus have a total mo-lecular weight of 1116 and total carbon atoms 66.However, the identification of the peaks in the massspectra shows the dominance of benzene rings withdifferent side chains. The presence of dioctyl phtha-late on one side and the absence of furan moiety aswell as coumarin and cyclopentanone on the otherhand may indicate the biodegradation of aflatoxins.

Compounds produced after bioremediation ofaflatoxins by Trichoderma viride. T. viride bio-remediated the aflatoxins B and G very extensively;10 peaks in addition to peak fragments as well as verysmall peaks as shown in GC analysis were detected(Fig. 2). The positions of these peaks were completelydifferent from peaks in control sample. Aspergillicacid as well as different degradation product exertedsuch as the dioctylphthalate, methyl jasmonate, buta-barbitol and cyclopentanione were detected. Cyclo-pentanetione indicate cleavage of cyclopentane ringof aflatoxins. The total molecular weight of the bio-remediated aflatoxins was high (1348) in contrast tothe control sample (1292), that was attributed to highlyfragmented aflatoxins by T. viride. Consequently, car-bon, hydrogen and oxygen atoms recorded 89, 136and 9, respectively. Identification of the peaks by massspectrometer showed the presence of benzene fusedwith furan moiety (C) as a dominant bioremediationproduct by T. viride. Limonene, jasmonate and otheressential oil compounds as well as benzene, 3-methyl-

butenyl- (with molecular weight 146) and tinuvin werealso detected. Furan moiety was detected in two peaks;weight peak 170 (R. time 30.64�31.25) for benzo-furanon and mass peak 221 (R. time 53.49�54.65)with androsanedione.

Compounds produced after bioremediation ofaflatoxins by Candida utilis. The chromatogram forbioremediation of aflatoxins by C. utilis shows thepresence of aspergillic acid, a metabolic product ofA. parasiticus, which is present in the control sample.The furan moiety was still present but cleaved from theaflatoxin structure since it appeared as a benzofuranderivative. Tinuvin was present as well. Dioctyl phtha-late as a degradation product was detected indicatingthe biodegradation process of the aflatoxins (Fig. 2).

Compounds produced after bioremediation ofaflatoxins by Saccharomyces cerevisiae. The GCanalysis of the compounds produced after the treatmentof aflatoxins by S. cerevisiae is given in Figure 2.Although, there are several peaks present, the mostimportant one is the methoxycoumarin. The presenceof furan moiety that appeared as furan, 4, 5-diethyl-2,3-dimethyl as well as the methoxycoumarin indicatethe cleavage of aflatoxins ring. Another peak pro-duced in figure could be also regarded as a benzenering degradation products identified as dimethyl-naphthalene, since its molecular weight 156 and mo-lecular formula C12H12. Also, dioctyl phthalate was de-tected. The chromatogram also shows the presence ofaspergillic acid, a metabolic product of A. parasiticusthat is present in the control sample.

Discussion

Biological detoxification or the biotransformationof aflatoxins by microbial systems to a metabolite(s)that is either nontoxic when ingested by animals orless toxic than the original toxin can be termedas bioremediation. The goal of bioremediation is todegrade organic pollutants to concentrations that areeither undetectable or, if detectable, to concentrationsbelow the accepted limits (McKane and Kandel, 1996).

The growth of certain fungal strains (P. griseo-fulvum, P. urticae, P. lilacinus, T. viride, C. utilis,S. cerevisiae as well as a non-toxigenic strain ofA. flavus) on the three aflatoxin B1 concentrations used(100, 250 and 500 ppb) indicated that these fungalstrains have the ability to tolerate or metabolize thetoxin. Other examined fungal strains showed variablecapabilities to tolerate aflatoxin B1 reflected in the dif-ference in their growth ability at the aflatoxin con-centrations used (250, and/or 100 ppb) with no growthat 500 ppb. T. viride, P. lilacinus, P. griseofulvum,S. cerevisiae, C. utilis, P. urticae and M. rouxii wereable to inhibit the growth of A. parasiticus and grew


in the presence of aflatoxins in its growth medium.The organism with the least efficiency to inhibitaflatoxins was Circinella sp. having a total inhibitionpercentage of 23%.

The reasons for the reduction in aflatoxin levelscan, therefore, be attributed to one or a combinationof the following factors: (1) physical competition forspace and nutrition; (2) test fungi may compete withA. parasiticus for a substrate required for toxin pro-duction but not the growth; (3) presence of test fungimight cause a change in the biochemical environmentdeciding, thereby, the metabolic pathway available tothe toxigenic fungi and (4) degradation of aflatoxinfollowing its formation. An inhibition in aflatoxinproduction by T. viride by more than 90% inhibitionwas also reported by Varma (1996). Shantha (1999)reported that several fungal cultures were found toprevent the synthesis of aflatoxin B1 by A. flavus inliquid medium. Among these Phoma spp., Mucor sp.,Trichoderma spp., Rhizopus sp., Alternaria sp. andSporotrichum spp. inhibited aflatoxin synthesis byabout 90% or more.

El-Sayed (1996) revealed the potential use of so-me phycomycetes (Absidia, Mucor, Cunninghamella,Rhizopus and Syncephalasrtum) to inhibit aflatoxins.Cole et al. (1972) and Nout (1989) studied a numberof Rhizopus species and indicated the accumulationof two fluorescent metabolites of aflatoxin B1 duringits degradation. These metabolites were identifiedas hydroxylated stereo isomers derived from thereduction of ketone function on the cyclopentanering of aflatoxin B1. Weckbach and Marth (1977)and Choudhary (1992) also found that Rhizopusnigricans inhibited both the growth and aflatoxin pro-duction by A. parasiticus.

Nour et al. (1982) reported that some species ofAspergillus, Mucor, Penicillium, Rhizopus genera areantagonizing fungi that seem capable of metabolizingaflatoxin B1 produced by A. flavus or probably pro-ducing some exudates that react with the toxin, trans-forming it into nontoxic compounds, degrading it ordeflecting the pathway of aflatoxin B1 synthesis. Theidentification of degradation products indicated thatthe aflatoxins were partly degraded with the exam-ined fungal strains at different manner. The presenceof furan moiety in the chromatograms of separationof most of the examined strains indicated degradationof aflatoxin. Three different furan moieties were ex-erted in case of M. rouxii.

The separation profile of R. nigricans exhibitedhigh efficacy to fatty acid formation, with high percent-age of linoleic acid. This result is in accordance withthat of Kim et al. (2000) where linoleic acid was iden-tified by GC-MS to be the main active component inaflatoxin degradation by soybean paste. The presenceof the fatty acids and subsequent appearance of the

degraded furan moiety proved this conclusion. How-ever, in case of P. griseofulvum, two forms of furanmoieties were appeared as degradation products aswell as the dioctyl phthalate. The GC chromatogramalso separated the antifungal metabolite of P. griseo-fulvum, the griseofulvin. The presence of aspergillicacid and its oxygenated form, oxyaspergillic acid, alsoconfirm the oxidation process that cleavage hydrogenatoms and lead to libration of H2O and CO2. WhileP. urticae exerted two forms of furan moieties, C. utilisand P. lilacinus, show lower efficiency in the de-gradation of aflatoxins. The T. viride strains showdifferent furan moieties as well as androstanedione,a non-active molecule that have a similar structureof aflatoxin with the active bonds. The presence ofbenzene rings along with the dioctyl phthalate alsoconfirms the degradation process. In case of S. cere-visiae, the mass analysis of the separated peaks indi-cates that degradation takes place in furan moiety aswell as the coumarin moiety. These moieties representthe main skeleton of aflatoxins.

The available literature indicates only five stra-tegies are known to reduce or detoxify aflatoxinscontamination in food, food and feed processing,biocontrol and microbial inactivation, dietary modifi-cation and chemoprotection, chemical degradation,and reduction in toxin bioavailability by selectivechemosorption (Smith et al., 1994). The presented re-sults may certainly suggest a sixth strategy; applica-tion of fungal biodegradation and/or bioremediation.It is worth mentioning that the genomic mechanismcontrolling aflatoxin biosynthesis by the same researchgroup is in progress and will be released shortly.T. viride, a classical fungal biocontrol agent that wasused in the treatment of infected plants either in sprayform or other formulation. Thus its utility can beexpanded to include aflatoxin bioremediation in theinfected plant, for being the most potent organism ca-pable of decreasing aflatoxin concentration.

Hence, this study emphasizes the new value ofS. cerevisiae used commercially as a rich sourceof protein and vitamins, which is its possible usage asa bioremedy for aflatoxins in a human being that can-not be treated with other treatments. Being used asa treatment for aflatoxins it will also provide the hu-man body with vitamins and proteins.

Literature

Atlas R.M. 1995. Bioremediation. Chemical and EngineeringNews. 73: 32�42.Bhatnagar D., E.B. Lillehoj and D.K. Arrora. 1992. Handbookof Applied Mycology. Vol. 5: Mycotoxins in Ecological Systems.Marcel Dekker, Inc., New York, USA.Boller R.A. and H.W. Schroeder. 1973. Influence of Aspergilluschevalieri on production of aflatoxin in rice by Aspergillus para-siticus. Phytopathology 63: 1507�1510.


Boller R.A. and H.W. Schroeder. 1974. Influence of Aspergilluscandidus on production of aflatoxin in rice by Aspergillus para-siticus. Phytopathology 64: 121�123.Buchi G. and I.D. Rae. 1969. The structure and chemistry of theaflatoxins. In: L.A. Goldblatt (ed.), Aflatoxins. pp. 55�75. Aca-demic Press, New York.Choudhary A.K. 1992. Influence of microbial co-inhabitants onaflatoxin synthesis of Aspergillus flavus on maize kernels. Lett.Appl. Microbiol. 14: 143�147.Christensen C.M., G.H. Nelson., G.M. Speers and C.H.Mirocha. 1973. Results of feeding tests with rations containinggrain invaded by a mixture of naturally present fungi plus Asper-gillus flavus. Feedstuffs 45: 20�41.Cole R.J., J.W. Kirksey and B.R. Blankenship. 1972. Conver-sion of aflatoxin B

1 to isomeric hydroxy compounds by Rhizopus

spp. J. Agric. Food Chem. 20: 1100�?Cotty P.J., P. Bayman, D.S. Egel and K.S. Elias. 1994. Agricul-ture, aflatoxins and Aspergillus. In: K. A. Powell, A. Fenwick andJ. F. Peberdy (eds), The Genus Aspergillus. pp. 1�27. PlenumPress, New York.Eaton D.L. and E.P. Gallagher. 1995. Mechanisms of aflatoxincarcinogenesis. Annu. Rev. Pharmacol. Toxicol. 34: 135�172.Ellis W.O., J.P. Smith, B.K. Simpson and J.H. Oldham. 1991.Aflatoxins in foods: occurrence, biosynthesis, effects on organismsand methods of detection and control. CRC Crit. Rev. Food Sci.30: 404�439.El-Sayed T. I. 1996. The potential of some phycomycetes to in-hibit aflatoxin production by Aspergillus flavus. Vol. II, p. 33�49.Proc. 1st International Conference on Fungi: Hopes and Challenges,Cairo. 2�5 Sept. 1996.Frisvad J.C. and U. Thrane. 1993. Liquid chromatography ofmycotoxins. In: V. Betina (ed.) Chromatography of Mycotoxins:Techniques and Applications. Elsevier Amsterdam.Hara S., D.I. Fennell and C.W. Hesseltine. 1974. Aflatoxin pro-ducing strains of Asergillus flavus detected by fluorescence of agarmedium under ultraviolet light. Appl. Microbiol. 27: 1118�1123.Kim J.G., Y.W. Lee, P.G. Kim, W.S. Roh and H. Shintani.2000. Reduction of aflatoxins by Korean soybean paste and itseffect on cytotoxicity and reproductive toxicity. I. Inhibition of

growth and aflatoxin production of Aspergillus parasiticus byKorean soybean paste (Doen-jang) and identification of theactive component. J. Food Prod. 63: 1295�1298.Kurtzman C.P., B.W. Horn and C.W. Hesseltine. 1987. Asper-gillus nomius, a new aflatoxin producing species related to Asper-gillus flavus and Aspergillus tamarii. Antonie van Leeuwenhoek53: 147�158.McKane L. and J. Kandel. 1996. Microbiology: Essentials andApplications (2nd ed.). McGraw Hill, Inc., New York.Nour M.A., B. Zienal and E. El-Sohili. 1982. Effect of storagecondition of fungal flora and aflatoxin accumulation in mixedpoultry feed. International Symposium on Mycotoxins 551�565.Nout M J.R. 1989. Effect of Rhizopus and Neurospora spp. ongrowth of Aspergillus flavus and A. parasiticus and accumulationof aflatoxin in groundnut. Mycol. Res. 93: 518�523.Park D.L., C.E. Ayla, S.E. Guzman-Perez, R. Lopez-Garciaand S. Trujillo. 2001. Microbial toxins in foods: algal, fungaland bacterial. In: W. Helferich and C. K. Winter (eds), Food Toxi-cology. pp.106�108. CRC Press, Boca Raton, Florida, USA.Paterson R.R. and P.D. Bridge. 1994. Biochemical Techniquesfor Filamentous Fungi. CAB International, Wallingford, UK.Shantha T. 1999. Fungal degradation of aflatoxin B

1. Nat. Toxins

7: 175�178.Smith J.E. 1997. Aflatoxins. In: J. P. Felix D�Mello (ed.), Hand-book of Plant and Fungal Ttoxicants. pp. 269�287. CRC Press,New York.Smith J.E., C.W. Lewis, J.G. Anderson and G. L. Solomons.1994. Mycotoxins in Human Nutrition and Health. EUR ReportNo. 16048. Luxembourg, European Commission.Smith J.E. and J. Bol. 1989. Biological detoxification of afla-toxin. Food Biotechnol. 3: 127�144.Smith J.E. and M.D. Moss. 1985. Mycotoxins: Formation, Analy-ses andSsignificance. John Wiley and Sons, New York.Varma C.S. 1996. Influence of some fungal species on aflatoxinproduction by toxigenic strain of Aspergillus parasiticus on natu-ral substrate. Indian Phytopathol. 49: 275�277.Weckbach L.S. and E.H. Marth. 1977. Aflatoxin production byAspergillus parasiticus in a competitive environment. Mycopatho-logia 62: 39�45.


Polish Journal of Microbiology, founded in 1953 as Acta Microbiologica Polonica, is a broad-based interna-tional microbiological journal covering: general microbiology, bacterial physiology, molecular biology andgenetics, virology, clinical bacteriology, environmental microbiology and biotechnology.

Polish Journal of Microbiology publishes mainly original research articles including short communica-tions, but also minireviews, book reviews and press releases. All the papers should be prepared accordingto �Instruction to Authors� which is included in the first issue of every volume and also available atwww.microbiology.pl/pjm. The papers appear usually within 3 months after final acceptance.

Abstracts and full texts of the papers (as pdf-s) are available under www.microbiology.pl/pjm directly after theprinted version appears.

Polish Journal of Microbiology is currently covered by Thomson Master List and is indexed in BIOSIS/Biolo-gical Abstracts, EMBASE/Excerpta Medica, Index Medicus/MEDLINE, Cambridge Scientific Abstracts,SCOPUS, Index Copernicus

Submissions of manuscripts in various areas of microbiology are very welcome.

After acceptance of the paper the copyrights are reserved to the Publisher. No part of the publication may bereproduced in any form without prior written permission of the Publisher, Polish Society of Microbiologists.

Subscription information for Polish subscribers under [email protected], for foreign subscribers underwww.arspolona.com.pl or [email protected]

Advertising offer

The Editorial Office of the Polish Journal of Microbiology encourages our Readers to place advertisementsin our quarterly. Our journal reaches scientists working in various research and educational institutions andinvolved in different fields of microbiology.We ensure rapid appearance of the ad.

Price of black and white ad on inside cover: half of page = 150 PLN, full page = 300 PLNPrice of ad inside an issue: half page = 75 PLN, full page 150 PLN.

Please send file with graphical layout directly to the editorial officeof the journal or contact us at [email protected].

The Editorial Office and Publisher do not take any responsibility for the contents of advertisementand announcements.

ISSN - 1733-1331Index - 35119

polish journal of microbiology · polish journal of microbiology 2007, vol. 56, no 3, 151Œ152...

Documents