lable at ScienceDirect

Food Microbiology 27 (2010) 493e502

Contents lists avai

Food Microbiology

journal homepage: www.elsevier .com/locate/ fm

Synthesis of cyclopropane fatty acids in Lactobacillus helveticusand Lactobacillus sanfranciscensis and their cellular fatty acidschanges following short term acid and cold stresses

Chiara Montanari a, Sylvain L. Sado Kamdema,b,*, Diana I. Serrazanetti a,François-Xavier Etoa b, M. Elisabetta Guerzoni a

aDipartimento di Scienze degli Alimenti, Alma Mater Studiorum, Università degli Studi di Bologna, Viale Fanin, 46, 40127 Bologna, Italyb Laboratoire de Microbiologie, Department of Biochemistry, University of Yaounde I, P.O. Box 812 Yaounde, Cameroon

a r t i c l e i n f o

Article history:Received 26 August 2009Received in revised form10 December 2009Accepted 14 December 2009Available online 6 January 2010

Keywords:Lactobacillus helveticusLactobacillus sanfranciscensisCyclopropane fatty acidsCold stressAcid stress

* Corresponding author at: Dipartimento di ScienzStudiorum, Università degli Studi di Bologna, Viale FaTel.: þ39 23796011559.

E-mail addresses: sylvain.sado@unibo.it, sadosylvKamdem).

0740-0020/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.fm.2009.12.003

a b s t r a c t

An implemented GC method to separate and quantify the cell cyclopropane fatty acids lactobacillic(C19cyc11) and dehydrosterculic (C19cyc9) was used to study the adaptive response to sublethal acid andcold stresses in Lactobacillus helveticus and Lactobacillus sanfranciscensis. The comparison of the compo-sition of cellular fatty acids of the two strains and their changes after 2 h of stress exposure under micro-aerobic and anaerobic conditions indicated that the aerobic biosynthetic pathway for unsaturated fattyacids is prevalent in L. sanfranciscensis, while the anaerobic pathway is prevalent in L. helveticus. Indeed inthe latter strain, in the presence of a source of oleic acid and under micro-aerobic conditions, C18:1n11 andits post-synthetic derivative C19cyc11 accounted for overall proportion ranging from 52 to 28% of the totalFAs. On the other hand L. sanfranciscensis synthesizes by aerobic pathway C18:1n9 and transforms it toC19cyc9. However in this species the cumulative level of these two FAs did not exceed 30%. The relevantproportion of dodecanoic acid in the latter species suggests that carbon chain shortening is the principalstrategy of L. sanfranciscensis to modulate fluidity or chemico-physical properties of the membranes.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Bacterial cytoplasmic membranes can be considered at the sametime as the Achilles heel and the strong point for the microbialsurvival in the growth environment. In fact, one of the most impor-tant adaptivemicrobial response to stress exposure, in addition to thesynthesis of specific proteins (Wouters et al., 1999, 2001), is relatedto changes in membrane fatty acid (FA) composition. The adaptivestrategies include alterations in saturation degree, carbon chainlength, branching position, cis/trans isomerisation and conversion ofunsaturated fatty acids (UFAs) into cyclopropanes (Kaneda, 1977;Russell, 1984; Heipieper et al., 1992; Loffeld and Keweloh, 1996). Inparticular, the ratio of unsaturated vs saturated fatty acids (USFAsratio) of the membrane lipids has been observed to vary both withthe growth temperature andwith the source of carbon and energy for

e degli Alimenti, Alma Maternin, 46, 40127 Bologna, Italy.

ain@hotmail.com (S.L. Sado

All rights reserved.

growth. The changes relative to the unsaturation level are accom-plished either by the oxygen independent synthesis of UFAs (theanaerobic mechanism) or by the oxygen dependent desaturation ofsaturated fatty acids (SFAs) (the aerobic mechanism) (Russell, 1984).In the first case vaccenic acid (cis-11-octadecenoic acid, C18:1n11) isformed by palmitoleic acid (cis-9-hexadecenoic acid, C16:1n9)(Keweloh and Heipieper,1996). In the second case the double bond isintroduced into the aliphatic acyl chains of SFAs, already integrated inthe phospholipids, by specific desaturases (Keweloh and Heipieper,1996). The so called anaerobicmechanism has been found not only inanaerobic bacteria (Clostridium, Lactobacillus) but also in aerobicones while the aerobic mechanism has been observed in species ofBacillus,Mycobacterium,Micrococcus (Magnuson et al.,1993). There issome evidence that bothmechanisms coexist in Pseudomonas, Vilniusand Acinetobacter (Härtig et al., 1999) and in Lactobacillus helveticus(Guerzoni et al., 2001).

Some specific FAs have a peculiar role in the bacterial growth orsurvival. In particular, Lactobacillus spp. are well known to needoleic acid (cis-9-octadecenoic acid, C18:1n9) for their growth(Johnsson et al., 1995). However it has been shown, at least inLactobacillus delbrueckii, that only some strains require exogenousFAs for their growth (Partanen et al., 2001).

C. Montanari et al. / Food Microbiology 27 (2010) 493e502494

Lactic acid bacteria (LAB), used as important starters in fermenta-tion of plants and animal origin foods, are exposed during theirproliferation and preservation, as well as during food fermentation, tounfavourable conditions such as oxidative, heat, cold, acid andosmoticstresses. Some specific FAs are reported to play an important rolein stress responses: the proportions of palmitic acid (hexadecanoicacid, C16:0) and linoleic acid (cis-9,cis-12-octadecadienoic acid, C18:2)increase in Lactobacillus acidophilus grown at low temperature (Fer-nandez Murga et al., 2000). The C18:1n9 concentration increases inresponse to low temperature in Lactobacillus plantarum (Russell et al.,1995), to acid pH in Streptococcus termophilus (Béal et al., 2001) and toosmotic stress in Lactococcus lactis (Guillot et al., 2000). On thecontrary C18:1n9 concentration decreases in response to freezing inlactic streptococci (Gilliland and Speck, 1974) and to spray drying inL. acidophilus (Brennan et al., 1986).

Also cyclopropane fatty acids (CFAs), lactobacillic acid (11,12-methyleneoctadecanoic acid, C19cyc11) and dehydrosterculic acid(9,10-methyleneoctadecanoic acid, C19cyc9), are regarded as key FAsin stress tolerance of LAB. These unusual FAs are formed “in situ”witha post-synthetic modification, by means of a transfer of a methylgroup from S-adenosyl-L-methionine (SAM) to a double bondof UFAsalready integrated into a phospholipid molecule (Law, 1971; Huanget al., 2002; Zhao, 2002). It has been reported that CFAs favour thestress tolerance of Lactobacillus bulgaricus, L. helveticus and L. acid-ophilus (Gómez Zavaglia et al., 2000), as well as the response to acid,osmotic and ethanol stresses (Guillot et al., 2000; Béal et al., 2001;Teixeira et al., 2002). However the two CFAs are generally not wellseparated by the gas-chromatographic conditions commonly adop-ted and are reported as cumulative values in the current literature.Since C19cyc11 and C19cyc9 are derived from different precursors,C18:1n11 and C18:1n9, synthesized respectively by the anaerobicand aerobic pathways, the uncertain assessment of their absolute andrelative extent does not allow a clear interpretation of the meaningand importance of their individual role in the stress response.

The principal aims of this paper were: i) the improvement ofa specific gas-chromatographic method to separate and quantifythe two CFAs; ii) the use of this implemented method to study theFAs composition and its modification in response to acid andcold stresses in Lactobacillus sanfranciscensis and L. helveticus.L. sanfranciscensis, whose exclusive habitat is wheat and othercereals sour dough (Ehrmann and Vogel, 2005), is predominantduring the continuous dough propagation under acid stress (Gänzleand Ehrmann, 1998; De Angelis et al., 2001). The cellular FAscomposition of the latter specie has not been previously studiedaccording to our knowledge. On the contrary, the FAs compositionof L. helveticus, which is usually propagated under selectiveconditions as natural whey starter during manufacturing of Swissand Italian hard and extra-hard cheeses (Di Cagno et al., 2006), hasbeenwidely investigated (Dionisi et al., 1999; Guerzoni et al., 2001).

In this paper the two species were exposed to the cold andacid sublethal stresses which are frequently associatedwith cell FAsadaptation and particularly cyclopropanation (Gómez Zavagliaet al., 2000; Béal et al., 2001; Zhao et al., 2009).

2. Materials and methods

2.1. Bacterial strains and growth conditions

L. helveticus CNBL 1156 from the collection of the Institute ofMicrobiology of the Catholic Sacred Heart University, Piacenza(Italy), stored at�80 �C, was pre-cultured for 24 h and subsequentlyinoculated and incubated for 24 h in MRS broth without Tween 80at 45 �C under anaerobic conditions. The MRS broth medium wasobtained by weighting single components as described by Oxoid(Unipath Ltd. Basingstoke, Hampsire, UK).

L. sanfranciscensis Bb12 from the collection of the Department ofFood Science of the University of Teramo (Italy), stored at �80 �C,was pre-cultured for 24 h and subsequently inoculated and incu-bated for 24 h in SDB medium (Kline and Sugihara, 1971) withoutTween 80 at 32 �C under anaerobic conditions. The pH of themedium was adjusted to 5.6 by addition of HCl 1 N.

2.2. Acid and cold stress exposure

The cells of both strains, obtained as above described, were usedfor the stress exposure experiment. L. helveticus culture was centri-fuged at 9000� g for 15min at 20 �C,while L. sanfranciscensis culture,due to its difficult sedimentation, was centrifuged at 12 000 � g for30 min at 20 �C. Cells were collected, inoculated (inoculum level8 log CFU/ml) in the same fresh medium and exposed for 2 h todifferent sublethal stresses, under static micro-aerobic (250 ml ofbroth in conical flasks of 500 ml of volume) or anaerobic conditions(250 ml of broth in conical flasks of 250 ml of volume, subsequentlyintroduced in anaerobic jar), in the presence or absence of Tween 80(0.3 g/l). Different pH values (5.0, 4.0, 3.0 for L. helveticus and 4.0, 3.6,3.0 for L. sanfranciscensis) were obtained by addition of lactic acid.Cold stress was applied incubating the cells at 10 �C, in comparisonwith the strains optimal growth temperatures (45 �C and 32 �Crespectively). Two independent experiments, with three repetitionsfor each experiment, were performed according to Table 1. After 2 hcell suspensions were centrifuged, as above described, and the cellswere harvested for FAs extraction.

Cell viability was determined by spread plating following serialdilutions, on MRS and SDB agar media (18 g/L agar concentration),for L. helveticus and L. sanfranciscensis respectively, immediatelyafter resuspension and at different time intervals, after 1, 2 or 3 h.Numbers of CFU were determined after 48 h incubation at therespective temperatures.

2.3. Development of separation and identification methodfor cyclopropane fatty acids (CFAs)

L. helveticus was pre-cultured for 24 h and subsequently inocu-lated and incubated for 24 h in MRS broth (250 ml) under anaerobicconditions at 45 �C. Tween 80 (0.3 g/l) was added to the culturein order to obtain cells rich in CFAs (Johnsson et al., 1995). The totalbroth (10 ml) containing the cells was resuspended in two flaskscontaining 250 ml of fresh medium and adjusted to obtain micro-aerobic and anaerobic conditions. The culture was centrifuged asabove described and cells were harvested for FAs extraction.

Since lipids of LAB are widely associated with the plasmamembrane (Kates, 1964) cell pellets were directly analysed. Cellularfatty acids extraction and methylation was performed using theMicrobial Identification System (MIS) protocol produced by Micro-bial ID (MIDI, Newark, DE, USA) also described in Welch (1991).

To evaluate the presence of CFAs hydrogenation and brominationwere performed using the method of Brian and Gardner (1968)modified as follows. After extraction 0.2ml of fatty acidmethyl esterswere evaporated with N2 and hydrogenated for 30 min in 10 ml ofchloroform:methanol (2:1 v/v, both from SigmaeAldrich, Milan,Italy) with 100 mg of 5% Pt on charcoal (Fluka, Milan, Italy). In thisselective procedure UFAs are converted to saturated ones, and CFAsare not affected. After paper filtration, sampleswere evaporatedwithN2, diluted in 1 ml of diethyl ether (reagent grade, SigmaeAldrich,Milan, Italy) and cooled to 0 �C. A 0.5-ml amount of bromine indiethyl ether (1:5 v/v, both from SigmaeAldrich, Milan, Italy) wasadded to each sample to break the cyclopropane ring. After 30min atroom temperature by stirring, ether and excess bromine wereevaporated at 50 �C with N2. The final sample was collected witha 0.1-ml amount of hexane and transferred to a GLC vial for injection.

Table 1Experimental design.

Strain Stress Growth conditions

Aerobic/anaerobic Tween80 pH Temperature (�C)

L. sanfranciscensis Control Micro-aerobic þ 5.60 32Micro-aerobic e 5.60 32Anaerobic þ 5.60 32Anaerobic e 5.60 32

Acid stress Micro-aerobic þ 4.00 32Micro-aerobic e 4.00 32Anaerobic þ 4.00 32Anaerobic e 4.00 32Micro-aerobic þ 3.60 32Micro-aerobic � 3.60 32Anaerobic þ 3.60 32Anaerobic � 3.60 32Micro-aerobic þ 3.00 32Micro-aerobic � 3.00 32Anaerobic þ 3.00 32Anaerobic � 3.00 32

Cold stress Micro-aerobic þ 5.60 10Micro-aerobic � 5.60 10Anaerobic þ 5.60 10Anaerobic � 5.60 10

L. helveticus Control Micro-aerobic þ 6.60 45Micro-aerobic e 6.60 45Anaerobic þ 6.60 45Anaerobic e 6.60 45

Acid stress Micro-aerobic þ 5.00 45Micro-aerobic e 5.00 45Anaerobic þ 5.00 45Anaerobic e 5.00 45Micro-aerobic þ 4.00 45Micro-aerobic e 4.00 45Anaerobic þ 4.00 45Anaerobic e 4.00 45Micro-aerobic þ 3.00 45Micro-aerobic e 3.00 45Anaerobic þ 3.00 45Anaerobic e 3.00 45

Cold stress Micro-aerobic þ 6.60 10Micro-aerobic e 6.60 10Anaerobic þ 6.60 10Anaerobic e 6.60 10

C. Montanari et al. / Food Microbiology 27 (2010) 493e502 495

2.4. Gas-chromatographic analysis

The GC analysis conditions, selected on the basis of differentexperiment, were the following. After extraction and/or bromina-tion fatty acid methyl ester analysis was performed both in SCANand SIM mode using an Agilent Technologies gas-chromatograph6890 N (Palo Alto, CA, US), equipped with an Agilent NetworkMass Selective detector HP 5973 (Palo Alto, CA, US) and a capillarycolumn SPB-5 30 m � 0.25 mm � 0.25 mm (Supelco Inc., Bellefonte,PA, US). The injector and the detector were both held at 250 �C. Thetemperature was programmed from 120 �C (held for 5 min) to215 �C at a rate of 3 �C/min, then from 215 �C to 225 �C at a rate of0.5 �C/min and the final temperature held for 5 min. The carriergas was helium with a rate of 1 ml/min and a split ratio of 1:10.Fatty acids were identified by comparing their retention time andmass fragmentation profiles with those of the standards mix FAME37 and BAME mix (SigmaeAldrich, Milan, Italy). Results wereexpressed as relative percentage of each FA. The USFAs ratio wasdetermined both without considering CFAs (USFAs ratio) andconsidering CFAs as UFAs (USFAs ratio cyc). This method was usedalso for the FAs analyses of the experiment of stress exposure bothof L. helveticus and L. sanfranciscensis.

2.5. Determination of free fatty acids (FFAs)

In order to assess the presence of cellular FFAs, cells obtained asabove described without stress exposure were used and extractionwas performed according to Patrignani et al. (2008).

2.6. Statistical analyses

The data presented are the mean of 2 independent experimentsand three replicates within each experiment. The coefficient ofvariability calculated as the ratio of the standard deviation ofsample estimates and the mean value of the sample estimates wasused to evaluate the data variability of the relative fatty acidspercentages with excel 2003.

3. Results

3.1. Separation and identification of dehydrosterculic acid(C19cyc9) and lactobacillic acid (C19cyc11)

The temperature programs for the GCeMS analysis described inMaterials and Methods was judged to be the best combination of

C. Montanari et al. / Food Microbiology 27 (2010) 493e502496

analysis time and good resolution of all FAs methyl esters peaks. Thedifferent chromatograms of the methylated FAs as well as the massfragmentation profiles were compared with that of the bacterialmethyl ester mix (BAME mix), which does not contain lactobacillicacid. Twowell separated peaks, having retention times of 36.27 minand 36.37 min, were present in the L. helveticus profile while, inthe same elution zone, only one peak, corresponding to the dehy-drosterculic acid (36.27 min), included in the standard mix, waspresent in the BAME mix profile (Fig. 1). On the basis of the massfragmentation profiles of the two peaks, which were identical(Fig. 2A and B), they could be identified as CFAs. Differences betweenthem were not easy to assess, even with the most recent chemicalcompounds database associated with GCeMS. The assumptionthat the peak having retention time 36.37 was lactobacillic acid wasverified by the following protocol. The methylated sample ofL. helveticus FAs extract was hydrogenated and brominated toconfirm the fact that the compounds at RT ¼ 36.27 and RT ¼ 36.37were cyclopropane ring FAs. In Figs. 3AeC, the chromatogramsobtained before hydrogenation, after hydrogenation and afterbromination are reported. Hydrogenation did not affect the twotarget peaks, while for example oleic acid (C18:1n9), cis-vaccenicacid (C18:1n11) and other UFAs disappeared (Fig. 3B compared to3A). The absence of the peaks at RT ¼ 36.27 and RT ¼ 36.37 (Fig. 3Ccompared to 3B) demonstrated the cyclic nature of the two compounds.Therefore on the basis of themass fragmentation spectra and of thislatter confirm of their cyclic nature, the two peaks were henceidentified as dehydrosterculic and lactobacillic acid respectively.

3.2. Influence of acid stress and cold stress on cellular fatty acidscomposition in L. helveticus and L. sanfranciscensis

Overnight cells of both strains, grown without Tween 80 underanaerobic conditions, were collected and exposed for 2 h (inoculumlevel 8 log CFU/ml) in a fresh medium to sublethal acid or coldstress in the presence or absence of Tween 80 and under micro-aerobic or anaerobic conditions. The viability changes of the cells ofboth strains during incubation at the various stress conditions werenot significant (data not shown).

The Table 2A and B report the relative percentage of the cellularFAs of L. helveticus and L. sanfranciscensis respectively in the pre-inoculum, in the control without stress exposure and in cells exposedfor 2 h to acid stress in micro-aerobiosis and in anaerobiosis inthe presence or absence of a source of oleic acid. It is important tomention that the preinoculum cells were in the stationary phase

.00 31.00 32.00 33.00 34.00 35

0.00 31.00 32.00 33.00 34.00 35

Retention T

dehydrosterculic acid (C1

dehydrosterculic acid (C1

.00 31.00 32.00 33.00 34.00 35

0.00 31.00 32.00 33.00 34.00 35

Retention T

dehydrosterculic acid (C1

.00 31.00 32.00 33.00 34.00 35

0.00 31.00 32.00 33.00 34.00 35

Retention T

dehydrosterculic acid (C1

.00 31.00 32.00 33.00 34.00 35

0.00 31.00 32.00 33.00 34.00 35

Retention T.00 31.00 32.00 33.00 34.00 35

0.00 31.00 32.00 33.00 34.00 35

ecnadnubA

Retention T

dehydrosterculic acid (C1dehydrosterculic acid (C1

dehydrosterculic acid (C1dehydrosterculic acid (C1

Fig. 1. Comparison between Bacterial Acid Methyl Ester (BAME

while the control (Co) corresponded to the same cells incubated infresh medium for 2 h. In fact the percentage of C19cyc11 in L. helve-ticus preinoculum presented the maximum level. The 2 h incubationresulted in a redistribution or metabolism of this CFA, accompaniedby a synthesis of C19cyc9 in the presence of oleic acid. The principalFAs of L. helveticus were in the order: C19cyc11, ranging between 41and 21%, C16:0 and C18:1n11. On the other hand in L. sanfranciscensisC12:0 was the prevalent FA, accompanied by C16:0, C18:0 andC18:1n9, presumably uptaken by the medium or synthesized by theaerobic pathway. In the former strain the cyclopropane C19cyc9did not exceed 15.0%. A first comparison indicated, on the basis of thedouble bond position, that the C18 monounsaturated FAs in the twostrains had different origin. In L. helveticus the prevalent occurrenceof C18:1n11 accounts for the anaerobic pathway while in L. san-franciscensis the predominance of C18:1n9 also in absence of Tween80 accounts for the aerobic pathway. Another significant differencebetween the two strains was the high proportion of the poly-unsaturated fatty acid (PUFA) C18:3 in L. sanfranciscensis. Theproportion of the medium chain C12:0, which attained level of 35%,unusual in bacteria, tended to increase with the acid stress inL. sanfranciscensis. On the other hand in L. helveticus C19cyc11,accompanied by relatively higher proportion of C18:1n11, increasedwith the acid stress in anaerobiosis. Lactobacillic acid was also thepredominant FA at pH 3 in aerobiosis, in the presence of Tween 80.

The Tables 3A and B report the changes of the proportion ofthe cellular FAs after 2 h of exposure to cold stress in L. helveticusand L. sanfranciscensis respectively.

In L. helveticus the principal changes, following the temperaturedecrease in the presence of Tween 80, regarded the reduction ofC19cyc9 and the increase of C19cyc11. The unsaturated/saturatedfatty acid ratio after 2 h in L. helveticus showed the same trend asL. sanfranciscensis.

In L. sanfranciscensis C12:0 significantly increased while C18:1n9generally diminished with temperature except in the presence ofTween 80 in anaerobiosis. Its diminution should not be attributedto cyclization because C19cyc9 decreased with temperature. Alsothe USFAs ratio decreased after 2 h at 10 �C with the exception ofthe cells incubated in the presence of Tween 80 in anaerobiosis.

In the strain L. sanfranciscensis the temperature response wasassociated principally with C12:0 increase while in L. helveticus it wasassociated with changes in CFAs. In fact C19cyc9 was present in non-negligibleconcentrationonly in thepresenceof anexogenous sourceofitsprecursor (oleic acid).On theotherhand in the latter strainC19cyc11was the principal CFA irrespectively of oleic acid supplementation.

.00 36.00 37.00 38.00 39.00

.00 36.00 37.00 38.00 39.00

ime (min)

9cyc9)

lactobacillic acid (C19cyc11)9cyc9)

.00 36.00 37.00 38.00 39.00

.00 36.00 37.00 38.00 39.00

ime (min)

9cyc9)

lactobacillic acid (C19cyc11)

.00 36.00 37.00 38.00 39.00

.00 36.00 37.00 38.00 39.00

ime (min)

9cyc9)

.00 36.00 37.00 38.00 39.00

.00 36.00 37.00 38.00 39.00

ime (min).00 36.00 37.00 38.00 39.00

.00 36.00 37.00 38.00 39.00

ime (min)

9cyc9)9cyc9)

lactobacillic acid (C19cyc11)lactobacillic acid (C19cyc11)9cyc9)9cyc9)

) standard (above) and a sample containing both the CFAs.

m/z

40 60 80 100 120 140 160 180 200 220 240 260 280 3000

10

20

30

40

50

60

70

80

90

A (RT = 36.26)

40 60 80 100 120 140 160 180 200 220 240 260 280 3000

10

20

30

40

50

60

70

80

90

B (RT = 38.36)

ytisnetnievitale

R

m/z

40 60 80 100 120 140 160 180 200 220 240 260 280 3000

10

20

30

40

50

60

70

80

90

A (RT = 36.26)

40 60 80 100 120 140 160 180 200 220 240 260 280 3000

10

20

30

40

50

60

70

80

90

B (RT = 38.36)

ytisnetnievitale

R

40 60 80 100 120 140 160 180 200 220 240 260 280 3000

10

20

30

40

50

60

70

80

90

A (RT = 36.26)

40 60 80 100 120 140 160 180 200 220 240 260 280 3000

10

20

30

40

50

60

70

80

90

B (RT = 38.36)

ytisnetnievitale

R

Fig. 2. Mass spectra of (A) dehydrosterculic (C19cyc9) and (B) lactobacillic acid (C19cyc11).

C. Montanari et al. / Food Microbiology 27 (2010) 493e502 497

In order to assess if the unusual concentration of C12:0 wasassociated with an accumulation of FFAs, their content in L. san-franciscensis and L. helveticus under standard conditions weredetected. According to the Table 4, the proportion of free C12:0 wasnegligible in both strains, suggesting that this FA is prevalentlyintegrated in phospholipids. On the contrary C18:0 was present at99.17% as FFA in L. sanfranciscensis.

4. Discussion

Both polar columns like carbowax and Hp-FFAP (Guerzoni et al.,2001; Teixeira et al., 2002) and non-polar columns like SE-30, OV1(Rizzo et al., 1987) are currently used to separate and detect cellularmethyl ester FAs. The temperature rate generally adopted rangesfrom3 �C/min to 10 �C/min in the zone of elution of the C19cyc9 andC19cyc11 andmaybe this is the reason for the absence of separationof these two CFAs. The procedure suggested in this paper isa development of a method proposed by Hanus et al. (2008) for theidentification of eight CFAs in walnuts. These Authors used a non-polar column HP-5MS, combined with a temperature rate goingfrom 60 �C to 260 �C, with a constant increase of 3 �C/min. For theseparation of C19cyc9 and C19cyc11, which have identical massspectra, a non-polar column, similar to that of Hanus et al. (2008),was used for this investigation. Moreover the innovative procedureto successfully detect the methyl esters of both CFAs, as well as ofthe other FAs, was based on a different modulation of the temper-ature rate. An initial temperature of 120 �C and an increase rate of3 �C/min in the zone of elution of the most common FAs methylesters and a subsequent decrease of the rate to 0.5 �C/min in the

zone of elution of the CFAs were used. The availability of an efficientand specific CFAs separation method allowed a better character-ization of lactobacilli fatty acids, and particularly of CFAs, withrespect to the currently used protocols (Rizzo et al., 1987; Partanenet al., 2001; Guerzoni et al., 2001; Teixeira et al., 2002).

L. sanfranciscensis and L. helveticus strains were chosen becauseof their prevalence in specific food habitats: L. sanfranciscensis isa dominant species in cereal sour dough and it has never beenisolated in other habitats up to now (Gobbetti et al., 1999; Hammeset al., 2005). L. helveticus is widely used as a starter in themanufacturing of Italian and Swiss cheeses and for the biotechno-logical production of lactic acid. During these industrial processesL. helveticus is exposed, under anaerobic conditions, to tempera-tures exceeding 53 �C (Guerzoni et al., 2001; Di Cagno et al., 2006).The improvement of the separation of the two CFAs and the positionof the double bond in the two C18:1 enabled the assumption thatthe aerobic biosynthetic pathway for UFAs is that prevalent inL. sanfranciscensis while the anaerobic pathway is that prevalent inL. helveticus. In fact in the latter strain also in the presence of Tween80 and under micro-aerobic conditions, C18:1n11 and its post-synthetic derivative C19cyc11 accounted for overall proportionsranging from 28 to 52% of the total FAs. On the other hand L. san-franciscensis assimilated or synthesized by aerobic pathwayC18:1n9 and transformed it in part into C19cyc9. However, in thelatter strain, the cumulative levels of the two FAs did not exceed30%. Moreover non-negligible proportions of C18:0 were observedin L. sanfranciscensis. This FAwas quite totally present, at least in thepreinoculum, as FFA. It derives or from the reduction of an exoge-nous source of C18:1n9 (Corcoran et al., 2007) or from the aerobic

Fig. 3. Chromatograms obtained before hydrogenation (A), after hydrogenation (B) and after bromination (C). cis-9-octadecenoic acid (C18:1n9), cis-11-octadecenoic acid(C18:1n11), dehydrosterculic (C19cyc9) and lactobacillic acid (C19cyc11).

C. Montanari et al. / Food Microbiology 27 (2010) 493e502498

synthetic pathway (Keweloh and Heipieper, 1996). In fact, in theanaerobic pathway of FAs the formation of the double bond antic-ipates the lengthening of the chain from C16:1n9 to C18:1n11(Keweloh and Heipieper, 1996). The prevalence of the aerobicpathway in L. sanfranciscensis can be associated with the originalhabitat of this strain, sour dough, which is characterized by aerobicor semi-aerobic conditions due to the continuous doughmixing. Onthe other hand the natural habitat of L. helveticus is characterized,

during cheese making, by negligible dissolved oxygen concentra-tion due to temperatures exceeding 53 �C, which reduce the oxygensolubility in the milk.

The occurrence of oxygen dependent desaturase in L. san-franciscensis is also proved by the presence, up to 11%, of C18:3which is absent in SDB. In the preparation of this growth mediumfresh cells of Saccharomyces cerevisiae were used. This speciesnotably does not produce PUFA (Kajiwara et al., 1996).

Table 2ARelative percentage of the cellular fatty acids of L. helveticus in the preinoculum (pre), in the control without stress exposure (co) and in cells exposed for 2 h at pH 5.0, 4.0 and3.0 in micro-aerobic and anaerobic conditions in the presence (t80) or absence (not) of a source of oleic acid. The data are means of three independent experiments. Thecoefficients of variability were lower than 5%.

Fatty acidsa Pre Micro-aero t80 Micro-aero not Ana t80 Ana not

co pH5 pH4 pH3 co pH5 pH4 pH3 co pH5 pH4 pH3 co pH5 pH4 pH3

C12:0 1.00 2.46 2.65 4.40 5.06 2.12 2.41 3.88 7.62 3.50 2.17 1.35 3.02 4.02 1.73 2.43 7.51C13:0 0.11 0.01 0.01 0.05 0.01 0.19 0.03 0.08 1.29 0.13 0.02 0.00 0.19 0.18 0.00 0.00 1.26C14:0 5.92 6.93 6.78 6.11 6.07 7.00 6.34 4.77 6.05 7.25 6.39 5.88 6.71 8.98 6.29 6.58 8.83C15:0 0.05 0.17 0.10 0.02 0.03 0.09 0.01 0.02 0.04 0.18 0.12 0.02 0.06 0.09 0.05 0.07 0.07C16:1 9.93 7.10 8.93 9.44 8.38 11.38 9.70 9.84 7.63 7.50 7.22 9.39 10.01 14.03 10.38 9.92 11.78C16:0 26.92 30.25 28.07 27.40 28.04 26.23 28.32 27.85 27.38 32.00 29.41 26.64 28.36 29.16 26.84 28.00 16.79C17:0 0.05 0.16 0.11 0.11 0.44 0.03 0.35 0.20 1.04 0.12 0.03 0.03 0.07 0.09 0.04 0.22 0.35C18:3 0.26 0.58 0.35 0.94 1.36 0.26 0.83 0.80 2.00 0.47 0.28 0.36 0.58 1.23 0.33 0.43 1.57C18:2 0.20 0.19 0.23 0.18 0.17 0.18 0.28 0.35 0.53 0.06 0.39 0.07 0.15 0.28 0.25 0.22 0.27C18:1n9 cis 0.86 8.72 4.03 4.51 6.90 1.70 1.85 2.33 6.27 9.90 6.93 3.61 4.88 1.80 1.28 1.94 2.50C18:1n11 11.83 6.26 9.86 11.40 10.07 13.19 11.54 11.75 9.22 6.65 7.08 11.92 10.72 12.75 11.33 11.38 11.84C18:0 1.23 1.16 1.59 1.61 1.64 1.50 1.87 2.04 1.73 1.30 1.40 1.48 1.17 0.82 2.17 1.86 2.17C19:0 cyc9 0.76 14.96 4.73 1.48 2.03 1.02 1.65 1.36 1.57 9.13 14.54 2.62 0.81 0.70 0.87 1.20 0.93C19:0 cyc11 40.83 21.04 32.56 32.20 29.70 35.11 34.79 34.68 27.63 21.70 24.02 36.63 33.17 25.68 38.45 35.67 33.85C19:0 0.04 0.00 0.00 0.13 0.11 0.00 0.03 0.08 0.02 0.13 0.01 0.00 0.10 0.20 0.00 0.05 0.25C20:0 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.03USFAs ratio 0.66 0.56 0.60 0.66 0.65 0.72 0.61 0.64 0.57 0.55 0.55 0.72 0.66 0.69 0.64 0.61 0.80USFAs ratio cyc 1.83 1.43 1.54 1.51 1.42 1.69 1.54 1.57 1.21 1.24 1.53 1.82 1.52 1.30 1.69 1.55 1.68

USFA ratio: Unsaturated/saturated fatty acid ratio determined without considering cyclopropane fatty acids.USFA ratio cyc: Unsaturated/saturated fatty acid ratio determined considering CFAs as unsaturated fatty acids.

a Fatty acids are: dodecanoic acid (C12:0), tridecanoic acid (C13:0) tetradecanoic acid (C14:0), pentadecanoic acid (C15:0), cis-9-hexadecenoic acid (C16:1), hexadecanoicacid (C16:0), heptadecanoic acid (C17:0), cis-9,cis-12,cis-15-octadecatrienoic acid (C18:3), cis-9,cis-12-octadecadienoic acid (C18:2), cis-9-octadecenoic acid (C18:1n9), cis-11-octadecenoic acid (C18:1n11), octadecanoic acid (C18:0), cis-9,10-methyleneoctadecanoic acid (C19cyc9), cis-11,12-methyleneoctadecanoic acid (C19cyc11),nonadecanoic acid (C19:0), eicosanoic acid (C20:0).

C. Montanari et al. / Food Microbiology 27 (2010) 493e502 499

In L. helveticus C19cyc11 dramatically increased after 2 h at 10 �Cand with the pH decrease, particularly in micro-aerobic conditions,in the presence of Tween 80, and in anaerobic conditions. On theother hand in L. sanfranciscensis the relevant proportion of C12:0,uncommon in bacteria (Ratledge and Wilkinson, 1989), suggeststhat the shortening of carbon chain is the principal strategy of thisstrain tomodulate the fluidity or other chemico-physical propertiesof the membranes. The prevalence of C12:0, not as FFAbut presumably integrated in the phospholipids, is rare in

Table 2BRelative percentage of the cellular fatty acids of L. sanfranciscensis in the preinoculum (preand 3.0 in micro-aerobic and anaerobic conditions in the presence (t80) or absence (noThe coefficients of variability were lower than 5%.

Fatty acidsa Pre Micro-aero t80 Micro-aero not

co pH4 pH3,6 pH3 co pH4 pH3,6

C12:0 26.76 28.86 28.89 32.43 35.68 33.40 29.00 33.35C13:0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.74C14:0 4.39 3.11 2.61 0.28 2.10 2.66 3.54 3.42C15:0 0.00 0.10 0.00 0.00 0.00 0.06 0.00 0.21C16:1 2.43 0.00 0.33 0.00 0.00 0.00 0.00 0.27C16:0 13.18 28.13 28.98 28.83 26.84 32.28 30.06 30.95C17:0 1.16 1.27 1.73 0.72 1.34 0.92 1.12 1.52C18:3 11.07 0.00 8.06 10.15 0.00 7.01 10.95 8.83C18:2 2.29 0.21 0.00 0.00 0.17 0.35 0.51 1.64C18:1n9 cis 15.51 23.28 17.67 17.51 17.25 7.34 10.04 8.65C18:1n11 10.45 5.27 0.00 0.00 5.08 0.10 0.66 0.44C18:0 10.36 3.60 7.42 4.41 6.76 10.07 9.35 8.67C19:0 cyc9 1.21 6.02 4.01 5.42 3.98 5.73 3.69 0.71C19:0 cyc11 1.18 0.15 0.30 0.05 0.24 0.00 0.48 0.32C19:0 0.00 0.00 0.00 0.21 0.55 0.09 0.61 0.29C20:0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00USFAs ratio 0.75 0.44 0.38 0.42 0.32 0.19 0.31 0.25USFAs ratio cyc 0.79 0.54 0.44 0.50 0.36 0.26 0.36 0.26

USFA ratio: determined without considering cyclopropane fatty acids.USFA ratio cyc: determined considering CFAs as unsaturated fatty acids.

a Fatty acids are: dodecanoic acid (C12:0), tridecanoic acid (C13:0) tetradecanoic acidacid (C16:0), heptadecanoic acid (C17:0), cis-9,cis-12,cis-15-octadecatrienoic acid (C18:3)octadecenoic acid (C18:1n11), octadecanoic acid (C18:0), cis-9,10-methyleneoctadecanoacid (C19:0), eicosanoic acid (C20:0).

microorganisms. The proportion of C12:0 in L. bulgaricus has beenreported to exceed levels of 18%, in the absence of a source of oleicacid (Smittle et al., 1974). Johnston et al. (2004) reported that themean concentration of dodecanoic acid in Clostridium perfringensphospholipids was of 29%. Moreover, in the yeast Kluyveromycesfragilis the survival under frozen conditions was linearly correlatedwith the proportion of the latter medium chain FA (ranging from 2to 27%) (Bostian and Gilliland, 1981). Another medium chainFA, C14:0, although fairly uncommon, has been identified, with

), in the control without stress exposure (co) and in cells exposed for 2 h at pH 4.0, 3.6t) of a source of oleic acid. The data are means of three independent experiments.

Ana t80 Ana not

pH3 co pH4 pH3,6 pH3 co pH4 pH3,6 pH3

34.68 27.78 28.42 22.44 30.40 27.29 29.84 26.69 30.560.00 0.19 0.00 0.00 0.00 0.26 0.00 0.00 0.232.35 2.84 2.98 15.36 2.67 3.26 2.45 3.34 2.070.00 0.12 0.00 0.00 0.96 0.16 0.00 0.00 0.000.16 0.00 0.18 0.00 0.07 0.40 0.00 0.00 0.11

29.37 24.17 29.38 25.61 24.34 27.60 30.50 29.94 27.914.11 2.34 1.16 0.31 3.30 3.03 0.95 0.00 0.004.19 7.16 2.56 0.39 6.47 6.23 0.96 2.60 1.821.02 0.85 0.27 0.33 0.64 1.20 0.88 0.82 3.868.42 18.57 23.16 20.90 15.30 8.53 11.49 12.00 1.230.10 4.75 1.14 0.00 4.97 6.15 1.96 0.00 10.767.34 6.11 6.55 5.98 7.37 9.91 7.64 10.06 3.668.13 4.97 3.68 8.49 2.90 4.32 13.20 14.56 8.490.13 0.16 0.00 0.00 0.62 0.29 0.12 0.00 8.280.00 0.00 0.53 0.20 0.00 1.16 0.00 0.00 0.530.00 0.00 0.00 0.00 0.00 0.21 0.00 0.00 0.510.18 0.49 0.40 0.31 0.40 0.31 0.22 0.22 0.270.28 0.57 0.45 0.43 0.45 0.37 0.40 0.43 0.53

(C14:0), pentadecanoic acid (C15:0), cis-9-hexadecenoic acid (C16:1), hexadecanoic, cis-9,cis-12-octadecadienoic acid (C18:2), cis-9-octadecenoic acid (C18:1n9), cis-11-ic acid (C19cyc9), cis-11,12-methyleneoctadecanoic acid (C19cyc11), nonadecanoic

Table 3ARelative percentage of the cellular fatty acids of L. helveticus in the preinoculum (pre), in the control without stress exposure (co) and in cells exposed for 2 h at 10 �C in micro-aerobic and anaerobic conditions in the presence (t80) or absence (not) of a source of oleic acid. The data are means of three independent experiments. The coefficients ofvariability were lower than 5%.

Fatty acidsa Pre Micro-aero t80 Micro-aero not Ana t80 Ana not

co 10 �C co 10 �C co 10 �C co 10 �C

C12:0 1.00 2.46 3.44 2.12 4.42 3.50 5.83 4.02 4.76C13:0 0.11 0.01 0.40 0.19 0.21 0.13 0.55 0.18 0.54C14:0 5.92 6.93 8.24 7.00 7.49 7.25 7.01 8.98 6.89C15:0 0.05 0.17 0.07 0.09 0.08 0.18 0.24 0.09 0.05C16:1 9.93 7.10 9.42 11.38 9.61 7.50 8.17 14.03 8.97C16:0 26.92 30.25 28.42 26.23 30.07 32.00 26.82 29.16 30.34C17:0 0.05 0.16 0.22 0.03 0.24 0.12 0.12 0.09 0.07C18:3 0.26 0.58 0.37 0.26 1.10 0.47 1.00 1.23 0.29C18:2 0.20 0.19 0.15 0.18 0.46 0.06 0.40 0.28 0.18C18:1n9 c 0.86 8.72 4.91 1.70 2.36 9.90 4.82 1.80 1.59C18:1n11 11.83 6.26 8.96 13.19 10.29 6.65 9.46 12.75 10.27C18:0 1.23 1.16 1.72 1.50 1.65 1.30 0.76 0.82 1.78C19:0 cyc9 0.76 14.96 3.08 1.02 0.95 9.13 3.61 0.70 0.93C19:0 cyc11 40.83 21.04 30.49 35.11 31.03 21.70 30.72 25.68 33.27C19:0 0.04 0.00 0.12 0.00 0.05 0.13 0.50 0.20 0.07C20:0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00USFAs ratio 0.66 0.56 0.56 0.72 0.54 0.55 0.57 0.69 0.48USFAs ratio cyc 1.83 1.43 1.35 1.69 1.26 1.24 1.39 1.30 1.25

USFA ratio: Unsaturated/saturated fatty acid ratio determined without considering cyclopropane fatty acids.USFA ratio cyc: Unsaturated/saturated fatty acid ratio determined considering CFAs as unsaturated fatty acids.

a Fatty acids are: dodecanoic acid (C12:0), tridecanoic acid (C13:0) tetradecanoic acid (C14:0), pentadecanoic acid (C15:0), cis-9-hexadecenoic acid (C16:1), hexadecanoicacid (C16:0), heptadecanoic acid (C17:0), cis-9,cis-12,cis-15-octadecatrienoic acid (C18:3), cis-9,cis-12-octadecadienoic acid (C18:2), cis-9-octadecenoic acid (C18:1n9), cis-11-octadecenoic acid (C18:1n11), octadecanoic acid (C18:0), cis-9,10-methyleneoctadecanoic acid (C19cyc9), cis-11,12-methyleneoctadecanoic acid (C19cyc11), nonadecanoicacid (C19:0), eicosanoic acid (C20:0).

C. Montanari et al. / Food Microbiology 27 (2010) 493e502500

proportion ranging from traces to 21%, in cellular FAs of lactobacilli(Smittle et al., 1974; Partanen et al., 2001; Wang, 2005; Behr et al.,2006; Streit et al., 2008).

The cereal sour dough can be regarded as a habitat characterizedby the occurrence of particles, starch granules and gluten reticuleswhich can harbour more or less attached dense microbialcommunities. The surfaces of this complex system are the sites ofbiological and microbial activity. FAs can modify surface tensionand increase the contact between microbial cells and particles of

Table 3BRelative percentage of the cellular fatty acids of L. sanfranciscensis in the preinoculum (prmicro-aerobic and anaerobic conditions in the presence (t80) or absence (not) of a sourceof variability were lower than 5%.

Fatty acidsa Pre Micro-aero t80 Micro-ae

co 10 �C co

C12:0 26.76 28.86 35.77 33.40C13:0 0.00 0.00 0.00 0.00C14:0 4.39 3.11 3.01 2.66C15:0 0.00 0.10 0.00 0.06C16:1 2.43 0.00 0.00 0.00C16:0 13.18 28.13 30.28 32.28C17:0 1.16 1.27 1.23 0.92C18:3 11.07 0.00 0.00 7.01C18:2 2.29 0.21 0.00 0.35C18:1n9 cis 15.51 23.28 14.73 7.34C18:1n11 10.45 5.27 5.65 0.10C18:0 10.36 3.60 6.40 10.07C19:0 cyc9 1.21 6.02 2.93 5.73C19:0 cyc11 1.18 0.15 0.00 0.00C19:0 0.00 0.00 0.00 0.09C20:0 0.00 0.00 0.00 0.00USFAs ratio 0.75 0.44 0.27 0.19USFAs ratio cyc 0.79 0.54 0.30 0.26

USFA ratio: Unsaturated/saturated fatty acid ratio determined without considering cycloUSFA ratio cyc: Unsaturated/saturated fatty acid ratio determined considering CFAs as u

a Fatty acids are: dodecanoic acid (C12:0), tridecanoic acid (C13:0) tetradecanoic acidacid (C16:0), heptadecanoic acid (C17:0), cis-9,cis-12,cis-15-octadecatrienoic acid (C18:3)octadecenoic acid (C18:1n11), octadecanoic acid (C18:0), cis-9,10-methyleneoctadecanoacid (C19:0), eicosanoic acid (C20:0).

different nature (Chang et al., 2009). In fact, the uncommon extentof this medium chain FA could be related also to the observeddifficulty of sedimentation of L. sanfranciscensis in SDB. Moreoverdodecanoic acid is reported to modulate the macromoleculehydrophobicity (Sallustio et al., 2004). In addition this FA wasone of the principal component of the cell-wall-bound FAs inRhodococcus erythropolis and it was involved in the hydrocarbonadhesion of the cells (Chang et al., 2009) and in their interactionswith particles.

e), in the control without stress exposure (co) and in cells exposed for 2 h at 10 �C inof oleic acid. The data are means of three independent experiments. The coefficients

ro not Ana t80 Ana not

10 �C co 10 �C co 10 �C

38.98 27.78 28.20 27.29 31.930.00 0.19 0.20 0.26 0.000.87 2.84 1.52 3.26 3.070.00 0.12 0.56 0.16 0.310.00 0.00 0.06 0.40 0.88

36.02 24.17 25.37 27.60 26.791.66 2.34 2.57 3.03 5.685.34 7.16 6.29 6.23 8.170.36 0.85 0.32 1.20 1.514.25 18.57 23.27 8.53 7.090.00 4.75 5.94 6.15 0.008.08 6.11 2.50 9.91 6.083.96 4.97 2.15 4.32 8.110.24 0.16 0.05 0.29 0.370.24 0.00 0.96 1.16 0.000.00 0.00 0.05 0.21 0.000.12 0.49 0.58 0.31 0.240.16 0.57 0.61 0.37 0.35

propane fatty acids.nsaturated fatty acids.(C14:0), pentadecanoic acid (C15:0), cis-9-hexadecenoic acid (C16:1), hexadecanoic, cis-9,cis-12-octadecadienoic acid (C18:2), cis-9-octadecenoic acid (C18:1n9), cis-11-ic acid (C19cyc9), cis-11,12-methyleneoctadecanoic acid (C19cyc11), nonadecanoic

Table 4Proportion of non-esterified fatty acids (free fatty acids) with respect to the totalamount of the same cellular fatty acids of preinoculum of L. helveticus and L. san-franciscensis. The data aremeans of three independent experiments. The coefficientsof variability were lower than 5%.

Fatty acidsa L. helveticus L. sanfanciscensis

C12:0 1.07 0.74C13:0 0.00 0.00C14:0 0.64 22.45C15:0 21.32 0.00C16:1 0.00 0.00C16:0 0.89 38.72C17:0 16.12 40.58C18:3 0.08 0.00C18:2 5.51 12.56C18:1n9 cis 12.74 10.81C18:1n11 0.22 2.19C18:0 11.24 99.17C19:0 cyc9 0.06 0.00C19:0 cyc11 0.10 0.00

USFA ratio: Unsaturated/saturated fatty acid ratio determined without consideringcyclopropane fatty acids.USFA ratio cyc: Unsaturated/saturated fatty acid ratio determined considering CFAsas unsaturated fatty acids.

a Fatty acids are: dodecanoic acid (C12:0), tridecanoic acid (C13:0) tetradecanoicacid (C14:0), pentadecanoic acid (C15:0), cis-9-hexadecenoic acid (C16:1), hex-adecanoic acid (C16:0), heptadecanoic acid (C17:0), cis-9,cis-12,cis-15-octadeca-trienoic acid (C18:3), cis-9,cis-12-octadecadienoic acid (C18:2), cis-9-octadecenoicacid (C18:1n9), cis-11-octadecenoic acid (C18:1n11), octadecanoic acid (C18:0), cis-9,10-methyleneoctadecanoic acid (C19cyc9), cis-11,12-methyleneoctadecanoic acid(C19cyc11), nonadecanoic acid (C19:0), eicosanoic acid (C20:0).

C. Montanari et al. / Food Microbiology 27 (2010) 493e502 501

The formation of CFAs is known as a post-synthetic modificationof cis-unsaturated fatty acids. This conversion does not significantlychanges, according to some Authors (Grogan and Cronan, 1986;Diefenbach et al., 1992; Denich et al., 2003) the physiological prop-erties of the membrane and its physiological function is still notunderstood. According to some Authors (Gómez Zavaglia et al.,2000) their increase is necessary to maintain the cell membrane ina suitable state of fluidity. However the changes of chemico-physicalproperties of membrane conferred by CFAs are not well character-ized (Grogan and Cronan, 1997). Because CFAs have biophysicalproperties similar to those of the correspondent UFAs, the reasonbehind the preference of CFAs, as the means of adaptation towardsstresses, can be found in the major chemical stability of thesecompounds compared to that of UFAs. In fact, CFAs are resistant toozonolysis, singlet oxygen, mild oxidative treatments (Grogan andCronan, 1986, 1997). The intrinsic resistance of CFAs to chemicalstresses accounts for their role in the improvement of themembranechemical properties, rather than of its physical properties.

In fact, in general CFAs seem to stabilizemembrane lipids againstturnover and degradation (MacDonald et al., 1985) and function asa resistance barrier against environmental stresses (Muñoz-Rojaset al., 2006). Indeed, the proportion of CFAs is modified in stressedcells, and the strains with altered CFA synthase activity exhibitaltered phenotypes towards diverse stressing situations (Budin-Verneuil et al., 2005). Moreover cyclopropanation of FAs was shownto be essential for acid resistance in Escherichia coli (Brown et al.,1997) and Salmonella enterica (Kim et al., 2005). In fact, an increasedproton permeability and decreased ability to extrude Hþ wereobserved in a mutant of E. coli unable to synthesize CFAs (ShabalaandRoss, 2008). The Authors suggested that CFAs protect cells actingas a permeability barrier to the protons or other chemicals.

Microbial FAs composition is the result of complex phenomenamaintaining optimal viability under different conditions. Moreover,the characteristic of growth medium or food systems considerablyaffect the FAs profile and their alterations. L. helveticus grown inwhey, rich of oleic and linoleic acid, incorporates and modulates

these FAs and their derivatives when exposed to acid and oxidativestresses (Guerzoni et al., 2001).

L. helveticus and L. sanfranciscensis showed different FAs compo-sition and environmental adaptation. The aerobic and anaerobic FAsbiosynthetic pathways coexist in both strains although the aerobicpathway is predominant in L. sanfranciscensis. The two speciesevolved in different habitats. In particular L. sanfranciscensis offers aninteresting perspective because it has evolved away in a stronglyspecialized environment created by man, i.e. fermented cereals.Traditional sour dough making is based on the sequential andrepeated transfer of samples to fresh dough. With records of sourdough dating to 1500 before Christ (von Stokar, 1956), one wouldlogically predict that over time L. sanfranciscensis adapted to thisenvironment by a specialization through loss (Serrazanetti, 2009)and/or remodulation of functions. The significant difference in cellFAs profiles and their changes following stress exposure of L. san-franciscensis, with respect to the major part of lactobacilli, accountsfor this long environmental adaptation.

References

Béal, C., Fonseca, F., Corrieu, G., 2001. Resistance to freezing and frozen storage ofStreptococcus thermophilus is related to membrane fatty acid composition.J. Dairy Res. 84, 2347e2356.

Behr, J., Gänzle, M.G., Vogel, R.F., 2006. Characterization of a highly hop-resistantLactobacillus brevis strain lacking hop transport. Appl. Environ. Microbiol. 72,6483e6492.

Bostian, M., Gilliland, S.E., 1981. Relationship of survival of Kluyveromices fragilis at�19 �C to cellular fatty acid composition. J. Food Sci. 46, 964e965.

Brennan, M., Wanismail, B., Johnson, M.C., Ray, B., 1986. Cellular damage in driedLactobacillus acidophilus. J. Food Prot. 49, 47e53.

Brian, B.L., Gardner, E.W., 1968. A simple procedure for detecting the presence ofcyclopropane fatty acids in bacterial lipids. Appl. Microbiol. 16, 549e552.

Brown, J.L., Ross, T., McMeekin, T.A., Nichols, P.D., 1997. Acid habituation ofEscherichia coli and the potential role of cyclopropane fatty acids in low pHtolerance. Int. J. Food Microbiol. 37, 163e173.

Budin-Verneuil, A., Maguin, E., Auffray, Y., Ehrlich, S.D., Pichereau, V., 2005. Tran-scriptional analysis of the cyclopropane fatty acid synthase gene of Lactococcuslactis MG1363 at low pH. FEMS Microbiol. Lett. 250, 189e194.

Chang, W.N., Liu, C.W., Liu, H.S., 2009. Hydrophobic cell surface and bioflocculationbehaviour of Rhodococcus erythropolis. Process. Biochem. 44, 955e962.

Corcoran, B.M., Stanton, C., Fitzgerald, G.F., Ross, R.P., 2007. Growth of probioticlactobacilli in the presence of oleic acid enhances subsequent survival in gastricjuice. Microbiology 153, 291e299.

De Angelis, M., Bini, L., Pallini, V., Cocconcelli, P.S., Gobbetti, M., 2001. The acid stressresponse in Lactobacillus sanfranciscensis CB1. Microbiology 147, 1863e1873.

Denich, T.J., Beaudette, L.A., Lee, H., Trevors, J.T., 2003. Effect of selected environ-mental and physico-chemical factors on bacterial cytoplasmatic membranes.J. Microbiol. Meth. 52, 149.

Di Cagno, R., De Angelis, M., Limitone, A., Fox, P.F., Gobbetti, M., 2006. Response ofLactobacillus helveticus PR4 to heat stress during propagation in cheesewheywitha gradient of decreasing temperatures. Appl. Environ. Microbiol. 72, 4503e4514.

Diefenbach, R., Heipieper, H.J., Keweloh, H., 1992. The conversion of cis into transunsaturated fatty acids in Pseudomonas putida P8: evidence for a role in theregulation of membrane fluidity. Appl. Microbiol. Technol. 38, 382.

Dionisi, F., Golay, P.A., Elli, M., Fay, L.B., 1999. Stability of cyclopropane and conju-gated linoleic acids during fatty acid quantification in lactic acid bacteria. Lipids34, 1107e1115.

Ehrmann, M.A., Vogel, R.F., 2005. Molecular taxonomy and genetics of sourdoughlactic acid bacteria. Trends Food Sci. Technol. 16, 31e42.

Fernandez Murga, M.L., Cabrera, G.M., Font De Valdez, G., Disalvo, A., Seldes, A.M.,2000. Influence of growth temperature on cryotolerance and lipid compositionof Lactobacillus acidophilus. J. Appl. Microbiol. 88, 342e348.

Gänzle, M.G., Ehrmann, M.A., 1998. Modeling of growth of Lactobacillus san-franciscensis and Candida milleri in response to process parameters of thesourdough fermentation. Appl. Environ. Microbiol. 64, 2616e2623.

Gilliland, S.E., Speck, M.L., 1974. Relationship of cellular components to the stabilityof concentrated lactic Streptococcus cultures at �17 �C. Appl. Microbiol. 27,793e796.

Gobbetti, M., Folkertsma, B., Fox, P.F., Corsetti, A., Smacchi, E., De Angelis, M.,Rossi, J., Kilcawley, K., Cortini, M., 1999. Microbiology and biochemistry of Fossa(pit) cheese. Int. Dairy J. 9, 763e773.

Gómez Zavaglia, A., Disalvo, A.E., De Antoni., L.G., 2000. Fatty acid composition andfreeze-thaw resistance in Lactobacilli. J. Dairy Res. 67, 241e247.

Grogan, D.W., Cronan, J.E., 1986. Characterization of Escherichia coli mutantscompletely detective in synthesis of cyclopropane fatty acids. J. Bacteriol.166, 872.

Grogan, D., Cronan, J., 1997. Cyclopropane ring formation in membrane lipids ofbacteria. Microbiol. Mol. Biol. Rev. 61, 429e441.

C. Montanari et al. / Food Microbiology 27 (2010) 493e502502

Guerzoni, M.E., Lanciotti, R., Cocconcelli, P.S., 2001. Alteration in cellular fatty acidcomposition as a response to salt, acid, oxidative and thermal stresses inLactobacillus helveticus. Microbiology 147, 2255e2264.

Guillot, A., Obis, D., Mistou, M.Y., 2000. Fatty acid membrane composition andactivation of glycine-betaine transport in Lactococcus lactis subjected to osmoticstress. Int. J. Food Microbiol. 55, 47e51.

Hammes,W.P., Brandt, M.J., Francis, K.L., Rosenheim, J., Seitter, M.F.H., Vogelmann, S.A.,2005.Microbial ecologyof cereal fermentations. Trends FoodSci. Technol.16, 4e11.

Hanus, L.O., Goldshlag, P., Dembitsky, V.M., 2008. Identification of cyclopropyl fattyacids in walnut (Juglans regia L.) oil. Biomed. Pap. Med. Fac. Univ. PalackyOlomouc Czech Repub. 152, 41e45.

Härtig, C., Loffhagen, N., Babel, W., 1999. Glucose stimulates a decrease of the fattyacid saturation degree in Acinetobacter calcoaceticus. Arch. Microbiol. 171,166e172.

Heipieper, H.J., Diefenbach, R., Keweloh, H., 1992. Conversion of cis unsaturated fattyacids to trans, a possible mechanism for the protection of phenol-degradingPseudomonas putida P8 from substrate toxicity. Appl. Environ. Microbiol. 58,1847e1852.

Huang, C.C., Smith, C.V., Glickman, M.S., Jacobs Jr., W.R., Sacchettini, J.C., 2002.Crystal structures of mycolic acid cyclopropane synthases from Mycobacteriumtuberculosis. J. Biol. Chem. 277, 11559e11569.

Johnsson, T., Nikkilä, P., Toivonen, L., Rosenqvist, H., Laakso, S., 1995. Cellular fattyacid profiles of Lactobacillus and Lactococcus strains in relation to the oleic acidcontent of the cultivation medium. Appl. Environ. Microbiol. 62, 4497e4499.

Johnston, N.C., Baker, J.K., Goldfine, H., 2004. Phospholipids of Clostridium per-fringens: a re-examination. FEMS Microbiol. Lett. 233, 65e68.

Kajiwara, S., Shirai, A., Fujii, T., Toguri, T., Nakamura, K., Ohtaguchi, K., 1996. Poly-unsaturated fatty acid biosynthesis in Saccharomyces cerevisiae: expression ofethanol tolerance and the FAD2 gene from Arabidopsis thaliana. Appl. Environ.Microbiol. 62, 4309e4313.

Kaneda, T., 1977. Fatty acids of the genus Bacillus: an example of branched-chainpreference. Bacteriol. Rev. 41, 391e418.

Kates, M., 1964. Bacterial lipids. In: Paleotti, R., Kritchevsky, D. (Eds.), Advances inLipid Research. Academic Press, London, pp. 17e90.

Keweloh, H., Heipieper, H.J., 1996. Trans unsaturated fatty acids in bacteria. Lipids31, 129e137.

Kim, B.H., Kim, S., Kim, H.G., Lee, I.S., Park, Y.K., 2005. The formation of cyclopropanefatty acids in Salmonella enterica serovar Typhimurium. Microbiology 151,209e218.

Kline, L., Sugihara, T.F., 1971. Microorganisms of the Sanfrancisco sourdough breadprocess. II. Isolation and characterization of bacterial species responsible for thesouring activity. Appl. Microbiol. 21, 459e465.

Law, J.H., 1971. Biosynthesis of cyclopropane rings. Acc. Chem. Res. 4, 199e203.Loffeld, B., Keweloh, H., 1996. Cis/trans isomerization of unsaturated fatty acids as

possible control mechanism of membrane fluidity in Pseudomonas putida P8.Lipids 31, 811e815.

MacDonald, P.M., Sykes, B.D., McElhaney, R.N., 1985. Florine-19 nuclear magneticresonance studies of lipid fatty acyl chain order and dynamics in Acholeplasmalaidlawii B membranes. A direct comparison of the effects of cis and transcyclopropane ring and double-bond substituents on orientational order.Biochemistry 24, 4651e4659.

Magnuson, K., Jackowski, S., Rock, C.O., Cronan Jr., J.E., 1993. Regulation of fatty acidbiosynthesis in Escherichia coli. Microbiol. Rev. 57, 522e542.

Muñoz-Rojas, J., Bernal, P., Duque, E., Godoy, P., Segura, A., Ramos, J., 2006.Involvement of cyclopropane fatty acids in the response of Pseudomonas putidaKT2440 to freeze-drying. Appl. Environ. Microbiol. 72 (1), 472e477.

Partanen, L., Marttinen, N., Alatossava, T., 2001. Fats and fatty acids as growthfactors for Lactobacillus delbrueckii. Syst. Appl. Microbiol. 24, 500e506.

Patrignani, F., Iucci, L., Belletti, N., Gardini, F., Guerzoni, M.E., Lanciotti, R., 2008.Effects of sub-lethal concentrations of hexanal and 2-(E)-hexenal on membranefatty acid composition and volatile compounds of Listeria monocytogenes,Staphylococcus aureus, Salmonella enteritidis and Escherichia coli. Int. J. FoodMicrobiol. 123, 1e8.

Ratledge, C., Wilkinson, S.G., 1989. In: Ratledge, C., Wilkinson, S.G. (Eds.), MicrobialLipids. Academic Press, London, pp. 525e566.

Rizzo, A.F., Korkeala, H., Mononen, I., 1987. Gas chromatography analysis of cellularfatty acids and neutral monosaccharides in the identification of lactobacilli.Appl. Environ. Microbiol. 53 (12), 2883e2888.

Russell, N.J., 1984. Mechanisms of thermal adaptation in bacteria: blueprints forsurvival. Trends Biochem. Sci. 9, 108e112.

Russell, N.J., Evans, R.I., Ter Steeg, P.F., Hellemons, J., Verheul, A., Abee, T., 1995.Membranes as a target for stress adaptation. Int. J. Food Microbiol. 28, 255e261.

Sallustio, S., Galantina, L., Gente, G., Masci, G., La Mesa, C., 2004. Hydrophobicallymodified pullulans: characterization and physicochemical properties. J. Phys.Chem. B. 108, 18876e18883.

Serrazanetti, D.I., 2009. Effects of acidic and osmotic stresses on flavour compoundsand gene expression in Lactobacillus sanfranciscensis Bb12. PhD Thesis in FoodBiotechnology, Università degli Studi di Teramo, Italy.

Shabala, L., Ross, T., 2008. Cyclopropane fatty acids improve Escherichia coli survivalin acidified minimal media by reducing membrane permeability to Hþ andenhanced ability to extrude Hþ. Res. Microbiol. 159, 458e461.

Smittle, R.B., Gilliland, S.E., Speck, M.L., Walter jr., W.M, 1974. Relationship of cellularfatty acid composition to survival of Lactobacillus bulgaricus in liquid nitrogen.Appl. Environ. Microbiol. 27, 738e743.

von Stokar, W., 1956. Der ursprung unseres hausbrotes. Brot Gebäck 10, 11e16.Streit, F., Delettre, J., Corrieu, G., Béal, C., 2008. Acid adaption of Lactobacillus del-

brueckii subsp. bulgaricus induces physiological responses at membrane andcytosolic levels that improves cryotolerance. J. Appl. Microbiol. 105, 1071e1080.

Teixeira, H., Goncalves, M.G., Rozes, N., Ramos, A., San Romao, M.V., 2002. Lacto-bacillic acid accumulation in the plasma membrane of Oenococcus oeni:a response to ethanol stress? Microb. Ecol. 43, 146e153.

Wang, Y., 2005. Fermentation pH and temperature influence in cryotolerance ofLactobacillus acidophilus RD758. J. Dairy Sci. 88, 21e29.

Welch, D.F., 1991. Applications of cellular fatty acid analysis. Clin. Microbiol. 4,422e438.

Wouters, J.A., Jeynov, B., Rombouts, F.M., de Vos, W.M., Kuipers, O.P., Abee, T., 1999.Analysis of the role of 7 kDa coldshock proteins of Lactococcus lactis MG1363 incryoprotection. Microbiology 145, 3185e3194.

Wouters, J.A., Frenkiel, H., de Vos, W.M., Kuipers, O.P., Abee, T., 2001. Cold shockproteins of Lactococcus lactis MG1363 are involved in cryoprotection and in theproduction of cold-induced proteins. Appl. Environ. Microbiol. 67, 5171e5178.

Zhao, Y., 2002. Cyclopropane Fatty Acid Derivatives: Biosynthesis, Function andTotal Synthesis. Personal Report.

Zhao, W., Li, H., Wang, H., Li, Z., Wang, A., 2009. The effect of acid stress treatmenton viability and membrane fatty acid composition of Oenococcus oeni SD-2a.Agric. Sci. China 8, 311e316.