clin ml leaders diagnosis, pathogenesis of bacterial infections … · bacterial pathogens. rather...

I Clin Pathol: Mol Pathol 1996;49:M1-M7

Leaders

Diagnosis, epidemiology and pathogenesis ofbacterial infections in the molecular era

S Sethi, T F Murphy, K L Klingman

IntroductionThe rapid development of techniques in mo-lecular biology over the past 20 years hasushered in a new era in microbiology and in-fectious diseases. The development of thesepowerful new methods has already had a pro-found impact in the clinical microbiologylaboratory and in research on the molecularepidemiology and pathogenesis of infectiousdiseases. The breadth of these advances willcontinue to expand at an even greater rate asnew tools are applied to the study of old prob-lems as well as to the study ofthe ever increasingnumber of emerging infectious diseases.The purpose of this review is to present

an overview of some of the applications ofmolecular techniques to the study of humanbacterial pathogens. Rather than dealing withdetails of methodology, this review will focuson the impact of new technology on clinicalapplications and research in the diagnosis, epi-demiology and pathogenesis of infectious dis-eases. While the main focus will be humanbacterial pathogens, it should be noted that thestudy of all microbes has been affected byapplication of new approaches. The selectedexamples which are discussed below do notrepresent an exhaustive list but illustrate im-portant areas in which the use of moleculartools has been most productive.

Diagnosis of bacterial infectionsFor the past 100 years, the laboratory diagnosisof infectious diseases has relied on directmethods such as bacterial culture. These welldescribed techniques can be used in most clin-ical laboratories. More recently, immunoassayssuch as immunofluorescence, immunoblot andenzyme linked immunosorbent assay have be-come useful adjuncts in the clinical laboratory.These techniques, usually used in the settingof large tertiary health care centres, facilitatethe aetiological diagnosis of infectious diseasesin many situations.

ADVANTAGES OF MOLECULAR BASEDTECHNIQUESThere are two significant limitations to theabove mentioned methods. Firstly, culturetechniques are limited because some pathogens

require a lengthy period oftime to be identified,exhibit similar characteristics to non-patho-genic species, thereby necessitating specialisedtesting, or have fastidious growth requirementsmaking detection by culture difficult. Secondly,the newer immunoassays are generally labourintensive or require expensive equipment, orboth. These two factors account for difficultyin applying these diagnostic methods in smallerhospitals, clinics, less developed countries, orto the field."2 Time is a crucial variable in anydiagnostic test, irrespective of the centre atwhich it is performed. Timely determinationof the specific aetiological agent ofan infectiousprocess is critical. Rapid, accurate diagnosis ofaetiology will improve the quality of health careand reduce health care costs in several ways:(1) earlier institution of appropriate therapywill reduce infection related morbidity andmortality; (2) early diagnosis in the field mayprevent the rapid dissemination ofan epidemic;(3) narrow spectrum antimicrobial agents canbe used earlier, decreasing expenditure and therisk of adverse drug events; and (4) invasivediagnostic procedures can be deferred oravoided.

Molecular biology based techniques offer thepotential for decreasing test time and broad-ening the scope of potential pathogens forwhich a clinical laboratory can test.2 The de-cision when to use a molecular technique suchas the polymerase chain reaction (PCR),restriction fragment length polymorphism(RFLP) analysis or nucleic acid probe tech-nology depends on the balance between thecost, difficulty, sensitivity, and specificity ofthe standard techniques compared with newermolecular techniques.

Molecular methods which will have broadapplicability in the diagnosis of infectious dis-eases include the use of nucleic acid probesand nucleic acid amplification with PCR.

Nucleic acid probesNucleic acid hybridisation techniques were thefirst application ofmolecular biology to be usedin the clinical microbiology laboratory. Nucleicacid probe test systems, such as GenProbe, arein widespread use for the diagnosis of sexuallytransmitted diseases from clinical samples, asculture confirmation assays for mycobacterialdiseases, and for the detection of Legionella

Division of Pulmonaryand Critical CareMedicine, StateUniversity of NewYork at BuffaloS Sethi

Division of InfectiousDiseases of theDepartment ofMedicineT F MurphyK L Klingman

Department ofMicrobiologyT F Murphy

Department ofVeterans AffairsMedical Center,Medical Research 151,3495 Bailey Avenue,Buffalo, NY 14215,USAT F MurphyK L Klingman

Correspondence to:T F Murphy MD.Accepted for publication26 September 1995

Ml

on August 1, 2020 by guest. P

rotected by copyright.http://m

p.bmj.com

/C

lin Mol P

athol: first published as 10.1136/mp.49.1.M

1 on 1 February 1996. D

ownloaded from

http://mp.bmj.com/

Sethi, Murphy, Klingman

from clinical specimens. Probe based test sys-tems for assaying various clinical samples havebeen described. These include studies of: spu-tum and bronchoalveolar lavage for the pres-ence ofMycobacterium tuberculosis, Mycoplasma,and other respiratory pathogens; urine for thepresence of urinary tract bacterial pathogens;stool for the presence of enteric pathogens; andblood and blood products for bacteria andother blood borne pathogens such as Babesiaand malaria."4 Some ofthese, such as the myco-bacterial probes for use in sputum samples,will become standard. This technology also hasgreat potential for use in outbreak situationswhere laboratories must be set up rapidly inless developed areas as minimal equipment isrequired for many probe based systems.'These test systems are designed to be used

either directly on clinical samples for diagnosisor to confirm culture results. The test usuallyinvolves a DNA or RNA oligomer (usually10-30 bases long) labelled with either a radio-isotope, an enzyme, or a chemiluminescentsubstrate. The probes are hybridised either insolution or on a nylon membrane after DNAor RNA has been extracted from the clinicalspecimen (for example, sputum, cerebrospinalfluid or leucocytes) or culture material. Theprobe must be specific for the organism beingtested and not bind to human DNA or RNAor other contaminating substances (such asproteins) in the sample. The benefits of thesetest systems are that they can be more sensitiveand specific than culture detection methods,and faster than culture based methods. Al-though individual assays may be more ex-pensive than traditional culture based methods,the fast turnaround time and the higher sensi-tivity translate into cost savings in the laborat-ory and in patient care. Invasive testingprocedures can be avoided, patients can betreated earlier in the course of their illness andthey can be treated with specific rather thanempirical drug regimens.'

In situ hybridisation is essentially a form ofDNA or RNA probe analysis carried out ontissue histopathological specimens rather thanon nylon membranes or in solution. In mostclinical situations this involves the applicationof a labelled oligomer to formalin fixed, paraffinwax embedded tissues or cytological speci-mens. These techniques are used clinically todetect a variety of viral, bacterial and fungalpathogens in tissue samples."' These methodshave been most useful in the detection of in-fection by human papilloma virus (HPV) andin advancing our understanding of the epi-demiology and pathogenesis of HPV, for whichthere is no culture technique yet available. Insitu hybridisation techniques have also beenimportant in linking Epstein-Barr virus (EBV)with various lymphoproliferative disorders andnasopharyngeal carcinoma. Because these tech-niques are more sensitive than routine histo-pathological stains, they are increasinglyimportant in the diagnosis of infections in im-munosuppressed subjects.' In situ hybrid-isation with the addition of PCR technologywill further enhance the sensitivity of the de-tection of pathogens in fixed tissue samples.5

Nucleic acid amplificationFor several clinical infections, the currentlyavailable diagnostic tests have limitations andthe change to a molecular based technique willbe rapid once sufficiently reliable tests becomeavailable. DNA and RNA amplification tech-niques use molecular methods to accomplishwhat older culture methods do; they amplify aprocess to a point where it can be detected.Culture media in Petri plates biologically amp-lifies one bacterium until it forms a colonywhich can be detected visually on the plate.PCR uses enzymatic duplication and detectionmethods to amplify one molecule of DNA.26Nucleic acid amplification systems do not re-quire an organism to be viable, one reason forits greater sensitivity over culture systems. Amajor drawback of PCR based systems is am-plification of contaminating, irrelevant DNA.'PCR is most likely to supplant current

methods in situations in which the diagnosisrelies on cultures of microbes with fastidiousgrowth requirements (for example, Bordetellapertussis, Borrelia burgdorferi, Pneumocystis ca-riniz), when culturing is dangerous (Bacillusanthracis, HIV, Lassa fever), the organismgrows slowly (Mycobacteria),'67 or the culturemethod is not sensitive enough to detect thepathogen in the clinical sample (herpes simplexvirus in spinal fluid).8 PCR technology hasgreat potential for screening a single clinicalsample, such as spinal fluid, a swab from agenital ulcer, or sputum69 for multiple potentialpathogens. For example, a sputum sample orbronchoalveolar lavage fluid from a patientwith pneumonia could be screened for variousbacteria, fungi and viruses which are most likelyto be the aetiological agent of pulmonary in-fection. Quantitative PCR can also be used tofollow a therapeutic response to treatment ofinfections such as malaria, tuberculosis, HIV,and hepatitis C.2'6 PCR has become themethod of choice for the laboratory diagnosisof HIV infection in neonates and infants.1'

In some clinical settings the type of organismpresent may not be known. Ribosomal RNAprobes have proved to be useful in the detectionof bacterial DNA. Oligonucleotide probesbased on shared 16S rRNA sequences amongall bacterial species (eubacterial primers) havebeen developed. These are amplified by PCRin a sample. The nucleotide sequence of theamplified DNA can be used to determine theidentity of the agent in the sample. This tech-nique has been successfully used to discovernew pathogens, such as those causing Whip-ple's disease, cat-scratch fever, and humanEhrlichiosis.8" 12 As our understanding of ge-netics improves, one can envisage an algorithmofPCR and a battery of primers based on thesetypes of shared sequences.Another area in which PCR techniques will

be used diagnostically is in the determinationof resistance to available anti-infective drugtherapies. The application of these techniqueswill greatly enhance our ability to treat patientseffectively early in the course of their illness. Amajor impediment to the current widespreaduse of molecular biological techniques to deter-

M2



p.bmj.com

/C

lin Mol P



ownloaded from

http://mp.bmj.com/

Bacterial infection in the molecular era

Table 1 Characteristics of typing assays

Discriminatory abilityTypabilityReproducibilitySensitivityEase of application

mine drug sensitivity is a lack of understandingof the genetic basis of drug resistance.3Although PCR in its various forms has great

promise, it probably will not totally replaceolder techniques in the clinical microbiologylaboratory. The major drawbacks are cost, thelabour intensive, technical nature of the workand lack, in some instances, of clear benefit toswitch from the older method to a newer PCRmethod. The lack of clear benefit is often re-lated to the observation that improved ana-lytical sensitivity of the PCR based test may notcorrelate with improved clinical sensitivity.' 7 Itmay not be clinically relevant if a small numberof live organisms or large numbers of non-viable organisms are present in a sample. Asmore PCR methods are used in the clinicallaboratory and automation improves, the costfactor will decline in importance.'

EpidemiologyMolecular epidemiology in relation to the studyof human disease is the use of molecular tech-niques to understand factors that influence therisk of developing a particular disease.'3 Mo-lecular epidemiology of infectious pathogensmainly deals with the differentiation of mi-crobial strains and species especially in tracingthe spread of a microbial species in an out-break.'" Wider application of molecular epi-demiology can permit an understanding of thepathogenesis and mechanisms of transmissionof pathogens."5 Insights into phylogenetic evol-ution of bacteria and species relatedness canalso be gained by molecular epidemiology.'6

CHARACTERISTICS OF TYPING SYSTEMSFor an epidemiological technique to be usefulin the study of infectious pathogens, it shouldhave certain characteristics; these are listed intable 1 and discussed below.'7 The superiorityof molecular or genotypic techniques over tra-ditional phenotypic techniques (for example,biotyping, phage typing, morphology) with re-gard to these characteristics has been clearlyestablished.

Discriminatory ability is the ability of anassay to discriminate between species and,more importantly, between clonally related butdistinct strains of a species. This is the mostvital characteristic of an epidemiological typingtechnique. Genotypic techniques have greaterdiscriminatory ability than phenotypic tech-niques.

Typability is the proportion of strains whichcan be typed by a particular typing method.Phenotypic methods depend on growth of thepathogen in vitro and expression of a particularcharacteristic. Thus, fastidious pathogens or

Table 2 Classificaton of molecular methods used inbacterial epidemiolgy

A. Plasmid analysisB. Restriction enzyme digestion (RFLP)

plasmid DNAchromosomal DNA

C. DNA amplification (PCR)of specific geneswith repetitive DNA element primers (Rep-PCR)with arbitrary primers (randomly amplifiedpolymorphic DNA or RAPD)with ribosomal RNA primers (PCR ribotyping)

D. Combination of B and CRFLP analysis of PCR products

pathogens lacking the relevant phenotypic char-acteristics are non-typable by the phenotypicmethods. Genotypic methods vary with respectto their typing ability. Plasmid DNA basedmethods often have low typability as somestrains may lack plasmids. Chromosomal DNAmethods have very high rates of typability be-cause ofthe universal presence ofchromosomalDNA and restriction sites.

Reproducibility is the ability of an assay togive the same result for a particular samplewhen performed many times. Before a tech-nique can be applied for general use, re-producibility in several laboratories needs tobe addressed. Phenotypic characteristics ofpathogens are notoriously variable in vitro andwith passage. The genotype is stable and there-fore, molecular methods have a distinct ad-vantage with regard to reproducibility.

Sensitivity of a molecular method is more ofan issue when the technique is used for diag-nosis. PCR based techniques can detect verylow concentrations of pathogens in clinicalsamples. If species specific primers are used forPCR, diagnosis and typing could be ac-complished in a single step. This is especiallyuseful for fastidious pathogens.

Ease of application is the factor governingwhether a typing method will remain confinedto research laboratories or become widely usedin clinical laboratories. PCR based methodsare extremely promising as they are easily ap-plied and can be used to process a large numberof specimens simultaneously. Problems of con-tamination by extraneous DNA is a limitationof these methods.

CLASSIFICATION OF MOLECULAR METHODSVarious typing methods have been developedfor and applied to epidemiological studies ofmicrobial pathogens. A simple classification ofmethods is presented in table 2.A DNA fingerprint is developed by each of

these methods. This DNA fingerprint differsamong strains because of polymorphism-thatis, length differences. Agarose gel electro-phoresis permits size fractionation of the DNAand detection of this polymorphism. The agar-ose gel can be stained and the banding patternof DNA studied directly. ' Alternatively, theDNA can be transferred to a nylon membraneand hybridised with DNA probes of specificsequences. 19 20

In the next section, two of these methodsas they have been applied to mucosal Gramnegative bacterial pathogens will be discussed.

M3



p.bmj.com

/C

lin Mol P



ownloaded from

http://mp.bmj.com/


RFLP ANALYSIS OF GENOMIC DNARFLP analysis (also known as restrictionendonuclease analysis) is one of the earliestmolecular methods to be applied to theepidemiological study of microbial pathogens.The organism is cultivated and chromosomalDNA is isolated and digested with restrictionenzymes. The resulting mixture of DNA frag-ments is separated by agarose gel electrophoresisto produce the DNA fingerprint.

In the early applications of this technique therestriction enzymes used were frequent DNAcutters whose recognition site is a 4-6 basepair DNA sequence. Therefore, the patternproduced has a large number of bands, makinginterpretation difficult and limiting discrim-inatory ability. A variation of this techniqueis RFLP analysis by pulsed field gel electro-phoresis (PFGE), also known as contourclamped homogeneous electric field (CHEF)electrophoresis. This method facilitates the sep-

aration of large fragments of DNA. Chro-mosomal DNA is purified and digested withinan agarose matrix. In this manner the chro-mosomal DNA is protected from the shearingforces inherent in conventional methods ofDNA purification. Restriction enzymes whichhave infrequent 8 base pair recognition sitesare used to produce large DNA fragments. Thebanding pattern produced on electrophoresis,therefore, is of a small number of large mo-

lecular weight bands which is easier to interpretand has a greater discriminatory ability.

Branhamella catarrhalis is a Gram negativemucosal pathogen which has become the thirdmost common cause of otitis media in childrenand acute exacerbations of chronic bronchitisin adults.2' There was no well established typingsystem available for this pathogen. On proteinelectrophoresis, the outer membrane proteinsof different strains of this bacteria are re-

markably similar. A high percentage of ,B-lac-tamase producing strains make antibioticresistance patterns insensitive for strain anal-ysis. Plasmids are present but not in greatenough numbers to allow plasmid analysis. Theepidemiology of this bacterium therefore hadnot been well established.Nosocomial outbreaks of B catarrhalis had

been studied successfully with conventionalRFLP analysis showing the epidemiological re-

latedness of the strains isolated and patient topatient transmission.22 However, as discussedabove, the large number ofbands seen precludeaccurate analysis of a large number of strainsisolated in an endemic setting. We had such a

collection ofB catarrhalis strains obtained frompatients with bronchiectasis over several years.PFGE distinguished the 37 strains accuratelyinto 13 classes and established the dynamicsof colonisation by B catarrhalis in this patientpopulation. 8

REPETITIVE DNA SEQUENCE PCR TYPING

Amplification in vitro of specific portions ofDNA by PCR has found wide applications inmolecular epidemiology of pathogenic or-

ganisms. Initial applications ofPCR used prim-ers based on species specific DNA sequences

followed by sequencing, RFLP or hybridisationof the PCR products. These techniques werelaborious and the primers could not be usedwith multiple bacterial species. Recent de-velopments, such as Rep-PCR,"6 RAPD23 andPCR ribotyping,24 use primers based on DNAsequences which are present in many speciesof bacteria, while still retaining their ability todistinguish the various strains within a species.These sequences are repetitive-that is, presentin multiple copies per genome but not constantin their position and frequency among variousstrains. Thus, the PCR amplifies genomic DNApresent between the repetitive sequences; thesize and number of these PCR products varyamong strains. On agarose gel electrophoresis,the PCR products produce a DNA fingerprint.Repetitive extragenic palindromic (REP) ele-ments and enterobacterial repetitive intergenicconsensus (ERIC) sequences are examples ofrepetitive sequences which have been used asprimers for typing.

Non-typable Haemophilus influenzae (NTHI)is a common mucosal pathogen in otitis mediaand chronic bronchitis.2' Several typing systemsutilising phenotypic characteristics-for ex-ample, biotyping and outer membrane proteinpatterns, have been developed. These systemshave drawbacks such as lack of reproducibility,phenotypic variation, and poor discriminatoryability. An efficient, reproducible typing systemwould be invaluable for the understanding ofthe dynamics of colonisation and pathogenesisof chronic infections by NTHI in chronic res-piratory diseases such as chronic obstructivepulmonary disease (COPD) and cystic fibrosis.Van Belkum et al25 used primers based on

ERIC sequences in a repetitive DNA sequencePCR assay to study 40 strains of NTHI frompatients with cystic fibrosis and COPD. Theyreliably and reproducibly distinguished strainswith the DNA fingerprints obtained. Whencompared with RFLP analysis, the PCR fin-gerprints were equally discriminatory and muchsimpler to perform and interpret. Ifthese resultsare substantiated with a larger number ofstrainsfrom varied epidemiological settings, an ad-equate typing system for NTHI could be es-tablished.The research potential of molecular epi-

demiology is well established. Genotypicmethods, though superior to phenotypicmethods, have limitations. Each molecularmethod examines the microbial genome in adifferent manner. Thus, when differentmethods are applied to a population of strains,varied results can be obtained.'9 The goldstandard among these methods has yet to bedefined. Interpretation of patterns obtainedwith genotypic techniques lacks stand-ardisation. What are the limits of similarity ina banding pattern which defines clonally relatedstrains versus different clones? These limits aresomewhat arbitrary and vary among in-vestigators. In general, the RFLP methods arecumbersome, need expensive equipment andare difficult to apply to large numbers of strains.The PCR based methods do not have thesedisadvantages but are prone to contaminationand need strict quality controls. Of the two

M4



p.bmj.com

/C

lin Mol P



ownloaded from

http://mp.bmj.com/


groups of methods, the PCR based methodsare more promising for application in clinicalmicrobiology laboratories. Cost effectiveness ofthese methods in a clinical setting needs to beaddressed.

PathogenesisThe application of molecular approaches to thestudy of bacterial infections has advanced theunderstanding of mechanisms of pathogenesisover the past decade. The ability to characterisethe molecular interaction between host andpathogen is critical for understanding patho-genesis so that better methods of preventionand treatment can be designed.

ADHERENCE OF BACTERIAL PATHOGENS TO HOSTCELLSThe initial step in the pathogenesis of bacterialinfections is often colonisation of a mucosalsurface. In order for colonisation to occurmolecules on the bacterial cell surface (ad-hesins) must bind to molecules on the epithelialcell surface (receptors). This adhesin-receptorinteraction partly accounts for the host andorgan specificity of bacterial infection. For ex-ample, bacterial species which are exclusivelyhuman pathogens (for example, Neisseriagonorrhoeae, H influenzae, B catarrhalis) mayderive their host specificity from the interactionof adhesin molecules with receptors which areunique to human epithelial cells. Adhesin-receptor interactions probably partly accountfor tissue specificity as well. Respiratory tractpathogens express adhesins which bind re-ceptors on epithelial cells ofthe respiratory tractwhereas genitourinary tract pathogens expressadhesins which bind receptors on urogenitalepithelial cells. Elucidating the mechanismsand specificity of these molecular interactionsis important for understanding the patho-genesis of bacterial infection.

Molecular studies of adhesins ofNTHI havehelped to elucidate the molecular mechanismsof adherence to human respiratory epithelialcells. The role of a high molecular weight outermembrane protein (HMW-1) in adherence tohuman cells was investigated.26 A kanamycinresistance cassette was placed into the HMW-1 gene which was in a plasmid vector. A strainof NTHI was transformed with the linearisedplasmid and a kanamycin resistant trans-formant which did not express the HMW-1was constructed. This isogenic mutant showednoticeably decreased adherence to Chang epi-thelial cells, a human cell line, compared withthe parent strain, which expressed HMW-1.26This series of experiments established thatHMW-1 is an important adhesin for NTHI.The relation between HMW-1 and other ad-hesins has been explored and the role of theadhesins in binding to other cell types hasbeen studied.27-29 The capability to constructisogenic mutants is a powerful method for elu-cidating the molecular mechanisms of ad-herence of pathogens to human cells.The differential adherence of bacteria to spe-

cific host cells plays a role in the pathogenesis of

infection. N gonorrhoeae expresses a family ofvariant outer membrane proteins (Opa) whichmediate adherence to human cells.303' The bac-terial chromosome contains 11 different copiesofthe gene. Each allele can be turned on and offindependently. Expression of some opa allelesenables the gonococcus to bind and invade epi-thelial cells. When the organism expresses otheropa alleles, the specificity for epithelial cells islost; instead, the bacterium interacts with poly-morphonuclear neutrophils.30 This host cellspecificity which is intrinsic to Opa proteins ex-pressed by the gonococcus confers importantvirulence properties on the organism. De-pending on the host tissue and the stage of in-fection, the bacterium expresses the Opa proteinwith appropriate host cell specificity. Un-derstanding these mechanisms is critical forcomprehending the pathogenesis ofinfection sothat new measures to treat and prevent bacterialinfections can be developed.

MOLECULAR STRUCTURE OF BACTERIAL OUTERMEMBRANE PROTEINSAn understanding ofthe molecular architectureofGram negative outer membrane proteins hashad an important influence on approaches tothe study of pathogenesis of infections causedby these organisms. A variety of molecularapproaches has been applied for studying theantigenic structure of outer membrane pro-teins. These include the development ofmono-clonal antibodies, cloning outer membraneprotein genes, sequence determination andanalysis, recombinant expression of selectedregions ofprotein molecules, mutagenesis stud-ies, and others.An understanding of molecular structure of

surface molecules is also essential. For example,investigators have been studying the role ofNTHI in lower respiratory tract infection inadults with chronic bronchitis for decades.2' Aserological approach has been used to evaluatea potential role of NTHI in these infections.The development of an antibody response tothe putative pathogen has been accepted asevidence that the organism causes the disease.Several studies from different centres have usedthis approach. Some studies showed no anti-body response, some studies showed highertitres of antibodies in patients compared withcontrols but no correlation with clinical ex-acerbations, and some studies showed an anti-body response to NTHI following infection(reviewed in 21). These apparently conflictingresults defied explanation until more wasknown about the antigenic structure of theouter membrane of this bacterium.The molecular architecture of the P2 porin

protein of NTHI accounts for the apparentlyconflicting results of serological studies inpatients with chronic bronchitis. P2 is the mostabundant protein in the outer membrane andcontains an immunodominant, strain specificepitope."" Preliminary studies indicate thatthe predominant immune response to NTHIis directed against this strain specific im-munodominant epitope.34 The host is therebyprotected from infection by the same strain but

M5



p.bmj.com

/C

lin Mol P



ownloaded from

http://mp.bmj.com/


remains susceptible to infection by other strainsof the species. Therefore, serological studieswhich used laboratory strains of NTHI ratherthan the patient's homologous strain failed todetect the predominant immune response. Fur-thermore, an understanding of the antigenicstructure of the major outer membrane proteinsand the host immune response to specific epi-topes on these outer membrane proteins allowsone to explain why adults with chronic bron-chitis experience recurrent infections by thesame bacterial species in spite of having anintact immune system.Over the past decade much has been learned

about the outer membrane structure of Gramnegative bacteria. This work has led to theidentification of a common theme in the mo-lecular structure of porin molecules.35-37 Theseprotein molecules consist of a P pleated sheetwith 16 strands that traverse and remain largelyburied within the outer membrane. Inter-spersed within the relatively conserved back-bone of the protein are eight peptide loops thatare potentially exposed on the bacterial surface.These surface accessible loops show antigenicdiversity among strains of a species. This struc-ture has important implications for the un-derstanding of the pathogenesis of recurrentinfections. As the surface exposed loops of themajor porin protein show antigenic hetero-geneity among strains, an immune response tothis part of the molecule will be relatively strainspecific.These two examples (molecular mechanisms

of adherence and molecular structure of outermembrane proteins) are just two of numerousobservations in which molecular approacheshave enabled observations regarding patho-genesis ofbacterial infections to be made. Stud-ies of the molecular mechanisms of bacterialpathogenesis promises to be a productive areaof investigation. The elucidation of these mech-anisms will lead directly to the development ofnew therapeutic and prophylactic modalitiesfor a variety of bacterial infections.

SummaryThe rapid progress in the development ofpowerful techniques in molecular biology overthe past 20 years has had an enormous impacton the disciplines of infectious diseases andmicrobiology. Our understanding of humanbacterial pathogens has undergone significantadvances, particularly in diagnosis, epidemi-ology and pathogenesis. Nucleic acid probesand amplification by PCR have forever changedthe laboratory diagnosis of many bacterial in-fections by providing specific, accurate andrapid means to identifyDNA from pathogens inclinical samples. The application of moleculartechniques has led to the development oftypingsystems based on the genetic structure ofbacterial pathogens. This has provided in-vestigators with highly effective tools for study-ing epidemiology, transmission, evolutionaryrelations, and pathogenesis of bacterial patho-gens. The ability to isolate and man-ipulate bacterial genes has resulted in a virtual

explosion of new information on molecularmechanisms of bacterial pathogenesis. Theseimportant advances in diagnosis, epidemiologyand pathogenesis will lead directly to bettermeans of treatment and prevention of a varietyof bacterial infections in the next decade.

1 Naber SP. Molecular pathology-diagnosis of infectiousdisease. NEnglJMed 1994;331:1212-15.

2 Persing DH. In vitro nucleic acid amplification techniques.In: Persing DH, Smith TF, Tenover FC, White TJ, eds.Diagnostic molecular nicrobiology: principles and applications.Rochester: Mayo Foundation, 1993:51-87.

3 Tenover FC, Unger ER. Nucleic acid probes for detectionand identification of infectious diseases. In: Persing DH,Smith TF, Tenover FC, White TJ, eds. Diagnostic nmolecularmticrobiology: principles atnd applications. Rochester: MavoFoundation, 1993:3-25.

4 Weiss JB. DNA probes and PCR for diagnosis of parasiticinfections. Cliii Microbiol Rev 1995;8: 113-30.

5 Mies C. Molecular biological analysis of paraffin-embeddedtissues. Humwl Pathol 1994;25:555-60.

6 Dale B, Dragon EA. Polymerase chain reaction in infectiousdisease diagnosis. Lab Med 1994;25:637-41.

7 Swaminathan B, Matar GM. Molecular typing methods.In: Persing DH, Smith TF, Tenover FC, White TJ, eds.Diagnostic nolecular nmicrobiology: principles and applications.Rochester: Mayo Foundation, 1993:26-50.

8 Whitley RJ, Lakeman F. Herpes simplex virus infections ofthe central nervous system: therapeutic and diagnosticconsiderations. Clin Infect Dis 1995;20:414-20.

9 Radstrom P, Backman A, Qian N, Kragsbjerg P, PahlsonC, Olcen P. Detection of bacterial DNA in cerebrospinalfluid by an assay for simultaneous detection of Neisseriameningitidis, Haemophilus influenzae, and streptococciusing a seminested PCR strategy. 7 Clini Microbiol 1994;32:2738-44.

10 De Rossi A, Ometto L, Zanotto C, Salvatori F, Masiero S,Mammano F, et al. Pediatric HIV-1 infection: advancesand perspectives in diagnosis and prognosis. AntibiotChemiother 1994;46:5-17.

11 Lowsky R, Archer GL, Fyles G, Minden M, Curtis J,Messner H, et al. Brief report: diagnosis of Whipple'sdisease by molecular analysis of peripheral blood. N E??gl_JMed 1994;331:1343-6.

12 Tompkins LS, Tenover F, Arvin A. New technology in theclinical microbiology laboratory: what you always wantedto know but were afraid to ask. 7 Infect Dis 1994;170:1068-74.

13 McMichael AJ. Molecular epidemiology: new pathway ornew travelling companion?Amj.Epidenmiol 1994;140: 1-11.

14 Jarvis WR. Usefulness of molecular epidemiology for out-break investigations. Infect Co??trol Hosp Epidemiol 1994;15:500-3.

15 Romling U, Fiedler B, Bobhammer J, Grothues D, GreipelJ, von der Hardt H, et al. Epidemiology of chronic Pseudo-monas aeruginosa infections in cystic fibrosis. 7 Infect Dis1994;170: 1616-21.

16 Versalovic J, Koeuth T, Lupski JR. Distribution of repetitiveDNA sequences in eubacteria and application to fin-gerprinting of bacterial genomes. Nuicleic Acids Res 199 1;19:6823-31.

17 Versalovic J, Woods CR, Jr, Georghiou PR, Hamill RJ,Lupski JR. DNA-based identification and epidemiologictvping of bacterial pathogens. Arch Pathol Lab Med 1993;117:1088-98.

18 Klingman KL, Pye A, Murphy TF, Hill SL. Dynamics ofrespiratory tract colonization by Moraxella (Branhamella)catarrhalis in bronchiectasis. AmJ_ Respir Crit Care Med1995;152: 1072-8.

19 Thorisdottir AS, Carias LL, Marshall SH, Green M, ZervosMJ, Giorgio C, et al. IS6770, an enterococcal insertion-like sequence useful for determining the clonal relationshipof clinical enterococcal isolates. _7 Infect Dis 1994;170:1539-48.

20 Shafer RW, Small PM, Larkin C, Singh SP, Kelly P, SierraMF, et al. Temporal trends and transmission pattemsduring the emergence of multidrug-resistant tuberculosisin New York City: a molecular epidemiologic assessment._7 Infect Dis 1995;171:170-6.

21 Murphy TF, Sethi S. Bacterial infection in chronic ob-structive pulmonary disease. Aml Re' Respir Dis 1992;146:1067-83.

22 Patterson TF, Patterson JE, Masecar BL, Barden GE, Hier-holzer WJ Jr, Zervos MJ. A nosocomial outbreak of Bran-hamella catarrhalis confirmed by restriction endonucleaseanalysis. _7 Infect Dis 1988;157:996-1001.

23 Wong NA, Linton CJ, Jalal H, Millar MR. Randomly amp-lified polymorphic DNA typing: a useful tool for rapidepidemiological typing of Klebsiella pneumoniae. Epi-demiol Infect 1994;113:445-54.

24 Kostman JR, Alden MB, Mair M, Edlind TD, LiPuma JJ,Stull TL. A universal approach to bacterial molecularepidemiology by polymerase chain reaction ribotyping..7Infect Dis 1995;171:204-8.

25 Van Belkum A, Duim B, Regelink A, Moller L, Quint W,van Alphen L. Genomic DNA fingerprinting of clinicalHaemophilus influenzae isolates by polymerase chainreaction amplification: comparison with major outer-membrane protein and restriction fragment length poly-morphism. .7 Med Microbi'l 1994;41:63-8.

M6



p.bmj.com

/C

lin Mol P



ownloaded from

http://mp.bmj.com/


26 St Geme JW III, Falkow S, Barenkamp SJ. High-molecular-weight proteins of nontypable Haemophilus influenzaemediate attachment to human epithelial cells. Proc NatlAcad Sci USA 1993;90:2875-9.

27 St Geme JW III. The HMW1 adhesin of nontypeableHaemophilus influenzae recognizes sialylated glycoproteinreceptors on cultured human epithelial cells. Infect Immun1994;62:3881-9.

28 Noel GJ, Barenkamp SJ, St Geme JW III, Haining WN,Mosser DM. High-molecular-weight surface-exposed pro-teins of Haemophilus influenzae mediate binding tomacrophages. Jf Infect Dis 1994;169:425-9.

29 Bakaletz LO, Barenkamp SJ. Localization ofhigh-molecular-weight adhesion proteins of nontypeable Haemophilusinfluenzae by immunoelectron microscopy. Infect Immun1994;62:4460-8.

30 Kupsch E-M, Knepper B, Toshikazu K, Heuer I, MeyerTF. Variable opacity (Opa) outer membrane proteins ac-count for the cell tropisms displayed by Neisseriagonorrhoeae for human leukocytes and epithelial cells.EMBOJ 1993;12:641-50.

31 Vlrji M, Makepeace K, Ferguson DJP, Achtman M, MoxonER. Meningococcal Opa and Opc proteins: their role

in colonization and invasion of human epithelial andendothelial cells. Mol Microbiol 1993; 10:499-5 10.

32 Haase EM, Campagnari AA, Sarwar J, Shero M, WirthM, Cumming CU, et al. Strain-specific and im-munodominant surface epitopes of the P2 porin proteinof nontypeable Haemophilus influenzae. Infect Immun199 1;59: 1278-84.

33 Haase EM, Yi K, Morse GD, Murphy TF. Mapping ofbactericidal epitopes on the P2 porin protein of non-typeable Haemophilus influenzae. Infect Immun 1994;62:3712-22.

34 Troelstra A, Vogel L, van Alphen L, Eijk P, Jansen H,Dankert J. Opsonic antibodies to outer membrane proteinP2 of nonencapsulated Haemophilus influenzae are strainspecific. Infect Immun 1994;62:779-84.

35 Weiss MS, Abele U, Weckesser J, Welte W, Schiltz E, SchulzGE. Molecular architecture and electrostatic properties ofa bacterial porin. Science 1991;254:1627-30.

36 Nikaido H, Saier MH Jr. Transport proteins in bacteria:common themes in their design. Science 1992;258:936-42.

37 Jeanteur D, Lakey JH, Pattus F. The bacterial porinsuperfamily: sequence alignment and structure prediction.Mol Microbiol 1991;5:2153-64.

M7



p.bmj.com

/C

lin Mol P



ownloaded from

http://mp.bmj.com/

clin ml leaders diagnosis, pathogenesis of bacterial infections … · bacterial pathogens. rather...

Documents