Molecular and Cellular Probes (1987) 1, 3-14

REVIEW

Molecular diagnosis of infectious diseases by nucleic acidhybridization

R. P. Viscidi* and R. G. YolkentInfectious Diseases Division, *Department of Medicine and EudowoodDivision of Infectious Diseases, tDepartment of Pediatrics, The Johns

Hopkins University School of Medicine, Baltimore, Maryland 21205, USA

(Received 5 November 1986, Accepted 17 December, 1986)

Diagnostic microbiology has traditionally relied on the cultivation of an infecting agent inan in vitro system. However, the limitations of these methods have led to thedevelopment of alternative, molecular techniques which can detect specific microbialproteins or nucleic acids directly in body fluids . One such method, nucleic acidhybridization, has many attractive features including the potential for high sensitivity andspecificity, the detection of a chemically stable analyte and the use of uniform reagents .Nucleic acid hybridization reactions have been used to detect a variety of microbialpathogens in assay formats using extracted nucleic acids or tissue sections . However,wider application of these techniques is limited by the fact that current methods employradioisotopically labelled probes and cumbersome assay procedures . Solutions to theseproblems have been sought by the development of non-radioactive hybridizationtechniques which utilize enzymatic detection methods and by the development of rapidhybridization assays performed in solution . Further improvements in labelling techniques,assay formats and enzymatic detection methods should allow for the wider application ofnucleic acid hybridization reactions in diagnostic microbiology .

KEYWORDS: nucleic acid hybridization, diagnostic microbiology, non-isotopic hybridization,biotinylated probes .

Sensitive, specific and rapid diagnostic techniques for the detection of microbialagents in body fluids are important for the prevention, control and management ofinfectious diseases in clinical medicine as well as for the large-scale study of theepidemiology of infectious agents . Traditionally, the diagnosis of infectious diseaseshas relied on the direct microscopic examination of clinical specimens and on thecultivation of an infecting agent in an in vitro system . However, there are a numberof limitations to these techniques . Direct microscopy is often insensitive and non-specific. While potentially extremely accurate, cultivation is often insufficiently

Address correspondence to : Raphael P. Viscidi, The Johns Hopkins Hospital, Blalock, Room 1111, 600 NorthWolfe Street, Baltimore, MD 21205, USA . Tel : (301) 955-3150.

0890-8508/87/010003+12 $03 .00/0 .

© 1987 Academic Press Inc . (London) Ltd

4

R. P. Viscidi and R. H . Yolken

rapid to be clinically useful, especially in the case of viral agents and other agentsthat grow slowly in in vitro systems . The requirement for the maintenance of celllines limits diagnostic virology to central laboratories . In addition, cultivation inavailable cell lines is often not possible for several viral agents of medicalimportance such as rotavirus, Norwalk agent, hepatitis A and B viruses and Epstein-Barr virus . Similarly, some important bacterial pathogens such as Treponemapallidum and Mycobacterium leprae cannot be cultivated on artificial media . Formany bacteria, such as enteric and respiratory pathogens, recovery is impairedwhen cultivation is attempted in the presence of a competing normal microbialflora. In the not uncommon clinical situation where antibiotics have been adminis-tered, cultivation can be inadequate in identifying a microbial pathogen inantibiotic-containing body fluids. In addition, cultivation techniques cannot alwaysdistinguish virulent from avirulent strains of bacteria, as in the case of enterotoxinproducing strains of Escherichia coli or toxigenic strains of Staphylococcus aureus .The most important limitation of cultivation techniques is the time required for thedetection and identification of an infecting micro-organism . This time can rangefrom 1 or 2 days to several weeks, depending upon the nature of the micro-organism . These inherent disadvantages of traditional methods have provided theimpetus to develop more rapid and more practical diagnostic techniques forinfectious diseases .

Molecular techniques directed at the detection of specific microbial proteins,lipopolysaccharides or nucleic acids provide an alternative method for detectinginfecting agents directly in body fluids . For a molecular diagnostic technique asystem that supports the replication of an organism is not necessary for itsidentification. It is only necessary to identify a specific molecule in a reaction withan appropriate molecular probe. Since most infectious agents have highly specificantigens against which antibodies can be produced and since antigen-antibodyreactions can be completed and quantitated in a relatively short period of time, ithas been possible to develop specific and rapid immunoassays for the detection of awide range of infectious agents. During the past 2 decades a number of immunoas-say systems have attained widespread usage for the detection of microbial antigensfrom clinical specimens . These systems include solid phase radioimmunoassays, andenzyme immunoaassays as well as assay systems based on immunodiffusion,immuno-electrophoresis, agglutination and immunofluorescence technologies .While immunoassays offer a number of advantages for the practical detection ofinfecting agents in body fluids, there are a number of limitations inherent in the useof immunological reactions for microbial diagnosis . Most importantly, the sensitivityof immunoassay systems is limited by the kinetics of the antigen-antibody reaction,the limiting quantities of immunoreactants present in the reaction and the rate ofnon-specific immunological reactions' . For this reason there is a need for thedevelopment of microbial detection systems which are not limited by the kineticsand binding characteristics of antibodies .

One non-immunological technique which offers great promise for the detectionof microbial agents in body fluids is nucleic acid hybridization . Nucleic acidhybridization has many features which make it an attractive method for thedetection of micro-organisms under clinical conditions (Table 1) . The hybridizationof complementary strands of DNA or RNA is a kinetically efficient reaction due tothe large number of hydrogen bonds involved in base pairing and to the stabilizing

Molecular diagnosis of infectious diseases

5

Table 1 . Advantages of nucleic acid hybridization o diagnostic microbiology

I . High sensitivity due to the favourable kinetics of hybridization reactions .IL High specificity due to the sequence of bases composing the genetic code .

A . Specific sequences unique to a particular strain .B . Specific yet highly conserved sequences shared among strains of a particular species .C . Specific sequences related to a virulence factor of pathogenic strains .

Ill . Chemical stability and uniformity of the analyte .A . Detection of nucleic acids in immune complexes by proteolytic treatments .B . Release of intracellular and sequestered nucleic acids by extraction procedures .C . Ease and safety of handling clinical specimens .D. Uniform processing of clinical samples .

IV . Ability to detect latent infections in which nucleic acid is present but in which antigen is notexpressed .

V. Ease with which standardized reagents can be produced .A. Recombinant DNA techniques .B. Solid phase oligonucleotide synthesis technology .

effect of the co-operative 'stacking-free energy' between bases in the double helicalstructure2 . This allows for the possibility of detecting very small amounts ofmicrobial nucleic acids in clinical samples in a short period of time . Anotherimportant benefit of nucleic acid hybridization in diagnosis is its high specificity .Most microbial organisms contain some genetic sequences which are highlyconserved in antigenically diverse strains of the organism . By means of molecularcloning, DNA fragments with the desired specificity can be selected and utilized inthe hybridization procedures under carefully controlled conditions. In contrast, thespecificity of the antigen-antibody reaction often limits the utility of immunoassaysfor the detection of organisms displaying antigenic diversity . For example, no singleassay can detect all the serotypes of antigenically diverse enteroviruses . Althoughimmunoglobulin pools can be utilized, this approach lowers the concentration ofeach specific antibody and thus decreases the sensitivity of the assay system . Onthe other hand, it is likely that genetic probes for a highly conserved non-codingregion of the genome of a prototype strain will be capable of binding to nucleicacids from all types of enteroviruses . In addition, when a gene associated withvirulence is used as a probe, pathogenic and non-pathogenic strains of a particularmicrobial species can be distinguished in mixed cultures without the need forselective enrichment. This method has been utilized for the specific identification ofE. coli expressing heat labile or heat stable enterotoxins 3 . Also, probes for antibioticresistance genes can detect resistant bacteria without the need to cultivate them inthe presence of the antibiotic. When hybridization reactions are performed underappropriate conditions, a nucleic acid probe can discriminate between twosequences of 20 or more bases which differ by a single base . This high degree ofspecificity has the potential to improve the sensitivity of nucleic acid hybridizationreactions by limiting the rate of non-specific base pairing between the probe andother nucleic acid sequences present in clinical samples . in contrast, low affinityreactions of polyclonal or monoclonal anti-microbial antibodies with extraneousantigens present in high concentration in clinical samples often are the major factorlimiting the sensitivity of immunoassays for microbial antigens . In addition, hybridi-zation reactions can allow for the detection of a microbial pathogen in the presence

6

R. P. Viscidi and R. H . Yolken

of co-existing antibody . This antibody could occupy antibody binding sites on theantigen, thus making it difficult to detect by means of immunoassay procedures .However, hybridization assays are not influenced by the presence of antibody sincenucleic acid present in immune complexes would be released by the proteolytictreatments which are used in the assay procedure. In addition, immunoassaysrequire that the antigen be present in the specimen in a micro-environment that isfavourable to antigen-antibody interactions . This can be a problem in the case ofintracellular antigens which are difficult to extract from sites inaccessible to labelledantibodies without altering their antigenicity . In contrast, the chemical stability ofnucleic acids preserves their reactivity even when clinical samples are subjected toharsh extraction procedures . In addition, this chemical stability allows for the easyand safe handling and storage of clinical specimens under diverse environmentalconditions. For immunoassays, the optimal procedure for preparing a clinicalspecimen may differ for each test system depending on the chemical nature andstability of the target epitope. In contrast, the nucleic acids of all organisms arechemically alike and thus the same procedure can be applied to all samples . Thiswould greatly simplify assay protocols when tests for several different pathogensneed to be performed on the same clinical specimen . Finally, neither cultivation norimmunoassays allow for the diagnosis of latent infections in which nucleic acid ispresent but in which antigen is not expressed . However, nucleic acid hybridizationhas been used to demonstrate the integration of hepatitis B virus DNA into thegenome of liver cells from patients with chronic hepatitis or hepatocellularcarcinoma and the presence of human papillomavirus DNA has been demonstratedin laryngeal papillomas and adjacent laryngeal tissue 4,5 . In addition to theseadvantages, recent advances in recombinant DNA technology and solid phaseoligonucleotide synthesis make it feasible to produce large quantities of nucleic acidprobes with defined sequences . The availability of techniques for readily producinglarge quantities of standardized reagents represents a distinct advantage of nucleicacid hybridization reactions over current immunoassay systems, which dependlargely on the vagaries of animal immunizations or hybridoma formation for theproduction of reagent antibodies .

The most commonly used hybridization assay formats for determining thepresence of a defined nucleic acid in clinical specimens include spot hybridization,in which nucleic acids extracted from the specimen are applied directly tonitrocellulose membranes, and Southern blot analysis in which restriction endonuc-lease digested nucleic acid is electrophoresced in an agarose gel and thentransferred to nitrocellulose' ,'. In both techniques, the presence of a specific nucleicacid sequence immobilized on the nitrocellulose is determined by a subsequentreaction with a labelled probe composed of complementary sequences . In addition,double-binding 'sandwich' techniques have been developed for the detection ofnucleic acids in crude clinical samples' . The nucleic acid sandwich assays are basedon the use of two separate nucleic acid fragments . The sequences of both reagentsare complementary to the nucleic acid to be identified, but contain few sequencesin common. One of the fragments is immobilized onto nitrocellulose while the otherfragment is utilized as a soluble labelled probe . Only if the correct nucleic acid ispresent in the sample is a hybrid formed between solid phase and labelled probe,and only in such cases is a measurable signal generated (Fig. 1). Sandwichhybridization offers several advantages over the previous methods . The tedious and

Molecular diagnosis of infectious diseases

7

Filter DNA

Sample DNA

Sandwich hybrid

Fig. 1 . Principle of sandwich hybridization assay .

time-consuming steps involved in pretreating samples is unnecessary because thesample is kept in solution . More importantly, extraneous components of the samplesuch as proteins, lipids and mucopolysaccharides cannot interfere with the bindingof nucleic acids to the filter or cause non-specific binding of probe DNA . Finally, thesensitivity of hybridization assays can be improved in a sandwich format by thepotential to increase the concentration of nucleic acid reagents . This is easilyachieved with the immobilized nucleic acid because the binding capacity ofnitrocellulose is very high . Sandwich hybridization assays have also been adapted toa format in which the two probe fragments are allowed to form hybrids with thetarget nucleic acid in solution . One fragment labelled with biotin is used to bindspecific hybrids to a strepavidin-agarose support after the reaction . A secondfragment labelled with radio-iodine allows the recovered hybrids to be measured .Because the rate of hybridization reactions in solution is faster than the rate ofmixed phase reactions on nitrocellulose, the assay can be completed in as short aperiod of time as 2 h. In an evaluation of this assay format with a mock sample, thedetection limit for target nucleic acid was 1 . 7 pg and the sensitivity was not affectedby the addition of cell lysates or serum to the reaction mixture 9 . We have recentlydeveloped another assay format for the performance of rapid hybridization assaysin solution phase . A biotinylated DNA probe is reacted in solution with a samplecontaining an RNA virus with sequences complementary to the probe . The reaction

Labeled

Probe DNA

8

R. P . Viscidi and R . H. Yolken

mixture is then incubated on a microtitre plate coated with an anti-biotin antibody .Any hybrids formed between the DNA probe and the viral RNA sequences arebound to the microtitre plate and they are detected in a subsequent reaction withan enzyme-labelled monoclonal antibody specific for DNA-RNA hybrids (gift of DrRobert J. Carrico, Ames Division, Miles Laboratories, Inc ., Elkhart, IN). In preliminarystudies we have found that this assay format is more sensitive than dot hybridiza-tion assays for the detection of viral sequences in clinical samples and tissue culturesupernatants. Another commonly used hybridization technique is in situ hybridiza-tion which involves performing reactions directly on tissue sections or cytologicsmears"'. This method has the advantage of providing information about the cellularspecificity and localization of an infecting organism . In addition, it may be moresensitive than the macroscopic detection techniques when an infecting agent isnon-uniformly distributed in a tissue specimen .

Most hybridization reactions utilize radio-labelled nucleic acid probes . A varietyof methods have been devised for preparing such probes, but the most widelyutilized technique is nick translation". This technique involved the introduction ofnicks into double-stranded DNA by partial digestion with pancreatic DNAse andsubsequent repair of the nicked DNA with E . coli DNA polymerase I . This enzymeadds sequential nucleotide residues to the 3'-hydroxyl end of the nicked DNA andsimultaneously eliminates the adjacent residue at the 5' phosphoryl end . In thismanner, probes can be labelled to specific activities up to 10 8 CPM µg - ' withdeoxynucleotide .5'-[a- 32P1 triphosphates . The complementary sequences of double-stranded probes prepared by nick translation can hybridize with each other insolution prior to hybridization with the immobilized nucleic acid . This self-hybridization can decrease the sensitivity of hybridization reactions both byreducing the concentration of probe available for hybridization with immobilizednucleic acid and by creating large nucleic acid aggregates with unfavourablereaction kinetics . One solution to this problem involves the preparation of single-stranded probes by cloning in the single-stranded phage, M13 12 . An M13 phagecontaining a probe insert can then be hybridized with a small oligonucleotidecomplementary to phage sequences upstream from the insert. Extension of thissmall oligonucleotide away from the insert using the Klenow fragment of E . coli DNApolymerase I, which lacks the 5'-3' exonuclease activity, generates probes of highspecific activity in the presence of radioactive nucleotides . Since the extendedsequences do not traverse the insert, a partially duplex probe is generated whichcan hybridize with target sequences complementary to the single-stranded insert .Another approach involves the preparation of single-stranded RNA probes fromcloning vectors which contain an SP6 or T7 RNA polymerase promoter locatedimmediately upstream from a region of multiple cloning sites 13 . These RNApolymerases exhibit a stringent promoter specificity, work very efficiently in simplebuffers and initiate and terminate RNA synthesis at precise positions on the DNAtemplate. For transcription the recombinant DNA is linearized by a restrictionenzyme which cleaves downstream from both the promoter and the DNA insertion .RNA probes of high specific activity are synthesized by the polymerase whenlimiting concentrations of radioactive ribonucleotides are used in an in vitro run-offtranscription reaction . Single-stranded RNA probes are 8-10 times more sensitivethan nick-translated DNA probes for the detection of RNA target sequences in RNAblot hybridization assays and in situ tissue section hybridization 14

Molecular diagnosis of infectious diseases

9

Nucleic acid hybridization reactions have been used to detect a variety ofmicrobial pathogens . Dot hybridization techniques have successfully identifiedcytomegaloviruses in urine and huffy coat cells, hepatitis B virus in serum, Epstein--Barr virus in cultured lymphoid cell lines, rotavirus in stool specimens, varicella-zoster virus in swabs and aspirates from vesicular skin lesions, and adenovirus instools, urine and swabs from the nasopharynx6,,5--2o One of the earliest practicalapplications of nucleic acid hybridization in diagnostic virology was for thedetection of potato spindle tuber viroid in potato tubers" . In diagnostic bacteriol-ogy, dot hybridization assays have been used to identify the gonococcus in swabsfrom the male urethrea and enterotoxigenic E . coli in stool specimens and stoolisolates' ," . Leishmaniasis can be diagnosed by nucleic acid hybridization ofkinetoplast DNA in tissue touch blots of cutaneous lesions" . In addition, thetechnique of Southern blot analysis has been used to detect varicella-zoster virus insensory ganglia, and human papillomaviruses in laryngeal papillomas and adjacentlaryngeal tissue 5,24 . Hepatitis B virus DNA has been demonstrated in liver cells frompatients with chronic liver disease4,25, the chronic carrier state and hepatocellularcarcinoma . Similarly, Epstein-Barr virus can be detected in tissue from patients withBurkitt's lymphoma, central nervous system lymphoma, nasopharyngeal carcinomaand infectious mononucleosis 26 . In situ hybridization techniques have revealed thepresence of measles virus in brain cells from patients with subacute sclerosingpanencephalitis and hepatitis B virus in hepatocytes and bile duct epithelium frompatients with chronic active hepatitis 27,28 . Since Southern blot analysis providesevidence of the integration of viral nucleic acid into host DNA and in situhybridization provides information about localized tissue and cellular viral infection,these techniques have applications not only in the diagnosis of viral diseases butalso in understanding their biology .

The major obstacle to the wider application of nucleic acid hybridizationreactions in diagnostic microbiology is the fact that current techniques routinelyemploy 32P- or 35S-labelled polynucleotide probes and rely on the subsequentdetection of radioactivity . These probes have a short functional half-life and mustbe frequently prepared and standardized . The expense, radiation exposure andisotope disposal problems associated with radioactive probes are additional disad-vantages. Assays using radionucleotides have the disadvantage that, for maximalsensitivity, prolonged exposure of photographic plates are necessary, thus makingthem impractical for rapid diagnosis . One alternative to radioactivity would be theuse of enzyme detection methods . The extensive experience over the past decadewith enzyme immunoassays provides a basis for the application of enzymaticanalysis to macromolecular reactions . Enzyme detection methods would obviatemany of the problems associated with isotopic methods . Enzyme reagents eliminatethe biohazard of radioisotopes . These reagents are also stable under a variety ofstorage conditions for a prolonged period of time . More importantly, the magnifyingnature of the enzyme substrate reaction raises the possibility that enzyme detectionmethods might be more sensitive within a clinically useful period of time thanradioisotopic methods . For example, with high energy fluorescent and chemilumi-nescent substrates attamole (10 - ' 7 mole) quantities of enzyme can be detected .Thus, fluorimetric assays of B-d-galactosidase from E. coli and alkaline phosphatasefrom calf intestine with 4-methyl-umbell iferyl derivatives are 1000- to 5000-foldmore sensitive than the currently more widely used colorimetric assays with

10

R. P . Viscidi and R . H. Yolken

nitrophenyl derivatives 29,3o The assay of horse-radish peroxidase with the lumines-cent substrate, luciferin, is reported to be capable of measuring 0 .125 amol of thisenzyme" . Based on a calculation of the specific radioactivity of carrier-freeradioisotopes, Ishikawa concludes that the fluorimetric assay of B-d-galactosidaseand the luminescent assay of horse-radish peroxidase are potentially more sensitivethan the measurement of radioisotopes using a gamma counter or scintillationcounter32 . Thus, enzymatic labels might provide for sensitive, practical methods forthe detection of nucleic acids in clinical specimens .

Several non-radioactive nucleic acid hybridization techniques which utilizeenzymatic detection methods have recently been developed . One method involvesthe incorporation of uridine triphosphate or thymidine triphosphate analoguescontaining biotin into nucleic acid probes by a standard nick translation reaction 33

When lightly labelled (2-12 .5% base substitution), biotin containing polynucleo-tides exhibit denaturation and reassociation characteristics that allow for their useas hybridization probes . Biotin is an attractive affinity label because it bindsspecifically, rapidly and tenaciously (K d15 =10-15) to avidin, a glycoprotein derivedfrom egg white or strains of Streptomyces 34 . Thus, minute quantities of biotin can bedetected by reaction with an enzyme coupled to avidin . Biotinylated DNA probeshave been applied for in situ hybridization to map genes on Drosphila polytenechromosomes, to detect the presence of actin mRNA and to visualize genes inwhole mount metaphase chromosomes by electron microscopy 35-37 The techniquehas also been extended to the detection of target sequences immobilized onnitrocellulose filters and it is sufficiently sensitive to visualize 1-2 pg of target DNA 38.There are, however, a number of limitations inherent in the use of assay systemswhich rely on the incorporation of modified nucleotides into probes by means ofnick translation procedures . For example, the system requires the performance ofincorporation reaction which can be difficult to standardize and control whenmodified bases are utilized . Also the reaction kinetics of several probe-generatingenzymes are markedly slowed when modified bases are utilized . Another approachto the development of non-radioactive hybridization techniques involves thechemical modification of guanine residues using N-acetoxy-N-2-acetylaminofluor-ine or its 7-iodo derivative39. The modified nucleic acids are recognized by specificantibodies which can then be detected with enzyme-labelled antiglobulins . Hybridi-zation reactions with such probes can detect 2-5 pg of target sequences immobi-lized in nitrocellulose using alkaline phosphatase linked second antibodies . Thedisadvantages of this approach are that the chemical compound is carcinogenicand the specific antisera is not commercially available . Biotinylated probes have alsobeen prepared by a direct reaction between DNA or RNA and a biotin derivativecontaining a photo activatable aryl azide functional group .". With this procedurethe extent of modification is limited by the occurrence of cross-linking at highconcentrations of the derivatizing reagent. A number of investigators have des-cribed chemical or enzymatic methods for labelling the 3' or 5' phosphate ofnucleic acid strands with either biotin or fluorescein 41,4z In general, end-labelledprobes are less sensitive for the detection of complementary sequences by a factorof 10-100 since each molecule of polynucleotide is labelled with only onedetectable molecule. Another method for the colorimeter detection of DNAhybridization reactions involves the covalent linkage of an enzyme to a modifiednucleic acid probe using glutaraldehyde43 . These DNA enzyme conjugates can

NH2

cytosine residue

(b)NH -CH2-CH2-NH2

Molecular diagnosis of infectious diseases

1 1

function in blot hybridizätiön assays with a sensitivity of 1-5 pg for nitrocellulosebound target sequences . Unfortunately, the procedure for preparing these conju-gates is difficult to apply in a reproducible fashion .

We have recently developed a novel chemical method for labelling nucleic acidswith affinity ligands such as biotin which avoids some of the problems inherent inthe above methods" . With this labelling procedure, cytosine residues in single-stranded regions of nucleic acids are modified by the addition of sodium bisulphiteto the 5,6 double bond of the pyrimidine base . The cytosine bisulphite adducts aresubsequently converted to N 4 substituted cytosine derivatives by transaminationwith ethylenediamine . The cytosine derivatives which have a side chain terminatingin a primary amino group are then labelled with biotinyl-e-amino-caproic acid N-hydroxysuccinimide ester (Fig . 2). This method is a simple and practical means ofpreparing non-isotopic hybridization probes . The reagents necessary for thelabelling procedures are inexpensive and commercially available and the methoduses basic laboratory procedures which can be easily performed in a reproduciblefashion. Biotinylated probes prepared by this method can readily detect 1-2 pg offilter bound DNA . This level of sensitivity is comparable to that achieved with

(a)

NoHS0 3

0 NN /'

0 2"' N"50 3

R

R

AN N

0

0n

nS

(CHZ)4-C-NH-(CH2)g-C-O-N

N 4-substituted

biotinyl-c-aminocaproic acidcytosine residue

N-hydroxysuccinimide ester

N N0

0

NH-CH2CH2NH-C-(CHZ)5NH-C-(CH2)4 S

NH2

bisulfite adduct

0

NHZ CH2 CH2 NH 2

-HS03

biotinylated cytosine residue

Fig. 2. Scheme for labelling nucleic acid with biotin . (a) Bisulphite catalysed transamination reactionwith ethylenediamine . (b) Biotinylation reaction R =RNA or DNA .

0A

NH-CH2-CHp-NH2

N 4 -substitutedcytosine residue

12

R. P. Viscidi and R. H . Yolken

biotinylated probes prepared by other published techniques. However, unlikemethods using nick translation protocols a non-enzymatic labelling procedure hasthe advantage of being less influenced by impurities such as phenol, salts or agarosewhich may be present in some nucleic acid preparations . Compared with otherchemical methods for labelling nucleic acids, the reagents required for this methodare inexpensive, non-carcinogenic and readily available .

Further improvements in labelling techniques, assay formats and enzymaticdetection methods should allow for the wider application of nucleic acid hybridiza-tion reactions in diagnostic microbiology and basic research in infectious diseases .The successful development of practical techniques for the detection of nucleicacids in clinical specimens thus might markedly improve the care of patients withinfectious diseases and allow for a greater understanding of the epidemiology andpathophysiology of human infections .

REFERENCES

1 . Yoken, R . H . (1982) . Enzyme immunoassays for the detection of infectious antigens in body fluids :current limitations and future prospects. Reviews of Infectious Diseases 4, 35-67 .

2 . Wetmur, J . G . & Davidson, W . (1968) . Kinetics of renaturation of DNA . Journal of Molecular Biology31, 349-70 .

3 . Moseley, S . L ., Escheverris, P., Seriwatana, J . et al. (1982) . Identification of enterotoxigenicEscherichia coli by colony hybridization using three enterotoxin gene probes . Journal of InfectiousDiseases 145, 863-9 .

4 . Schafritz, D. A ., Shouval, D ., Sherman, H . I . et al . (1981) . Integration of hepatitis B virus DNA intothe genome of liver cells in chronic liver disease and hepatocellular carcinoma . New EnglandJournal of Medicine 305, 1067-73 .

5 . Steinberg, B . B ., Topp, W. C ., Schneider, P . J . et al. (1983) . Laryngeal papillomavirus infection duringclinical remission . New England Journal of Medicine 308, 1261-4 .

6. Brandsma, J. & Miller, G . (1980) . Nucleic acid spot hybridization : Rapid quantitative screening oflymphoid cell lines for Epstein-Barr viral DNA . Proceedings of the National Academy of ScienceUSA 77,6851-5 .

7. Southern, E . M . (1975) . Detection of specific sequences among DNA fragments separated by gelelectrophoresis . Journal of Molecular Biology 98, 503-17.

8 . Ranki, R ., Palva, A ., Virtanen, M . et al . (1982) . Sandwich hybridization as a convenient method forthe detection of nucleic acids in crude sample . Gene 21, 77-85 .

9 . Syvanen, A . C ., Loaksanen, M . & Soderlund, H . (1986) . Fast quantification of nucleic acid hybrids byaffinity-based hybrid collection . Nucleic Acid Research 14, 5037-48 .

10. Brahic, M . & Hasse, A . T . (1978). Detection of viral sequences of low reiteration frequency by in situhybridization . Proceedings of the National Academy of Science, USA 75, 6125-9 .

11 . Rigby, P . W . J ., Dieckmann, M ., Rhodes, C . et al . (1977). Labeling deoxyribonucleic acid to highspecific activity in vitro by nick translation with DNA polymerase I . Journal of Molecular Biology113, 237-51 .

12. Hy, N . & Messing, J . (1982) . The making of strand-specific M13 probes . Gene 17, 271-7.13. Melton, D . A ., Krieg, P. A ., Rebagliati, M . R . et al . (1984) . Efficient in vitro synthesis of biologically

active RNA and RNA hybridization probes from plasmids containing bacteriophage SP6 promoters .Nucleic Acid Research 12, 7035-56.

14. Cox, K . H ., DeLeon, D. V ., Angerer, L. M . et al. (1984) . Detection of mRNA's in sea urchin embryosby in situ hybridization using asymmetric RNA probes . Developmental Biology 101, 485-502 .

15 . Chou, S. & Merigan, T . C. (1985) . Rapid detection and quantitation of human cytomegalovirus inurine through DNA hybridization . New England Journal of Medicine 308, 921-5 .

16. Berninger, M ., Hammer, M., Hoyer, B . et al . (1982) . An assay for the detection of the DNA genome ofhepatitis B virus in serum . journal of Medical Virology 9, 57-68 .

Molecular diagnosis of infectious diseases

13

17. Flores, J ., Purcell, R . G ., Perez, I . et al. (1983) . A dot hybridization assay for detection of rotavirus .Lancet i, 555-9.

18. Dimitrou, D . H ., Graham, O . Y . & Estes, M. K . (1985) . Detection of rotaviruses by nucleic acidhybridization with cloned DNA of simian rotavirus SA11 genes . Journal of Infectious Diseases 152,293-300.

19 . Seidlin, M., Takoff, H . E ., Smith, H . A . et al . (1984) . Detection of varicella-zoster virus by dot-blothydridization using a molecularly cloned viral DNA probe . Journal of Medical Virology 13, 53-61 .

20. Stalhandski, P ., Hyypia, T., Godlec, H . et al. (1983) . The use of molecular hybridization fordemonstration of adenoviruses in human stools . Current Topics in Microbiology and Immunology104,299-306.

21. Owens, R. A . & Diener, T. O . (1981) . Sensitive and rapid diagnosis of potato spindle tuber viroiddisease by nucleic acid hybridization . Science 213, 670-2 .

22 . Perine, P . L., Totten, P . A ., Holmes, K . K . et al . (1985) . Evaluation of a DNA hybridization method fordetection of African and Asian strains of Neisseria gonorrhoeae in men with urethritis . Journal ofInfectious Diseases 152, 59-63 .

23. Wirth, D . F . & Pratt, D . M. (1982) . Rapid identification of Leishmanic species by specific hybridiza-tion of kinetoplast DNA in cutaneous lesions . Proceedings of the National Academy of Science,USA 79,6999-7003 .

24. Gildren, O . H ., Vafai, A ., Shtram, Y . et al . (1983) . Varicella-zoster virus DNA in human sensoryganglia . Nature 306, 478-80 .

25 . Miller, R . H ., Lee, S . C ., Law, Y . F . et al . (1985) . Hepatitis B viral DNA in infected human liver and inhepatocellular carcinoma . Journal of Infectious Diseases 151, 1081-92 .

26. Andiman, W ., Gradoville, L . & Heston, L . (1983) . Use of cloned probes to detect Epstein-Barr viralDNA in tissues of patients with neoplastic and lympho-proliferative diseases . Journal of InfectiousDiseases 148, 967-77 .

27 . Haase, A . T ., Ventura, P ., Gibbs, C . J . et al . (1981) . Measles virus nucleotide sequences : detection byhybridization in situ . Science 212, 672-4 .

28. Blum, H . E ., Stowring, L ., Figus, A . et al. (1983) . Detection of hepatitis B virus DNA in hepatocyates,bile duct epithelium, and vascular elements by in situ hybridization . Proceedings of the NationalAcademy of Science, USA 80,6685-8.

29. Ishikawa, E. & Kato, K . (1978) . Ultrasensitivie enzyme immunoassay . Scandinavian Journal ofImmunology 8 (Suppl . 7), 43-55 .

30. Yoshitake, S ., Hamaguchi, Y . & Ishikawa, E. (1979). Efficient conjugation of rabbit Fab' with B-D-galactosidase from Escherichia coli . Scandinavian Journal of Immunology 10, 81-6 .

31. Puget, K ., Michelson, A . M. & Avrameas, S . (1977) . Light emission techniques for the micro-estimation of femtogram levels of peroxidase . Analytical Biochemistry 79, 447-56 .

32. Ishikawa, E ., Imagawa, M . & Hashide, S . (1983) . Ultrasensitive enzyme immunoassay usingfluorogenic, immunogenic, radioactive and related substrates and factors to limit the sensitivity . InImmunoenzymatic Techniques ( Avrameas, S., Druet, P ., Masseyeff, R . & Feldman, G ., eds), pp . 219--32. Amsterdam : Elsevier .

33 . Langer, P . R ., Waldrop, A . A. & Ward, D . C . (1981) . Enzymatic synthesis of biotin-labelledpolynucleotides : novel nucleic acid affinity probes . Proceedings of the National Academy ofScience, USA 78, 6633-7 .

34. Bayer, E . A . & Wilchek, M . (1980) . The use of avidin-biotin complex as a tool in molecular biology .Methods in Biochemical Analysis 26, 1-45 .

35. Langer-Safer, P . R ., Levin, M . & Ward, D. C . (1982) . Immunological method for mapping genes onDrosophila polytene chromosomes . Proceedings of the National Academy of Science, USA 79,4381-5 .

36 . Singer, R . H . & Ward, D . C . (1982) . Actin gene expression visualized in chicken muscle tissue cultureby using in situ hybridization with a biotinated nucleotide analog . Proceedings of the NationalAcademy of Science, USA 79, 7331-5 .

37. Hutchison, M. H ., Langer-Safer, P . R., Ward, D . C . et al . (1982) . In situ hybridization of the electronmicroscope level : hybrid detection by autoradiography and colloidal gold . Journal of Cell Biology95,609-18 .

38 . Leary, J . J ., Brigati, D . J . & Ward, D . C . (1983) . Rapid and sensitive colorimetric method for visualizingbiotin-labeled DNA probes hybridized to DNA or RNA immobilized on nitrocellulose : bio-blots .Proceedings of the National Academy of Science, USA 80, 4045-9 .

14

R. P. Viscidi and R. H . Yolken

39 . Tchen, P ., Fuchs, R . P . P ., Sage, E . et al. (1984) . Chemically modified nucleic acids as immunodetec-table probes in hybridization experiments . Proceedings of the National Academy of Science, USA81,3466-70.

40. Forster, A . C ., McInnes, J . L . & Skingle, O . C . (1985) . Non-radioactive hybridization probes preparedby the chemical labelling of DNA or RNA with a novel reagent photobiotin . Nucleic Acid Research13, 745-61 .

41 . Chu, B . C . F ., Wahl, G. M. & Orgel, L . E . (1983) . Derivatization of unprotected polynucleotides .Nucleic Acid Research 11, 6513-29 .

42 . Agrowal, S., Cristodoulou, C . & Gait, M . (1986) . Efficient methods for attaching non-radioactivelabels to the 5' ends of synthetic aligodeoxy ribonucleotides . Nucleic Acid Research 14, 6227-45 .

43 . Renz, M . & Kurz, C . (1984). A colorimetric method for DNA hybridization . Nucleic Acid Research12,3435-44.

44 . Viscidi, R . P., Connelly, C . J . & Yolken, R. H . (1986) . Novel chemical method for the preparation ofnucleic acids for non-isotopic hybridization . Journal of Clinical Microbiology 23, 311-17 .