Volume 61, Number 2Printed in the U.S.A.

INTERNATIONAL JOURNAL OF LEPROSY

Molecular Cloning and Characterization ofContiguously Located Repetitive and Single CopyDNA Sequences of Mycobacterium tuberculosis:Development of PCR-Based Diagnostic Assay'

P. Prabhakara Reddi, Gursaran P. Talwar, andP. S. Khandekar2

Tuberculosis continues to be a globalhealth problem. Nearly one third of theworld's population is reported to be infectedwith the causative agent M ycobacterium tu-berculosis ("). In India alone there are 9—10 million people suffering from the diseaseat any given time, with an annual mortalityrate of 500,000 (13). The white plague hasreappeared in the West as a result of theincrease in the numbers of patients withAIDS. While several immunological studiesare currently in progress to identify pro-phylactic/immunotherapeutic agents, anequally important task in combating thedisease is to gain an understanding of thecomplex mycobacterial genome at the mo-lecular level. There have been many reportsof reiterated DNA sequences in AI. tuber-culosis (6, 14, 2I )) Nucleotide sequence datahave shown some of these to be insertionelements (9. 17). Hermans, et al., however,reported a 10-bp tandemly repeated DNAsequence which showed significant homol-ogy to the "chi" (crossover hotspot insti-gator) sequence of Escherichia coli (b0).Characterization of repetitive elements andadjacently located genes may give insightinto some of the molecular mechanisms thatmake mycobacteria so unique.

We have described earlier the identifi-cation of a 5.6-kb Alu fragment which is

' Received for publication on 11 March 1991; ac-cepted for publication in revised form on 22 January1993.

= P. P. Reddi, Ph.D.; G. P. Talwar, D.Sc.; P. S. Khan-dekar, Ph.D., National Institute of Immunology, NewDelhi 110067, India.

Present address for Dr. Reddi: Department of Anat-omy and Cell Biology, Box 439, School of Medicine,University of Virginia, Charlottesville, Virginia 22908,U.S.A.

Reprint requests to Dr. Khandekar.

present in multiple copies in the genome ofM. tuberculosis (14)• We now report the a)molecular cloning and characterization ofthis repetitive element and a contiguouslylocated single copy region of Al. tubercu-losis, and b) development of a highly specificand sensitive polymerase chain reaction(PCR)-based diagnostic assay for the detec-tion of Al, tuberculosis.

MATERIALS AND METHODSBacterial strains, phage library and plas-

mids. The bacterial strains listed in TheTable were obtained from the standard col-lection of the Trudeau Institute, SaranacLake, New York, U.S.A., except for Al. tu-berculosis south Indian isolate and Myco-bacterium tv, which are available at the Na-tional Institute of Immunology, New Delhi,India. The Al. tuberculosis Xgt 1 1 genomiclibrary was obtained from the World HealthOrganization. R18.8.2 is a phage clone whichstrongly hybridized to the 5.6-kb Alu frag-ment (14). The 2-kb and 1.4-kb Eco RI in-serts of R18.8.2 were subcloned in blue-script plasmid vector, and the resultingclones were named pKP4 and pKS I 0, re-spectively.

Recombinant DNA techniques, such asextraction of genomic and plasmid DNA,restriction endonuclease cleavage, purifi-cation of fragments from low melting point(LMP) agarosc and Southern hybridization,were performed as described earlier (14).

Replica plaque lifts. A Al. tuberculosisphage library was plated such that a 90-mmpetri dish contained 103 plaque formingunits (pfu). Genescreen nylon membrane(Dupont NEN, Boston, Massachusetts,U.S.A.) 88 mm in diameter was layered ontop of the soft agarose and allowed to absorb

227

228^ International Journal of Leprosy^ 1993

THE TABLE. Results of hybridization of2-kb repetitive and 1.4-kb single copy se-quences of Mycobacterium tuberculosis toDNA from other mycobacteria."

2-kb 1 . 4-kb

Itivcobacterium tuberculosis H37Rv^+ +Al. tuberculosis H37Ra^+ +Al. tuberculosis, south Indian isolate^+ +M. &Nis BCG^ + +M. africanum^ ++

aviumAl. int racellulare^ —Al. scrofulaceumM. kansasii

smegmatisM. cheloni

fortuitumM. vaccaeMu

+ = Positive hybridization; + + = intense hybrid-ization; — = no hybridization even after long exposuretimes.

phage for 60 sec. Then it was carefully peeledoff the plate. A second membrane (the rep-lica) was then placed on the same plate andallowed to sit for 90 sec before it was liftedoff. The two membranes were processed forplaque hybridization as per Maniatis, et al.("). The first membrane was probed with a5.6-kb Alu fragment; the replica lift wasprobed with a 2-kb insert of pKP4.

Polymerase chain reaction (PCR). Theoligonucleotide primers used for PCR —5'-CAACCAACATCGCGCTG-3' (KS3) and5'-ATAGCCGATGACGGTCG-3' (KS4)—were synthesized using a Gene Assembler(Pharmacia, Sweden). Typically, a 100-p1PCR reaction volume contained 1 x PCRbuffer [10 mM Tris-HC1 at pH 8.3, 50 mMKC1, 1.5 mM MgC17, 0.01% (w/v) gelatin];200 AM of deoxynucleotide triphosphates(dNTP), 20 AM each of primers, 50 ng ofgenomic or 1 ng of plasmid DNAs and 2.5units of Taq DNA polymerase. The mixturewas overlaid with 100/11of mineral oil. EachPCR cycle was performed at 94°C for 1 min,55°C for 2 min, and 72°C for 2 min; 30 suchcycles were repeated in a thermal DNA cy-cler (Perkin-Elmer Cetus Instruments, Nor-walk, Connecticut, U.S.A.). The amplifiedproduct was analyzed by electrophoresis onethidium bromide containing 1.2% agarosegels, followed by Southern hybridizationwith the 169-bp product of pKSIO. For this

purpose, the 169-bp fragment was elutedfrom LMP agarose gel and 32P-dCTP (alpha'2P-deoxycytidine triphosphate) labeled byprimer extension using KS3 and KS4 prim-ers. The filters in Figures 4B and 5B werehybridized in 5x Denhardt's [0.1% Ficoll(Pharmacia), 0.1% poly vinylpyrrolidone,and 0.1% bovine serum albumin (FractionV; Sigma)] and 6x SSC (1 x SSC: 0.15 Msodium chloride, 0.015 M sodium citrate)at 65°C overnight. Washings were done with2 x SSC and 1% SDS (sodium dodecyl sul-fate) at room temperature for 30 min, thenat 65°C for 1 hr and, finally, with 0.2 x SSCat 65°C for 30 min. The specificity of thehybridization signal (as shown in Fig. 4B)was obtained after the wash with 2 x SSCat 65°C for 1 hr.

RESULTSSubcloning. In order to identify recom-

binant clones containing repetitive DNA, aXgt 11 genomic library of Al. tuberculosisH37Rv was screened with the 5.6-kb Alufragment (t4). One of the strongly hybrid-izing phage clones, R18.8.2, was plaque pu-rified and characterized. Eco RI digestionof R18.8.2 DNA released two insert frag-ments of 2-kb and 1.4-kb that were sub-cloned in blucscript plasmid vectors. Thesubclones have been referred to as pKP4and pKS10, respectively.

Comparative Southern hybridization. Acomparative Southern hybridization exper-iment was designed to determine whetherone or both inserts of R18.8.2 possess arepetitive element. Equal amounts of Alu l-and Eco RI-digested M. tuberculosis DNAwere run in duplicate in parallel lanes (seelegend to Fig. 1). Following Southern trans-fer, lanes I and 2 were hybridized with thepKP4 insert and lanes 3 and 4 with thepKS 10 insert. Southern hybridization re-sults have shown the pKP4 insert to be hy-bridizing to multiple DNA fragments in theAlu I digest and to several high molecularsize bands in the Eco RI digest (Fig. I, lanes1, 2). On the other hand, the pKSIO inserthybridized to two bands in the Alu I digestand to one band in the Eco RI digest (Fig.1, lanes 3, 4). The hybridization pattern ofthe pKP4 insert resembled that of the pre-viously reported 5.6-kb Alu fragment (14) inthat it hybridized to multiple DNA frag-ments of similar size; whereas that of the

Species

61, 2^Reddi, et al.: DNA Sequences in M. tuberculosis^229

FIG. 1. Comparative Southern hybridization. Equalamounts (3 pg each) of M. tuberculosis H37Rv DNAdigested to completion with Alu I (lanes I, 3) or EcoRI (lanes 2, 4) were separated by 1.2% agarose gelelectrophoresis, Southern transferred and hybridizedto 'P-dCTP labeled 2-kb (lanes 1, 2) or 1.4-kb (lanes3, 4) inserts of pKP4 and pKSIO, respectively. Arrow= position of 3.5-kb Alu fragment in lanes 1 and 3 towhich both probes arc seen to hybridize; molecularsize markers in kilobase pairs are indicated on the left.

pKSIO insert was of a type consistent withsingle copy genes. The results of this exper-iment, therefore, suggested that Al. tuber-culosis repetitive DNA sequence, homolo-

FIG. 2. Copy number and similarity of the repeat-ing unit shared by the 5.6-kb Alu I genomic fragmentand 2-kb cloned insert. M. tuberculosis phage librarywas plated at 10' pfu per 90-mm petri dish, replicaplaque lifts were probed with either 5.6-kb Alu frag-ment (A) or 2-kb cloned insert (B). A total of 1018plaques were present on this petri dish. Of the 40%recombinants, 48 hybridized with either probe. Plaquetransfer was more efficient in A; hence the intensity ofthe signal as compared to the second plaque lift in panelB. Striking similarity in hybridization signals is no-ticeable in both A and B.

gous to the genomic 5.6-kb Alu fragment,is contained within the cloned 2-kb insertof pKP4.

Comparative plaque hybridization. TheSouthern hybridization of the pKP4 insertto multiple bands in the genomic digestsindicated the high copy number of the re-peat. In order to examine the extent of rep-resentation of this repeat in the gcnome ofM. tuberculosis and to further confirm thesimilarity of the nature of the repetitive el-ement shared by the 5.6-kb Alu fragmentand the cloned 2-kb insert, a comparativeplaque hybridization was performed. A Al.tuberculosis phage library was plated at adensity of 10 pfu per 90-mm petri plate.Replica plaque lifts were then probed witheither the pKP4 insert or the 5.6-kb Alufragment. As expected, the plaque hybrid-ization signals obtained with the two probeswere identical. Typical results are shown inFigure 2. Out of the 1018 plaques on thisplate, of which only 40% were recombinant,as many as 48 strong super imposable hy-bridization signals were observed with ei-ther probe, indicating that nearly 10% ofthe recombinant clones in the Al. tubercu-losis genomic library contained the repeti-tive DNA.

2-kb Repetitive and 1.4-kb single copy in-serts of R18.8.2 are contiguously located inM. tuberculosis genome. The Xgt 1 1 geno-

A

genomic DNA

A

cloned insert

of R16.6.2

repetitive(2 kb)

single copy(1.4 kb)

rl^

230^ International Journal of Leprosy^ 1993

0.5 kb

K S3 —IIPPAI— KS 4_;

(169bp PCR product)

FIG. 3. Repetitive (2-kb) and single copy (1.4-kb) sequences are contiguously located in the Al. tuberculosisgenome. Physical maps of complete insert of R18.8.2 (thick bar) and chromosomal region (— — — bar) fromwhere cloned insert originated. Parts of the chromosomal DNA between Alu I sites correspond to 3.5-kb and0.8-kb Alu bands seen in Fig. 1, lane 3. Location and direction of synthesis of PCR primers KS3 and KS4 isalso depicted. A = Alu I, R = Eco RI, r = linked Eco RI. There were single sites for Sac II, list EH and Sma Ion the 2-kb insert. The following sites were not present on the 2-kb repetitive DNA: Alu I, Kpn I, Xho I, AceI, Sal I, Cla I, Hind III, Pst I, Bain HI and Sau 3A.

mic library used in this study was con-structed by adding Eco RI linkers to sheared,end repaired and Eco RI methylase pro-tected genomic DNA (20). Two possibilitiestherefore existed: 1) The two Eco RI frag-ments of R18.8.2 (2 and 1.4 kb) could haveoriginated from different regions of the ge-nome but got ligated together becausc of EcoRI linkers, or 2) the entire 3.4-kb insert ofR18.8.2 is one contiguous piece of genomicDNA and an internal Eco RI site cleaved itinto the 2-kb repetitive and 1.4-kb singlecopy fragments. That the latter is true isindicated by the hybridization of 2-kb and1.4-kb inserts to a common 3.5-kb Alu frag-ment in the genomic DNA digest of M. tu-berculosis (Fig. 1, arrow). A partial restric-tion map showed that there are no Alu Isites on the pKP4 insert and one Alu site ispresent on the pKS 10 insert. By virtue ofthe absence of the Alu I site, the pKP4 inserthybridized to the larger 3.5-kb genomic Alufragment from which it originated. ThepKS10 insert, which had an internal Alu Isite, also hybridized with this 3.5-kb band(of which it forms a part) and with an ad-ditional genomic band of 0.8 kb (Fig. 1, lane3). The hybridization of the pKP4 andpKSIO insert DNAs to the common 3.5-kbgenomic Alu fragment, therefore, provedthat the repetitive element (pKP4 insert) andthe single copy region (pKS10 insert) orig-

mated from this contiguous piece of geno-mic DNA, as has been schematically rep-resented in Figure 3.

Diagnostic potential. To evaluate theusefulness of the cloned inserts in the di-agnosis of Al. tuberculosis, the specificity ofhybridization was determined for pKP4 andpKSIO. Two kig each of genoniic DNAs fromAl. tuberculosis and several other mycobac-teria were dotted in duplicate and hybrid-ized with radiolabeled probes. The repeti-tive pKP4 insert hybridized strongly withAl tuberculosis strains, Al. bovis BCG andAl. africanum DNA, but not with Al. avium,M. intracellulare, M. scrofulaceum, Al.smegmatis, Al. cheloni, Al. fortuitum, Al.vaccae or Mycobacterium w. However, therewas a weak crosshybridization with M. kan-sasii DNA. On the other hand, the singlecopy pKS10 insert appeared to be specificin its hybridization with members of the M.tuberculosis complex. There was no reactiv-ity with any of the other mycobacteria, in-cluding Al. kansasii. These results are sum-marized in The Table. Neither clonedfragment showed hybridization with geno-mic DNA from a number of nonmycobac-terial organisms or human DNA (data notshown).

PCR-based diagnostic assay. The spec-ificity of the 1.4-kb single copy insert to theAl. tuberculosis complex suggested that a

61, 2^Reddi, et al.: DNA Sequences in M. tuberculosis^231

FIG. 4. Specificity of PCR. Agarosc gel electrophoresis of amplified DNAs of different mycobacterial strainsusing oligonucleotide pair KS3 and KS4 (A) and autoradiograph of the transferred DNA hybridized to theradiolabeled 169-bp product of pKS 10 (B). Lanes 1 = 1-kb size marker; 2 = plasmid clone pKS10; 3 = Al.vaccae; 4 = Al. stnegmatis; 5 = Al. kansasii; 6 = M. scrofulaccum; 7 = Al. intracellularc; 8 = Al. (triton; 9 =M. bovis BCG; 10 = Al. tuberculosis south Indian isolate; Ii = Af. tuberculosis H37Ry; 12 = M. tuberculosisH37Ra. Arrow = position of the I69-bp band; molecular size markers are listed in basepairs.

target region of appropriate length could bechosen from this sequence for PCR ampli-fication (''). The pKS 10 insert has beencompletely sequenced (Reddi, et al., un-published results). Oligonucleotide primersKS3 and KS4 were designed such that a169-bp band could be amplified from thegenome of Al. tuberculosis as shown in Fig-ure 3, lower panel.

Specificity of the PCR amplification. Gc-nomic DNAs from members of the Al. tu-berculosis complex as well as from othermycobacteria (see legend to Fig. 4) were PCRamplified using the KS3 + KS4 primercombination. The expected 169-bp frag-ment was amplified in M. tuberculosisH37Ra, Al. tuberculosis H37Rv and Al. bo-vis BCG (Fig. 4A, lanes 9, 11, 12, arrow)which was of the same size as the amplifiedproduct of plasmid clone pKSIO, used as apositive control (Fig. 4A, lane 2). DNAsfrom other strains of mycobacteria were alsoamplified (Fig. 4A, lanes 3-8). However,the sizes of the amplified fragments weredifferent from those obtained with Al. tu-berculosis DNA. In order to differentiate theM. tuberculosis complex-specific PCR am-plified product from the rest, the 169-bpPCR product of pKS 10 was radiolabeledand used as probe to hybridize to the gel inFigure 4A. On Southern hybridization, a169-bp hybridization signal was seen onlyin lanes containing Al. tuberculosis H37Rv,

Al. tuberculosis H37Ra, Al. tuberculosissouth Indian isolate and Al. bolls BCGDNAs (Fig. 4B, lanes 9-12). Even thoughseveral amplified fragments were visualizedon ethidium-bromide-stained gel, there wasno hybridization with any of the other my-cobacteria tested (Fig. 4B, lanes 3-8) there-by putting in evidence the specificity of PCR.It is also of relevance that Southern hybrid-ization revealed the presence of 169-bp am-plified product in the Indian isolate of ALtuberculosis (Fig. 4B, lane 10). The targetsequence chosen for PCR amplification withamplimers KS3 and KS4 appeared to beconserved among strains ail/. tuberculosis,thereby indicating that the PCR test is suit-able for application to clinical specimens.

Sensitivity of PCR. The sensitivity of thePCR with primers KS3 and KS4 was de-termined by adding to the reaction mixturesserial dilutions of purified Al. tuberculosisDNA ranging from 300 ng to 30 fg per assay.Up to 300 pg of chromosomal DNA, theamplified product could be detected by aga-rose gel electrophoresis followed by ethid-ium-bromide staining (Fig. 5A); whereas theamplified product from 30 fg of chromo-somal DNA could be detected by a simpledot hybridization using a radiolabeled 169-bp PCR product as probe (Fig. 5B). Giventhe size of the Al. tuberculosis genome as2.5 x 109 Da, 30 fg of DNA correspondsto approximately 9-10 bacilli. This exper-

232^ International Journal of Leprosy^ 1993

FIG. 5. Level of detection by PCR amplification.Tenfold serial dilutions of.lf. tuberculosis genomic DNAranging from 300 ng to 30 fg (3 x 10 ' to 3 x 10-8pg) were PCR amplified using KS3 and KS4 primers.One tenth volume of PCR products were electropho-resed on 1.2% agarose gels (A) or dotted and hybridizedwith radiolabeled 169-bp probe (B). Lane M = 1-kbladder as idicated in Fig. 4. Lane P has the 1 69-bp PCRamplification product of pKS10.

iment demonstrated that with PCR ampli-fication followed by hybridization, one coulddetect as few as 9-10 tubercle bacilli. Inorder to examine the factors that could in-hibit the PCR, unrelated DNA from eitherM. smegmatis or human leukocytes wasadded to simulate the presence of extrane-ous DNA in the PCR reaction mixtures.This did not affect the level of sensitivity ofthe assay (data not shown). To rule out PCRcontamination and false-positive amplifi-cation, we routinely included a tube withno template DNA and constantly found thisto be negative for amplification (data notshown).

DISCUSSIONScreening of a M. tuberculosis genomic

library with the 5.6-kb Alu fragment led tothe identification of a clone bearing 2-kband 1.4-kb inserts. Hybridization to M. tu-berculosis genomic DNA digests showed that

the 2-kb fragment (pKP4) hybridized tomultiple bands of sizes ranging from 6.5 to0.6-kb. This indicated the presence of a re-petitive sequence within this cloned frag-ment. The 1.4-kb fragment (pKS10), on theother hand, has shown a hybridization pat-tern which is typical of single copy genes. Itis to be noted that the 2-kb probe hybridizedintensely with the 5.6-kb Alu fragment. Fur-ther evidence that the cloned pKP4 and theoriginally identified 5.6-kb genomic Alufragment contain essentially the same re-petitive element came from the plaque hy-bridization experiment. Replica plaque liftsof the M. tuberculosis library gave identicalhybridization signals when probed with ei-ther the 5.6-kb Alu fragment or the cloned2-kb probe. The ratio of plaques hybridizedto the total number of recombinant plaquesplated gave an indication that the repetitivesequence is found on 10% of the cloned ge-nomic fragments of the Al. tuberculosis li-brary.

There have been several reports of well-characterized, repetitive DNA sequences ofAl. tuberculosis (17)• 1S6110 and IS986 havebeen shown to be mobile elements. Boththese sequences contain several Alu I sites;whereas the pKP4 repeat does not containany Alu I site and is, therefore, not the sameas 1S6110 or IS986. Hermans, et al. (8) havereported cloning of the Al. tuberculosis re-petitive fragment pPH7301 based on thereactivity with monoclonal antibody F116-5. The nucleotide sequence of pPH7301showed the presence of 10-bp repeats thatwere termed as major polymorphic tandemrepeats (MPTR) (10). Amplimer pairs A andD or I and F (8' '°) which amplify 158- and181-bp fragments of pPH7301, respective-ly, failed to amplify pKP4 DNA (data notshown). Further, the presence of two Sau3A sites in pPH7301 and none in pKP4proved that these two repeated sequencesare not the same. Ross, et al. have reportedthe cloning of high molecular weight Nde IIfragments of M. tuberculosis which containrepetitive DNA (15). These investigators ap-peared to have cloned essentially the samerepetitive element as is present on the 5.6-kb Alu fragment previously reported by ourgroup (14). This is evident from the Southernhybridization profiles of pTBN12 andpTBN52 to Alu I-digested M. tuberculosisDNA (15) which are identical to that of the

61, 2^Reddi, et al.: DNA Sequences in M. tuberculosis^233

5.6-kb Alu fragment (14) and the presentpKP4 repeat. Interestingly enough, the re-striction maps of the three clones, pKP4,pTBN12 and pTBN52, are distinctly dif-ferent. We believe that this argues againstthe presence of a coding region in these re-peats. It could be conceived then that theactual repetitive unit, which is likely to becommon to pKP4, pTBN12 and pTBN52,is small in size and is present at multipleloci in the genome of Al. tuberculosis. Byemploying different routes of molecularcloning, our group and Ross, et al. haveidentified genomic fragments originatingfrom different loci but yet containing thesame repetitive unit. While these repeatedelements have proven to be useful epide-miological markers (3' 111' 15' 18), the function-al significance of the presence of such pre-sumably noncoding repeats at multiple lociin the genome remains to be elucidated. Itis also in this context that the characteriza-tion of contiguously located gene sequences,such as the 1.4 kb of the present study, be-comes relevant. Grossinsky, et al. (7) re-ported the location of a species-specific re-petitive DNA downstream of the codingregion of the 65-kDa antigen gene in Al.leprae. We are further investigating if thepKP4 M. tuberculosis repeat is located neara particular class of genes by characterizingadditional clones which contained the re-peat sequence.

Conventional methods used for the di-agnosis of tuberculosis infection have in-herent limitations. Bacteriological exami-nation by microscopy is simple, rapid andeconomical, but is nonspecific and can de-tect only the heavily infected cases. The cul-ture technique, on the other hand, is highlysensitive and specific but it takes 6-8 weeks.DNA probe technology combines the speedof microscopy with the reliability and ac-curacy of the culture technique and should,therefore, be considered as an alternativemethod for the diagnosis of tuberculosis.

Diagnostic potential has been assessed forthe cloned Al. tuberculosis DNA sequencespKP4 and pKSIO identified in the presentstudy. The repetitive as well as the singlecopy region were specific in hybridizationto members of the M. tuberculosis complex,except that the former exhibited weakcrosshybridization with Al. kansasii. In thisregard, pKS 1 0 showed the specificity de-

sired of a diagnostic probe. The reporteddetection limit of a radiolabeled DNA probeis approximately 10,000 bacilli per assay(9. In most clinical situations, especially inextra pulmonary tuberculosis and pauci-bacillary cases, the bacilli in a test sampletend to be fewer than 1000 which cannotbe picked up by the direct probe assay. Inorder to be able to detect M. tuberculosisdirectly in a clinical specimen without theneed for culture, we have developed a PCR-based diagnostic assay.

PCR amplification of a defined target se-quence and hybridization with a specificprobe is a highly sensitive and reliablemethod for the early detection of M. tuber-culosis. We have chosen a 169-bp regionfrom the 1.4-kb single copy fragment as atarget for PCR amplification by amplimersKS3 and KS4. By PCR amplification fol-lowed by hybridization with the radiola-beled 169-bp PCR product of pKS 10, wewere able to demonstrate that 1) the assayis highly specific to Al. tuberculosis and Al.bovis and 2) the sensitivity of the assay hasimproved from roughly 1000 bacilli de-tectable by ethidium-bromide-stained gel toas few as 9-10 bacilli.

Several reports have appeared describingthe PCR assay for the detection of Al'. tu-berculosis (1.2,4,5,8

) . These differ in the targetregion chosen for amplification. Althoughthe PCR technique offers a high degree ofsensitivity, care should be taken to elimi-nate false-negative and false-positive re-sults. Some investigators read the PCR re-sults from ethidium-bromide-stainedagarose gels alone (2); others introduced anadditional confirmatory step either in theform of restriction endonuclease digestionof PCR product (4) or hybridization with anoligonucleotide probe internal to the am-plified fragment ('). However, impurities of-ten inhibit cleavage by restriction endonu-cleases which makes interpretation of theresults difficult. It should also be borne inmind that Taq polymerase, which lacksproofreading ability, may introduce mis-matched bases into the amplified productwhich, theoretically, would alter a restric-tion enzyme site or the nucleotide sequenceto be hybridized with the internal oligoprobe. We believe that our method of usingthe radiolabeled, whole 169-bp probe forhybridization to the PCR products not only

234^ International Journal of Leprosy^ 1993

safeguards against these possible artifacts butoffers the additional advantage of highersensitivity. Conversion of the labeling of the169-bp probe to a nonisotopic modality andanalysis of a large number of clinical sam-ples from proven and suspected cases of tu-berculosis to evaluate the utility of our PCRdiagnostic assay is currently in progress.

SUMMARYScreening of a Ägt11 genomic library has

been used as an approach for molecularcloning of the Mycobacterium tuberculosisrepetitive DNA shown to be present on apreviously described 5.6-kb Alu I genomicfragment. The recombinant clone R18.8.2,which strongly hybridized with the radio-labeled 5.7-kb Alu fragment, carried twoEco RI inserts of 2 kb and 1.4 kb in size.Southern hybridization of each of thesefragments to restriction endonuclease-cleaved Al. tuberculosis DNA clearly dem-onstrated the 2 kb to contain the repetitiveDNA sequence, while the 1.4 kb is repre-sented in a single copy. When replica plaquelifts from the genomic library were probed,the 5.6-kb genomic fragment and the cloned2-kb repetitive insert hybridized to an iden-tical number of plaques, indicating the sim-ilarity and the high copy number of the re-peating unit shared by the two fragments.Restriction mapping and Southern hybrid-ization patterns indicated that the 2-kb re-petitive and the 1.4-kb single-copy DNAsequences originated from a contiguouspiece of genomic DNA. Both fragments werefound to be unique to members of the Al.tuberculosis complex, except that the 2-kbinsert exhibited a weak hybridization withM. kansasii DNA. Finally, a 169-bp regionfrom one end of the single-copy sequencehas been amplified by polymerase chain re-action (PCR) in a manner specific to Mem-bers of the Al. tuberculosis complex. Thesensitivity of the PCR was such that as fewas 9-10 bacilli could be detected.

RESUMENSc hizo el tamisaje de una genoteca preparada en

Xgt 11 para despues intentar la clonaciOn de la secuenciarepetitiva del DNA del Mycobacterium tuberculosispresente en el fragment° genOrnico Alu I de 5.6 kb. Laclona recombinante R18.8.2, la cual hibridizO fuerte-mente con el fragment° Alu 5.7-kb radiomarcado, con-tuvo dos insertos Eco RI de 2 kb y 1.4 kb, respecti-

vamente. La hibridizaciOn de cada uno de estosfragmentos con el DNA del Al. tuberculosis fraccionadopor endonucleasas de restricción, demostrO claramenteque el fragment° de 2 kb contuvo la secuencia repe-titiva de DNA, mientras que el framento de 1.4 kb seexpresó en una sola copia. El andlisis de la replicas delas placas de la genoteca, mostrO que el fragment° ge-nómico de 5.6 kb y el insert° repetitivo de 2 kb, hi-bridizaron con un nUmero identico de placas, indican-do la similitud y el alto nUmero de copias de la unidadrepetitiva compartida por los 2 fragmentos. Los patro-nes de restricción y de hibridización indicaron quetanto la secuencia de DNA repetitiva (2 kb) como lacontenida en una sola copia (1.4 kb), Sc originaron desecuencias contiguas en el DNA genómico. Sc encontrOque ambos fragmentos fueron exclusivos de los miem-bros del complejo del Al. tuberculosis, aunque el insert°de 2 kb exhibió una débil hibridización con el DNAde Al. kansasii. Finalmente, utilizando la reacción encadena de la polimerasa (PCR) se ha amplificado unaregion de 169 pb de uno de los extremos de la secuenciacontenida una sola copia (1.4 kb), la cual ha resultadosec especifica para los miembros del complejo Al. tu-berculosis. La sensibilidad del PCR fue tai que permitiOdetectar desde 9-10 bacilos.

RESUMEL'analyse d'une librairie génomique Xgtll a été uti-

lisée comme approche pour le clonage moléculaire del'ADN répetitif de Mycobacterium tuberculosis donton a montre la presence sur un fragment génomiqueAlu I de 5,6 kb précédemment decrit. Le clone recom-binant R18.8.2, qui s'hybridisait fortement avec lefragment Alu de 5,7 kb marque par radioactivite,contenait deux fragments Eco RI d'une taille de 2 kbet 1,4 kb. L'hybridisation de chacun de ces fragments

l'ADN de Al. tuberculosis dive par une endonucléasede restriction a clairement démontre quc le fragmentde 2 kb contenait la sequence d'ADN repetitif, tandisqu'un seul exemplaire du fragment de 1,4 kb est re-presente. Lorsque des répliques en provenance de lalibrairie genomique ont été sondees, le fragment gé-nomique de 5,6 kb et Einsert repétitif clone de 2 kbs'hybridisaient avec un numbre identique de plaques,indignant la similarité et lc grande nombre d'exem-plaires de l'unité repetitive que les deux fragmentsavaient en commun. La cartographic par restriction etl'hybridisation indiquaient que le fragment répétitifd'ADN de 2 kb et l'exemplaire unique de 1,4 kb pro-venaient de parties contigues de l'ADN genomique.Les deux fragments se révélèrent spécifiques auxmembres du complexe Al. tuberculosis, si cc n'est quele fragment de 2 kb montrait une hybridisation faibleavec l'ADN de Al. kansasii. Finalement, une region de169 bp d'une extremité de la sequence unique a étéamplifiée par une reaction de polymerase en chaine(PCR) d'une maniere specifique aux membres ducomplexe Al. tuberculosis. La sensibilite de la PCR étaittelle que des bacilles en nombre aussi faible que 9 ou10 pouvaient etre detectés.

61, 2^Reddi, et al.: DNA Sequences in M. tuberculosis^235

Acknowledgment. We thank Ashis Das (NII) for thesynthesis of PCR primers and Dr. Rama Mukherjee(NII) for providing us with mycobacterial strains.

REFERENCES1. BRISSON-NOEL, A., GICQUEL, B., LECOSSIER, D.,

LEVY-FREBAULT, V., NASSIF, X. and HANCE, A. J.Rapid diagnosis of tuberculosis by amplificationof mycobacterial DNA in clinical samples. Lancet2(1989) 1069-1071.

2. BUCK, G. E., O'HARA, L. C. and SUMMERSGILL, J.T. Rapid, simple method for treating clinicalspecimens containing Mycobacterium tuberculosisto remove DNA for polymerase chain reaction. J.Clin. Microbiol. 30 (1992) 1331-1334.

3. CAVE, M. D., EISENACH, K. D., MCDERMOTT, P.F., BATES, J. H. and CRAWFORD, J. T. IS6110:conservation of sequence in Mycobacterium tu-berculosis complex and its utilization in DNA fin-gerprinting. Mol. Cell. Probes 5 (1991) 73-80.

4. COUSINS, D. V., WILTON, S. D., FRANCIS, B. R. andGow, B. L. Use of polymerase chain reaction forrapid diagnosis of tuberculosis. J. Clin. Microbiol.30 (1992) 255-258.

5. DE WIT, D., STEYN, L., SHOEMAKER, S. and SCOGIN,

M. Direct detection of Mycobacterium tubercu-

losis in clinical specimens by DNA amplification.J. Clin. Microbiol. 28 (1990) 2437-2441.

6. EISENACH, K. D., CRAWFORD, J. T. and BATES, J.H. Repetitive DNA sequences as probes for My-cobacterium tuberculosis. J. Clin. Microbiol. 26(1988) 2240-2245.

7. GROSSINSKY, C. M., JACOBS, W. R., CLARK-CURTISS,

J. E. and BLOOM, B. R. Genetic relationshipsamong Mycobacterium leprae„ifycobacterizun tu-berculosis, and candidate leprosy vaccine strainsdetermined by DNA hybridization: identificationofan Al. teprae-specific repetitive sequence. Infect.Immun. 57 (1989) 1535-1541.

8. HERMANS, P. W. M., SCHUITEMA, A. R. J., VAN

SOOLINGEN, D., VERSTYNEN, C. P. H. J., BIK, E.M., THOLE, J. E. R., KOLK, A. H. K. and VAN

EMBDEN, J. D. A. Specific detection of Mycobac-terium tuberculosis complex strains by polymerasechain reaction. J. Clin. Microbiol. 28 (1990) 1204-1213.

9. HERMANS, P. W. M., VAN SOOLINGEN, D., DALE,

J. W., SCHUITEMA, A. R. J., McADAm, R. A., CATTY,

D. and VAN EMBDEN, J. D. A. Insertion elementIS986 from Afrcobacterium tuberculosis: a usefultool for diagnosis and epidemiology of tubercu-losis. J. Clin. Microbiol. 28 (1990) 2051-2058.

10. HERMANS, P. W. M., VAN SOOLINGEN, D. and VAN

EMBDEN, J. D. A. Characterization of a majorpolymorphic tandem repeat in Mycobacterium tu-berculosis and its potential use in epidemiology of

Mycobacterium kansasii and Mycobacteriunz gor-donae. J. Clin. Microbiol. 12 (1992) 4157-4165.

11. MANIATIS, T., FRITSCH, E. F. and SAN1BROOK, J.

Molecular Cloning: A Laboratory Manual. ColdSpring Harbor, New York: Cold Spring HarborLaboratory, 1982.

12. PAO, C. C., LYN, S. S., Wu, S. Y. and JUANG, W.

M. The detection of mycobactcrial DNA se-quences in uncultured clinical specimens withcloned Mycobacterium tuberculosis DNA as probes.Tubercle 69 (1988) 27-36.

13. PARK, J. E. and PARK, K. Textbook of Preventiveand Social Medicine. Jabalpur, India: M/S Ban-arsidas Bhanot, 1986, pp. 164-177.

14. REDDI, P. P., TALWAR, G. P. and KHANDEKAR, P.S. Repetitive DNA sequence of Mycobacteriumtuberculosis: analysis of differential hybridizationpattern with other mycobacteria. Int. J. Lepr. 56(1988) 592-597.

15. Ross, B. C., RAIOS, K., JACKSON, K. and DWYER,

B. Molecular cloning of a highly repeated DNAelement from Mycobacterium tuberculosis and itsuse as an epidemiological tool. J. Clin. Microbiol.30 (1992) 942-946.

16. SAIKI, R. K., GELFAND, D. H., STOFFEL, S., SCHARF,

S. J., HIGUCHI, R., HORN, G. T., Minus, K. B.and ERLICH, H. A. Primer directed enzymatic am-plification of DNA with a thermostable DNApolymerase. Science 239 (1988) 487-491.

17. THIERRY, D., CAVE, M. D., EISENACH, K. D.,CRAWFORD, J. T., BATES, J. H., GICQUEL, B. andGUESDON, J. L. 1S6 110, an IS-like element of Aly-

cobacterium tuberculosis complex. Nucleic AcidsRes. 18 (1990) 188.

18. VAN SOOLINGEN, D., HERMANS, P. W. M., DE HAAS,

P. E. W., SoLL, D. R. and VAN EMBDEN, J. D. A.Occurrence and stability of insertion sequences inMycobacterium tuberculosis complex strains: eval-uation of an insertion sequence-dependent DNApolymorphism as a tool in epidemiology of tu-berculosis. J. Clin. Nlicrobiol. 29 (1991) 2578-2586.

19. WORLD HEALTH ORGANIZATION. Progress andevaluation report, Tuberculosis Control Pro-gramme, EB87/4. Geneva: World Health Orga-nization, November 1990.

20. YOUNG, R. A., BLOOM, B. R., GRossimicv, C. M.,IVANYI, J., THOMAS, D. and DAVIS, R. W. Dis-section of Mycobacterium tuberculosis antigens us-ing recombinant DNA. Proc. Natl. Acad. Sci.U.S.A. 82 (1985) 2583-2587.

21. ZAINUDDIN, Z. F. and DALE, J. W. Polymorphicrepetitive DNA sequences in Mycobacterium tu-berculosis detected with a gene probe from My-cobacterium fortuittou plasmid. J. Gen. Microbiol.135 (1989) 2347-2355.