Proc. Natl. Acad. Sci. USAVol. 90, pp. 2608-2612, April 1993Microbiology

Mycobacterium tuberculosis expresses two chaperonin-60 homologsT. H. KONG*, A. R. M. COATES*, P. D. BUTCHER*, C. J. HICKMANt, AND T. M. SHINNICKt*Division of Molecular Microbiology, Department of Medical Microbiology, St. George's Hospital Medical School, London, SW17 ORE United Kingdom; andtDivision of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, GA 30333

Communicated by Barry Bloom, December 14, 1992 (received for review October 16, 1991)

ABSTRACT A 65-kDa protein and a 10-kDa protein aretwo of the more strongly immunoreactive components of My-cobacterium tuberculosis, the causative agent of tuberculosis.The 65-kDa antigen has homology with members of the GroELor chaperonin-60 (Cpn6O) family of heat shock proteins. The10-kDa antigen has homology with the GroES or chapero-nin-10 family of heat shock proteins. These two proteins areencoded by separate genes in M. tuberculosis. The studiesreported here reveal that M. tuberculosis contains a secondCpn6O homolog located 98 bp downstream of the 10-kDaantigen gene. The second Cpn6O homolog (Cpn6O-1) displays61% amino acid sequence identity with the 65-kDa antigen(Cpn6O-2) and 53% and 41% identity with the Escherichia coliGroEL protein and the human P60 protein, respectively.Primer-extension analysis revealed that transcription starts 29bp upstream of the translation start of the Cpn6O-1 homologand protein purification studies indicate that the cpn6O-1 geneis expressed as an -60-kDa polypeptide.

The diseases caused by mycobacteria are important sourcesof morbidity and mortality in the world today. More than onebillion persons have been infected with Mycobacterium tu-berculosis, the causative agent of tuberculosis. This diseaseaccounts for -25% of all preventable deaths worldwide-more than 3.5 million deaths per year (1). Much work over thepast few years has gone into the identification and charac-terization of immunoreactive proteins ofM. tuberculosis (forreview, see refs. 2-4). Two of the more strongly immunore-active components of the pathogenic mycobacteria are a65-kDa protein and a 10-kDa protein. The immune responseto the 65-kDa antigen has received much attention because ofpossible roles in autoimmunity and arthritis and as an antigenof y8 T cells (for review, see refs. 5 and 6). Recently, the10-kDa antigen has been shown to be an important T-cellantigen in tuberculosis patients (7).The genes encoding these proteins have been isolated and

characterized (8-13). The 65-kDa antigen has homology withmembers of a family of heat shock proteins that has beencalled the GroEL, HSP60, or chaperonin-60 (Cpn6O) family(13-16). The 10-kDa antigen has homology with the GroES orchaperonin-10 (CpnlO) family of heat shock proteins (10, 11,16). An unusual feature of these M. tuberculosis CpnlO andCpn6O homologs is that they are encoded by nonadjacentgenes. That is, most bacteria express the CpnlO and Cpn6Ohomologs as an operon (e.g., ref. 17). The studies reportedhere reveal that M. tuberculosis does contain a potentialoperon encoding a CpnlO homolog (the 10-kDa antigen),t aCpn6O homolog,* and the separate gene encoding the 65-kDaantigen. A similar arrangement of cpnlO and cpn6O genes hasbeen described in Streptomyces spp. (18) and Mycobacte-rium leprae (19).To avoid confusion in referring to the two M. tuberculosis

Cpn6O homologs, particularly in regard to the extensiveexisting literature, we suggest that the 65-kDa antigen retain

its designation as the 65-kDa antigen and the correspondinggene be designated cpn60-2. We suggest that the homologencoded by the gene adjacent to the cpnlO gene be designatedthe Cpn60-1 product (cpn60-1 gene) in accord with currentnomenclature for chaperonins.

MATERIALS AND METHODSBacteria, Phage, and Plasmids. The isolation of a recom-

binant phage expressing the CpnlO homolog (A-SK24) fromthe A-gtll M. tuberculosis recombinant DNA library (20) hasbeen described (21). The 3.4-kb EcoRI fragment from thisrecombinant phage was subcloned into pUC19 to generatepRL4. In this plasmid, transcription of the M. tuberculosisgene encoding the CpnlO homolog is from the pUC19 lacpromoter. Portions ofthe mycobacterial DNA insert in pRL4were also subcloned into M13mp18, M13mpl9, and pBlue-script KS (Stratagene) vectors using standard procedures(22). A set of nested deletions of this fragment was generatedusing the Erase-a-Base kit (Promega). Plasmid pTB12 carriesthe gene encoding the M. tuberculosis 65-kDa antigen (9).Phage and plasmids were propagated in Escherichia colistrain XL-1 (Stratagene). Sequences of the second Cpn60homolog were also obtained independently from two M13clones (11) containing the 2.25-kbp Sal I M. tuberculosisgenomic DNA fragment.

Nucleotide Sequencing. Single-stranded DNA was isolatedas described (22) and nucleotide sequences were determinedusing the Sanger dideoxynucleotide chain-terminationmethod (23), deoxyadenosine 5'-[a-[35S]thio]triphosphate(Amersham), and a Sequenase kit (United States Biochem-ical). The products of the sequencing reactions were sepa-rated on 6% polyacrylamide/7 M urea/l x TBE gels (22)poured with 0.4- to 1-mm wedge spacers (Bio-Rad) or 0.4-mmflat spacers. After electrophoresis, the gels were fixed in 10%(vol/vol) acetic acid/10% (vol/vol) methanol, washed inH20, dried under vacuum, and exposed to Kodak XRP-1film. The nucleotide sequences were determined indepen-dently for both strands of the mycobacterial DNA. Comput-er-aided analysis of the nucleic acid and deduced proteinsequences was performed using the GCG sequence analysissoftware package (ref. 24; University of Wisconsin Biotech-nology Center, Madison).Primer Extension. Mycobacterium bovis bacillus Cal-

mette-Gudrin (BCG) cells were grown to late-exponentialgrowth phase in Dubos broth supplemented with Dubosmedium albumin (Difco) in 200-ml bottles. Total cellularRNA was extracted by sonication in 4 M guanidinium iso-thiocyanate and CsCl ultracentrifugation (25). Total RNAwas digested with 10 units of RNase-free DNase I for 20 minat 37°C. (The RNase control was digested with 10 ug ofDNase-free pancreatic RNase for 20 min at 37°C.) The

Abbreviations: RBS, ribosome binding site; BCG, bacillus Calmette-Gderin; CpnlO and Cpn6O, chaperonin-10 and chaperonin-60, re-spectively; IMAC, immobilized metal ion affinity chromatography.*The sequence reported in this paper has been deposited in theGenBank data base (accession no. X60350).

2608

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 27

, 202

0

Proc. Natl. Acad. Sci. USA 90 (1993) 2609

resulting RNA was then subjected to primer extension usingthe oligonucleotide 5'-GCGGTTTCGTCGTATTCGAT-3'(bases 13-32 downstream of the Cpn6O-1 ATG start codon).The primer was annealed to the RNA by heating to 70°C andthen slowly cooling to 50°C. Primer extension was performedat 50°C for 15 min in a final volume of 20 ,ul of 50 mM Tris,pH 8.4/10 mM MgCl2/15 mM KCl/1 mM dithiothreitol/0.01mM dGTP/0.01 mM dATP/0.01 mM TTP/0.01 mM[a-32P]dCTP. Reverse transcriptase was added and the reac-tion was continued for 15 min. The primer-extension prod-ucts and sequencing products were electrophoresed on a 6%polyacrylamide/8 M urea gel, and the gel was fixed and driedas described above and exposed to Hyperfilm-MP (Amer-sham) for 2 days.

Preparation of Lysates. M. bovis BCG was grown at 37°Cin Middlebrook 7H9 medium containing 0.05% Tween 80 andoleic acid/albumin/dextrose/catalase supplement (Difco) in490-cm2 tissue culture roller bottles (Coming) to an OD600 of0.5. Bacteria were harvested by centrifugation and the pelletwas washed with phosphate-buffered saline (PBS). The finalpellet from 100 ml of culture was resuspended in 2.5 ml ofbuffer A (6 M guanidine hydrochloride/0.1 M NaH2PO4/0.01M Tris, pH 8.0) and transferred to a 50-ml Oakridge centri-fuge tube (Nalge). Sterile glass beads (0.1 mm, diameter; 0.5ml, packed volume) were added and the sample was soni-cated in a cup horn sonicator for 10 min (40W in 3-sec pulses).The glass beads and cellular debris were removed by cen-trifugation at 10,000 x g for 10 min and the supernatant wastransferred to a fresh tube and stored at -20°C.

Strains XL-l(pRL4) and XL-l(pTB12) were grown at 37°Cin Luria broth containing ampicillin (33 ,ug/ml) to an OD600 of0.5. Expression of the mycobacterial genes in pRL4 wasinduced by adding isopropyl j3-D-thiogalactopyranoside(United States Biochemical) to a final concentration of 2 mM.Both cultures were then incubated an additional 2 h at 37°C.Cells were harvested by centrifugation and lysed as describedabove.

Immobilized Metal Ion Affinity Chromatography (IMAC).Proteins were purified using a nitrilotriacetic acid resin (26)and a guanidine hydrochloride/urea/phosphate buffer sys-tem as described by the supplier (Qiagen, Chatsworth, CA;ref. 27). Briefly, columns containing -2 ml of the nitrilotri-acetic acid resin were equilibrated with 10 ml of buffer A. Thecrude bacterial lysates (2.5 ml in buffer A) were then loadedonto the columns. The samples were washed with (i) 10 ml ofbuffer A, (ii) 10 ml of buffer B (8 M urea/0.1 M NaH2PO4/0.01 M Tris, pH 8.0), and (iii) 10 ml of buffer C (buffer B, pH7.0). The columns were stored overnight at 4°C in buffer C.The bound proteins were eluted with buffer D (buffer B, pH5.9). Samples (1 ml) were collected and analyzed by electro-phoresis on 10% polyacrylamide/SDS Laemmli gels (28)followed by silver staining (29). Fractions containing the-60-kDa protein (typically samples two to four) were pooled,dialyzed against PBS, and lyophilized.

RESULTSA chance observation by one of us (A.R.M.C.) during a DNAsequencing project revealed that there might be an openreading frame downstream from the 10-kDa open readingframe that could potentially encode a protein homologous tothe 65-kDa antigen. This observation led us to reexamine theavailable nucleotide sequences of genes encoding the M.tuberculosis CpnlO homolog (EMBL accession nos. M25258and X13739; refs. 11 and 12) and the 65-kDa antigen (EMBLaccession no. M15467; ref. 9). Indeed, the beginning of anopen reading frame was found _100 bases downstream fromthe end of the 10-kDa antigen open reading frame. This framewas preceded by a 5/6 match with the consensus sequencefor a ribosome binding site (RBS) and translation of it

produced a 55-residue sequence that had 34 amino acidsidentical to those in the N-terminal sequence of the 65-kDaantigen. On the other hand, no open reading frame was foundin the 575 bp of sequence upstream of the 65-kDa antigenopen reading frame that could encode a CpnlO homolog.The nucleotide sequence of the remainder of the second

open reading frame was determined from subclones ofpRL4(12) and a 2.25-kbp Sal I M. tuberculosis genomic DNAfragment (11). The nucleotide sequence and deduced proteinsequences are shown in Fig. 1. Also, as is typical for M.tuberculosis genes (9), the overall G + C content of the twoopen reading frames is =65%, and the G + C content of thethird position of the codons is 83%. Overall, this region oftheM. tuberculosis genome displays the typical groE operonstructure of promoter-RBS-GroES-(50-100 bases)-RBS-GroEL-transcription terminator (Fig. 2).To map the start site of the mRNA encoding the second

Cpn6O homolog (hereafter designated Cpn60-1), primer-extension studies were done using a primer located 13-32bases downstream ofthe presumed ATG start codon. A 61-bpproduct was produced in these experiments (Fig. 3), whichindicates that the transcription start site was 29 bp upstreamof the ATG start codon. No extension products were ob-served after RNase treatment or omission of the primer,confirming that the bands produced were mRNA extensionproducts. Thus, the second cpn60 gene (hereafter, cpn60-1)is expressed at the level of RNA but does not appear to beexpressed as part of an operon with the cpnl0 gene.At the nucleotide level, the open reading frame for the

Cpn60-1 homolog displays -66% sequence identity with thatof the 65-kDa antigen gene. The sequences outside the openreading frames show little similarity. The deduced amino acidsequences of Cpn6O-1 and the 65-kDa antigen display excel-lent alignment with 61% sequence identity and 76% sequencesimilarity (Fig. 4).When compared to other members of the Cpn6O family of

heat shock proteins, the 65-kDa antigen consistently shows5-10% greater sequence identity and similarity than does theCpn6O-1 protein (14/14 comparisons; P < 0.002, Wilcoxonsigned rank sum test). For example, the 65-kDa antigendisplays 59% sequence identity and 76% similarity to the E.coli GroEL protein (17), and Cpn60-1 displays 53% sequenceidentity and 70% similarity. Similarly, the 65-kDa antigendisplays 47% sequence identity to the human P60 protein (31)and Cpn60-1 displays 41% sequence identity. However, itshould be noted that the differences are statistically signifi-cant for only 5 ofthe 14 comparisons (P < 0.01, Fisher's exacttest).One apparently unique feature ofthe Cpn60-1 protein is the

histidine-rich sequence at the C terminus, -DHDHH-HGHAH. Typically, Cpn60 homologs contain a sequencerich in glycine and methionine residues at the C terminus(e.g., -DMGGMDF, for the 65-kDa antigen), although theRickettsia tsutsugamushi Cpn60 homolog does contain twohistidine residues near the C terminus along with the MGGMmotif (32) and one of the Streptomyces Cpn60 homologscontains four closely spaced histidine residues (18). Oneconsequence of this is that the Cpn60-1 protein should bindto metal chelate adsorbents. The 65-kDa antigen should notbind to the metal chelate adsorbents since binding requirestwo or more adjacent histidine residues (26, 27).To determine whether the Cpn60-1 protein was actually

expressed in mycobacteria, a lysate of M. bovis BCG wassubjected to IMAC (26, 27). The M. bovis BCG lysate wasused since (i) the M. bovis BCG 65-kDa and 10-kDa proteinshave amino acid sequences identical to those of the M.tuberculosis homologs (9-13) and (ii) the coding regions ofthe cpn60-J gene of M. bovis BCG have identical restrictionmaps to the M. tuberculosis cpn60-J gene and produce similaramplicons when amplified using the PCR and eight primers

Microbiology: Kong et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 27

, 202

0

2610 Microbiology: Kong et al.

161

121181241

CAGCTGGTGGCCGCGGGAGCGCCGCCGTCGCCGCTGGACGTGCCCAGTGACCCGGGTCTGCCCGTGATGCAGGGGCAGGTGCGGTGACCCGGACATTGCACCTGGCGTACCCGGACACGCTGGCAACCAGGAAGCAAGGGGGCGCCCTTGAGTGCTAGCACTCTCATGTATAGAGTGCTAGATGGCAATCGGCTAACCCCTGCGTCGGCACCCGCGACGACGGCGCGAGGGCGCGGACGA

1 RB9 fM A K301 CCGGTCCGGGCGAGCGTCCCGGGCTCTGATCCAAATAGTGGAGGGCTCCAATCGTGGCGA

4 V N I K P L E D K I L V Q A N E A E T T361 AGGTGAACATCAAGCCACTCGAGGACAAGATTCTCGTGCAGGCCAACGAGGCCGAGACCA

24 T A S G L V I P D T A K E K P Q E G T V421 CGACCGCGTCCGGTCTGGTCATTCCTGACACCGCCAAGGAGAAGCCGCAGGAGGGCACCG

44 V A V G P G R W D E D G E K R I P L D V481 TCGTTGCCGTCGGCCCTGGCCGGTGGGACGAGGACGGCGAGAAGCGGATCCCGCTGGACG

64 A E G D T V I Y S K Y G G T E I K Y N G541 TTGCGGAGGGTGACACCGTCATCTACAGCAAGTACGGCGGCACCGAGATCAAGTACAACG

84601

661

1721

E E Y L I L S A R D V L A V V S K * *GCGAGGAATACCTGATCCTGTCGGCACGCGACGTGCTGGCCGTCGTTTCCAAGTAGTAGA

GCGTGTTCCGCCCCGGCGATCCCCGTGCTCACCACGGGTGATTTCCGGGGCGGCATGCGT

r-> RIB M S K L I E Y D E TTAGCGGACTAGCCCTGCGTAGAGGAGCCTGATGAGCAAGCTGATCGAATACGACGAAACC

11 A R R A M E V G M D K L A D T V R V T L781 GCGCGTCGCGCCATGGAGGTCGGCATGGACAAGCTGGCCGACACCGTGCGGGTGACGCTG

31841

51901

G P R G R H V V L A K A F G G P T V T NGGGCCGCGCGGCCGGCATGTGGTGCTGGCCAAGGCGTTTGGCGGACCCACGGTTACCAAC

D G V T V A R E I E L E D P F E D L G AGACGGCGT GTCACGTGAGATCGAGCTGGAAGATCCGTTTGAAGACTTGGGCGCC

71 Q L V K S V A T K T N D V A G D G T T T961 CAGCTGGTGAAGTCGGTGGCCACCAAGA-CAACGATGTGGCCGGTGACGGCACCACCACC

91 A T I L A Q A L I K G G L R L V A A G V1021 GCAACCATCTTGGCGCAGGCACTGATCAAGGGCGGCCTGAGGCTAGTGGCCGCCGGCGTC

111 N P I A L G V G I G K A A D A V S E A L1081 AACCCGATCGCGCTCGGCGTGGGAATCGGCAAGGCCGCCGACGCGGTATCCGAGGCGCTG

131 L A S A T P V S G K T G I A Q V A T V S1141 CTGGCATCGGCCACGCCGGTGTCCGGCAAGACCGGCATCGCGCAGGTGGCGACGGTGTCC

1511201

1711261

S R D E Q I G D L V G E A M S K V G H DTCGCGCGACGAGCAGATCGGTGACCTGGTTGGCGAAGCGATGAGCAAGGTCGGCCACGACG V V S V E E S S T L G T E L E F T E GGGCGTGGTCAGCGTCGAAGAATCCTCGACGCTGGGCACCGAGTTGGAGTTCACCGAGGGT

191 I G F D K G F L S A Y F V T D F D N Q Q1321 ATCGGCTTCGACAAGGGCTTCTTGTCGGCATACTTCGTTACCGACTTCGATAACCAGCAG

Proc. Natl. Acad. Sci. USA 90 (1993)

211 A V L E D A L I L L H Q D K I S S L P D1381 GCGGTGCTCGAGGACGCGTTGATCCTGCTGCACCAAGACAAGATCAGCTCGCTTCCCGAT

2311441

2511501

2711561

2911621

3111681

3311741

3511801

3711861

3911921

4111981

4312041

4512 101

4712161

4912221

L L P L L E K V A G T G K P L L I V A ECTGTTGCCATTGCTGGAAAAGGTTGCAGGAACGGGTAAGCCACTACTGATCGTGGCTGAA

D V E G E A L A T L V V N A I R K T L KGACGTGGAGGGCGAAGCGTTGGCGACGCTGGTCGTCAACGCGATTCGCAAGACGTTGAAA

A V A V K G P Y F G D R R K A F L E D LGCGGTCGCGGTCAAGGGGCCGTACTTCGGTGACCGCCGTAAGGCGTTCCTTGAGGACCTG

A V V T G G Q V V N P D A G M V L R E VGCGGTGGTGACGGGTGGCCAGGTGGTCAACCCCGACGCCGGCATGGTGCTGCGCGAGGTGG L E V L G S A R R V V V S K D D T V IGGCTTGGAGGTGCTGGGCTCGGCCCGACGCGTGGTGGTCAGCAAGGACGACACGGTCATTV D G G G T A E A V A N R A K H L R A EGTCGACGGCGGCGGCACCGCAGAAGCGGTGGCCAACCGGGCGAAGCACTTGCGTGCCGAG

I D K S D S D W D R E K L G E R L A K LATCGACAAGAGCGATTCGGATTGGGATCGGGAAAAGCTTGGCGAGCGGCTGGCCAAACTG

A G G V A V I K V G A A T E T A L K E RGCCGGCGGGTTGCTGTCATCAAGGTGGGTGCCGCCACCGAGACCGCACTCAAGGAGCGCK E S V E D A V A A A K A A V E E G I VAAGGAAAGCGTCGAGGATGCGGTCGCGGCCGCCAAGGCCGCGGTCGAGGAGGGCATCGTC

P G G G A S L I H Q A R K A L T E L R ACCTGGTGGGGGAGCCTCGCTCATCCACCAGGCCCGCAAGGCGCTGACCGAACTGCGTGCGS L T G D E V. L G V D V F S E A L A A PTCGCTGACCGGTGACGAGGTCCTCGGTGTCGACGTGTTCTCCGAAGCCCTTGCCGCGCCG

L F W I A A N A G L D G S V V V K K V STTGTTCTGGATCGCCGCCAACGCTGGCTTGGACGGCTCGGTGGTGGTCAAGAAGGTCAGC

E L P A G H G L N V N T L S Y G D L A AGAGCTACCCGCCGGGCATGGGCTGAACGTGAACACCCTGAGCTATGGTGACTTGGCCGCTD G V I D P V K V T R S A V L N A S S VGACGGCGTCATCGACCCGGTCAAGGTGACTAGGTCGGCGGTGTTGAACGCGTCATCGGTT

511 A R N V L T T E T V V V D K P A K A E D2281 GCCCGGATGGTACTCACCACCGAGACGGTCGTGGTCGACAAGCCGGCCAAGGCAGAAGAT

5312341

240124612521258126412701276128212881294 1

H D HH H G H A H *CACGACCATCACCACGGGCACGCGCACTGAACTCCGGTCAGCAGACGCAAAAGCCCCCGA

CACGCCGAGCGTGCGGGGGCTTTAGCGTCTGCTCGCGCGGCTTAAGCTGTGCGGCGGATGCCGCGGGTGCGTCCCATGGTGCCCTTGAGGAGTAGGTCGCGCTCGGATTCGGACAGGCCACCCCAAACGCCATAGGGCTCACCGACCTCTAACGCATGGGATCGGCACGCCTCGATCACGGGGCAGCGCCGACACATTTCCTTGGCGCGTTGTTCGCGCTGCGTTCGGGCACGGCCACGCTCGCCGTCGGGATGGAAGAACATCGATGAGTCCATGCCGCGACACAGGCCTTGCAATTGCCAGTTCCAGATGTCTGCGTTGGGTCCCGGTAGCTGCTCCGGCTGTGGCATTGCTGATTCCCTCTCTGCACGACACGCCCAAGTGGAGGTTGGGTTCGCGGACGCGCGAGTTGTGCAACTGAAGTGGAGTGGCGTCACCGATGACTACCGTCAAGGATAGACGCTAGGGGCGCTGCCGAATTTCCGTCAATAGGCGTCCGTTGGCGTGCAATATCGGACCGTTGCGTGAGAATTAACTCTGCGTTCATCTGCGCCGCATTCGCCGCCTGCGCCGCATGCTATCCCGGG

FIG. 1. M. tuberculosis cpnlO and cpn60-1 genes. Nucleotides and amino acids are numbered to the left of each sequence. Only the upper(coding) strand of the DNA sequence is shown. The deduced amino acid sequences of the 10-kDa antigen (CpnlO homolog) and the Cpn60-1protein are shown above the nucleotide sequence. Also shown are (i) matches with consensus sequences for the prokaryotic RBSs, (ii) imperfectinverted repeats reminiscent of prokaryotic factor independent RNA polymerase transcription termination sequences, and (iii) the 5' end of thecpn60-1 mRNA (nt 722).

(data not shown). As seen in Fig. 5, the M. bovis BCG lysatedoes contain an -60-kDa protein that binds to the metalchelate column (lane C). A similarly sized protein is alsorecovered from a lysate of E. coli containing a plasmid(pRL4) that expresses the M. tuberculosis cpn60-J gene usingthe E. coli lac operon promoter (lane B) but not from a lysateofE. coli containing a plasmid (pTB12) expressing the 65-kDa

Chaperonin-10

E. coli

M. tuberculosis

Chaperonin-60

P R R T

GroES GroEL

P R PR T

0 -->IICpnlO

P R

Cpn6O-l

65-kDa antigen(Cpn6o-2)

FIG. 2. Organization of the genes encoding the M. tuberculosisand E. coli Cpn6O homologs. Matches with consensus sequences forpromoters (P), RBSs (R), and rho-independent transcription termi-nation sequences (T) are indicated above the double lines, whichrepresent the DNA sequences. Arrows represent the open readingframes.

antigen gene (lane A). Importantly, both of the =60-kDaIMAC-purified proteins reacted with a polyclonal antiserumraised against the Legionella Cpn6O homolog (data notshown). Also, as expected (27), both E. coli lysates containthree or four other proteins (one is superoxide dismutase)that bind to the column. The crude lysates of M. bovis BCGand XL-1(pTB12) did contain the 65-kDa antigen as deter-mined by a Western blot analysis using monoclonal antibod-ies specific for the 65-kDa antigen (data not shown).

DISCUSSIONPrevious studies have shown that both M. tuberculosis andM. leprae encode their CpnlO and Cpn6O homologs in sep-arate genes (8-13). This was in contrast to the observationthat the CpnlO and Cpn6O homologs in other bacteria wereexpressed as part of an operon (e.g., ref. 17). The resultspresented here show that M. tuberculosis does contain apotential operon that could encode a CpnlO homolog, aCpn6O homolog, and a separate gene encoding a secondCpn6O homolog.The strong band aligning with the thymidine (T) in the

sequence ladder (Fig. 3) most likely represents the 5' end ofthe mRNA for the cpn60-1 gene. Since the sequence ladderrepresents the complement of the sequence shown in Fig. 1,this corresponds to the adenosine (A) at position 722 in Fig.1. Minor primer-extension products of 50, 86, and 107 bp

Dow

nloa

ded

by g

uest

on

Dec

embe

r 27

, 202

0

Proc. Natl. Acad. Sci. USA 90 (1993) 2611

A B C

97 -

66 -

45 -

31 -

21 -

FIG. 3. Primer-extension analysis of total M. bovis BCG RNA.Primer-extension products from DNase-digested RNA (lane 1), fromRNase-digested RNA (lane 2), and in the absence of primer (lane 3)were electrophoresed on 6% polyacrylamide/urea gels. The se-

quence ladder was derived using the same oligonucleotide as theprimer in Sanger dideoxynucleotide sequencing reactions (lanesmarked T, C, G, and A).

were also observed and may represent alternative 5' ends.Several matches with consensus promoter sequences (bothor-70 and a-32 promoters) are present in this region of thesequence, suggesting that the transcription initiation point ofthe cpn6O-1 mRNA is at position 722. However, it must benoted that the relationship between the prokaryotic consen-sus sequences and mycobacterial promoters is not known.Alternatively, the 5' end of the cpn60-1 mRNA could beproduced by a specific cleavage from a larger transcript. Ineither case, despite having a structure resembling an operon,primer-extension studies indicate that the Cpn6O-1 homologis not translated from a polycistronic mRNA that also con-tains the 10-kDa antigen gene.At the protein level, we have shown that a member of the

M. tuberculosis complex expresses both the 65-kDa antigenand an "60-kDa protein that binds to IMAC columns. Weconclude that the -60-kDa IMAC-binding protein expressedin M. bovis BCG is the product of the cpn60-1 gene because

FIG. 5. Expression of Cpn6O-1. Proteins were purified by IMAC.Samples were electrophoresed on a l1o polyacrylamide/SDS Laem-mli gel (28) and the proteins were visualized using a silver stainingtechnique (29). Lanes: A, lysate of E. coli XL-1 containing plasmidpTB12, which expresses the M. tuberculosis 65-kDa antigen (9); B,lysate of E. coli XL-1 containing plasmid pRL4, which expresses theM. tuberculosis cpnlO and cpn6O-1 genes using the plasmid lacpromoter; C, lysate of M. bovis BCG. Positions of the molecularmass standards are shown on the left (kDa).

(i) the deduced amino acid sequence of the protein is con-sistent with its size and ability to bind to IMAC columns, (ii)the IMAC-purified protein of M. bovis BCG comigrates withthe IMAC-purified recombinant-produced Cpn6O-1 protein,and (iii) both the recombinant-produced protein and theIMAC-purified mycobacterial protein react with a polyclonalanti-Cpn6O serum.The deduced amino acid sequences of the Cpn6O-1 protein

and the 65-kDa antigen display 61% sequehce identity. In-terestingly, this is much less than the 95% identity betweenthe 65-kDa antigens of M. tuberculosis and M. leprae (34),strongly suggesting different functional roles for the twoCpn6O homologs in M. tuberculosis. In addition, the distinctbiochemical characteristics of the two Cpn6O homologs raiseinteresting questions as to their potential interactions andbiological functions as chaperonins. It will be of particularinterest to determine whether the two M. tuberculosis ho-mologs can associate to form heteromultimers. [The E. coli

Cpn6O-1 NSKLIEYDETARRAMEVGMDKLADTVRVTLGPRGRHVVLAKAFGGPTVTNDGVTVAREIELEDPFEDLGAQLVKSVATKT 80

65-kDa MAKTIAYDEEARRGLERGLNALADAVKVTLGPKGRNVVLEKKWGAPTITNDGVSIAKEIELEDPYEKIGAELVKEVAKKT 80

NDVAGDGTTTATILAQALIKGGLRLVAAGVNPIALGVGIGKAADAVSEALLASATPVSGKTGIAQVATVSSRDEQIGDLV 120:1111111111 1111 11 111111::1 11:11-: 1.1.11 :I..I..I..II...:I. 1: 1111:

DDVAGDGTTTATVLAQALVREGLRNVAAGANPLGLKRGIEKAVEKVTETLLKGAKEVETKEQIAATAAISAGDQSIGDLI 120

GEAMSKVGHDGVVSVEESSTLGTELEFTEGIGFDKGFLSAYFVTDFDNQQAVLEDALILLHQDKISSLPDLLPLLEKVAG 240

AEAMDKVGNEGVITVEESNTFGLQLELTEGMRFDKGYISGYFVTDPERQEAVLEDPYILLVSSKVSTVKDLLPLLEKVIG 240

TGKPLLIVAEDVEGEALATLVVNAIRKTLKAVAVKGPYFGDRRKAFLEDLAVVTGGQVVNPDAGMVLREVGLEVLGSARR 320 FIG. 4. Comparison of the

AGKPLLIIAEDVEGEALSTLVVNKIRGTFKSVAVKAPGFGDRRKAMLQDMAILTGGQVISEEVGLTLENADLSLLGKARK 320 amino acid sequences of the M.tuberculosis Cpn6O homologs.

VVVSKDDTVIVDGGGTAEAVANRAKHLRAEIDKSDSDWDREKLGERLAKLAGGVAVIKVGAATETALKERKESVEDAVAA 400 The sequence of the Cpn60-1 ho-111-11:1-11:1:1-:1:1-. :1-11:-~1111:11111.11111111111111.11 III. 1111..:1 . molog is the upper sequence andVVVTKDETTIVEGAGDTDAIAGRVAQIRQEIENSDSDYDREKLQERLAKLAGGVAVIKAGAATEVELKERKHRIEDAVRN 400 that of the 65-kDa antigen is the

lower sequence. A vertical lineAKAAVEEGIVPGGGASLIHQARKALTELRASLTGDEVLGVDVFSEALAAPLFWIAANAGLDGSVVVKKVSELPAGHGLNV 480

1 11 *-1-1: 1.111. 1.::-. 11-111 11II.I:.:11--11-:11111111- indicatesidenticalresiduesaandaAKAAVEEGIVAGGGVTLL. QAAPTLDELK. . LEGDEATGANIVKVALEAPLKQIAFNSGLEPGVVAEKVRNLPAGHGLNA 477 colon indicates synonymous sub-

stitutions. The sequences wereNTLSYGDLAADGVIDPVKVTRSAVLNASSVARMVLTTETVVVDKPAKAEDHDHHHGHAH 539aligned using the algorithm of.1 1:11 1.11 111111 III: 1II.I:1 54 aNeedl usingdtheIIorithms(QTGVYEDLLAAGVADPVKVTRSALQNAASIAGLFLTTEAVVADKPEKEKASVPGGGDMGGMDF 540 Needleman and Wunsch (30).

T C G A

Microbiology: Kong et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 27

, 202

0

2612 Microbiology: Kong et al.

GroEL protein forms a multimer containing 14 GroEL mono-mers (17).] The ability to obtain pure preparations of each ofthe Cpn6O homologs should allow dissection of their bio-chemical and structural properties.

Despite the expression of two highly conserved Cpn6Ohomologs in M. tuberculosis, it appears that the extensiveimmunologic and biochemical analyses of a cross-reactive65-kDa protein of mycobacteria (for review, see refs. 2-6)have been directed at the protein encoded by the 65-kDaantigen gene and not at the cpn60-1 gene. This is due in partto much of the work being done using proteins purified fromE. coli recombinants expressing the 65-kDa antigen gene orsynthetic peptides corresponding to the 65-kDa antigen se-quence. However, even studies using whole bacteria orlysates seem to have involved the 65-kDa antigen. Forexample, the N-terminal sequence determined by De Bruynet al. (35) matches that of the 65-kDa antigen. All but one(SL-22) of the seven monoclonal antibody binding sitesmapped on the 65-kDa antigen sequence map to a sequencethat is not shared with the Cpn6O-1 protein (36). More than 20T-cell epitopes have also been mapped onto the 65-kDaantigen sequence (37, 38). Again, most of these regions mapto areas that are unique to the 65-kDa antigen. Perhaps ofgreatest interest here is the observation that the sequences ofthe epitopes involved in adjuvant arthritis (residues 180-188)and in yOT-cell reactivity (residues 180-195) are unique to the65-kDa antigen (39, 40). However, we must note that it is notpossible at this stage to predict the immunologic role of theCpn6O-1 protein. For example, two-dimensional SDS/polyacrylamide gels of M. tuberculosis extracts contain sev-eral spots in the 60- to 65-kDa range that stimulate theproliferation of human T cells (33).The Cpn60-1 protein and the 65-kDa antigen do share

B-cell epitopes since a polyclonal antiserum raised againstthe Legionella Cpn6O homolog does cross-react with both M.tuberculosis proteins (C.J.H., unpublished results). Thiscross-reactivity raises the possibility that other species mightbe assayed for expression of a polyhistidine-containingGroEL homolog by a combination of IMAC and immunoblotanalysis.

Finally, it should be emphasized that, although the Cpn6O-1protein was purified by IMAC, we do not know whether theCpn6O-1 protein binds metal ions in vivo. That is, in theguanidine/urea buffer system used in this study, binding tothe metal chelate adsorbent simply requires the presence ofthree or more adjacent histidine residues (26). The techniquedoes not address activities in vivo.

A.R.M.C. is grateful to Dr. Lucinda Hall and Mr. J. Cookson atLondon Hospital Medical College (United Kingdom) for help withDNA sequencing and computer searches. This work was supportedin part by grants from the National Institutes of Health (T.M.S. andA.R.M.C.) and the Science and Technology for Development Pro-gramme of the European Community (A.R.M.C.). C.J.H. was sup-ported by a National Research Council, Centers for Disease Controland Prevention, National Center for Infectious Diseases ResearchAssociateship.

1. Murray, C. J. L., Styblo, K. & Rouillon, A. (1990) Bull. Int.Union Tuberc. 65, 6-24.

2. Young, R. A. (1990) Annu. Rev. Immunol. 8, 401-420.3. Young, D. B., Mehlert, A., Bal, V., Mendez-Samperio, P.,

Ivanyi, J. & Lamb, J. R. (1988) Antonie van Leeuwenhoek 54,431-439.

4. Shinnick, T. M. (1991) Curr. Top. Microbiol. Immunol. 167,145-160.

5. Kaufmann, S. H. E. & Kabelitz, D. (1991) Curr. Top. Micro-biol. Immunol. 167, 191-208.

6. Winfield, J. B. & Jarjour, W. N. (1991) Curr. Top. Microbiol.Immunol. 167, 161-190.

Proc. Natl. Acad. Sci. USA 90 (1993)

7. Barnes, P. F., Mehra, V., Riviore, B., Fong, S.-J., Brennan,P. J., Voegtline, M. S., Minden, P., Houghton, R. A., Bloom,B. R. & Modlin, R. L. (1992) J. Immunol. 148, 1835-1840.

8. Mehra, V., Sweetser, D. & Young, R. A. (1986) Proc. Natl.Acad. Sci. USA 83, 7013-7017.

9. Shinnick, T. M. (1987) J. Bacteriol. 169, 1080-1088.10. Baird, P. N., Hall, L. M. C. & Coates, A. R. M. (1988)

Nucleic Acids Res. 16, 9047.11. Baird, P. N., Hall, L. M. C. &Coates, A. R. M. (1989) J. Gen.

Microbiol. 135, 931-939.12. Shinnick, T. M., Plikaytis, B. B., Hyche, A. D., Van Land-

ingham, R. M. & Walker, L. L. (1989) Nucleic Acids Res. 17,1254.

13. Thole, J., Dauwerse, H. G., Das, P. K., Groothius, D. G.,Schouls, L. M. & van Embden, J. D. A. (1987) Infect. Immun.55, 1466-1475.

14. Shinnick, T. M., Vodkin, M. H. & Williams, J. C. (1988)Infect. Immun. 56, 446-451.

15. Young, D. B., Lathriga, R., Hendrix, R. & Young, R. A. (1988)Proc. Natl. Acad. Sci. USA 85, 4267-4270.

16. Ellis, R. J. & van der Vies, S. M. (1991) Annu. Rev. Biochem.60, 321-347.

17. Hemmingsen, S. M., Woolford, C., van der Vies, S. M., Tilly,K., Dennis, D. T., Georgopoulos, C. P., Hendrix, R. W. &Ellis, R. J. (1988) Nature (London) 333, 330-334.

18. Mazodier, P., Guglielme, G., Davies, J. & Thompson, C. J.(1991) J. Bacteriol. 173, 7382-7386.

19. Rinke de Wit, T. F., Bekelie, S., Osland, A., Miko, T. L.,Hermans, P. W. M., van Soolingen, D., Drijflhout, J.-W.,Schoningh, R., Janson, A. A. M. & Thole, J. E. R. (1992) Mol.Microbiol. 6, 1995-2007.

20. Young, R. A., Bloom, B. R., Grosskinsky, C. M., Ivanyi,J. D., Thomas, D. & Davis, R. J. (1985) Proc. Natl. Acad. Sci.USA 82, 2583-2587.

21. Shinnick, T. M., Krat, C. & Schadow, S. (1987) Infect. Immun.55, 1718-1721.

22. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) MolecularCloning:A Laboratory Manual (Cold Spring Harbor Lab., ColdSpring Harbor, NY), 2nd Ed.

23. Sanger, F., Nicklen, S. & Coulson, A. (1977) Proc. Natl. Acad.Sci. USA 74, 5463-5467.

24. Devereux, J., Haeberli, P. & Smithies, 0. (1984) Nucleic AcidsRes. 12, 387-395.

25. Patel, B. K. R., Banerjee, D. K. & Butcher, P. D. (1991) J.Microbiol. Methods 13, 99-111.

26. Ljungquist, C., Breitholtz, A., Brink-Nilsson, H., Moks, T.,Uhlen, M, & Nillson, B. (1989) Eur. J. Biochem. 186, 563-569.

27. Hochuli, E., Dobeli, H. & Schlacher, A. (1987) J. Chromatogr.411, 177-184.

28. Laemmli, U. K. (1970) Nature (London) 227, 680-685.29. Heukeshoven, J. & Dernick, R. (1985) Electrophoresis 6,

103-112.30. Needleman, S. & Wunsch, C. (1970) J. Mol. Biol. 48, 443-453.31. Jindal, S., Dudani, A. K., Singh, B., Harley, C. B. & Gupta,

R. S. (1989) Mol. Cell. Biol. 9, 2279-2283.32. Stover, C. K., Marana, D. P., Dasch, G. A. & Oaks, E. V.

(1990) Infect. Immun. 58, 1360-1368.33. Guile, H., Schoel, B. & Kaufmann, S. H. E. (1990) J. Immu-

nol. Methods 133, 253-261.34. Shinnick, T. M., Sweetser, D., Thole, J., van Embden, J. &

Young, R. A. (1987) Infect. Immun. 55, 1932-1935.35. De Bruyn, J., Bosmans, R., Turneer, M., Weckx, M., Mya-

benda, J., van Vooren, J. P., Falmange, P., Wiker, H. G. &Harboe, M. (1987) Infect. Immun. 55, 245-252.

36. Anderson, D., Barry, M. & Buchanan, T. (1988) J. Immunol.141, 607-613.

37. Munk, M. E., Shinnick, T. M. & Kaufmann, S. H. E. (1990)Immunobiology 180, 272-277.

38. Van Schooten, W., Elferink, D., van Embden, J., Anderson, D.& De Vries, R. (1989) Eur. J. Immunol. 19, 2075-2079.

39. Van Eden, W., Thole, J. E., van der Zee, R., Noordzij, A., vanEmbden, J. D. A., Hensen, E. J. & Cohen, I. R. (1988) Nature(London) 331, 171-173.

40. Born, W., Hall, L., Davis, A., Boymel, J., Shinnick, T. M.,Young, D. B., Brennan, P. & O'Brien, R. (1990) Science 249,67-69.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 27

, 202

0