Gene. 117 (1992) 53-60 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0378-1119/92/$05.00

53

GENE 06529

The Mycobacterium tuberculosis 38-kDa antigen: overproduction in Escherichia coli, purification and characterization (Recombinant DNA; prokaryotic expression; inclusion bodies; diagnostic reagent; vaccine)

M. Singh"', A.B. A n d e r s e n c, J . E . G . M c C a r t h y a, M . R o h d e a, H. Sch~tte b, E. S a n d e r s b a n d K.N. Timmis ~ Departments of" Microbiology attd b Bioengineering. GBF-National Research Center.for Biotechnology, Braunschweig (Germany). and " Mycobacteria Departmere. Statens Seruminstitut, Copenhagen S (Denmark). Tel. (45)32683721

Received by H.M. Krisch: 9 August 1991; Revised/Accepted: 12 October/19 October 1991; Received at publishers: 31 March 1992

SUMMARY

The 38-kDa protein (Ag38) of the Gram + bacterium, Mycobacterium tuberculosis H37Rv, is an immunodominant antigen of potential utility for diagnosis and vaccine development. Assessment of this potential requires large amounts of the purified protein that would be difficult, if not impossible, to obtain from M. tuberculosis itself. The gene coding for Ag38 had been previously cloned and in the present study was expressed as an unfused protein in Escherichia coli under the control of strong transcriptional (bacteriophage AptpR) and translational (atpE) signals. Fermentation of the recombinant E. coil K-12 strain CAG629[pMS9-2], which is deficient in Lon protease and the heat-shock response, produced recombinant Ag38 (reAg38) at high levels (about 10% of total cellular protein). The reAg38, which accumulated as inclusion bodies, was completely solubilized in 6 M guanidine. HCI, refolded and purified to apparent homogeneity. The product showed the expected amino acid composition and M r, and had similar reactivities as the native protein with three different mAb. Polyclonal antibodies raised against reAg38 reacted strongly with the native antigen in enzyme-linked immunosorbent assay. These results demonstrate that reAg38, which cannot be distinguished antigenically from the native protein of M. tuberculosis, can be prepared in quantity from E. coll.

INTRODUCTION

M. tuberculosis is the causative agent of tuberculosis, a widespread human disease claiming about 3 million lives

Correspondence to: Dr. M. Singh, Bereich Mikrobiologie, GBF, Mascheroder Weg 1, D-3300 Braunschweig (Germany). Tel. (49-531)6181-415; Fax (49-531)6181-411.

Abbreviations: A, absorbanee (1 cm); aa, amino acid(s); Ag38, 38-kDa antigen; Ap, ampicihin; bp, base pair(s); DTT, dithiothreitol; ELISA, enzyme-linked immunosorbent assay; FPLC, fast protein liquid chromatography; IB, inclusion body(ies); kb, kilobase(s) or 1000 bp; LB, LuriaBertani (medium); M., Mycobacterium; mAb, monoclonal antibody; nt, nucleotide(s); oligo, oligodeoxyribonucleotide;ORF, open reading fame; p, plasmid; PAGE, polyacrylamide-gelelectrophoresis: re, recombinant; SDS; sodium dodecyl sulfate; [ ], denotes plasmid-carrier state.

each year. Important goals of mycobacterial research are to provide protective immunity against tuberculosis through more effective vaccines, and to develop specific skin-test/ sero-diagnostic reagents. A pre-requisite for the attainment of such goals is the characterization and evaluation of the immunological role of individual mycobacterial antigens. For this purpose, substantial amounts of such antigens are required. Purification of antigens directly from M. tuberculosis is difficult because of low cell yields, slow growth rates and the virulent nature of the organism (Young et ai., 1986; Kadival et al., 1987). A potential solution to this problem is the production of recombinant antigens in organisms such as E. coll. The immunological and diagnostic relevance of the Ag38 of M. tuberculosis has been shown previously (Andersen et al., 1986; Young et al., 1986). The protein contains

54 species-specific B-cell epitopes (Andersen et al., 1986), and T-cells isolated from immunized mice, guinea pigs or humans proliferate when cultured in its presence (Young et al., 1986; Kadival et al., 1987; Worsaae et al., 1987). The majority of humans (especially of the HLA type DR2) suffering from active tuberculosis develop antibodies against the Ag38 (Bothamley et al., 1989). Here, we describe high-level production of the reAg38 of the virulent M. tuberculosis as an unfused protein in E. coli, its purification and the demonstration of its immunological similarity to native antigen.

start codon (Andersen et al., 1989). There is also a 24-aa long signal sequence (Fig. 1A) which had similarity to those of bacterial lipoproteins. In the present study, the gene was manipulated so that high-level synthesis of the unfused reAg38 could be achieved in the pJLA603 expression vector (Schauder et al., 1987). For cloning and expression in these vectors, it is required that the foreign gene contains a restriction site, e.g., Ncol, Ndel, or Sphl, which has an in-frame ATG within its recognition sequence. Since there is no such site around the start codon of the Ag38 gene, oligo mutagenesis in M 13mp 19 was carried out to create an NdeI site at the N terminus by changing the start codon from GTG to ATG (Fig. 1). The 1.2-kb Ndel-Sphl fragment that could be excised from the M I3 derivative after mutagenesis was then cloned between the Ndel and Sphl sites of pJLA603. The re-plasmid, which was designed to produce the reAg38 protein with its original signal peptide

RESULTS AND DISCUSSION

(a) Construction of expression plasmids The nt sequence of the Ag38gene had an ORF encoding a polypeptide of 374 aa and containing GTG as the

#

All s' aPE r,a

AGG~,.,CAA^AAAGu uU" ~UUl~C C

.

,~ ".

atpETIR _

"G

A

Be

UAuUU CG CGCA UCGUCGCCGCGUCGCCAGUUGUGCCGGUUGU U ACCAACA AC UU CU AU UA AG UCG

(AG = - 5.5 kcallmol)

5' "

A A

M L L A V L T A A P L L L A A A G C G CATATGCTGI"I'QGCCGTGTrGACCGCTGCGCCGCTGCTGCTAGCAGCGGCGGGCTGTGGC C . Nde I

1

CGU

(AG = . 18,7 kcallmol)

U AGCcGUCGG,, c C _P~uA U " '-QcQ u A_ ~; C G r,CGCCAGUUGUGCCGGUUQU U '= "C H .CCAAC A .C Uu AAU

AQu uca

D.

UA C GA~' U GA ~

cAYc, %UuAAC A U A

A A AC UG

Fig. 1. Nucleotide sequences of the 5'-ends of engineered genes encoding thc prc-form (A) and truncated form (C) of Ag38. The Ndel (CATATG) site was created by oligo-mutagcnesis using the oligo 5'-AGCACAGAAAGGTATCATATGAAAATTCGTTTGCATA (for A) and the oligo 5'-ATrCGTI'I'GCATATGCTG-I-rGGCCGTGT (for C). The predicted aa sequence is presented in the one-letter code and the short arrow above the sequence shows the putative processing site. The large arrows show the corresponding RNA secondary structures in the region of the ATG start codon of the pre-form (B) and the truncated form (D) as calculated by the method of Zucker and Stiegler (1981). In B and D, the ribosome-binding site of the atpK gene and the start codon of the Ag38 gene are marked with a line and an arrow, respectively.

55 intact, was designated pMS9-2 (Fig. 2). In the same way, another re-plasmid pMS10-4 containing a deletion of the first 6 aa in the signal sequence was constructed (Fig. 1C). Computer analysis of the translation start region of pMS92-specified mRNA predicted a loose secondary structure (Fig. IB) which should be compatible with high level expression in E. coil (McCarthy and Bockelmann, 1988). On the other hand pMS 10-4 showed a more stable secondary EcoRI I b l a ~ ~ ~ ~

Sphl EcoRI -~I I

ts857 cloned in M13

r

1 1

oligo mutagenesis o~d-b'anscriptional terminator

Ndel I

Sphl .A 1.2 kb

T4 DNAUgase

!

c [ ts857 bl

structure (Fig. ID) indicating that the expression from this plasmid might not be as good as from pMS9-2. Both replasmids were used for expression studies.

(b) Expression of the reAg38 in small-scale culture Several E. coli strains were tested for production of reAg38 encoded by pMS9-2 and pMS10-4. The strain CAG629 (Ion, htpR::TnlO; C. Gross) showed the strongest expression. Extracts of CAG629[pMS9-2 ] cells induced at 42°C contained substantial amounts of the re-protein (Fig. 3A, lanes 3 and 4), whereas it was absent in extracts from uninduced cells (Fig. 3A, lane 1). The re-strain exhibited no obvious growth perturbations after induction (data not shown). Immunoblotting with mAbs HBTI2 (Fig. 3A, lanes 10, and 11), HAT2 and HYT28 (data not shown) yielded positive reactions with reAg38. Most of the reAg38 was present in the cell pellet fraction of disrupted cells and only a small fraction was detected in the supernatant fluid (Fig. 3A, lanes 10 and 11). The r~-clone CAG629[pMSI04], as expected from the secondary structure prediction, produced considerably less protein than pMS9-2 (Fig. 3A, lanes 13 and 14). The faint bands seen on the immunoblots which ran slower than Ag38 correspond to SDS-insoluble, aggregated forms of the reAg38. The reAg38 was produced at high levels (about 10% of the total cellular protein) as measured by laser densitometry of the silver-stained S D S - P A G E gels. As is often the case with re-proteins produced at high levels in E. coil, reAg38 was mostly contained in cytoplasmic aggregates or IB (data not shown).

(c) Fermentation of recombinant Escherichia coli and purification of reAg38 [

pMSg-2 \

\

~

Ndel

t 1 ~S-"I

\ fd-transcripUonal terminator Fig. 2. Construction of expression plasmids. A 2.0-kb EcoRl fragment containing the Ag38 gene was cloned from 21059 (Andersen ct al., 1986) into M13mp19 and mutagenized using the Amersham kit (RPN1523) to create a Ndel (CATATG) in such a way that it included the start codon of the gene. A 1.2-kbNdel-Sphl fragment was then subcloned between the Ndel and Sphl sites of the expression vector, pJLA603, which contains the bacteriophage 2 PLand PRpromoters in tandem, the atpE translation initiation region (white box), the fd transcriptional terminator and translational stop codons in all the three reading frames (Schauder et al., 1987). Piasmidsare not drawn to scale. Preparation and handlingof DNA was according to standard protocols (Maniatis et al., 1982).Transformation was performed as described by Hanahan (1983). DNA sequencing was done by the dideoxynucleotide chain-termination method (Sanger et al., 1977).Oligoswere synthesized using an Applied Biosystemsmodel 380B DNA synthesizer and purified with OPC columns (Applied Biosystems Inc.).

Re-clone CAG629[pMS9-2] was used for a 30-liter fermentation because it produced and tolerated reAg38 at high levels in batch cultures, and coded for reAg38 with an intact signal sequence. The time course of production of reAg38 in the bioreactor was monitored (Fig. 3B). S D S PAGE of the whole cell extracts showed that within 3,0 min after increasing the temperature from 30°C to 42 'C, a doublet of approx. 38 kDa was clearly visible on silverstained gels. Washing of the IB obtained from the fermentor culture with buffer containing Triton X-100 resulted in the removal of some of the contaminating proteins (Fig. 4A, lane 3) without any apparent loss of reAg38 (Fig. 4A, lane 2). Using a Sephadex G-25 column, reAg38 was desalted and renatured (Fig. 4A, lane 6); we did not observe any reaggregation of reAg38 at this stage. Further purification of the antigen was obtained by multiple rounds of FPLC-anion-exchange chromatography (Fig. 4B) where the antigen was found to elute at 100 mM NaCI (Preparation I) and between 130-200 mM NaCl (Preparation III). Preparation II shown in Fig. 4B represents the antigen pu-

56

A 1 2 3 4 5 6 7 8 9 1011121314

t

l-W/

-___ -

B kDa

......

""'r,

66- I 45"I

36~-~ 24 Fig. 3. Expression of the re-Ag38 gene in E. coil CAG629 (/on, htpR) in batch cultures (A) and in a bioreactor (B). (Panel A) Silver-stained gels (lanes 1.-7) and the corresponding immunoblots with mAb HBTI2 (lanes 8-14) showing the proteins produced by CAG629[pMS9-2] after growth at 30°C (lanes 1/8 and 2/9), and at 42°C (lanes 3/10 and 4/11). The protein patterns of CAG629[pMS 10-4] after induction at 42°C are shown in lanes 6/13 and 7/14. Lanes I, 3, and 6 represent the soluble fractions and lanes, 2, 4, and ? the IB. Lanes: 5, standard (values given on the left are in kDa); 12 (prestained standard). Strains were grown in LB medium containing Ap (I00 pg Ap/ml) at 30°C to A.~Hoof 0.6, Cultures were induced by shifting to 42°C for 3 h in a shaking water bath. Bacteria were harvested from I ml culture, suspended in I00141 of sample buffer (62 mM Tris, HCI pH 6.8/2% SDS/0,7 M 2-mercaptoethanol/ I0~ glycerol/0.002% bromophcnol blue) and broken by sonication in ice (3 × 30 s; 50 W) using a Braun Labsonic 2000, Samples were heated at 95°C for I0 rain and I0 l~l were analysed by 0,1% SDS-12.5% PAGE, After eleetrophoresis the polypeptides were visualized by silver staining (Damerval et al., 1987). Protein concentrations were determined by the method of Lowry et al. (1951). Proteins were transferred to a nitrocellulose membrane (BioRad) with a home-made semi-dry blotting apparatus using 25 mM Tris/192 mM glycine/20~ methanol pH 7,4, Non-specific binding was blocked by incubating the filter in TBS (50 mM Tris.HCl/ 200 mM NaCI pH 7.5) containing a I0~ solution of milk (0.3% fat). The primary antibodies (mouse mAbs) were diluted 1000-fold in TBS and incubated with the filters overnight at 4°C. The filters were washed three times with TBS and immunodetection was carried out with a biotinylated anti-mouse IgG and streptavidin-alkaline phosphatase conjugate (BRL, Gaithersburg, MD). For the immunodot.blot assay, protein samples were filtered through a nitrocellulose membrane using a BioRad bio-dot apparatus and processed further as described above. The mAbs HAT2, HBT12, HYT28 have been described earlier (Schou et ai., 1985; Andersen et al., 1986; Ljungquist et al., 1988). Densitometric measurements of silver-stained gels and Western blots were done on a laser densitometer (LKB). (Panel B) Accumulation of the reAg38 in the bioreactor at 30°C (lane 3) and after induction at 42 °C (lanes 4-9). Of a modified concentrated LB medium (tryptone, 40 g per l/yeast extract, 20 g per liter/

rifled on F P L C - a n i o n - e x c h a n g e column in presence o f Triton X-100 from the aggregated and contaminated fractions obtained from earlier anion-exchange c h r o m a t o g r a p h i c steps. The native Ag38 runs as a doublet on S D S - P A G E . The reason for this is not k n o w n but it might be due to acylation or processing of the pre-pro*ein. In the large-scale preparations we did not remove the signal sequence before producing reAg38 in large a m o u n t s because the presence of the lipoyl moieties on the N-terminal cystein m a y play important role in the immunogenicity of protein antigens (Deres et al., 1989). All the reAg38s purified in this study showed the doublet characteristic. Preparation I contains mostly the upper band and only minor amounts of the lower band. Preparation II contains both the upper and the lower b a n d s in almost equal amounts. Preparation III represents a proteolyticaUy truncated derivative (34 k D a ) of reAg38.

(d) Structure, immunological reaction and immunogenicity of purified reAg38 The immunological reactions of the purified antigen preparations were tested with m A b s H A T 2 , H B T I 2 and H Y T 2 8 . All the three antibodies reacted with reAg38 preparations (Fig. 4B). W h e n the affinity-purified native antigen w a s c o m p a r e d with the re-antigen (Preparation I) by S D S P A G E and immunoblotting, no difference w a s observed (Fig. 5A). The silver-stained gel and the immunoblot shown in Fig. 5A were also traced with a laser densitometer and, after normalizing for differences in the a m o u n t s o f protein present, the native and the re-antigen were seen to have given identical reactions with the m A b s H B T I 2 (Fig. 5B), H A T 2 and H Y T 2 8 ( d a t a not shown), The immunogenicity of the re-antigen was also tested by raising polyclonal antisera in rabbits against the native and the re-antigen under similar conditions. The two sera were then assayed by E L I S A against both reAg38 (Fig. 6A) and native Ag38 (Fig. 6B). The slopes of the curves are identical showing that the two sera bind native Ag38 a n d reAg38 equally well, The aa composition of the reAg38 closely resembles the

NaCl, 5 g per liter) 30 liter were sterilized in a 50-1iter bioreactor (Biostat U30D, Braun Melsungen, Germany) in the presence of 3.5 ml Ucolub N38 antifoam (Brenntag, M01heim, Germany), Cultivation was initiated by inoculation with a 0,5-1iter overnight culture of the organism in LB medium to give an initial As46= 0.04. Stirrer speed was maintained constant at 300 rpm, which ensured a dissolved oxygen concentration close to saturation, and pH was held at 6.9. After 4 h of fermentation (A54o = 0.4) the temperature was shifted from 30°C to 42°C, and mainrained at this level for a further 4 h. Samples were withdrawn every 30 min after induction. Lane I shows the standard and lane 2 a small amount of reAg38 protein purified from IB of a batch culture of CAG629[pMS9-2]. The arrow marks the double band corresponding to reAg38. E. coil CAG629 is described in section b.

57

A 1

2

3

4

5

6

7

,~mmmmkmp

66_qb

•

45---

m

36--

II L

m

B S1

23

P4

5

6

P 7 8

9

P 10 1112

~.~,

B

45-i 36"BgP ~ a

N I

"" 4ib

Fig. 4. 0.1% SDS-12.5% PAGE (silver stain) analysis of different purification steps (panel A), and final purification and analyses of the purified reAg38 preparations (panel B). (Panel A) At the end of the induction period the broth from the bioreactor was concentrated in a closed system by crossflow microfiitration (Enka module type A7 ABA 3A) with 0.23 m 2 Accural membrane (0.2 #m pore diameter; Enka, Wuppertal, Germany) until a final volume of 10 liter was obtained. Cells obtained from the bioreactor were disrupted by a single pass through a high-pressure homogenizer LAB 60/500/2 (A.P.V.-SchrOder, L0beck, Germany) at 500 bar with a flow rate of 60 liter/h. IB were crudely separated using a centrifugal separator SA 1-01-175 (Wcstfalia, Oclde, Germany). This device allows isolation of IB from the broth at a flow rate of 15-20 liter/h. The IB were washed twice with 200 ml of buffer L (50 mM Tris. HCI/10 mM EDTA pH 8.0) containing 2% Triton X-100. Washed pellets were resuspended in 3 liter of buffer L containing 6 M guanidine. HCI/20 mM DTT for 16 h at 4°C with slow stirring. After centrifugation at 7000 x g, the supernatant fluid was passed through a Sephadex G-25 gel filtration column (10 x 90 cm) equilibrated with 10 mM Tris. HCI buffer pH 7.0, containing 100 mM NaCi to remove the guanidine. HC! and effect renaturing of the reAg38. The solution was applied in 2-liter aliquots with a flow rate of 7 liter/h and the eluate was monitored for A20o and conductivity. The desalted and renatured antigen solution was recovered in starting buffer with a conductivity of 8 mho, well separated from the salt peak containing mainly guanidine.HCI. The antigen peaks were combined and diluted with distilled water to obtain a conductivity of 5 mho and the pH adjusted to 8.5 with 1 M Tris base. The solution was ~.ivided into two parts for the following purification step. One part of the solution was applied to an FPLC column (QAE-Scpharose FF; 5 x 18 cm) equilibrated with 20 mM Tris. HCI buffer pH 8.0. ,The flow rate was 1.72 liter/h corresponding to a linear flow rate of 86.6 cm/h. After extensive washing with the starting buffer, elution of the antigen was performed by application of a step gradient consisting of 50 raM, 100 mM, 250 raM, 500 mM and ] M NaCI in starting buffer. Afterwards the column was re-equilibrated with starting buffer, the second portion from the gel filtration column was applied and cluted as described before. The antigen-containing fractions from the two QAE-Sepharose runs were pooled, concentrated and diafiltered (2 mho) by ultrafiltration using ~.a Amicon Hollowfiber Cartridge (type H IP 10; cutoff at M r 10000). Lanes 1 and 4 show the standard. Lanes: 2, solubilized IB; 3, proteins released during washing of IB before solubilization; 5, proteins renatured on Sephadex G25; 6+7, QAE-Sepharose eluates. The arrowhead in panel A indicates the reAg38. (Panel B) Final purification and SDS-PAGE and immunoblot analyses of the FPLC-purified reAg38 preparations. Around 20 mg of the protein obtained from QAE-Sepharose step (see legend to panel A, above) was applied to a Mono Q HR (5/5) FPLC-column equilibrated with 20 mM Tris. HCI pH 8.0. The column was washed with starting buffer at a flow rate of 2 ml/min and at a pressure of 25 bar. Thereafter, the antigen was eluted by gradually increasing the NaCI concentration to 0.5 M in starting buffer. Altogether three Mono Q HR (5/5) runs were carried out under the conditions described above. Samples containing about 1/~g proteins were separated by 0.1% SDS-12.5% PAGE and either silver-stained (lanes 1-3), or immunoblotted using anti-Ag38 mAbs: HAT2 (lanes 4--6), HBT12 (lanes 7-9), and HYT28 (lanes 10-12). The re-protein that reacted with the three mAbs was found in two main peaks: one at 100 mM NaCI (lanes I, 4, 7, and 10), and the other at 130-200 mM NaC! (lanes 3, 6, 9, and 12). Lanes 2, 5, 8, and I1 show the protein purified in the presence of Triton X-100 (eluting at 170 mM NaCI). The standard is in lane S (in kDa, as shown on the left) and lane P represents the pre-stained marker used during immunoblotting.

58

A

A 1

2

3

4P

kDa

At.9o

" - - - - 47

3.0 2.0

----

33

1.0

B Re. 811v~

I / \~

/""X

~R

reeal'.ml.m~l°

two fold dilutions

B At,90 3.0 2.0

Fig. 5. Comparison of the native Ag38 and the reAg38. (Panel A) About ! pg of the affinity purified native Ag38 (lanes 1 and 3) and the reAg38 preparation-I (lanes 2 and 4) were separated on 0.1% SDS- 12.5% PAGE and either silver-stained (lanes I and 2) or immunoblotted using mAb HBTi2 (lanes 3 and 4). Lane P indicates the pre-stained standard, Arrowhead indicates the Ag38 bands, as in Fig, 4A. (Panel B) Laser densitometric comparison of the immunoreactivity of the native Ag38 and the reAg38, The silver-stained gel and the immunoblot shown in panel A were scanned with a laser densitometer, Peak areas of the silver-stained native (Nat, Silver) and the reAg38 (Re, Silver), as well as those of the immunoblotted native Ag38 (Nat, Immano) and the reAg38 (Re. Immune) are shown, Protein samples were mixed 1:1 with 2 x sample buffer and heated at 95°C for 10 min before loading onto the gels,

1.0

Fig. 6

Fig. 6. Comparison of the native Ag38 and tile reAg38 using polyelonal sera from rabbits immunized with native Ag38 (Q) or reAg38 ( I ) . Serial two-fold dilutions starting with a 1:100 dilution were titrated in microtiter plates coated with the preparation ! (panel A) of reAg38 (0.1/zg/ well). Bound immunoglobulinswere detected by horse radish peroxidaseconjugated swine anti-rabbit immunoglobulins diluted !: 1000 (P 217,

two fold dilutions

Dakopatts, Glostrup, DK). (Panel B) As for panel A except that the microtitcr plates were coated with the native Ag38 (0. I l~g/well). Methods. To obtain polyclonal sera, rabbits were immunized subcutaneously with either affinity-purified native Ag38 (Worsaae et al,, 1988) or with reAg38 (Preparation I). The antigens (10 leg per dose) were adsorbed to aluminium hydroxide (2.4 rag) and subsequently mixed with 1 ml of Freund's incomplete adjuvant. The rabbits were immunized three times with intervals of two weeks. The blood was drawn ten days after the last immunization and the IgG fraction was purified as described by Harboe and Ingild (1983).

59 aa composition derived from the nt sequence (data not shown).

(e) Conclusions (1) A strategy was developed and implemented to clone an M. tuberculosis DNA fragment in expression vector such that native, unfused Ag38 would be produced at high levels. About 15 mg reAg38 per liter was produced under the conditions given. It should be emphasized, however, that the objective of the present work was to determine whether overproduced reAg38 could be recovered in an antigenic form immunologically indistinguishable from native antigen obtained from M. tuberculosis, and not to optimize fermentation yields. (2) Most of the reAg38 accumulated as IB. 6M guanidine.HCl in the presence of a reducing agent (DTT) at high concentration was found to be suitable for the solubilization of the IB. Renaturation of IB proteins, specially hydrophobic, membrane proteins, is often difficult and needs careful optimization of experimental conditions. In our case, the usual procedure of dialysis or step-dialysis resulted in the precipitation of reAg38 even at very low protein concentrations (0.05mg/ml). Renaturation of reAg38 on Sephadex G-25 proved on the other hand to be effective and no significant reaggregation of the protein was observed. Starting with IB containing about 200 mg total protein, we v, ere able to purify 19 mg of reAg38 with more than 95~o purity as judged by silver staining of S D S polyacrylamide gels. Part of this has been deposited in the WHO-Mycobacterial Protein Bank in Bilthoven, The Netherlands (J. van Embden). (3) The native Ag38 isolated from the culture supernatant of M. tuberculosis and the purified re-proteins present in preparations I and II are of the same size and show the doublet character on SDS-PAGE. Furthermore, the purified proteins and the native antigen showed identical reactions with mAbs HAT2, HBTI2, HBT28 and polyclonai serum raised against native or re-protein (Preparation I) demonstrating that the reAg38 possesses similar epitopes and is as immunogenic as the native Ag38 purified from M. tuberculosis.

(4) We have developed an expression system and production and purification procedure for Ag38 of M. tuberculosis in E. coil that permits the ready isolation of significant quantities of reAg38. The reAg38 is immunologically indistinguishable from the native antigen. These results should facilitate the assessment of the utility of reAg38 for diagnosis and vaccine development. ACKNOWLEDGEMENTS

We thank Dr. U. Menge and R. Getzlaff for aa composition analysis; E. Joud and R. Kraume-Fltlgel for expert

technical assistance, and W.-D. Deckwer for his support of the bioreactor and IB isolation part of the effort. This work was partially supported by a WHO grant.

REFERENCES Andersen, A.B. and Hansen, E.B.: Structure and mapping of antigenic domains of protein antigen b, a 38000-molecular-weight protein of Mycobacterium tuberculosis. Infect. Immun. 57 (1989)2481-2488. Andersen, A.B., Yuan, Z.-L., Haslov, K., Vergmann, B. and Bennedsen, J.: Interspecies reactivity of five monoclonal antibodies to Mycobacterium tuberculosis as examined by immunoblotting and enzyme-linked immunosorbent assay. J. Clin. Microbiol. 23 (1986) 446-451. Andersen, A.B., Worsaae, A. and Chaparas, S.D.: Isolation and characterization of recombinant lambda gtl I bacteriophages expressing eight different mycobacterial antigens of potential immunological relevance. Infect. Immun. 56 (1988) 1344-1351. Bothamley, G.H., Beck, J.S., Schreuder, G.M.T., D'Amaro, J., De Vries, R.R.P., Kardijito, T. and lvanyi, J.: Association of tuberculosis and Mycobacterium tuberculosis.specific antibody level with HLA. J. Infect. Dis. 159 (1986) 549-555. Damerval, C., le Guillnux, M., Blaisonneau, J. and de Vienne, D.: A simplification of Heukeshoven and Dernick's silver staining of proteins. Electrophoresis 8 (1987) 158-159. Deres, K., Schild, H., WiesmUller, K.-H., Jung, G. and Ramensee, H.G.: In vivo priming of virus-specific cytotoxic T lymphocytes with synthetic lipopeptide vaccine. Nature 342 (1989) 561-564. Hanahan, D.: Studies on transformation ofEscherichia coli with plasmids. J. Mol. Biol. 166 (1983) 557-580. Harboe, N. and Inglid, A.: Immunization, isolation of immunoglobulins and antibody titre determination. Scand. J. Immunol. 17S 10 (1983) 345-35 I. Kadival, G.V., Chaparas, S.D. and Hussong, D.: Characterization of serologic and cell-mediated reactivity of a 38kDa antigen isolated from Mycobacterium tuberculosis. J. Immunol. 139 (1987) 2447-2451. Laemmli, U.K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 (1970) 680-683. Ljungqvist, L., Worsaae, A. and Heron, I.: Antibody responses against Mycobacterimn tuberculosis in 11 strains of inbred mice: noval monoclonal antibody specificities generated by fusions, using spleen from BALB.BI0 and CBA/J mice. Infect. Immun. 56 (1980) 1994-1998. Lowry, O.H., Farr, A.L., Rosebrough, M.J. and Randall, R.: Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193 (1951) 265-275. McCarthy, J.E.G. and Bokelmann, C.: Determinants of translational initiation efficiency in the atp operon of Escherichia coll. Mol. Microbiol. 2 (1988) 455-465. Maniatis, T., Fritsch, E.F. and Sambrook, J.: Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982. Roth, J., Bendayan, M., Carlemalm, E., Villiger, W. and Garavito, M.: Enhancement of structural preservation and immunocytochemical staining in low temperature embedded pancreatic tissue. J. Histochem. Cytochem. 29 (198I) 663-669. Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing with chainterminating iahibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 54635467. Schauder, B., BlOcker, H., Frank, R. and McCarthy, J.E.G.: Inducible expression vectors incorporating the Escherichia coli atpE translational initiation region. Gene 52 (1987) 279-283.

60 Schou, C., Yuan, Z.-L., Andersen, A.B. and Bennedsen, J.: Production and partial characterization of monoclonal hybridoma antibodies to Mycobacterium tuberculosis. Acta Pathol. Microbiol. lmmunol. Scand. Sect. C 93 (1985) 265-272. Styblo, K.: Overview and epidemiological assessment of the current global tuberculosis situation with an emphasis on control in developingcountries. Rev. Infect. Dis. 11 (1989) 5339-5346. Worsaae, A., Ljungqvist, L., Haslov, K., Heron, I. and Bennedsen, J.: Allergenic and blastogenic reactivity of three antigens from Mycobacterium tuberculosis in sensitized guinea pigs. Infect. lmmun. 55 (1988) 2922-2927. Worsaae, A., Ljungqvist, L and Heron, I.: Monoclonal antibodies produced in BALB.BI0 mice define new antigenic determinants in cul-

ture filtrate preparations of Mycobacwrium tuberculosis. Infect. lmmun. 26 (1988) 2608-2614. Young, R.A., Bloom, B.R., Grosskinsky, C.M., Ivanyi, J., Thomas, D.D. and Davis, R.D.: Dissection of Mycobacterium tuberculosis antigens using recombinant DNA. Proc. Natl. Acad. Sci. USA 82 (1985) 2583-2587. Young, D., Kent, L., Rees, A., Lamb, J. and Ivanyi, J.: Immunological activity of a 38-kilodaiton protein purified from M.t,cobacterium tuberculosis. Infect. lmmun. 54 (1986) 177-183. Zucker, M. and Stiegler, P.: Optimal computer folding of large RNA sequences using thermodynamics and auxiliary information. Nucleic Acids Res. 9 (1981) 133-148.