OUP user menu

Assessment of the serodiagnostic potential of nine novel proteins from Mycobacterium tuberculosis

Alison J. Moran , Janice D. Treit , Janice L. Whitney , Bassam Abomoelak , Raymond Houghton , Yasir A.W. Skeiky , Diana Pedral Sampaio , Roberto Badaró , Francis E. Nano
DOI: http://dx.doi.org/10.1111/j.1574-6968.2001.tb10615.x 31-36 First published online: 1 April 2001


To identify antigens that would improve the accuracy of serological diagnosis of active tuberculosis, we cloned the genes encoding nine potentially immunogenic secreted or surface-associated proteins of Mycobacterium tuberculosis. Recombinant proteins were reacted with sera from HIV-negative individuals with extrapulmonary tuberculosis (EP-TB) or HIV-positive individuals with pulmonary tuberculosis (TBH). Specific and high level antibody responses were obtained for four recombinant proteins, of which antigen GST-822 was recognized by 60% of EP-TB and 42% of TBH and antigen MBP-506 was recognized by 45% of EP-TB and 61% of TBH. These results suggest that these proteins are strong candidates as subunits in a polyvalent serodiagnostic test.

  • Serodiagnosis
  • Tuberculosis
  • Antigen

1 Introduction

Approximately one third of the world's population are infected with Mycobacterium tuberculosis, and there are 7–8 million cases of active tuberculosis (TB) per year [1]. Among the challenges represented by this problem is the need to develop easy to use, cost effective diagnostic assays [1]. The most common diagnostic assay for M. tuberculosis infection is the tuberculin skin test. This test measures the delayed-type hypersensitivity response to an intradermal injection of purified protein derivative [2]. In areas where TB is endemic, or there are abundant environmental non-tuberculous mycobacteria, the tuberculin skin test is compromised and thus, has poor diagnostic value [2].

TB is a common HIV-related opportunistic infection [3]. In some populations up to 30% of individuals that have pulmonary TB are also HIV-positive [4]. Furthermore, dually infected individuals have an increased chance of developing extrapulmonary disease [5]. The serodiagnosis of individuals with extrapulmonary M. tuberculosis or those dually infected with M. tuberculosis and HIV is specifically complicated by a decreased cell-mediated response to some M. tuberculosis antigens [68].

Serologic tests, using an enzyme-linked immunosorbent assay (ELISA), are a simple and inexpensive alternative to the tuberculin skin test. However, there is considerable heterogeneity of both antigen recognition and response among individuals with pulmonary TB and this heterogeneity has severely hampered the performance of single antigen serodiagnostic assays [911]. It is now generally recognized that serodiagnosis may be improved by the identification and inclusion in a cocktail of a number of antigens that react with the sera from a high proportion of infected individuals [12,13]. As yet, however, the multi-antigen cocktails have not attained the sensitivity levels available using rapid molecular testing [12,14]. Thus, the requisite improvements in sensitivity may be achieved by the identification of additional immunodominant M. tuberculosis antigens, which could be included in the multi-antigen cocktails.

2 Materials and methods

2.1 Growth media and plasmids

Table 1 lists plasmids used and constructed in this study. Luria-Bertani medium [15] with appropriate antibiotics (ampicillin 250 μg ml−1; chloramphenicol 10 μg ml−1) was used for the growth of Escherichia coli strains. Preparation and transformation of competent cells were carried out as previously described [16].

View this table:
Table 1

Recombinant plasmids for cloning and expression of the full-length proteins

Plasmid constructExpression vectorCloning sitesSanger IDFusion productPredicted Mr
pAM825EpMAL-c2E, HMTCY31.03cMBP-82560
  • E, EcoRI; B, BamHI; H, HindIII; c, complementary direction.

2.2 Selection and DNA sequence analysis of mtb-phoA fusions

Random M. tuberculosis-alkaline phosphatase (Mtb-PhoA) fusion proteins were selected from a clone bank by plate assay, the encoding DNA sequenced, and identified as previously described [17]. Signal peptide determination was made using the SignalP neural network trained on Gram-positive data [18].

2.3 Novel antigen open reading frame (ORF) analysis

DNA sequences from the mtb-phoA clones were aligned with the DNA sequence generated by the Sanger Centre [19] and the TIGR (http://www.tigr.org/tdb/CMR/gmt/htmls/SplashPage.html) M. tuberculosis genome projects. In many cases we had to determine the extent of the ORF and verify that the translational reading frame was the same as in our phoA fusion. In some cases, the extent of the ORF had already been assigned and the genome project's assessment was used for expression cloning experiments.

2.4 PCR primers and conditions, cloning and expression

Oligonucleotide primers used for PCR amplification are shown in Table 2. The oligodeoxynucleotide primers were designed with restriction sites (underlined) for cloning of amplified fragments into expression vectors.

View this table:
Table 2

Oligonucleotide primers used for PCR amplification

PrimerSequence 5′ to 3′
  • Restriction sites are underlined. BamHI, GGATCC; EcoRI, GAATTC; HindIII, AAGCTT.

Each PCR reaction used different conditions and the specific reaction conditions will be supplied on request. All PCR reactions were conducted in 20 μl using a 6:1 Taq polymerase:Pfu polymerase enzyme combination. The reaction mixes contained either 1 μl DMSO or 4 μl of Q solution (Qiagen) as a denaturant, primer concentrations were usually 0.75 μM and the DNA source was chromosomal M. tuberculosis H37Rv DNA. A manual hot start was used for all PCR reactions which consisted of an initial denaturation (95°C, 4 min) followed by 25 cycles of denaturation (95°C, 1 min), annealing (between 60°C and 70°C, 30 s) and then extension (72°C, 2 min) followed by a final extension (72°C, 4 min). Standard protocols were followed for cloning [15].

The different expression vectors used included pMAL-c2 (New England Biolabs), pGEX-4T3 (Pharmacia) and pET-17xb (Novagen). Respectively, these expression vectors enabled the N-terminal fusion of the maltose binding protein (MBP), glutathione S-transferase (GST) or the 260-amino acid T7 gene 10 product (PET) containing the T7-Tag? to the products of the cloned DNA.

Fusion proteins were expressed in the commercial E. coli strains E. coli BL21 (Novagen), with pLysS in E. coli BL21 (DE3) (Novagen), or in E. coli SURE (Stratagene). The following recombinant proteins were overexpressed in E. coli BL21 pLysS: MBP-506, MBP-825, PET-639, PET-47, PET-916, PET-1084; in E. coli BL21: GST-152, GST-822; and in E. coli SURE: GST-206. Table 1 identifies the cloning strategy including the vectors used, as well as the constructs and the fusion proteins obtained.

2.5 Western immunoblotting of novel antigens

Sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) was carried out using 10% slab gels in a continuous buffer system [20], and proteins were electrophoretically transferred from the gel to a nitrocellulose membrane using standard protocols [15]. Antibody detection for the GST (Pharmacia), T7-Tag? (Novagen) and MBP (New England Biolabs) tagged proteins was conducted as per the supplier's instructions. The chemiluminescent RenaissanceR system (DuPont NEN Renaissance, NEL-201) was used to image bound antibody.

2.6 Antigen preparation

Over-night cultures of recombinant E. coli strains were diluted 1:100, grown to mid-exponential phase and induced with IPTG (1 mM) for 3 h. The bacterial pellet was harvested, sonicated in phosphate-buffered saline (PBS), subjected to centrifugation and the appropriate fraction separated by preparative SDS–PAGE. Fractions containing full-length proteins for serodiagnosis were collected using a whole gel eluter (Bio-Rad) in a 50 mM Tris, 25 mM boric acid (pH 8.7) buffer.


Sera were obtained from 38 Brazilian individuals with TBH, from 20 HIV-negative individuals with extrapulmonary TB (EP-TB) and from 17 healthy volunteers. The collection of sera followed the ethical guidelines of the host institution.

Microtiter wells were coated with 200 ng of antigen, either M. tuberculosis cell lysate or fusion protein, in 50 μl of coating buffer (15 mM Na2CO3, 35 mM NaHCO3 adjusted to pH 9.6) and incubated for 1 h. Plates were then aspirated and 250 μl of blocking buffer (0.5% bovine serum albumin and 0.01% Thimerosal in PBS at pH 7.4) added to each well and incubated for a further 2 h. Plates were washed six times with 350 μl well−1 of washing solution (2 ml l−1 Tween 20 in PBS at pH 7.4) and serum added at a 1:100 dilution in serum diluting buffer (blocking buffer with 2 ml l−1 Tween 20). Plates were incubated for 30 min and washed as before.

Fifty μl of a 1:50 000 dilution of HRP-Protein A (Zymed, VWR) was added to each well, incubated for 30 min and washed as before. One hundred μl well−1 of TMB Microwell Peroxidase Substrate (Kirkegaard and Perry Laboratories) was added and incubated for 15 min in the dark. The reaction was stopped with 100 μl of 0.5 M H2SO4 and read immediately at 450 nm. The mean and standard deviations (S.D.) were calculated from the sera of uninfected control subjects (n=17) and the cut-off for positive results was calculated as greater than the mean plus 3 S.D., and for high level responses, as the mean plus 6 S.D.

3 Results

3.1 Sequence identification

Previously constructed clone banks [21] that generated fusions between partial M. tuberculosis genes and the E. coli phoA gene were used to select proteins that could be significant antigens [17]. The M. tuberculosis DNA encoding the amino-termini of the partial Mtb-PhoA fusion proteins of nine selected clones was sequenced in order to match the clones with ORFs described in the two M. tuberculosis genome projects.

The majority of the clones were predicted to encode proteins with lipoprotein precursors, including UgpB (pJTA1-152), GgtB (pJTA2-506), OppA (pJDTA3-47) and two proteins for which the products are as yet unknown, LpqD (pJTA2-822) and LprF (pJTA2-639). The ORFs which were predicted to encode secreted proteins were MTCY31.03c (pJTA2-825), and MTCY21D4.03c (pJTA2-916). The latter protein has since been identified as MTC28 [22]. Other ORFs were predicted to encode membrane proteins including FtsQ (pJTA3-206) and the hypothetical protein MTV023.03c (pJTA2-1084).

3.2 Expression of full-length recombinant M. tuberculosis proteins

To maximize the production of the selected M. tuberculosis proteins we chose to use cloning vectors with amino-terminal protein tags. The vector-encoded tags enhance translation of the M. tuberculosis genes, resulting in a higher level of recombinant product. To engineer fusion between the vector-encoded tags and the M. tuberculosis genes we used PCR to amplify precise fragments from the M. tuberculosis chromosome. The products from PCR reactions were ligated into expression vectors and the inserts in the resulting recombinants were checked for orientation and reading frame by sequencing both the 5′ and 3′ junctions with the vector. The expression vector that provided the greatest expression of each particular M. tuberculosis protein was chosen for further use (Table 1).

The putative ORFs of genes matching those encoding the Mtb-PhoA fusion proteins were determined from preliminary sequence provided primarily by the Sanger Centre M. tuberculosis genome project. Hence, there are some differences between what we cloned as full-length ORFs and the final assignments of ORFs by the Sanger Centre. Also, in some cases, we chose to leave out the lipoprotein signal sequences or the first few amino acids of a predicted protein for ease of cloning. The full-length predicted amino acid sequence was contained in GST-206, GST-152, GST-511, PET-916. Relative to the sequence predicted by the Sanger M. tuberculosis genome project, the following N-terminal amino acids were excluded: the first 48 amino acids of MBP-506, the first amino acid of MBP-825 and PET-47, the first four amino acids of PET-639, and the first 10 amino acids of PET-1084. GST-822 does not contain the 32 amino acids at the C-terminal of the predicted sequence. As well GST-822 has an additional 27 N-terminal amino acids from the 3′ end of the ORF adjacent to the 5′ end of MTV004.48. These errors were due to changes in the DNA sequence of the region surrounding MTV004.48 made at the Sanger Centre.

Table 1 provides the genome projects’[19] identity of each recombinant clone as well as plasmid names and cloning vectors.

3.3 Protein overexpression

The recombinant proteins MBP-506, MBP-825, GST-152, GST-822, PET-47, PET-639 and PET-1084 formed inclusion bodies, which were harvested from the pellet following centrifugation of the bacterial sonicate. The recombinant proteins PET-916 and GST-206 did not form inclusion bodies and were found primarily in the supernatant of the bacterial sonicate. Fig. 1 shows a Western blot of the recombinant antigens.

Figure 1

Western immunoblot of the overexpressed fusion proteins. Lanes A–E, antibody reaction to T7-Tag?. A, T7 gene 10 (PET); B, PET-639; C, PET-916; D, PET-1084; E, PET-47. Lanes F–H, antibody reaction to MBP. F, MBP; G, MBP-825; H, MBP-506. Lanes I–L, antibody reaction to GST. I, GST; J, GST-822; K, GST-152; L, GST-206. Molecular mass standards are indicated on the left.

3.4 Serological recognition of recombinant proteins

Using ELISA, we compared the humoral responses of healthy individuals to those of patients with TBH and to HIV-negative patients with active EP-TB. Serum antibodies to a lysate of M. tuberculosis and to the nine fusion proteins were measured from the negative control group of 17 healthy individuals. The antibody response to each antigen by sera from infected individuals was compared to a cut-off value for positivity which was determined as the mean OD450 plus 3 S.D. of negative control group (Fig. 2). In the negative control group (group A), only MBP-825 elicited a single response above the cut-off value and that was within 10% of the cut-off. Thus, the cut-off values were sufficient for determination of a positive response.

Figure 2

Levels of serum antibodies to antigens in M. tuberculosis lysate and to the fusion proteins in healthy blood donors (A), patients with pulmonary TB and infected with HIV, TBH (B), and patients with EP-TB and HIV-negative, EP-TB (C). Each point represents one serum sample tested by an ELISA. The horizontal dotted line denotes the cut-off value determined as mean OD450 plus 3 S.D. using the negative control sera (healthy blood donors).

For patients with EP-TB (group C), antibodies against GST-822 were found in 60% of individual sera and a third of these were high level responses. Specific antibody responses (% positive response, % high level response, respectively) to PET-639 (40%, 10%), MBP-825 (35%, 5%) and MBP-506 (45%, 15%) were also found in the EP-TB group. The other five antigens, as well as the M. tuberculosis lysate elicited responses in fewer sera from EP-TB patients (35% or less).

For patients with TBH (group B), antibodies against MBP-506 were found in 61% of individual sera and 42% of individuals elicited high level responses. GST-822 was recognized in 42% of sera and 26% of individuals responded at a high level. The other seven antigens, as well as the M. tuberculosis lysate, elicited responses in fewer sera from TBH patients (35% or less).

By following the reactivity of individual serum samples with all of the antigens we were able to discern what percent of sera reacted with at least one antigen. In the EP-TB group, 71% of sera contained antibodies against at least one antigen (82% including TB lysate) and in the TBH group, 66% of sera contained antibodies against at least one antigen (84% including TB lysate). Thus, specific antibody responses can be identified in the majority of the individual sera.

4 Discussion

The groups of patients with M. tuberculosis-induced disease in this study responded immunologically by producing serum antibodies to a variety of M. tuberculosis antigens. As has been seen for individuals with pulmonary TB [9], the sera responses in this study confirm that there was heterogeneity of antigen recognition and strength of response in both EP-TB and TBH groups. Encouragingly, the majority of sera from diseased individuals contained specific antibodies to the small set of M. tuberculosis antigens tested. This finding suggests that for EP-TB and TBH patients, previously considered refractory to serodiagnosis, the combination of only a few well-recognized antigens might greatly improve diagnostic success.

Other researchers independently identified the M. tuberculosis antigen contained in the fusion PET-916, as the proline rich, immunogenic protein, MTC28 [22]. In their report, the reactivity of MTC28 in sera from patients with pulmonary TB was assessed and found to be recognized by 27% of responders. This result concurs with our finding that 29% of sera from HIV-positive patients with pulmonary TB were reactive to PET-916.

This is the first reporting of specific and high level antibody responses to GgtB (MBP-506), LprF (PET-639), LpqD (GST-822) and MTC31.03c (MBP-825). MBP-506 and GST-822, the two highly reactive and most frequently recognized recombinant antigens identified in this study, are potentially valuable candidates for inclusion in a serodiagnostic test.


Thanks to Jeffery Guderian and Lisa Reynolds for assistance with protein purification and ELISA and Jonathan Moran for proof-reading the draft. This work was supported by the Glaxo-Wellcome Action TB Initiative, the Medical Research Council of Canada (PA12992), the Canadian Bacterial Diseases Network, the British Columbia Lung Association, and N.I.H. Grant AI75320.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
View Abstract