Single nucleotide polymorphism in Ag85 genes of Mycobacterium tuberculosis complex: analysis of 178 clinical isolates from China and 13 BCG strains.

Host immune pressure and associated immune evasion of pathogenic bacteria are key features of host-pathogen co-evolution. Human T-cell epitopes of Mycobacterium tuberculosis (M. tuberculosis) were evolutionarily hyperconserved and thus it was deduced that M. tuberculosis lacks antigenic variation and immune evasion. However, in our previous studies, proteins MPT64, PstS1, Rv0309 and Rv2945c all harbored higher numbers of amino acid substitutions in their T cell epitopes, which suggests their roles in ongoing immune evasion. Here, we used the same set of 180 clinical M. tuberculosis complex (MTBC) isolates from China, amplified the genes encoding Ag85 complex, and compared the sequences. The results showed that Ag85 were hyperconserved in T/B cell epitopes and the genes were more likely to be under purifying selection. The divergence of host immune selection on different proteins may result from different function of the proteins. In addition, A312G of Ag85A and T418C of Ag85B may represent special mutations in BCG strains, which may be used to differentiate M.bovis and BCG strains from MTB strains. Also, C714A in Ag85B seems to be a valuable phylogenetic marker for Beijing strains.

INTRODUCTION

Tuberculosis (TB) is one of the most important issues of public health worldwide. About one third of the world population has been infected with M. tuberculosis, over 8.7 million new cases and 1.4 million deaths each year 1. The current efforts to reduce the global problem have been focused on improving the diagnosis methods and effective vaccines. The biochemical, immunological, and molecular biological characteristics of M. tuberculosis have led to the identification of several antigens which may be useful in the development of improved diagnostic methods and/or vaccines 2.

In 2010, Inaki Comas et al reported that human T cell epitopes of M. tuberculosis were evolutionarily hyperconserved and thus deduced that M. tuberculosis was lack of antigenic variation and immune evasion 3. However, our previous studies showed that there were polymorphisms existing in two important antigens, MPT64 4 and PstS1 5 in clinical M. tuberculosis strains isolated from China. This may be the reason for changes in the antigens produced, which may in turn cause alteration of related functions, thereby allowing immune evasion. Some other proteins such as Rv2945c and Rv0309 also owned polymorphisms, which suggest their roles in diversifying selection to evade host immunity 6. The antigen 85 complex (Ag85) consists of three predominantly secreted proteins (Ag85A, Ag85B, and Ag85C), which plays a key role in the mycobacterial pathogenesis and also possesses enzymatic mycolyltransferase activity involved in cell wall synthesis 7. Disruption of the gene encoding Ag85A in M. tuberculosis produces a strain that fails to replicate in human or mouse macrophages indicating that Ag85A may play a key role in M.tuberculosis pathogenesis. Knockout of the gene encoding Ag85C results in 40% reduction of M.tuberculosis the cell wall mycoloylation. Ag85 complex contribute to adherence, invasion, and dissemination of mycobacteria in host cells 8. By virtue of their strong potential to induce Th1-type immune responses, important for the control of intracellular infections, Mycobacterium Ag85 complex rank among the most promising TB vaccine candidate antigens. 9-14. Recently, Modified-Vaccinia-Ankara (MVA)85A vaccine became the first TB vaccine since BCG itself to complete an efficacy trial 15.

Here, we used the same set of clinical M. tuberculosis complex (MTBC) isolates(including two BCG strains) from China in our previous study 4, amplified genes of the antigens Ag85 (Ag85A, Ag85B, and Ag85C) and compared the sequences to explore the genetic diversity of them and to evaluate the impact of immune recognition on sequence variation of these three genes. In addition, we analyzed changes in protein level which was induced by single nucleotide polymorphism in Ag85 genes.

MATERIALS AND METHODS

Strains and DNA preparation

The first set of strains consisted of 180 clinical isolates that were selected from 2346 MTBC strains isolated in China genotyped by spoligotyping previously 16. All major and rare genotyping strains in China were included (Table 1). Considering the predominance of the Beijing family strains in China, we chose about half of the Beijing family strains (92 strains) and half non-Beijing family strains (88 strains). We randomly selected the 92 Beijing family strains from 1738 Beijing strains among 2346 strains. The other 88 strains were selected from 608 non-Beijing family isolates. Further, we attempted to purposely include strains representing different spoligotypes that were isolated from different regions. Table 2 showed the numbers of strains used in this study that were obtained from different provinces in China. A second set of strains contained 11 BCG strains, each of which originated from different places around the world. The strain names were showed in Table 3.

Table 1

No. of the strains of each Spoligotype pattern

Spoligotyping	No. of strains
Beijing	92
T	13
U	28
MANU	11
Haarlem	5
EAI	1
LAM	2
H37Rv family	1
BCG	2
S	1
CAS	4
new	20

Table 2

No. of the strains of different provinces in China

Places	No. of isolates
Anhui Province	12
Shannxi Province	17
Beijing Municipality	11
Fujian Province	29
Gansu Province	12
Guangxi Zhuang Autonomous Region	29
Sichuan Province	1
Henan Province	12
Hunan Province	7
Xizang (Tibet) Autonomous Region,	11
Xinjiang Uygur Autonomous Region	13
Jilin Province	14
Zhejiang Province	12

Table 3

Strains of Mycobacterium bovis and Bacillus Calmette Guerin (BCG).

ID No.	Strain name
1	BCG Birkhaug
2	BCG China
3	BCG Danish
4	BCG Frappier
5	BCG Glaxo
6	BCG Moreau
7	BCG Phipps
8	BCG Prague
9	BCG Swedens
10	BCG Tice
11	BCG Russia
12	BCG Tokyo*
13	BCG Paster*
14	BCG Mexco*
15	M. bovis AF2122/97*

* Data were obtained from the NCBI genome website

These strains were cultured using the standard Löwenstein-Jensen medium method, the genomic DNA were prepared according to previously reported and then used directly in polymerase chain reactions (PCRs).

The following Ag85 genes of the four published M.bovis and BCG strains were obtained from the NCBI genome website: M. bovis AF2122/97 (NC_002945), BCG Pasteur 1173P2 (NC_008769), BCG Tokyo 172 (NC_012207) and BCG Mexico (NC_016804).

Primers

The nucleotide sequences of the primers (from the 5' to 3' end) used in this study were designed with DNAstar software according to H37Rv genome sequence and showed in Table 4.

Table 4

The primers used in this study for PCR amplification

Gene	Locus tag	Length(bp)	Primers
Ag85A	Rv3804c	1166	5'- CACCGCCGCTAGATGTTGTG-3'F
			5'- CGCCCGAAGTTGTGGTTGAC-3'R
Ag85B	Rv1886c	1234	5'- ACTCGGCTAACTGGCTGGT-3'F
			5'- CGGTAACCGATACGGAAATG -3'R
Ag85C	Rv0129c	1509	5'- TGGTCGGCAGTAAGCATAGG-3'F
			5'- ACTGGTTGGGAGCGGCC -3'R

Polymerase Chain Reaction

The PCR were performed in a total volume of 20μl. The PCR mix contained 10μl PCR buffer, 100nM each primer, 200μM each of the four dNTPs and 0.5U DNA Taq Polymerase (Takara). An initial denaturation of 5min at 94℃ was followed by 35 cycles of denaturation at 94℃ for 45s, annealing at 62℃ for 45s and extension at 72℃ for 1min, followed by a final extension at 72℃ for 10min.

Negative controls using ddH2O instead of DNA were included each time when the PCR was performed. The positive control was 500pg DNA from M. tuberculosis H37Rv. The presence and size of each PCR product were determined by electrophoresis on 2% agarose gel in Tris/boric acid/EDTA buffer followed by staining with ethidium bromide.

We performed all of the PCRs at least twice to validate the reproducibility. The variants were confirmed by sequencing of the new PCR products.

Sequence and data Analysis

The sequences of the PCR products were determined by ABI 3730xl DNA Analyzer.

The sequences were first aligned by ClustalW 17 software with the Ag85 genes sequence from M. tuberculosis H37Rv genome to determine the regions of the genes, and then these regions were split out by a personalized PERL script. The sequence compare and translation were carried out by Bioedit software. Values of dN and dS were calculated by MEGA5. In addition, SPSS 14.0 (SPSS, Inc.) was used to perform chi-square analysis, and differences were considered to be statistically significant when P<0.05.

RESULTS

Mutations in gene sequences of Ag85

All 180 strains presented relative PCR products of antigens Ag85A, Ag85B and Ag85C. Table 5 showed the mutations in the gene sequences of Ag85A, Ag85B and Ag85C. Ag85A harbored three nonsynonymous mutations and two synonymous mutations. Ag85B owned two nonsynonymous mutations and three synonymous mutations. There were one nonsynonymous mutation and one synonymous mutation in antigen Ag85C.

Table 5

Changes in antigen Ag85A, Ag85B and Ag85C among 180 clinical strains*

Genes	Isolates	Base change	AA change	Spoligotypes
Ag85A	FJ05009	T12C	No change	New
	FJ06038ShanX05098	C139TG141C	P47SNo change	HaarlemBeijing
	HuN06009	C734G	A245G	Beijing
	FJ07113	C935G	A312G	BCG
	JL06005
Ag85B	AH03031	C131T	P44L	Beijing
	FJ07113	T418C	F140L	BCG
	JL06005
	GS05127	C666T	No change	Beijing
	GS05129
	93 strains	C714A	No change	-#
	AH03037	C786G	No change	Beijing
Ag85C	JL06007	G420A	No change	Beijing
	FJ05009	G472A	G158S	New
	FJ06159			Beijing
	XZ06003			CAS
	XJ06018			CAS
	XJ06153			CAS
	XJ06188			CAS

*: Use the CDS of Ag85A, Ag85B and Ag85C of M. tuberculosis H37Rv strain as the reference sequence.

#: Details are showed in Supplementary Material: Table S1

A total of 16 BCG strains (14 different BCG strains described above and 2 BCG strains from the clinical sample in China in the first isolates set) and one M.bovis (M. bovis AF2122/97, NC_002945) were included in this study. All of the M.bovis and BCG strains had two unique SNPs located in C935G of Ag85A and T418C of Ag85B. In the gene sequence of Ag85B, position 714 presented higher polymorphisms, as 93 strains owned an sSNPs (C-A). Six strains, i.e. four CAS family strains, one Beijing strain and one New strain presented same nonsynonymous mutation (G472A) in gene sequence of antigen Ag85C.

Changes in T/B cell epitopes

Table 6 presented T cell epitopes and B cell epitopes in Ag85A, Ag85B and Ag85C according to the Immune Epitopes Database (IEDB) 18. In the IEDB database, the impact on immune recognition was tested by bioinformatic approaches, and not in vivo or in vitro. Ag85A and Ag85B owned both T cell epitopes and B cell epitopes, while Ag85C only had three T cell epitopes (See Figure 1). For Ag85A and Ag85B, some epitopes were both T cell epitopes and B cell epitopes. Among all of the strains in this study, 12 of 64 T cell epitopes accounting for 18.75% and one of seven B cell epitopes accounting for 14.29% in Ag85A showed AA changes resulting from nucleotide alterations. Ag85B owned 37 B cell epitopes and 54 T cell epitopes, while 23 of them are both. Three B cell epitopes and three T cell epitopes altered from AA changes in Ag85B. There was no change in T cell epitopes of Ag85C.

Table 6

Amino acid changes of human T/B cell epitopes in antigen Ag85A, Ag85B and Ag85C*,‡

T or B epitope	IEDB_ID	Epitope	Rv locus	Base change	AA change	Antigen
T	56994	SASMGRDIKVQFQG	Rv0129c	No	No	Ag85C
T	72965	WPTLIGLAM	Rv0129c	No	No	Ag85C
T	74768	YLLDGLRAQ	Rv0129c	No	No	Ag85C
B	503	AAVVLPGLVGLAGGAATAGA	Rv1886c	No	No	Ag85B
B	34776	LAGGAATAGAFSRPGLPVEY	Rv1886c	CCG-TGG	P-W	Ag85B
B	42790	MTDVSRKIRAWGRRLMIGTA	Rv1886c	No	No	Ag85B
B	43514	NDPTQQIPKLVANNTRLWVY	Rv1886c	CCC-CCA	No	Ag85B
B	48646	PNGTHSWEYWGAQ	Rv1886c	No	No	Ag85B
B	72515	WGRRLMIGTAAAVVLPGLVG	Rv1886c	No	No	Ag85B
B	103272	IGLSMAGSSAMILAA	Rv1886c	No	No	Ag85B
B	103457	PAEFLENFVRSSNLK	Rv1886c	No	No	Ag85B
B	103530	QSGGNNSPAVYLLDG	Rv1886c	No	No	Ag85B
B	103532	QSSFYSDWYSPACGK	Rv1886c	No	No	Ag85B
B	103578	SAAIGLSMAGSSAMI	Rv1886c	No	No	Ag85B
B	103668	TSELPQWLSANRAVK	Rv1886c	No	No	Ag85B
B	103729	WGPSSDPAWERNDPT	Rv1886c	CCC-CCT	No	Ag85B
B	103732	WLSANRAVKPTGSAA	Rv1886c	No	No	Ag85B
T	3094	AMGDAGGYK	Rv1886c	No	No	Ag85B
T	5623	AVYLLDGLR	Rv1886c	No	No	Ag85B
T	8685	DIKVQFQSG	Rv1886c	No	No	Ag85B
T	13215	ELPQWLSANR	Rv1886c	No	No	Ag85B
T	13473	ENFVRSSNL	Rv1886c	No	No	Ag85B
T	15116	EYWGAQLNAMKGDLQSSLGA	Rv1886c	No	No	Ag85B
T	16332	FIYAGSLSA	Rv1886c	No	No	Ag85B
T	16333	FIYAGSLSAL	Rv1886c	No	No	Ag85B
T	16924	FLTSELPQW	Rv1886c	TTC-CTC	F-L	Ag85B
T	18276	FVRSSNLKF	Rv1886c	No	No	Ag85B
T	21078	GLPVEYLQV	Rv1886c	No	No	Ag85B
T	21275	GMGPSLIGL	Rv1886c	No	No	Ag85B
T	21780	GPSLIGLAM	Rv1886c	No	No	Ag85B
T	26269	IGLSMAGSSAMILAAY	Rv1886c	No	No	Ag85B
T	27786	IPAEFLENF	Rv1886c	No	No	Ag85B
T	27901	IPKLVANNT	Rv1886c	CCC-CCA	No	Ag85B
T	29558	IYAGSLSAL	Rv1886c	No	No	Ag85B
T	32213	KLVANNTRL	Rv1886c	No	No	Ag85B
T	38049	LMIGTAAAV	Rv1886c	No	No	Ag85B
T	42342	MPVGGQSSF	Rv1886c	No	No	Ag85B
T	43950	NFVRSSNLKFQDAYNAAGGH	Rv1886c	No	No	Ag85B
T	49862	PVEYLQVPSPSMGRD	Rv1886c	No	No	Ag85B
T	52025	QQFIYAGSLSALLDPSQGM	Rv1886c	No	No	Ag85B
T	59627	SMAGSSAMI	Rv1886c	No	No	Ag85B
T	60262	SPSMGRDIKVQFQS	Rv1886c	No	No	Ag85B
T	67695	VANNTRLWVYCGNGT	Rv1886c	No	No	Ag85B
T	73306	WYYQSGLSI	Rv1886c	No	No	Ag85B
T	76455	YWGAQLNAMKGDLQSSLGAG	Rv1886c	No	No	Ag85B
T	92817	GLAGGAATA	Rv1886c	No	No	Ag85B
T	174019	VEYLQVPSPSMGRDI	Rv1886c	No	No	Ag85B
T	174021	VPSPSMGRDIKVQFQ	Rv1886c	No	No	Ag85B
T/B	223	AAIGLSMAGSSAMILAAYHP	Rv1886c	No	No	Ag85B
T/B	1545	AGGYKAADMWGPSSDPAWER	Rv1886c	CCC-CCT	No	Ag85B
T/B	2695	ALLDPSQGMGPSLIGLAMGD	Rv1886c	No	No	Ag85B
T/B	3400	ANRAVKPTGSAAIGLSMAGS	Rv1886c	No	No	Ag85B
T/B	6323	CGNGTPNELGGANIPAEFLE	Rv1886c	No	No	Ag85B
T/B	8688	DIKVQFQSGGNNSPAVYLLD	Rv1886c	No	No	Ag85B
T/B	10841	DWYSPACGKAGCQTYKWETF	Rv1886c	TTC-CTC	F-L	Ag85B
T/B	18700	GANIPAEFLENFVRSSNLKF	Rv1886c	GCC-GCG	No	Ag85B
T/B	18898	GCQTYKWETFLTSELPQWLS	Rv1886c	TTC-CTC	F-L	Ag85B
T/B	21096	GLRAQDDYNGWDINTPAFEW	Rv1886c	No	No	Ag85B
T/B	21797	GPSSDPAWERNDPTQQIPKL	Rv1886c	CCC-CCT	No	Ag85B
T/B	40165	LTSELPQWLSANRAVKPTGS	Rv1886c	No	No	Ag85B
T/B	43332	NAVFNFPPNGTHSWEYWGAQ	Rv1886c	No	No	Ag85B
T/B	45250	NNSPAVYLLDGLRAQDDYNG	Rv1886c	No	No	Ag85B
T/B	49421	PSLIGLAMGDAGGYKAADMW	Rv1886c	No	No	Ag85B
T/B	49872	PVGGQSSFYSDWYSPACGKA	Rv1886c	No	No	Ag85B
T/B	50442	QDAYNAAGGHNAVFNFPPNG	Rv1886c	No	No	Ag85B
T/B	52026	QQFIYAGSLSALLDPSQGMG	Rv1886c	No	No	Ag85B
T/B	56895	SAMILAAYHPQQFIYAGSLS	Rv1886c	No	No	Ag85B
T/B	64079	THSWEYWGAQLNAMKGDLQS	Rv1886c	No	No	Ag85B
T/B	67697	VANNTRLWVYCGNGTPNELG	Rv1886c	No	No	Ag85B
T/B	72314	WDINTPAFEWYYQSGLSIVM	Rv1886c	No	No	Ag85B
T/B	76584	YYQSGLSIVMPVGGQSSFYS	Rv1886c	No	No	Ag85B
B	1522	AGGGHNGVFDFPDSG	Rv3804c	No	No	Ag85A
B	3402	ANRHVKPTGSAVVGL	Rv3804c	No	No	Ag85A
B	49333	PSDLGGNNLPAKFLE	Rv3804c	No	No	Ag85A
B	51790	QPACRKAGCQTYKWE	Rv3804c	No	No	Ag85A
B	70010	VMPVGGQSSFYSDWY	Rv3804c	No	No	Ag85A
T	1546	AGGYKASDMWGPKEDPAWQR	Rv3804c	No	No	Ag85A
T	3403	ANRHVKPTGSAVVGLSMAAS	Rv3804c	No	No	Ag85A
T	3422	ANSPALYLLDGLRAQDDFSG	Rv3804c	No	No	Ag85A
T	6901	CQTYKWETF	Rv3804c	No	No	Ag85A
T	8686	DIKVQFQSGGANSPALYLLD	Rv3804c	No	No	Ag85A
T	10838	DWYQPACGKAGCQTYKWETF	Rv3804c	No	No	Ag85A
T	18896	GCQTYKWETFLTSELPGWLQ	Rv3804c	No	No	Ag85A
T	19646	GFVRTSNIKFQDAYNAGGGH	Rv3804c	No	No	Ag85A
T	20979	GLLDPSQAMGPTLIGLAMGD	Rv3804c	No	No	Ag85A
T	21093	GLRAQDDFSGWDINTPAFEW	Rv3804c	No	No	Ag85A
T	21439	GNGKPSDLGGNNLPAKFLEG	Rv3804c	No	No	Ag85A
T	21482	GNNLPAKFLEGFVRTSNIKF	Rv3804c	No	No	Ag85A
T	21670	GPKEDPAWQRNDPLLNVGKL	Rv3804c	No	No	Ag85A
T	21960	GQSSFYSDWY	Rv3804c	No	No	Ag85A
T	25363	IANNTRVWVYCGNGKPSDLG	Rv3804c	GCC-GGC	A-G	Ag85A
T	31902	KLIANNTRV	Rv3804c	GCC-GGC	A-G	Ag85A
T	34823	LAIYHPQQFVYAGAMSGLLD	Rv3804c	No	No	Ag85A
T	40162	LTSELPGWLQANRHVKPTGS	Rv3804c	No	No	Ag85A
T	41872	MKPDLQRALGATPNTGPAPQGA	Rv3804c	No	No	Ag85A
T	43504	NDPLLNVGKLIANNTRVWVY	Rv3804c	GCC-GGC	A-G	Ag85A
T	44100	NGVFDFPDSGTHSWEYWGAQ	Rv3804c	GCG-GGG	A-G	Ag85A
T	49699	PTLIGLAMGDAGGYKASDMW	Rv3804c	No	No	Ag85A
T	49870	PVGGQSSFYSDWYQPACGKA	Rv3804c	No	No	Ag85A
T	50444	QDAYNAGGGHNGVFDFPDSG	Rv3804c	No	No	Ag85A
T	52030	QQFVYAGAMSGLLDPSQAMG	Rv3804c	No	No	Ag85A
T	52431	QSSFYSDWY	Rv3804c	No	No	Ag85A
T	56884	SALTLAIYHPQQFVYAGAMS	Rv3804c	No	No	Ag85A
T	64081	THSWEYWGAQLNAMKPDLQR	Rv3804c	GCG-GGG	A-G	Ag85A
T	72312	WDINTPAFEWYDQSGLSVVM	Rv3804c	No	No	Ag85A
T	73578	YDQSGLSVVMPVGGQSSFYS	Rv3804c	No	No	Ag85A
T	103416	MQLVDRVRG	Rv3804c	GTT-GTC	No	Ag85A
T	103423	MSRRLVVGA	Rv3804c	No	No	Ag85A
T	173920	AMGPTLIGLAMGDAG	Rv3804c	No	No	Ag85A
T	173921	AMSGLLDPSQAMGPT	Rv3804c	No	No	Ag85A
T	173924	AYNAGGGHNGVFDFP	Rv3804c	No	No	Ag85A
T	173925	DSGTHSWEYWGAQLN	Rv3804c	GCG-GGG	A-G	Ag85A
T	173926	FEWYDQSGLSVVMPV	Rv3804c	No	No	Ag85A
T	173927	FLEGFVRTSNIKFQD	Rv3804c	No	No	Ag85A
T	173932	FSGWDINTPAFEWYD	Rv3804c	No	No	Ag85A
T	173933	FVYAGAMSGLLDPSQ	Rv3804c	No	No	Ag85A
T	173934	FYSDWYQPACGKAGC	Rv3804c	No	No	Ag85A
T	173935	GAQLNAMKPDLQRAL	Rv3804c	GCG-GGG	A-G	Ag85A
T	173936	GGQSSFYSDWYQPAC	Rv3804c	No	No	Ag85A
T	173937	GKAGCQTYKWETFLT	Rv3804c	No	No	Ag85A
T	173938	GKLIANNTRVWVYCG	Rv3804c	GCC-GGC	A-G	Ag85A
T	173941	KEDPAWQRNDPLLNV	Rv3804c	No	No	Ag85A
T	173943	KVQFQSGGANSPALY	Rv3804c	No	No	Ag85A
T	173944	LLDGLRAQDDFSGWD	Rv3804c	No	No	Ag85A
T	173948	MGRDIKVQFQSGGAN	Rv3804c	No	No	Ag85A
T	173950	NNTRVWVYCGNGKPS	Rv3804c	No	No	Ag85A
T	173953	PLLNVGKLIANNTRV	Rv3804c	GCC-GGC	A-G	Ag85A
T	174003	QTYKWETFLTSELPG	Rv3804c	No	No	Ag85A
T	174004	RAQDDFSGWDINTPA	Rv3804c	No	No	Ag85A
T	174007	SELPGWLQANRHVKP	Rv3804c	No	No	Ag85A
T	174010	SPALYLLDGLRAQDD	Rv3804c	No	No	Ag85A
T	174014	SWEYWGAQLNAMKPD	Rv3804c	GCG-GGG	A-G	Ag85A
T	174016	TGSAVVGLSMAASSA	Rv3804c	No	No	Ag85A
T	174020	VFDFPDSGTHSWEYW	Rv3804c	No	No	Ag85A
T	174022	WLQANRHVKPTGSAV	Rv3804c	No	No	Ag85A
T	174023	WQRNDPLLNVGKLIA	Rv3804c	GCC-GGC	A-G	Ag85A
T	174024	WVYCGNGKPSDLGGN	Rv3804c	No	No	Ag85A
T	174025	YHPQQFVYAGAMSGL	Rv3804c	No	No	Ag85A
T/B	17838	FSRPGLPVEYLQVPSPSMGR	Rv3804c	CCG-TCGCCG-CCC	P-SNo	Ag85A
T/B	39011	LQVPSPSMGRDIKVQFQSGG	Rv3804c	No	No	Ag85A

*The CDS of Ag85A, Ag85B and Ag85C of M. Tuberculosis H37Rv strain has been used as the reference sequence.

‡Bold and underlined AA indicates locations of amino acid changes.

Figure 1

Genetic diversity of antigens Ag85A, Ag85B and Ag85C among 180 strains. T/B cell epitope region are marked in the sequences.

dN/dS values of proteins, epitope region and non-epitope region

Table 7 showed the distribution of synonymous and nonsynonymous SNPs in Ag85A, Ag85B and Ag85C among all 180 strains. The dN/dS value of Ag85A and Ag85B were 0.23 and 0.021, both much lower than 1, suggesting that these two proteins are likely appeared to be under purifying selection (Form of natural selection that acts to eliminate selectively deleterious mutations). For Ag85A, T/B cell epitope regions harbored higher dN/dS values than non-epitope regions, which mean the formers had accumulated significantly more amino acid changes than the latters. For Ag85B, non-T-cell-epitope regions had higher dN/dS value than T cell epitope regions, while B cell epitope regions owned higher dN/dS than non-B-cell-epitope regions. dN/dS of Ag85C was 2.04, higher than 1. All changes in Ag85C were from non epitope regions.

Table 7

Distribution of synonymous and nonsynonymous SNPs in gene sequence of Ag85A, Ag85B and Ag85C among 180 strains*

Gene		Length(bp)	SNPs			dN	dS	dN/dS
			Nonsyn	Syn	All
Ag85A	T cell epitope region	939	3	2	5	0.000032	0.000190	0.17
	Non-T-cell-epitope region	75	0	0	0	0	0	0
	B cell epitope region	315	1	1	2	0.000047	0.000146	0.32
	Non-B-cell-epitope region	699	2	1	3	0.000022	0.000183	0.12
	All	1014	3	2	5	0.000031	0.000134	0.23
Ag85B	T cell epitope region	897	1	3	4	0.000034	0.002458	0.014
	Non-T-cell-epitope region	78	1	0	1	0.000219	0	NA
	B cell epitope region	930	2	3	5	0.000049	0.002355	0.021
	Non-B-cell-epitope region	45	0	0	0	0	0	NA
	All	975	2	3	5	0.000047	0.002230	0.021
Ag85C	T cell epitope region	96	0	0	0	0	0	0
	Non-T-cell-epitope region	924	1	1	2	0.000122	0.000061	2
	All	1020	1	1	2	0.000108	0.000053	2.04

* H37Rv was used as reference to base the change in allele for the SNPs

NA, not applicable

Changes in protein level

All of the M.bovis and BCG strains presented two unique mutations in A312G of Ag85A and F140L of Ag85B, which might represent special mutations in BCG strains. Six strains with mutation of G158S in Ag85C included four CAS strains (XZ06003, XJ06018, XJ06153 and XJ06188), one Beijing strain (FJ06159) and one new spoligotype strain (FJ05009). As the C714A in Ag85B showed high polymorphism, we counted the frequencies of the synonymous mutation (Table 7). Among the 92 isolates of Beijing genotype, 95 % (n = 87) of the isolates presented A; meanwhile, among the 88 non-Beijing isolates, only ten isolates were A in the position 714 of Ag85B.

DISCUSSION

In this study, we chose 180 clinical MTBC strains which were originated from a very large geographical area and have different spoligotyping patterns in China; hence the data provided by them could be representative of genetic diversity that might be present within China, at least to some extent.

Studies in human pathogenic viruses, bacteria and protozoa have revealed that genes encoding antigens tend to be highly variable as a consequence of diversifying selection to evade host immunity 19-22. Comas et al reported that human T cell epitopes of M. tuberculosis were evolutionarily hyperconserved and thus deduced that M. tuberculosis was lack of antigenic variation and immune evasion3. However, in our previous studies, some proteins, such as MPT64, PstS1, Rv0309 and Rv2945c, harbored higher numbers of amino acid substitutions in their T cell epitopes, which suggesting their role in ongoing immune evasion 4,5,6. In this study, we found that Ag85A, Ag85B and Ag85C on the contrary showed lower substitution of amino acid in T/B cell epitopes. The dN/dS value of Ag85A and Ag85B were 0.23 and 0.021, both lower than 1, suggesting that these two proteins were likely appeared to be under purifying selection. dN/dS of Ag85C was 2.04, higher than 1. Yet all changes in Ag85C were from non epitope regions. Our data indicated that Ag85 were hyperconserved in T/B cell epitopes and the genes were more likely to be under purifying selection, which is in line with Comas' study.

Mycobacterium Ag85 complex consists of Ag85A, Ag85B, and Ag85C, which play an important role in cell wall biosynthesis by catalyzing the synthesis of the cord factor (trehalose 6,6′-dimycolate, TDM) by mycolyltransferase activity and serve as fibronectin-binding proteins that interact with host macrophage to trigger host immune response23,24. M. tuberculosis secretes many proteins into the extracellular environment, which can be recognized by the host immune system and induce protective immunity and immune responses with diagnostic values. Our findings that Ag85 complex are highly conserved in T/B cell epitopes also indicate they are suitable for diagnose and vaccine for TB. Currently, there are several vaccine candidates undergoing clinical trials that represent vaccines with different immune profiles and modes of action. In a study, 2797 BCG-vaccinated infants were boosted with MVA expressing the MTB antigen 85A or a placebo control and thereafter followed for 3 years. However, the outcome of the trial was very disappointing with no detectable improvement of protection against TB 15. It was proposed that rather than boosting Th1 responses, we should focus on understanding protective immune responses that are lacking or insufficiently promoted by BCG that can intervene at critical stages of the TB life cycle 25.

Backus KM et al reported that the three Mycobacterium tuberculosis antigen 85 isoforms have unique substrates and activities determined by non-active site regions, which reflect the differences among the three antigens 26. Our study showed that each of these three antigens was conserved, suggesting similar activity and substrate selectivity among different strains.

Genes in M. tuberculosis was assigned functional categories according to Tuberculist (http://tuberculist.epfl.ch/). MPT64, PstS1, Rv2945c and Rv0309 all belong to virulence, detoxification, adaption proteins, while Ag85 are lipid metabolism proteins. Different functions between individual genes indicate they are under distinct selection pressures. The former category of proteins are the first proteins interact with host immune system after M. tuberculosis strains infected, which made them easier to be under host immune selection to induce immune evasion. However, Ag85 showed lower polymorphisms than the former proteins. The data showed that proteins Ag85 were under purifying selection. We assume that M. tuberculosis proteins in different categories are suffered divergence from host immune selection. More proteins should be included to clarify this point.

In position 312 of Ag85A, a nonpolar amino acid Ala was substituted by a polar one (Gly) in M.bovis and BCG strains, the hydrophilicity increased, which suggests that the protein is more likely to combine the membrane structures related to function. The mutation in F140L of Ag85B rarely affects antigen function since it was changed between two nonpolar amino acids, it could still be used as a good phylogenetic marker to differentiate M.bovis and BCG strains from M.tuberculosis strains. Six strains with mutation of G472A in Ag85C included four CAS strains (XZ06003, XJ06018, XJ06153 and XJ06188), one Beijing strain (FJ06159) and one new spoligotype strain (FJ05009). Therefore, the SNPs were acting as phylogenetic markers for the four CAS strains. For strains in different spoligotypes, i.e. non-closely related strains, we had one homoplastic SNP(convergent evolution), which usually is a strong indicator of selection. As the mutation were changed between two polar amino acids (G-A), it nearly did not affect the protein function. 95 % Beijing isolates presented A in position 714 of Ag85B, thus C714A in Ag85B seemed to be a valuable phylogenetic marker for Beijing strains.

In conclusion, proteins Ag85 are highly conserved in T/B cell epitopes and the genes are more likely to be under purifying selection. The divergence of host immune selection on different proteins may result from different function of the proteins. A312G of Ag85A and T418C of Ag85B may represent special mutations in BCG strains, which may be used to differentiate M.bovis and BCG strains from MTB strains.

Table S1.

Click here for additional data file.