IgG, IgM and IgA Antibodies Against the Novel, Polyprotein in Active Tuberculosis

 

	R E S E A R C H A R T I C L E	Open Access

2 IgG, IgM and IgA antibodies against the novel

3 polyprotein in active tuberculosis

4 Xiaoyan Feng^1†, Xiqin Yang^1†, Bingshui Xiu¹, Shuang Qie¹, Zhenhua Dai¹, Kun Chen¹, Ping Zhao², Li Zhang³,

56789 Russell A Nicholson⁴, Guohua Wang¹, Xiaoguo Song¹ and Heqiu Zhang^1*

10  Abstract

11 Background: The present study was aimed to evaluate whether IgG, IgM and IgA antibodies levels detected against a

12 novel Mycobacterium tuberculosis polyprotein 38 F-64 F (with 38 F being the abbreviation for 38kD-ESAT6-CFP10 and

13 64 F for Mtb8.4-MPT64-TB16.3-Mtb8) are suitable for diagnosing active tuberculosis, and for monitoring the efficacy of

14 chemotherapy on TB patients.

15 Methods: In this study, a total of 371 active TB patients without treatment were selected and categorized into

16 S+/C+ group (n = 143), S-/C+ group (n = 106) or S-/C- group (n = 122). A series of serum samples were collected from

17 82 active TB patients who had undergone anti-TB chemotherapy for 0–6 months at one month interval. Humoral

18 responses (IgG, IgM and IgA) were determined for the novel Mycobacterium tuberculosis polyprotein using indirect

19 ELISA methods in all of serum samples.

20 Results: For S+/C+, S-/C+ and S-/C- active tuberculosis patients before anti-TB chemotherapy, the sensitivities of tests

21 based on IgG were 65.7%, 46.2% and 52.5% respectively; the sensitivities based on IgM were 21.7%, 24.5% and 18.9%;

22 and the sensitivities based on IgA were 25.2%, 17.9% and 23.8%. By combination of three isotypes, for all active

23 tuberculosis patients, the test sensitivity increased to 70.4% with the specificity being 91.5%. After anti-TB chemotherapy,

24 there were no significant differences between groups with different courses of anti-TB chemotherapy.

25 Conclusions: The novel Mycobacterium tuberculosis polyprotein 38 F-64 F represents potential antigen suitable for

26 measuring IgG, IgM and IgA antibodies. However, the serodiagnostic test based on the 38 F-64 F polyprotein appears

27 unsuitable for monitoring the efficacy of chemotherapy.

28			Keyword: Tuberculosis, Serodiagnosis, Polyprotein
29	Background			readily used at all levels of the health system and in		40
30	Tuberculosis (TB) remains the leading single microbial			the community [2,3].		41
31	illness globally, with one-third of the world’s population			The identification of the bacillus by microscopic exam-		42
32	infected with Mycobacterium tuberculosis (M. tubercu-			ination of sputum smear or by culture, however, presents		43
33	losis, Mtb) complex. In 2009, there were over 9.4 million			certain limitations. Around 30-50% of TB patients are		44
34	new cases and 1.7 million deaths from M. tuberculosis			negative in the microscopy examination, and culture re-		45
35	[1]. Over 90% of the worldwide burden of tuberculosis			quires a long time for the growth of M. tuberculosis,		46
36	is in low-income and middle-income countries where			which maybe lead to delay diagnosis [4,5]. The tubercu-		47
37	the diagnosis of tuberculosis still relies heavily on spu-			lin skin test has long been used for the diagnosis of TB.		48
38	tum smear microscopy and chest radiology. There is a			While this test is the recommended diagnosis test for la-		49
39	great need for rapid point-of-care tests that can be			tent TB infection, it requires standardized application		50
				and interpretation, and a positive result depends on an		51
				adequate immune response [6]. The usefulness of the		52
	* Correspondence: zhangheqiu2004@126.com			IFN-γ-release assays in the diagnosis of active TB re-		53
	†Equal contributors			mains questionable [7]. Nucleic acid amplification tests,		54
	1Department of Bio-diagnosis, Beijing Institute of Basic Medical Sciences,
	Beijing 100850, China			for example the Xpert MTB/RIF assay, are the most		55

Full list of author information is available at the end of the article

 

56	promising development in tuberculosis diagnostics in	symptoms of TB, by chest radiography, and by symp-	108
57	the USA and Europe [8]. However, uses of such tests are	tomatic  improvement  after  chemotherapy.  Sera  were	109
58	restrictive because this assay requires dedicated and ex-	stored at −70°C until start of testing.	110
59	pensive equipment. The serological test based on the de-	A total of 371 active TB patients at their first visit to	111
60	tection of circulating antibodies against M. tuberculosis-	the outpatient clinic without treatment were selected	112
61	specific antigens could represent a useful complement to	and categorized into three groups, i.e.	113
62	microscopic examination for screening active tubercu-	S+/C+:  Smear-positive  and  culture-positive  group	114
63	losis [9]. This method is quite attractive because of its	(n = 143), with 82 males and 61 females and the age range	115
64	easy application, low cost of testing many serum samples	from 19 to 67, with a median of 42.	116
65	in parallel and relatively low invasiveness.	S-/C+:  Smear-negative  and  culture-positive  group	117
66	A comprehensive insight to immunoprofiling of anti-	(n = 106), with 55 males and 51 females and the age range	118
67	gen specific responses is critical for TB diagnosis and	from 22 to 70, with a median of 46.	119
68	therapeutic monitoring. Currently, gold standard methods	S-/C-:  Smear-negative  and  culture-negative  group	120
69	to diagnosis TB and monitor treatment response in-	(n = 122), with 65 males and 57 females and the age range	121
70	clude sputum smear microscopy and culture conversion	from 19 to 78, with a median of 40.	122
71	after 2 months of TB treatment. But, for patients in	Sputum samples were processed using standard NALC-	123
72	whom such sputum sample is not available, alternative	NaOH method and smears were examined after Ziehl–	124
73	serological tests are needed. Some results showed that	Neelsen staining. Processed samples were inoculated	125
74	combined use of different antibody isotypes allow an	in MGIT (Mycobacterial growth indicator tube) 960 non-	126
75	increased accuracy in diagnostic of tuberculosis [10],	radiometric automated isolation system (BD, USA) in	127
76	and the levels of antibody against to some antigens de-	accordance with the standard procedure.	128
77	creased together with treatment [11]. On the contrary,	To assess whether determination of 38 F-64 F specific	129
78	some results showed that combination of IgG with IgA	antibody responses could be useful for monitoring the	130
79	and/or IgM does not improve its sensitivity, and the	efficacy of chemotherapy, a series of serum samples were	131
80	levels of antibody against to other antigens were not as-	obtained from 82 active TB patients, which including 30	132
81	sociated with anti-TB treatment [12].	IgG positive patients, 28 IgA positive patients and 24	133

 

82 In our previous study, two novel M. tuberculosis poly-   IgM positive patients at their first visit to the outpatient  134

83 proteins, 38 F (38kD-ESAT6-CFP10) and 64 F (Mtb8.4-   clinic. The diagnosis of active TB in these patients was  135

84	MPT64-TB16.3-Mtb8), were expressed as antigens with
85	multiepitopes, and evaluated for serodiagnosis of TB.
86	The novel 38 F-64 F indirect ELISA assay based on
87	measuring IgG antibody has potential to achieve higher
88	sensitivity and specificity, and the ROC analysis indi-

89 cated that the novel 38 F-64 F indirect ELISA assay had

90 a better overall diagnostic performance [13].

91 The goals of the present study were to evaluate serum

92 levels of all of three isotype antibodies, IgG, IgM and

93 IgA, specific to 38 F-64 F, in patients with active TB and

94 in BCG-vaccinated healthy individuals, and to assess

95 whether determination of 38 F-64 F specific antibody re-

96 sponses could be useful for monitoring the efficacy of

97 chemotherapy.

98 Methods

99 Ethics statement

100 The study was approved by the ethics committee of the

101 Beijing Chaoyang District Centre for Disease Control

102 and Prevention, and Tianjin Haihe Hospital. Written in-

103 formed consent was obtained from all participants.

104 Study population

105 Sera were obtained from patients with active TB and

106 from normal individuals, as described below. Diagnosis

107 of active TB was established by the presence of clinical

based on sputum smear positive and/or sputum culture	136
positive and/or X-ray clinical findings. All patients had	137
received combination antituberculosis chemotherapy with	138
isoniazid and rifampin and pyrazinamide and ethambutol/	139
streptomycin for 2 months followed by admission isoniazid	140
and rifampin for 4 months. All patients were HIV sero-	141
negative and there were no other disease states accom-	142
panying the TB. A series of serum samples were collected	143
at one month interval, and divided into seven groups, i.e.,	144
0 M, 1 M, 2 M, 3 M, 4 M, 5 M and 6 M.	145
Ninety-four BCG-vaccinated healthy blood donors were	146
included. Their age ranged from 20 to 55, with a median of	147
38. The healthy subjects had normal findings on chest	148
radiogram and no history of close contract with TB pa-	149
tients and no family history of tuberculosis. These subjects	150
were apparently normal without HIV infection and other	151
diseases that might be confused with TB such as pneumo-	152
nia, fungal infections, lung cancer, etc.	153
ELISA	154
Microplates  were  coated  with  individual  antigens  at	155
5 μg/ml (3 μg/ml 38 F and 2 μg/ml 64 F) in coating buf-	156
fer (0.05 M carbonate/bicarbonate, pH 9.6) and stored at	157
4°C overnight. The plates were washed three times with	158
phosphate-buffered saline (PBS) containing 0.05% Tween	159
20 (PBST). Two hundred microliters of PBST containing	160

161 1% bovine serum albumin was added to each well, and the

162 plates were sealed and incubated at 37°C for 1 h. The plates

163 were washed three times. One hundred microliters of

164 serum diluted 1:10 in PBST containing 1% BSA was added

165 to each antigen-coated well. The plates were sealed and in-

166 cubated at 37°C for 30 min and then washed three times.

167 One hundred microliters of horseradish peroxidase-

168 conjugated anti-human IgG antibody (Sigma, USA), anti-

169 human IgM antibody (Bethyl, USA) or anti-human IgA

170 antibody (Bethyl, USA) was added to each well respect-

171 ively, and the plates were sealed and incubated at 37°C

172 for another 30 min. Again, the plates were washed three

173 times. The bound enzyme was detected by freshly-

174 prepared tetramethyl benzidine (TMB) substrate. After

175 20 min incubation in room temperature, the stop solution

176 (0.1 N sulfuric acid) was added and the optical density was

177 determined at 450 nm using an automatic microplate

178 reader (Bio-Rad, USA).

179 Statistical analysis

180 Data processing was performed using GraphPad Prism 4.0

181 (GraphPad Software Inc., San Diego, CA) and SPSS16.0

182 software package (SPSS Inc., Chicago, IL). The diagnostic

183 value of the novel 38 F-64 F indirect ELISA assay was eval-

184 uated by the receiver operating characteristic (ROC) curve

185 performed with the data from patients with active TB and

186 94 BCG-vaccinated healthy blood donors. The cutoff value

187 of 38 F-64 F IgG, IgM and IgA were chosen according

188 to the ROC analysis. Comparisons between treatment

189 groups were done by the Mann Whitney test.

190 Results

191 IgG, IgM and IgA antibody responses against 38 F-64 F

192 antigens in active TB without treatment

193 In our previous study, the novel 38 F-64 F antigens were

194 confirmed to have potential for higher sensitivity and

195 specificity based on measuring IgG antibody [13]. Con-

196 sidering the heterogeneity of the humoral response, in

197 this study, serum levels of all of three isotype antibodies,

198 IgG, IgM and IgA, specific to 38 F-64 F were evaluated

199 in active TB patients at their first visit to the outpatient T1 200 clinic without treatment. The data are shown in Table 1 F1 201 and presented as scattergrams in Figure 1.

Before anti-TB chemotherapy, the sensitivities of the	202
tests based on IgG were 65.7% (95% CI, 57.3% to 73.5%),	203
46.2% (95% CI, 36.5% to 56.2%) and 52.5% (95% CI,	204
43.2% to 61.6%) in S+/C+, S-/C + and S-/C- active tuber-	205
culosis patients respectively; the sensitivities of the tests	206
based on IgM were 21.7% (95% CI, 15.2% to 29.3%),	207
24.5% (95% CI, 16.7% to 33.8%) and 18.9% (95% CI,	208
12.3% to 26.9%) respectively; and the sensitivities of the	209
tests based on IgA were 25.2% (95% CI, 18.9% to 33.9%),	210
17.9% (95% CI, 11.2% to 26.6%) and 23.8% (95% CI,	211
16.5% to 32.3%) respectively. In S+/C+ group, 53, 4 and	212
8 sera were IgG-, IgM- and IgA-positive only, respect-	213
ively, and the number of combination positive sera was	214
108, which including anyone, any two and all of three	215
isotype of antibodies positive sera. By combination of all	216
the three isotype an increased sensitivity from 65.7% to	217
75.5% was obtained. In S-/C+ group and S-/C- group,	218
the sensitivity increased from 46.2% to 64.2% and from	219
52.5% to 69.7% by combination of all the three isotype	220
respectively.	221
To evaluate the specificity, all the three isotype anti-	222
bodies responses were examined in 94 BCG-vaccinated	223
healthy blood donors. The specificity of the tests based	224
on IgG, IgM and IgA were 94.7% (95% CI, 88.0% to	225
98.3%), 92.6% (95% CI, 85.3% to 97.9%) and 95.7% (95% CI,	226
89.5% to 98.8%) respectively (Figure 1). By combination of	227
three isotypes, the specificity of the tests decreased slightly	228
to 91.5% (95% CI, 87.0% to 94.8%). These results are con-	229
sistent with the results of Uma Devi et al. [14]. In that art-	230
icle, when IgG was taken individually, the specificity was	231
100%, when IgG + IgA were taken, the specificity reduced	232
to 96%, and when IgG + IgA + IgM were taken, the specifi-	233
city reduced to 90%.	234
The serodiagnosis performance of the novel 38 F-64 F	235
polyprotein on measuring IgG, IgM and IgA antibody	236
was analyzed with all TB patients and healthy controls	237
(Figure 1). The area under the ROC curve (AUC) of the	238
novel 38 F-64 F indirect ELISA assay on measuring IgG,	239
IgM and IgA antibody were 0.81 (95% CI, 0.75 to 0.85),	240
0.68 (95% CI, 0.61 to 0.74) and 0.62(95% CI, 0.56 to 0.69)	241
respectively. The ROC analysis indicated that the novel	242
38 F-64 F polyprotein had a better diagnostic performance	243
on measuring IgG than IgM and IgA.	244

t1:1 Table 1 ELISA results of the 38 F-64 F antigens in the serodiagnosis of active pulmonary TB

 

t1:2	Group		S+/C + (n = 143)				S-/C + (n = 106)				S-/C- (n = 122)
t1:3
		IgG	IgM	IgA		IgG	IgM	IgA		IgG	IgM	IgA
t1:4
	# Pos. (%)	94 (65.7)	31 (21.7)	36 (25.2)		49 (46.2)	26 (24.5)	19 (17.9)	64 (52.5)		23 (18.9)	29 (23.8)
t1:5	# Pos. for one isotype of antibody	53	4	8		29	11	5	39		9	11
t1:6	# Combin. Pos. (%)		108 (75.5)				68 (64.2)				85 (69.7)
t1:7
	S+/C+: smear-positive and culture-positive active pulmonary TB patients; S-/C+: smear-negative and culture-positive active pulmonary TB patients; S-/C-: smear-negative
t1:8	and culture-negative active pulmonary TB patients; # Pos. (%): number of positive sera (percentage of positive sera); # Combin. Pos. (%): number of combination positive
t1:9	sera for any of three isotype antibodies (percentage of combination positive sera).

 

Figure 1 The serodiagnostic performance of the novel 38 F-64 F indirect ELISA assay in active TB patients at their first visit to the outpatient clinic without treatment. Levels of serum IgG, IgM and IgA against the novel 38 F-64 F polyprotein in S+/C+, S-/C + and S-/C- TB patients (A1, B1 and C1) and in total TB patient (A2, B2 and C2) were presented as scatter grams respectively. Each bar represents the mean

OD value. The ROC curves of the novel 38 F-64 F indirect ELISA assay based on measuring IgG (A3), IgM (B3) and IgA (C3) were shown and the areas under the curve (AUC) for IgG, IgM and IgA were 0.81, 0.68 and 0.62 respectively.

245 Integration of serodiagnostic test and bacteriology

246 examination to screen for active TB

247 By combination of all of three isotypes (IgG or IgM

248 or IgA), a sensitivity of 70.4% (261/371) was obtained

249	with  a  specificity  of  91.5%  (86/94)  in  all  of  three
250	groups of TB patients without treatment. Moreover,

251 the positive rates of antibodies specific to 38 F-64 F

252 for smear-positive, smear-negative, culture-positive or

253 culture-negative patients were not significantly different.

254 In smear-negative or culture-negative patients, the posi-

255 tive rates of antibodies specific to 38 F-64 F were 67.1%

256 and 69.7%, slightly lower than the positive rates for

257 smear-positive (75.5%) or culture-positive (70.68%) patients T2 258 

258 (Table 2).

t2:1 Table 2 Serodiagnosis results of the 38 F-64 F antigens in t2:2 the active pulmonary TB patients with different sputum t2:3 bacteriology results

 

t2:4	Group	S + (n = 143)	S- (n = 228)	C + (n = 249)	C- (n = 122)
t2:5
	# Pos. (%)	108 (75.5)	153 (67.1)	176 (70.7)	85 (69.7)

 

t2:6 S+: smear-positive active pulmonary TB patients; S-: smear-negative active t2:7 pulmonary TB patients; C+: culture-positive active pulmonary TB patients; t2:8 C-: culture-negative active pulmonary TB patients; # Pos. (%): number of positive t2:9 sera (percentage of positive sera).

 

272 efficacy of chemotherapy, a series of serum samples were

273 obtained from active TB patients who had undergone

274 anti-TB chemotherapy for 0–6 months at one month

275 interval and the serum levels of all of three isotype anti-

276 bodies, i.e., IgG, IgM and IgA, specific to 38 F-64 F, were

277 determined. The results are presented as scatter grams F3 278 in Figure 3.

279 In this study, the mean levels of IgG, IgM and IgA

280 antibodies against 38 F-64 F polyprotein in sera of active

281 TB patients with different courses of anti-TB chemo-

282 therapy were shown to be no significant difference when

283 compared with results from the same patients at their

284 first visit to the outpatient clinic without anti-TB chemo-

285 therapy (P > 0.05). The mean antibody levels and the

286 standard deviation of each groups are shown in Figure 3.

287 After anti-TB chemotherapy was initiated, the IgG, IgM

288 and IgA antibody responses were heterogeneous and

289 differed considerably from patient to patient, although

290 the  response  patterns  can  be  categorized  into  three

291 types. The antibody levels against 38 F-64 F in some

292 patients increased with the course of anti-TB chemo-

293 therapy. In contrast, in some patients, the antibody levels

294 against 38 F-64 F decreased with the course of anti-TB

 

Figure 2 The positive rates of sputum smear/culture with or without integration of serodiagnostic test. The integration of serodiagnostic test with sputum smear or sputum culture increased the positive rates from 38.54% to 79.78% or from 67.12% to 90.03%, respectively.

chemotherapy. The third type of response pattern indi-	295
cated that the antibody levels against 38 F-64 F were more	296
stable with slightly fluctuations. Some corresponding ex-	297
amples are shown in Figure 4.	298	F4

Discussion	299
An effective in vitro diagnostic for TB based on serological	300
methods could be an attractive area of exploration because	301
immunoassays have advantages of simplicity, rapidity and	302
low cost, and have possibility to find the cases missed by	303
standard sputum smear microscopy [2,3,15]. Nevertheless,	304
considerable progress has been made in the identification	305
of many new serological antigens in recently years, and	306
the sensitivity and specificity of serological tests are im-	307
proving. It is therefore generally accepted that it will be	308
advantageous to include several antigens in a future sero-	309
diagnostic assay. In our previous study, the novel polypro-	310
tein 38 F-64 F including seven Mtb antigens had been	311
found to achieve higher sensitivity and specificity based	312
on measuring IgG antibody [13]. The knowledge of the	313
humoral immune responses at various stages of TB in-	314
fection and disease could help to elucidate the complex	315
interaction between host and pathogen [16-18]. A com-	316
prehensive insight to immunoprofiling of antigen spe-	317
cific responses is critical not only for the understanding	318
of disease pathogenesis, but also for development of	319
diagnostic tests. In view of the heterogeneity of the	320
humoral response, in this study, serum levels of three	321
38 F-64 F-specific antibody isotypes, IgG, IgM and IgA,	322
were evaluated in active TB patients at their first visit to	323
the outpatient clinic without treatment and those who	324
had undergone anti-TB treatment.	325
IgG antibody level is higher in most advanced and ex-	326
tensive forms of the disease. Patients with active TB usu-	327
ally exhibited strong IgG responses but poor IgM and	328
IgA responses [19]. While the function of anti-M. tuber-	329
culosis antibodies in providing protective immunity is	330
still under investigation, it has been proposed that they	331
may be utilized as a diagnostic marker of active disease	332
[18,19]. Previous studies analyzing IgG antibodies showed	333
that anti-M. tuberculosis IgG antibodies increased in pa-	334
tients with active disease [20-22]. In our results, IgG anti-	335
body level is higher in the most active TB patients with or	336
without the anti-TB chemotherapy than that in health con-	337
trol, and the positive rate of IgG was highest among the	338
three isotypes, indicating that the IgG antibody was the	339
most extensive antibody isotype.	340
Usually, IgM antibodies appear first and are produced	341
in large quantities in response to any antigen and decline	342
in more advanced phases. Several authors suggested that	343
IgM antibodies are produced mainly during the early	344
phase of primary TB infection [23,24]. Therefore, the	345
IgM-positive patients were diagnosed at an early stage of	346
the infection process [19,25]. Our results demonstrate	347

 

Figure 3 Levels of serum IgG (A), IgM (B) and IgA (C) against the novel 38 F-64 F polyprotein in active TB patients with different course of chemotherapy. 0 M = patients at their first visit to the outpatient clinic, 1 M-6 M = patients received anti-TB chemotherapy for 1–6 months. The mean antibody level and the standard deviation of each group were shown. There was no significant difference between the patients at their first visit to the outpatient clinic without anti-TB chemotherapy and those receiving anti-TB chemotherapy for 1–6 months. All of the p values between each of two groups were greater than 0.05

Figure 4 Changes in IgG (A1, A2 and A3), IgM (B1, B2 and B3) and IgA (C1, C2 and C3) antibody levels against the novel 38 F-64 F polyprotein after initiation of anti-TB chemotherapy. After initiation of anti-TB chemotherapy, patients showed three different IgG, IgM and IgA antibody responses patterns, i.e., decrease, increase and maintaining at stable level.

348 that the positive rate of IgM is lowest among the three

349 isotypes. In China, most people visit a doctor only after

350 symptoms appear. Accordingly, at their first visit to the

351 outpatient clinic most TB patients are over the early

352 stage of the infection process. Interestingly, following

353 different stages of the infection process, three types of

354 response patterns of IgM antibody were observed. The

355 antibody levels of IgM against 38 F-64 F in some pa-

356 tients became lower with the course of treatment, but

357 IgM antibody levels in some patients were elevated or

358 were maintained at more stable levels throughout the

359 course of anti-TB chemotherapy. Several authors have

360 observed that IgM antibody production was not associ-

361 ated with any clinical phases and radiological factor [26].

362 IgA is secreted during the contact of M. tuberculosis

363 and/or its antigens with the mucosal surface, which

364 stimulates the release of cytokines [27]. Several authors

365 have described the presence of IgA against the mycobac-

366 terial antigens in the serum of TB patients [26,28-31],

367 suggesting that this immunoglobulin isotype is present

368 in both  patients with newly acquired  infections  and

369 those of a longer duration [32]. In the present investiga-

370 tion, the positive rate of IgA was not too high in active

371 TB patients at their first visit to the outpatient clinic and

372 the IgA antibody levels were maintained for a longer

373 duration until the completion of anti-TB chemotherapy.

374 Secondly, from our results, we found that some sera

375 are IgM- or IgA-positive only. Though most of the active

376 TB patients (70.4%) could be diagnosed by the detection

377 of IgG antibody isotype, few TB patients showing IgM-

378 or IgA-positive only [about 13% (48/371)], were left out,

379 which would lead to delay in treatment and more exten-

380 sive transmission of TB. Thus, it should detect different

381 types of antibodies simultaneously. And in our results,

382 the test sensitivity was improved for IgG + IgA + IgM

383 without significantly compromising the specificity. These

384 results are consistent with the results of Uma Devi et al.

385 [14]. On the other hand, if one were to test all of three

386 antibody isotypes, the expense of diagnosis for TB pa-

387 tient would increase. Therefore, in order to increase the

388 positive rate and decrease the expense of diagnosis, it

389 would be necessary for researchers to develop the total

390 antibody detection method, i.e., double antigen sandwich

391 ELISA.

392 The other significant result obtained in this study was

393 that the positive antibody responses in smear-negative

394 and/or culture-negative active TB patients were not dif-

395 ferent significantly with those in smear- and culture-

396 negative active TB patients. This is in uncoincidence with

397 some other studies results showing that smear-positive

398 TB patients generate more proteinaceous antigens than

399 smear-negative TB patients [33,34]. On the other hand,

400 there are a few studies showing that the antibody level

401 does not always correspond to the bacterial load in TB

patients [35,36]. Thus, for serodiagnostic methods, there	402
is greater value for combination with, not replacement	403
of, the sputum smear and sputum culture test. The posi-	404
tive rate would increase in paucibacillary TB patients,	405
which are difficult to diagnose by bacteriological tests	406
[12,26]. The findings of Kanaujia GV et al. [9] also dem-	407
onstrated that the effectiveness of the screening strategy	408
was improved by integrating multi-antigen ELISA with	409
microscopic examination of sputum for acid-fast bacilli	410
in areas with high TB prevalence.	411
In order to assess whether determination of 38 F-64 F	412
specific antibody responses could be useful for monitor-	413
ing the efficacy of chemotherapy, the levels of three iso-	414
type of antibodies were measured in active TB patients	415
with different course of anti-TB chemotherapy. On one	416
hand, the heterogeneity of the antibody response be-	417
tween patients was confirmed by several researchers in	418
previous studies [37-39]. We also observed the antibody	419
response to be heterogeneous in patients with active TB	420
undergoing anti-TB chemotherapies. On the other hand,	421
many investigations concerning the changes in anti-	422
body levels showed that antibody levels decrease after	423
initiation of anti-TB chemotherapy [40,41]. Our results	424
did not support such a conclusion. Based on our statis-	425
tical results, there was no significant difference between	426
groups with different course of anti-TB chemotherapy	427
for all of subjects, and the response patterns could be	428
grouped into three types, i.e. increase, decrease and main-	429
taining at a stable level. Moreover, the antibody against	430
Mtb antigen continued to be present throughout the anti-	431
TB chemotherapy.	432
A wide spectrum of humoral responses exist in TB pa-	433
tients, depending on different antigens, the expression	434
levels of different antigens, the patient’s immunological	435
background, the disease stage, and/or the type of anti-	436
TB therapy [42,43]. Although higher sensitivity and spe-	437
cificity would be obtained based on measuring antibody	438
against polyprotein, such as 38 F-64 F including seven	439
proteins from Mtb, the regular change in the level of	440

antibody against only one antigen would be covered up.  441

Therefore, the polyprotein may not be suitable for moni-	442
toring the efficacy of chemotherapy.	443

Conclusions	444
Based on our present experiments we can make recom-	445
mendations that may be helpful in subsequent studies	446
on the serological diagnosis of TB. First, all of three	447
38 F-64 F-specific antibody isotypes, IgG, IgM and IgA,	448
were detected in the sera of TB patients, indicating that	449
the novel 38 F-64 F polyprotein was suitable for diagnos-	450
ing active TB and it is necessary to develop the double	451
antigen sandwich ELISA to detect total antibodies. Sec-	452
ond, the positive rate would increase in the pauciba-	453
cillary TB patients by serodiagnostic methods. Thus, an	454

455 improved screening strategy can be created by integrating

456 multi-antigen ELISA with bacteriology tests in areas with

457 high TB prevalence. Third, the serodiagnostic test based

458 on polyprotein antigens does not appear to be useful for

459 monitoring the efficacy of chemotherapeutics or as a cri-

460 terion to determine whether the patient should continue

461 or terminate anti-TB chemotherapy. Therefore, it will be

462 necessary to evaluate novel predictors of the efficacy of

463 anti-TB drugs based on other immunological markers,

464 such as the level of antigen, cytokine, microRNA, alone

465 or in combination.

466 Abbreviations

467 TB: Tuberculosis; ROC: Receiver operating characteristic; AUC: Area under the

468 ROC curve..

469 Competing interests

470 The authors declared that they have no competing interests.

471 Authors’  contributions

472 The conception and design of the study: XYF, XQY, HQZ; Acquisition of data:

473 XQY, BSX, SQ, ZHD, KC, XGS, GHW; Collection of samples: PZ, L Zhang;

474 Analysis and interpretation of data: XYF, XQY, HQ Zhang; Drafting the article

475 or revising it critically for important intellectual content: XYF, RAN; Final

476 approval of the version to be submitted: XYF, HQZ.

477 Acknowledgment

478 This work was supported by the grants from the National Natural Science

479 Foundation of China [30772067], National S & T Major Project for Infectious

480 Diseases Control [2009ZX10004-718] and the National High Technology

481 Research and Development Program of China [2011AA02A120].

482 Author details

483 ¹Department of Bio-diagnosis, Beijing Institute of Basic Medical Sciences,

484 Beijing 100850, China. ²Chaoyang District Centre for Disease Control and

485 Prevention, Beijing 100029, China. ³Tianjin Haihe Hospital, Tianjin 300350,

486 China. ⁴Department of Biological Sciences, Simon Fraser University, Burnaby,

487 B.C V5A 1S6, Canada.

488 Received: 2 December 2013 Accepted: 12 June 2014

489 Published: 17 June 2014

490 References

491 1.   WHO: Global tuberculosis control. Geneva: Switzerland; 2010.

492 2.   Ai-Zamel FA: Detection and diagnosis of Mycobacterium tuberculosis.

493 Expert Rev Anti Infect Ther 2009, 7:1099–1108.

494 3.   Wallis RS, Pai M, Menzies D, Doherty TM, Walzl G, Perkins MD, Zumla A:

495 Biomarkers and diagnostics for tuberculosis: progress, needs, and

496 translation into practice. Lancet 2010, 375:1920–1937.

497 4.   American Thoracic Society: Center for Disease Control and Prevention:

498 Diagnostic standards and classification of tuberculosis in adults and

499 children. Am J Respir Crit Care Med 2000, 161:1376–1395.

500 5.   Teixeira HC, Abramo C, Munk ME: Immunological diagnosis of

501 tuberculosis: problems and strategies for success. J Bras Pneumol 2007,

502 33:323–334.

503 6.   Huebner RE, Schein MF, Bass JB Jr: The tuberculin skin test. Clin Infect Dis

504 1993, 17:968–975.

505 7.   Chegou NN, Hoek KGP, Kriel M, Warren RM, Victor TC, Walzl G: Tuberculosis

506 assays: past, present and future. Expert Rev Anti Infect Ther 2011, 9:457–469.

507 8.   Lawn SD, Zumla AI: Tuberculosis. Lancet 2011, 378:57–72.

508 9.   Kanaujia GV, Lam PK, Perry S, Brusasca PN, Catanzaro A, Gennaro ML:

509 Integration of microscopy and serodiagnostic tests to screen for active

510 tuberculosis. Int J Tuberc Lung Dis 2005, 9:1120–1126.

511 10.  Kochak HE, SeyedAlinaghi S, Zarghom O, Hekmat S, Jam S, Sabzvari D,

512 Abdi Z: Evaluation of serological tests using A60 antigen for diagnosis

513 of tuberculosis. Acta Med Iran 2010, 48:21–26.

11.	Goodridge A, Zhang T, Miyata T, Lu S, Riley LW: Antiphospholipid IgM	514
	antibody response in acute and chronic Mycobacterium tuberculosis	515
	mouse infection model. Clin Respir J 2014, 8:137–144.	516
12.	Ben-selma W, Harizi H, Marzouk M, Ben Kahla I, Ben Lazreg F, Ferjeni A,	517
	Boukadida J: Evaluation of the diagnostic value of measuring IgG, IgM,	518
	and IgA antibodies to mycobacterial A60 antigen in active tuberculosis.	519
	Diagn Microbiol Infect Dis 2010, 68:55–59.	520
13.	Feng X, Xiu B, Chen K, Yang X, Zhang H, Yue J, Tan Y, Li H, Nicholson RA,	521
	Tam AW, Zhao P, Zhang L, Liu J, Song X, Wang G, Zhang H: Enhanced	522
	serodiagnostic utility of novel Mycobacterium tuberculosis polyproteins.	523
	J Infect 2013, 66:366–375.	524
14.	Uma Devi KR, Ramalingam B, Brennan PJ, Narayanan PR, Raja A: Specific	525
	and early detection of IgG, IgA and IgM antibodies to Mycobacterium	526
	tuberculosis 38kda antigen in pulmonary tuberculosis. Tuberculosis 2001,	527
	81:249–253.	528
15.	Mattos AM, Almeida Cde S, Franken KL, Alves CC, Abramo C, de Souza MA,	529
	L’Hotellier M, Alves MJ, Ferreira AP, Oliveira SC, Ottenhoff TH, Teixeira HC:	530
	Increased IgG1, IFN-gamma, TNF-alpha and IL-6 responses to	531
	Mycobacterium tuberculosis antigens in patients with tuberculosis	532
	are lower after chemotherapy. Int Immunol 2010, 22:775–782.	533
16.	Glatman-Freedman A, Casadevall A: Serum therapy for tuberculosis	534
	revisited: reappraisal of the role of antibody-mediated immunity against	535
	Mycobacterium tuberculosis. Clin Microbiol Rev 1998, 11:514–532.	536
17.	Azzurri A, Kanaujia GV, Sow OY, Bah B, Diallo A, Del Prete G, Gennaro ML:	537
	Serological markers of pulmonary tuberculosis and of response to anti-	538
	tuberculosis treatment in a patient population in Guinea. Int J	539
	Immunopathol Pharmacol 2006, 19:199–208.	540
18.	Brust B, Lecoufle M, Tuaillon E, Dedieu L, Canaan S, Valverde V, Kremer L:	541
	Mycobacterium tuberculosis lipolytic enzymes as potential biomarkers	542
	for the diagnosis of active tuberculosis. PLoS One 2011, 6:e25078. 10.1371/	543
	journal.pone. 0025078.	544
19.	Welch RJ, Lawless KM, Litwin CM: Antituberculosis IgG antibodies as a	545
	marker of active Mycobacterium tuberculosis disease. Clin Vaccine	546
	Immunol 2012, 19:522–526.	547
20.	Gupta S, Shende N, Bhatia AS, Kumar S, Harinath BC: IgG subclass antibody	548
	response to mycobacterial serine protease at different stages of	549
	pulmonary tuberculosis. Med Sci Monit 2005, 11:585–588.	550
21.	Hoff ST, Abebe M, Ravn P, Range N, Malenganisho W, Rodriques DS,	551
	Kallas EG, Søborg C, Mark Doherty T, Andersen P, Weldingh K: Evaluation of	552
	Mycobacterium tuberculosis-specific antibody responses in populations	553
	with different levels of exposure from Tanzania, Ethiopia, Brazil, and	554
	Denmark. Clin Infect Dis 2007, 45:575–582.	555
22.	Mizusawa M, Kawamura M, Takamori M, Kashiyama T, Fujita A, Usuzawa M,	556
	Saitoh H, Ashino Y, Yano I, Hattori T: Increased synthesis of anti-	557
	tuberculous glycolipid immunoglobulin G (IgG) and IgA with cavity	558
	formation in patients with pulmonary tuberculosis. Clin Vaccine	559
	Immunol 2008, 15:544–548.	560
23.	Gupta S, Kumari S, Banwalikar JN, Gupta SK: Diagnostic utility of the	561
	estimation of mycobacterial antigen A60 specific immunoglobulins IgM,	562
	IgA and IgG in the sera of cases of adult human tuberculosis. Tuber Lung	563
	Dis 1995, 76:418–424.	564
24.	Shin AR, Lee KS, Lee JS, Kim SY, Song CH, Jung SB, Yang CS, Jo EK, Park JK,	565
	Paik TH, Kim HJ: Mycobacterium tuberculosis HBHA protein reacts	566
	strongly with the serum immunoglobulin M of tuberculosis patients. Clin	567
	Vaccine Immunol 2006, 13:869–875.	568
25.	Rabahi MF, Junqueira-Kipnis AP, Dos Reis MC, Oelemann W, Conde MB:	569
	Humoral response to HspX and GlcB to previous and recent infection by	570
	Mycobacterium tuberculosis. BMC Infect Dis 2007, 7:148.	571
26.	Demkow U, Filewska M, Michalowska-Mitczuk D, Kus J, Jagodzinski J,	572
	Zielonka T, Zwolska Z, Wasik M, Rowinska-Zakrzewska E: Heterogeneity	573
	of antibody response to myobacterial antigens in different clinical	574
	manifestations of pulmonary tuberculosis. J Physiol Pharmacol 2007,	575
	58(Suppl 5):117–127.	576
27.	Reljic R, Williams A, Ivanyi J: Mucosal immunotherapy of tuberculosis: is	577
	there a value in passive IgA? Tuberculosis 2006, 86:179–190.	578
28.	Bezerra JM, Beck ST, Kanunfre KA, Leite OM, Ferreira AW: A study of IgA	579
	antibody response to different Mycobacterium tuberculosis antigens in	580
	the diagnosis and monitoring of pulmonary tuberculosis. Braz J Infect Dis	581
	2009, 13:53–58.	582
29.	Schuler NW, Rom WN: The host immune response to tuberculosis. Am J	583
	Respir Care Med 1998, 157:679–691.	584

Feng et al. BMC Infectious Diseases 2014, 14:336

http://www.biomedcentral.com/1471-2334/14/336

585 30.  Araujo LS, Maciel RM, Trajman A, Saad MH: Assessment of the IgA

586 immunoassay diagnostic potential of the Mycobacterium tuberculosis

587 MT10.3-MPT64 fusion protein in tuberculous pleural fluid. Clin Vaccine

588 Immunol 2010, 17:1963–1969.

589 31.  Siddiqi UR, Leano PS, Chagan-Yasutan H, Shiratori B, Saitoh H, Ashino Y,

590 Suzuki Y, Hattori T, Telan EF: Frequent detection of anti-tubercular-

591 glycolipid-IgG and -IgA antibodies in healthcare workers with latent

592 tuberculosis infection in the Philippines. Clin Dev Immunol 2012,

593 610707:. 10.1155/2012/610707.

594 32.  Kaisermann MC, Sardella IG, Trajman A, Coelho TL, Kämpfer S, Jonas F,

595 Singh M, Saad MH: IgA antibody responses to Mycobacterium

596 tuberculosis recombinant MPT-64 and MT-10.3 (Rv3019c) antigens

597 in pleural fluid of patients with tuberculous pleurisy. Int J Tuberc

598 Lung Dis 2005, 9:461–466.

599 33.  Samanich KM, Keen MA, Vissa VD, Harder JD, Spencer JS, Belisle JT,

600 Zolla-Pazner S, Laal S: Serodiagnostic potential of culture filtrate antigens of

601 Mycobacterium tuberculosis. Clin Diagn Lab Immunol 2000, 7:662–668.

602 34.  Moran AJ, Treit JD, Whitney JL, Abomoelak B, Houghton R, Skeiky YA,

603 Sampaio DP, Badaró R, Nano FE: Assessment of the serodiagnostic

604 potential of nine novel proteins from Mycobacterium tuberculosis. FEMS

605 Microbiol Lett 2001, 9875:1–6.

606 35.  Fujita Y, Doi T, Sato K, Yano I: Diverse humoral immune responses and

607 changes in IgG antibody levels against mycobacterial lipid antigens in

608 active tuberculosis. Microbiology 2005, 151:2065–2074.

609 36.  Li X, Xu H, Jiang S, Jing K, Wang L, Liu X, Li W, Zhang H, Wang L: TB-SA

610 antibody test for diagnosis and monitoring treatment outcome of

611 sputum smear negative pulmonary tuberculosis patients. Southeast Asian

612 J Trop Med Public Health 2011, 42:1147–1153.

613 37.  Imaz MS, Comini MA, Zerbini E, Sequeira MD, Spoletti MJ, Etchart AA,

614 Pagano HJ, Bonifasich E, Diaz N, Claus JD, Singh M: Evaluation of the

615 diagnostic value of measuring IgG, IgM and IgA antibodies to the

616 recombinant 16-kilodalton antigen of mycobacterium tuberculosis in

617 childhood tuberculosis. Int J Tuberc Lung Dis 2001, 5:1036–1043.

618 38.  Raja A, Ranganathan UD, Bethunaickan R: Improved diagnosis of

619 pulmonary tuberculosis by detection of antibodies against multiple

620 Mycobacterium tuberculosis antigens. Diagn Microbiol Infect Dis 2008,

621 60:361–368.

622 39.  Goodridge A, Cueva C, Lahiff M, Muzanye G, Johnson JL, Nahid P, Riley LW:

623 Anti-phospholipid antibody levels as biomarker for monitoring

624 tuberculosis treatment response. Tuberculosis 2012, 92:243–247.

625 40.  Sireci G, Dieli F, Di Liberto D, Buccheri S, La Manna MP, Scarpa F, Macaluso P,

626 Romano A, Titone L, Di Carlo P, Singh M, Ivanyi J, Salerno A: Anti-16-kilodalton

627 mycobacterial protein immunoglobulin m levels in healthy but purified

628 protein derivative-reactive children decrease after chemoprophylaxis. Clin

629 Vaccine Immunol 2007, 14:1231–1234.

630 41.  Baumann R, Kaempfer S, Chegou NN, Nene NF, Veenstra H, Spallek R,

631 Bolliger CT, Lukey PT, van Helden PD, Singh M, Walzl G: Serodiagnostic

632 markers for the prediction of the outcome of intensive phase

633 tuberculosis therapy. Tuberculosis 2013, 93:239–245.

634 42.  Lyashchenko K, Colangeli R, Houde M, Al Jahdahli H, Menzies D,

635 Gennaro ML: Heterogeneous responses to tuberculosis. Infect Immun

636 1998, 66:3936–3940.

637 43.  Kunnath-Velayudhan S, Salamon H, Wang HY, Davidow AL, Molina DM,

638 Huynh VT, Cirillo DM, Michel G, Talbot EA, Perkins MD, Felgner PL, Liang X,

639 Gennaro ML: Dynamic antibody responses to the Mycobacterium

640 tuberculosis proteome. Proc Natl Acad Sci U S A 2010, 107:14703–14708.

 

641    doi:10.1186/1471-2334-14-336

642 Cite this article as: Feng et al.: IgG, IgM and IgA antibodies against the

643 novel polyprotein in active tuberculosis. BMC Infectious Diseases

644 2014 14:336.