Detection of natural infection with Mycobacterium intracellulare in healthy wild-caught Chacma baboons (Papio ursinus) by ESAT-6 and CFP-10 IFN-γ ELISPOT tests following a tuberculosis outbreak

Background Both tuberculous and non-tuberculous mycobacteria can cause infection in nonhuman primates (NHP), indicating the existence of potential zoonotic transmission between these animals and visitors to zoos or animal handlers in primate facilities. Screening of mycobacterial infections in NHP is traditionally done by tuberculin skin test (TST), which is unable to distinguish between pathogenic and non-pathogenic mycobacterial infections. In this study, we investigated the use of ESAT-6 and CFP-10 for detection of mycobacterial infections in a wild-caught baboon colony after one baboon died of tuberculosis (TB). Methods Peripheral blood lymphocytes for interferon-gamma enzyme-linked immunospot assay (IFN-γ ELISPOT) assay were obtained from TST positive baboons and those in contact with tuberculous baboons before being euthanased, autopsied and lung tissues taken for histology and mycobacterial culture. Results Both ESAT-6 and CFP-10 IFN-γ ELISPOT assays were able to detect early M. tuberculosis but also M. intracellulare infection. Although this indicates potential cross-reactivity with M. intracellulare antigens, the method was able to distinguish M. bovis BCG vaccination from M. tuberculosis infection. This assay performed better than the TST, which failed to detect one M. tuberculosis and two early M. intracellulare infections. Conclusion These results suggest that the IFN-γ ELISPOT assay could improve the detection of M tuberculosis infections when screening NHP. There is some doubt, however, concerning specificity, as the assay scored positive three animals infected with M. intracellulare.


Background
The majority of the disease-causing species of the genus Mycobacterium belong to two groups, the M. tuberculosis complex (MTC) and the M. avium-intracellulare complex (MAC). MTC comprises of M. tuberculosis, M. bovis, M africanum and the attenuated M. bovis bacille Calmette-Guerin (BCG). With the exception of BCG, these species are pathogenic and can cause tuberculosis (TB) in humans and animals [1][2][3]. Disease caused by MTC pathogens is characterised by tubercle formation in the infected tissues. These bacteria are considered obligate intracellular pathogens and the most efficient mode of transmission is via respiratory route [4]. In contrast, environmental mycobacteria have the capacity to survive and multiply under a wide range of environmental conditions [5] and to interact with a variety of environmental reservoirs, including protozoa and insects [6]. The majority of environmental mycobacteria are considered non-pathogenic. However, members of MAC are capable of causing nontuberculous infections and disease in a wide range of animal species [5,6]. MAC comprises of two species, M. avium and M. intracellulare, and although these species are considered opportunistic pathogens, their role in human and animal diseases and their pathogenic potential have become increasingly recognised, especially in immunocompromised persons with HIV-1 infection [7,8].
Non-human primates (NHP) are susceptible to infections caused by members of MTC [9][10][11][12][13] and MAC [14,15], including M. intracellulare [16,17], indicating the existence of potential zoonotic transmission between NHP and visitors to zoos and game parks as well as animal handlers and laboratory workers using biological products from NHP. The tuberculin skin test (TST) is the traditional in vivo screening test for detecting TB in NHP [18]. This test utilizes PPDs, which are crude antigen preparations, usually made from M. tuberculosis, M. bovis or M. avium. As PPD contains cross-reactive antigens [19], a positive TST could indicate tuberculous and non-tuberculous disease, prior vaccination with BCG, or immune sensitisation with non-pathogenic environmental mycobacteria. In addition, BCG vaccine vectors are increasingly being used in NHP studies [20][21][22][23], making it important to distinguish such vaccinations from natural pathogenic mycobacterial infections.
The 6-kDa early secretory antigenic target (ESAT-6) and culture filtrate protein 10 (CFP-10) are well defined mycobacterial proteins [24][25][26]. Both are low molecular-weight proteins which are secreted in the culture filtrate by M. tuberculosis and other pathogenic mycobacteria, including M. bovis [27], following short-term axenic culture. These proteins are absent from the genomes of all BCG substrains and those of several non-pathogenic environmental mycobacteria including M. avium avium and M. avium paratuberculosis [24,28]. Both ESAT-6 and CFP-10 induce strong T-cell IFN-γ responses [27,[29][30][31], prompting their proposed use as diagnostic markers for M. tuberculosis and M. bovis infections [25,32,33]. However, the presence of ESAT-6 and CFP-10 orthologues in other mycobacterial species, including non-pathogenic species such as M. smegmatis and M. gastri [34] and environmental species like M. flavescens [28] has raised some concerns regarding the potential use of these antigens as diagnostic markers [35]. Paradoxically, the presence of ESAT-6 and CFP-10 in other mycobacterial species does not seem to confound the detection of M. tuberculosis-and M. bovis-associated specific responses in clinical and epidemiological practice [29,30,32,[36][37][38], although these studies were mostly performed in low-incidence settings and patients may not have had significant previous contact with non-tuberculous mycobacteria. A study by Arend et al [39] showed that when ESAT-6 and/or CFP-10 enzyme-linked immunosorbent assay (ELISA) and enzyme-linked immunospot assay (ELISPOT) were used to measure interferon-gamma production, most M. kansasii-or M. marinum-infected patients and several persons exposed to environmental mycobacteria responded to ESAT-6 and/or CFP-10. IFN-γ production by T cells from leprosy patients, TB patients and unexposed controls in response to the M. leprae homologue of CFP-10 also showed significant crossreactivity with CFP-10 of M. tuberculosis [40].
In the present study, we evaluated ESAT-6 and CFP-10 for use in the interferon-gamma (IFN-γ) ELISPOT assay for diagnosis of mycobacterial infections in baboons held in a research facility in South Africa after the death of one baboon due to TB.

Baboons
Ninety wild-caught Chacma baboons (Papio ursinus) were involved in this study. Of these, 34 baboons had been newly introduced into the colony from the wild, while the other 56, including the one that died of M. tuberculosis infection, had been kept in the colony for over 3 years. Of the 56 baboons, 10 had been sensitised to BCG by experimental inoculation, 2 to 3 years prior to the current study. Ethical approval for the experimentation on the baboons was obtained from the Animal Ethics Committee of the University of Cape Town.

Tuberculin Skin test
Baboons were anaesthetised with ketamine hydrochloride (10 mg/kg body mass; intramuscular injection) and injected intradermally with 100 μL (1000 IU in saline) of bovine tuberculin PPD (Institute of Animal Science and Health, Lelystad, The Netherlands) in the upper palpebrum of the eyelid. Reaction to tuberculin was checked by visual observation at 24, 48 and 72 hours and interpreted as TST positive or negative. The TST positivity reactions comprised of erythema of palpebrum alone, various degrees of erythema with minimum swelling or slight swelling of palpebrum without erythema, and obvious swelling of palpebrum with drooping of the eyelid or swelling and/or necrosis of the palpebrum with eyelid closed. Bruise-extravasation of blood associated with injection of tuberculin or no detectable reaction on the palpebrum were interpreted as TST negative.

IFN-γ ELISPOT assay
Peripheral blood mononuclear cells (PBMC) were isolated from heparinised blood by standard Ficoll-gradient centrifugation. Freshly isolated or cryo-preserved PBMC were used in a standard IFN-γ ELISPOT assay as previously described [41]. Briefly, ELISPOT plates (MultiScreen-IP, Millipore) were coated overnight at 4°C with purified anti-human IFN-γ monoclonal antibody (clone 1-D1K, Mabtech). PBMC were incubated for 24 h at 37°C in triplicate at 200,000 cells per well with PPD (4 μg/mL), ESAT-6 (4 μg/mL), or CFP-10 (4 μg/mL) as the stimulant in a total volume of 0.1 mL. Triplicate wells with cells and PHA (4 μg/mL) and culture medium alone, served as positive and background stimulation controls respectively. The average of triplicate counts of IFN-γ spot forming cells (SFC) was calculated for each stimulant and normalised to 10 6 PBMC to give IFN-γ SFC/10 6 PBMC. The results were reported as net IFN-γ SFC/10 6 PBMC after subtracting the background SFC/10 6 PBMC obtained in the absence of any stimulant (PBMC plus culture medium alone).
To determine if a response was positive, a cut-off value was established for PPD, ESAT-6 and CFP-10 using PBMC from 27 TST negative healthy baboons whose background response or response to ESAT-6 or CFP-10 was <50 IFN-γ SFC/10 6 PBMC. The cut-off value was defined as the mean net IFN-γ SFC/10 6 PBMC after subtracting the value of background reaction plus 3 standard deviations of the mean or twice the highest value of background reaction, whichever was greater. The mean net responses plus three standard deviations for PPD, ESAT-6 and CFP-10 were 8 + 35, 12 + 31 and 5 + 22 IFN-γ SFC/10 6 PBMC respectively while the highest value of background reactions were 43, 35 and 30. Thus, the cut-off values for PPD, ESAT-6 and CFP-10 were determined as 86, 70 and 60 IFN-γ SFC/10 6 PBMC, respectively. The mean background stimulation in the absence of any stimulant and in response to the PHA positive controls were 5 ± 4 and 1562 ± 697 SFC/10 6 PBMC respectively.

Euthanasia and necropsy
Baboons with a positive TST were euthanased for necropsy using sodium pentobarbital (200 mg/kg body mass). In addition, baboons that were caged adjacent to those which were found to have tubercles at necropsy (here referred to as contact baboons) were also euthanased for necropsy. All the TST-positive baboons had been kept in the colony for over 3 years while 6 of 8 contact baboons were less than 3 months in the colony. Before euthanasia, blood for PBMC isolation was collected from these animals. Also, as part of ongoing research, cryo-preserved PBMC were available from some of these animals. Organs in the thoracic and abdominal cavities were examined macroscopically for the presence of tubercles, adhesions or enlargement and/or caseation of lymph nodes. Samples for histology, mycobacteria-specific staining and culture were taken from the lung tissue and/or bronchial lymph nodes (BLN).

Mycobacteria culture and identification
Fresh tissues from the lung and/or BLN collected at necropsy were cultured for mycobacterial culture, using standard methods in the TB research laboratory of the Stellenbosch University. Standard precautions were followed to limit cross-contamination. All cultures were genotyped by the IS6110 Restriction Fragment Length Polymorphism (RFLP) method [42]. Isolates were also typed according to the Mycobacterium tuberculosis complex typing method developed in our laboratory [43]. Mycobacterium tuberculosis complex negative samples were further subjected to a 5'-16S rRNA gene PCR-sequencing assay, which is able to identify and speciate Mycobacterium spp. [44,45]. Used in combination these methods would be able to identify superinfections between mycobacterial species of MTC and MAC.

Tuberculin Skin test
Following the death of one baboon (B662) due to pulmonary TB diagnosed at necropsy, the remaining 89 baboons in the colony were screened for mycobacterial infection using bovine PPD tuberculin. Initially, only eight baboons tested positive, including three (B454, B536 & B548) with known prior vaccination with BCG or recombinant BCG. One additional baboon (B369) tested positive three months after initially testing negative, resulting in a total of nine tuberculin reactors ( Figure 1 and Table  1). The test reaction ranged from severe (drooping of the eyelid, necrosis of the palpebrum with eyelid completely closed; B524, B531 & B629) to slight reactions (erythema of palpebrum with minimal swelling; B369, B454, B536 & B548). Reactions on eyelids of baboon B659 and B697 were moderate (obvious swelling of palpebrum without drooping or necrosis).

Necropsy
The nine TST reactors and an additional eight contact baboons were euthanased for necropsy and further investigations. Of the nine TST positive baboons, four had classical tubercles in the lungs including at least one caseous BLN, one had a non-classical tubercle-like lesion (B697) with an enlarged but not caseous BLN and four had no macroscopic lesions (Table 1). Paradoxically, one of eight TST negative baboons (B673) had a single caseous BLN with no visible tubercles. The severity of lesions varied widely, ranging from numerous small tubercles in the lungs and one or more caseous BLN (B524, B531 and B629) to just a single caseous BLN (B673). These macroscopic lesions were restricted to the pulmonary system.

Histology, ZN test and M. tuberculosis immunoperoxidase staining
As shown in the Table 1, necrotic granulomas and areas of necrosis were found in the lung and BLN tissues of three of nine TST positive baboons while other lesions such as lymphocytic infiltration and granulomatous inflammation, suggestive of mycobacterial infection, were found in four other TST reactors. Also, five of seven histologic lesions were confirmed by ZN and MTIP staining. For the TST negative baboons, only mild histological lesions were found in four of eight baboons and one (B673) was positive by ZN and MTIP staining.

Mycobacterial culture and identification
Of the 17 baboons' specimens that were cultured, eight yielded mycobacterial growth. M. tuberculosis was identified in five and M. intracellulare in three of these cultures by PCR-based methods and 16S rRNA sequencing. M. tuberculosis was cultured from four of the nine TST reactors and one of eight TST negative baboons while M. intracellulare was cultured from one TST reactor and two TST negative baboons (Figure 1 and Table 1). It was unlikely that the culture of M. intracellulare represented laboratory cross-contamination as this species is rarely identified in sputum specimens routinely cultured in our laboratory. No superinfection was detected in any of these samples.

IFN-γ ELISPOT assay
PBMC were isolated from some baboons at the time of TST or just before euthanasia for determination of IFN-γ ELISPOT response to bovine PPD, ESAT-6 and CFP-10. As shown in Figure 1 and Table 1, PPD ELISPOT test was positive for all TST positive reactors and negative for all TST non-reactors, indicating agreement between these two tests. However, two of the tuberculin reactors (B536 and B697) were negative for both ESAT-6 and CFP-10 ELIS-POT test. The mycobacterial culture in these animals was also negative. The ESAT-6 and CFP-10 ELISPOT test was positive for B454, another TST reactor which had previ-Summary of the TST, ESAT-6 and CFP-10 IFN-γ ELISPOT assay and mycobacterial culture outcome   ously been vaccinated with BCG, although no mycobacteria was isolated from tissues from this animal. The ESAT-6 and CFP-10 ELISPOT test was also positive for B369, which was initially TST negative and mycobacteria culture positive for M. intracellulare. For the TST non-reactors, ESAT-6 and CFP-10 ELISPOT test was positive for four baboons, two (B694 and B709) from which M. intracellulare was isolated. However, no mycobacteria was isolated from the other two baboons (B679 and B689), which were also positive by ESAT-6 and CFP-10 ELISPOT.
In an attempt to understand the kinetics of immune responses to M. tuberculosis and M. intracellulare, we performed retrospective measurements of T-cell IFN-γ responses in an ELISPOT assay using cryo-preserved PBMC, previously obtained from baboons B369 and B662. As depicted in Figure 2, the responses to ESAT-6 for B369 could be detected as early as 18 weeks before euthanasia but the PPD (and CFP-10) responses were detectable only 6 weeks later. Thus, the TST was initially negative at 16 weeks before euthanasia although the ELISPOT assay value was above the positive cut-off value for ESAT-6 (but not PPD and CFP-10). This delay in induction of a PPD response was not observed for B662 that died of TB.

Discussion
The TST, which is a delayed-type hypersensitivity (DTH) reaction, is partially T cell-mediated [46] and IFN-γ plays a major role [47]. The major drawback of this test is the inability to distinguish active pathogenic infection from healthy carrier status, or from non-pathogenic exposure to environmental mycobacteria or BCG. The majority of TB outbreaks in NHP are associated with human contact or TB infected animals and carcasses [9,10,12] and rarely give rise to latent TB infection [11]. Thus, MTC infections in NHP are considered progressive and almost always fatal. However, infection with MAC is usually chronic [14][15][16][17] and sometimes refractory to TST [16]. In order to eliminate their zoonotic potential, early and reliable detection of potential pathogenic mycobacterial infections, including members of MAC in the nonhuman primate colonies is crucial.
Both ESAT-6 and CFP-10 have been proposed as diagnostic markers for pathogenic infection caused by M. tuberculosis and M. bovis in human and animals [25,[31][32][33] because they are well defined antigens [24][25][26], absent from BCG vaccine strains and some non-pathogenic environmental mycobacteria [28,36], and they associate in M. tuberculosis-exposed people with a higher risk of developing active tuberculosis [29,30].
In our study, the IFN-γ ELISPOT response to ESAT-6, CFP10 and PPD were compared to the TST. There was concordance between the TST and PPD ELISPOT tests.
However the results for ESAT-6 and CFP-10 ELISPOT and TST were discordant in two of seven TST-positive and four of seven TST-negative animals. One of the two TST-positive but ESAT-6 and CFP-10 negative baboon (B536) could be a result of previous exposure to BCG via inoculation with a recombinant BCG vaccine 2-3 years previously, but the other baboon (B697) possibly represents environmental sensitisation as this animal had been caught from the wild less than 3 months before the test and had no known exposure to BCG.

Competing interests
The author(s) declare that they have no competing interests.

Authors' contributions
GKC participated in collection and subsequent delivery of animals' samples to the laboratory, processing PBMC and performing of the ELISPOT tests, helped in TST and autopsies, compiled data and drafted primarily the manuscript. RMW performed cultures, mycobacteria speciation and helped drafting the manuscript. NCGvP helped in IFN-γ ELISPOT responses to mycobacterial proteins at various times prior to death (B662) or euthanasia (B369) Figure 2 IFN-γ ELISPOT responses to mycobacterial proteins at various times prior to death (B662) or euthanasia (B369). Three mycobacterial antigen preparations (bovine PPD, ESAT-6 and CFP-10) were used in IFN-γ ELISPOT assay using cryo-preserved PBMC from baboon B662 (infected with M. tuberculosis) and B369 (infected with M. intracellulare). The IFN-γ response to each protein was determined by subtracting the background (response to PBMC in the culture medium only) from the antigen response. Data points are the magnitudes of response to each antigen preparation at various time points. The dotted lines indicate the cut-off value for ESAT-6 (70 IFN-γ SFC/10 6 PBMC). PBMC mycobacteria speciation and drafting the manuscript. WAB helped in ELISPOT assay and compiling of data. RJW participated partially in necropsies and helped in drafting the manuscript. EGS contributed in ELISPOT assay analysis and facilitating histopathology. A-LW was the principal investigator and was responsible for the study design. All the authors read and approved the final manuscript.