• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of clinexpimmunolLink to Publisher's site
Clin Exp Immunol. Nov 2002; 130(2): 190–195.
PMCID: PMC1906511

Exposure to Mycobacterium avium primes the immune system of calves for vaccination with Mycobacterium bovis BCG


The objective of the investigation was to provide data on how a prior exposure of cattle to Mycobacterium avium, used here as a model of exposure to an environmental mycobacterium, affected the cellular immune response that follows vaccination with Mycobacterium bovis BCG. The assessment of cellular immune responses included lymphocyte proliferation assays, the delayed hypersensitivity skin test and IFN-γ synthesis in whole blood cultures. One group of calves was inoculated subcutaneously with M. avium followed 12 weeks later by M. bovis-BCG. The other group was vaccinated subcutaneously with BCG alone. Calves previously exposed to M. avium responded more rapidly, as assessed in the in vitro assays, to purified protein derivative (PPD) from M. avium (PPD-A) or M. bovis (PPD-B) than did calves inoculated with BCG only, indicating that the exposure to M. avium had primed the immune response in these calves. Following inoculation of BCG the intensity of the in vitro responses and the delayed hypersensitivity skin test to PPD-A was higher for the M. avium-primed animals while the responses to PPD-B were similar in the M. avium-primed and BCG-only groups. The results are consistent with a model in which prior exposure to environmental mycobacteria does not necessarily inhibit the immune response to the vaccine strain, BCG. They suggest that M. avium infection primes the immune system of calves and that the detection of an immune response specific for M. bovis BCG is masked by reactivity to antigens also present in M. avium.

Keywords: BCG, bovine TB, interferon-γ, Mycobacteria, skin test


Mycobacterium bovis is the cause of bovine tuberculosis (TB) a progressive and potentially fatal zoonotic disease of cattle. Data from investigations in rodent models, as well as humans, support the conclusion that cellular immune responses are critical for immunity to mycobacteria. IL-12 synthesis by antigen-presenting cells (APC) and interferon-γ (IFN-γ) synthesis by NK cells initially, as part of the innate response, and subsequently by T cells, as part of the adaptive response, have been established as pivotal mediators of immunity [13]. Diagnosis of infection or vaccination is also based on the assessment of cellular immune responses. These include the delayed type hypersensitivity (DTH) response detected by skin testing and the synthesis of IFN-γ detected by enzyme-linked immunosorbent assay (ELISA), the two validated and established methods for the diagnosis of bovine TB [4,5].

Vaccination of humans against TB has been undertaken for a century. The attenuated M. bovis strain BCG generates a significant degree of protection against Mycobacterium tuberculosis in rodent models. However, vaccine trials in humans have not given consistent results. In some a high degree of protection has been shown, in others none. A reason that has been put forward to explain this is that previous exposure to environmental mycobacteria compromises the immune response to BCG [68]. It has also been reported that humans with a pre-existing response to environmental mycobacteria were more resistant to TB than non-responding humans, indicating that exposure to environmental mycobacteria provided a degree of immunity to M. tuberculosis [8]. Two alternative, but not mutually exclusive, explanations have been suggested to account for these observations. One is that prior exposure might reduce immunity induced by BCG vaccination; effective vaccine take may be prevented, or prior exposure might prime for an inappropriate bias in the immune response generated subsequently by the inoculation of BCG. Alternatively, different levels of baseline immunity are present in populations before vaccination as a consequence of the varied exposure to bacteria in the environment and the differences between the vaccinated and non-vaccinated groups are potentially smaller in populations that have a high incidence of environmental responders. These two hypotheses can be described simply as ineffectiveness or masking, respectively [8].

Studies in rodents confirm that exposure to environmental bacteria affects subsequent responses. Some suggest that BCG is not ineffective in these animals but that only a certain level of immunity is achievable and that a level of protection is already present due to exposure to environmental mycobacteria [6]. Others suggest that exposure to environmental bacteria prevents BCG from inducing immunity [9].

A number of vaccination studies have been reported in cattle. These show that a significant degree of immunity to respiratory challenge with M. bovis can be induced by subcutaneous (s.c.) vaccination with BCG [10,11]. Although it is clear that this immunity is not complete it does provide a model for exploring correlates with immunity. It has also been proposed that in cattle exposure to environmental mycobacteria reduced the effectiveness of BCG vaccination, the implication being that the animals did not respond effectively to BCG [12].

The objective of the reported investigation was to provide data on how a prior exposure of cattle to M. avium – as a model of exposure to an environmental mycobacterium – affected the cellular immune response to vaccination with BCG. These cellular immune responses would include lymphocyte proliferation assays, the DTH skin test and IFN-γ synthesis by blood cells. The questions that were addressed were: (1) did exposure to M. avium prevent BCG from inducing an immune response, indicating protection against infection by the vaccine strain? (2) Did exposure to M. avium prime the immune system and so modulate the cellular immune response to BCG?


Animal groups

Calves were conventionally reared Bos taurus held in appropriate secure accommodation and were aged about 6 months. Five calves were inoculated s.c. with 105 colony-forming units (cfu) M. avium strain D4ER (kindly provided by M. Vordermeier, Veterinary Laboratories Agency (VLA), Weybridge, UK) grown in Middlebrook 7H9 medium and suspended in phosphate buffered saline (PBS). A second group of five calves was inoculated with PBS. Blood was taken into heparin (10 U/ml) before the calves were inoculated (week 0) and 3, 6, 9 and 12 weeks post-inoculation. On week 12 all 10 calves were inoculated s.c. with 106 cfu M. bovis BCG strain Pasteur. Blood was taken into heparin on weeks 13, 14, 15, 16, 18, 21 and 24. After the final blood sampling on week 24 the DTH response of calves was tested in the standard comparative skin test [13]. The skin at the base of the neck was shaved, the thickness measured with callipers and 100 µl of PPD from M. avium and M. bovis (PPD-A or PPD-B, respectively, containing 1000 units/ml, VLA, Weybridge, UK) was inoculated intradermally (i.d.). Skin thickness was measured 3 days later at the sites of the i.d. inoculations. All statistical comparisons were by Student's t-test. The experiment was approved according to national regulations and local ethical committee approval.

In vitro assays

Assays for IFN-γ synthesis were as follows. PPD-A or PPD-B diluted in RPMI (Gibco), or RPMI in a total volume of ≤100 µl was added to 2 ml undiluted whole-heparinized blood to give a final concentration of 20 µg/ml. After incubation at 37°C in a CO2 incubator for 24 h the supernatant plasma was removed after centrifugation and stored at −20°C until assayed for IFN-γ by ELISA. The standard capture ELISA has been described in detail [14]. Monoclonal antibody (MoAb) CC330 at 2 µg/ml was used as the capture MoAb and biotinylated MoAb CC302 at 2 µg/ml to detect bound cytokine. Tetramethylbenzidine was used as substrate. Estimates of amount of IFN-γ/ml were made by comparing O.D. values with those of a standard curve of recombinant bovine IFN-γ.

The lymphocyte proliferation assay was performed with diluted blood. Heparinized blood was diluted 1/5 with RPMI containing gentamycin (50 µg/ml) and 100 µl volumes added in triplicate to a 96-well U-plate. Antigen was 100 µl PPD-A or PPD-B, added to give a final concentration in wells of 10 µg/ml diluted in RPMI, or RPMI. Plates were incubated for 5 days in a CO2 incubator and 0·037MBq [3H]thymidine added for the final 18 h of incubation prior to harvesting. Results are expressed as mean cpm ± s.d. All statistical analyses were by Student's t-test.



The IFN-γ content of culture supernatants is shown in Fig. 1 for weeks 0, 3, 6, 9 and 12, i.e. after exposure to M. avium, and in Fig. 2 for weeks 12, 13, 14, 15, 16, 18, 21 and 24, i.e. after the inoculation of BCG on week 12. The histograms have been separated because of the different scales necessary to show the responses. Following the inoculation of calves with M. avium higher levels of IFN-γ were present in supernatants of cells cultured with PPD-A on weeks 3 (P = 0·025), 6 (P = 0·043), 9 (P = 0·012) and 12 (P = 0·025) for animals inoculated with mycobacteria compared to the control animals. No significant difference was evident at week 0 when the two animal groups were compared (Fig. 1). Thus, there was clear evidence for this dose of M. avium inducing an immune response and priming the immune system of calves.

Fig. 1
IFN-γ levels in supernatants from whole blood after PPD restimulation, weeks 0–12. Whole blood was incubated with medium (grey columns), PPD-A (black columns) or PPD-B (open columns) for 24 h. Supernatants were removed and assayed by ELISA ...
Fig. 2
IFN-γ levels in supernatants from whole blood after PPD restimulation, weeks 12–24. Blood incubated with medium (grey columns), PPD-A (black columns) or PPD-B (open columns) and supernatants assayed for IFN-γ as in Fig. 1. Mean ...

IFN-γ levels evident in supernatants of blood taken from BCG-only vaccinates and M. avium + BCG vaccinates incubated with PPD-A or PPD-B increased following the inoculation of BCG, reaching levels that were about 10-fold higher than that seen after M. avium alone (Fig. 2). In the M. avium + BCG group the IFN-γ synthesis with PPD-A was higher than in the BCG-only group on weeks 14 (P = 0·02) and 15 (P = 0·05) and in the same group the IFN-γ synthesis with PPD-B was higher on week 14 (P = 0·04). Thus, the onset of the response to PPD appeared more rapidly in M. avium-primed animals than in BCG-only animals, indicative of an anamestic response, although the levels reached from week 16 were the same.

When IFN-γ synthesis with PPD-A and PPD-B for blood from individual animals was compared marked differences were evident between for the two calf groups. Thus, at week 24 for the M. avium + BCG group a comparison of the levels of IFN-γ in supernatants from cultures incubated with PPD-A was found to be greater for 5/5 individual animals than the level in supernatants incubated with PPD-B. In contrast for individual animals inoculated with BCG alone the levels of IFN-γ in supernatants incubated with PPD-B was higher in 5/5 cases than that seen in supernatants from cultures incubated with PPD-A. Based on this, exposure to M. bovis BCG would not have been diagnosed in the M. avium-primed animals.

Proliferation responses

The mean proliferative response in the two groups of calves is shown in Fig. 3. An enhanced proliferative response was evident to PPD-A for calves inoculated with M. avium prior to the inoculation of BCG (P = 0·022) at week 9 when the means of the two groups were compared. After inoculation of BCG the proliferative response to PPD-A was higher for the M. avium + BCG group compared to the BCG-only group within 1 week following BCG vaccination. This difference appeared to be maintained throughout the course of the experiment with differences being significant on weeks 13 (P = 0·0005), 14 (P = 0·01), 18 (P = 0·005), 21 (P = 0·0013) and 24 (P = 0.0005). The proliferative response with PPD-B in the M. avium + BCG group that developed 1 week after inoculation of BCG was significantly different at this time, week 13 (P = 0·006) from the BCG-only group, but not subsequently.

Fig. 3
Proliferation of lymphocytes after PPD restimulation. Diluted blood (1 : 10) incubated with medium (grey columns), PPD-A (black columns) or PPD-B (open columns) for 5 days. [3H]thymidine was added for the final 18 h of culture. Mean counts per minute ...

Skin test reactions

The mm increase in skin thickness for the individual animals 3 days after i.d. inoculation of PPD-A or PPD-B is shown in Fig. 4. For calves inoculated with M. avium + BCG the increase in thickness to PPD-A was significantly higher than that in the BCG-only group (mean of 10·6 ± 1·52 mm compared to 2·6 ± 1·34 mm, respectively, P = 0·0002). The increase in skin thickness was not significantly different in the two calf groups following i.d. inoculation of PPD-B (7·4 ± 3·36 mm compared to 7·6 ± 2·41 mm). However, if a comparison of the increase in skin thickness to PPD-A and PPD-B is made within either calf group then all five BCG-only vaccinates had more intense responses to PPD-B compared to the response with PPD-A, group means 7·6 ± 2·41 mm and 2·6 ± 1·34 mm, respectively (P = 0·002). The responses to PPD-A in the five M. avium + BCG calves (10·6 ± 1·52 mm) was not significantly different (P = 0·066) from the response to PPD-B (7·4 ± 3·36 mm). Using the standard interpretation of the avian/bovine tuberculin test for the United Kingdom the difference seen for the BCG-only vaccinates would be taken to indicate infection with M. bovis for three of the five animals, with retesting required for the other two. For the M. avium + BCG group none would be diagnosed as having been exposed to M. bovis BCG.

Fig. 4
Delayed hypersensitivity response to i.d. inoculation of PPD. Calves were inoculated with PPD from M. avium (black columns) or M. bovis (open columns). Skin thickness was measured at day 0 and day 3. Skin thickness on day 3 minus the value at day 0 is ...


The objective of the investigation was to provide data on how a prior exposure of cattle to an environmental mycobacterium could affect the cellular immune response following vaccination with BCG. Inoculation of M. avium s.c. was selected as the model. Strain D4ER was chosen, as this is the strain from which the avian PPD is derived that is used as the basis of the comparative DTH skin test in the United Kingdom [13]. DTH reactivity to this preparation is common in UK cattle, indicating that the antigens contained in it are representative of common environmental mycobacteria. The s.c. route was selected to reduce variation as far as was possible in small groups of outbred animals and as this route has been used commonly in small animal models. The cellular immune responses assessed included lymphocyte proliferation assays, the comparative skin test and IFN-γ synthesis. One possibility was that prior exposure to M. avium would prevent the induction of an immune response by BCG. Support for this hypothesis, it was argued, would be derived if the response to BCG was less in the calves that had experienced M. avium. If a lesser response was observed it would probably be the result of cross-protection and imply less replication of, and resistance to, BCG. A second possibility was that exposure to M. avium would prime the immune system and as a consequence modulate the cellular immune response that was induced following the inoculation of BCG. Support for this hypothesis would be derived if an enhanced or modified response to BCG was seen following exposure to M. avium.

Taken together, the results of all three immunological assays indicate that prior inoculation with M. avium primed the calves and influenced the response following vaccination with BCG. This response appeared to be biased towards antigens that were common to the two species of mycobacteria. Thus, a secondary response to PPD-A was evident in the two in vitro assays following the inoculation of BCG. The response to PPD-B was similar for both groups of calves following the inoculation of BCG, except for an enhancement in M. avium-primed animals noted 1 and 2 weeks after BCG vaccination.

Overall, no evidence was obtained for the hypothesis that prior exposure to M. avium reduced efficacy of BCG vaccination by restricting the multiplication of BCG in vivo and as a consequence stimulated a poorer immune response to the vaccine. However, it is possible that exposure to M. avium reduces BCG replication but without preventing BCG from boosting M. avium-induced immune responses, as tested in this study. Recently reported inhibitory effects of the prior s.c. inoculation of M. avium complex organisms on the replication of BCG following its s.c. inoculation and the immunity generated in mice [9], in which prior inoculation of M. avium did not result in priming and abrogated and the immune response induced by BCG, were not seen in the calves. Rather, the results indicated that exposure to M. avium primed the calves to antigens common to both M. avium and BCG. The contrasting findings may reflect differences between rodents and cattle and emphasize the need to examine a variety of models including, where possible, the natural host.

Data from human vaccine trials with BCG have revealed differences in efficacy against pulmonary tuberculosis in different areas of the world. There is much support for the view that regional differences in exposure to environmental mycobacteria are responsible for much of the variation observed. Thus in the United Kingdom, where the vaccine appeared effective, there were few skin test positive individuals before vaccination, while in the southern United States and India, where it did not, there were many. Although the simple interpretation is protection or not, care needs to be taken to determine whether this is due to interference or masking of protection by environmental mycobacterial infections. Masking and interference are quite different mechanisms, either of which could reduce protection compared to controls [8].

Data from cattle studies also suggests that a pre-existing immune response resulting from prior exposure to environmental mycobacteria compromises immunity-generated by vaccination with BCG [12]. In a series of trials investigating BCG as a vaccine carried out over 3 years it was reported that in the first 2 years BCG gave a significant degree of protection against intratracheal challenge with virulent M. bovis, while in the third year it did not. In the calves used in the third year a higher IFN-γ response was evident to M. avium PPD prior to vaccination with BCG when compared to the two previous years. It was suggested that the weaker induction of cytokine responses seen in the animals that had relatively high prevaccination IFN-γ responses to avian PPD compared with animals in previous years may have resulted in restricted multiplication of BCG in vivo after vaccination.

Experimental investigations in rodents add further support for the concept that exposure to environmental mycobacteria affects immunity induced by BCG. Studies of M. tuberculosis in guinea pigs [6] suggested that exposure to environmental mycobacteria imparted a degree of resistance to TB and that BCG could only improve a little on this natural immunity. Thus, the i.d. inoculation of several different mycobacteria species gave a significant degree of protection against i.p. challenge with virulent M. tuberculosis and the immunity seen with BCG after exposure to environmental bacteria was not independent of the first exposure, but complemented it. Further experiments in guinea pigs [15] and mice [16] indicated that exposure to environmental bacterial could impart a significant degree of protection against respiratory challenge with M. tuberculosis. Supporting evidence for the concept that exposure to environmental mycobacteria protects against M. tuberculosis comes from studies of US navy recruits. These showed that higher skin test reactivity related to lower infection in the absence of vaccination. A general conclusion that has been made, taking these various studies and investigations into account, is that the degree of protection imparted by exposure to both environmental mycobacteria and to BCG together is neither additive nor multiplicative but equivalent to the maximum protection imparted by either alone [8].

The results noted here for calves are consistent with this last paradigm and not one in which exposure to environmental bacteria compromises the immunity induced by BCG by prevention of an immune response. Although a more rapid decline in the synthesis of IFN-γ following the incubation of cells with PPD-B may be evidence for a shorter duration of immunity in animals with a prior exposure to M. avium, this possibility needs to be addressed. It should also be noted that the immunity from environmental exposure is only partial; there is no claim that exposure will give total protection against virulent M. bovis or M. tuberculosis. However, it should be emphasized that immunity to TB in humans is considered as controlling the disease rather than eliminating infection. Furthermore, there is evidence that reinfection is possible after antibiotic treatment has been used to cure a first infection, indicating that after infection with a virulent strain a high level of protective immunity does not exist [17]. Thus, it has been suggested that it may not be possible to generate solid immunity unless new vaccination strategies are adopted [1].

It should be emphasized that the reported experiments were with a particular strain of M. avium, dose and route. Indeed, the dose of M. avium used induced a much less intense immune response that did BCG. Higher doses, or the use of a more virulent strain, might generate a more pronounced response. Studies in cattle suggest that animals that had naturally acquired a higher level of response to M. avium showed a relatively poor response to BCG vaccination and a lower level of immunity compared to animals with a lower level of naturally acquired response to M. avium tested on previous occasions [12,18]. Results from studies of BCG and M. tuberculosis in rodent models imply that different effects might also occur in cattle with different species of environmental Mycobacteria [6]. This could provide an alternative explanation for our failure to show a reduced level of immune resoponse in contrast to observation with naturally responsive animals. Furthermore, natural oral or respiratory infection may induce a different immune response compared to s.c. exposure and have different consequences for vaccination, although this route is likely to produce a more consistent effect in models with small groups of outbred calves. Thus, mice exposed to M. avium by the s.c. route were refractive to BCG vaccination by s.c. and intravenous routes [9] while mice with a pulmonary infection with M. avium were not refractive to intravenous BCG vaccination, as evidenced by susceptibility to aerogenic challenge with M. tuberculosis [19]. To define precisely how these variations affect immunity following BCG vaccination in calves experiments comparing different parameters, such as oral/respiratory infection with M. avium and dose, are necessary in the natural host. It is also recognized that although a type 1 immune response and the cytokines associated with it, particularly IFN-γ and IL-12 are necessary for immunity to mycobacteria IFN-γ levels, DTH responses and proliferation assays do not correlate clearly with immunity, being evident in humans that have been exposed but have controlled infection as well as those that have not [1,20]. Protection studies are necessary to determine precisely how infection with M. avium affects the ability of BCG to protect calves against TB.

There was evidence that the bias of the antigens recognized during the immune response following BCG vaccination was affected by prior exposure to M. avium. This may affect immunity to virulent M. bovis. It certainly affects the development of the DTH response that is used as a diagnostic test and the comparative levels of IFN-γ produced following incubation of blood with PPD-A and PPD-B. This indicates that animals with a significant response to PPD-A may be infected with M. bovis but not be diagnosed positively in the comparative skin test.


This work was supported by the BBSRC and DEFRA UK.


1. Kaufmann SH. How can immunology contribute to the control of tuberculosis. Nature Immunol Rev. 2001;1:20–30. [PubMed]
2. Collins HL, Kaufmann SH. The many faces of host responses to tuberculosis. Immunology. 2001;103:1–9. [PMC free article] [PubMed]
3. Bancroft GJ. The role of natural killer cells in innate resistance to infection. Curr Opin Immunol. 1993;5:503–10. [PubMed]
4. Wood PR, Corner LA, Rothel JS, Baldock C, et al. Field comparison of the interferon-gamma assay and the intradermal tuberculin test for the diagnosis of bovine tuberculosis. Aust Vet J. 1991;68:286–90. [PubMed]
5. Adams LG. In vivo and in vitro diagnosis of Mycobacterium bovis infection. Rev Sci Off Int Epiz. 2001;20:304–24. [PubMed]
6. Palmer CE, Long MW. Effects of infection with atypical mycobacteria on BCG vaccination and tuberculosis. Am Rev Resp Dis. 1966;94:553–68. [PubMed]
7. Stanford JL, Shield MJ, Rook GA. How environmental mycobacteria may predetermine the protective efficacy of BCG. Tubercle. 1981;62:55–62. [PubMed]
8. Fine PEM. Variation in protection by BCG. implications of and for heterologous immunity. Lancet. 1995;346:1339–45. [PubMed]
9. Brandt L, Cunha JF, Olsen AW, et al. Failure of the Mycobacterium bovis BCG vaccine: some species of environmental mycobacteria block multiplication of BCG and induction of protective immunity to tuberculosis. Infect Immun. 2002;70:672–8. [PMC free article] [PubMed]
10. Buddle BM, Keen D, Thomson A, et al. Protection of cattle from bovine tuberculosis by vaccination with BCG by the respiratory or subcutaneous route, but not by vaccination with killed Mycobacterium vaccae. Res Vet Sci. 1995;59:10–6. [PubMed]
11. Buddle BM, de Lisle GW, Pfeffer A, Aldwell FE. Immunological responses and protection against Mycobacterium bovis in calves vaccinated with a low dose of BCG. Vaccine. 1995;13:1123–30. [PubMed]
12. Buddle BM, Wards BJ, Aldwell FE, Collins DM, de Lisle GW. Influence of sensitisation to environmental mycobacteria on subsequent vaccination against bovine tuberculosis. Vaccine. 2002;20:1126–33. [PubMed]
13. Monaghan ML, Doherty ML, Collins JD, Kazda JF, Quinn PJ. The tuberculin test. Vet Microbiol. 1994;40:111–24. [PubMed]
14. Kwong LS, Hope JC, Thom ML, et al. Development of an ELISA for bovine IL-10. Vet Immunol Immunopathol. 2002;85:213–23. [PubMed]
15. Edwards ML, Goodrich JM, Muller D, Pollack A, Ziegler JE, Smith DW. Infection with Mycobacterium avium-intracellulare and the protective effects of Bacille Calmette–Guerin. J Infect Dis. 1982;145:733–41. [PubMed]
16. Orme IM, Collins FM. Efficacy of Mycobacterium bovis BCG vaccination in mice undergoing prior pulmonary infection with atypical mycobacteria. Infect Immun. 1984;44:28–32. [PMC free article] [PubMed]
17. van Rie A, Warren R, Richardson WR, et al. Exogenous reinfection as a cause of recurrent tuberculosis after curative treatment. N Engl J Med. 1999;341:1174–9. [PubMed]
18. Skinner MA, Wedlock DN, Buddle BM. Vaccination of animals against Mycobacterium bovis. Rev Sci Techn Off Epiz. 2001;20:112–32. [PubMed]
19. Orme IM, Collins FM. Efficacy of Mycobacterium bovis BCG vaccination in mice undergoing prior pulmonary infection with atypical mycobacteria. Infect Immun. 1984;44:28–32. [PMC free article] [PubMed]
20. Ellner JJ, Hirsch CS, Whalen CC. Correlates of protective immunity to Mycobacterium tuberculosis in humans. Clin Infect Dis. 2000;30:S279–82. [PubMed]

Articles from Clinical and Experimental Immunology are provided here courtesy of British Society for Immunology
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...