Lepr Rev (2002) 73, 254±261

Anti-leprosy protective vaccination of rhesus monkeys with BCG or BCG plus heat-killed Mycobacterium leprae: skin test results

B. J. GORMUS*, G. B. BASKIN**, K. XU*, M.S.RATTERREE***,P.A.MACK***, R. P. BOHM, JR.***, W. M. MEYERS‡ & G. P. WALSH‡‡ Departments of Microbiology*, ** and Veterinary ***, Tulane Regional Primate Research Center, 18703 Three Rivers Road, Covington, LA 70433 ‡The Armed Forces Institute of Pathology, Washington, DC and ‡‡American Leprosy Foundation, Rockville, MD, USA

Accepted for publication 9 April 2002

Summary Groups of rhesus monkeys (RM) were vaccinated and boosted with living Mycobacterium bovis Bacillus Calmette-Guerin (BCG) or BCG ‡ low dose (LD) heat-killed Mycobacterium leprae (HKML) or high dose (HD) HKML or were unvaccinated. Animals vaccinated with BCG ‡ LD and HD HKML were lepromin skin tested 2 weeks after boosting. All groups were lepromin tested 37 and 46 months after challenge with live M. leprae. Fernandez (72 h) and Mitsuda (28 day) responses were recorded. Ten of 10 rhesus monkeys in each of the two BCG ‡ HKML- vaccinated groups signi®cantly converted to strong positive Fernandez status within 2 weeks of boosting, compared to one of six positives in the unvaccinated unchallenged normal control group. Both BCG ‡ HKML groups were signi®cantly protected from clinical leprosy. Six of 10 in each of the two BCG ‡ HKML groups signi®cantly converted to Mitsuda positivity within 2 weeks of boosting compared to zero of six in the normal control group. The sizes of the Mitsuda responses were larger in the LD group than the HD HKML vaccinated/boosted group, suggesting suppression by vaccination with higher doses of HKML in combination with BCG. Fernandez responses were negative in normal RM as well as in the unvaccinated, ML-challenged group and the BCG-vaccinated, ML-challenged group at 37 or 46 months after ML inoculation, although the BCG-vaccinated group was signi®cantly protected from leprosy and the unvaccinated group was not. In contrast, at 37 months the Fernandez reaction was positive in the BCG plus LD and the BCG plus HD HKML-vaccinated groups, both of which were signi®cantly protected from clinical leprosy. By 46 months, the Fernandez responses were below signi®cance in all groups. Thus, Fernandez reactivity is not a reliable correlate to protection from experimental leprosy in RM. Mitsuda responses became strongly positive in all four

Correspondence to: B. J. Gormus, Tulane Regional Primate Research Center, 18703 Three Rivers Road, Covington, LA 70433, USA; e-mail: [email protected]

254 0305-7518/02/073254+08 $1.00 q Lepra Protective vaccination of monkeys 255

ML-challenged groups by 37 months and remained strongly positive at 46 months after ML inoculation, suggesting that strong Mitsuda reactivity re¯ects responses to living ML. BCG or BCG ‡ LD or HD HKML vaccination/boosting of RM produced signi®cant clinical protection from leprosy and there was a good correlation between protection from LL forms of leprosy and positive Mitsuda skin test responses after challenge with live ML. Positive Mitsuda responses were generated in essentially all individuals after challenge with live ML, and this response was primed by prior vaccination/boosting with BCG ‡ HKML as shown by conversion to positivity 2 weeks after boosting. The data show that resistance to clinical leprosy is re¯ected by Mitsuda responses in ML-exposed RM, similar to results from human studies, and con®rm the suitability of RM as a model for leprosy vaccine studies.

Introduction

The precise relationship between anti-mycobacterial immunity and delayed-type hyper- sensitivity (DTH) responses is unclear. In leprosy, this relationship is further complicated by the existence of an early (Fernandez) and a late (Mitsuda) response to lepromin. The Fernandez response is measured 48±72 h after skin sensitization and is regarded as a true DTH response. A positive Fernandez response is, therefore, indicative of a pre-existing sensitivity to Mycobacterium leprae (ML) or to cross-reacting mycobacterial antigens. The Mitsuda response, measured at 21±28 days, is regarded as a granulomatous response secondary to sensitization by the lepromin skin test antigen and appears to re¯ect the potential to generate cell-mediated immunity (CMI) to ML antigens. Mitsuda responses require the presence of intact bacilli, whereas the Fernandez reaction requires either intact bacilli or antigen extracts. The Fernandez response may be negative in normal individuals or positive due to cross-reacting antigens of environmental mycobacteria. The Mitsuda response, however, is positive in most normal individuals.1 In ML exposed or infected individuals, a Fernandez response can be positive or nega- tive together with either a strong, weak or absent Mitsuda response. Strong Mitsuda can be indicative of sub-clinical leprosy exposure or leprosy near the tuberculoid end of the leprosy spectrum, whereas weak or absent Mitsuda responses in ML exposed individuals suggest leprosy risk nearer the lepromatous end of the spectrum.1,2 In spite of these generalizations, the exact relationships between Fernandez or Mitsuda reactions, disease susceptibility in ML exposed individuals and immunity to leprosy are not well de®ned. One advantage of using animal models to study leprosy is that one can follow changes in immune parameters before and at intervals after inoculation with ML, which cannot be done with leprosy patients. Another advantage is that one can be con®dent that these animals are not harboring sub-clinical ML infections prior to inoculation. In the present study, we have followed Fernandez and Mitsuda responses in groups of rhesus monkeys (RM) beginning after vaccination and boosting with BCG or BCG plus heat-killed ML (HKML) and at intervals after subsequent challenge with live ML. In parallel, we followed the clinical progress of leprosy, changes in immunological parameters and changes in Fernandez and Mitsuda status in these groups of monkeys with time. The clinical and immunological results were published previously.3,4 This approach permits the observation of associations between the different activities studied in relation to disease progress. The results are valuable in formulating testable hypotheses concerning relationships between immune activities, DTH responses and disease susceptibility using a nonhuman primate model. 256 B. J. Gormus et al. We have previously demonstrated the utility of the RM model in the study of leprosy and in studying the effectiveness of potential vaccines.3±6 In most manifestations, RMs are similar to humans in their susceptibility and reactions to infection with ML.3,4,6 Acaseof naturally-acquired leprosy in a Macaque has been reported.7 One point of difference between RM and humans is the requirement for approximately 10-fold higher doses of lepromin to elicit positive skin test responses.5,6 Moreover, normal RM generally fail to give a positive Mitsuda response,5 whereas most normal humans produce a positive response to lepromin whether they are from a leprosy endemic area or not.1 After infection with ML, however, responses of RM to 10 lepromin appear to mimic those of humans described above.5,6 Thus, lepromin skin testing of ML infected RM can provide information similar to that obtained in humans.

Materials and methods

ANIMALS The methods and procedures utilized herein have been previously described.3,4,6,8±10 Brie¯y, 40 Chinese RM (Macaca mulatta) were divided into four experimental groups of 10 (three vaccine groups and one unvaccinated control group, all ML challenged). There were three females and seven males/group. There was also one group of six unvaccinated, non-ML challenged normal controls.

IMMUNIZATIONS WITH BCG OR BCG PLUS HKML AND MONKEY INOCULATIONS These methods have been reported in detail previously.3,4 Monkeys were vaccinated with BCG alone (1±2´6 ´ 106 viable units) or BCG ‡ high-dose (HD) HKML (3´2 ´ 109)or BCG ‡ low-dose (LD) HKML (1´6 ´ 109) by intra-cutaneous (IC) injection of 0´1mlofthe appropriate vaccine. Primary vaccinations were followed by boosting at approximately 7 weeks and challenge with live ML at 10 weeks (4´3 ´ 109 total ML). Monkeys were inoculated with live, freshly prepared ML suspensions by combined IC (1´7 ´ 109, distributed among nine sites) and intravenous (IV) (2´6 ´ 109) routes. Animals were observed daily and examined in detail 3±4times per year or more, depending on the status of the animal, and the clinical aspects of the disease were recorded. Biopsies were taken at intervals from dermal inoculation sites and/or from sites of dissemination for analysis of the lesions and for documentation of clinical leprosy, as previously reported.3,5,6 Criteria for the diagnosis and classi®cation of leprosy in RM were described previously.3,5,6,8,10,11

LEPROMIN SKIN TESTING Brie¯y, abdominal areas were shaved and 0´1 ml of lepromin was injected IC, according to our published procedures.3,5,6 Fernandez reactions were read at 72 h and Mitsuda responses were read on day 28. Lepromin was of sooty mangabey monkey origin after inoculation of and re-isolation from armadillos by Drs Wayne M. Meyers and Gerald P. Walsh at the Armed Forces Institute of Pathology, Washington, D.C. At each skin test time, the Fernandez and the Mitsuda results were read; sites were biopsied for histopathology at 28 days post- inoculation (PI). Protective vaccination of monkeys 257

STATISTICAL ANALYSES All statistical calculations were performed using statistical programs for the Macintosh Computer. The results of skin testing were analysed by Fisher's Exact (chi square) method to compare numbers of positives in various groups and by the Kruskal-Wallis ANOVA by ranks method.

Results

The groups of RM that were vaccinated with BCG plus LD or HD HKML were skin tested with lepromin 2 weeks after boosting to determine if the vaccination was effective. The unvaccinated group and the group vaccinated with BCG alone were not lepromin skin tested at this time, to avoid introduction of ML antigens and possible confusing vaccine effects. The Fernandez results 2 weeks PI are shown in Table 1. All 10 RM in each of the two groups signi®cantly converted to strong positive Fernandez status within 2 weeks of boosting compared to one of six positives in the unvaccinated unchallenged normal control group. Mitsuda results from testing 2 weeks PI are shown in Table 2. Six of 10 in each of the two groups signi®cantly converted to Mitsuda positivity within 2 weeks of boosting compared to zero of six in the normal control group. The sizes of the Mitsuda responses were larger in the LD group than the HD HKML vaccinated/boosted group (ANOVA, P <0´01). Lepromin skin testing was undertaken in all four experimental (ML-challenged) groups and a normal control group at 37 and 46 months after live ML challenge. Compared to normal control RM, Fernandez responses (Table 3) were not signi®cantly different in the unvacci- nated, ML-challenged group or the BCG-vaccinated, ML-challenged group at 37 or 46 months PI. In contrast, at 37 months the Fernandez reaction was signi®cantly positive in both the BCG plus LD and the BCG plus HD HKML-vaccinated groups. By 46 months PI, the Fernandez responses in the BCG ‡ LD and BCG ‡ HD HKML vaccinated groups were no longer signi®cantly different from the unvaccinated control group. Mitsuda responses became strongly positive in all four ML-challenged groups by 37 months and remained strongly positive at 46 months PI (Table 4).

Discussion

Signi®cant positive Fernandez conversion by all 10 RM within 2 weeks of boosting with BCG plus LD or HD HKML demonstrates that the vaccines were effective in eliciting a DTH

Table 1. Fernandez lepromin skin test results post-boosting1

Group Average size (mm) Range (mm) No. positive/no. tested P2

BCG ‡ LDHKML 12´2 5±17 10/10 <0´001 BCG ‡ HDHKML 11´2 6±19 10/10 <0´001 Normal controls 2´7 1±5 1/6 ±

172 h observations of skin test sites 2 weeks post-boosting. 2Fisher's exact method versus normal control group. 258 B. J. Gormus et al.

Table 2. Mitsuda lepromin skin test results post-boosting1

Group Average size (mm) Range (mm) No. positive/no. tested P2

BCG ‡ LDHKML 7´5 0±21 6/10 <0´03 BCG ‡ HDHKML 3´6 0±7 6/10 <0´03 Normal controls 0´5 0±1 0/6 ±

128 day observations of skin test sites 2 weeks post-boosting. 2Fisher's exact method versus normal control group. response. This result correlates with the fact that there was highly signi®cant protection in each of these two groups (89% in BCG ‡ LD HKML and 78% in BCG ‡ HD HKML)3 from clinical leprosy after subsequent challenge with live ML. It also correlates with the observed stimulation of signi®cant in vitro blastogenic responses previously reported in these two groups after vaccination with BCG ‡ LD or HD HKML. These powerful, pre- viously reported, lepromin blastogenic responses (SI ˆ 28 and 13, respectively) were short- lived, however, returning to baseline after boosting with BCG + HKML and remaining close to baseline after live ML challenge.4 Smaller signi®cant in vitro responses by blood MNC to lepromin (SIs approaching 5) were observed in the BCG ‡ LD and BCG ‡ HD HKML groups by 24 months post-live ML challenge,4 suggesting that an immunomodulating mechanism post-boosting had dampened the systemic blastogenic response to ML antigens induced by vaccination with BCG ‡ HKML. Similar immunomodulating effects were not observed in the BCG ‡ HKML-vaccinated groups with regard to the DTH reaction. Fernandez reactions remained positive in both BCG ‡ LD and HD HKML groups 37 months after ML inoculation, but by 46 months PI

Table 3. Fernandez skin test results post-challenge

Group No. positive1/Total P2

37 months PI Normal controls 1/6 ML-challenged Unvaccinated 4/10 NS BCG only 4/10 NS BCG ‡ LDHKML 8/93 <0´01 BCG ‡ HDHKML 7/93 <0´04 46 months PI Normal controls 1/6 ML-challenged Unvaccinated 1/83 NS BCG only 3/83 NS BCG ‡ LDHKML 6/93 NS BCG ‡ HDHKML 5/93 NS

1Positive by WHO criteria (5 mm or greater indura- tions). 2Fisher's exact method compared to normal controls. 3By 37±46 months PI, several animals had died, reducing the total number/group. Protective vaccination of monkeys 259

Table 4. Mitsuda skin test results post-challenge

Group No. positive1/total P2

37 months PI Normal controls 0/6 ML-challenged Unvaccinated 10/10 <0´0001 BCGonly 10/10 <0´0001 BCG ‡ LDHKML 9/93 <0´0002 BCG ‡ HDHKML 9/93 <0´0002 46 months PI Normal controls 0/6 ML-challenged Unvaccinated 8/83 <0´0003 BCGonly 7/8 3 <0´002 BCG ‡ LDHKML 9/93 <0´0002 BCG ‡ HDHKML 9/93 <0´0002

1Positive by WHO criteria (5 mm or greater indura- tions). 2Fisher's exact method compared to normal controls. 3By 37±46 months PI, several animals had died, reducing the total number/group.

Fernandez reactivity appeared to wane, as evidenced by the declining numbers of Fernandez- positive animals. This suggests that the DTH response to lepromin induced by BCG ‡ HKML is long-lived, but may have a ®nite existence. Fernandez reactions in the unvaccinated, ML-challenged and the BCG-only vaccinated, ML-challenged groups were not signi®cantly different from unvaccinated, unchallenged normal control RM at 37 and 46 months PI, although the BCGgroup was highly protected (70%) from clinical leprosy, but the unvaccinated, ML challenged control group was not protected.3 These results show that con- version to Fernandez positivity can correlate with protection from leprosy, but that protection can occur in the absence of a Fernandez conversion, in accord with the results from leprosy patients.1,12,13 Both Fernandez reactions and blastogenic responses to lepromin can be misleading indicators of leprosy susceptibility or infection, and the nature of any possible mechanistic relationships between these two activities is not apparent from these studies. Six of 10 animals in each BCG ‡ HKML group converted to Mitsuda positivity 2 weeks post-boosting. Moreover, the sizes of the Mitsuda reactions were dose-dependent, with BCG ‡ LD HKML-vaccinated and boosted animals giving the largest reactions compared to the group vaccinated and boosted with BCG ‡ HD HKML. Thus, immunization/boosting with higher doses of HKML in combination with BCGappears to suppress the Mitsuda response prior to challenge with live ML. By 37 months PI, Mitsuda reactions were strongly positive in 100% of the RM in each of the four ML-challenged groups, including the unvaccinated and the BCG-only-vaccinated group. These latter two groups had never been exposed to ML other than living ML in the challenge inoculation. Mitsuda results were virtually unchanged at 46 months PI. These observations are consistent with suggestions that persisting Mitsuda positivity is due to prior exposure to living ML and with the observation that the majority of animals developing leprosy in this study had paucibacillary forms and/or other indications of some degree of 260 B. J. Gormus et al. resistance to clinical leprosy, as discussed below. The fact that six of 10 BCG ‡ LD or HD HKML animals converted to Mitsuda positivity 2 weeks post-boosting, prior to inoculation with live ML, is most likely the result of inoculation with living BCG together with cross- reacting killed ML antigens. In spite of the strong persisting positive Mitsuda conversions in live ML-challenged unvaccinated and BCG-only-vaccinated ML-challenged RM, nine of 10 unvaccinated and three of 10 BCG-only-vaccinated animals developed leprosy.3 Only two of the nine unvaccinated leprosy positives had multibacillary (MB) forms of leprosy, however, and none of the three BCG-only vaccinated leprosy positive RMhad MBleprosy. 3 A total of three of 20 RMin the BCG ‡ LD and BCG ‡ HD HKML groups developed leprosy.3 Of these three, all became positive for live ML in nasal secretions, indicative of ML dissemination, but no gross evidence of MB leprosy dissemination to uninoculated dermal sites were observed. Two of the three leprosy positive BCG ‡ HKML vaccinated animals developed MB or neuritic MB leprosy, respectively, at dermal inoculation sites.3 More- over, as previously reported, there was no detectable PGL-I antigen in the plasma of RMin any group in this study.3 These observations suggest that there was some degree of resistance to leprosy among all groups of ML-challenged RM, consistent with the observation of continuing Mitsuda positivity for at least 46 months after live ML challenge in all four groups. These data are consistent with our prior reports that RMs as a species are more resistant to leprosy and appear to be more similar to humans in this regard than sooty mangabey monkeys.3;4;6;14 The results suggest that lepromin skin testing in the rhesus monkey model is of similar value to that in human leprosy studies.1;2;12;13;15;16

Acknowledgements

Support for this study was provided by grants #AI-19302 from NIAID and #RR-00164 from the NCRR. We acknowledge the expert technical assistance of the following persons: Mss Cynthia Trygg, Carolyn Coyne, Doris O'Leary and Renee Grow. We thank Ms Ann Bennett for secretarial contributions.

References

1 Harboe, M. Leprosy, 2nd edn. Churchill Livingstone, London, 1994, pp 87±112. 2 Mitsuda K. On the value of a skin reaction to a suspension of leprous nodules. Int J Lepr Other Mycobact Dis, 1953; 21: 347±358. 3 Gormus BJ, Baskin GB, Xu K et al. Protective immunization of monkeys with BCG or BCG plus heat-killed Mycobacterium leprae: clinical results. Lepr Rev, 1998; 69: 6±23. 4 Gormus BJ, Baskin GB, Xu K et al. Antileprosy protective vaccination of rhesus monkeys with BCG or BCG plus heat-killed Mycobacterium leprae: immunologic observations. Int J Lepr Other Mycobact Dis, 2000; 39: 27±39. 5 Baskin GB, Gormus BJ, Martin LN et al. The lepromin test in rhesus monkeys. Int J Lepr, 1986; 54: 427±436. 6 Gormus BJ, Xu K, Baskin GB et al. Experimental leprosy in rhesus monkeys: transmission, susceptibility, clinical and immunological ®ndings. Lepr Rev, 1998; 69: 235±245. 7 Valverde CR, Can®eld D, Tarara R et al. Spontaneous leprosy in a wild-caught cynomolgus macaque. Int J Lepr Other Mycobact Dis, 1998; 66: 140±148. 88Gormus BJ, Murphey-Corb M, Martin LN et al. Impaired responses to Mycobacterium leprae antigens in rhesus monkeys experimentally inoculated with simian immunode®ciency virus and M. leprae. Lepr Rev 1998; 69: 24±39. 9 Gormus BJ, Ohashi DK, Ohkawa S et al. Serologic responses to Mycobacterium leprae-speci®c phenolic Protective vaccination of monkeys 261

glycolipid-I antigen in sooty mangabey monkeys with experimental leprosy. Int J Lepr Other Mycobact Dis, 1988; 56: 537±545. 10 Gormus BJ, Murphey-Corb M, Baskin GB et al. Interactions between Mycobacterium leprae and simian immunode®ciency virus (SIV) in rhesus monkeys. J Med Primatol, 2000; 29: 259±267. 11 Gormus BJ, Murphey-Corb M, Martin LN et al. Interactions between simian immunode®ciency virus and Mycobacterium leprae in experimentally inoculated rhesus monkeys. J Infect Dis, 1989; 160: 405±413. 12 Sharma P, Kar HK, Kaur H et al. Induction of lepromin positivity and immunoprophylaxis in household contacts of multibacillary leprosy patients: a pilot study with a candidate vaccine, Mycobacterium w. Int J Lepr Other Mycobact Dis, 2000; 68: 136±142. 13 Chaudhury S, Hazra SK, Saha B et al. An eight-year ®eld trial on antileprosy vaccines among high-risk household contacts in the Calcutta metropolis. Int J Lepr Other Mycobact Dis, 1994; 62: 389±394. 14 Gormus BJ, Xu K, Baskin GB et al. Experimental leprosy in monkeys. I. Sooty mangabey monkeys: transmission, susceptibility, clinical and pathological ®ndings. Lepr Rev, 1995; 66: 96±104. 15 Convit J, Sampson C, Zuniga M et al. Immunoprophylactic trial with combined Mycobacterium leprae/ BCG vaccine against leprosy: preliminary results. Lancet, 1992; 339: 446±450. 16 Rees RJW. The signi®cance of the lepromin reaction in man. Int J Lepr Other Mycobact Dis, 1964; 8: 224±258.