Skin testing with recombinant Mycobacterium intracellulare antigens

Tuber&

and Lung LXsmse

(1992) 13,129-133

Skin testing with recombinant Mycobacterium intracellulare antigens S. L. Morris, J. C. Mackall, A. Malik, D. A. Rouse, S. D. Chaparas Center for Biologics Evaluation

and Research Food and Drug Administration

Bethesda, Maryland,

USA

S U MM A R Y. The hnmunoreactivity of four recombinant Mycobacterium intracellulure P-galactosidase fusion proteins, which correspond to 22, 40, 43 and 85 kDa M. intracellulure antigens, was assessed. Lymphoproliferative assays demonstrated that Escherichiu coli lysates containing each of the fusion proteins stimulated T cells in vitro. Purified preparations of three of these recombinant M. intruceUulure antigens (22,43 and 85 kDa) also induced delayed-type hypersensitivity (DTH) reactions in sensitized guinea pigs. However, the skin test responses evoked by each of these antigens was not species-specific. Given these results, the potential utility as skin test reagents of the purified antigens or peptides derived from these proteins is discussed. R l? S U M I?. L’immunori?activit6 de 4 proteines de fusion P-galactosidase Mycobacterium intracellulure obtenus par recombinaison (correspondant a des antigenes M. intracellulare de 22,40,43 et 85 Kilodaltons @Da)) a Cte Cvaluee. Des dosages lymphoproliferatifs ont dCmontr& que des lysats d’Escherichia coli contenant chacune des proteines de fusion ont stimule des cellules T in vitro. Des preparations purifi&s de 3 de ces antigenes M. intruceZZuZure obtenues par recombinaison (22, 43 et 85 kDa) ont induit Cgalement des reactions d’hypersensibilite de type differ6 (HTD) chez des cobayes sensibilises. NCamnoins, les reponses aux tests cutan& provoquees par chacun de ces antigenes n’ont pas temoigne de @chicit d’esp&ces. En fonction de ces resultats se discute l’utilite potentielle d’antigenes ou peptides purifiCs derives de ces proteines comme reactifs dans les tests cuta&. R ES U M E N . Se estudio la immunorreactividad de 4 proteinas recombinantes por fusion, beta-galactosidasa de M. intracellulare, correspondientes a antigenos de M. intracellulare de 22, 40,43 y 85 kilodaltons (kDa). Los tests linfoproliferativos demostraron que 10s lisados de Escherichiu coli que contenian cada una de las proteinas de fusion mencionadas eran capaces de estimular in vitro a 10s linfocitos T. Las preparaciones purificadas de 3 de estos antigenos recombinantes de M. intrucellulare (22, 43 y 85 kDa) tambien eran capaces de inducir reacciones de hipersensibilidad de tipo retardada en cobayos sensibilizados. Sin embargo, las respuestas a 10s tests cut6neos por cada uno de estos antigenos no evocaban una especificidad de especie. Debido a estos resultados, se discute la utilidad potential, coma reactivos de tests cutaneos, de 10s antigenos purificados o de peptidos derivados de estas proteinas.

tuberculosis because the clinical and radiologic manifestations of the two diseases are very similar.5 The different clinical courses, responses to chemotherapy and prognosis of M, tuberculosis and MAC infections emphasize the need to develop better methods to precisely identify the infecting mycobacterial species. Skin testing done with the tuberculin and Battey (developed from the Boone Battey strain of M. intracellulare) preparations has been an invaluable methodology in the diagnosis of mycobacterial disease and for studying mycobacterial epidemiology.h,7 The Center for

In the past decade, the rate of Mycobacterium tuberculosis and Mycobacterium avium, Mycobacterium intracellulare complex (MAC) infections in the USA has increased dramatically primarily because of the AIDS epidemic.‘” Hospital-based estimates suggest that about 10% of American AIDS patients will develop tuberculosis and 50% will develop MAC disease.4 Protocols that permit the rapid detection of MAC disease and allow the rapid differentiation between M. tuberculosis and MAC infections are currently unavailable. In fact, many patients with MAC infections have an initial diagnosis of

Correspondence to: Sheldon L. Morris, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, Maryland 20892, USA. 129

130

Tubercle and Lune Disease

Disease Control has recently recommended skin testing for the detection of mycobacterial infections even in AIDS patients where HIV infections can depress tuberculin responsiveness.’ PPD skin testing in conjunction with delayed-type hypersensitivity (DTH) anergy testing is advised for immunocompromised individuals. However, testing with the current tuberculin and Battey skin test reagents is problematic because these preparations do not elicit monospecific responses. Both of these reagents give significant DTH reactions to heterologous mycobacterial infections.9x10 This cross-reactivity can confound interpretation of patient skin test responses and delay proper drug therapy. The availability of monospecific specific skin test reagents for detecting and differentiating M. tuberculosis and MAC infections would, therefore, greatly improve clinical management of mycobacterial diseases. The wide experience reported with the Battey antigen has prompted us to examine DNA from the source strain for genes encoding antigens which may be useful in the preparation of monospecific skin test reagents. We have previously described the isolation of five recombinant bacteriophages from a A4. intracellulare hgtll gene library which express M. intracellulare antigens in Escherichia coli.“S’2 In this report, we have evaluated the immunoreactivities of four of these recombinant proteins in vitro with T cell proliferation assays and in vivo with skin tests. We demonstrate that these purified mycobacterial proteins elicit moderate nonspecific skin test reactions. Given these results, the possible utility of these antigens or peptides derived from these antigens as components of monospecific diagnostic reagents is discussed.

Production of recombinant mycobacterial proteins Recombinant bacteriophages expressing M. intraP-galactosidase fusion proteins were isolated and purified as described previously.“~‘* Lambda E. lysogens of these bacteriophages were generated in coli strain Y1089. The recombinant mycobacterial proteins were then overproduced in the h lysogens as described by Huynh et a1.13After the induction protocol, the cells were frozen at -70°C and thawed quickly at 37°C. The lysates were sonicated for 1 min to reduce viscosity. Finally, cellular proteins were precipitated with an equal volume of ammonium sulfate. The protein concentrations were determined using the Bio-Rad protein assay (Bio-Rad Laboratories, Richmond, CA). cellulare

Lymphoproliferative assays for assessing the immunoreactivity of recombinant antigens of the recombinant antigens assayed in vitro with

Skin testing of purified recombinant antigens The mycobacterial P-galactosidase fusion proteins were purified to a single spectrophotometric peak by gel permeation high performance liquid chromatography (HPLC) as reported previously.‘4 To prepare animals for skin testing, Hartley guinea pigs were sensitized with either M. avium (serovar 4 strain isolated from an AIDS patient), M. intracellular-e (Battey strain), or M. tuberculosis (Erdman strain) as described earlier.18 After a minimum of 6 weeks post initial immunization, the guinea pigs were given intradermal injections of 3 p.g of each recombinant antigen dissolved in 100 pl of 0.85% sodium chloride, 0.001% Tween 80. As positive skin test controls, 0.1 pg of homologous antigen preparations [M. avium (serovar 4 AIDS-associated strain), M. intracellulare (Battey strain) sonic extract or tuberculin] were injected into the appropriate animals. Analogous skin tests were done in control guinea pigs which had not been sensitized with mycobacterial extracts. The size of the resulting erythemas were measured 18-24 h after injection.

RESULTS Expression of recombinant M. intracellulure fusion proteins in E. coli

MATERIALS AND METHODS

The capacity cells was

intracellulare T cell line. The lymphocyte proliferation assay used in these experiments has been described in a previous report. I4 The data was statistically analyzed using the t-test comparisons of the Stat View 512 Macintosh computer program (Brain Power, Inc., Calabasas, CA).

to activate T an anti-M.

Recombinant bacteriophages expressing immunoreactive mycobacterial proteins were isolated from a M. intracellulare gene library using anti-M. intracellulare monoclonal antibodies or an absorbed polyclonal antiM. intracellular-e serum.“‘* The monoclonal antibodies (MAbs) 7F7, 6B8, and IE5 recognized M. intracellulare proteins of 22, 40, and 43 kDa, respectively. The polyclonal serum, which was absorbed with M. tuberculosis, M. kansasii, and M. fortuitum sonic extracts, bound a hybrid protein derived from an 85 kDa M. intracellulare polypeptide. To increase expression of the recombinant antigens, 3L lysogens were generated by infecting E. coli strain Y 1089 with recombinant bacteriophages expressing the immunoreactive proteins. Figure a Coomassie blue stained SDS 1A shows polyacrylamide gel electrophoresis (SDS-PAGE) gel containing ammonium sulfate precipitates of h lysogen sonic extracts. Each of the recombinant h lysogen extracts (lanes l-4) has an overexpressed high molecular weight protein which is not seen in the E. coli control (lane C). The h lysogen extracts were also characterized by

Skin testing with recombinant

kDa

C

1234

mycobacterium

intracellulare

antigens

fusion proteins, we also reacted nitrocellulose strips containing the h lysogens with anti-a-galactosidase antibody (data not shown). In these immunoblot assays, each recombinant antigen was bound by the anti-pgalactosidase antibody. Therefore, four different recombinant P-galactosidase fusion proteins derived from four different M. intracellulare antigen genes (MI 22, MI 40, MI 43, and MI 85) have been overexpressed in E. coli.

llO-

84-

Lymphoproliferative activity of the recombinant intracellulare proteins

3324A

/

-110 -mm

Yfimhm

13 I

7

. -33 i - 24

M.

To identify potential immunodiagnostic antigens, the fusion proteins were initially tested for their capacity to stimulate T cells in vitro. The proliferative stimulation of an anti-M. intracellulare T cell line by increasing doses of ammonium sulfate precipitates of h lysogen extracts containing the recombinant antigens is seen in Figure 2. The dose response curves demonstrate that the lymphoproliferative activity of each of the recombinant lysates is significantly greater than the E. coli control assayed at analogous concentrations. The largest responses for each of the recombinant lysates were seen at 5-10 pg of total cellular protein. At these concentrations, the maximum relative stimulation increases of the recombinant antigens compared to the E. coli control were 9.8 for MI 8.5, 6.5 for MI 43, 2.7 for MI 22, and 2.5 for MI 40. Statistical analyses of the thymidine uptake data indicated that the maximal T cell stimulation determined for each of the recombinant lysates was significantly greater (P < 0.05) than the E. coli control. Suppression of T cell activation was apparent when the T cell line was challenged with 20 l.tg of protein. Inhibition of T cell stimulation by high concentrations of E. coli

14000 12000

10000

6000

B Fig l-Expression of recombinant M. infrucellulure P-galactosidase fusion proteins in E. coli lysogens. (A) SDS-PAGE of ammonium sulfate precipitates of h lysogen sonic extracts. Lane C is the E.coli control. Lanes 14 are ammonium sulfate precipitates of h lysogen sonic extracts containing M140( l), M143(2), M185(3) and M122(4). (B) Immunoblot analyses of h lysogen ammonium sulfate precipitates. The 1, lysogen sonic extracts shown in Figure 1A blotted against their corresponding antibody: lane 1 = MAb 6B8, lane 2 = MAb lE5, lane 3 = absorbed polyclonal anti-M. intracellulare serum, and lane 4 = MAb 7F7.

6000

4000

2000

0

Antigen

immunoblot analyses. A single high molecular weight reactive band was identified in each Western blot (Fig. 1B). The electrophoretic mobilities of the antibodyreactive proteins and overexpressed proteins seen by Coomassie blue staining were identical. Since the hgtl 1 expression system is designed to yield P-galactosidase

(ug)

Fig %-The stimulation of an anti-M. intracellulare T cell line by recombinant M. intracellulare B-galactosidase fusion proteins. The ‘T cell proliferation assay protocol has been described previously.” Ammonium sulfate precipitates of E. coli lysates containing the recombinant antigens - M 122 (t), M 140 (-o_), M I43 (ix) and Ml85 (4) -were tested. Lysates from the host strain E. coli Y 1089 (a served as controls. The error bars represent the standard error for each antigen dose.

132

Tubercle and Lung Disease

lysates has been previously described.‘* As a control, the anti-M. intrucellulure T cell line was challenged with 3 pg of purified P-galactosidase. No stimulatory response to P-galactosidase was demonstrated.

reactive reaction in guinea pigs sensitized with M. avium (6 mm) and M. tuberculosis (6 mm).

DISCUSSION Induction of delayed-type hypersensitivity reactions in guinea pigs using M. intracelldare recombinant proteins The T cell line proliferation studies suggested that each of the recombinant M. intracellulare antigens had the capacity to evoke T cell responses in vitro. To evaluate their in vivo cell-mediated reactivity, these antigens were skin tested in sensitized guinea pigs. Prior to skin testing, the antigens were purified to a single peak by high performance liquid chromatography as described in the Methods section. Three micrograms of each purified antigen were then injected into guinea pigs sensitized to M. avium, M. intracellulare, and M tuberculosis as well as non-sensitized controls. To establish that the guinea pigs had been properly sensitized, the animals were also challenged with 0.1 pg of homologous antigen preparation. The size of the erythemas resulting from the antigenic challenge were measured 24 h later. Purified recombinant antigen preparations did not elicit significant skin test responses in non-sensitized guinea pigs (Table). Purified P-galactosidase also did not induce a DTH response in control or sensitized animals (data not shown). Challenge with homologous sonic extract did evoke a strong DTH response in sensitized animals (11-12 mm). Furthermore, although one of the recombinant M intrucellulure antigens (MI 40) did not give a significant DTH response in sensitized guinea pigs, the other three recombinant mycobacterial proteins did elicit skin test responses. However, the skin test responses evoked by the purified antigens were crossreactive and of weak or moderate intensity. The MI 43 antigen evoked a weak reaction in M. intracellulare (5 mm), M. avium (4 mm), and M. tuberculosis sensitized animals (4 mm). The antigen MI 22 was moderately active in M. intrucellulure (7 mm) and M. tuberculosis (9 mm) sensitized animals. MI 85 evoked a moderate 9 mm response in guinea pigs sensitized with M. intracellulure. MI 85 also elicited a weaker cross-

Table. Shin test reactivities Sensitizing extract

For decades, mycobacteriologists have attempted to isolate antigens capable of eliciting species-specific mycobacterial skin test reactions. Despite many years of effort, this goal has not been achieved. Several recent studies with purified recombinant antigens have suggested that isolated proteins, native or recombinant, may have limited utility as monospecific skin test reagents.15-17In this report, we have extended these conclusions. We have shown that purified M. intracellulare recombinant antigens derived from 22, 40, 43 and 85 kDa MAC proteins will not be satisfactory as skin test antigens in their present form. Thus far, none of the purified mycobacterial proteins which have been tested for delayed-type hypersensitivity have yielded specific cellular immune responses. Even purified protein molecules which contain specific B cell epitopes may have cross-reactive T cell determinants and elicit non-specific skin test responses. For example, although the 38 kDa and 14 kDa M. tuberculosis proteins have B cell specific epitopes, these antigens evoke crossreactive T cell responses.‘5*‘7Y’8 Our study confirms these observations. Our purified recombinant proteins did not elicit monospecific DTH responses in sensitized guinea pigs. Even MI 43, which has a MAC specific B cell epitope, evoked weak non-specific T cell responses in vivo. These non-specific DTH responses were not unexpected because previous skin tests with mycobacterial sonicates had shown extensive skin test crossreactivity among mycobacterial species. In these studies, M. intrucellulure sonicates elicited moderate DTH reactions in M, tuberculosis sensitized guinea pigs and strong reactions in animals immunized with M. avium.lo Most of the purified native or recombinant mycobacterial proteins that have been assayed, besides being cross-reactive, have also given weak or modest skin test responses.‘5-‘7 In our study, 3 pg of MI 43 gave a weak 5 mm response and the same amount of MI 22 and MI 85 gave moderate 9 mm maximal reactions. In contrast, 0.1 pg of the homologous antigens evoked considerably

of M. intracehlare

Recombinant antigen challenge 22 kDa 40 kDa 43 kDa

fusion proteins

85 kDa

Homologous sonicate

Controls

3

3

3

3

3

M. avium

ND

3

4+/- 1

&I- 1+

12+/- 1

M. intracellulare 7+/2+

4+/-l

5+/-o*

9+/-2*

12+/-2

M. tuberculosis

4+/-l

4-G 1

6+/-l+

11+/-l

9+l-2+

Three guinea pigs were tested in each group. The average diameters [mean +/- standard error (mm)] are reported. t-test statistical analysis showed that some DTH assays were significantly different than controls at P
Skin testing with recombinant

stronger DTH responses. Although approximately twothirds of each of these fusion proteins consists of pgalactosidase sequences, approximately 1 pg microgram of recombinant mycobacterial antigen was injected in each assay. This recombinant antigen skin test dose was IO-fold greater than the dose given of homologous antigen. Worsaae et al I6 and Young et al I7 have also demonstrated the superior potency of homologous mycobacterial antigen preparations (such as PPD) compared to purified mycobacterial proteins. The increased molar immunogenicity of PPD and other mycobacterial extracts is probably a consequence of the complexity of antigens present in these preparations. While purified antigens induce a select subset of T cells with limited reactivities, the homologous antigen mixtures stimulate numerous T cells with multiple specificities. This activation of large subsets of T cells by PPD and other complex antigen mixtures probably results in the relatively larger DTH reactions. Although we have not identified an appropriate monospecific MAC skin test reagent in these experiments, the results have yielded a clearer definition of what may be needed to achieve this goal. The ideal skin test reagent may be a polymer consisting of several mycobacterial peptides which encode species-specific T cell epitopes and which are able to bind to major histocompatability complex (MHC) molecules. Because of the multiple specific T cell epitopes present, such a polymer should evoke a significant monospecific DTH response. Although the T cell response to mycobacterial epitopes is haplotype dependent, it is possible that a properly constructed multi-epitopic copolymer will be active within the genetically diverse human population. The elucidation of M. leprue specific T cell epitopes on the 65 and 36 kDa M. leprue proteins suggests that the identification of specific mycobacterial T cell epitopes and the development of specific multi-epitopic skin test reagents is likely.‘” Furthermore, recent studies demonstrating the immunogenicity of multi-epitope peptide vaccines against viral diseases support this synthetic peptide antigen approach.“‘,” We are currently delineating the DNA sequences of genes encoding M. intrucellulare and M. tuberculosis antigens, synthesizing peptides inferred from these DNA sequences as well as published sequences, and evaluating the immunogenicity of these synthetic peptides. Our preliminary experiments indicate that peptides that have been polymerized can induce DTH responses in guinea pigs12. When peptides which elicit specific T cell responses have been defined, the rational design of monospecific skin test reagents for the detection and differentiation of mycobacterial infections may be possible. References

4.

5.

6.

7.

8.

9.

10

I1

12

13

14.

15

16.

17.

18.

19.

20.

21.

22.

I. Collins F M. AIDS-related mycobacterial lmmunopathol

3.

disease. Springer Semin

1988; 10:375-391. disease, immunosuppression

2. Collins F M. Mycobacterial

and the

mvcobacterium

intracellulare

antieens

133

acquired immunodeficiency syndrome. Am Rev Respir Dis 1989; 137: 149-152. Horsbaugh C R. Mycobaclerium avium complex infections in the acquired immunodeficiency syndrome. New Engl J Med 199 1: 324: 1332-I 338. Selwyn PA, Hartel D, Lewis V A et al. A prospective study of the risk of tuberculosis among intravenous drug users with human immunodeficiency virus infection. N Engl J Med 1989; 320: 545-550. Conteras A.. Cheung 0 T, Sanders D E, Goldstein R S. Pulmonary infection with nontuberculosis mycobacteria. Am Rev Respir Dis 1988; 137: 149-152. Chaparas S D. Immunologically based diagnostic tests with tuberculin and other mycobacterial antigens. In Kubica G P and Wayne L G, eds. The mycobacteria. A sourcebook. Part A. New York: Marcel Dekker, 1984: pp 195-2 19. Lind A, Larson L 0, Bentzon M Wet al. Sensitivity to sensitins and tuberculin in Swedish children I. A study of schoolchildren in an urban area. Tubercle 199 I; 72: 29-36. CDC. Purified Protein Derivative (PPD) tuberculin anergy and HIV infection: Guidelines for anergy testing and management of anergic persons at risk of tuberculosis. MMWR 1991; 40: 27-35. Chaparas S D, Maloney C J, Hedrick, S R. Speciticity of tuberculins and antigens from various species of mycobacteria. Am Rev Respir Dis 1970; 101: 74-83. Vandiviere H M, Melvin I G, Narain R, Harris W D M, Chaparas S D. Profiles of skin test reactivity to various mycobacterial species in a human population and in experimental infections. Tubercle 1980; 61: 245-257. Morris S L, Rouse DA, Hussong D, Chapards S D. Isolation and characterization of a recombinant kgt 1 I bacteriophage which expresses an immunoreactive Myobncrerium intrucelluhrr protein in Eschen’chiu coli. Infect lmmun 1988; 56: 3026-3 I Morris S L, Rouse D A. Hussong D, Chaparas S D. Isolation and characterization of recombinant hgtl 1 bacteriophages expressing four different Mycobnc/eriurn irzrrclcellulare antigens. Infect Immun 1990; 58: 17-20. Huyhn T V, Young R A, Davis R W. Construction and screening cDNA libraries in hgtl0 and )igtl 1.In: Glover D M. cd. DNA cloning, vol 1. Oxford: IRL Press 1984; 49-78. Rouse D A, Morris S L, Karpas A B, Mackall .I C, Probst P G, Chaparas S D. Immunological characterization of recombinant antigens isolated from a Myxbacferium atrium hgt 1 1 expression library using antibody probes. Infect Immun 199 1: S9: 2595-26(X). Kingston A E, Salgami P R, Mitchison N A, Colston. MT. Immunological activity of a 14.kilodalton recombinant protein of Mycobac~ferium tuberculosis H37 Rv. Infect lmmun 1987: 55:3 149-3 154. Worsaae A, Ljungquist L, Haslov K, Heron 1. Bennedsen J. Allergenic and blastogenic reactivity of three antigens from Mycobncterium tuberculosi.~ in sensitized guinea pigs. Infect Immun 1987; 55: 2922-2927. Young D, Kent L, Rees A, Lamb J, Ivanyi J. Immunological activity of a 38 kDa protein purified from M?c,oh~ic,tt,rirrr,~ tuberculosis. Infect Immun 1986; 54: 177-l 83. Kadival G B, Chaparas S D, Hussong D. Characterization of the serologic and cell mediated reactivity of a 38 kDa antigen isolated from Mycobacterium tuberculosis. J Immunol 1987: 139: 2447-245 1. van Schooten W C A, Ottenhoff T H M, Klaster P R, Thole J, Devries R R P, Kolk A H J. T cell epitopes on the 36 K and 65 K Mycobacterium leprae antigens defined by human T cell clones. Eur J Immunol 1988; 18: 849-854. Tam J P, Lu Y. Vaccine engineering: enhancement of immunogenicity of synthetic peptide vaccines related to hepatitis in chemically defined models consisting of T and B cell epitopes. Proc Nat1 Acad Sci U S A 1989; 86: 9084-9088. Leverly ME, Mitchell MM, Nicholas JA. Synthetic immunogena constructed from T-cell and B-cell stimulating peptides (T:B chimeras): Preferential stimulation of uniqueT- and B- cell speciticities is influenced by immunogen coniipuration. Ccl1 Immunol 1990; 125: 65-78. Morris S L, Rouse D A, Chaparas S D. B and T cell epitopes of Mycobacrerium ovium complex proteins. Abstracts of the tirst international conference on the pathogenesis of mycobacterial infections 1990; IO.