Resistance to tuberculosis (TB) is cell-mediated but a humoral response is common and may be ... critical for the protective immunity to M. tuberculosis (2-4).
putative open reading frames were compared to mass spectrometry data from M. haemofelis protein spots identified as being .... use in general practice, for example, the SNAP ELISAs (Idexx. Labs.). The ideal serological assay for ...... We thank Ben C
protein in the pathogenesis of tubal infertility. Key words: Chlamydia trachomatis infection/chlamydial anti- bodies/heat shock protein/salpingitis/tubal infertility.
www.annualreviews.org/aronline. Annu. Rev. Biochem. 1986.55:1151-1191. Downloaded from arjournals.annualreviews.org by MASSACHUSETTS INSTITUTE ...
Aug 5, 1999 - in an Experimental Model of Chronic Pelvic Inflammatory Disease in ... A primate model of chlamydial pelvic inflammatory disease was used to ...
Apr 2, 1999 - Dhurjati, and R. A. Larossa. 1994. Rapid and sensitive pollutant detection by induction of heat shock gene-bioluminescence gene fusions.
OBJECTIVES. This study was designed to determine the role of cellular and humoral immune responses to heat shock protein 65 (HSP65) in murine ...
to follow the cross-regulation of protein breakdown and rebuild- ing, including novel ..... larly, knockdown of an essential autophagy gene (ATG5) results.
sulfur-oxidizing hyperthermophile Sulfolobus shibatae and in the anaerobic sulfur ... In both S. shibatae and ES4, proteins were produced during exposure to a ...
The heat shock response is characterized by a rapid and robust increase in heat shock proteins upon exposure to protein-damaging stresses. This evolutionarily conserved cellular protection mechanism is primarily regulated at the level of transcriptio
Abstract. The humoral immune response against synthetic peptides of two ... in the capacity to mount immune responses against P. falciparum antigens among ...
molecular chaperones, called heat shock proteins (Hsps), which consist of several ... The human hsp70.1 promoter contains two stress-responsive heat shock ...
Apr 1, 1993 - The LPS molecules of C. jejuni and C. cob organism Campylobacter coli ..... Blar, M. J., E. Sazie, and L. P. Williams. 1987. to homologous and ...
Abstract. Hepatocellular carcinoma (HCC) is a worldwide prevalent cancer with an exremely poor prognosis. Detection of serum Î±-fetoprotein (AFP) and liver ...
Keywords: heat shock, heat shock proteins, thermophiles, archaea, ... homology with the chaperonins of the eukaryotic cytosol, and constitutes together with these proteins the Group II chaperonins, whereas the chaperonins from Bacteria and.
weighed and fixed by immersion in buffered 10% for- malin. In hearts ...... Fishbein MC, Meerbaum S, Rit J, Lando U, Kanmatsuse K, Mer- cier JC, Corday E, ...
Bacterial heat-shock response is a global regulatory system required for ... In low-G+C gram positive bacteria, such as Bacillus subtilis, the heat-shock genes are.
Mar 30, 2006 - ndhR, a LysR family regulator. Journal of Biological Chemistry. 279, 5739â5751. Yamaguchi K, Suzuki I, Yamamoto H, Lyukevich A, Bodrova I,.
Jul 26, 2016 - weekly ixazomib, an oral proteasome inhibitor, in relapsed/ refractory multiple myeloma patients. Blood. 2014;. 124:1038-1046. 8. O'Connor OA ...
IPOCAMP, chaperone, Rimicaris exoculata, shrimp, heat shock. Summary ... questions, we undertook several experiments with live shrimps ... means of optical fiber light-guides. .... only the individuals for which the final oxygen concentration.
supports the protein labeling data: hsp 70 gene induction is first detectable in gastrulae, whereas hsp'30 genes remain uninducible until the tadpole stage.
heat stress (Boorstein and Craig, 1990). These data suggest that HSF ..... SchÃ¶ffl F, Rieping M, Baumann G, Bevan M, AngermÃ¼ller S. (1989) The function of heat ...
Oct 26, 2011 - Adrenaline. HR (6). 1.4. 1.0a. 15.6. 6.0a. 113.9. 64.5b. 23.1. 12.0a. 8.3. 1.5a ..... ments and the staff of DTU Aqua for fish rearing and care.
May 17, 1988 - 100 ,ug of P. aeruginosa RNA polymerase and Freund incomplete adjuvant ..... carried out with antibodies prepared against the E. ccli a b ..... Top. Microbiol. Immu- nol. 96:47-67. 4. Bardwell, J. C. A., and E. A. Craig. 1984.
Vol. 1, No. 5
CLINICAL AND DIAGNOSTIC LABORATORY IMMUNOLOGY, Sept. 1994, p. 552-555
1071-412X/94/$04.00+0 Copyright C 1994, American Society for Microbiology
Humoral Response to Mycobacterial Heat Shock Proteins in Patients with Constrictive Pericarditis Caused by Tuberculosis and Its Implications for Pathogenesis T. T. C. NG,1,2* J. I. G. STRANG,3 AND E. G. L. WILKINS' Monsall Unit, Department of Infectious Diseases and Tropical Medicine, North Manchester General Hospital, Manchester M8 6RL, 1 and Department of Medical Microbiology, University of Manchester Medical School, Manchester M13 9PS,2 United Kingdom, and Cecilia Makiwane Hospital, Ciskei, South Afrtica3 Received 2 February 1994/Returned for modification 6 May 1994/Accepted 31 May 1994
MATERIALS AND METHODS
Mycobacteria are rich in antigens capable of stimulating both T- and B-cell responses in mammals. A significant proportion (up to 20%) of the cell-mediated response generated is directed at the mycobacterial heat shock proteins (M-hsp) (6, 8). The humoral response to some of these proteins such as M-hsp70 (71/70 kDa) is also exaggerated in patients with active tuberculosis (TB) (4). Because of the antigenic cross-reactivity shared between these proteins and their human homologs (reviewed in reference 20), considerable interest has been generated in establishing their role in the pathogenesis of autoimmune diseases such as rheumatoid arthritis. In a previous study (4), we demonstrated that elevated antibody titers against the 71-kDa M-hsp were also present in a small group of patients with nontuberculous lung fibrosis (of unknown etiology) and proposed that this specific serological response could be used as a marker of fibrosis. Chronic inflammation of unknown etiology which results in substantial fibrosis is usually due to an aberrant immune response or hypersensitivity reaction (e.g., in autoimmune diseases). Fibrosis of this type is also seen in tuberculous constrictive pericarditis, which is caused by a hypersensitivity reaction resulting from the rupture of tuberculous lymph nodes (mediastinal or paratracheal) into the pericardial sac. In this study we investigated the serum antibody responses against the 65- and 71-kDa M-hsp by an indirect enzyme-linked immunosorbent assay (ELISA) in 25 patients with microbiologically confirmed tuberculous pericardial effusion and found a significant association between high anti-65-kDa M-hsp antibody titers and subsequent development of pericardial fibrosis and constriction.
Subjects and sera. Sera from a total of 65 subjects were tested. Of the 65 subjects, 25 patients from a previously reported clinical trial (14) had clinically diagnosed and microbiologically confirmed (by culture of pericardial fluid) tuberculous pericardial effusion and 40 were controls who did not have TB. Fourteen of the patients with pericardial TB were randomized in the original trial to receive prednisolone (in addition to standard antituberculous therapy), while the remainder 11 received placebo tablets. The sera were collected from Northwick Park Hospital, London, United Kingdom (15 controls), and Umtata Hospital, Transkei (25 pericardial TB patients and 25 controls). Of the 40 control subjects, 15 (from Northwick Park Hospital) had known diagnoses other than TB which were divided into five categories and are summarized in Table 1. The remaining 25 controls (from Transkei) had diseases simulating tuberculosis (unspecified) and no history of recent or earlier TB. All sera from TB patients were collected before the institution of antituberculous therapy. All sera were stored in aliquots at -20°C and tested blind to eliminate operator bias. Purified recombinant antigens. Both the Mycobacterium bovis BCG recombinant 65-kDa (r65-kDa) and Mycobacterium tuberculosis r71-kDa antigens were kindly prepared and purified by J. Kamerbeek and J. van Embden, National Institute of Public Health and Environmental Protection, Bilthoven, The Netherlands. The r65-kDa protein was obtained from Escherichia coli K-12 strain M1164, carrying plasmid pRIB1300 (16, 18), and the r71-kDa protein was obtained from E. coli K-12 strain M1485, carrying plasmid Y3111/8 (11). Indirect ELISA. Flat-bottomed flexible polyvinyl microtiter plates (Falcon 1413 Microtest III flexible assay plates; Becton Dickinson Labware, Oxnard, Calif.) were used. These were coated with 50 ,ul of 10-,ug/ml recombinant mycobacterial antigen (dissolved in phosphate-buffered saline [PBS]) or PBS
* Corresponding author. Present address: Department of Immunology, St. Bartholomew's Hospital, 38 Little Britain, West Smithfield, London EClA 7BE, United Kingdom. Phone: 44-71-601-8104. Fax: 44-71-606-0845.
Downloaded from http://cvi.asm.org/ on November 3, 2017 by guest
Tuberculous pericarditis is one of the commonest causes of cardiac failure in Transkei and the surrounding regions in southeast Africa. About 20%o of patients with clinically diagnosed tuberculous pericardial elfusion go on to develop pericardial fibrosis (i.e., constriction), a complication which is associated with significant mortality and morbidity. The pathological mechanisms underlying this aberrant inflammatory response are poorly understood, and there is a lack of reliable pointers (clinical or laboratory) in predicting the likelihood of development of constriction. We studied the humoral response to mycobacterial heat shock proteins (65 and 71 kDa) in 25 patients with culture-positive tuberculous pericardial elfusion and found a significant correlation between high anti-mycobacterial hsp60 antibody titers (before treatment) and subsequent development of fibrosis (P = 0.035 by logistic regression), which is independent of the effect of the use of prednisolone as adjuvant therapy. Possible mechanisms underlying the pathogenesis of pericardial constriction in tuberculosis are postulated.
HUMORAL RESPONSE TO M-hsps IN TUBERCULOSIS
VOL. 1, 1994
TABLE 1. Known diagnoses of 15 control subjects Group
washed five times with PBST and patted dry. Affinity-isolated Fc-specific goat anti-human immunoglobulin G (IgG) peroxidase conjugate (Sigma, St. Louis, Mo.) diluted in BSA-Tween to 1/1,000 was added at 50 pul per well, and the plates were incubated for 1 h at 37°C. After a further five washes, the plates were dried and 50 ,u1 of 2'2'-azinobis(3-ethylbenzthiazolinesulfonic acid) (ABTS) solution (0.5 mg of ABTS per ml in phosphate-citrate buffer [pH 4] with 0.01% hydrogen peroxide) was added per well. The A405 was read (40 min after the addition of ABTS) in a Titertek Multiskan spectrophotometer. Antibody titers were expressed as the dilution of serum giving 50% of the plateau binding value of a standard hyperimmune serum (ABT5O) (4) after the value of nonspecific binding to wells coated with PBS had been subtracted from all values (to keep mean optical density ± standard deviation due to nonspecific binding amounts to 0.1 ± 0.03). Antibody titers of less than 50 were calculated by extrapolation from the linear regression lines (derived from the three serum dilutions; r 2 0.9 in most cases). All titers of less than 1 were recorded as zero.
Statistical analysis. Statistical analysis was done by using the SPSS/PC+ Statistics 4.0 (SPSS Inc., Chicago, Ill.) package as specified in Results. Curve-fitting by linear regression was done by using Fig.P (Biosoft, Cambridge, United Kingdom).
alone per well and held overnight at 4°C. After being washed once with PBS containing 0.5% Tween 20 (BDH, Poole, United Kingdom) (PBST), the wells were incubated with 200 pI of 2% skimmed milk in PBST at 37°C for 1 h. The liquid was tipped off, and the plates were patted dry. Three dilutions of human sera (1:50, 1:250, 1:1,250; 50 pI each) in 3% bovine serum albumin (BSA) solution containing 0.2% Tween 20 (BSA-Tween) were added to triplicate wells (one coated with PBS only) and incubated for 1 h at 37°C. The wells were then
FIG. 1. IgG ABT50 values to M-hsp6O (65 kDa) in patients with pericardial TB (12 of whom subsequently developed constriction [Constr.] and 13 of whom did not [Eff.]) and controls (25 African [Afr.] and 15 British controls). Group means are represented by the horizontal bars. The abbreviations for the five categories of British controls are defined in Table 1.
patients subsequently developed signs of pericardial constriction (constrictors) despite standard antituberculous therapy (14) (time to development of constriction ranged from 2 weeks to 4 months [mean, 6 weeks]). Of these, four received prednisolone and eight received placebo tablets initially (among those who did not develop constriction, 10 were given prednisolone and 3 were given placebo). The mean r65-kDa ABT50 values for the constrictor (n = 12) and nonconstrictor (n = 13) groups were 36.7 (range, 0 to 280) and 6.6 (range, 0 to 50), respectively. The results are shown in Fig. 1. (ii) r71-kDa IgG antibody. The mean ABT50 value was 41 (range, 0 to 137). The mean r71-kDa ABT50 values for the constrictor and nonconstrictor groups were 39 (range, 0 to 106) and 42 (range, 0 to 137), respectively (Fig. 2). Statistical analysis. The effects of r65-kDa IgG seropositivity and prior prednisolone therapy on the development of constriction were further analyzed by using a logistic regression
RESULTS Non-TB controls. For the African controls (n = 25), the mean ABT50 values against the r65-kDa and r71-kDa antigens were 13 (range, 0 to 155) and 52 (range, 0 to 223), respectively. For the British controls (n = 15), the mean ABT50 values were 59 (range, 0 to 349) and 92 (range 0 to 267), respectively (Fig. 1 and 2). Tuberculous pericarditis patients. (i) r65-kDa IgG antibody. The mean ABT50 value for patients with tuberculous pericarditis (n = 25) was 21 (range, 0 to 280). Twelve (48%)
Meig. Others -
FIG. 2. IgG ABT50 values to M-hsp7O (71 kDa) in patients with pericardial TB and controls. Abbreviations are defined in Table 1 and the legend to Fig. 1.
Downloaded from http://cvi.asm.org/ on November 3, 2017 by guest
NG ET AL.
DISCUSSION The 65-kDa M. bovis antigen shares at least four identical regions (of 10 or more amino acids) with the human GroEL protein (also known as the human hsp58 [H-hspS8]) (12). Both are homologous with the E. coli GroEL protein (21), a cytosolic chaperonin protein which functions primarily to mediate the folding of newly synthesized polypeptides into their final conformation before export (10). The surface expression of the human GroEL protein (normally located in the mitochondrial compartment) or its cross-reactive epitopes can be induced by cytokines such as gamma interferon (19) and may serve as a marker of active inflammation in certain tissues such as alveolar lining or synovium (5). It is conceivable (but not demonstrated) that the same process may also occur in chronically inflammed pericardium as a result of tuberculous infection. The surface expression of H-hsp58 or its epitopes may then provide potential targets, by virtue of the crossreactivity of H-hsp58, for the host cytototoxic T lymphocytes generated in response to the invading mycobacteria (9). We have shown in this study a previously undocumented association (independent of the effect of steroid) between high IgG antibody titers against the M. bovis 65-kDa antigen and subsequent development of fibrosis (constriction) in patients with tuberculous pericarditis. It remains to be seen whether the antibodies produced are autoreactive and capable of binding to the autologous epitope sites expressed in the diseased pericardium, thereby providing an alternative (antibody-dependent) mechanism of cell-mediated cytotoxicity. These two (T-cellmediated and antibody-dependent) mechanisms may operate synergistically in bringing about the fibrotic sequelae in pericardial TB. Regardless of the effector cell type, any aberrant immune response raised against H-hsp58 (a dominant self-antigen) would normally be tightly controlled by the naturally existing anti-idiotypic networks to prevent the development of pernicious autoimmune disease (reviewed in references 1 and 2). Factors which suppress these regulatory networks and hence facilitate the development of pericardial fibrosis are at present unknown.
The measurement of serum antibody titers against the M. bovis 65-kDa antigen may be used to assist in the clinical management of patients with tuberculous pericarditis. Despite antituberculous chemotherapy, about 15 to 20% of patients who present with clinically diagnosed tuberculous pericardial effusion will subsequently develop constriction (13), a complication which is associated with significant mortality (4 to 11 %) and morbidity (15). The ability to identify the group of patients with an increased propensity to develop constriction may enable us to tailor the use of immunomodulatory agents to the individual's "susceptibility." Diagnostically, the antibody response to neither of the stress proteins could be used to differentiate pericardial TB from other nontuberculous inflammatory conditions. This is somewhat surprising in view of our previous finding of elevated antibody titers against the M tuberculosis 71-kDa antigen in patients with active TB (4). However, nearly two-thirds of these patients had pulmonary disease. The apparent discrepancy between the results of the two studies may therefore be accounted for by the sitedependent property of B-cell activation in tuberculous infections (7; our unpublished data). The close correlation between the anti-65-kDa and anti-71kDa antibody titers in pericardial TB is also found in other nontuberculous mycobacterial lung diseases but not in Crohn's disease (3). The finding of no detectable IgG antibody titer against the mycobacterial 65-kDa antigen in our control subject with rheumatoid arthritis (a putative mycobacterial condition) is surprising in view of the results of a previous study (17). The immunopathogenesis of tuberculous pericarditis is not yet fully understood. Clinically, the use of corticosteroids in addition to antituberculous therapy for tuberculous pericardial effusion has led to a reduction in the mortality and morbidity associated with the development of fibrosis (and possibly its incidence as shown in our study) (14, 15). The exact mechanism by which prednisolone, a nonspecific immunosuppressive agent, halts the progression of the aberrant inflammatory response is, however, unknown. A better understanding of these immunopathological processes may enable us to develop more-specific immunomodulatory agents than are currently available. ACKNOWLEDGMENTS We are indebted to Sally Hollis, Department of Statistics, University of Manchester, Hope Hospital, Salford, United Kingdom, for statistical advice. This work was supported by the Wellcome Trust (grants 031463/2/ 90/2/1.5 and 019317/2/89/B/1.5/REH/CS to J.I.G.S.) and the British Society for the Study of Infection/Bayer Gerhard-Domagk award (to E.G.L.W.). Preparation of the r71-kDa and r65-kDa antigens was supported by the UNPD/World Bank/WHO Special Programme for Research and Training in Tropical Diseases. REFERENCES 1. Cohen, I. R. 1991. Autoimmunity to chaperonins in the pathogenesis of arthritis and diabetes. Annu. Rev. Immunol. 9:567-589. 2. Cohen, L. R., and D. B. Young. 1991. Autoimmunity, microbial immunity and the immunological homunculus. Immunol. Today 12:105-110. 3. Elsaghier, A., C. Prantera, G. Bothamley, E. Wilkins, S. Jindal, and J. Ivanyi. 1992. Disease association of antibodies to human and mycobacterial hsp70 and hsp60 stress proteins. Clin. Exp. Immunol. 89:305-309. 4. Elsaghier, A. A. F., E.G. L. Wilkins, P. K. Mehrotra, S. Jindal, and J. Ivanyi. 1991. Elevated antibody levels to stress protein hsp70 in smear-negative tuberculosis. Immunol. Infect. Dis. 1:323-328. 5. Evans, D. J., P. Norton, and J. Ivanyi. 1990. Distribution in tissue sections of the human GroEL stress protein homologue. APMIS 98:437-441.
Downloaded from http://cvi.asm.org/ on November 3, 2017 by guest
model (with the cutoff titer for r65-kDa IgG arbitrarily chosen at 4). The r65-kDa IgG seropositive rates for the constrictor and nonconstrictor groups were 58% (7 of 12) and 15% (2 of 13), respectively (P = 0.035 by logistic regression). The effect of adjunctive steroid therapy resulting in a reduction in fibrotic sequelae was also statistically significant (P = 0.04). The effects of seropositivity and steroid were found to be independent of each other (P = 0.3426 for the interaction by logistic regression). This analysis was not performed on the r71-kDa IgG data because of the similarity in the distribution of antibody titers between the constrictor and nonconstrictor groups. The use of the r65-kDa ABT50 value alone in predicting the likelihood of development of constriction was associated with a positive predictive value of 0.8 and negative predictive value of 0.7. There was no significant difference in the mean r65-kDa or r71-kDa ABT50 values among the African controls (n = 25), British controls (n = 15), and the patients with tuberculous pericarditis (n = 25) (P = 0.4945 and 0.5152 by the KruskallWallis test for r65-kDa and r71-kDa values, respectively). The r65-kDa and r71-kDa titers were closely correlated among the pericarditis patients (Spearman correlation coefficient = 0.6432; P < 0.002) but not in the corresponding African control group (Spearman correlation coefficient = 0.2974; P = 0.07).
CLIN. DIAGN. LAB. IMMUNOL.
VOL. 1, 1994 6. Havlir, D. V., R. S. Wallis, W. H. Boom, T. M. Daniel, K. Chervenak, and J. J. Ellner. 1991. Human immune response to Mycobacterium tuberculosis antigens. Infect. Immun. 59:665-670. 7. Ivanyi, J., E. Krambovitis, and M. Keen. 1983. Evaluation of a monoclonal antibody (TB72) based serological test for tuberculosis. Clin. Exp. Immunol. 54:337-345. 8. Kaufmann, S. H. E., U. Vath, J. E. R. Thole, J. D. A. van Embden, and F. Emmrich. 1987. Enumeration of T cells reactive with Mycobacterium tuberculosis organisms and specific for the recombinant mycobacterial 64-kDa protein. Eur. J. Immunol. 17:351-
HUMORAL RESPONSE TO M-hsps IN TUBERCULOSIS
9. Kiessling, R., A. Gronberg, J. Ivanyi, K. Soderstrom, M. Ferm, S. Kleinau, E. Nilsson, and L. Klareskog. 1991. Role of hsp60 during autoimmune and bacterial inflammation. Immunol. Rev. 121:91-
Controlled clinical trial of complete open surgical drainage and of prednisolone in treatment of tuberculous pericardial effusion in Transkei. Lancet ii:759-764. Strang, J. I. G., H. H. S. Kakaza, D. G. Gibson, D. J. Girling, A. J. Nunn, and W. Fox. 1987. Controlled trial of prednisolone as adjuvant in treatment of tuberculous constrictive pericarditis in Transkei. Lancet ii:1418-1422. Thole, J. E. R., W. J. Keulen, J. De Bruyn, A. H. Kolk, D. G. Groothuis, L. G. Berwald, R. H. Tiesjema, and J. D. van Embden. 1987. Characterization, sequence determination, and immunogenicity of a 64-kilodalton protein of Mycobacterium bovis BCG expressed in Escherichia coli K-12. Infect. Immun. 55:1466-1475. Tsoulfa, G., G. A. W. Rook, G. M. Bahr, M. A. Sattar, K. Behbehani, D. B. Young, A. Mehlert, J. D. A. van Embden, F. C. Hay, D. A. Isenberg, and P. M. Lydyard. 1989. Elevated IgG antibody levels to the mycobacterial 65-kDa heat shock protein are characteristic of patients with rheumatoid arthritis. Scand. J. Immunol. 30:519-527. Van Eden, W., J. E. Thole, R. van der Zee, A. Noordzij, J. D. van Embden, E. J. Heusen, and I. R. Cohen. 1988. Cloning of the mycobacterial epitope recognized by T-lymphocytes in adjuvant arthritis. Nature (London) 331:171-173. Wand-Wurttenberger, A., B. Schoel, J. Ivanyi, and S. H. E. Kaufmann. 1991. Surface expression by mononuclear phagocytes of an epitope shared with mycobacterial heat shock protein 60. Eur. J. Immunol. 21:1089-1092. Winfield, J., and W. Jarjour. 1991. Do stress proteins play a role in arthritis and autoimmunity? Immunol. Rev. 121:193-220. Young, D. B., and A. Mehlert. 1989. Serology of mycobacteria: characterization of antigens recognised by monoclonal antibodies. Rev. Infect. Dis. 11:431-435.
Downloaded from http://cvi.asm.org/ on November 3, 2017 by guest
10. Martin, J., S. Geromanos, P. Tempst, and F. U. Harti. 1993. Identification of nucleotide-binding regions in the chaperonin proteins GroEL and GroES. Nature (London) 366:279-282. 11. Mehiert, A., and D. B. Young. 1989. Biochemical and antigenic characterisation of the Mycobacterium tuberculous 7lkD antigen, a member of the 7OkD heat-shock protein family. Mol. Microbiol. 3:125-130. 12. Munk, M. E., B. Schoel, S. Modrow, R. W. Karr, R. A. Young, and S. H. E. Kaufmann. 1989. T lymphocytes from healthy individuals with specificity to self-epitopes shared by the mycobacterial and human 65-kilodalton heat shock protein. J. Immunol. 9:28442849. 13. Strang, J. I. G. 1984. Tuberculous pericarditis in Transkei. Clin. Cardiol. 7:667-670. 14. Strang, J. I. G., H. H. S. Kakaza, D. G. Gibson, B. W. Allen, D. A. Mitchison, D. J. Evans, D. J. Girling, and A. J. Nunn. 1988.