Search

Both experimental and clinical forms of chronic GVHD have unique immunological features. The affected animals/individuals suffer from autoimmune disorders such as systemic lupus erythematosus (SLE), and yet they are unable to mount a self MHC-restricted T cell response to foreign antigens. Pathogenesis of the latter phenomenon was investigated in an experimental model of chronic GVHD. Chronic GVHD was induced in 8-10-week-old (B6xC3H)F1 mice by tail vein injection of 5 x 10(7) spleen cells of C3H parental strain. The recipients, when tested 3 months later, were unable to mount a T helper (Th) cell response to a randomly selected immunogen, a vaccine of l0(8) killed Mycobacterium vaccae. The animals showed evidence of generalized lymphoid hyperplasia, as indicated by GVH index >1.34, and also revealed autoantibodies against erythrocytes and dsDNA, indicating establishment of chronic GVHD. However, mice with chronic GVHD of only 3 weeks duration were able to mount the Th cell response to M. vaccae. Three consecutive immunizations of these mice at 1-week intervals, with the same immunogen, resulted in the mice becoming non-responsive to the antigen. All the three responses tested, namely the DTH, lymphoproliferation and the antibody responses, were adversely affected. The non-responsiveness induced was antigen-specific. Mice receiving two immunizations with M. vaccae responded normally to Salmonella enteritidis. Pulse treatment with cyclosporin A 0.5 mg/mouse by the i.p. route, on days 0, 1, 2, 3 and 4 at the time of immunization with M. vaccae on day 1, prevented emergence of non-responsiveness. Based on this evidence, it was concluded that repeated activation of T cells of mice with chronic GVHD induces non-responsiveness. Extent of clonal loss due to activation-induced cell death (AICD) caused by i.p. injection with a superantigen Staphylococcal enterotoxin B (SEB) was investigated in F1 mice with chronic GVHD. I.p. injection of 25 microg/mouse of SEB induced loss of SEB responding clones in both normal F1 mice and those having chronic GVHD; however, the extent of loss was much greater in the latter. In vitro antigen-specific proliferation of primed splenic T cells of normal F1 mice was observed to be quite poor when antigen was presented by APC of mice with chronic GVHD of 3 weeks duration. Proliferation profiles of T cells of normal F1 mice, in response to stimulation with concanavalin A (Con A) or SEB, were studied, using as APC irradiated spleen cells of normal F1 mice or of F1 mice with chronic GVHD of 3 weeks duration. With Con A and APC of normal F1 mice, peak proliferation was observed at 48 h, which remained at the same level up to 72 h and declined thereafter, possibly due to AICD. With SEB and the normal APC, proliferation progressively peaked at 72 h and declined thereafter. With APC of mice with chronic GVHD, the 48 h proliferative responses of both Con A and SEB were comparable to those caused by APC of normal F1 mice; however, thereafter the responses declined steeply, suggesting greater AICD. Based on these results, it was concluded that APC of mice with chronic GVHD are functionally altered to induce greater AICD.

In earlier work, intraperitoneal (i.p.) immunization with Mycobacterium vaccae was shown to generate a T-suppressor (Ts) response but intradermal (i.d.) immunization did not. We have now studied the major histocompatibility complex (MHC) restriction of this Ts response. The ability of C57BL/6 (H-2b), BALB/c (H-2d), and the (C57BL/6 x BALB/c) F1 mice to generate suppression after i.p. immunization with 10(8) killed M. vaccae was investigated. The BALB/c and the F1 mice generated suppression, but the C57BL/6 mice failed to do so. The suppression could be ascribed to Lyt-2+, L3T4- antigen-specific T cells. The F1 suppressors generated after i.p. immunization could suppress the generation of T-cell responses to i.d. immunization with M. vaccae in the parental BALB/c but not in the C57BL/6 mice. Monoclonal anti-I-A antibody could suppress the antigen-induced proliferative response of mice primed i.d. with M. vaccae. In contrast, monoclonal anti-I-E antibody enhanced antigen-specific proliferation of spleen cells primed i.p. with M. vaccae. The suppressors generated by i.p. priming of mice with M. vaccae could also suppress the in vitro antigen-induced proliferative response of i.d.-primed spleen cells; the suppression could be blocked by anti-I-E antibody. Thus, the T-cell-mediated suppression in the above experimental model was I-E restricted. The inability of the C57BL/6 mice to generate suppression after i.p. immunization with M. vaccae was ascribed to the lack of I-E expression by mice of H-2b strain.

Markers of hepatitis B virus (HBV) and immune response against them were studied in 18 chronic asymptomatic carriers, 8 patients of the virus induced chronic liver disease (CLD), and 7 patients of chronic alcoholic liver cirrhosis, who were also chronic HBV carriers (CALC). The LMI responses to HBeAg were elevated in HBeAg and/or HBV-DNA positive chronic asymptomatic carriers, (median response 31.5%), along with elevation of serum alanine aminotransferase (sALT) levels (59-150 IU/l). On the other hand the LMI responses to this antigen, in HBeAg and HBV-DNA negative chronic carriers were in the normal range (median response 12%) and their sALT levels were also normal (7-50 IU/l). The CLD and CALC patients did not show any relation between their LMI to HBeAg and sALT levels. In contrast no relation between LMI to HBsAg and sALT levels was observed in any group. The LMI responses to HBsAg in CLD patients were elevated (median response 38%) and the responses of chronic asymptomatic carriers and CALC patients were either in the normal range or poor (median responses, 18 and 7% respectively), irrespective of their sALT levels. These results suggest that T cell responses to both the antigens may be involved in liver cell damage.

The antigen-presenting efficiency of peritoneal cells and irradiated spleen cells was compared using Mycobacterium tuberculosis- and M. vaccae-primed T cells and corresponding sonicates as antigens in an in vitro lymphocyte transformation test. The presentation efficiency of irradiated spleen cells was reasonably good for both antigens. However, with peritoneal cells as the antigen-presenting cells, the proliferative response against only M. tuberculosis sonicate was good. Proliferation of M. vaccae-primed T cells was very poor when the antigen was presented by peritoneal cells. Poly I:poly C treatment of mice prior to harvesting the peritoneal cells resulted in distinct improvement in their efficiency to present M. vaccae sonicate; maximal proliferative response was obtained with peritoneal cells from mice receiving two and three doses of poly I:poly C 24 hr apart. Even paraformaldehyde-fixed peritoneal cells from poly I:poly C-treated mice gave an efficient M. vaccae-specific stimulation to primed T cells. Based on these data, it was concluded that failure of mice to respond to M. vaccae by intraperitoneal immunization is the result of the poor efficiency of presentation of M. vaccae antigen.

The route of immunization was observed to play a significant role in deciding the T-cell response to immunization with killed mycobacterial vaccines. Slow-growing mycobacteria were found to be immunogenic by both the intraperitoneal (i.p.) and intradermal (i.d.) routes; rapid-growing mycobacteria were immunogenic by the i.d. route only. The nonresponder state following i.p. immunization with Mycobacterium vaccae could be corrected by treatment of the mice with poly I:poly C or indomethacin prior to immunization. Both poly I:poly C, an interferon inducer, and indomethacin, a prostaglandin inhibitor, are known to enhance the expression of major histocompatibility complex glycoproteins. Since they are so important in antigen preparation, it was concluded that the inability of mice to respond to M. vaccae by the i.p. route is likely due to defective presentation of the bacterial antigens by the antigen-presenting cells at the site, namely, the peritoneal macrophages. These findings are significant because M. leprae has been reported to be antigenically similar to M. vaccae, and the response of mice to i.p. immunization with both of these mycobacteria is very similar.

The route of immunization was observed to play a significant role in deciding the outcome of immunization with killed mycobacterial vaccines. Earlier we reported that the slow growers were immunogenic by both the intraperitoneal (i.p.) and intradermal (i.d.) routes. In contrast, the rapid growers were immunogenic by the i.d. route only. Both rapid and slow growers generated the classical, antigen-specific Lyt-2 positive, T-cell-mediated suppression after i.p. immunization but not after i.d. immunization. Thus, in the case of the slow growers, T-cell-mediated suppression was only a component of the immune response generated after i.p. immunization. In contrast, in the case of Mycobacterium vaccae and the other rapid growers, the T-cell-mediated suppression was the predominant response with i.p. immunization. The T-cell-mediated suppression generated by i.p. immunization exhibited crossreactivity, the spectrum of which was dependent upon the dose of the immunization.

The route of immunization was observed to play a significant role in deciding the outcome of immunization with killed mycobacterial vaccines. Whereas the slow growers were immunogenic by both intraperitoneal and intradermal routes, the rapid growers were immunogenic only by intradermal route. The non-responder state of mice to Mycobacterium vaccae by i.p. route of immunization could be corrected by prior treatment with poly I:poly C, an interferon inducer, or indomethacin, a prostaglandin inhibitor. Antigen-presenting efficiency of peritoneal and spleen cells were compared employing M. vaccae and M. tuberculosis H37Rv primed T cells and corresponding sonicates as antigens in an in vitro lymphocyte transformation test. Irradiated spleen cells presented both the antigens efficiently. However, with peritoneal cells as antigen-presenting cells, proliferative response against only M. tuberculosis was observed; proliferation of M. vaccae primed T cells was very poor. Peritoneal cells of poly I:poly C treated mice showed distinct improvement in their efficiency of presentation; even paraformaldehyde-fixed peritoneal cells gave an efficient stimulation with M. vaccae. The percentage of Ia-positive fraction in peritoneal cells was very low (5.95%) in comparison with spleen cells (38.37%). Poly I:poly C treatment resulted in increase in the Ia-positive cell fraction of the peritoneal cells to 24.5%.

The in-vitro adhesion of Vibrio cholerae to intestinal mucous membrane was studied in isolated adult-rabbit ileal loops. Antisomatic antiserum Against V. cholerae Inaba could inhibit adhesion of three different strains of V. cholerae Inaba but had no effect on the adhesion of two different strains of enterotoxigenic NAG vibrios. The antiserum's bacterial agglutinin titre was 320, its anti-Inaba lipopolysaccharide (LPS) titre was 16 000 and its anti-flagellar antibody titre was 3200. Conversely, anti-live V. Cholerae Inaba antiserum absorbed with boiled cells of Inaba and devoid of antisomatic antibody, could not inhibit adhesion of the same three strains of V. cholerae Inaba. This antiserum had no anti-LPS or bacterial agglutinin activity, but its anti-flagellar antibody titre was 32 000. Thus, ability to inhibit adhesion of V. cholerae could be correlated only with antisomatic (anti-LPS) antibody activity. Antisomatic antiserum had no activity against 'adhesion', a V. cholerae surface antigen described by Freter. Conversely anti-live V. cholerae antiserum absorbed with boiled cells showed anti-adhesion activity even at a dilution of 1 in 200. LPS preparations from V. cholerae strain 569B Inaba could inhibit adhesion of two different Inaba strains to the intestinal mucous membrane. It is concluded that the somatic antigen plays a major role in the adhesion of V. cholerae to the intestinal mucous membrane.

In studies of the role of surface antigens of Shigella flexneri in HeLa cell invasion, three antisera were employed to block the invasion. Antisera against live (ALS) and boiled (ABS) S. flexneri blocked invasion very effectively. Reduction in the numbers of intracellular shigellae was always accompanied by reduction in the number adherent to the cells, indicating the importance of adhesion in the invasive process. Anti-live absorbed antiserum (ALAS) prepared by exhaustive absorption of ALS with boiled S. flexneri blocked adhesion and invasion at dilutions of 20 or 50; the efficiency of the absorption was indicated by absence of agglutinating and anti-lipopolysaccharide (LPS) antibodies. S. flexneri LPS did not block adhesion and invasion even at a concentration of 1.0 mg/ml. Hence it was concluded that heat-labile surface antigens are important in adhesion and invasion of HeLa cells by S. flexneri. Antiserum against heat stable antigen (ABS) probably blocks adhesion by steric hindrance.

Delayed type hypersensitivity (DTH) and protective cell mediated immunity showed different profiles with respect to time following intraperitoneal immunization with live Salmonella enteritidis. Whereas the DTH response decreased with time the Protection Index increased. The decline in DTH response was found to be associated with suppressor cells generated by intraperitoneal immunization and could be prevented by cyclophosphamide treatment prior to immunization. It was concluded that the two effector responses of cell mediated immunity were under differential regulation.

Clinical profiles, serological markers, and antibody responses to antigens of hepatitis B virus (HBV) were studied in patients with fulminant viral hepatitis. Whereas hepatitis A and B were found to be uncommon causes (6.9% and 12.2%, respectively), non-A, non-B (NANB) hepatitis was found to be the most common cause of fulminant hepatitis (80.9%). As against this, the incidence of hepatitis B and NANB hepatitis was very similar in nonfulminant acute viral hepatitis in adults (41.2% and 51.9%, respectively). Pregnancy with labour was an important precipitating factor for development of fulminant hepatitis of the NANB type only; 32% of fulminant NANB hepatitis patients were pregnant women and 22.6% had a history of labour preceding hepatic coma. Only 0.8% of nonfulminant NANB hepatitis cases were pregnant women. Another major precipitating factor for the development of the fulminant form of NANB hepatitis was concomitant chronic HBV carrier state. A total of 38.6% of fulminant NANB hepatitis patients were HBV carriers, whereas only 19.2% of nonfulminant acute NANB hepatitis cases were HBV carriers. Sera of 32 chronic HBV carriers with fulminant NANB hepatitis and 10 cases of fulminant hepatitis B were tested for delta antibody, and all were nonreactive. The antibody responses of the fulminant hepatitis B patients to the antigens of HBV were found to be greater compared to those of patients with nonfulminant acute hepatitis B. Antibody responses of chronic HBV carriers with fulminant NANB hepatitis to antigens of HBV were found to be depressed in comparison with those of chronic asymptomatic carriers.

Cross-reactivity in a delayed-type hypersensitivity (DH) response was studied in mice immunised with live Salmonella typhi, S. paratyphi A and S. paratyphi B. Extensive cross-reactions outside the serogroup limits were observed. The ability of DH cross-reacting and non-cross-reacting sonicates to generate activated macrophages was studied in mice immunised 3 months earlier with S. paratyphi B. Whereas DH cross-reacting S. poona sonicate generated activated macrophages the non-cross-reacting S. typhi sonicate did not. To determine whether infections due to diarrhoea-causing salmonellae generated cross-reactive cell-mediated immune responses against enteric fever-causing organisms, similar reverse experiments were performed in mice immunised with S. enteritidis. S. paratyphi A sonicate generated both effector responses, i.e., DH and activated macrophages.

Effector responses of cell-mediated immunity were studied in mice immediately and 3 months after immunisation with different kinds of salmonella vaccine. Whereas immunisation with live Salmonella enteritidis generated delayed hypersensitivity and activated macrophages, immunisation with formalised vaccine generated neither. Although activated macrophages had declined considerably by 3 months after immunisation, the delayed-hypersensitivity response had persisted. It was possible to generate activated macrophages in these mice by stimulation with sonicate antigens of the immunising strain or Salmonella paratyphic C. Delayed-hypersensitivity responses were elicited in mice, immunised with live S. enteritidis, by sonicate antigens of the homologous strains, S. gallinarum or S. paratyphi C.

Markers of hepatitis B viral infection and the evolution of immune response to these were compared with serum alanine aminotransferase (ALT) levels in adult male and non-pregnant and pregnant female patients with acute hepatitis B from the time of onset of disease to the seventh week. In the adult male and non-pregnant female patients, the peak ALT levels of about 360 IU/litre, seen at the time of onset, gradually declined during the course of the disease. Significantly, even in the seventh week, the median ALT level was abnormal (80 IU/litre). In contrast, the disease was mild in pregnant patients and the ALT levels declined rapidly, returning to normal by the third week. Markers associated with HBV replication, i.e., serum HBV-DNA and HBeAg, declined early in the course of the disease in both groups. The anti-HBc-IgM and anti-HBe responses were well evolved early in the course of the disease in both groups. HBsAg was present in the serum in large amounts (1-1.5 X 10(4) AU/100 microliter) early in the course of the disease and remained so up to the seventh week. Even the pregnant patients who had recovered clinically by the fourth week continued to have HBsAg in their sera in large amounts in spite of normal ALT levels. LMI and LTT responses to HBsAg, which were practically absent in the first week, gradually increased to a peak during the fourth week and remained elevated up to the seventh week in adult male and non-pregnant female patients. In contrast, LMI response to HBsAg was absent in pregnant patients with acute hepatitis B even up to the fourth week Thus, continued liver cell necrosis after the fourth week, as indicated by raised ALT levels, may be associated with T cell responses to HBsAg.

The cross-reactivity of crude sonic extracts of six species of mycobacteria was studied by skin tests in mice and guinea pigs immunised with live Mycobacterium tuberculosis strain H37Rv, and in patients with pulmonary tuberculosis. All slowly growing mycobacteria elicited a strong delayed hypersensitivity response, M. vaccae a moderate response and M. phlei a poor or no response. A specific target antigen for cell-mediated immunity, known to be present in M. tuberculosis, was also present in all the mycobacteria studied. This shared antigen was shown, by immunoprecipitation tests, to be identical in all the slowly growing species but only a partial reaction of identity was obtained with M. phlei. It is concluded that the antigen shared by the slowly growing mycobacteria is immunodominant in cell-mediated immunity.

In studies with the adult-rabbit ileal-loop model, antibodies to the lipopolysaccharide somatic-antigen component of Vibrio cholerae gave passive protection against challenge with live V. cholerae. The antisomatic antibodies had no effect on bacterial proliferation and toxin production either in vivo or in vitro; after challenge, antibody-protected and non-protected rabbit ileal loops developed almost identical amounts of cholera toxin and numbers of V. Cholerae. The protection could be correlated only with a 10-15-fold reduction in the number of V. cholerae adherent to the mucous membrane of the antibody-protected loops. The amount of cholera toxin in the two sets of loops ranged from 1600 to 3200 units. In contrast, when biologically active cholera toxin was prepared in vitro, the amount required to induce ileal-loop secretion was very large (25,600 units). These findings indicate that toxin production by adherent vibrios on the surface of the mucous membrane is an important factor in the pathogenesis of cholera.

In order to investigate regulatory significance of humoral and cellular responses to the idiotypic (Id) determinants on the antibody to hepatitis B surface antigen (anti-HBs), they were studied in acute hepatitis B and in chronic HBV infection. The results were compared with humoral and cellular responses of the same patients to hepatitis B surface antigen (HBsAg). In acute hepatitis B, the responses to HBsAg, were delayed until 3-4 weeks after the onset of clinical symptoms. However, the leucocyte migration inhibition (LMI) and the lymphocyte transformation (LTT) responses to affinity purified anti-HBs were found to be evolved very early in the course of acute hepatitis B, though anti-Id antibodies were absent. The majority of chronic HBV carriers showed a poor humoral and cellular response to HBsAg. Ten out of 38 chronic carriers showed anti-Id antibodies which recognized a major cross-reactive idiotype (CRI) on the anti-HBs molecule. Twenty-five out of 38 chronic carriers also showed LMI response to the Id determinants on the anti-HBs. LMI response induced by anti-HBs could be blocked by a specific Balb/c anti-Id antibody which also recognized the CRI. Thus, in both acute and chronic HBV infections, the anti-Id humoral and cellular responses correlated with poor humoral and cellular responses to HBsAg, indicating regulatory significance.

In order to assess the role of the route of immunization on the immunogenicity of killed Salmonella vaccine, mice were immunized with killed S. enteritidis by intraperitoneal (i.p.) and intradermal (i.d.) routes. Whereas the former was non-immunogenic, the i.d. immunization generated an excellent delayed-type hypersensitivity response; further, i.p. immunization could even suppress the subsequent i.d. immunization. Since the peritoneal macrophages (MO) are known to be particularly low in Ia or MHC-class II antigens, so essential for antigen presentation, the non-immunogenicity by i.p. route was thought to be due to their poor presentation efficiency. Poly I: poly C, an interferon inducer, is known to enhance the MHC-class II expression; hence effect of poly I: poly C treatment on the immunogenicity of the killed vaccine by i.p. route was tested and indeed the non-immunogenicity was corrected. Poor efficiency of presentation of S. enteritidis antigen by peritoneal cells and its improvement by prior poly I: poly C treatment was further confirmed by in vitro lymphocyte transformation test using primed T cells and peritoneal cells from normal and poly I: poly C treated mice. Poly I: poly C treatment also enhanced expression of Ia antigens on peritoneal cells.

The relative magnitude by hepatitis A virus (HAV), hepatitis B virus (HBV) and hepatitis Non-A, Non-B virus (HNANBV) was determined in 496 children from three different parts of India suffering from acute viral hepatitis by tests for specific IgM class anti-HAV and anti-HBV antibodies in the serum. HAV, HBV and NANB infections accounted for 55.8 per cent, 20.2 per cent and 23.2 per cent of cases respectively. Hepatitis A largely (59.5%) affected younger children of 1-5 yr. Nearly a third of children affected by NANB hepatitis were additionally positive for HBsAg. The proportions of HAV and HBV infected cases respectively decreased and increased with increasing age whereas the incidence of HNANBV infection remained almost constant throughout childhood. Acute NANB hepatitis, a major health problem in the adults of India is also common throughout childhood. This study suggests that this infection does not impart long lasting protective immunity.

Antigen participating in cell-mediated immunity was isolated from a sonic extract of Mycobacterium tuberculosis H37Rv. The purified antigen had a mol. wt 150 000-200 000, based upon its Rf in 7.5% polyacrylamide gel in the presence of sodium dodecyl sulphate. In simple agarose gel electrophoresis it had an Rf of 0.85. The protein in its native state was identified in the sonic extract with an antiserum to the purified antigen; it too had an Rf of 0.85 in agarose gel electrophoresis. The purified antigen elicited a delayed hypersensitivity response and stimulated the generation of activated macrophages in mice immunised with M. tuberculosis H37Rv.

Swiss white mice were immunized with different mycobacteria and delayed type hypersensitivity (DTH) responses were studied by the foot-pad swelling technique of Gray and Jennings (1955). Extensive cross-reactions in DTH, outside the limits of Runyon's groups were observed. As a general trend slow growing mycobacteria showed greater cross-reactivity with slow growers than with rapid growers and vice versa. The implied cross-protective significance of DTH cross-reactions was further confirmed by demonstration of the ability of DTH cross-reacting sonicates to generate activated macrophages in M. avium immunized mice. An antiserum was raised against the earlier reported DTH eliciting antigen of M. tuberculosis H37Rv (DTH-H37Rv). The sero-reactivity of anti-DTH-H37Rv against the sonicates of different mycobacteria was studied with the objective of investigating the molecular basis of DTH cross-reactivity. Immunoprecipitation reactions of different mycobacterial sonicates with anti-DTH-H37Rv showed that the antigen was shared by all the mycobacteria tested irrespective of their cross-reactivity in a DTH response. All of the slow growers showed reactions of total identity with DTH-H37Rv. However with rapid growers DTH-H37Rv showed only a partial identity. From these data it was concluded that an antigen participating in DTH response is shared by all mycobacteria and that it is polymorphous, having genus specific and group specific (as slow and rapid grower groups) determinants.

Cross-reactivity in the delayed hypersensitivity response to mycobacteria of different Runyon groups was tested in Swiss white mice immunised with live mycobacteria. All the strains tested gave cross-reactions and, generally, slow growers gave stronger cross-reactions with other slow growers than with rapid growers and vice versa. Sonicates of cross-reacting mycobacteria were also tested for their ability to generate activated macrophages in mice immunised with Mycobacterium avium. All the mycobacterial sonicates generated activated macrophages but a sonicate of Salmonella typhi did not. The sonicate of M. tuberculosis H37Rv also generated activated macrophages, which indicates that there might be protective cross-reactions between M. tuberculosis and atypical mycobacteria.

The effects of diverse factors, such as route of immunisation, composition of immunogen and administration of interferon inducer, on the expression of cell-mediated immune responses against Salmonella enteritidis were investigated in BALB/c and Swiss white mice. Immunisation with live cells of S. enteritidis by the intraperitoneal route (ip) generated both delayed type hypersensitivity (DTH) and protective cell-mediated immunity (CMI). However, the two responses showed diametrically opposite time kinetics. The decline and disappearance by 9 weeks after ip immunisation of DTH and the rise of protective immunity in the same period suggested the possibility that the two responses were mediated by different subsets of T cells. Immunisation by the intradermal (id) route with a sonicate of S. enteritidis generated only DTH; id immunisation also suppressed the development of the protective response following ip immunisation with live S. enteritidis. Both responses were not seen when T cells were eliminated with anti-T cell serum. Oral immunisation with live cells of S. enteritidis induced excellent CMI expressing both DTH and protective responses. On the other hand, oral immunisation with the sonicate of S. enteritidis not only did not induce CMI, but also prevented the development of the DTH and protective response to ip immunisation with live S. enteritidis. Induction of interferon by the administration of poly I: poly C for four consecutive days after id immunisation with killed S. enteritidis suppressed the generation of DTH.