INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Clostridium piliforme is the agent of Tyzzer's disease. Although C. piliforme is classified as a Clostridium sp. based upon 16S rRNA analysis, it is unlike other clostridia in several respects. The organism is an obligately intracellular pathogen, consistently stains gram negative, and is positive for lipopolysaccharide by the Limulus amebocyte lysate assay (34). In many animal species, C. piliforme infections induce rapidly fatal enterohepatic disease (26, 35, 37, 38, 42), and the bacterium has recently been reported as a cause of subcutaneous infection in immunocompromised humans (14, 31). However, infections in mice are most often subclinical (26). Factors which mediate host resistance to clinical C. piliforme are not well defined but apparently involve both bacterial and host factors (8, 19, 30, 33, 36).

Bacterial factors that influence disease severity are not well characterized but may include recently described toxins (30). Previous studies in our laboratory have documented that the severity of experimental murine C. piliforme infection varies with the toxicity of the bacterial isolate (34). Toxic isolates induce pronounced hepatic lesions with histologically discernible bacteria up to 14 days postinoculation, whereas nontoxic isolates rarely induce lesions or histologically visible bacteria (34).

Several clinical and experimental findings suggest that the host immune system may also regulate C. piliforme expression. For example, susceptibility to Tyzzer's disease varies with the host's strain, age, and immune status; the severity of the disease is greatest in weanling and immunocompromised animals (19, 37). Although murine C. piliforme infection is often subclinical, mouse strains differ in susceptibility to disease (9, 12, 33, 37). DBA/2 mice are susceptible to C. piliforme infection and typically develop multifocal hepatic necrosis following natural or experimental infection. Conversely, C57BL/6 mice develop only a few small hepatic lesions following exposure to C. piliforme and are considered resistant.

Previous investigators have indicated that murine susceptibility to C. piliforme may be mediated by leukocytes such as B and T lymphocytes and natural killer (NK) cells (19, 33, 36). One of the most important cytokines regulating lymphocyte and NK cell activity is interleukin-12 (IL-12), a heterodimeric 70-kDa protein, which was originally named NK cell stimulatory factor. IL-12 promotes development of the Th1-type response by T lymphocytes and NK cells and is known to mediate resistance to pathogens such as Brucella abortus, Chlamydia trachomatis, Listeria monocytogenes, Candida albicans, Mycobacterium tuberculosis, and Salmonella spp. (3-5, 16, 20, 39, 43). Given the apparent importance of lymphocytes and NK cells in mediating the susceptibility of murine strains to C. piliforme infection (19, 33, 38), this study was undertaken to investigate the role of IL-12 in Tyzzer's disease.

In this study, we examined the IL-12 response to both a toxic (H1) and a nontoxic (M1) C. piliforme isolate in susceptible DBA/2 and resistant C57BL/6 mice. Results of these studies demonstrated that hepatic IL-12 mRNA levels were upregulated and serum IL-12 levels were elevated for at least 28 days by both C. piliforme isolates. Serum IL-12 levels were significantly (P < 0.05) higher in C57BL/6 mice than in DBA/2 mice.

To further evaluate the role of IL-12 in Tyzzer's disease, polyclonal anti-IL-12 antibodies were used to neutralize IL-12 in both mouse strains prior to C. piliforme H1 challenge. While IL-12 neutralization increased the severity of disease in both mouse strains, the increase was significantly (P < 0.05) greater in C57BL/6 mice than in DBA/2 mice. These data document the importance of IL-12 in the murine immune response to C. piliforme challenge and suggest a possible role for IL-12 in mediating the susceptibility of murine strains to infection.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Mice. Female DBA/2 and C57BL/6 mice were obtained at 3 weeks of age from Charles River Laboratories (Wilmington, Mass.). Mice were allowed to acclimate for 3 days prior to use.

Bacteria. The toxic H1 isolate of C. piliforme, which induces experimental liver lesions in mice, was isolated from the livers of hamsters with clinical Tyzzer's disease (30). The nontoxic M1 isolate of C. piliforme, which does not induce experimental liver lesions in mice, was isolated from the livers of subclinically infected mice (30). Both isolates were subsequently maintained in tissue culture (30). For animal inoculation, bacteria were grown on murine liver cells (BNL cell line; ATCC TIB 73) and processed as previously described (33).

Plaque assay. To determine the infectivity of bacteria from tissue culture, plaque assays were performed (17, 33, 40). Plaque assay results indicated 25 to 35% viability among C. piliforme preparations.

Experimental design. Groups of 10 or more DBA/2 mice and 10 or more C57BL/6 mice were inoculated intravenously with 10⁵ C. piliforme H1 or M1 organisms and compared to control animals inoculated intravenously with BNL cells in order to evaluate the in vivo IL-12 response to Tyzzer's disease. Hepatic and blood samples were collected prior to inoculation (day 0) and at days 1, 3, 7, 14, and 28 post-bacterial inoculation. Hepatic samples were snap frozen in liquid nitrogen and stored at 80°C until they were evaluated for IL-12 mRNA production by reverse transcription-PCR (RT-PCR), as described below. Blood samples from at least two (time 0) or four (all other time points) groups of five animals at each time point were pooled, allowed to clot at room temperature, centrifuged at 5,000 × g for serum collection, and stored at 20°C until they were evaluated for IL-12 by enzyme-linked immunosorbent assay (ELISA) as described below.

Reverse transcription for preparation of cDNA. Hepatic tissues collected at necropsy were processed for mRNA isolation as previously described (16). Briefly, 10 to 50 mg of hepatic tissue was pulverized under liquid nitrogen and resuspended in Ultraspec solution (Biotecx Corp., Houston, Tex.), and RNA was extracted with chloroform. RNA concentration and purity were determined spectrophotometrically by using 260 and 280 nm readings. Reverse transcription was performed with 1 µg of total RNA by using oligo(dT) primers (Superscript II Reverse Transcription kit; Gibco BRL Corp., Gaithersburg, Md.) according to the manufacturer's instructions.

Cytokine-specific PCR. Cytokine-specific PCR was performed with 1 µg of cDNA by using primers specific for murine -actin or IL-12 p40. Primers for -actin were synthesized by Integrated DNA Technologies (Coralville, Iowa) by using previously described sequences (2, 27). Primers for IL-12 p40 were synthesized with sequences kindly provided by Steven Kunkel (University of Michigan). The sequences for cytokine primers were as follows: for -actin, 5'-GTGGGCCGCTCTAGGCACCAAGGT-3' and 5'-CTGGATGGCTACGTACATGGC-3', and for IL-12 p40, 5'-CTCACCTGTGACACGCCTGA-3' and 5'-CAGGACACTGAATACTTCTC-3'. Briefly, samples were amplified with Perkin-Elmer reagents and Taq polymerase in 50 µl of PCR solution in thin-walled reaction vials. Samples were denatured at 94°C for 2 min, followed by 35 cycles of 94°C for 30 s, 58°C for 45 s, and 72°C for 2 min. A 7-min final extension was performed at 72°C.

Cytokine ELISA. Serum samples collected at necropsy were quantitatively assayed for total-IL-12 levels by using a commercially available ELISA kit (Genzyme Corp., Cambridge, Mass.) according to the manufacturer's instructions. Briefly, 50 µl of pooled sera was diluted in 50 µl of buffer and loaded into commercially prepared antibody-coated 96-well plates. Serum cytokine levels were determined by comparing the intensity of the ELISA signal to standards of known concentrations of IL-12. Each pooled sample was run at least in duplicate.

Tissue harvest for evaluation of lesions and bacterial loads. Hepatic lesions and bacterial loads were evaluated at day 3 postinoculation as previously described (33). Briefly, gross hepatic surface lesions were scored on a scale of 0 to 10 as follows: 0, no hepatic lesions; 1, 1 to 10 lesions per liver; 2, 11 to 20 lesions, etc. Livers with >90 lesions were given a score of 10. Histologic lesions were evaluated by using 5-µm-thick sections stained with hematoxylin and eosin and were assigned cumulative scores from 35 high-power fields (400×). Histologic lesions were scored on a scale from 0 to 10 as follows: 0, no lesions; 2, 1 to 3 coagulative necrotic lesions per liver; 4, 4 to 6 lesions; 6, 7 to 10 lesions; 8, >10 lesions; 10, >10 lesions more than half of which had caseous necrosis. Attempts in our laboratory to quantitate C. piliforme from infected-liver homogenates by plaque assays have been unsuccessful; however, the large size of C. piliforme (10 to 40 µm in length) allows visualization of individual bacteria within hepatocytes. Therefore, bacterial load was assessed by counting the number of organisms in 35 high-power silver-stained hepatic fields by using the following scale: 0, no bacteria; 2, 1 to 50 bacteria; 4, 51 to 100 bacteria; 6, 101 to 500 bacteria; 8, 501 to 1,000 bacteria; 10, >1,000 bacteria.

Statistical analysis. Data were evaluated for normal distribution by the Kolmogorav-Smirov test. When normally distributed, data were analyzed by one-way analysis of variance. When data were not normally distributed, analysis was performed by analysis of variance of Wilcox ranks. All analyses were performed with SigmaStat (SSPS Corp., San Diego, Calif.) software. Results are reported as means ± standard errors of the means. P values of <0.05 were considered significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Effect of C. piliforme challenge on hepatic IL-12 p40 mRNA expression. IL-12 is a heterodimeric protein (p70) composed of p40 and p35 subunits. Both subunits may be constitutively expressed, but exposure to foreign antigens reportedly induces greater alterations in p40 mRNA expression than in p35 mRNA expression (28). DBA/2 and C57BL/6 mice were injected intravenously with 10⁵ organisms of either the toxic H1 or the nontoxic M1 C. piliforme isolate in order to evaluate whether C. piliforme induces hepatic expression of IL-12 p40 and whether this expression differs between strains of mice or between mice inoculated with different bacterial isolates. Hepatic IL-12 p40 mRNA was not detectable in uninoculated mice or in mice inoculated intravenously with uninfected BNL cells (Fig. 1 and 2). Hepatic IL-12 p40 mRNA expression could be demonstrated at all time points after C. piliforme challenge. The cytotoxic H1 isolate induced a higher level of IL-12 p40 mRNA expression than did the nontoxigenic M1 isolate.

View larger version (58K): [in this window] [in a new window]   FIG. 1.   Results of RT-PCR for hepatic IL-12 p40 (A) and -actin (B) in C57BL/6 mice at days 0, 1, 3, 7, 14, and 28 post-C. piliforme H1 or M1 inoculation and at day 3 post-BNL cell inoculation in control animals. M, molecular weight marker; B, hepatic samples from BNL-inoculated control mice 3 days postinoculation. No cytokine induction was seen at any time in BNL-inoculated control mice. Similar results were obtained from five mice at each time point.

View larger version (53K): [in this window] [in a new window]   FIG. 2.   Results of RT-PCR for hepatic IL-12 p40 (A) and -actin (B) in DBA/2 mice at days 0, 1, 3, 7, 14, and 28 post-C. piliforme H1 or M1 inoculation and at day 3 post-BNL cell inoculation in control animals. No cytokine induction was seen at any time in BNL-inoculated control mice. Similar results were obtained from five mice at each time point.

Effect of C. piliforme challenge on serum IL-12 levels. Serum samples were collected from mice at necropsy to determine whether hepatic IL-12 p40 mRNA expression was related to a systemic increase in total-IL-12 levels in serum. Examination of serum samples by ELISA demonstrated that levels of total (p40 and p70) IL-12 in uninoculated animals were similar to those previously reported for uninfected mice (10) and were consistent with levels reported by the distributor of the ELISA (11). Inoculation of mice with BNL cells did not alter serum IL-12 levels (Fig. 3). However, significant (P < 0.05) elevations in serum total-IL-12 levels were seen by day 1 postinoculation in both mouse strains inoculated with either C. piliforme isolate. Serum total-IL-12 levels were significantly (P < 0.05) higher in C57BL/6 mice than in DBA/2 mice at all times post-H1 inoculation and by day 3 post-M1 inoculation.

View larger version (22K): [in this window] [in a new window]   FIG. 3.   Production of IL-12 following infection of C57BL/6 (C57) and DBA/2 (DBA) mice with either the H1 (A) or the M1 (B) isolate of C. piliforme (means ± standard errors of the means). Serum samples taken from these mice at days 0, 1, 3, 7, 14, and 28 postinoculation and from control (BNL cell-inoculated) mice were evaluated by a cytokine-specific ELISA as described in Materials and Methods. Pooled sera from at least two (time 0) or four (all other time points) groups of five animals were run in duplicate for each time point.

Effect of IL-12 depletion on the course of in vivo C. piliforme infection. Mice were treated with a polyclonal antibody for IL-12 neutralization and were then inoculated with 10⁵ C. piliforme H1 organisms in order to evaluate whether endogenous IL-12 regulates the response to C. piliforme infection. Anti-IL-12 treatment resulted in serum IL-12 levels which were undetectable in both strains of mice and induced significant (P < 0.05) increases in hepatic-lesion and bacterial-load scores compared to scores for control mice (Fig. 4). At day 3 postinoculation, the magnitudes of the increases in lesion and bacterial-load scores were significantly (P < 0.05) greater in C57BL/6 mice than in DBA/2 mice. In IL-12-depleted C57BL/6 mice, gross and histologic lesion scores increased 4.5- and 6.4-fold, respectively, and bacterial-load scores increased 5.6-fold. IL-12 depletion in DBA/2 mice resulted in gross and histologic lesion score increases of 2.4- and 1.8-fold, respectively, while bacterial-load scores increased 3.2-fold.

View larger version (23K): [in this window] [in a new window]   FIG. 4.   Gross hepatic surface lesion scores, histologic lesion scores, and bacterial-load scores in weanling DBA/2 mice treated with anti-IL-12 (n = 16) or IgG (control group; n = 12) (A) and in weanling C57BL/6 mice treated with anti-IL-12 (n = 15) or IgG (control group; n = 14) (B).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

In this study we used an approach similar to that reported by others, in which RT-PCR data were used to qualitatively demonstrate mRNA upregulation while ELISA data were used to quantitate the IL-12 response to C. piliforme (7, 13, 20, 21). These studies demonstrated that C. piliforme induced hepatic IL-12 p40 mRNA expression and increased levels of total IL-12 in serum. Consistent with previous reports by other investigators (13), tissue expression of IL-12 p40 mRNA was not detected preinoculation, in spite of the fact that total IL-12 was detectable in serum. Alterations in hepatic p40 mRNA and total IL-12 were apparent from day 1 postinoculation until day 28 postinoculation, the last point evaluated. Total-IL-12 levels after inoculation with C. piliforme were significantly (P < 0.05) higher in C57BL/6 mice than in DBA/2 mice. Previous investigators have reported that results of ELISAs which detect IL-12 p40 alone or total IL-12 demonstrate close correlations with those of IL-12 bioassays (1, 23). In this study, IL-12 neutralization increased lesion and bacterial-load scores in both strains of mice, but these increases were larger in C57BL/6 mice than in DBA/2 mice. IL-12-depleted C57BL/6 mice developed gross and histologic hepatic lesions and bacterial loads comparable to those typically seen in DBA/2 mice. These data suggest that IL-12 may regulate the susceptibility of murine strains to C. piliforme infection.

It is interesting that the intensities of the serum total-IL-12 response did not differ significantly (P > 0.05) between mice inoculated with the toxic C. piliforme isolate and those inoculated with the nontoxic isolate. Although recent data from our laboratory indicate that both isolates colonize the liver and persist for up to 28 days postinoculation (34), the M1 isolate produces no discernible hepatic lesions or leukocytic influx. In contrast, inoculation of mice with the H1 isolate produces dramatic hepatic lesions and leukocytic responses. Thus, one might have expected higher total-IL-12 levels in H1-infected mice than in M1-infected mice. However, induction of similar total-IL-12 levels by pathogens which differ in virulence is not without precedent. Previous investigators have suggested that with bacteria such as Listeria spp. and M. tuberculosis, levels of total-IL-12 induction may be similar for isolates that differ in virulence or even viability (10, 41). Other investigations have shown that phagocytosis of latex beads by macrophages may induce total-IL-12 levels as high as those seen with viable M. tuberculosis (18). The conclusion from these studies is that phagocytosis itself maybe sufficient to induce macrophage expression of total IL-12 (10, 18).

IL-12 regulates host response to many facultatively and obligately intracellular pathogens, including Coccidioides immitis, L. monocytogenes, Mycobacterium avium, Salmonella spp., Toxoplasma gondii, and B. abortus (4, 5, 15, 22, 24, 32, 43). IL-12 is an important regulator of NK cell activity, and most studies which have demonstrated a role for IL-12 in the murine response to these pathogens also implicate upregulation of NK cells in the control of infection (22, 24, 43). Previous studies in our laboratory documented that NK cells are important in C. piliforme infection and probably play a role in determining the resistance of certain mouse strains to Tyzzer's disease (33). The resistance of C57BL/6 mice and the susceptibility of DBA/2 mice to C. piliforme infection are consistent with this hypothesis, since the NK cell cytotoxic activity of C57BL/6 mice is higher than that of DBA/2 mice, whose NK cell activity is low compared to that in most mouse strains (29). However, NK cell cytotoxic activity does not fully explain the resistance of murine strains to C. piliforme. C57BL/6 mice are one of the most C. piliforme-resistant mouse strains (37); however, their NK cell cytotoxic activity is intermediate among mouse strains (29). C3H mice are relatively susceptible to C. piliforme infection (37), yet their NK cell cytotoxic activity is higher than that in most mouse strains, including C57BL/6 mice. Thus, there is clearly more involved in IL-12-mediated C. piliforme resistance than NK cell activity alone, and the role of IL-12 is likely mediated through several cell types and cytokine responses.

In conclusion, we have shown that C. piliforme inoculation induces prolonged IL-12 upregulation in both C57BL/6 and DBA/2 mice. IL-12 levels remained elevated 28 days postinoculation, regardless of bacterial toxicity or the development of hepatic lesions. The C. piliforme-induced IL-12 response was significantly higher in C57BL/6 mice than in DBA/2 mice, and neutralization of IL-12 resulted in comparable lesion and bacterial-load scores in C57BL/6 mice and non-IL-12-neutralized DBA/2 mice. Based on these findings, it appears that IL-12 is involved in mediating the resistance of murine strains to Tyzzer's disease. These studies also indicate that subclinical C. piliforme infection may have a prolonged impact on IL-12 levels in the host. Such cytokine perturbations have the potential to significantly alter physiologic responses in subsequent research uses of mice. For example, investigators have documented that subclinical infections with pathogens such as L. monocytogenes (6) and C. albicans (20) may prolong host survival and alter immune responses of mice to subsequent experimental infections or transplanted tumors (25). On the basis of this study, it appears that subclinical murine C. piliforme infections may have a significant impact on research mice and on investigations which utilize infected animals, regardless of whether any gross or histologic evidence of infection is ever seen.