Abstract
Alcoholism increases susceptibility to fatal infections caused by Streptococcus pneumoniae, or the pneumococcus (PNC). Alcoholics with PNC pneumonia have an increased likelihood of developing bacteremia, and alcohol ingestion in combination with leukopenia increases the likelihood of fatality due to systemic PNC infection. We have shown previously that ethanol-ingesting rats, like their human counterparts, develop prolonged bacteremia and have a higher mortality after transtracheal PNC infection. All clinical PNC isolates produce the toxin pneumolysin that has been shown to reduce PNC bloodstream clearance. The goal of this study was to determine whether pneumolysin production hampers the removal of PNC from the bloodstream of the alcohol-ingesting host. The pair-feeding regimen employed in our experiments has been shown previously to induce leukopenia in both Ethanol-fed and Pair-fed rats. This enables us to specifically examine the effects of ethanol ingestion in resistance to PNC bacteremia in the leukopenic alcoholic host. Male Sprague-Dawley rats were pair-fed a liquid diet containing 36% of its calories as either ethanol or dextrin-maltose for 7 days. Chow-fed rats were included to control for nutritional deprivation inherent in this feeding model. On day 8, rats were infected intravenously with 2 x 105 colony forming units (cfu) / ml blood of either a PNC strain producing wildtype pneumolysin (H+C+) or an isogenic pneumolysin- negative mutant (PLY-). Mortality was followed for 14 days. Clearance of the PNC from the bloodstream (as determined by plate counts) was also analyzed at 0.25 to 27 h post-infection. In general, pneumolysin production reduced PNC clearance and increased total mortality significantly for all animals except those in the Chow-fed group. Surprisingly, Ethanol-fed rats cleared both the PNC strains and survived infection at least as well as the Pair-fed and Chow-fed controls. In contrast, Chow-fed rats infected with the H+C+ strain had significantly higher bacteremia counts at 27 h post-infection than rats in the other diet groups. Although nearly all H+C+ infected animals died, the mean day of death for Chow-fed rats was 2.3 ± 0.4 days compared to 6.7 ±1.4 and 9.2 ±1.6 days for the Ethanol- and Pair -fed rats, respectively. Furthermore, 86% of the PLY - infected Chow-fed rats died by day 14, whereas at least 80% of both Ethanol- and Pair-fed rats survived PLY- infection. To determine whether pnuemolysin production or ethanol ingestion influences the trafficking of PNC through the organs of the reticuloendothelial system, numbers of organisms in the lung, spleen and a lobe of the liver were determined 2 h post-infection. Although there were no differences in organ burdens among the diet groups, each diet group had fewer PLY- organisms than H+C+ organisms in their spleens, possibly suggesting that the PLY- organisms may be more susceptible to killing. In order to further understand the increased susceptibility of the Chow- fed rats to the H+C+ strain, serum levels of TNFa, IL-1 p, IL-6 and IL-10 were analyzed by ELISA at 15 min, 2 h and 27 h post-infection. Levels of all four cytokines were minimal early after infection in all feeding groups infected with either strain. Levels of IL-1 p, IL-6 and IL-10 were mildly elevated 27 h after infection with the PLY-strain, with all groups having similar levels. By contrast, 27 h after infection with the H+C+ strain, serum levels of TNFa, IL- 1 p, IL-6 and IL-10 were all significantly higher in Chow-fed than in Ethanol- or Pair-fed rats. This is consistent with the rapid deaths seen in H+C+ infected Chow-fed rats. Since ethanol ingestion and malnutrition have been shown to be associated with increased corticosterone production, we hypothesized that the immunosuppressive effects of this hormone may have suppressed the inflammatory response in the Ethanol- and Pair-fed rats. This would explain why they do not experience the overwhelming lethal cytokine response seen in the Chow-fed rats. To determine if corticosterone differed among the feeding groups, a radioimmunoassay was performed on serum samples taken preinfection, 2 h post-infection and 27 h post-infection to quantify serum corticosterone levels. Although there were minor and conflicting differences that varied with diet group and infecting strain at 2 h post-infection, there were no significant differences in serum corticosterone levels among the feeding groups pre-infection or at 27 h post-infection. However, both Ethanol-fed and Pair-fed rats had significantly reduced corticosterone levels at 2 h and 27 h post-infection with the H+C+ strain in comparison to their pre-infection levels. In contrast, Chow-fed rats had significantly lower levels of corticosterone 2 h post-infection with the PLY- strain. Because we only have corticosterone levels for three rats per diet group at 27 h post-infection, it is difficult to ascertain the exact role of corticosterone after the initial hours in our model of infection. In conclusion, although pneumolysin production contributed to PNC virulence for rats in all feeding groups, ethanol ingestion was not associated with more severe disease.|Furthermore, Chow-fed rats are inherently more susceptible to PNC bacteremia in general than rats in either of the other diet groups, although the reason for this is not apparent from our results. Our results suggest that caloric restriction may contribute to increased resistance to fatal PNC bacteremia.