Immune responses to plague infection in wild Rattus rattus, in Madagascar: a role in foci persistence?

Abstract

Background: Plague is endemic within the central highlands of Madagascar, where its main reservoir is the black rat, Rattus rattus. Typically this species is considered susceptible to plague, rapidly dying after infection inducing the spread of infected fleas and, therefore, dissemination of the disease to humans. However, persistence of transmission foci in the same area from year to year, supposes mechanisms of maintenance among which rat immune responses could play a major role. Immunity against plague and subsequent rat survival could play an important role in the stabilization of the foci. In this study, we aimed to investigate serological responses to plague in wild black rats from endemic areas of Madagascar. In addition, we evaluate the use of a recently developed rapid serological diagnostic test to investigate the immune response of potential reservoir hosts in plague foci.

Methodology/principal findings: We experimentally infected wild rats with Yersinia pestis to investigate short and long-term antibody responses. Anti-F1 IgM and IgG were detected to evaluate this antibody response. High levels of anti-F1 IgM and IgG were found in rats one and three weeks respectively after challenge, with responses greatly differing between villages. Plateau in anti-F1 IgM and IgG responses were reached for as few as 500 and 1500 colony forming units (cfu) inoculated respectively. More than 10% of rats were able to maintain anti-F1 responses for more than one year. This anti-F1 response was conveniently followed using dipsticks.

Conclusion/significance: Inoculation of very few bacteria is sufficient to induce high immune response in wild rats, allowing their survival after infection. A great heterogeneity of rat immune responses was found within and between villages which could heavily impact on plague epidemiology. In addition, results indicate that, in the field, anti-F1 dipsticks are efficient to investigate plague outbreaks several months after transmission.

Figure 1. Short term kinetic of anti-F1 IgM in R. rattus after inoculation with Y. pestis.

Wild healthy rats (Rattus rattus) were inoculated with different doses of living Y. pestis (0 to 30,000 cfu) and blood sampled over one month. Anti-F1 IgM was detected by a sandwich ELISA. Results were expressed as the ratio of the mean optical density (OD) of the sample on the mean OD obtained for negative control rats +3 SD. Rats are from two villages Ambohimasina and Maromanana. (median; bar: 10% and 90% percentiles) A/Time course of the median OD ratio of each group: OD ratio peaked at Day 13 and decreased rapidly after B/Median of OD ratio obtained at Day13 and Day 25 for the two villages according to the dose inoculated. Rats from Ambohimasina produced a stronger IgM response against Y. pestis than Maromanana rats C/Median fraction of maximal anti-F1 IgM titer (OD ratio, mostly at Day 13) remaining at Day 18 and Day 25 according to the dose inoculated. For each rat the maximal OD ratio is noted as well as the OD ratio at Day 18 and Day 25, the fractions are calculated and the median of each group is plotted according to the dose inoculated. The speed of decay of IgM increases with the dose inoculated. D/Percent of rats remaining seronegative after inoculation. For each dose the fraction of rats remaining negative for anti-F1 IgM at Day 13 and IgG at Day 25 is plotted. It decreases from 15 to 30,000 cfu.

Figure 2. Short term kinetic of anti-F1 IgG in R. rattus after inoculation with Y. pestis.

Wild healthy rats (Rattus rattus) were inoculated with different doses of living Y. pestis and blood sampled over one month. Anti-F1 IgG was detected by ELISA. Results were expressed as the ratio of the mean optical density (OD) of the sample on the mean OD obtained for negative control rats +3SD. (bar: 10% and 90% percentiles) A/Time course of the median anti-F1 IgG OD ratio of each group: OD ratio peaked between Day 13 and Day 25 and remained stable. Rats are from the two villages Ambohimasina and Maromanana B/Median anti-F1 IgG OD ratio at Day 13 and Day 25 according to the dose of Y. pestis inoculated. OD reaches a plateau for 500 cfu. Rats are from the two villages Ambohimasina and Maromanana C/Scatter plot of the OD ratio (IgM at Day 13/IgG at Day25) of rats from Ambohimasina and Maromanana. Logarithmic regression highlighted positive correlations between IgM and IgG responses to Y. pestis, (for Maromanana (solid line) Day 25-IgG  = 0,5322+8,7431*log10(Day 13-IgM); for Ambohimasina (dashed line) Day 25-IgG  = 0,4428+6,4026*log10(Day 13-IgM) ). D/Comparison of the dose/anti-F1 IgG titer relation for four different villages. For each group of dose, median OD ratio at Day 13 and Day 25 are plotted. Different sensitivity of the rats according to the village can be shown.

Figure 3. Long term kinetic of anti-F1 IgG in R. rattus after inoculation with 125 cfu of Y. pestis.

Rats captured in three villages were inoculated with 125 cfu of living Y. pestis and blood sampled over 12.5 months using seropads. Anti-F1 IgG measured by ELISA delta OD are plotted. A/Mean of delta OD of survival rats were plotted according to time. Columns represent the number of deaths registered at each time point. Data are separated in two groups, i.e. rats with long-lasting IgG response and those with short-lasting response. Almost no death was registered in the long-lasting group. (n: number of remaining rat in the group at time point) B/Same plot but with data separated according to the villages. No difference between villages was found C/Same plot according to the capture site of the rats showing no difference.

Figure 4. Comparison of detection of anti-F1 IgG by dipstick and ELISA methods.

A/Seropads obtained from the long-term follow up are analyzed using both dipstick and ELISA The percent of positive samples by dipstick is plotted according to the delta OD obtained by ELISA. Both percent per class and cumulative percent of positive samples are plotted. B/Comparison of the percent of positive rat detected using ELISA and the expected positive detection using dipsticks. Using the relationship shown in A and the delta OD values represented in Figure 3A, we predicted the percent of positive rats expected using the dipstick method.