Résumé
When hosts are simultaneously infected by different pathogens, the severity of the disease might be altered compared to hosts harboring single infections. The reasons underlying these changes in parasite virulence are manifold. Here, we investigated the importance of order and timing of infection. We used a model of rodent coinfection between two parasites that do not compete for common resources, an intestinal nematode ( Heligmosomoides polygyrus, Hp) and an apicomplexan protozoan ( Plasmodium yoelii, Py). During single infections, Hp induced only mild disease symptoms. Plasmodium produced a substantial reduction in the number of red blood cells but all mice recovered from the infection. A different picture emerged in coinfected hosts. Hp maintained a profile of mostly asymptomatic infection when infecting hosts that had been previously infected with Py. On the contrary, Py incurred substantially higher costs in hosts that had been previously infected with Hp. We then investigated the possible reasons underlying the increase of Py virulence in hosts that had been previously infected with Hp. We found that coinfected hosts were less able to control Py multiplication and to recover from infection-induced anemia. Coinfected hosts had similar levels of erythropoietin and similar renewal of lost red blood cells compared to single Py infected hosts, resulting in decreased tolerance to Py infection. Experimental administration of erythropoietin in coinfected (Hp infecting first) hosts, partially decreased the severity of disease symptoms and improved tolerance. The detoxification of free heme released during the lysis of red blood cells, and the expression of Th1 and anti-inflammatory cytokine genes were also similar between coinfected and single infected hosts. However, coinfected mice had higher proportions of regulatory T cells expressing the CTLA-4 immune checkpoint, suggesting an enhanced immunosuppressive activity of Tregs. Py infection also induced the exhaustion of CD8 + T cells, as coinfected mice had higher proportions of both PD-1 + and LAG-3 + CD8 + T cells, and an increase in the CD4 + /CD8 + ratio. Overall, these results stress the importance of the order of infection as a major determinant of malaria severity in hosts harboring a gastrointestinal nematode infection. We discuss the possible epidemiological and evolutionary consequences of these results.