Concomitant influence of helminth infection and landscape on the distribution of Puumala hantavirus in its reservoir, Myodes glareolus

Abstract

Background: Puumala virus, the agent of nephropathia epidemica (NE), is the most prevalent hantavirus in Europe. The risk for human infection seems to be strongly correlated with the prevalence of Puumala virus (PUUV) in populations of its reservoir host species, the bank vole Myodes glareolus. In humans, the infection risks of major viral diseases are affected by the presence of helminth infections. We therefore proposed to analyse the influence of both helminth community and landscape on the prevalence of PUUV among bank vole populations in the Ardennes, a PUUV endemic area in France.

Results: Among the 313 voles analysed, 37 had anti-PUUV antibodies. Twelve gastro-intestinal helminth species were recorded among all voles sampled. We showed that PUUV seroprevalence strongly increased with age or sexual maturity, especially in the northern forests (massif des Ardennes). The helminth community structure significantly differed between this part and the woods or hedgerows of the southern cretes pre-ardennaises. Using PUUV RNA quantification, we identified significant coinfections between PUUV and gastro-intestinal helminths in the northern forests only. More specifically, PUUV infection was positively associated with the presence of Heligmosomum mixtum, and in a lesser extent, Aonchotheca muris-sylvatici. The viral load of PUUV infected individuals tended to be higher in voles coinfected with H. mixtum. It was significantly lower in voles coinfected with A. muris-sylvatici, reflecting the influence of age on these latter infections.

Conclusions: This is the first study to emphasize hantavirus--helminth coinfections in natural populations. It also highlights the importance to consider landscape when searching for such associations. We have shown that landscape characteristics strongly influence helminth community structure as well as PUUV distribution. False associations might therefore be evidenced if geographic patterns of helminths or PUUV repartition are not previously identified. Moreover, our work revealed that interactions between helminths and landscape enhance/deplete the occurrence of coinfections between PUUV and H. mixtum or A. muris-sylvatici. Further experimental analyses and long-term individual surveys are now required to confirm these correlative results, and to ascertain the causal links between helminth and PUUV infection risks.

Figure 1

Sampling localities for M. glareolus in the French Ardennes. Forests and wooded areas are indicated in grey. White circles correspond to forested areas of the Northern massif des Ardennes. White and dashed circles respectively correspond to wooded areas and hedge networks of the Southern crêtes pré-ardennaises. The dashed line indicates the limit between the Northern massif des Ardennes and the Southern crêtes pré-ardennaises. Numbers refer to site codes indicated in Table 1.

Figure 2

Relationships between the mass (g) of bank voles and their seroprevalence with regard to PUUV (0: no anti-PUUV antibodies detected, 1: anti-PUUV antibodies detected) for each landscape configuration. Grey bars represent data from the Northern sites (massif des Ardennes) and dashed bars correspond to the Southern sites (crêtes pré-ardennaises).

Figure 3

Correspondence analysis of the helminth community structure. a) Factorial plan (F1 × F2) showing the relationships between the helminth species. b) Factorial plan of the landscape according to its effect on the helminth community. The grey circles represent the gravity centres of the three landscapes considered, forest (F), wood (W) and hedge network (H). The lines show the variation within each site. c) Schematic representation of the site map based on helminth community characteristics. Sites represented with circles have above average F1 factorial values, whereas sites represented with squares have below-average F1 factorial values. Hedge networks are indicated with black dashed lines. Circle or square sizes are proportional to the distance of the value above or below the average value.

Figure 4

Results of the discriminant analysis performed on the helminth community of PUUV-seronegative and PUUV-seropositive bank voles sampled in the northern sites of the transect. a) Sample scores of the discriminant function for PUUV-seronegative and PUUV-seropositive bank voles. The symbols (-) and (+) represent the group averages of these two classes of individuals. b) Coefficient of the discriminant scores on this axis.

Figure 5

Comparison of PUUV viral load in bank voles infected with H. mixtum or A. muris-sylvatici and in those not infected by these helminth species. "0" indicates bank voles that are not infected with H. mixtum (resp. A. muris-sylvatici) and "1" indicates bank voles that are infected with at least 1 H. mixtum helminth (resp. A. muris-sylvatici). Only samples from the massif des Ardennes are considered. N indicates the sampling size for each category.