In the context of low malaria transmission, the current methods used to evaluate the intensity of transmission, such as EIR or Plasmodium parasitemia, present substantial limitations. Alternative methods to estimate Anopheles density and human exposure would be of great value, allowing epidemiological studies when the use of classical methods may not be relevant such as in low transmission settings. In this respect, our present study investigated whether the gSG6-P1 salivary peptide could be a sensitive and reliable biomarker allowing the detection of micro-geographical heterogeneity of human exposure to Anopheles bites in particular settings.
Our results showed that the IgG Ab levels to gSG6-P1 peptide and the percentage of immune responders varied between the five studied villages. These results suggest that the immune response to gSG6-P1 salivary peptide could identify villages more at risk of malaria than others even in an area presenting low exposure to vectors. The high heterogeneity of exposure to Anopheles bites observed in the studied villages may be explained by the presence of different surrounding landscapes among them [39, 40] and/or the proximity of the river. Indeed, several studies have shown a positive correlation between malaria transmission and the distance to a river as a potential breeding site [41, 42]. Nevertheless, this factor is not sufficient to explain the variation of Anopheles exposure between villages. In fact, Agniam, Guede and Niandane, three villages located near the river, presented different levels of IgG responses to gSG6-P1. A possible explanation of the high human-Anopheles contact in Agniam compared to Guede, for instance (villages only separated by the river) could be the influence of man-made conditions. In this area, Ndiath et al. have shown that Anopheles density variation was more related to the presence of ditch water used for gardening, rice cultivation, manufacture of bricks or animal watering, than the proximity of the river . Thus, Anopheles exposure could be influenced by local human activity and/or household factors which may favor the development of artificial breeding sites, allowing the emergence of adults . Overall, these results confirm that the gSG6-P1 peptide could measure the real human-vector contact and is sensitive enough to detect small-scale variations of vector bites in areas with very low-exposure [32, 33, 43].
However, inconsistencies observed between immunological parameters and entomological results in some villages (e.g.: Agniam vs. Guede) and/or according to seasons suggest that other factors could be taken into account in measuring the heterogeneity of man-vector contact. The use of vector control tools (spray, ITNs or LLINs, personal protection…) as well as household characteristics (traditional, modern…) could significantly reduce the human-vector contact. Moreover, individual exposure to Anopheles was evaluated by HLC (using adult volunteers) and therefore, could present considerable limitations for evaluation of entomological exposure in children [44, 45]. Entomological data was also collected in limited number of areas (five randomised houses/village) and cannot represent the individual exposure to Anopheles bites and its micro-geographical variations. The use of the gSG6P1 biomarker could therefore be relevant and useful for assessing human exposure to Anopheles bites at the population and individual level.
We also reported that both IgG Ab levels and the percentage of immune responders to the salivary biomarker varied according to the season and remained high in October corresponding to the peak of exposure to Anopheles bites. This seasonal variation of specific IgG response was associated to the similar variations of entomological data, as has been reported in previous studies [31, 38, 46]. Moreover, the drop of IgG response observed in January 2009, only three months after the peak of exposure (October 2008), confirmed that the anti-gSG6-P1 IgG responses were short-lived, decreasing after a few months of no exposure [33, 47]. These observations indicate that the Ab responses to gSG6-P1 antigen are transient and sensitive to the seasonal variations of human exposure to Anopheles bites. Similar seasonal fluctuations of Ab to gSG6-P1 were reported in some other epidemiological studies taking place elsewhere in Africa [33, 47]. Indeed, a drop in the anti-saliva IgG response has been previously described in soldiers exposed to An. gambiae, three months after their return from a travel period in tropical Africa  and in children from Angola 6 weeks after the interruption of exposure by bednet use . This concomitant variation of IgG response according to season and consequently to Anopheles exposure indicates that the gSG6-P1 biomarker could be used as an alternative tool when trapping methods are difficult to apply, are fastidious and not applicable at large scale [49, 50], particularly in low or very low endemic areas.
In addition, we observed village-dependent differences between 2009 and 2008. Indeed, specific IgG Ab levels to gSG6-P1 significantly increased only in Agniam, Pendao and Guede in October 2009 compared to June 2009, but not in Niandane village. Moreover, specific IgG responses were lower in October 2009 compared to October 2008 in Agniam and Niandane, whereas entomological data indicated that Anopheles populations remained similar at these two time periods. As gSG6-P1 biomarker has clearly been shown as a pertinent indicator for measuring the efficacy of LLINs , the present data could indicate a change in human-vector contact between both periods. Indeed, human-vector contact may be influenced by several human or environmental factors . The large scale implementation of LLINs conducted by the NMCP in June 2009  could explain the reduced probability of being bitten by Anopheles and partly explain the observed immunological results. However, it is necessary to indicate that despite this implementation of LLINs, some individuals presented gSG6-P1 specific IgG responses indicating that they were still exposed to Anopheles bites. Altogether, we can hypothesize that the observed differences between villages could be due to a different distribution, owners or real use of LLINs by children according to villages. For instance, it could be hypothesized that the efficacy of LLINs implementation and use by children could be higher in Niandane village, and also in Agniam to a lesser extent, compared to other villages. Some factors such as genetics, nutritional status, population displacement (holidays spent in an area of low/high exposure for school-age children), micro-climatic and micro-habitat variations from the studied villages, could not be excluded to explain the observed differences. Nevertheless, these data suggest that the gSG6-P1 biomarker could represent an alternative tool for evaluating the effectiveness of vector control strategies by The National Malaria Control Programmes and variations of effectiveness between villages and environmental and epidemiological contexts [38, 43]. Future studies evaluating the impact of LLINs on malaria transmission in children could be performed by a multi-disciplinary approach where immune response to salivary biomarker would be integrated.