TO THE EDITOR Powassan virus (POWV) is an emerging North American tick-borne flavivirus transmitted to humans by infected tick bites. Ticks transmit pathogens during the complex feeding process of penetrating the skin and stay attached for several days to acquire blood. This process is facilitated by a repertoire of pharmacologically active proteins/factors in tick saliva (Ribeiro et al., 2006; Kazimirova and Stibraniova, 2013). Thus, skin acts as the interface of the host–pathogen–vector interactions (Wikel, 2013). Skin provides the first line of defense against mechanical and environmental damage and infectious agents (Nestle et al., 2009). In a previous study, which examined cutaneous bite-site lesions from uninfected Ixodes scapularis nymphs, a rapid, proinflammatory progression of the early host response was identified, culminating in the infiltration of innate immune cells by 12 hours after tick infestation (Heinze et al., 2012). Successful transmission of tick-borne POWV has been shown to occur within 15 minutes of I. scapularis attachment (Ebel and Kramer, 2004). In addition, it was demonstrated that during early feeding time points the viral load in the tick salivary glands increases (Alekseev and Chunikhin, 1990). Therefore, the early cutaneous interactions between host immunity and initial tick-mediated immunomodulation are central to successful disease-causing agent transmission. In this study, we sought to characterize tick-induced changes in cutaneous gene expression at the early stages of attachment and feeding by POWV-infected I. scapularis nymphs. This will allow us to demonstrate the effect of a tick-borne virus on immune response at the tick–host interface. In this study, we generated POWV-infected I. scapularis nymphs by synchronous infection (McNally et al., 2012) and allowed them to feed on 6-week-old female Balb/C mice. Uninfected ticks were used as control. Each treatment group consisted of four mice, each with a capsule containing one tick. At least three out of four mice had successful tick attachment at each time point, providing us with sufficient sample sizes to perform statistical analyses. Three and six hours after tick attachment (hours post infection, h.p.i.), 4 mm mouse skin biopsies were harvested along with the feeding ticks. Ticks and skin were checked for POWV infection, and all the infected ticks/skin biopsies used in this experiment contained POWV RNA. Total RNA was extracted from each skin biopsy and cutaneous immune responses were analyzed by pathway-specific PCR arrays (Supplementary Table S1 online). In total, 456 genes were analyzed with these arrays. Relative fold differences of the immune genes were calculated as previously described (Heinze et al., 2012). These data were then uploaded to ingenuity pathway analysis software for further analysis. Comparative analysis between POWV-infected and uninfected tick attachment sites at 3 and 6 h.p.i. was performed (Supplementary Table S2 online). When all significantly modulated (P⩽0.05) host genes in the uninfected versus POWV-infected 3 h.p.i. tick-feeding sites were taken into account, there were 40 upregulated genes and 11 downregulated genes (Figure 1a). Of all significantly modulated host genes in the 6 h.p.i. uninfected versus POWV-infected tick-feeding sites, 13 were upregulated and 46 were downregulated. Figure 1 An overview of host gene modulation at 3 and 6 hours post infection (h.p.i.). (a) Number of significantly up- and downregulated genes measured at 3 and 6 h.p.i. during tick infestations of mice with Powassan virus (POWV)-infected Ixodes scapularis nymphs. ... The majority of modulated genes in the 3 h.p.i. comparison were significantly upregulated and ‘inflammatory response' was the top associated reaction at this time point. Several proinflammatory cytokines, such as IL1B and IL6, were significantly upregulated (Figure 2a). IL36A, which is involved in the positive regulation of IL6 production, was also upregulated. IL1B, IL6, and IL36A all influence the quantity of phagocytes and neutrophils during the inflammatory response (Fielding et al., 2008; Rider et al., 2011). TLR4 is another molecule, which was significantly upregulated and associated with the inflammatory response (Figure 2d). Similar to IL1B and IL6, TLR4 is linked to innate immunity. It has been shown that the TLR4 signaling pathways are involved in the innate immune response to viral infections, leading to the induction of additional proinflammatory cytokines (Okumura et al., 2010). In addition, CCR3 was upregulated and it contributes to the chemotaxis of lymphocytes and eosinophils (Fahy et al., 2001). To various extents, IL1B, IL6, IL36A, TLR4, and CCR3 all help establish the proinflammatory environment, which is generated by the chemotaxis of certain immune cells. These relationships are best shown in Figure 1c network, which illustrates their central role in inflammatory response regulation during POWV-infected tick feeding at 3 h.p.i. Figure 2 Changes in the host immune genes expressed in response to Powassan virus (POWV)-infected versus uninfected tick feeding. Genes that were modulated at both 3 hours post infection (h.p.i.) and 6 h.p.i. were compared for difference in fold-change expression ... In contrast to 3 h.p.i. comparison, the majority of significantly modulated genes at 6 h.p.i. were downregulated, including several proinflammatory cytokines associated with the inflammatory response reaction: IL1B, IL18, IFNγ, and tumor necrosis factor (TNF; Figure 2a). In Figure 1d, IFNG, IL1B, and TNF are the most connected to other molecules in this network, indicating that these three cytokines are heavily influenced by molecules that regulate the inflammatory response and cell-to-cell signaling during POWV-infected tick feeding at 6 h.p.i. Although the majority of genes modulated at 6 h.p.i. were downregulated, CCL2 was slightly upregulated (Figure 2a). As CCL2 has chemotactic activity for monocytes and basophils, this suggests that such immune cells are being recruited to the bite site after a POWV-infected tick has been feeding for 6 hours. Ingenuity pathway analysis identified several significantly modulated molecules, which can either be classified under cell death and apoptosis (Figure 2b) or under signal transduction (Figure 2c). The majority of all molecules in Figure 2b and c were upregulated at 3 h.p.i. and downregulated at 6 h.p.i. Trem1 (triggering receptor expressed on myeloid cells 1) and Traf6 (TNF receptor-associated factor 6) are two molecules associated with signal transduction. In Figure 1d network, it is evident that the 3 h.p.i. upregulation of these molecules is closely tied to proinflammatory cytokines such as IL1B and IL6. Specifically, Trem1 stimulates the release of such proinflammatory cytokines, which in turn amplify the neutrophil and monocyte-mediated inflammatory response. Traf6 is a protein that facilitates signaling from the Toll/IL1 family and the TNF receptors. In response to proinflammatory cytokines, Traf6 transduces signaling in the NF-κB pathway (Wong et al., 1998). Traf6 and the NF-κB pathway are also linked with iNOS (inducible nitric oxide synthase) signaling. NOS2 (inducible nitric oxide synthase 2) is a reactive free radical, which acts as a biologic mediator in antimicrobial processes. It is inducible by a combination of lipopolysaccharide and certain cytokines such as IFNγ and TNF (Lau et al., 1995). IFNGR2 and TNF were both significantly upregulated (Figure 2a and d), linking these cytokines to the host's induction of NOS2 after a POWV-infected tick has been feeding for 3 hours. This NOS2 pattern is consistent with the overall 3 h.p.i. upregulation of other molecules associated with cell death (Figure 2b). Finally, our PCR array data demonstrated that at both 3 and 6 h.p.i., IL2 and IL4 were downregulated (Figure 2a). Therefore, our data suggest that the host's induced immune response to POWV-infected tick feeding does not have a defined Th1 or Th2 profile. This further supports our conclusion that a complex proinflammatory environment exists, with increased granulocyte recruitment, migration, and accumulation, specifically of neutrophils, at the POWV-infected tick feeding loci. Macrophages were also predicted to undergo apoptosis. Comparing POWV-infected with uninfected tick feeding loci at 6 h.p.i. predicts decreased recruitment of neutrophils and phagocytes. Our data clearly indicate that POWV-infected tick feeding recruits immune cells much earlier than the uninfected tick feeding (Figures 1 and and2).2). This could be directly attributed to POWV infection or changes in tick saliva secretion, or a synergistic effect of both. Further research needs to be conducted to elucidate this phenomenon. To our knowledge, this is the first report of the early cutaneous response during POWV transmission.