Evaluation of cutaneous immune response in a controlled human in vivo model of mosquito bites

Abstract

Mosquito-borne viruses are a growing global threat. Initial viral inoculation occurs in the skin via the mosquito 'bite', eliciting immune responses that shape the establishment of infection and pathogenesis. Here we assess the cutaneous innate and adaptive immune responses to controlled Aedes aegypti feedings in humans living in Aedes-endemic areas. In this single-arm, cross-sectional interventional study (trial registration #NCT04350905), we enroll 30 healthy adult participants aged 18 to 45 years of age from Cambodia between October 2020 and January 2021. We perform 3-mm skin biopsies at baseline as well as 30 min, 4 h, and 48 h after a controlled feeding by uninfected Aedes aegypti mosquitos. The primary endpoints are measurement of changes in early and late innate responses in bitten vs unbitten skin by gene expression profiling, immunophenotyping, and cytokine profiling. The results reveal induction of neutrophil degranulation and recruitment of skin-resident dendritic cells and M2 macrophages. As the immune reaction progresses T cell priming and regulatory pathways are upregulated along with a shift to T_h2-driven responses and CD8⁺ T cell activation. Stimulation of participants' bitten skin cells with Aedes aegypti salivary gland extract results in reduced pro-inflammatory cytokine production. These results identify key immune genes, cell types, and pathways in the human response to mosquito bites and can be leveraged to inform and develop novel therapeutics and vector-targeted vaccine candidates to interfere with vector-mediated disease.

Fig. 1. Graphical representation of study design.

Thirty Cambodian participants were enrolled in the study. Each patient underwent a mosquito feeding (five female Aedes aegypti) for 10 min. Each volunteer underwent a 3-mm biopsy from a mosquito bite site at 30 min, 4 h, and 48 h. Distinct bite sites were biopsied at different time points. Additionally, two 3-mm biopsies of normal skin (NSK) were taken from the opposite forearm at baseline. Blood was collected to obtain serum. Designed with Biorender.

Fig. 2. Clinical observations of skin reaction pre- and post-bite on gross examination and H&E staining.

a Skin immune responses to mosquito bites (left to right) normal skin (NSK), 30 min, 4 h, and 48 h after exposure to five Ae. aegypti mosquitos. b Representative cross-sectional of 3-mm punch biopsies H&E stained at (left to right) 0 h, 30 min, 4 h, and 48 h after exposure to Ae. aegypti bites. Distinct bite sites were biopsied at different time points. Scale bar is 1 mm. Black arrows indicate area of infiltrate. c Magnification of infiltrate indicated by black areas in B, scale bar 100 µM.

Fig. 3. Differential expression gene (DEG) induced by Aedes aegypti mosquito bites over time.

a 3D principal component analysis showing clustering of individual biopsies taken at 30 min (green dots), 4 h (blue dots), 48 h (orange dots) post-bite or normal skin (NSK; pink dots). Blue ellipse indicates clustering at 4 h, while orange ellipse at 48 h. b Volcano plot comparing DEG from skin biopsies post-bite with NSK (significant genes with −1 < log2 fold change (FC) > 1 and FDR < 0.05 are shown in red). FDR were accounted for by adjusting p-values using Benjamini-Hochberg multiple testing correction method (DESeq2 R-package). c Venn diagram summarizing the relationships between significantly upregulated genes over NSK (log2FC > 1 and FDR < 0.05) and time post-bite.

Fig. 4. Pathway analysis resulting from Aedes aegypti mosquito bites over time.

A Chord plot shows significantly upregulated genes at 4 h and 48 h post-bite over NSK (left side of chord plot) and overrepresented pathways (right side of chord plot). Outer left ring shows genes and color range displays log2 fold change (FC) (left, key at bottom) or Reactome terms (outer right ring). Chords connect gene names with Reactome term pathways. B Heatmap (scaled to log2FC) displaying a list of DEGs based on the common pathways overrepresented at 4 and 48 h post-bite.

Fig. 5. Early innate immune responses in the skin are characterized by increased proportions of DCs and M2 macrophages, and a decreased proportion of NK cells at 4 h post-bite.

a Pie charts showing changes in the frequencies of skin immune cells during early innate immune response to mosquito bite at 30 min and 4 h after exposure. Frequencies are reported as percentages of total leukocytes (CD45⁺). b Total frequency of dendritic cells calculated as the sum of Langerhans (CD207⁺), dermal (CD1c⁺), and plasmacytoid (CD123⁺) DCs. c, d Frequency of plasmacytoid dendritic cells and Langerhans cells populations reported as percentage of total CD45⁺ cells shows significant increase of the latter, at 4 h post mosquito bites. e Frequency of activated dermal DCs (CD1c⁺CD69⁺) from the total dermal DCs population (CD1⁺). f Frequency of NK cells (CD56) reported as percentage of total CD45⁺ cells. g Frequency of cytotoxic NK cells (CD56⁺CD16⁺) as percentage of total NK cells. h–j Frequency of M2 macrophages and MFI of their expressed activation markers CD16 and CD69. Statistical analyses were performed with Chi-square test, two tailed (a) and Friedman + Dunn’s multiple comparisons test, two tailed with adjusted p-values reported (b–j). Bars indicate median and interquartile range. N = 10 individuals. Only p-values < 0.05 are reported. Source data are provided as a Source Data file.

Fig. 6. Changes in the T cell compartment are observable at 48 h after mosquito exposure. CD8 T cells exhibit a high activation phenotype, while CD4 T effector memory cells shift from a T_h1/T_h17 phenotype to a T_h2/T_h17 phenotype.

a Increased frequency of CD25⁺ and PD1⁺ CD8 T cells at 48 h post mosquito exposure. b Frequencies of CD8 T cells naïve (CCR7⁺CD45RA⁺), central memory (CCR7⁺CD45RA^-), effector memory (CCR7^-CD45RA^-), and TEMRA (CCR7^-CD54RA⁺). c Decreased T_h1/T_h17 and increased T_h2/T_h17 effector memory CD4 T cell compartment at 48 h. d CD4 T cells naïve (CCR7⁺CD45RA⁺), central memory (CCR7⁺CD45RA^-), effector memory (CCR7^-CD45RA^-) and TEMRA (CCR7^-CD54RA⁺) frequencies. Statistical analyses were performed with Wilcoxon signed-rank test two tailed (a–d). Bars indicate median and interquartile range. N = 8 individuals. Only p-values < 0.05 are reported. Source data are provided as a Source Data file.

Fig. 7. Human skin cells obtained from Ae. aegypti bitten skin stimulated by Ae. aegypti salivary gland extract (SGE) produce significantly less pro-inflammatory cytokines (IL-2 and IFN-γ).

Cytokine production measured in skin cell culture supernatant after treatment with PMA/Ionomycin in the presence or absence of SGE. Skin dissociated cells were seeded on round-bottom 96-well plates (50,000 cells/well) and treated with SGE (10 µg/mL) or PBS for 24 h. Cells were stimulated with PMA (0.1 µg/mL) and Ionomycin (1 µg/mL) or left unstimulated for the last 6 h of the culture. Statistical analysis was performed with Wilcoxon signed-rank test, two tailed. N = 9 individuals. Only p-values < 0.05 are reported. Source data are provided as a Source Data file.

Fig. 8. The cutaneous immune response to Aedes aegypti mosquito ‘bites’ evolves over time.

Little activity in gene regulation and cell populations were noted at 30 min. At 4 h post-bite, a clear innate immune signature emerged followed by a strong adaptive response at 48 h.