Tick extracellular vesicles enable arthropod feeding and promote distinct outcomes of bacterial infection

Abstract

Extracellular vesicles are thought to facilitate pathogen transmission from arthropods to humans and other animals. Here, we reveal that pathogen spreading from arthropods to the mammalian host is multifaceted. Extracellular vesicles from Ixodes scapularis enable tick feeding and promote infection of the mildly virulent rickettsial agent Anaplasma phagocytophilum through the SNARE proteins Vamp33 and Synaptobrevin 2 and dendritic epidermal T cells. However, extracellular vesicles from the tick Dermacentor andersoni mitigate microbial spreading caused by the lethal pathogen Francisella tularensis. Collectively, we establish that tick extracellular vesicles foster distinct outcomes of bacterial infection and assist in vector feeding by acting on skin immunity. Thus, the biology of arthropods should be taken into consideration when developing strategies to control vector-borne diseases.

Fig. 1. Ticks secrete EVs with a distinct cargo profile.

a Transmission electron microscopy of EVs released by I. scapularis (ISE6) cells. Scale bar: 100 nm, large panel = 11,000x; small panel = 30,000x. The image is representative of two independent experiments. b SDS-PAGE immunoblot showing CD63⁺ EVs purified from AAE2, ISE6, and DAE100 tick cells (70 µg of protein). Images are a representation of three independent experiments. c I. scapularis ISE6 cells transfected with synaptobrevin 2 small interfering (Synaptobrevin 2 si) (green) or scrambled control RNA (Synaptobrevin 2 sc) (gray) to reduce the EV release. Concentration and size distribution of EVs were measured by nanoparticle tracking analysis (NTA). Mean ± standard error of the mean (SEM) of technical replicates plotted from one of two independent experiments are shown. Similar results were obtained in both independent studies. Two-way ANOVA using size and treatment as variables followed by Sidak’s multiple comparison statistical test of the same size vesicles between treatments. *p < 0.05; **p < 0.005; ***; p < 0.001;  NS not significant. d EVs originated from I. scapularis salivary gland (SG) cultures. Mean ± SEM are plotted. Data is representative of three independent experiments. e–g SDS-PAGE immunoblots showing CD63⁺, ALIX⁺, and TSG101⁺ EVs purified from I. scapularis SGs (24 µg of protein). h CCT7, CTNNB1, DHX16, PLS3, and GFPT1 expression in EVs purified from partially engorged I. scapularis SG and uninfected tick cells. i Glycosylation of proteins in EVs purified from partially engorged I. scapularis SG and tick cells. HRP (+) represents the horseradish peroxidase positive control and STI (−) indicates the soybean trypsin inhibitor negative control for the assay. j, k Phosphorylation and carbonylation of proteins in EVs purified from partially engorged I. scapularis SG and tick cells. l EVs purified from I. ricinus saliva and immunoblotted against CD63, Sialostatin L2 (SL2), and calnexin (15 µg of protein). EVs from HL^-60 were used as control cells, whereas calnexin was used as a technical negative control. Blots are representative of two independent experiments. Source data are provided as a Source Data file.

Fig. 2. Tick EVs bind to mammalian immune cells.

a Overrepresentation of proteins involved in cell adhesion signaling networks in I. scapularis salivary gland EVs. The biological relationship between proteins was determined using the right-tailed Fisher’s exact test with Benjamini–Hochberg multiple-testing correction. p < 0.05; −log (p value) >1.5. b Graphic representation of focal adhesion and integrin signaling proteins (purple) found in I. scapularis salivary gland EVs. Confocal images of c murine bone marrow-derived macrophages (BMDMs) (bar = 20 µm) and g macrophages derived from human peripheral blood mononuclear cells (green) bound to I. scapularis EVs labeled with PKH26 (orange) (bar = 20 µm), respectively. Cells were pre-treated with cytochalasin D (5 µM). (−) No EVs; (+) EVs. Images are representative of two independent experiments. d Arbitrary fluorescent units (AFU) of murine BMDMs bound to PKH26-labeled tick EVs (red). BMDMs incubated in the absence of EVs (black). Each time point represents the mean fluorescence from seven different cells minus background and normalized to time 0. Graphs show one of two independent experiments. Flow cytometry with e murine BMDMs (F4/80⁺-APC) and h human macrophages (CD11b⁺-APC/Cy7) bound to DiO-labeled tick EVs (40 µg), respectively. No EVs (−); EVs (+). Figures are representative of three biological replicates. f Flow cytometry analysis of unlabeled BMDMs (black), F4/80-APC-labeled BMDMs (blue), F4/80-APC-labeled BMDMs bound to 40 µg DiO-labeled tick EVs (red), and F4/80-APC-labeled BMDMs bound to 20 µg of DiO-labeled and 20 µg unlabeled tick EVs (gray). The histogram shows one of three biological replicates. PKH26 is a red fluorescent dye with long aliphatic tails. In c, g, PKH26 was artificially transformed to an orange color to be visualized by color-blind readers. Source data are provided as a Source Data file.

Fig. 3. Tick EVs affect the skin resident immune environment.

a Attachment and b weight of I. scapularis ticks microinjected with vamp33 small interfering (Vamp33 si) (green) (n = 26) or scrambled (Vamp33 sc) (gray) (n = 20) RNA on C57BL/6 mice (n = 2 each treatment) after 3 days of feeding. Mean ± standard error of the mean (SEM) are plotted. Data represent two independent experiments. Statistical differences were evaluated by two-tailed t test. *p < 0.05. Effect of vamp33 si I. scapularis ticks on c interferon (IFN)-γ, d interleukin (IL)-5, and e IL-13 in the murine skin after 3 days of feeding. Increased f γδ T cells, g IL-2 and h–j chemokine release at the bite site after Vamp33 si I. scapularis fed for 3 days on C57BL/6 mice. Mean ± SEM are plotted. Cytokine and chemokine data represent a single experiment with multiple biological replicates, whereas data with skin immune cells represent three independent experiments combined. Statistical differences in d–j were evaluated by two-tailed t test. *p < 0.05. NS not significant. n number of samples. Source data are provided as a Source Data file.

Fig. 4. Tick EVs affect DETCs in the murine skin.

a Mice injected with anti-γδ antibodies (500 µg) display decreased γδ T cells numbers at the skin site (right). The graph is representative of two independent experiments. b Vamp33 silencing (squares) resulted in similar tick bloodmeal intake when compared to the control treatment (circles) on γδ T cell-depleted mice (green). The graph represents two independent experiments combined. c Weight of Vamp33 si or Vamp33 sc ticks placed on wild type (WT; gray), tcrβ^−/− (green), and tcrδ^−/− (red) mice for 3 days. The graph represents four independent experiments combined. d Ticks were placed on FTY720-(squares) (1 mg/kg) or PBS-(circles) (−) treated C57BL/6 mice and allowed to feed for 3 days. Skin biopsies at the tick bite site and the draining lymph nodes were taken and immune cell populations were assessed by flow cytometry. The graph represents one of two independent experiments. e Vamp33 sc (green) or Vamp33 si (red) ticks were placed on PBS- (circles) (−) or FTY720- (squares)-treated (1 mg/kg) C57BL/6 mice and their weight was measured three days post-infestation (two independent experiments combined). γδ T cell graph (left) represents one of two independent experiments. f Ticks were placed on C57BL/6 mice and allowed to feed for 3 days and their weight was measured (two independent experiments combined). Biopsies were taken from the tick bite site and the DETC (Vγ5⁺Vδ1⁺) population was assessed by flow cytometry (one of two independent experiments). g Flow cytometry measurement of TCR γδ^hi immune population in a murine skin sample when compared to a TCR γδ^lo immune population in the draining lymph node. h Skin biopsies were obtained from FVB-Jax (gray) and FVB-Tac (green) mice. Vγ5⁺ T cell population was assessed by flow cytometry (two independent experiments). Vamp33 sc (circles) or Vamp33 si (squares) ticks were placed on FVB-Jax (gray) or FVB-Tac (green) mice. Tick weight measurement was obtained 3 days post-infestation (two independent experiments combined). Differences between treatments were evaluated with two-tailed Student’s t test for all data sets, except one-tailed in d. Statistical significance for all experiments are shown *p < 0.05; NS not significant, n number of samples analyzed. Ab antibody, AM morning, PM afternoon. Source data are provided as a Source Data file. Mean ± standard error of the mean (SEM) are plotted for all experiments except in g.

Fig. 5. Tick EVs are modified by the intracellular bacterial lifestyle.

a Transmission electron microscopy of EVs purified from I. scapularis ISE6 cells infected with A. phagocytophilum or B. burgdorferi. Scale bars: 100 nm, large panel = 11,000x; small panel = 30,000x. The image is representative of two independent experiments. b I. scapularis EVs released from A. phagocytophilum or B. burgdorferi-stimulated ISE6 cells. Means ± standard error of the mean (SEM) are plotted. (−) unstimulated cells. (*p < 0.05). Data is representative of three technical and two biological replicates. SDS-PAGE immunoblots showing increased release of c 6-phosphogluconate dehydrogenase (PGD) (50 µg for lysates and 40 µl for EV immunoblots, respectively); NRF2 nuclear factor erythroid 2-related factor 2. d 15 µg of EVs derived from tick ISE6 cells infected with A. phagocytophilum (MOI 50) showing carbonylation. e Carbonylation in EVs derived from A. phagocytophilum-infected IDE12 cells (MOI 50; 15 µg of EVs for ELISA). DNPH = 2,4-dinitrophenylhydrazine. (−) represents uninfected ISE6 cells lysate. f, g glycosylation and phosphorylation of proteins in uninfected and A. phagocytophilum-infected (MOI 10 and 50) EVs purified from tick ISE6 cells at indicated amounts. HRP (+) represents the horseradish peroxidase positive control while STI (−) indicates the soybean trypsin inhibitor negative control. MOI multiplicity of infection. f, g blots and gels were repeated at least twice obtaining the same results. Statistics in b was done by using Two-way ANOVA using size and treatment as variables followed by the post-hoc Sidak test for comparisons. (*p < 0.05). Source data are provided as a Source Data file.

Fig. 6. D. andersoni EVs mitigate F. tularensis infection in the mammalian host.

In a and b C57BL/6 mice were intradermally inoculated with three doses of F. tularensis live vaccine strain: 7.5 × 10⁶ (black), 7.5 × 10⁷ (blue), and 7.5 × 10⁸ (red). Survival and loss of weight were monitored daily for 10 days. Data represents the mean ± SEM; n number of mice. c Quantity and size distribution of EVs purified from D. andersoni salivary glands, as judged by nanoparticle tracking analysis. Bars represent the standard error from three technical replicates. In d, e survival and weight loss in C57BL/6 mice inoculated with 8.4 × 10⁷ CFU of F. tularensis co-injected with 1 × 10⁸ D. andersoni EVs. EV treatment (blue); control (red). For e, data presents the mean ± SEM. In f–i F. tularensis (8 × 10⁷ CFU) was co-injected in the presence (green) or absence (gray) of D. andersoni salivary gland EVs (1 × 10⁸). f splenomegaly, g cytokine measurement made by a multiplex cytokine ELISA and h, i plaque assays at day 5 post-infection. Data is presented as a mean and standard error of the mean (±SEM). Statistical significance was determined using a two-tailed t test of Francisella and Francisella + EVs. f p = 0.03; g p = 0.0004; h p = 0.05; i p = 0.0017. j D. variabilis nymphs were placed onto naïve female C3H/HeN mice on day -5 and allowed to feed for three days. Mice were intravenously infected on day −2 with 1 × 10⁷ CFU of F. tularensis. Engorged (repleted) ticks and the mouse blood were collected 2 days later (day 0) and the CFU were obtained at indicated time points. The graph is representative of two independent experiments. k Infected adult D. variabilis ticks (week 14) were individually placed onto naïve mice to examine F. tularensis infection. Ticks completed their blood meal by day 8 and the mouse blood was harvested to quantitate bacterial numbers. Data is presented as mean. Five mice were monitored through day 18 with animals being euthanized when moribund. Skull and bones denote a single mouse death, whereas skull and bones with the pound sign indicates death of two mice. Survival was analyzed with the Kaplan–Meier curve. Statistical analysis was performed with the a, d Log-rank (Mantel-Cox) or f–i the two-tailed t test. In b, e statistical analysis of weight data was not done due to the differential animal mortality during experiment. EVs extracellular vesicles, CFU colony-forming units. *p ≤ 0.05. NS not significant. Source data are provided as a Source Data file.

Fig. 7. I. scapularis EVs enable A. phagocytophilum infection in the mammalian host.

A. phagocytophilum (1 × 10⁷) was intradermally injected in the presence or absence of I. scapularis salivary gland EVs (4 × 10⁷). The presence (red) or absence (green) of A. phagocytophilum in a skin or b spleen was measured by quantitative real-time PCR at day 3 post-infection. Cytokine measurements were done in the mouse c skin and d blood by a multiplex ELISA assay. Baseline treatment denotes skin and mouse blood without bacterial infection (gray bars). Data is presented as a mean and standard error of the mean (±SEM). (*p < 0.05; NS not significant). Graphs are representative of one in two independent experiments. Statistics in c were done by using one-way ANOVA, whereas d was done by two-tailed t-test between Anaplasma and Anaplasma + EVs. For c n = skin samples and d n = blood samples. Source data are provided as a Source Data file.