Ixodes scapularis saliva mitigates inflammatory cytokine secretion during Anaplasma phagocytophilum stimulation of immune cells

Background Ixodes scapularis saliva enables the transmission of infectious agents to the mammalian host due to its immunomodulatory, anesthetic and anti-coagulant properties. However, how I. scapularis saliva influences host cytokine secretion in the presence of the obligate intracellular rickettsial pathogen Anaplasma phagocytophilum remains elusive. Methods Bone marrow derived macrophages (BMDMs) were stimulated with pathogen associated molecular patterns (PAMPs) and A. phagocytophilum. Cytokine secretion was measured in the presence and absence of I. scapularis saliva. Human peripheral blood mononuclear cells (PBMCs) were also stimulated with Tumor Necrosis Factor (TNF)-α in the presence and absence of I. scapularis saliva and interleukin (IL)-8 was measured. Results I. scapularis saliva inhibits inflammatory cytokine secretion by macrophages during stimulation of Toll-like (TLR) and Nod-like receptor (NLR) signaling pathways. The effect of I. scapularis saliva on immune cells is not restricted to murine macrophages because decreasing levels of interleukin (IL)-8 were observed after TNF-α stimulation of human peripheral blood mononuclear cells. I. scapularis saliva also mitigates pro-inflammatory cytokine response by murine macrophages during challenge with A. phagocytophilum. Conclusions These findings suggest that I. scapularis may inhibit inflammatory cytokine secretion during rickettsial transmission at the vector-host interface.


Background
Hematophagy occurred independently in more than 14,000 arthropod species and this adaptation required physiological, morphological, behavioral and biochemical changes [1,2]. Salivary gland secretion is among the most common physiological and biochemical adaptation in hematophagous arthropods and salivary proteins from ticks, mosquitoes, biting flies, fleas and other bloodfeeding arthropods ensue defense against host homeostasis and inflammation [1][2][3][4][5][6][7][8]. Combating inflammation is particularly problematic for ixodid ticks because these arthropods have to feed for a prolonged period of time and are exposed to a wide range of immune cells [9,10]. Pioneering studies have characterized the physiology of tick salivary glands [11] and raised the importance of saliva as an instrumental force for immune evasion [12][13][14][15]. Several groups have demonstrated that both proteinaceous and non-proteinaceous components of tick saliva impair the complement system and the function of macrophages, dendritic, T and B cells [8,[16][17][18][19][20][21][22][23][24][25]. From these studies it also became apparent that pathogens take advantage of the immunomodulatory properties of vector saliva to colonize the host.
Titus and Ribeiro were the first researchers to describe the role of arthropod salivary glands facilitating parasite transmission. In this seminal work, the authors showed that sandfly salivary glands facilitate Leishmania spp. transmission to the mammalian host [26]. Subsequent studies by many other groups demonstrated that the saliva of arthropod vectors potentiates the transmission of a wide-range of arthropod-borne pathogens, including bacteria, viruses and protozoal organisms. For example, in Ixodes spp., Borrelia burgdorferi outer surface protein C (OspC) binds to salivary protein 15 (Salp15), which then acts as a barrier to protect this spirochete against the host immune response [27]. Sialostatin L2, an I. scapularis cystatin protein also facilitates the growth of the Lyme disease agent B. burgdorferi [28]. In addition, I. ricinus saliva inhibits interferon and Toll-like receptor (TLR) signaling during cell stimulation with B. afzelli [29,30]. Finally, infection by the tick-borne encephalitis virus can be prevented by immunizing animals against a truncated recombinant form of a tick salivary protein named 64P [31].
Anaplasma phagocytophilum is a pathogen transmitted by ixodid ticks and causes human granulocytic anaplasmosis, an emerging infectious disease in the United States, Europe and Asia [32]. However, it remains mostly unknown whether tick saliva inhibits cytokine secretion by macrophages during stimulation with A. phagocytophilum. Macrophages were recently shown to be important for defense against A. phagocytophilum colonization [33]. In this study, we show that I. scapularis saliva inhibits extracellular and intracellular receptor signaling in both murine and human immune cells. We also demonstrate that tick saliva mitigates A. phagocytophiluminduced cytokine secretion by murine macrophages. This report expands previous scientific knowledge on the immunomodulatory properties of tick saliva and suggests that rickettsial agents may use I. scapularis saliva to inhibit inflammation at the vector-host interface.

Ethical statements
Blood samples were obtained from healthy, non-pregnant adults. This procedure was approved by the Human Research Review Board (HRRB number: HS-08-135) at the University of California-Riverside. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC number: A-20110030BE) at the University of California-Riverside. We used C57BL/6 mice at 6-10 weeks of age purchased from Jackson Laboratories. Experimentation with A. phagocytophilum (HZ strain) was approved by the Biological Use Authorization Committee (BUA number: 20120020) at the University of California-Riverside. A. phagocytophilum was grown in HL-60 cells (ATCC CCL-240). HL-60 cells were maintained in Iscove's Modified Dulbecco's Media (IMDM) with L-glutamine and hydroxyethyl piperazineethanesulfonic acid (HEPES) (Thermo Scientific), 20% heat-inactivated fetal bovine serum (FBS) (Sigma) in 5% CO 2 and humidified air at 37°C, as previously described [33].
We collected I. scapularis saliva 4-5 days after feeding because studies suggest that transmission of A. phagocytophilum initiates slowly between 24 and 48 hours and is enhanced during rapid feeding to repletion around 72 h-96 h post tick attachment [36][37][38]. Therefore, saliva from I. scapularis would reflect actual conditions during A. phagocytophilum transmission at the vector-host interface. In addition, I. scapularis saliva collection at 24-48 hours is technically very challenging. The alternative would be using salivary glands. However, salivary glands bring a technical artifact to the system because this organ in ticks is rich in intracellular proteins and other immune effectors such as nucleotides, which may skew cytokine response in immune cells. To isolate vector saliva, I. scapularis ticks were allowed to feed on New Zealand white rabbits. A restraining collar was placed around the neck of each rabbit, and their ears were covered prior to tick exposure. Ticks were permitted to engorge for 4-5 days on the ear of a rabbit. Upon harvesting, ticks were rinsed in distilled water and were immediately fixed to glass slides with double-sided tape. A sterile glass micropipette was placed around the hypostome to collect saliva. Salivation was induced by the application of pilocarpine to the scutum of the tick. Saliva was pooled and stored at −80°C for use.

ELISA
Mouse TNF-α, IL-1β, IL-6 and human IL-8 were measured with the BD OptEIA Set from BD Biosciences. Mouse IL-12p40 was measured with capture and detection antibodies from eBiosciences. For the ELISA assays, wells were coated with recommended capture antibody dilutions in freshly prepared coating buffer (0.1 M sodium carbonate, pH 9.5). Plates were sealed and incubated overnight at 4°C followed by aspiration. Wells were then washed 3 times with ≥ 300 μL/well of freshly prepared wash buffer (phosphate buffered saline (PBS) with 0.05% Tween-20). After the last wash, plates were inverted and blotted on absorbent paper to remove any residual buffer. Wells were then blocked with ≥ 200 μL/well of assay diluent (PBS with 10% FBS, pH 7.0) and incubated at room temperature for 1 hour. Plates were washed 3 times with wash buffer. Then, standards and sample dilutions were prepared in assay diluent, as recommended (BD Biosciences). 100 μL of samples and standards were pipetted into the wells, incubated for 2 hours at room temperature followed by 5 washes. 100 μL of detection antibodies were diluted in assay diluent and added to each well. Plates were sealed and incubated for 1 hour at room temperature. Plates were washed 5 times. Then, 100 μL of enzyme reagent (BD OptEIA Set from BD Biosciences) were diluted in assay diluent, pipetted into each well and incubated for 30 minutes at room temperature. Wells were aspirated and washed 7 times with wash buffer. 100 μL of substrate solution (BD OptEIA Set from BD Biosciences) were added to each well and incubated for 30 minutes (without plate sealer) at room temperature in the dark. 50 μL of 2 N H 2 SO 4 was added to each well. Absorbance was read in the ELISA plate reader (Bio-Rad) at 450 nm within 30 minutes. Background was corrected by reading the subtract absorbance at 570 nm.

Cell death assay
Cell death was assayed by measuring lactate dehydrogenase (LDH), as recommended by the manufacturer (Takara). Briefly, 100 μL of each sample was placed into a well. Then, a catalyst solution (Takara) was added to the samples and controls. Samples were incubated for 10-30 minutes, at room temperature, protected from light. Reactions were stopped at the end of the incubation period by adding 50 μL of 1 N HCl. Absorbance was measured at 490 nm. The iMark microplate absorbance reader (Bio-Rad) was used according to the manufacturer's instruction.

Statistical analysis
Data were expressed as means ± standard errors of the means (SEM). The following parametric analyses were used: unpaired Student's t test (two-group comparisons); one-way analysis of variance (ANOVA) (comparisons of three or more groups); Bonferroni post hoc multiplecomparison tests were used following ANOVA. All statistical calculations were performed using GraphPad Prism version 5.04. Graphs were made using GraphPad Prism version 5.04. P < 0.05 was considered statistically significant.

I. scapularis saliva diminishes inflammatory cytokine secretion by murine macrophages
Tick saliva has immunomodulatory properties [1,2]. To determine whether tick saliva inhibits macrophage function, we first stimulated mouse BMDMs with LPS in the presence or absence of different dilutions of I. scapularis saliva. Cytokine levels were not altered during BMDM stimulation with tick saliva alone, suggesting that the saliva does not carry any contaminants, pathogenassociated molecular (PAMPs) or danger-associated molecular (DAMPs) patterns. As expected, LPS induced high levels of cytokine secretion in murine BMDMs (Figures 1 and 2). However, I. scapularis saliva inhibited secretion of both TNF-α and IL-12p40 by BMDMs after stimulation with LPS ( Figure 1). This effect was more pronounced for TNF-α, as a tick saliva dilution of 1:10000 (v/v) still affected cytokine secretion by BMDMs during LPS stimulation ( Figure 1A). A reduction in IL-12p40 secretion by macrophages after LPS stimulation was only observed for a tick saliva dilution of 1:1000 (v/v) and below ( Figure 1B).
We then stimulated mouse BMDMs with a wide range of TLR ligands and measured IL-6 and IL-12p40 secretion in the presence of I. scapularis saliva. Tick saliva 1:500 (v/v) inhibited IL-6 and IL-12p40 secretion by BMDMs when stimulated with TLR agonists, such as LPS, Pam3CSK4, Zymosan and PG-LPS ( Figure 2). Next, we stimulated BMDMs with Nod1 and Nod2 agonists to determine whether the effect of I. scapularis saliva on cytokine secretion was restricted to TLRs. Nod1 and Nod2 are considered cytosolic receptors and are part of the Nod-like receptor (NLR) protein family [39]. The Nod1 and Nod2 agonists iE-DAP and MDP did not induce cytokine production in BMDMs during extracellular stimulation (Figure 3; data not shown). However, transfection of MDP to the cytosol using the cationic lipid DOTAP led to secretion of both IL-6 and IL-12p40 by BMDMs (Figure 3). Further, I. scapularis saliva inhibited IL-6 and IL-12p40 secretion mediated by the transfected Nod2 agonist MDP at the saliva dilution of 1:500 (v/v). Taken together, we report that I. scapularis saliva mitigates cytokine secretion by BMDMs during extracellular and cytosolic stimulation of TLR and NLR pathways.

I. scapularis saliva inhibits chemokine secretion by human peripheral blood mononuclear cells
We then tested whether the effect observed for I. scapularis during cytokine secretion was restricted to murine BMDMs. We isolated human PBMCs from the blood of healthy and immunocompetent adults and stimulated these immune cells with the pro-inflammatory cytokine TNF-α. TNF-α has been extensively used to stimulate cytokine secretion in immune cells [40]. We included human PBMCs in our studies for two reasons. First, we would like to show that the inhibitory aspect of tick saliva is not only restricted to murine immune cells. To the contrary, it can also occur during stimulation of human immune cells. Second, PBMCs are a heterogenous group of immune cells. Therefore, we reasoned that PBMCs would be ideal to demonstrate the inhibitory cytokine capacity of tick saliva in a mixed group of immune cells. The chemokine IL-8 (CXCL-8) was used as a read-out in our analysis. PBMCs produced large amounts of IL-8 when stimulated with TNF-α ( Figure 4). Similar to our findings observed in murine BMDMs, IL-8 (CXCL8) secretion by PBMCs stimulated with recombinant human TNF-α was diminished in the presence of tick saliva ( Figure 4). Overall, our results show that I. scapularis saliva is able to mitigate cytokine and chemokine secretion by both murine and human immune cells.

I. scapularis saliva lessens inflammatory cytokine secretion by murine macrophages during A. phagocytophilum stimulation
To determine whether the effect of tick saliva was restricted to TLR and NLR ligands, we stimulated BMDMs with the I. scapularis rickettsial pathogen A. phagocytophilum. A. phagocytophilum induced the secretion of large amounts of cytokines by murine macrophages ( Figures 5 and 6A-B). Nonetheless, tick saliva was also efficient in reducing cytokines by BMDMs, such as IL-6, IL-12p40 and TNF-α during A. phagocytophilum stimulation ( Figure 5). Similar to BMDM stimulation with LPS, the effect of tick saliva was more pronounced on TNF-α secretion by BMDMs during pathogen stimulation. TNF-α secretion was completely abolished when BMDMs were stimulated with A. phagocytophilum in the presence of tick saliva ( Figure 5C).
Overall, the inhibitory effect of tick saliva on cytokine secretion by BMDMs during A. phagocytophilum stimulation was best observed at a 1:150 dilution (v/v). Experiments performed with tick saliva at a 1:300 dilution (v/v) also showed reduction of cytokine secretion by BMDMs during A. phagocytophilum stimulation ( Figure 5). However, the effect was milder (although statistically significant) for IL-6 and IL-12p40 ( Figure 5A and B). We did not detect a dilution effect on TNF-α secretion by BMDMs at a 1:300 dilution (v/v) ( Figure 5C). These results suggest that the inhibition threshold for tick saliva on TNF-α secretion by BMDMs during A. phagocytophilum stimulation is greater than 1:300 (v/v) compared to 1:150 dilution (v/v). It is unclear why more concentrated tick saliva is required to inhibit cytokine secretion during A. phagocytophilum stimulation when compared to individual TLR or NLR agonists. We reasoned that the presence of multiple PAMPs in a pathogen, such as A. phagocytophilum, may require a stronger dose of tick saliva to mitigate cytokine secretion by BMDMs.

I. scapularis saliva inhibits IL-1β secretion by murine macrophages during A. phagocytophilum stimulation
A dogma has emerged in the last few years in which the production and release of IL-1β are the result of a twotier system: one signal is induced by pattern recognition receptors or pro-inflammatory cytokines, which activates the transcription and translation of IL-1β via the transcription factor nuclear factor (NF)-κB. This signal is also referred to as priming and it is typically done by LPS stimulation of immune cells. The second signal is mediated by the inflammasome, a protein scaffold that cleaves IL-1β into its mature form through caspase-1 activation [41][42][43]. Our studies show that IL-1β secretion triggered by A. phagocytophilum during BMDM  stimulation does not require LPS priming. To the contrary, LPS priming in BMDMs before A. phagocytophilum stimulation inhibited IL-1β secretion in a dosedependent manner ( Figure 6A). These results are consistent with the lack of genes for LPS and peptidoglycan synthesis in the A. phagocytophilum genome [44] and suggest that molecules other than LPS may prime BMDMs for the production of pro-IL-1β. Further, tick saliva inhibited A. phagocytophilum-induced IL-1β secretion by BMDMs ( Figure 6B). IL-1β secretion may be coupled to a cell death phenomenon named pyroptosis [45]. We did not detect high levels of cell death during A. phagocytophilum stimulation of BMDMs in the absence of tick saliva (~10-15%) ( Figure 6C). However, this is not entirely surprising, as A. phagocytophilum does not cause much cell death during infection [32,46]. Based on these findings, we conclude that in the A. phagocytophilum model, cytokine secretion is a better indicator of the anti-inflammatory features of tick saliva when compared to cell death.

Discussion
Inflammation is characterized by complex interactions between innate and adaptive immunity [47]. Pro-inflammatory cytokines and chemokines recruit immune cells to the site of tick feeding. Tick salivary proteins then mitigate the secretion of cytokines by immune cells, thereby, diminishing inflammation [1,2,4]. Despite significant progress in the past decades, how ectoparasites, such as ticks, regulate host innate immune signaling during transmission of the rickettsial agent A. phagocytophilum to the mammalian host remains mostly elusive. In this study, we demonstrate that I. scapularis saliva has the ability to inhibit cytokine secretion by murine and human immune cells. These findings are supported by our results showing that extracellular and cytosolic  stimulation of macrophages with PAMPs can be inhibited by I. scapularis saliva. We also performed experiments with A. phagocytophilum and demonstrated that similar mitigation effects occur in macrophages. The implications for these findings are wide in scope as ticks, mosquitoes, biting flies, fleas and blood-feeding bugs have also evolved similar strategies for modulating host defenses [2].
We reported an effect of tick saliva on TLR and NLR signaling in macrophages. Although the effect of tick saliva was previously demonstrated during TLR stimulation of dendritic cells [29,30,[48][49][50], whether I. scapularis saliva affects the response to stimulation of murine macrophages and human peripheral blood mononuclear cells had not been determined. Macrophages and peripheral blood mononuclear cells are important because these immune cells respond to A. phagocytophilum infection [33,51]. To our knowledge, we describe for the first time that secretion of IL-6 and IL-12p40 after stimulation with the Nod2 ligand MDP was diminished in macrophages during treatment with tick saliva. Nod2 has emerged as a critical regulator for immunity and inflammation since it activates canonical and non-canonical NF-κB signaling, mitogenactivated protein kinases, cytokines, chemokines and antimicrobial reactive oxygen species [39].
Previously, we showed that A. phagocytophilum is partially recognized by the NLRC4 inflammasome [52], a protein scaffold that regulates the secretion of IL-1β and IL-18 [45]. We also demonstrated that mice deficient in caspase-1 and asc, essential components of the inflammasome, were more susceptible than wild-type animals to A. phagocytophilum infection. These findings were due to the absence of IL-18 secretion and reduced interferon (IFN)-γ levels in the peripheral blood. It is unclear how I. scapularis saliva regulates IL-1β secretion by macrophages during A. phagocytophilum stimulation. However, it is possible that multiple salivary proteins regulate IL-1β secretion during hematophagy. Ticks have large genomes and carry many gene paralogs [53]. These gene paralogs may act redundantly to provide inhibition of immune protein scaffolds in the mammalian host. Two earlier articles provided experimental support for this hypothesis. Ramachandra and Wikel showed that salivary gland extracts from the tick Dermacentor andersoni reduced IL-1 levels during the early phases of tick feeding [54], whereas Fuchsberger et al., 1995 determined that human IL-1β secretion was mitigated when treated with LPS and salivary gland extracts from partially fed adult female Rhipicephalus appendiculatus [55].
In addition, A. phagocytophilum may need redundant mechanisms of innate immune recognition to trigger IL-1β secretion. Secretion of IL-1β requires NF-κB activation to generate pro-IL-1β [56,57]. Dumler and colleagues demonstrated that A. phagocytophilum triggers TLR2 activation during immune cell stimulation [58]. TLR activation is known to initiate NF-κB signaling in immune cells [59]. More recently, our group participated in a study showing that receptor interacting protein-2 (RIP2) affects A. phagocytophilum infection in mice [60]. RIP2 is an adaptor molecule for the innate immune receptors Nod1/2, which also regulates NF-κB signaling [61]. Finally, assembly of a multi-protein complex coined "inflammasome" is critical for IL-1β secretion [56,57]. We previously demonstrated that the inflammasome is critical for immunity against A. phagocytophilum infection [52]. Taken together, our findings reinforce the notion that A. phagocytophilum immunity is multi-factorial, and suggests a holistic inhibitory effect of tick saliva on innate immunity. This is important because a pathogen such as A. phagocytophilum may need multiple layers of immune evasion during transmission. Therefore, the holistic properties of tick saliva may be a major strategy of host immune evasion during pathogen transmission.
The above hypothesis is supported by several lines of evidence. Post-genomic approaches show that A. phagocytophilum actively modulates gene expression in ticks [62,63]. Recently, the P11 salivary protein was shown to be required for A. phagocytophilum migration from hemocytes to the salivary glands in ticks [64]. Another salivary gland protein named SALP16 was deemed important for A. phagocytophilum survival within the tick vector [65]. A. phagocytophilum alters the monomeric/ filamentous (G/F) actin ratio leading to the translocation of phosphorylated/G-actin to the nucleus [66]. This event affects salp16 gene transcription in association with the RNA polymerase II (RNAPII) and the TATAbinding protein. Fikrig and colleagues have also demonstrated that A. phagocytophilum appears to increase the ability of I. scapularis to survive in cold temperatures by up-regulating an antifreeze glycoprotein [67] and α1, 3fucosyltransferases, which are important for pathogen colonization [68]. On the other hand, the Janus kinase (JAK)-signaling transducer activator of transcription (STAT) pathway seems to be important for the restriction of A. phagocytophilum infection in ticks. Clearly, further studies are necessary to determine the contribution of salivary proteins to A. phagocytophilum pathogenesis and immunity.
Elucidating the underlying effect of tick saliva in vivo should be considered a high-priority in tick research. Several experiments show that findings obtained in vitro may sometimes differ from those occurring in vivo, and mRNA and protein levels do not necessarily correlate well [1]. In vivo characterization of tick salivary proteins, however, is not a trivial task. The presence of multiple paralogues in the I. scapularis genome [53] make the use of siRNA or dsRNA technology challenging because of known off-target effects [69]. Another technical limitation is the lack of strategies to introduce and/or delete genes in ticks. Thereby, we were unable to generate knock-out, knock-in or transgenic I. scapularis. The development of this technology would enable researchers to characterize tick salivary proteins and clarify underlying events at the vector-pathogen-host interface. Despite all these technical restrictions, we posit that I. scapularis saliva may inhibit cytokine secretion during rickettsial transmission. We look forward with reasonable confidence that our findings may be used as a prelude for future in-vivo experimentation.

Conclusions
To facilitate blood-feeding habits and counteract homeostasis and inflammation in the host, arthropod vectors have evolved a sophisticated pharmacological portfolio. Our studies add to a continuum of research over the past 20-30 years and show that I. scapularis saliva diminishes cytokine secretion by mouse and human immune cells during TLR and NLR stimulation. We also show that cytokine response by macrophages is diminished in the presence of I. scapularis saliva during A. phagocytophilum stimulation. Taken together, our findings should pave the ground for new research directions in the field of A. phagocytophilum pathogenesis and immunity and may have direct implications for understanding how ticks circumvent defenses promoted by the mammalian immune system.

Competing interests
The authors declare that they have no competing interests.
Authors' contributions GC, MSS, MS and OSS carried out the experimental work. SKW and MK analyzed the data and provided intellectual support. JHFP directed the project and wrote the manuscript. All authors read and approved the final manuscript.