Function and evolution of the aquaporin IsAQP1 in the Lyme disease vector Ixodes scapularis

Abstract

Ticks are important vectors of pathogenic viruses, bacteria, and protozoans to humans, wildlife, and domestic animals. Due to their life cycles, ticks face significant challenges related to water homeostasis. When blood-feeding, they must excrete water and ions, but when off-host (for stretches lasting several months), they must conserve water to avoid desiccation. Aquaporins (AQPs), a family of membrane-bound water channels, are key players in osmoregulation in many animals but remain poorly characterized in ticks. Here, we bioinformatically identified AQP-like genes from the deer tick Ixodes scapularis and used phylogenetic approaches to map the evolution of the aquaporin gene family in arthropods. Most arachnid AQP-like sequences (including those of I. scapularis) formed a monophyletic group clustered within aquaglycerolporins (GLPs) from bacteria to vertebrates. This gene family is absent from insects, revealing divergent evolutionary paths for AQPs in different hematophagous arthropods. Next, we sequenced the full-length cDNA of I. scapularis aquaporin 1 (IsAQP1) and expressed it heterologously in Xenopus oocytes to functionally characterize its permeability to water and solutes. Additionally, we examined IsAQP1 expression across different life stages and adult female organs. We found IsAQP1 is an efficient water channel with high expression in salivary glands prior to feeding, suggesting it plays a role in osmoregulation before or during blood feeding. Its functional properties are unique: unlike most GLPs, IsAQP1 has low glycerol permeability, and unlike most AQPs, it is insensitive to mercury. Together, our results suggest IsAQP1 plays an important role in tick water balance physiology and that it may hold promise as a target of novel vector control efforts.

INTRODUCTION

Ticks are obligate blood-feeding arachnids that are not only pests of livestock and humans, but also vectors of numerous dangerous pathogens including viruses, bacteria, protozoans and nematodes. The deer tick, Ixodes scapularis, is the most important tick vector of pathogens in the eastern United States and a major vector of Borrelia burgdorferi, the Lyme disease agent. The increasing burden of tick-borne diseases—on people and animals alike—necessitates new methods for controlling this pernicious vector (Boulanger, Boyer, Talagrand-Reboul, & Hansmann, 2019; Dantas-Torres, Chomel, & Otranto, 2012). However, an obstacle to developing new tools is a lack of molecular detail in our current understanding of tick biology.

Ticks are long-lived arthropods (living up to several years) with a multistage life history that creates significant and opposing water balance challenges. On the one hand, while attached to hosts, ticks must excrete excess water and ions from the host blood they consume—a major physiological challenge because a bloodmeal may be 100 times their unfed body weight (Kaufman & Phillips, 1973). This excess water is removed via salivary secretion back into the host (Bowman & Sauer, 2004). On the other hand, ticks face desiccation risk during the extended periods they spend off-host (Benoit & Denlinger, 2010), during which water can only be obtained by absorbing it from the air into salivary tissues (Needham and Teele, 1991). “Three-host ticks” including I. scapularis require a host bloodmeal during each of three distinct life stages, with extended periods spent living off-host without feeding between these meals (Boulanger et al., 2019). The water homeostasis needs of I. scapularis, therefore, fluctuate dramatically across life stages, and these fluctuating needs may be reflected in stage- or tissue-specific expression of osmoregulatory molecules.

Aquaporins (AQPs) are a family of tetrameric transmembrane water channels found in the genomes of most organisms, from bacteria to plants and animals (Abascal, Irisarri, & Zardoya, 2014). They facilitate the movement of water across cells and membranes and have traditionally been classified as either water-specific (‘classical AQPs’), or as additionally permeable to small molecules such as glycerol or urea (the aquaglyceroporins, or GLPs; Abascal et al., 2014). Generally, but not always, these traditional functional categories comport with AQP gene trees—i.e., phylogenetic analysis results in distinct clades of AQPs and GLPs (e.g. Finn, Chauvigné, Stavang, Belles, & Cerda, 2015; Tsujimoto, Sakamoto, & Rasgon, 2017). The permeabilities and other properties of each channel are determined by a handful of residues within the channel pore, and changes to a few amino acids can alter channel permeability (Fu et al., 2000; Sui, Han, Lee, Walian, & Jap, 2001; Beitz et al., 2006; Campbell, Ball, Hoppler, & Bowman, 2008). Indeed, mismatches between measured channel function and traditional categories (i.e. classical AQP or GLP) have been described. For example, insect entomoglyceroporins (Eglps) have GLP-like permeability to solutes yet they evolved from classical water-specific AQPs (Finn et al., 2015), and Big brain (Bib) gene products appear to have lost permeability to water and other substrates but gained a new role in embryonic development (Doherty, Jan, & Jan, 1997; Rao, Jan, & Jan, 1990).

As water channels, AQPs mediate essential homeostatic processes including water balance (e.g. Liu, Tsujimoto, Cha, Agre, & Rasgon, 2011; Drake et al., 2010; Drake et al., 2015). Glycerol conductance (e.g. by GLPs) can also be important for survival (Philip, Kiss, & Lee Jr, 2011; Philip, Yi, Elnitsky, & Lee Jr, 2008), as some arthropods accumulate glycerol and other polyols to survive cold or desiccating conditions (e.g. Duman, 2001; Yoder et al., 2006). Knocking down AQPs can negatively impact hematophagous insects. In one case, reduction of an AQP and an Eglp decreased the ability of the bed bug, Cimex lectularius, to excrete waste following a bloodmeal (Tsujimoto et al., 2017). Other studies functionally verified a role for AQPs in modulating desiccation status in the mosquitoes Anopheles gambiae and Aedes aegypti (Drake et al., 2010; Liu et al., 2011), and in osmoregulation in the ticks Ixodes ricinus and Rhipicephalus microplus (Campbell, Burdin, Hoppler, & Bowman, 2010; Hussein et al., 2015). Outside of insects, less is known about arthropod AQPs (but see Stavang et al., 2015; Benoit et al., 2014, Campbell et al., 2008). Of the >700 hard tick and nearly 200 soft tick species (Boulanger et al., 2019), an aquaporin channel has thus far been functionally characterized in only one: Rhipicephalus sanguineus (Ball, Campbell, Jacob, Hoppler, & Bowman, 2009). We, therefore, studied putative AQP-like genes in I. scapularis, examined the pattern of AQP evolution in arthropods more broadly, and characterized both the functional properties and organ- and stage-specific expression patterns of the I. scapularis aquaporin IsAQP1.

RESULTS

Identification of I. scapularis AQPs and characterization of IsAQP1

We identified 18 putative aquaporin-like sequences for Ixodes scapularis including ISCW003957 (IsAQP1; Table S1). For in-depth functional analysis, we focused on one of the identified genes, Ixodes scapularis aquaporin-1 (IsAQP1), for which we acquired the full-length cDNA sequence using RACE. The IsAQP1 mRNA consists of 1481 nucleotides (excluding the poly-A tail) with a 885-nt coding sequence encoding a 294-aa polypeptide. When mapped to the genome, it spans 62,141 nucleotides on the Scaffold:IscaW1:DS692803 positive strand and comprises seven exons and six introns (Figure 1a). Our analysis adds two exons (one to the 5′ end and one to the 3′ end) to a previous genome annotation by VectorBase (geneset: IscaW1.5). The start and stop codons are located in these new exons, which is now reflected in the geneset IscaW1.6. The sequence information was submitted to GenBank with an accession number KT988052. The polypeptide contains two NPA motifs—a conserved feature of AQPs (Benga, 2012; Murata et al., 2000)—complete with RD residues after the second NPA motif uniquely conserved in aquaglyceroporins (GLPs) (Fu et al., 2000; Sui et al., 2001).

FIGURE 1

Gene structure of IsAQP1 and the evolution of AQP-like sequences in arthropods. (a) ISCW003957 (IsAQP1) gene structure in the I. scapularis genome. Top: the structure obtained by this study; bottom: the structure by the previous annotation (IscaW1.5). In both, the start (ATG) and stop (TGA) codons are indicated. (b) Phylogenetic tree of selected aquaporins (AQPs) and aquaglyceroporins (GLPs) from arthropods and other organisms. Bootstrap values (1000 replicates) are indicated at each node; unlabelled nodes have bootstrap values ≥85. Sequence IDs are as in Table S1. The scale bar represents amino acid substitutions.

Phylogenetic analysis of arthropod AQPs reveals distinct evolution in arachnids

We examined the relationship of the newly identified I. scapularis AQPs within the broader gene family using maximum likelihood phylogenetic analysis. We constructed a gene tree for AQP and AQP-like genes from diverse arthropods including hexapods, crustaceans and chelicerates that includes 12 orders and 48 species (Figure 1b). We additionally included 17 aquaporin sequences from outgroup taxa Escherichia coli, Milnesium tardigradum and Homo sapiens. The data set included 176 total AQP and AQP-like sequences. Most (68/94, 72%) arachnid AQPs (represented by Aranae, Ixodida, Mesostigmata, Scorpiones and Trombidiformes) formed a clade nested within the classical aquaglyceroporins (GLPs). This pattern was even more pronounced among tick sequences: only six of 57 tick AQPs (10.5%) did not cluster within the arachnid GLP clade, five of which appeared to be Bib genes. As expected, insect AQPs and insect-specific Eglps both clustered with classical water-specific AQPs (Figure 1b).

IsAQP1 is a functional water channel with limited glycerol permeability and no urea permeability

We expressed IsAQP1 heterologously in Xenopus laevis oocytes to assess its water and solute permeability and to compare it to channels characterized in our previous work. IsAQP1 exhibited significantly greater water permeability in comparison to vehicle (nuclease-free water) injected controls (Figure 2a, one-way ANOVA with post hoc Tukey test, p < 0.05). Indeed, IsAQP1 moved water more efficiently (1.47×) than the positive control classical AQP channel AgAQPB1 (Liu et al., 2011) from the malaria vector mosquito, An. gambiae (Figure 2a, one-way ANOVA with post hoc Tukey test, p < 0.05). However, unlike many characterized AQPs (e.g. Campbell et al., 2008; Hirano et al., 2010), it was not significantly inhibited by Hg²⁺ (Figure 3a, p > 0.05). IsAQP1 also had low glycerol permeability (0.28× that of the positive control channel (Tsujimoto et al., 2017); Figure 2b, one-way ANOVA with post hoc Tukey test, p < 0.05). It was not permeable to urea (Figure 2c; p > 0.05).

FIGURE 2

Water, glycerol and urea permeability of IsAQP1 (ISCW003957) alongside selected insect channels (positive controls) previously characterized in Liu et al. (2011) and Tsujimoto et al. (2017). (a) Water permeability of Xenopus oocytes expressing IsAQP1 in the presence or absence of conventional inhibitor Hg²⁺. (b) Glycerol permeability. (c) Urea permeability. Bars show means for each group while error bars depict ± standard deviation. H₂O: water-injected negative control; AgAQP1B: Anopheles gambiae AQP1-expressing oocyte. ClAQP1: Cimex lectularius AQP1 expressing oocyte. ClAQP1 + ClGlp1: Cimex lectularius AQP1 and Glp1 co-expressing oocyte. Letters indicate statistical grouping according to one-way ANOVA with post hoc Tukey tests.

FIGURE 3

Quantification of IsAQP1 mRNA expression by qRT-PCR. Organ-specific expression in unfed females (left) and engorged females (right). SG: salivary glands, MG: Midguts, MT: Malpighian tubules. Carcass is the remainder of the body after removal of organs (SG, MG, MT and ovaries). Shaded bars depict means while error bars show ± standard deviation. Significance was evaluated by two-way ANOVA followed by post hoc Tukey tests (to evaluate expression in four organs across two feeding statuses), and by one-way ANOVA with post hoc Tukey test (to compare five organs in engorged females only; MT expression data was available for engorged females only)

IsAQP1 shows organ and life-stage specific expression differences

Using qPCR, we examined IsAQP1 expression in the major organs of adult females including salivary gland, ovary and midgut in both unfed and engorged animals. We additionally measured expression in Malpighian tubules in fed females only as tubules from unfed ticks were very small and yielded almost no detectable RNA. Across major female organs, IsAQP1 expression differed significantly by both organ type and feeding status, with a significant organ × feeding status interaction (Figure 3, two-way ANOVA, p < 0.0001, p = 0.011, p = 0.030, respectively). IsAQP1 expression was higher in salivary glands compared to other organs, both in the full two-way model (post hoc Tukey tests, p < 0.001 for each pairwise comparison), as well as in a one-way analysis of engorged females that included Malpighian tubules (post hoc Tukey tests, p < 0.0001 for each comparison), but there were no other differences between organs (p > 0.05 for all). After females fed on blood and became engorged, expression in salivary gland fell significantly to 36.5% of the levels of unfed females (Figure 3, post hoc Tukey test, p = 0.013). Expression of IsAQP1 in other organs did not change with feeding (post hoc Tukey test, p > 0.05 for all).

We additionally measured transcript expression across life stages including egg, larva, nymph and adults of both sexes. Eggs and nymphs showed minimal IsAQP1 expression, while expression at other stages (larva, adults) was higher on average, but highly variable across biological replicates (Figure S1). Expression in male and female adults was also characterized by high variation across replicates, but the sexes did not appear to differ (Figure S1). Due to high variability between replicates, we did not statistically compare IsAQP1 expression in these stages or sexes.

DISCUSSION

Here, we identify AQP-like genes in I. scapularis and use these sequences and others (176 total) to study AQP evolution in arachnids and other arthropods using phylogenetic methods. Consistent with previous work, insect-specific aquaglyceroporins (Entomoglyceroporins, or “Eglps”) clustered together within classical water-specific AQPs (Finn et al., 2015). In contrast to previous work (Stavang et al., 2015), we find that arachnid AQP-like sequences formed a major clade which seems to derive from classical GLPs. More specifically, our analysis placed a majority of arachnid AQP-like sequences (n = 68/94), including 51/52 non-Bib AQPs from 13 tick species, into a monophyletic clade nested within GLPs (Figure 1b). We also found a few putative classical AQP and Bib-like genes in arachnids—though none of these has been functionally described to date. Our findings, therefore, suggest that in arachnids (and potentially in their closest relatives as well; Ballesteros et al., 2022), the GLP gene family has expanded considerably while some (but not all) classical water-specific AQP genes were lost, perhaps beginning soon after they split from insects and other mandibulates more than 550 million years ago (Giribet & Edgecombe, 2019). Further work—for example, to carefully inventory and characterize the AQPs of focal taxa—is needed to understand the evolution of this gene family in arachnids including ticks.

We examined and characterized both the IsAQP1 genomic locus and gene product. We found the IsAQP1 locus to be intron-rich, as expected (Gulia-Nuss et al., 2016). Next, we functionally characterized the permeability of IsAQP1 to water, glycerol and urea. Biochemical characterization showed IsAQP1 has high water channel activity—moving water even more efficiently than a control classical AQP from an insect (Figure 2a). Yet unlike most GLPs, IsAQP1 was a poor conductor of glycerol and was impermeable to urea (Figure 2b,c). Thus, despite being phylogenetically a GLP, IsAQP1 behaves similarly to a classical AQP (though with some remaining glycerol permeability). This finding mirrors a recent study in insects, in which Finn et al. (2015) described the evolution of GLP-like channels (i.e. permeable to both water and other specific solutes) from classical AQP precursors. Together, our phylogeny and functional analyses imply the converse: that a (largely) water-specific channel—IsAQP1—evolved from a GLP precursor in this arachnid. In other words, arachnids (and especially ticks) may have co-adapted classical GLP homologues to function as water-specific AQPs, an evolutionary path likely related to the depauperate condition of the classical AQP gene family in this clade. Only one other tick AQP, RsAQP1 from the brown dog tick, Rhipicephalus sanguineus, has been biochemically characterized. RsAQP1 has water channel activity that was inhibited by mercury (Hg²⁺, 100 μM), but it did not have glycerol or urea channel activity (Ball et al., 2009). The same authors mentioned in a review as unpublished data that another R. sanguineus AQP exhibited high glycerol and urea channel activity (Campbell et al., 2008). These findings suggest that AQP-like channels may be more evolutionarily labile than was previously assumed, though more studies are needed to functionally characterize AQPs in diverse arachnids to verify the hypothesis that arachnid AQPs are derived primarily from the GLP lineage.

The diversity of AQPs across arthropods may reflect the diverse roles of these channels in maintaining homeostasis and many other aspects of their biology. AQPs function not only in diuresis and desiccation resistance (e.g. Drake et al., 2010; Liu et al., 2011), but also heat tolerance, intrauterine lactation and cold tolerance (Benoit et al., 2014; Philip et al., 2008, 2011). The unusual osmoregulatory challenges faced by ticks across their life cycle could have shaped the evolution of their AQPs. Broad scale patterns of evolution may be visible in AQP gene evolution (e.g. Finn, Chauvigné, Hlidberg, Cutler, & Cerdà, 2014), but more work is needed to understand the drivers of AQP evolution in arthropods including hematophagous insects and arachnids.

Unlike many AQPs (e.g. Campbell et al., 2008; Kuwahara, Gu, Ishibashi, Marumo, & Sasaki, 1997; Shi & Verkman, 1996), the water channel activity of IsAQP1 was not inhibited by mercury (Figure 2a). Mercury inhibition is thought to rely on the ion binding to specific cysteine residues (Hirano et al., 2010; Jung, Preston, Smith, Guggino, & Agre, 1994; Kuwahara et al., 1997; Savage & Stroud, 2007), though the effects of mercury on AQPs can be complex (e.g. Frick et al., 2013). IsAQP1 has three cysteines including Cys²⁰⁰, a residue that may be homologous to the verified mercury-sensitive cysteine of other AQPs (e.g. Cys¹⁸⁹ of human AQP1 and Cys¹⁸¹ of human AQP2; Preston, Jung, Guggino, & Agre, 1993), both of which sit three residues proximal to the second NPA motif (Figure S2; Shi & Verkman, 1996). However, Cys²⁰⁰ of IsAQP1 is positioned further upstream from the NPA domain (Figure S2) and this positional shift may explain the loss of mercury sensitivity. Publicly available structural predictions made with AlphaFold2 (Jumper et al., 2021; Varadi et al., 2022) suggest Cys²⁰⁰ may lie within an $a$-helix while Cys¹⁸⁹ and Cys¹⁸¹ are exposed residues within the channel pore, allowing them to bind mercury. Explicit functional studies would be needed to test this hypothesis—for example, to functionally test the mercury sensitivity of IsAPQ1 with Cys²⁰².

Our expression data also shed light on the function of IsAQP1. We detected some IsAQP1 expression at multiple life stages and within multiple organs, but expression was strikingly enriched in adult female salivary glands (Figure 3). IsAQP1 expression was especially elevated in the salivary glands of un-engorged females (Figure 3), that is, expression was highest ahead of an anticipated bloodmeal. High AQP expression in salivary glands has been noted in other tick species including Haemaphysalis qinghaiensis (Niu et al., 2022), Ixodes ricinus (Campbell et al., 2010) and Rhipicephalus sanguineus (Ball et al., 2009). IsAQP1 additionally showed low expression in the Malpighian tubules (Figure 3), but very little expression in gut and ovary—two organs commonly reported to have high AQP expression in ticks (Niu et al., 2022; Ball et al., 2009; Holmes et al., 2008; Campbell et al., 2010). AQPs are typically expressed in a tissue-specific manner (e.g. Campbell et al., 2008; Tsujimoto, Liu, Linser, Agre, & Rasgon, 2013; Van Ekert et al., 2016), and expression level may be adjusted dynamically to meet the fluctuating water balance needs of the animal (e.g. Hu et al., 2020; Liu et al., 2011; Martini et al., 2004; Rosendale, Dunlevy, McCue, & Benoit, 2019). Because salivary glands are the major osmoregulatory organ in ticks (Bowman & Sauer, 2004; Kim, Šimo, Vancová, Urban, & Park, 2019), this expression pattern suggests that IsAQP1 functions in water balance before or during blood feeding. Prior to blood feeding, ticks may use this water channel to sequester water vapour from the air. Alternatively, while feeding on hosts, ticks must return 70% of the water and ions from acquired blood back into the host (Bowman & Sauer, 2004), which implies a need for an efficient water channel such as IsAQP1. To shed further light on this question, future work could examine how the expression of the complete repertoire of AQPs found in I. scapularis changes (or remains stable) across life stages and in response to feeding.

We found high variability among biological replicates when examining IsAQP1 expression across life stages in ticks sourced from BEI Resources. The source of this expression variation is unclear, but it could be related to uncontrolled variation in age, housing conditions (e.g. Hu et al., 2020; Rosendale et al., 2019), or shipping conditions prior to arrival in our laboratory. Repeating this work using animals kept in controlled laboratory conditions would be necessary to determine if this variation arises from inherent biological sources or from environmental perturbation.

Currently, tick control primarily involves the use of chemical acaricides, yet these tools lead to the evolution of resistance and environmental contamination (e.g. Murgia et al., 2019), leaving us in urgent need of new methods of tick control. Our work suggests tick AQPs, including IsAQP1, may be effective molecular targets for new control efforts aimed at disrupting essential biological processes in ticks. Because pathogens are passed from ticks to hosts via saliva, the AQPs involved in salivation remain attractive targets for the development of novel control methods. Indeed, antitick vaccination strategies targeting AQPs have proven effective in field trials (de la Fuente, Kopáček, Lew-Tabor, & Maritz-Olivier, 2016; Guerrero et al., 2014; Scoles, Hussein, Olds, Mason, & Davis, 2022). We attempted to knock down IsAQP1 expression using RNAi for in vivo functional characterization that could give insight into potential control targets, but gene transcription was unaffected by dsRNA injection compared to controls (data not shown). Thanks to the development of new methods for targeted genetic manipulations this species (Sharma et al., 2022), future research could functionally test the importance of this gene in tick physiology and explore its potential for use as a tool of vector control.

EXPERIMENTAL PROCEDURES

All methods were carried out in accordance with The Pennsylvania State University institutional guidelines and regulations.

Animal and tissue handling

Ticks were obtained from both the field and from BEI Resources. Adult Ixodes scapularis were collected in State College, PA using a standard canvas-dragging method. Eggs, larvae, nymphs, male and female adults and engorged female Ixodes scapularis were obtained from BEI Resources (NIAID, NIH), and kept under standard conditions by the supplier (Troughton & Levin, 2007). Field-collected ticks were used for cloning and sequencing IsAQP1, while purchased ticks were used for organ- and life stage-specific expression analyses.

For cloning and sequencing, field-collected ticks were either processed immediately or stored at −80°C until processing. For gene expression analysis, (including organ dissection) ticks were kept alive at room temperature (22–25°C) and ambient conditions for 0–4 days prior to processing. The majority of whole animals were shipped in a single batch from BEI Resources and were processed on the day of arrival (N = 507/537), with 30 adult females arriving in a later shipment that was processed four days postdelivery. Organs were harvested from both unfed and engorged adult females, which arrived in separate shipments from BEI Resources. Organs were dissected from animals 0–3 days postdelivery, using a balanced design such that any effects of time were balanced across groups. Organs (salivary glands, midgut, Malpighian tubules and ovaries) were dissected in PBS, pooled, and either immediately processed or stored in TRI Reagent (Sigma Aldrich, Saint Louis, MO) at −80 °C until use. Pools were composed as follows: 120–160 individual larvae, 25 nymphs, 8–20 adult males (whole animals or organs) and 8–10 adult females (whole animals or organs). A volumetrically similar clump of eggs comprised a replicate of pooled eggs.

RNA extraction, cDNA synthesis and RACE

Whole ticks or dissected organs were homogenized in TRI Reagent (Sigma Aldrich, Saint Louis, MO) in a 1.5 mL microcentrifuge tube with a pestle (Kimble Kontes, Fisher Scientific, Hampton, NH). Whole RNA was extracted following the manufacturer's instructions. Extracted RNA was treated with DNase (TURBO DNAfree, Life Technologies, Carlsbad, CA) for 1 hr at 37°C. First-strand cDNA synthesis was performed using Accuscript High Fidelity Reverse Transcriptase (Agilent Technologies, Santa Clara, CA) with oligo d(T)₂₀ primer on 1 μg of total RNA at 42°C for 2 h.

We used RLM-RACE (RNA Ligase-Mediated Rapid Amplification of cDNA Ends) to determine the unknown sequence at the 5′ and 3′ ends of IsAQP1 cDNA as previously described (Tsujimoto et al., 2013). Briefly, RNA was extracted from adult ticks (both male and female adults) using TRI Reagent. RNA processing for RACE was done according to the manufacturer's instructions (GeneRacer, Life Technologies, Carlsbad, CA). RACE PCR was performed with gene-specific primers based on the predicted gene sequence at VectorBase.org (ISCW003957) plus an anchor primer (3′ RACE) or GeneRacer 5′ primer/GeneRacer 5′ nested primer (5′ RACE) (Table S1). RACE PCR products were separated by agarose gel electrophoresis, and bands were excised for gel extraction using a Gel Extraction Kit (QIAGEN, Valencia, CA). Gel-extracted PCR products were cloned into pJET1.2/blunt vector using CloneJET PCR Cloning Kit (Fermentas, Glen Burnie, MD). The inserts were then Sanger sequenced, and the resulting full length cDNA sequence was submitted to GenBank (accession number KT988052).

Mapping exons in genomic sequence

We mapped the IsAQP1 cDNA sequence to the genome to determine the location and structure of the gene locus. We completed a BLASTn search of the I. scapularis genome sequence at VectorBase (vectorbase.org) using the full-length cDNA as query. We then manually adjusted exon-intron boundaries so that all introns have a canonical splice signal (GT-AG) and submitted our revised annotation to VectorBase. Visualization of the IsAQP1 locus was accomplished using the web-based software FancyGene (Rambaldi & Ciccarelli, 2009).

Phylogeny estimation

We obtained AQP sequences from public sources (e.g. GeneBank, FlyBase, VectorBase, transcriptome shotgun assemblies, whole genome shotgun contigs). Sequences were identified via arthropod-specific searches for “aquaporin”, BLAST searches using arthropod aquaporin sequences as query, and from previous studies (Stavang et al., 2015; Niu et al., 2022). Scorpion transcriptome contigs were provided by Dr. Thorsten Burmester (Zoological Institute and Museum, University of Hamburg, Germany; Roeding et al., 2009). All taxon and accession information is given in Table S1. Polypeptide sequences (translated from nucleotide sequence as needed at expasy.org) were used as input, and all sequences ≥200 amino acids with at least one NPx motif (where x = any amino acid) were included in phylogeny estimation. Sequences were aligned using SeaView (Gouy, Guindon, & Gascuel, 2010) built in Clustal Omega (Sievers et al., 2011), where unaligned regions were filtered using the “Gblocks” option with the least stringent options (i.e. allowing smaller final blocks, gaps and less strict flanking positions). For simplicity of the resulting tree, when more than one AQP-like sequence from the same species showed greater than 95% identity, only the longest sequence was kept. The phylogenetic tree was constructed using maximum likelihood in PhyML 3.1 with 500 bootstraps, and plotted using FigTree v1.4.4.

Cloning, expression in Xenopus laevis oocytes and osmotic swelling assays

We used X. laevis (hereafter, Xenopus) oocyte swelling assays to evaluate the channel properties of the cloned I. scapularis AQP-like gene. The sequence was cloned as previously described (Tsujimoto et al., 2013). Briefly, the complete coding sequence was then amplified using primers with a restriction site at the 5′ end (MfeI for upstream primers and NheI for downstream primers). The resulting PCR products were initially cloned into pJET1.2/blunt vector (Fermentas, Glen Burnie, MD) and after restriction digestion of the cloned insert, the IsAQP1 CDS was ligated into EcoRI/NheI-treated pXβG-myc. The purified construct was linearized using XbaI for in vitro complementary RNA (cRNA) transcription with T3 RNA polymerase (Agilent Technologies, Inc., Santa Clara, CA). Synthesized cRNA was purified with the RNeasy Mini kit (Qiagen), and its integrity was analysed by denaturing agarose gel electrophoresis. For comparison controls, we additionally evaluated AQP genes AgAQP1B, ClAQP1 and ClAQP1/ClGlp1 co-expression (Tsujimoto et al., 2013, 2017) using the same methods, allowing for direct comparison of channel properties. Xenopus ovaries were purchased from Xenopus I (Dexter, MI), defolliculated with collagenase I (Sigma) and injected with 50 nL of 100 ng/μL cRNA solution (5 ng) or 50 nL of nuclease-free water as control. cRNA-injected oocytes were incubated for 3 days in ~200 mOsm Modified Barth's Solution (MBS: 88 mM NaCl, 1.0 mM KCl, 2.4 mM NaHCO₃, 15 mM Tris, 0.8 mM MgSO₄, 0.4 mM CaCl₂, 0.3 mM Ca(NO₃)₂), pH 7.6, containing 0.5 mM theophylline, 100 U/mL penicillin and 100 μg/mL streptomycin) at 16°C to allow for overexpression of aquaporin in the oocyte plasma membrane. The oocytes were then subjected to a swelling assay as previously described (Tsujimoto et al., 2017). Briefly, we transferred oocytes to MBS diluted to 70 mOsm. The change of oocyte volume was monitored at room temperature using a digital camera (Olympus DP-21) equipped stereo-microscope (Olympus SZX-16) for 60 s with the relative volume (V/V₀) calculated from the cross-sectional area at the initial time (A₀) and after the specified time interval (A_t): V/V₀ = (A_t ∕A₀)^3∕2. The coefficient of osmotic water permeability (P_f) was determined from the initial slope of the time course [d(V/V₀)/dt], average initial oocyte volume (V₀ = 9 × 10⁻⁴ cm³), average initial oocyte surface area (S = 0.045 cm²), the molar volume of water (V_w = 18 cm³/mol) and the osmotic solute gradient (osm_in − osm_out: P_f = (V₀ × d(V/V₀)/dt)/(S × V_w × (osm_in − osm_out)). A minimum of six (range 6–12) individual oocytes were tested for each group.

To test the effect of mercury (Hg²⁺) on water permeability (Preston, Carroll, Guggino, & Agre, 1992), oocytes were placed in 500 μM HgCl₂ in MBS for 5 min prior to the swelling assay. We tested the reversibility of Hg²⁺ treatment by transferring Hg²⁺-treated oocytes (500 μM, 5 min) to 5 mM 2-mercaptoethanol in MBS for 10 min.

Glycerol and urea permeability were also assessed in separate assays by replacing diluted MBS (i.e. 70 mOsm) with MBS that was mixed with an equal volume of iso-osmotic (200 mOsm) glycerol or urea solution. Data were compared by one-way ANOVA followed by Tukey's correction for multiple comparisons. All analyses were completed using Graphpad Prism software.

Real-time quantitative PCR (qRT-PCR)

We used qRT-PCR to quantify the expression of IsAQP1 in different life stages and organs. Primers were designed using Primer3 (Koressaar & Remm, 2007; Untergasser et al., 2012), and designed to amplify 53–145 bp fragments of the cDNA (Table S2). We empirically verified that all primer pairs have an amplification efficiency (E) between 0.9 and 1.1 using the same qPCR protocol. cDNA was diluted to 1/50 in nuclease-free H₂O, of which 2.5 μL was used for 10 μL reactions using the RotorGene Q qPCR system (QIAGEN). Reactions were performed with 5 min at 95°C, followed by 45 cycles of 95°C for 5 s, 60°C for 10 s and melt curve analysis from 50 to 95°C. Expression was calculated using the Pfaffl (2001) method relative to the housekeeping gene (Glyceraldehyde 3-phosphate dehydrogenase, GAPDH), a gene shown to have stable expression across tissues, life stages and feeding status in ticks (Browning, Adamson, & Karim, 2014; Bullard, Williams, & Karim, 2016; Koči, Šimo, & Park, 2013; Nijhof, Balk, Postigo, & Jongejan, 2009). Expression levels in organs (salivary gland [SG], ovary, midgut [MG], and remainder of carcass) $\times$ feeding status, as well as their interaction, were statistically compared using two-way ANOVA followed by post hoc Tukey test using the aov() and TukeyHSD() functions in R. Post hoc, we compared IsAQP1 expression in organs (including Malpighian tubules [MT]) in engorged females only using a one-way ANOVA with post hoc Tukey test, using the same functions. Malpighian tubule samples from unfed ticks yielded no RNA, likely due to the small size of this organ in unfed animals.

AUTHOR CONTRIBUTIONS

Hitoshi Tsujimoto: Conceptualization; investigation; writing – original draft; writing – review and editing; formal analysis. Hillery C. Metz: Writing – original draft; writing – review and editing; formal analysis. Alexis A. Smith: Investigation. Joyce M. Sakamoto: Investigation. Utpal Pal: Investigation; supervision; funding acquisition. Jason L. Rasgon: Conceptualization; funding acquisition; writing – original draft; writing – review and editing; formal analysis; project administration; supervision.

ACKNOWLEDGEMENTS

This study was supported by USDA Hatch funds (Project PEN04769), a grant with the Pennsylvania Department of Health using Tobacco Settlement Funds, and funds from the Dorothy Foehr Huck and J. Lloyd Huck Endowment to JLR, NSF/BIO grant 1645331 to JLR and JMS, and NIH/NIAID grant R01AI080615 to UP. BEI Resources (NIAID, NIH) provided Ixodes scapularis used in this research (NR-42510). We thank Dr. Thorsten Burmester (Zoological Institute and Museum, University of Hamburg, Germany) for providing the contig set from the scorpion transcriptome.

CONFLICT OF INTEREST

The authors have no conflicts of interest to disclose

Open Research

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are openly available in Genbank at https://www.ncbi.nlm.nih.gov/genbank/, reference number K988502.