The insulin-like growth factor (IGF) binding proteins (IGFBP) family is the regulatory arm of the IGF signaling system that control mitogenic and anabolic actions of IGF peptide hormones. This study describes cloning and biological characterization of three Amblyomma americanum (L.) (Aam) proteins that show amino-terminal sequence and secondary structure similarity to the IGFBP superfamily. The three molecules here provisionally identified as AamIGFBP-rP1 and short (S) and long (L) AamIGFBP-rP6 are expressed in multiple tick organs and are responsive to tick feeding activity with the former being upregulated and the latter being downregulated. We show that they regulate tick physiological functions that may be related to A. americanum tick feeding success as revealed by RNAi-mediated dual silencing of AamIGFBP-rP6S and AamIGFBP-rP6L or AamIGFBP-rP1 alone, which caused a reduction in blood meal size compared to the controls. Additionally, in the case of AamIGFBP-rP1 silencing, 47% of ticks died while attempting to feed and those that did survive and spontaneously detached from the host failed to lay eggs. Although AamIGFBP-rP6S and AamIGFBP-rP6L show overall identities of 49% and 59%, respectively, to Rhipicephalus microplus C protein, the identity level jumps to ~84% when the comparison is restricted to first 70 amino acids of the mature protein. Similarly, the AamIGFBP-rP1 mature protein is ~72%, 87%, 88% and 92% identical to that of Ixodes scapularis S, R. microplus, R. appendiculatus N and A. variegatum F, respectively. The observed across-tick-species conservation suggests that the three molecules (AamIGFBP-rP1, AamIGFBP-rP6S and AamIGFBP-rP6L) represent target for development of vaccines to protect animals against multiple tick species. The data are discussed with reference to advances in tick molecular biology and the potential of the three proteins as targets for immunizing animals against tick feeding.
Ticks are among the most important arthropods affecting humans and animals (Sonenshine, 1993). Direct tick feeding activity can cause damage to their hosts, paralysis, irritation and allergy. However, ticks are mostly known for their role as vectors of numerous tick borne disease (TBD) agents: rickettsiae, protozoa, spirochetes and viruses (Sonenshine, 1993; Jonejan and Uilenberg, 2004). Until the 1980s and early 1990s, when many emerging human TBD agents were described (Spach et al., 1993; Azad and Beard, 1998; Bratton and Corey, 2005; Perlman et al., 2006), ticks were mostly regarded as a veterinary problem. Management and treatment of major animal TBD agents, Babesia spp. (babesioses), Theilleria spp. (theillerioses), Anaplasma spp. (anaplasmoses) and Ehrlichia ruminatium (heartwater) continue to be a source of significant monetary losses in the livestock industry especially in tropical and sub-tropical countries (Jongejan and Uillenberg, 2004). In public health, cases of human TBDs, Lyme disease (Borrelia burdoferi), human granulocytic anaplasmosis (A. phagocytophilum) and human monocytic ehrlichiosis (Ehrlichia chaffensis) have been on the rise (Branton and Corey, 2005).
With the exception of Francisella tularensis (Petersen et al., 2009) and Coxiella burnetti (Simor et al., 1984; Reimer, 1993), which can be acquired through other means, all TBD agents require successful tick feeding to be transmitted. Thus, the avoidance of tick bites through suppression of tick vector populations is the method of choice to prevent TBD transmission. Tick vector populations have traditionally been controlled by acaricide application. Acaricides are essential in the short term, but they do not offer a permanent solution because of disadvantages such as widespread tick resistance to acaricides and contamination of the environment and food chain (Frisch, 1999). Vaccination of animals against tick feeding has been shown to be a practical and sustainable alternative (Frisch, 1999; Willadsen, 2004; Willadsen, 2006; de la Fuente et al., 2007). We are interested in understanding the molecular mechanisms that underlie the regulation of early stage tick feeding. Understanding the feeding process will lead to discovery of tick proteins that are important to pathogen transmission by ticks.
Tick feeding involves multiple behavioral changes that begin with hunger and ends with satiation (Walade and Rice, 1982). The sequence of events leading to successful tick feeding were categorized into nine behavioral changes: (1) attainment of appetence, (2) engaging the host, (3) exploring the host for a feeding site, (4) penetration, (5) attaching onto host skin and establishing the feeding site, (6) ingestion of host blood, (7) engorgement, (8) detaching from the host and (9) disengaging from the host (Walade and Rice, 1982). Identifying tick proteins that regulate various phases of the feeding cycle should enable us to develop novel strategies for controlling ticks and tick-borne pathogens. Several studies have focused on characterizing proteins expressed by ticks during the slow feeding phase (3- to 5-day fed ticks) (Mulenga et al., 2001; Mulenga et al., 2003; Valenzuela et al., 2002; Ribeiro, 2006) primarily because of the ease of obtaining their tissues, in that they are larger ticks. Multiple bioactive enzymes, anti-inflammatory agents, anti-complement, and anti-coagulants, have been identified in this manner (Ribeiro et al., 2006). We are interested in unfed tick proteins that may regulate the preparatory feeding phase as this precedes the key facets of tick parasitism, blood meal feeding and pathogen transmission. We previously identified 40 genes that were differentially upregulated or induced in unfed Amblyomma americanum ticks that were stimulated to start feeding (Mulenga et al., 2007). While analyzing these sequences, we identified two sequence fragments that showed significant amino-terminal structural similarity to the insulin-like growth factor (IGF) binding protein (IGFBP) family.
Members of the IGFBP family are evolutionary conserved cysteine-rich extracellular proteins that have been identified in a wide range of organisms from humans, fish and chickens (Rodgers et al., 2008; Kamangar et al., 2006) to, fruit fly, army fall worm and shrimps (Castellanos et al., 2008; Honegger, 2008; Alic and Partridge, 2008; Andersen et al., 2000). In vertebrates where most studies have been conducted, they are the regulatory arm of the IGF signaling system, a complex interaction that includes IGF-I and IGF-II peptide hormones, IGF receptors and IGFBP proteases (Hwa et al., 1999; Burger et al., 2005; Delaney, 2005). The IGFBP family include six proteins, IGFBP-1 to IGFBP-6 that bind IGF peptide hormones with high affinity and a group of 10 IGFBP-related proteins (IGFBP-rP-1 to -10) that show significant amino-terminal sequence similarity to IGFBP-1 to -6 but bind IGF peptide hormones with low affinity (Hwa et al., 1999; Burger et al., 2005; Delaney, 2005). The six high-affinity binding proteins, IGFBP-1 to IGFBP-6 control anabolic and mitogenic functions of the IGF peptide hormones through sequestration, and indirectly by acting as their transporters, protecting them from degradation, limiting their binding to receptors and maintaining a reservoir of biologically inactive IGFs (Hwa et al., 1999; Delaney, 2005). In other studies, some IGFBPs and IGFBP-rPs have been shown to have intrinsic cell proliferation and tissue growth regulatory functions that are independent of their ability to bind IGF proteins (Hwa et al., 1999; Burger et al., 2005; Delaney, 2005; Sato et al., 2007; Ahmed et al., 2003). Although at the sequence level IGFBP-like proteins have been identified in arthropods, very few functional studies have been reported. A 27 kDa protein that shows sequence similarity to vertebrate IGFBP-rP1 and can bind insulin and IGF peptide hormones has been characterized in Spodotera frugiperda (Anderson et al., 2000) and Drosophila melanogaster (Honegger, 2008; Alic and Partridge, 2008). The Drosophila protein, ImPL-2 was recently shown to negatively regulate growth similar to vertebrate IGFBP-rP1 protein (Alic and Partridge, 2008). In ticks, two indirect studies that reported insulin-like reactivity in the synganglia of Ornithodoros parkeri (Zhu and Oliver, 1991) and nymphal and adult Dermacentor variabilis ticks (Davis et al., 1994), but there are no reports of the IGFBP family in ticks. In this study we have, for the first time, characterized three tick IGFBP sequences that we have provisionally identified as A. americanum (Aam) short (S) and long (L) IGFBP-rP6 and AamIGFBP-rP1. We provide evidence suggesting that biological functions regulated by the three molecules are important to A. americanum tick feeding success as revealed by RNAi-mediated silencing of the three genes, which caused ticks to obtain significantly smaller blood meals. Most interestingly, AamIGFBP-rP1 silencing also caused 47% mortality, and ticks that did survive and spontaneously detached from the host failed to lay eggs. The data are discussed with reference to advances in tick molecular biology and the potential of the three molecules as target anti-tick vaccine antigens.
MATERIALS AND METHODS
Tick dissections and total RNA isolation
Adult female ticks used in this study were obtained from a colony of A. americanum that is maintained on cattle and chickens for use in the laboratory of Dr Pete Teel, in our department. Tick dissections were routinely done as previously described (Mulenga et al., 2003). Briefly, 10 ticks that were partially fed for 24, 48, 72, 96 and 120 h on cattle were rinsed in 70% ethanol. Subsequently ticks were held onto a glass slide with a pair of soft tissue forceps and then the alloscutum edges were trimmed off using a sharp and sterile razor blade. Under a dissection microscope, the dorsal cuticle flap was lifted and tick organs, salivary glands (SG), midgut (MG) and ovary (OV) were teased out using an 18-gauge needle and a soft tissue forceps. Tick organs including the carcass (CA) representing the tick remnant after removal of SG, MG and OV pooled in groups of three or four were stored in RNAlater (Ambion, Austin, TX, USA) at −80°C until used for RNA extraction.
Extraction of total RNA from whole ticks and dissected tick organs was done using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Briefly, within the first hour of being detached from the host, whole ticks that were partially fed for 24–120 h were rinsed in 70% ethanol, dried between paper towels, pulverized in liquid nitrogen, and transferred to TRIzol reagent for total RNA extraction. Similarly, for dissected tick organs, they were rinsed in DEPC water to remove the storage solution and then transferred to the TRIzol reagent for RNA extraction. Tissue lysis was accomplished either by repeated pipetting (SG, MG and OV) or homogenization (CA and whole ticks) using a sonic dismembrator, model 100 (Fisher Scientific, Pittsburgh, PA, USA). Extracted total RNA was reconstituted in RNase-free water and stored at −80°C until used.
Cloning and sequence analysis of full-length Amblyomma americanum insulin-like growth factor binding proteins-related proteins
Templates for 3′ and 5′ RACE were synthesized from 5 μg of total RNA extracted from whole ticks that were partially fed for 5 days using the RLM RACE cDNA synthesis kit (Ambion) according to the manufacturer's protocol. Primers based on the short (S) and long (L) partial AamIGFBP-rP6 cDNA sequences that were identified in a previous study (Mulenga et al., 2007) were used with 3′ and 5′ RACE templates to amplify full-length cDNAs. TBLASTN scanning of the tick ESTs in the Gene Indices database at the computational biology and functional genomics laboratory (http://compbio.dfci.harvard.edu/tgi/cgi-bin/tgi/Blast/index.cgi) using AamIGFB-rP6S or AamIGFB-rP6L cDNAs as query, retrieved a tick EST sequence that is highly conserved in A. variegutum (accession no. TC737), Rhipicephalus microplus (accession no. TC17305), R. appendiculatus (accession no. TC5751) and Ixodes scapularis (accession no. TC43412). RACE PCR primers based on consensus sequences of these ESTs were used to clone the third molecule. PCR products were routinely cloned in TOPO TA cloning vectors (Invitrogen) and sequenced using M13 TOPO plasmid-specific primers.
Sequence and bioinformatics analyses
DNA and deduced protein sequences were routinely analyzed using the Vector NTI (Invitrogen) or the MacVector (Accelrys, San Diego, CA, USA) software packages. For comparison with other proteins and provisional identification, DNA and deduced protein sequences were scanned against known entries in GenBank or the Gene Indices database at the computational biology and functional genomics laboratory using the BLASTX, BLASTP and TBLASTN homology search program. Additionally deduced amino sequences were submitted to the ExPASY Proteomics Server (http://ca.expasy.org/) for prediction of signal peptides, and identification of domains and amino acid motifs.
To gain insight into secondary structures of AamIGFBP-rPs, the crystal structure of human IGFBP-4 complexed with human IGF-I [2DSP chain B (Sitar et al., 2006)] was chosen as a molecular template based on PDB-BLAST. Structure-based sequence alignment was performed using Expresso (Armougom et al., 2006) with some slight manual adjustment. Comparative modeling was performed by MODELLER 9v1 (Sali and Blundell, 1993) using sequence alignment in PIR format (generated by GeneDoc) and 2DSP chain B (human IGFBP-4, chain B subtracted by WebLab Viewer as input data. To assess the quality, the AamIGFBP model was evaluated using Verify3D (Luthy et al., 1992) and secondary structures visualized by PyMol 0.99 (DeLano, 2002).
Expression analyses by RT-PCR
To determine temporal and spatial expression patterns, oligo(dT) primed cDNA templates of whole ticks that were fed for 24, 48, 72, 96 and 120 h and tick organs, SG, MG, OV and CA of 120 h of ticks fed ticks were subjected to standard two-step RT-PCR. About 5 μg of total RNA was used to synthesize template cDNA using the VersoMax cDNA synthesis kit (ThermoScientific, New York, NY, USA). Prior to synthesis of cDNA, total RNA was treated with RQ1 DNAse according to the manufacturer's instructions (Promega, Madison, WI, USA) to remove genomic DNA contamination. A 1μl aliquot of template cDNA was used in a PCR reaction with gene-specific PCR primers. PCR products were electrophoresed on a 2% agarose gel containing 1 μgml−1 ethidium bromide. The tick actin gene (Mulenga et al., 2009) was used for PCR template load control. To determine mRNA abundance, densitograms of amplified PCR bands were determined using the web based ImageJ software (http://rsb.info.nih.gov/ij/). To correct for variations in PCR template concentrations, PCR band densities were normalized as previously described (Mulenga et al., 2008).
RNA interference-mediated gene silencing
To validate their significance in tick feeding, RNA interference (RNAi)-mediated silencing of AamIGFBP-rP6S and AamIGFBP-rP6L together, or AamIGFBP-rP1 alone, was conducted as previously published (Mulenga et al., 2008). Briefly, double stranded (dsRNA) was synthesized in vitro using the MegaScript RNAi kit according to instructions by the manufacturer (Ambion). The rationale to silence AamIGFBP-rP6S and AamIGFBP-rP6L mRNA together in the same tick was based on their high sequence similarity. Templates for dsRNA synthesis were amplified from cloned plasmid DNA using gene-specific PCR primers (AamIGFBP-rP6S/L, For: 5′-GGAACCGCACTGCGAGGGCGTGA-3′, Rev: 5′-CCGGGAAGAACCCTCAGCGAGG-3′; AamIGFBP-rP1, For: 5′-GACGGGCCGTGCGGCGAGGA-3′, Rev: 5′-CCAGTAGGTGGCCGTGTCCTCTGG-3′ with added T7 promoter sequence (5′-TAATACGACTCACTATAGGG-3′). Green fluorescent protein (GFP, For: 5′-TCACGAACTCCAGCAGGACCATGTGATC-3′, Rev: 5′-ACGTAAACGGCCACAAGTTCAGCGTGTC-3′) PCR primers with added T7 promoter sequence were used to amplify templates for control dsRNA synthesis. Following purification of dsRNA and adjustment of concentration, three groups of 25 unfed A. americanum female ticks were microinjected with ~1 μl (~2–3 μg μl−1 in TE buffer pH 8.0) of a mixture of AamIGFB-rP6S and AamIGFB-rP6L dsRNA, AamIGFBP-rP1 dsRNA alone or GFP dsRNA via the groove between the basis capituli and the scutum using half-inch 33 gauge needles attached to a 10μl gas-tight syringe (Hamilton, NV, USA). Injected ticks were kept overnight at 22°C to observe any mortality due to injection injury. Injected ticks that were alive the next morning (AamIGFBP-rP1-dsRNA-injected N=21, AamIGFBP-rP6S/L-dsRNA-injected, N=22) and control ticks (no-injected, N=25, GFP-dsRNA-injected, N=22) were placed on cattle to feed. Ticks were confined to feeding area in cells that were secured on the back of the calf using livestock identification cement (Nasco, Ft Atkinson, WI, USA). Ticks were allowed to feed until they naturally detached from the host. Prior to placing female ticks in cells, male ticks (25 per cell) were pre-fed for 3–4 days in to stimulate the release of aggregation pheromones. We have observed that this arrangement synchronizes the female tick feeding cycle.
To validate whether the injection of dsRNA triggered the disruption of target gene mRNA, three ticks per treatment were detached after 48–60h. The 48–60h time point was chosen to ensure that the ticks that were sampled for the assay were alive and had a chance to attempt feeding. These ticks were individually processed for total RNA extraction. Total RNA was treated with RQ1 DNAse (Promega, Madison, WI, USA) to remove genomic DNA contamination and then subjected to standard two-step semi-quantitative RT-PCR. To forestall the possibility of re-amplifying the injected dsRNA during validation of target gene silencing, PCR primers were set to cDNA regions flanking the domain that was used for dsRNA synthesis.
Photographs of ticks that had spontaneously detached from the host were used to document the observed physical phenotypes. To assess the effects of gene silencing on the ability of ticks to complete feeding, tick feeding parameters, numbers of ticks that attached at 24 h following placement of ticks on the animal, mortality during feeding, time to spontaneous detachment from the host and engorgement masses (EM; as an index of amount of blood imbibed) were evaluated. For statistical analysis, the observed EMs were subjected to one way analysis of variance (ANOVA) followed by pair-wise comparison of means using the Tukey HSD (honestly significant differences) test with confidence levels set by default to studentized range statistic (Q) of 0.05 and 0.01 using the web-based statistical calculator, VassarStats (http://faculty.vassar.edu).
cDNAs cloning, provisional identification, sequence analysis and comparative modeling
We cloned three ~1300–1400 base pair full-length cDNAs encoding 152, 168 and 248 amino acid residue polypeptides that we have provisionally identified as AamIGFBP-rP6S (accession number GU907778), AamIGFBP-rP6L (accession no. GU907779; Fig. 1A) and AamIGFBP-rP1 (accession no. GU907780; Fig. 1B), respectively. Provisional identification was based on conservation of the IGFBP domain (InterPro pattern ‘PS51323’, ExPASY, http//ca.expasy.org), which characterizes the amino terminus region of the IGFBP superfamily (first 70–80 amino acid residues of the mature protein) and the carboxyl-terminal domain structures described by Hwa et al. (Hwa et al., 1999). As shown in Fig. 1B, the carboxyl-terminal regions of AamIGFBP-rP6S and AamIGFBP-rP6L do not show any similarity to known domains or protein motifs, consistent with human IGFBP-rP6 (Hwa et al., 1999). Additionally, similar to human IGFBP-rP6 cDNA, which is characterized by a long 3′ untranslated region (UTR) (Lassalle et al., 1996), both AamIGFBP-rP6S and AamIGFBP-rP6L cDNAs have ~800 bp 3′ UTR (not shown). However, unlike human IGFBP-rP6 3′ UTR, which has a series of AU rich elements, tick sequences do not to contain any (results not shown). Furthermore human IGFBP-rP6 has an open reading frame of 552 (Lassalle et al., 1996) compared to 459 and 507 for AamIGFBP-rP6S and AmIGFBP-rP6L, respectively (not shown). The short and long designation of AamIGFBP-rP6S and AamIGFBP-rP6L is based on the fact that they are highly similar and the major difference is the 14 amino acids (Fig. 1A) or 42 nucleotides (not shown) deletion in the C-terminal region. In the case of AamIGFBP-rP1 (Fig. 1B) its C-terminal domain possess the Kazal-type serine proteinase inhibitor (InterPro pattern ‘PF076648’) and immunoglobulin-like (InterPro pattern ‘PS50835’) domain (Fig. 1B), consistent with human IGFBP-rP1 sequence features (Hwa et al., 1999). Sequence alignments shown in Fig. 1A,B, show that the 12 cysteine residues and the signature amino acid motif ‘CGCCXXC’ that characterize the amino terminus region of the IGFBP superfamily (Hwa et al., 1999) are conserved in ticks (Fig. 1A,B). Scanning of all three putative proteins on the SignalP server (http://www.cbs.dtu.dk/services/SignalP/) revealed a 19 amino acid signal peptide (SP) in AamGFBP-rP6S and AamIGFBP-rP6L, and a 23 SP in AamIGFBP-rP1.
Given this conservation, we thought to develop comparative models of AamIGFBPs based on the first 90 amino acids of the mature protein. The comparative model (Fig. 2) was confirmed to be of high quality as determined by a ‘verify3D’ score of +0.11 to +0.52 for 86 of the 90 residues. Verify3D scores below +0.1 are indicative of serious problems in the model (Janusz et al., 2002), which in this case accounted for only 4 of the 90 residues involved in the model (not shown). Based on the model in Fig. 2A, AamIGFBP amino-terminal domain has secondary structures that are similar to the human IGFBP-4 (Sitar et al., 2006), except that the tick has two α-helices whereas human IGFBP-4 has three (Sitar et al., 2006).
AamIGFBP-rPs are conserved in other ticks
When compared with each other, putative AamIGFBP-rP6S and AamIGFBP-rP6L proteins are ~86 and 89% identical at the amino acid and nucleotide (not shown) levels, respectively. The identity levels drops to ~20% and 28% when both AamIGFBP-rP6S and AamIGFBP-rP6L are compared with human IGFBP-rP6 (accession no. NM_007036) and AamIGFBP-rP1 (not shown), respectively. When scanned against tick EST sequences on the Gene Indices database (http://compbio.dfci.harvard.edu), AamIGFBP-rP6S and AamIGFBP-rP6L showed 49% and 59% overall amino acid identity to a Rhipicephalus microplus (Rm) EST (TC890) translational product (Fig. 1A). However, when sequence comparison was restricted to the first 70 amino acid residues, the identity levels between AamIGFBP-rP6S or AamIGFBP-rP6L and RhmIGFBP-rP6S sequences jumped to 84% and 82%, respectively (not shown). The TC890 EST is provisionally identified as RmIGFBP-rP6S in this study because of closer similarity to AamIGFBP-rP6S (Fig. 1A) and the 27 amino acid deletions in its carboxyl-terminal region (Fig. 1A). In the case of AamIGFBP-rP1, it shows ~92, 88, 87, 72 and 34% amino acid identity to A. variegutum (accession no. TC737), R. microplus (accession no. TC17305), R. appendiculatus (accession no. TC5751), Ixodes scapularis (accession no. TC43412) and human IGFBP-rP1 (accession no. NM_001553), respectively.
AamIGFBPs mRNAi expression analyses by RT-PCR
In a previous study, we used real-time RT-PCR analyses to determine that both AamIGFBP-rP6S and AamIGFBP-rP6L were expressed in the SG, MG, OV and CA (the remaining carcass) and that their transcript abundance decreased with tick feeding (Mulenga et al., 2007). In this study, results summarized in Fig. 3A show that AamIGFBP-rP1 is expressed in SG, MG, OV and CA as revealed by semi-quantitative RT-PCR. AamIGFBP-rP1 is expressed to a similar degree in SG, MG and CA; in the OV, however, the expression level was about one sixth that of the other tissues (Fig. 3B). At the temporal level, AamIGFBP-rP1 is expressed in ticks that were fed for 1–5days (3C) with transcript abundance apparently increasing at least fourfold by day 5 of feeding (Fig. 3D).
RNAi-mediated dual silencing of AamIGFBP-rP6S and AamIGFBP-rP6L or AamIGFBP-rP1 alone caused ticks to obtain significantly smaller blood meals
Qualitative two-step RT-PCR expression analysis summarized in Fig. 4 was used to validate whether or not the injection of dsRNA into ticks triggered disruption of the target gene mRNA. The expectation was that if the mRNA of the target gene, was completely disrupted, its cDNA would not be amplified. On this basis, both AamIGFBP-rP6S and AamIGFBP-rP6L mRNA were partially disrupted as faint PCR bands were detectable (Fig. 4). However, in the case of AamIGFBP-rP1, its mRNA was apparently completely silenced as no bands were amplified (Fig. 4). In order to assess the effects of silencing, ticks were allowed to feed on cattle until spontaneous detachment. At 24 h after ticks were placed on cattle to feed, we observed that both control and treatment ticks were attached to host skin, indicating that silencing of the three molecules did not affect the ability of ticks to start feeding (not shown). However, results summarized in Figs 5 and 6 show that dual silencing of AamIGFBP-rP6S and AamIGFBP-rP6L mRNA or AamIGFBP-rP1 mRNA alone caused ticks to obtain smaller blood meals. Fig. 5 documents the observed physical phenotypes. During the first 4 days of feeding, phenotypes of treatment groups were similar to controls. However, from day 5 of feeding, AamIGFBP-rP1-dsRNA-injected ticks assumed a darkish coloration, which became progressively darker (* in Fig. 5). To measure the effects of silencing on the tick's ability to take in host blood, engorgement mass (EM) was determined. The range of EMs, 780–940mg (N=12) and 750–1070 (N=11) for non-injected and GFP-dsRNA-injected controls, 270–640 (N=11) and 285–541 (N=8) for AamIGFBP-rP6S/L- and AamIGFBP-rP1-dsRNA-injected, respectively. A one-way analysis of variance to examine differences among EMs revealed that there were significant differences among the four groups (F3, 38=63.72, P<0.0001). A post-hoc pair-wise comparison of mean EMs summarized in Fig. 6 with Tukey's HSD (α=0.05, Q=2.7) revealed that both the AamIGFBP-rP6S and AamIGFBP-rP1-dsRNA-injected group (512±34) and the AamIGFBP-rP1-dsRNA-injected (417±39) were significantly smaller than non-injected (883±12) and GFP-dsRNA-injected (830±27) control groups (P<0.01). Although apparently smaller, the mean EM of AamIGFBP-rP1-dsRNA-injected ticks was not statistically different from ticks injected with the dual mixture of AamIGFBP-rP6S/L and AamIGFBP-rP1 dsRNA. When incubated at 25°C for egg laying, all AamIGFBP-rp1-dsRNA-injected ticks failed to lay eggs, whereas all of those injected with AamIGFBP-rP6S and AamIGFBP-rP6L dsRNA laid eggs (results not shown).
This study describes, for the first time, cloning, bioinformatics analyses and biological characterization of three A. americanum IGFBP-like proteins, here provisionally identified as AamIGFBP-rP6L, AamIGFBP-rP6S and AamIGFBP-rP1. Based on data in this study, we are unable to conclusively determine whether or not the three molecules in this study are functionally similar to vertebrate proteins. However, the adoption of secondary structures (Sitar et al., 2006) and conservation of key amino acid residues that characterize the IGFBP protein superfamily (Hwa et al., 1999) as revealed by our comparative modeling and sequence alignment analyses data indicates that AamIGFBP-rP6S, AamIGFBP-rP6L and AamIGFBP-rP1 proteins, are likely to regulate biological functions that are similar to or related to their mammalian counterparts. Temporal gene expression patterns in total RNA of ticks at different feeding phases have been used as valuable tool to gain insight on probable role(s) or establish a biological relationship to tick feeding regulation (Mulenga et al., 2001; Mulenga et al., 2008; Leboule et al., 2002). On the basis of our temporal mRNA expression analysis data we speculate that AamGFBP-rP6S and AamIGFBP-rP6L, which are highly expressed during the first 24 h of feeding, but are downregulated as ticks continue to feed (Mulenga et al., 2007), were associated with regulating early stage tick feeding events such as formation of the feeding site, whereas AamIGFBP-rP1, which is upregulated in response to feeding may be associated with regulating continuous events such as blood meal feeding and/or maintenance of the tick-feeding lesion.
From the perspective of our long-term interest to find tick proteins that regulate feeding as targets for immunizing animals against tick feeding and prevent pathogen transmission, an important goal of this study was to investigate the effect of RNAi-mediated silencing of the three molecules on A. americanum feeding and reproduction. The substantially smaller blood meals obtained by ticks as a result of either dual silencing of AamIGFBP-rP6L and AamIGFBP-rP6S or AamIGFBP-rP1 alone suggest that these three proteins may be associated with physiological functions that are related to the tick feeding pathway. Of particular interest to us was the observation of 47% mortality and a failure to lay eggs in AamIGFBP-rP1-silenced ticks, all of which will lead to reduced tick populations and lower frequency of tick-borne disease transmission, the ultimate goal of any tick control program.
The observation that EMs of ticks that were injected with AamIGFBP-rP1 dsRNA was not statistically significantly different from those injected with the dual mixture of AamIGFBP-rP6L and AamIGFBP-rP6S dsRNA, rules out the possibility that the failure to lay eggs in the former was caused by ticks taking in an insufficient amount of blood. Indirect evidence in rainbow trout in which IGFBP-rP1 is upregulated during vitellogenesis and oocyte maturation (Kamangar et al., 2006) suggest the possibility that AamIGFBP-rP1 silencing may have disrupted the vittelogenesis and/or oocyte maturation process of ticks leading to failure to lay eggs. It is important to point out here that, based on our RT-PCR data, AamIGFBP-rP1 is weakly expressed in the ovary compared with salivary glands and midguts. This may be explained by the fact that we tested AamIGFBP-rP1 mRNA expression in RNA of ticks that were fed for up to 120 h. This period is well before the rapid tick feeding phase when vittelogenesis and oocyte maturation starts (Friesen and Kaufman, 2009; Coons et al., 1989). It is possible that similar to rainbow trout (Kamangar et al., 2006), AamIGFBP-rP1 mRNA is upregulated from the rapid tick feeding phase, which was not investigated in this study.
The EM as used in this study is not an accurate measure of the amount of blood imbibed by the tick (Kaufman, 2007). The tick feeding process consists of alternating episodes of blood sucking and the tick secreting excess fluids through its saliva back into the host, which results in an approx. two- to threefold concentration of the final blood meal (Kaufman and Philips, 1973; Koch et al., 1974; Koch and Sauer, 1984). Thus it has been estimated that the total volume of whole blood taken in by the tick is somewhere between 200 and 300 times the unfed female body mass although the EM is only about 100 times the unfed mass (Kaufman, 2007). It is not known at present whether or not silencing of the three molecules affected the tick's ability to feed per se, or if there is some other physiological process that affects feeding indirectly by interfering with the uptake of blood in or the process to concentrate the blood meal. Being among the recently described members of the IGFBP family, very little is known about biological functions of IGFBP-rP6 (Sarrazin et al., 2006) and thus we unable to predict the probable role (s) of AamIGFBP-rP6S and AamIGFBP-rP6L in regulating tick feeding. However, there is potential for AamIGBP-rP1 to be indirectly involved in regulating blood meal feeding by mediating the anti-coagulant and vasodilation functions of tick saliva. Vertebrate IGFBP-P1 is involved in several biological functions, including stimulating the secretion of prostacylin (Hwa et al., 1999; Burger et al., 2005). Prostacyclin (PGI2) is a potent inhibitor of platelet aggregation and a vasodilator (Rosenblum and El-Sabban, 1979) that has been found in huge amounts in the saliva of I. scapularis (Ribeiro et al., 1988). Similarly, Aljamali et al. (Aljamali et al., 2002) found 6-keto-PGF(1alpha), a breakdown product of PGI2 in salivary glands of A. americanum PGI2 is thought to facilitate tick feeding by blocking platelet aggregation to prevent blood coagulation and causing vasodilation to increase blood flow to the tick-feeding site (Ribeiro et al., 1988). The expected response to the tick feeding style of lacerating host tissue to create its feeding site and then sucking up blood from the hematoma that forms in the wounded area should stimulate vasoconstriction to slow the blood flow to wounded area and blood coagulation to stop further blood loss. Ticks successfully feed by secreting numerous molecules, such as PGI2, that keep host blood in a fluid state at the tick-feeding site and in the gut. Studies have shown that platelet aggregation function, and not blood coagulation in general, was critical to controlling bleeding of small blood vessels such as those injured by ticks during tick feeding (Mustard and Packham, 1977; Packham, 1994). Thus it is not surprising that tick saliva contains a high titer of PGI2 [535 ng ml−1 in I. scapularis (Ribeiro et al., 1988) and 124–134ngml−1 in A. variegatum (Martinez et al., 1993)]. PGI2 can disaggregate clumped platelets at concentrations of as low as 1ngml−1 (Randomski et al., 1987). Bowman et al. (Bowman et al., 1995) have proposed a possible pathway of PG synthesis in ticks from blood meal-derived arachidonic acid. However, specific mechanisms regulating PGI2 secretion in ticks are not known. It will be exciting to investigate whether or not, similar to vertebrates, AamIGFBP-rP1 is involved the secretion pathway of PGI2 in ticks. The possibility of AamIGFBP-rP1 being involved in the secretion pathway of PGI2 raises exciting prospects to indirectly block the role of PGI2 in tick feeding by immunizing animals against AamIGFBP-rP1 function. The expectation is that during tick feeding, antibodies to recombinant AamIGFBP-rP1 will bind and interfere with biological functions of native AamIGFBP-rP1 in the tick to block PGI2 secretion.
Although massive tick infestations can affect the productivity of animals, ticks cause more damage because of their role as vectors of tick-borne diseases (Jonejan and Uilenberg, 2004). Except for certain tick borne viruses such as Powassan (POW) virus that are transmitted into the host within minutes of the tick attaching onto host skin (Ebel and Kramer, 2004), there is clear evidence that numerous tick-borne pathogens such as agents of Rocky Mountain spotted fever (Rickettsia rickettsii), Lyme disease (Borrelia burgdorferi), human babesiosis (Babesia microti) and human granulocytic ehrlichiosis (Anaplasma phagocytophilum) require ticks to feed for at least 48 h before these agents are reactivated, start to replicate and are able to be transmitted to the vertebrate host (Katavolos et al., 1998; Spencer and Parker, 1923; Piesman and Spielman, 1980; Piesman et al., 1987).
From the perspective of ticks being important because of their role as vectors, an ideal target anti-tick vaccine antigen should be one that if disrupted blocks ticks from starting to feed and implicitly prevent tick-borne pathogen transmission. In this light one may argue against the potential of the three proteins in this study being suitable targets for anti-tick vaccine antigens as their silencing did not stop ticks from starting to feed. It is also noteworthy that the 47% tick mortality observed in AamIGFBP-rP1 silencing occurred after 5days of feeding, by which time most pathogens will have been transmitted. We are, however, interpreting these observations with caution, for two reasons. The first is that tick-feeding regulation is likely to be a multi-protein-regulated activity and preventing ticks from starting to feed may require knocking out multiple, yet unknown, tick proteins alongside the three described in this study. Secondly and most importantly, the RNAi methodology as used in tick research has limitations, as it does not block functions of proteins that are already expressed prior to silencing the target mRNA. The implication of this is that for genes that are expressed in unfed ticks such as the three molecules in this study, silencing of their mRNA will not block the functions of the pre-existing protein at the start of feeding and thus the effects of silencing will be delayed until the pre-existing protein become degraded. This means that, if any of the three molecules in this study played a role in regulating early-stage tick feeding events such as tick attachment onto host skin or formation of the tick feeding lesion, the RNAi-mediated silencing of the mRNA will not prevent ticks from starting to feed. This may explain why the effects of silencing the three molecules in this study were only observable from the fifth day of feeding. Given the diversity of tick species that infest animals in nature, tick vaccines targeting highly conserved antigens such as the three molecules described in this study are desirable. Studies are underway to evaluate the utility of the three proteins in this study as anti-tick vaccine antigens.
We thank Dr Pete Teel in our department for providing ticks and space in his lab to conduct the RNAi-mediated gene silencing experiments. We also would like to thank Otto Strey for his help with tick feeding. This project was funded by start up funds from the Department of Entomology, Texas A&M AgriLife Research and the office of the vice president for research at Texas A&M University.