Transcriptional Profile of Aedes aegypti Leucine-Rich Repeat Proteins in Response to Zika and Chikungunya Viruses

Aedes aegypti (L.) is the primary vector of chikungunya, dengue, yellow fever, and Zika viruses. The leucine-rich repeats (LRR)-containing domain is evolutionarily conserved in many proteins associated with innate immunity in invertebrates and vertebrates, as well as plants. We focused on the AaeLRIM1 and AaeAPL1 gene expressions in response to Zika virus (ZIKV) and chikungunya virus (CHIKV) infection using a time course study, as well as the developmental expressions in the eggs, larvae, pupae, and adults. RNA-seq analysis data provided 60 leucine-rich repeat related transcriptions in Ae. aegypti in response to Zika virus (Accession number: GSE118858, accessed on: August 22, 2018, GEO DataSets). RNA-seq analysis data showed that AaeLRIM1 (AAEL012086-RA) and AaeAPL1 (AAEL009520-RA) were significantly upregulated 2.5 and 3-fold during infection by ZIKV 7-days post infection (dpi) of an Ae. aegypti Key West strain compared to an Orlando strain. The qPCR data showed that LRR-containing proteins related genes, AaeLRIM1 and AaeAPL1, and five paralogues were expressed 100-fold lower than other nuclear genes, such as defensin, during all developmental stages examined. Together, these data provide insights into the transcription profiles of LRR proteins of Ae. aegypti during its development and in response to infection with emergent arboviruses.


Introduction
The leucine-rich repeats (LRR)-containing domain is noted to be evolutionarily conserved among many proteins correlated with innate immunity in an array of organisms, including invertebrates, vertebrates, and plants [1]. Innate immunity is a conserved host response that serves as the host's first line of defense by sensing pathogen-associated molecular patterns through germline-encoded pattern recognition receptors [2]. For example, in mammals, activation of antimicrobial defenses and stimulation of the adaptive immune response is mediated by Toll-like receptors (TLRs) and nucleotide-binding oligomerization domain (NOD)-like receptors through the LRR domain in response to pathogens [3].
Mosquitoes lack an adaptive immune system [4] and so they are fully dependent on mounting an innate immune response to fight infection [5][6][7][8]. Research indicates that mosquito vectors and mosquito cell lines produce both humoral and cellular immune responses against invading pathogens [9,10]. Leucine-rich repeat immune proteins (LRIMs) are a mosquito-specific family of putative innate receptors [11][12][13]. Two LRR proteins, LRIM1 (leucine-rich repeat immune protein 1) and APL1 (Anopheles Plasmodium-responsive leucine-rich repeat 1), have been recognized as major mosquito our understanding of the transcription profiles of Ae. aegypti LRR proteins during development and in response to arbovirus infection.

Leucine-Rich Repeat Proteins Changes in Transcriptome of the Ae. aegypti Female Adult in Response to ZIKV Infection
To better our understanding of molecular interactions and the immune response to arbovirus of Ae. aegypti from Florida, we re-examined RNA-seq data to explore the changes in leucine-rich repeat proteins in the Ae. aegypti (Key West and Orlando strains) transcriptome in response to oral ingestion of ZIKV infected blood and ZIKV infection (Accession number: GSE118858, https://www.ncbi.nlm.nih.gov/gds/?term=GSE11). Data were provided by RNA-seq analysis that generated 60 leucine-rich repeat related transcriptions in the Ae. aegypti genome in response to Zika virus (Table S1A-D and Table 1A-D). Specifically, female Ae. aegypti transcriptomic RNA-seq data showed that 23 genes related to leucine-rich repeat proteins (LRRP) were significantly upregulated during infection by ZIKA in 7-days post infection (dpi) Ae. aegypti Key West strains compared with Orlando strains. Additionally, 17 of these genes between the two strains were upregulated more than 2-fold (p-adj ≤ 0.01; log2 fold change > ±2.0) in response to ZIKV 7 dpi (Table 1A). AaeLRIM1 (AAEL012086-RA) and AaeAPL1 (AAEL009520-RA) were significantly upregulated 2.5 and 3-fold (Table 1A). When comparing transcriptome profiles of two Ae. aegypti strains in response to the control (blood-feeding only), only three genes related to leucine-rich repeat proteins were significantly upregulated/downregulated in 7-days post infection Ae. aegypti in Key West strains compared with Orlando strains (Table 1B). AaeLRIM1 (AAEL010286-RA) was significantly upregulated 3-fold (Table 1B). Comparing ZIKV infected Key West Ae. aegypti with the Key West control at 7 dpi, four differentially expressed (DE) transcripts related to LRRP were significantly dysregulated (two upregulated and two downregulated, Table 1C). AaeLRIM1 (AAEL010286-RA, p-adj 5.0 × 10 −9 , log2 fold change −3.3918) was significantly down-regulated (Table 1C). Analysis and comparison of mRNA expression profiles of Ae. aegypti Orlando strains following ZIKV infection indicated five LRRP related genes, including AaeAPL1 (AAEL009520-RA, p-adj 2.8 × 10 −4 , log2 fold change −2.2074), were significantly dysregulated (downregulated) 7-days post infection (Table 1D). Table 1. (A) Female Aedes aegypti transcriptomic RNA-seq data show leucine-rich repeat proteins related genes significantly upregulated in the Zika infection in the Key West strain compared with the Orlando strain Aedes aegypti 7-days post infection (Log2FC, Log2 fold change); (B) female Aedes aegypti transcriptomic RNA-seq data show leucine-rich repeat proteins related genes significantly upregulated/downregulated in the control (uninfected blood-feeding only) in the Key West strain compared with the Orlando strain Aedes aegypti 7-days post injection; (C) female Aedes aegypti transcriptomic RNA-seq data show leucine-rich repeat proteins related genes significantly dysregulated in the Key West strain Aedes aegypti 7-days post infection with ZIKV compared with control (uninfected blood-feeding only) in the Key West strain; (D) female Aedes aegypti transcriptomic RNA-seq data show leucine-rich repeat proteins related genes significantly dysregulated in the Orlando strain Aedes aegypti 7-days post infection with ZIKV compared with control (uninfected blood-feeding only) in the Orlando strain. 1 AAEL009520-RA is the same gene as AAEL024406. 2 AAEL009520-RA is the same gene as AAEL024406.

Leucine-Rich Repeat Proteins AaeAPL1 Paralogues of Aedes aegypti
Evolutionary analysis of eight paralogues of AeaAPL1 of the Ae. Aegypti were conducted in MEGA7 [38] ( Figure S1). The data showed that AaeAPL1 was closely related to AaeLRIM4. The DNA sequences producing significant alignments between AaeAPL1 and AaeLRIM4 showed 44% identity.

Developmental Regulation of AaeLRIM1 and AaeAPL1
To understand how AaeLRIM1 and AaeAPL1 are regulated during the development of Ae. aegypti, qPCR was performed to examine the relative transcription levels of AaeLRIM1 and AaeAPL1 in eggs, larvae, pupae, and male and female adults ( Figure 1A,B).     In addition, we also examined five paralogues of AeaAPL1, i.e., AaeLRIM3, AaeLRIM4, AaeLRIM15, AaeLRIM16, and AaeLRIM17, during developmental stages using qPCR ( Figure S2A-E).  (Table 2A). For the mosquito strain effect, Key West (permethrin resistant) had higher AaeLRIM1 than Orlando (permethrin susceptible strain) Ae. aegypti [34] (Figure 2). Standardized canonical coefficients showed that AaeLRIM1 and AaeAPL1 contributed similarly, but in opposite directions (Table 2A). In contrast, Orlando had higher AaeAPL1 than Key West Ae. aegypti ( Figure 1). Standardized canonical coefficients showed that AaeLRIM1 contributed twice as much as AaeAPL1 for the significant developmental stage effect ( Figure 2). For the developmental stage effect, LRIM1 was significantly different between developmental stages with increases associated between each stage ( Figure 1). Gene expression of AaeAPL1 was significantly higher for the larval and pupal stages compared to the egg stage. However, AaeAPL1 was lower among pupae than larvae ( Figure 1). For the significant interaction, we compared less than all possible treatment groups by which developmental stage was compared within a given strain (e.g., Key West eggs vs. Key West larvae). Standardized canonical coefficients showed a similar contribution of AaeLRIM1 and AaeAPL1 to the significant interaction, in opposite directions (Table 2A). Gene expression of AaeLRIM1 was higher for Key West than Orlando Ae. aegypti and occurred over a greater range. We observed significant increases in gene expression of AaeLRIM1 for each developmental stage for both Key West and Orlando Ae. aegypti ( Figure 2). In contrast, gene expression of AaeAPL1 was higher for Orlando than Key West Ae. aegypti and occurred over a greater range. Gene expression of AaeAPL1 significantly increased for each developmental stage for both Key West and Orlando Ae. aegypti ( Figure 2).

183
Orlando strain female Ae. aegypti. showed that AaeLRIM1 contributed much more to all significant effects than AaeAPL1 (Table 2B).

193
Because the three-way interaction was significant, we focused on pairwise comparisons of treatment MANOVA showed significant main effects of mosquito strain, sex, and age, as well as the two-way interaction and three-way interaction (strain × sex × age). Standardized canonical coefficients showed that AaeLRIM1 contributed much more to all significant effects than AaeAPL1 (Table 2B). Because the three-way interaction was significant, we focused on pairwise comparisons of treatment groups for this effect. Specifically, we compared mosquito strains of a given sex and age (e.g., Key West, female, 1-day old vs. Orlando, female, 1-day old). For expression of AaeLRIM1, Key West was higher for 1-day, 3-day, and 10-day old mosquitoes than Orlando female Ae. aegypti. In contrast, AaeLRIM1 was higher for 5-day and 7-day Orlando than Key West female Ae. aegypti. Gene expression for AaeLRIM1 was higher for all ages of Key West male Ae. aegypti except for 1-day old males. Lowest rates of expression were observed for 1-day old male Ae. aegypti for both strains and highest rates were observed for intermediate aged mosquitoes.
For gene expression of AaeAPL1, Orlando was higher for all ages than Key West female Ae. aegypti. Rates of expression were highest for 7-day old female mosquitoes and expression was lower among older females (10-day old) for both mosquito strains. For gene expression of AaeAPL1, Orlando was higher for all ages than Key West male Ae. aegypti, except 5-day old males. Rates of expression were highest for 5-day old and 3-day old male mosquitoes for Key West and Orlando Ae. aegypti, respectively. Gene expression was low for young and old male mosquitoes (Figure 1).

Infection in Ae. aegypti Exposed to CHIKV and ZIKV
In Florida, and other geographic regions of the world, there is a heterogeneity in the susceptibility of Ae. aegypti to insecticides, in part, attributable to the evolution of insecticide resistance. In particular, pyrethroid insecticides, such as permethrin, continue to be the primary means to control Ae. aegypti [39]. For this study, we deliberately included both susceptible and permethrin resistant strains of Ae. aegypti to allow for the possibility to observe differences in gene expression between the strains following ingestion of arbovirus infected blood. The susceptible strain was a long-standing laboratory colony of Ae. aegypti originating from Orlando, Florida (28.53 • N, 81.37 • W). The resistant strain was created based on a collection of Ae. aegypti from Key West, Florida (24.55 • N, 81.78 • W). We first tested Ae. aegypti from Key West for permethrin resistance following the use of World Health Organization (WHO) bottle bioassays (WHO 2016). Next, we subjected this population to 15 generations of permethrin selection and re-tested for permethrin resistance (resistant strain). Aedes aegypti females aged 4-days as adults were allowed to feed on blood consisting of either 6.4 log10 pfu/mL of ZIKV (Table 3A) or 8.0±0.09 and 8.3 ± 0.08 log10 pfu/mL of CHIKV for each strain (Table 3B). The amount of virus imbibed by mosquitoes and the viral titer at 3, 7, and 10, days post infections are described in Table 3A, B. After 10-days post infection, but not at any earlier times, we detected 100-fold higher ZIKV titer in the permethrin resistant strain than the susceptible strain (t4 = 8.12, p = 0.001) [34]. For the CHIKV infection study, we observed no significant differences in the viral titer between the susceptible and resistant strains of Ae. aegypti (all p > 0.05).

AaeLRIM1 and AaeAPL1 Transcriptional Induction of ZIKV Infections in Orally Infected Ae. aegypti Females
To characterize AaeLRIM1 and AeaAPL1 expression in response to ZIKV exposure, we measured AaeLRIM1 and AeaAPL1 expressions in orally infected Ae aegypti. MANOVA showed the significant effects of strain of Ae. aegypti, time, and their interaction (Table 4A). For the strain effect, SCCs showed that AaeLRIM1 contributed more to the significant effect than AaeAPL1 (Table 4A). Gene expression of AaeLRIM1 was significantly higher for Key West than Orlando strains. Similar contributions of gene expression of AaeAPL1 were observed for Orlando and Key West Ae. aegypti. For the significant time effect, SCCs showed that AaeLRIM1 contributed approximately 8-fold higher than AaeAPL1 (Table 4A). The highest gene expression of AaeLRIM1 was 3-hours post infection with the later time points having lower levels. For AaeAPL1, the highest levels were observed for 72, 120, and 168-hours post infection with other points being lower (Figure 2).
For the interaction, SCCs showed that AaeAPL1 contributed approximately 2-fold greater than AaeLRIM1 (Table 4A). We made less than all possible comparisons by comparing between strains and holding time constant (e.g., Key West, 3-hours vs. Orlando, 3-hours). Gene expression for AaeLRIM1 was highest for 3-hours post infection compared to other time points for both Key West and Orlando Ae. aegypti. All pairwise contrasts of treatment groups were significantly different from one another, except for 120 h post infection (Figure 2A,B).
For the significant interaction, gene expression for AaeAPL1 was highest for Key West at 168-hours post infection and for Orlando at 72-and 120-hours post infection for Ae. aegypti. All pairwise comparisons were significantly different between each other except for 24-and 48-hours post infection between the two mosquito strains (Figure 2A,B). For comparison, the amount of ZIKV imbibed by mosquitoes (3-hours) and the viral titer at 72-hours, 168-hours, and 240-hours days post infection are described in Table 3A.
To determine paralogues of AeaAPL1 expression in response to ZIKV exposure, we measured AeaLRIM3, AeaLRIM4, AeaLRIM15, AeaLRIM16, and AeaLRIM17 expressions in Ae. aegypti. MANOVA showed significant effects of the strain of Ae. aegypti, time, and their interaction (Table 4B). For the strain effect, gene expression of all five paralogues of AeaAPL1 were higher for Key West than Orlando strains, with AeaLRIM16 being the largest contributor. For the time effect, AeaLRIM17 made the largest contribution with the highest expression at 3 h post infection. Gene expression for the remaining paralogues were highest between 3-and 48-hours post infections ( Figure 3A,B).   For the interaction, SCCs showed that AeaLRIM16 contributed approximately 2-to 8-fold greater than the other paralogues. As previously described, we made comparisons between strains and held time constant (e.g., Key West, 3-hours vs. Orlando, 3-hours). Gene expression for AeaLRIM16 was significantly higher at 3-and 48-hour post infection for Key West than Orlando strains of Ae. aegypti. No other comparisons were significantly different from one another after adjusting the experimentwise alpha for multiple comparisons.

AaeLRIM1 and AaeAPL1 Transcriptional Induction of CHIKV Infections in Orally Infected Ae. aegypti Females
To characterize AaeLRIM1 and AaeAPL1 expression in response to CHIKV exposure, we measured AaeLRIM1 and AaeAPL1 expressions in orally infected Ae aegypti. MANOVA showed the significant effects of the strain of Ae. aegypti, time, and their interaction. For all significant treatment effects, SCCs showed that AaeLRIM1 contributed approximately 2-6-fold greater than AaeAPL1 (Table 5A).  Table 3B.   For the significant interaction, gene expression of AaeLRIM1 was highest for 24-hours and 240-hours post infection for Key West and Orlando Ae. aegypti, respectively ( Figure 4). All pairwise comparisons of treatment groups were significantly different from one another except for 48-hours and 168-hours post infection (Figure 4). The timing of the highest gene expression of AaeAPL1 was slightly different to observations of AaeLRIM1, with the highest levels observed for 72-hours and 240-hours post infection for Key West and Orlando Ae. aegypti, respectively ( Figure 4). All pairwise comparisons of treatment groups were significantly different from one another except for 3-hours, 120-hours, and 168-hours post infection (Figure 4). For comparison, the amount of CHIKV imbibed by mosquitoes (3-hours) and the viral titer at 72-hours, 168-hours, and 240-hours days post infection are described in Table 3B. To determine paralogues of AeaAPL1 expression in response to CHIKV exposure, we measured AeaLRIM3, AeaLRIM4, AeaLRIM15, AeaLRIM16, and AeaLRIM17 expressions in Ae. aegypti. Similar to responses to ZIKV exposure, MANOVA showed significant effects of strain of Ae. aegypti, time, and their interaction (Table 5B). In all cases, gene expression of all five paralogues of AeaAPL1 were higher for Key West than Orlando strains with AeaLRIM3 and AeaLRIM16 contributing far more than the other paralogues. For the time effect, AeaLRIM3 and AeaLRIM16 contributed far more than the other paralogues, with by far the highest expressions at 24-hours post infection. Gene expression for the remaining paralogues were similarly highest at 24-hours post infection. Gene expression for all the paralogues tended to be lowest later during the infection process (168-and 240-hour post infection).

307
For the interaction, SCCs showed that AeaLRIM3 contributed 2-to 76-fold greater than the other paralogues. As previously described, we made comparisons between strains and held time constant. Gene expression for AeaLRIM3 was significantly higher at 3-, 24-, and 72-hours post infection for Key West than Orlando strains of Ae. aegypti. In contrast, gene expression for AeaLRIM3 was significantly higher at 120-hours post infection for Orlando than Key West strains of Ae. aegypti ( Figure 5A,B). No other comparisons were significantly different from one another.

346
quinquefasciatus revealed that mosquito LRIM proteins can be classified into four distinct subfamilies 347 by a variable number of LRRs [16]. Our phylogenetic tree of paralogue showed that AaeAPL1 (or 348 AaeLRIM2) sequence-structure-function was most closely related to AaeLRIM4.

349
We analyzed developmental changes in the gene expression of LRR-containing proteins in Ae.

Discussion
Identification and characterization of genes related to LRIM1 and APL1C revealed novel innate immune factors and furthered our understanding of their presumed molecular functions. Waterhouse et al. (2010) used comparative sequenced genomes: An. gambiae, Ae. aegypti, and Cx. quinquefasciatus revealed that mosquito LRIM proteins can be classified into four distinct subfamilies by a variable number of LRRs [16]. Our phylogenetic tree of paralogue showed that AaeAPL1 (or AaeLRIM2) sequence-structure-function was most closely related to AaeLRIM4.
We analyzed developmental changes in the gene expression of LRR-containing proteins in Ae. aegypti eggs, larvae, pupae, and adults. The nucleus gene LRR-containing proteins, AaeLRIM1, AaeAPL1, and the other five paralogues are expressed <100-fold lower than the other nuclear genes, such as defensin, during all developmental stages examined [40]. Our data show that the expression of AaeLRIM1 (AAEL012086), AaeAPL1 (AAEL009520-RA), and the other five paralogues is not only regulated by development, but also by the varying environmental origin (or permethrin resistant selected strain) of mosquito strains in Ae. aegypti. For both immature and adult stages, we observed higher expression of AaeLRIM1 than AaeAPL1 in Key West Ae. aegypti and higher expression of AaeAPL1 than AaeLRIM1 in Orlando Ae. aegypti. These differences in responses may be attributable to differences in insecticide resistance among the Orlando (permethrin susceptible) and Key West (permethrin resistant) strains of Ae. aegypti. However, we are unable to rule out that other differences between these two strains of Ae. aegypti (e.g., geographic origin and founder effects) may contribute to the observed differences in gene expressions. For the immature, but not adult stages, expression of AaeAPL1 and AaeLRIM1 increased with the developmental stage. The strain effect for the analyses were modified by interactions with other factors, suggesting complex interactions between gene expression and immature stage, adult age, and sex. To our knowledge, this is the first exploration of AaeLRIM gene expression during Ae. aegypti development.
In Anopheles, two leucine-rich repeat (LRR) protein related genes, LRIM1 and APL1, have been shown to strongly affect P. berghei development in the mosquito midgut and have been identified as major mosquito factors that regulate parasite loads [11,[13][14][15][41][42][43][44]. In Ae. aegypti, the likely LRIM1 orthologue is upregulated with other immune genes following infection with Wolbachia bacteria, resulting in immune activation and shortened mosquito life spans [45]. However, there is no information available for revealing LRIM1 and APL1 affected by arboviruses. Our data demonstrates that gene expression of LRR-containing proteins, AaeLRIM1 and AaeAPL1, not only affect regulated parasites in the Anopheles, but are also altered by arbovirus infection in Ae. aegypti. Biophysical analysis of An. gambiae LRR proteins, APL1A1, APL1B, and APL1C, can arrange an extended, flexible heterodimer with LRIM1, providing a repertoire of functional innate immune complexes to protect An. gambiae from a diverse array of pathogens [14]. Future studies are needed to identify how proteins, AaeLRIM1 and AaeAPL1, may influence progression of infection of arboviruses in Ae. aegypti. Some of the highest changes in expression of AaeAPL1 and AaeLRIM1 for both strains of Ae. aegypti occurred 24-72 h post infection with CHIKV, which approximates the time when Ae. aegypti acquire disseminated infections [24]. However, this response appears to be earlier for the Key West strain (approximately 24 h), which may suggest alterations in immune responses between permethrin susceptible and resistant Ae. aegypti (Shin et al. unpublished data). Changes in the expression of AaeLRIM1 and AaeAPL1 followed a different pattern for ZIKV infected mosquitoes, suggesting gene expression changes depend on the particular arbovirus. For both strains of Ae. aegypti, AaeLRIM1 tended to be highest early during infection and declined at later points. In contrast, expression of AaeAPL1 was low early during infection and higher at later measured times.
We also described the gene expression of five paralogues of AaeAPL1 (AaeLRIM3, AaeLRIM4, AaeLRIM15, AaeLRIM16, and AaeLRIM17) in Ae. aegypti following ingestion of arbovirus infected blood. The relative contributions of the five paralogues of AaeAPL1 differed in response to infection with either ZIKV or CHIKV, which may suggest differences in mounting an immune response between viruses of different families. Gene expression of AeaLRIM16 and AeaLRIM17 made the largest contributions to significant treatment effects for mosquitoes orally exposed to ZIKV. Along the same lines, AeaLRIM3 and AeaLRIM16 made the largest contributions to significant treatment effects for mosquitoes following oral ingestion of CHIKV infected blood. However, there were some common features observed in the gene expression of the paralogues of AaeAPL1. Gene expression of all five paralogues of AaeAPL1 were usually the highest within the first 48-hours following ingestion of ZIKV or CHIKV infected blood, with lower rates of expression later during the infection process.
Studying the immune system of mosquitoes provides insights into significant opportunities to bridge tissue damage, immune invasions mechanisms, and the immune response against pathogens [4]. More importantly, unveiling a newfound understanding of mosquito immunity will shed light on the fight against disease-spreading pathogens, including ZIKV and CHIKV. Understanding the mechanisms that allow pathogens to grow and replicate in mosquitoes will provide insights into the mechanisms of mosquito-pathogen interactions. Finding exact immune evasion strategies of pathogens will help produce novel strategies that are effective at controlling them.

Mosquito Strains and Developmental Stages of Aedes aegypti
Ae. aegypti larvae were collected from Key West (24.55 • N, 81.78 • W), Florida, USA since 2011 and were initially tested for permethrin resistance, then subjected to permethrin selection for 15 generations and again assayed for resistance strain [34]. Key West strain Ae. aegypti, referred to as the resistant strain, was maintained at the Florida Medical Entomology Laboratory (FMEL) in Vero Beach, FL, USA. The Orlando population of Ae. aegypti was collected from Orlando, FL, USA and reared in the Mosquito and Fly Research Unit, Center for Medical, Agricultural and Veterinary Entomology, ARS-USDA in Gainesville, FL since 1952. The Orlando strain is recognized as a permethrin susceptible strain of Ae. aegypti [46]. For the experiments, mosquito eggs were hatched and reared in a rearing chamber at 27 • C [40]. We sampled eggs, larvae, pupae, and adults to measure developmentally regulated gene expression of Leucine-Rich Repeat Proteins in Ae. aegypti. We collected 100 µg of eggs, 20 larvae at the third instar stage, 20 pupae, 10 adult mosquitoes, 10 males and 10 females at 1-day-old (teneral), 3-day-old, 5-day-old, 7-day-old, and 10-day-old adults. The experiments were repeated three times.

Chikungunya Virus and Zika Virus Infection
Four-day-old female adults were fed blood containing either CHIKV, ZIKV, or blood without virus as the control [34,40]. Isolates of the Indian Ocean lineage of CHIKV (LR2006-OPY1, GenBank accession: KT449801) from Réunion and the Asian lineage of ZIKV (strain PRVABC59, GenBank accession # KU501215.1) from Puerto Rico were cultured in African green monkey (Vero) cells and used in the mosquito infection study. Ten mosquitos were used in each sample and three independent experiments were conducted for each time point. The detailed procedures were described in the previous publication [34,40].
Arboviruses were propagated on monolayers of African green monkey kidney (Vero) cells at a multiplicity of infection of 0.1 and environmental conditions of 37 • C and 5% CO 2 . Monolayers of cells were covered with 24 mL media (M199 medium supplemented with 10% fetal bovine serum, penicillin/streptomycin, and mycostatin). For CHIKV and ZIKV, we collected the media with virus on days three and six, respectively, and mixed it with defibrinated bovine blood containing 0.005 M ATP. Mosquitoes were allowed to feed on infected blood for one hour feeding trials using a Hemotek membrane feeding system. Samples of blood were taken during feeding trials to allow for quantification of the amount of virus presented to mosquitoes. We also allowed mosquitoes to feed on a blood and media mixture without the presence of arboviruses (controls). Mosquitoes were fed 8.0-8.3 log10 pfu/mL of CHIKV for each strain and 6.4 log10 pfu/mL of ZIKV.
Following feeding trials, fully engorged mosquitoes were sorted using light microscopy (10×) (Carson, COPYRIGHT 2006 USA, UK, HK, CHINA, Made in China) and held in cages (h by d: 10 cm by 10 cm) maintained at a 12:12 h light:dark photoperiod and 30 • C. Mosquitoes were provided with an oviposition substrate and 10% sucrose solution on cotton pads. Cohorts of mosquitoes were killed and stored at −80 • C at the following sample periods after ingesting infected blood: 3-h, 12-h (ZIKV only), 24-hours, 48-hours, 72-hours, 120-hours, 168-hours, and 240-hours. Mosquitoes were deprived of sucrose, but not water 1-day before trials used to measure transmission on 3, 5, 7, and 10 days following ingestion of ZIKV and CHIKV infection blood.

RNA-Seq Library Sequencing, Data Mining, and RNA-seq Analysis
RNA sequencing and RNA-seq analysis were featured in a previous publication [34]. Gene expression was assessed by counting the number of mapped reads for each transcript [52]. Significantly up-and downregulated genes were determined using the adjusted p-value (p-adj), log2 fold-change (log2FC), or both for downstream analysis. The RNA-seq data have been deposited to NCBI (https: //www.ncbi.nlm.nih.gov/gds/?term=GSE118858). RNA-seq analysis data provided 60 leucine-rich repeat related transcriptions in the Ae. aegypti in response to Zika virus (Table S1A-D).

Statistical Analysis
For this study, we were interested in measuring the magnitude of gene expression of components of the leucine-rich repeats (LRR)-containing domain as well as the relationship between gene expressions. So, we used a multivariate analysis of variance approach (MANOVA), which enabled us to simultaneously measure various treatment effects (e.g., developmental stage, infection with ZIKV and CHIKV) on the expression of multiple genes. Further, the MANOVA is associated with standardized canonical coefficients, which describe the relative contribution of each of the variables (genes) to a treatment effect and their direction to one another (positive or negative) [53].
Multivariate analysis of variance (MANOVA) and ANOVA were used to measure developmentally regulated gene expression of AaeLRIM1 and AaeAPL1. The relative contribution and relationship of AaeLRIM1 and AaeAPL1 to developmental treatment effects were assessed using standardized canonical coefficients (SCC) (PROC GLM, SAS 9.22). When significant effects were detected, we used univariate comparisons among treatment least-squares means for the developmental stages (Tukey-Kramer method). Separate analyses were performed for the immature stages and adult stages. Similarly, we used MANOVA to measure the expression of AaeLRIM1 or AaeAPL1 following ingestion of CHIKV and ZIKV infected blood. We tested for all main treatment factors and interactions.