Transcriptional Mediators Kto and Skd Are Involved in the Regulation of the IMD Pathway and Anti-Plasmodium Defense in Anopheles gambiae

The malarial parasite Plasmodium must complete a complex lifecycle in its Anopheles mosquito host, the main vector for Plasmodium. The mosquito resists infection with the human malarial parasite P. falciparum by engaging the NF-κB immune signaling pathway, IMD. Here we show that the conserved transcriptional mediators Kto and Skd are involved in the regulation of the mosquito IMD pathway. RNAi-mediated depletion of Kto and Skd in the Anopheles gambiae cell line L5-3 resulted in a decrease in the transcript abundance of Cec1, which is controlled by the IMD pathway. Silencing the two genes also resulted in an increased susceptibility of the mosquito to bacterial and Plasmodium falciparum infection, but not to infection with the rodent malaria parasite P. berghei. We also showed that Kto and Skd are not transcriptional co-activators of Rel2 or other key factors of the IMD pathway; however, they participate in the regulation of the IMD pathway, which is crucial for the mosquito’s defense against P. falciparum.


Introduction
Malaria, one of the deadliest diseases in the world, is responsible for the deaths of over one million people annually. Anopheles mosquitoes are the main vectors for protozoan parasites of the genus Plasmodium, which cause the disease. About 24 h after the female mosquito ingests a blood meal from an infected mammalian host, the parasites develop into ookinetes that invade the epithelium of the mosquito midgut. The innate immune system of the mosquito is the main defense against the Plasmodium parasites [1,2,3,4]. Therefore, a better understanding of the interaction between the parasite and the mosquito's immune system could facilitate the development of novel disease control and prevention strategies. Recent studies have shown that the IMD pathway is the most important immune pathway in the mosquito's defense against the human pathogen P. falciparum [5,6]. Several anti-Plasmodium immune effectors controlled by the IMD pathway, such as TEP1, APL1, LRRD7 and FBN9, have also been characterized with regard to their anti-parasitic activity [3,7,8,9,10,11,12,13].
Previous studies of Kto and Skd in flies have focused on their function in the wing and eye disks [14,15]. Kto-and Skdmutant cells proliferate, survive, and initiate but do not complete differentiation; most notably, these cells do not respect compartment boundaries, leading to a disorganized tissue architecture [14,15]. Kto and Skd have been shown to be essential for the function of the transcription factor Atonal (Ato) in the spatial patterning of proneural clusters in the morphogenetic furrow [25]. In C. elegans, Kto is an essential gene [20,24] and is required for asymmetric cell division in the T blast cell lineage [18]. In zebrafish, mutation of Kto results in abnormal development of the brain, neural crest, and kidney [17].
Here we show that Kto and Skd regulate the A. gambiae IMD immune pathway. Silencing of Kto or Skd in a mosquito cell line resulted in a decrease in the transcript abundance of Cec1, which is known to be controlled by the IMD pathway. Kto and Skd silencing in vivo increased the mosquitoes' susceptibility to infection with Staphylococcus aureus and Escherichia coli, as well as with P. falciparum, but not P. berghei. Our study shows, for the first time, that the transcriptional mediators Kto and Skd are involved in the regulation of the IMD immune signaling pathway.

Kto Regulates the IMD Pathway
In order to determine the role of Kto in the regulation of the IMD pathway, we monitored the activation of the immune pathway after using RNAi to silence the expression of Kto in the L3-5 mosquito cell line [31], which expresses firefly luciferase under a IMD pathway-regulated Cec1 promoter. We used Renilla luciferase under the control of the Drosophia Actin5c promoter as an internal control [32]. This RNAi treatment resulted in a prominent down-regulation of the Cec1-driven luciferase gene (as measured by relative light units), indicating a down-regulation of the IMD pathway (Fig. 1A).
We then wanted to investigate which components of the IMD pathway are regulated by Kto (i.e., at which level of the IMD pathway Kto exerts its regulatory activity). Since Kto functions as a transcriptional co-activator in other organisms, we hypothesized that it may serve as a transcriptional co-activator of Rel2, the transcription factor of the IMD pathway. To test this possibility, we knocked down Kto in L3-5 cells that over-expressed the active form of Rel2; this form of Rel2 lacks the inhibitory domain and therefore can be translocated into the nucleus and directly activate gene expression [11]. Over-expression of the activated form of Rel2 strongly activated the IMD pathway, as has been reported previously ( [11]; Fig. 1B, black bars). However, silencing of Kto had no measurable effect on the Cec1 promoter activity in the Rel2 over-expressing cells (Fig. 1B), suggesting that Kto may not function as a co-activator of Rel2 but rather plays a role upstream of this NF-kB transcription factor, although it is possible that overexpression of Rel2 in the cells could have masked the effect of Kto silencing to some degree. We have previously shown that overexpression of the IMD pathway pattern recognition receptor PGRP-LC1 increases the expression of Cec1 [33,34]. Silencing of Kto in PGRP-LC1 over-expressing cells impaired this IMD pathway activation, as measured by Cec1 expression (Fig. 1A), suggesting that Kto acts downstream of PGRP-LC1.
To further investigate which part of the IMD pathway is targeted by Kto, we silenced it in IMD-over-expressing L3-5 cells; IMD acts downstream of PGRP-LC1 and upstream of Rel2. In response to Kto silencing, we saw impairment in the boosting effect of over-expressing IMD (Fig. 1C). These results suggest that Kto targets one or several components of the IMD pathway that act downstream of PGRP-LC1/IMD and upstream of Rel2.

Skd Regulates the IMD Pathway
Since previous studies conducted in Drosophila have shown that Kto acts together with Skd, and that Kto and Skd mutants have similar phenotypes in abnormally developing eyes [25], we wanted to investigate whether Skd acts together with Kto in the regulation of the IMD immune signaling pathway. For this purpose, we silenced Kto and Skd separately and together in the L3-5 cell line and measured Cec1 expression by the luciferase assay. The results indicated that single and double silencing had similar effects on the IMD pathway (Fig. 1D). Silencing Kto and Skd separately showed a similar degree of Cec1 activity repression with or without PGRP-LC1 over-expression, and simultaneous silencing of both genes did not further decrease the Cec1 activity. Our results and those previous studies in other species, indicate that Kto and Skd are likely to act together on the same target factor, or alternatively, on different targets of the IMD pathway.

Kto and Skd do not Regulate the Transcription of IMD Pathway Factors
Kto and Skd are known to be involved in gene transcription, acting as co-activators of transcription factors. However, our results did not indicate that they are transcriptional co-activators of Rel2 (Fig. 1B). It was possible that they are instead involved in the transcription of IMD pathway factors, thereby influencing the pathway's activity. In order to investigate this possibility, we studied the effect of Kto and Skd silencing at 48 h after dsRNA treatment on the transcript abundance of five major IMD pathway factors that act downstream of PGRP-LC and upstream of Rel2. At a Kto and Skd silencing efficiency of approximately 50% (Fig.  2AB) the mRNA abundance of Imd, Dredd, Fadd, Ikk-c,Tak1 and Rel2 was not altered from that of GFP dsRNA-treated control cells ( Fig. 2C-H). We also tested whether Kto and Skd silencing had any effect on the transcript abundance of the anti-Plasmodium immune effector Fbn9, which is controlled by the IMD pathway [8]. Indeed the expression of Fbn9 was down-regulated in both Kto and Skd silenced groups (Fig. 2I).

Kto and Skd Influence Mosquitoes' Resistance to Bacterial Challenges
Previous studies have shown that the IMD pathway mediates the mosquitoes' defense against infections with both Gram-positive and Gram-negative bacteria. For example, transgenic mosquitoes over-expressing Rel2 show enhanced resistance to both Grampositive and Gram-negative bacteria [11]. To investigate the potential role of Kto and Skd in the mosquitoes' resistance to challenge with the Gram-positive bacterium S. aureus and the Gram-negative bacterium E. coli DH5a, we silenced Kto and Skd prior to injection of the mosquitoes with live bacteria, and then monitored survival rates for 8 days. Silencing of either Kto or Skd impaired the mosquitoes' capacity to defend against both E. coli and S. aureus (Fig. 3, Table S1 and S2), thereby indicating a role for these factors in the anti-bacterial defense, most likely mediated through the IMD pathway.
Kto and Skd Influence Mosquitoes' Resistance to P. falciparum and P. berghei Infection by Affecting the IMD Pathway We have previously shown that the Anopheles IMD pathway mediates resistance to infection with P. falciparum but not P. berghei [5]. To investigate whether Kto and Skd influence the mosquitoes' susceptibility to infection with these parasite species, we performed gene-silencing experiments in conjunction with infection assays. Independent antibiotic-treated mosquito cohorts were injected with dsRNAs targeting each gene and were then fed on either a P. falciparum gametocyte culture or P. berghei-infected mouse 4 days later. At 8 days after feeding, the mosquito midguts were dissected, and infection intensity (as indicated by parasite oocyst number) was determined. Silencing either Kto or Skd resulted in an increased susceptibility to P. falciparum (Fig. 4B) but not P. berghei infection (Fig. 4A), when compared to the GFP dsRNA-treated controls. Antibiotic-treated mosquitoes were used because we had observed a rather high mortality of gene-silenced non-antibiotictreated mosquitoes after feeding when the midgut microbiota proliferates to high numbers [35] (data not shown). The IMD pathway is known to be implicated in the control of the midgut microbiota and its impairment may thus have led to mortality caused by bacterial infection. (data not shown). However, the gene silencing-mediated resistance to P. falciparum infection was in the same range for both antibiotic-treated and non-treated mosquito cohorts (Fig. 4C). The resistance specificity of Kto or Skd genesilenced mosquitoes to the two parasite species was therefore consistent with that seen for the Imd pathway.

Discussion
Mediator complexes represent a group of evolutionally conserved proteins involved in transcriptional activation. Studies in C. elegans, zebrafish, and Drosophila have shown that Kto and Skd are involved in regulating several specific developmental processes.
Here we have shown for the first time that Kto and Skd also play key roles in regulating the IMD pathway in the mosquito, an essential part of the innate immune system's defense against bacterial and P. falciparum infections.
Although Kto and Skd are transcriptional co-activators, our results show that they are not likely to act as such for Rel2 (Fig. 1B), the transcription factor of the IMD pathway. Nor are they involved in the transcription of major known IMD pathway factors (Fig. 2). However, our results clearly show that they regulate the IMD pathway by targeting factors that are downstream of PGRP-LC/IMD and upstream of Rel2; furthermore, our results indicate that they are likely to act together, since we saw no synergistic effect when the two genes were simultaneously silenced (Fig. 1) and studies in other species have also showed the two proteins act together in regulating gene expression. The two genes could be involved in the transcription of some other unknown IMD pathway factor(s), or function in ways other than regulation of the transcriptional machinery. Recent studies have shown that mediator complexes can also regulate alternative splicing of premature RNAs through the MED23 subunit [36]. Although being components of the mediator complex, Kto and Skd may regulate the IMD pathway through other unknown mechanisms.
The IMD pathway is of vital importance for the mosquitoes' resistance to P. falciparum infection [5,6], and a better understanding of its regulation may aid in the development of novel malaria control strategies.

Ethics Statement
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Animal Care and Use Committee of the Johns Hopkins University (Permit Number: M006H300). Commercially obtained human blood from Interstate Blood Bank Inc was used for parasite cultures and mosquito feeding, and informed consent was therefore not applicable. The Johns Hopkins School of Public Health Ethics Committee approved the protocol.

Mosquito Rearing
A. gambiae Keele strain mosquitoes were maintained on sugar solution at 27uC and 70% humidity with a 12-h light/dark cycle according to standard procedures. Antibiotic treatment of the mosquitoes was performed according to a previous protocol to obtain mosquitoes from which the LB-culturable midgut microbial flora had been eliminated [37].
About 69 nl of dsRNA (2-3 mg/ml) in water was introduced into the thorax of cold-anesthetized 2-to 4-day old female mosquitoes using a nano-injector (Nanoject, Drummond) with glass capillary needles according to established methodology. Gene silencing was verified by qRT-PCR.

qRT-PCR
At 48 h after dsRNA injection, 10 mosquitoes from each replicate were collected and homogenized in lysis buffer. For cell culture, samples were collected 48 h after adding dsRNA. RNA was extracted using the RNeasy kit (Qiagen). Reverse transcription was carried out at 42uC for 2 h using a SuperScript II kit (Invitrogen) and 20-ml reaction mixtures containing oligo(dT) primers and 2 mg of total RNA. qRT-PCR assays were performed according to a standard protocol [12] using SybrGreen PCR Master Mix (Applied Biosystems) and the ABI StepOne real-time PCR system. The relative -fold induction or repression of gene expression in the experimental samples was determined by comparing these values to their respective controls after normalizing the transcript levels with the A. gambiae ribosomal S7 gene. The primers used for qRT-PCR are listed in Table S3.

Transfection
The A. gambiae cell line L3-5 [31] was grown in S2 medium (Sigma Aldrich, St. Louis, MO) supplemented with 10% fetal bovine serum (FBS, Invitrogen, San Diego, CA). Approximately 5610 5 cells were seeded per well in 24-well plates and maintained until they reached 70%-90% confluency. Transfections were carried out with Effectene (Qiagen). For the over-expression of various genes, expression plasmids were transfected into the cells together with: (i) a plasmid carrying the firefly luciferase gene under the control of the Cec1 promoter element, and (ii) pRL-Act5C carrying the Renilla luciferase (Promega, Madison, WI) gene under the control of the Drosophila melanogaster Actin 5C promoter [32]. Cells were then harvested 24 h later for dual luciferase assays. If RNAi-mediated inhibition was used, it was carried out 24 h before transfection. Double-stranded RNA (7 mg per well) was added to the medium without FBS and used to replace the old medium of the confluent cells. Verification of gene silencing was done via qRT-PCR.

Dual Luciferase Assay
Cells were lysed in a passive lysis buffer at 24 h after the transfection and assayed with the dual luciferase system according to the manufacturer's instructions (Promega). Each experiment was repeated three times, with three independent measurements in each repeat.

Challenge with Bacteria and Plasmodium
Gram-positive (S. aureus) and Gram-negative (E. coli DH5a) bacteria were cultured in LB broth overnight, washed three times with PBS, and resuspended in PBS. At 4 days after dsRNA treatment, the anesthetized mosquitoes were injected with 69 nl of either S. aureus (CFU/mL = 2.5610 9 ) or E. coli (CFU/ mL = 4.0610 9 ) into the hemolymph, using a microcapillary Nanoject II injector (Drummond). Control dsRNA-treated mosquitoes were injected with 69 nl of sterile PBS. Dead mosquitoes were counted and removed daily over an 8-day period. The results shown here were representative of 40-50 mosquitoes for each treatment and at least three independent experiments per tested group. A Kaplan-Meier survival analysis was used to determine the significance of the differences observed.
P. falciparum and P. berghei infections were administered according to standard protocols [12]. For P. falciparum infections, mosquitoes were fed on NK54 gametocytes (provided by the Johns Hopkins Malaria Institute Core Facility) in human blood through a membrane feeder at 37uC 4 days after dsRNA treatment. Unfed mosquitoes were removed within 24 h after feeding, and the rest were maintained at 27uC for 7 days. For P. berghei infections, mosquitoes were fed on Swiss-Webster mice infected with the wildtype ANKA strain of P. berghei [38] at 21uC 4 days after dsRNA treatment. Unfed mosquitoes were removed from the group within 24 h after feeding, and the rest were maintained at 21uC for 14 days. For both infections, mosquito midguts were dissected. P.falciparum-infected midguts were stained with mercurochrome, and oocyst numbers were recorded using a light-contrast microscope (Olympus). P. berghei oocyst numbers were directly recorded under a fluorescent microscope (Leica) without staining. Each assay was done with at least 25 mosquitoes, and the data represent the results of three independent assays. P-values were determined using a Mann-Whitney test.

Supporting Information
Table S1 Survival analysis of control GFP dsRNA-injected mosquitoes compared to Kto dsRNA-or Skd dsRNA-injected mosquitoes after E. coli challenge. (DOCX)