Manipulation of pantothenate kinase in Anopheles stephensi suppresses pantothenate levels with minimal impacts on mosquito fitness

Abstract

Pantothenate (Pan) is an essential nutrient required by both the mosquito vector and malaria parasite. We previously demonstrated that increasing pantothenate kinase (PanK) activity and co-enzyme A (CoA) biosynthesis led to significantly decreased parasite infection prevalence and intensity in the malaria mosquito Anopheles stephensi. In this study, we demonstrate that Pan stores in A. stephensi are a limited resource and that manipulation of PanK levels or activity, via small molecule modulators of PanK or transgenic mosquitoes, leads to the conversion of Pan to CoA and an overall reduction in Pan levels with minimal to no effects on mosquito fitness. Transgenic A. stephensi lines with repressed insulin signaling due to PTEN overexpression or repressed c-Jun N-terminal kinase (JNK) signaling due to MAPK phosphatase 4 (MKP4) overexpression exhibited enhanced PanK levels and significant reductions in Pan relative to non-transgenic controls, with the PTEN line also exhibiting significantly increased CoA levels. Provisioning of the PTEN line with the small molecule PanK modulator PZ-2891 increased CoA levels while provisioning Compound 7 decreased CoA levels, affirming chemical manipulation of mosquito PanK. We assessed effects of these small molecules on A. stephensi lifespan, reproduction and metabolism under optimized laboratory conditions. PZ-2891 and Compound 7 had no impact on A. stephensi survival when delivered via bloodmeal throughout mosquito lifespan. Further, PZ-2891 provisioning had no impact on egg production over the first two reproductive cycles. Finally, PanK manipulation with small molecules was associated with minimal impacts on nutritional stores in A. stephensi mosquitoes under optimized rearing conditions. Together with our previous data demonstrating that PanK activation was associated with significantly increased A. stephensi resistance to Plasmodium falciparum infection, the studies herein demonstrate a lack of fitness costs of mosquito Pan depletion as a basis for a feasible, novel strategy to control parasite infection of anopheline mosquitoes.

Graphical Abstract

1. Introduction:

Malaria kills over 600,000 people annually (World Health Organization, 2021) and while current control strategies such as insecticide treated nets and artemisinin combination therapies have kept this disease in check, there is a growing concern that these control strategies will prove ineffective in the future. As malaria parasites continue to develop resistance against front line drug treatments (Plowe, 2022), mosquitoes evolve resistance against available insecticides (Hancock et al., 2020; Susanna and Pratiwi, 2021), and new zoonotic malaria parasite species increase pressure on global control campaigns (Voinson et al., 2022), the importance of developing novel strategies for mitigating malaria parasite transmission continues to increase. One such novel approach is to starve malaria parasites developing in either the mosquito vector or vertebrate host of essential nutritional resources. Pantothenate (Pan), also known as vitamin B5, is one such essential nutrient for the malaria parasites. Plasmodium spp. cannot synthesize Pan de novo (Divo et al., 1985), and must acquire exogenous Pan from stores in either the mosquito or vertebrate host in order to survive. Malaria parasites, like mosquitoes and humans, convert Pan to coenzyme A (CoA), a critical co-factor for metabolism, via the pantothenate kinase (PanK) pathway. In turn, CoA is used to synthesize acetyl-CoA (AC), which is a central metabolic intermediate, second messenger and regulator of a variety of cellular processes (Pietrocola et al., 2015). Notably, Hart et al. confirmed that mosquito-stage parasites cannot utilize CoA directly from the mosquito host and must instead synthesize it through their own PanK pathway after acquiring the Pan precursor from the mosquito (Hart et al., 2016). Thus, depletion of Pan stores in the mosquito, through increased synthesis of CoA, could conceivably inhibit parasite development, while minimally impacting mosquito fitness.

In light of studies demonstrating the critical requirement of Pan by malaria parasites, Schalkwijk et al. (2019) developed parasite-targeted Pan analogs capable of directly inhibiting the parasite PanK pathway or producing non-functional analogs of parasite CoA or AC. A novel pantothenamide MMV689258 effectively suppressed Plasmodium falciparum parasitemia in a humanized mouse model (Schalkwijk et al., 2019). The effectiveness of disrupting CoA biosynthesis in P. falciparum parasites using Pan analogs suggests that a related approach may also be effective at controlling the parasite during the mosquito stages. Our earlier studies suggested that shifting mosquito Pan stores to CoA might reduce parasite infection in A. stephensi (Souvannaseng et al., 2018). We subsequently demonstrated that manipulation of A. stephensi PanK, the rate limiting enzyme in the CoA biosynthesis pathway, through the use of the pantazines PZ-2891 and Compound 7 significantly decreased and increased, respectively, the prevalence and intensity of P. falciparum oocysts in the mosquito midgut (Simão-Gurge et al., 2021). Specifically, we determined that provisioning A. stephensi with PZ-2891 (at 1.3 or 13 nM) led to increased CoA levels, whereas Compound 7 (at 62 or 620 nM) suppressed CoA levels (Simão-Gurge et al., 2021). We proposed that the observed reduction in parasite load was due to increased A. stephensi PanK activity, which in turn reduced Pan availability to the parasite and led to parasite starvation. Further, we confirmed that knockdown of A. stephensi PanK via RNA interference led to significantly reduced levels of CoA (Simão-Gurge et al., 2021). We did not, however, directly demonstrate that Pan levels were reduced in A. stephensi with chemical or genetic manipulation of PanK.

Studies in both mice and Caenorhabditis elegans have shown a link between the PanK pathway and insulin/insulin growth factor 1 signaling (IIS) (Jackowski and Leonardi, 2014; Lee et al., 2003; Leonardi et al., 2014; Samuelson et al., 2007). Our previous work in A. stephensi demonstrated that both IIS and MAP kinase (MAPK) signaling were involved in regulating PanK expression and CoA biosynthesis (Souvannaseng et al., 2018). In the work herein, we assessed the impact of these two signaling pathways on PanK activity and CoA biosynthesis utilizing two previously generated transgenic (TG) mosquito lines from our lab. The first line, CP-AstePTEN-HA, expresses the IIS inhibitor phosphatase and tensin homologue (PTEN), an inhibitor of the IIS cascade (Hauck et al., 2013) that catalyzes dephosphorylation of phosphatidylinositol (3,4,5)-trisphosphate. The second line, CP-AsteMKP4-HA, expresses MAP kinase phosphatase 4 (MKP4), an inhibitor of the c-Jun N-terminal kinase (JNK) signaling pathway (Souvannaseng et al., 2018). The transgenes in both lines are driven by the midgut specific carboxypeptidase promoter. Both TG lines exhibited significantly reduced P. falciparum oocyst loads (Hauck et al., 2013; Souvannaseng et al., 2018). Intriguingly, metabolic assays in the CP-AsteMKP4-HA line demonstrated that the most enriched metabolic processes were associated with Pan and the CoA biosynthesis pathway (Souvannaseng et al., 2018). We also reported that PanK expression was significantly increased in the MKP4 line, as was PTEN expression (Souvannaseng et al., 2018). Not surprisingly, manipulation of IIS and JNK signaling, major regulatory pathways of many physiological processes, also extended mosquito lifespan (Hauck et al., 2013; Souvannaseng et al., 2018). Few studies, however, have examined the direct effects of manipulating Pan and CoA levels on fitness, but studies in Drosophila melanogaster revealed that manipulations to increase CoA levels in the fly can reduce the toxic effects of caloric overload and enhance the formation of triacylglycerols (TGs) (Musselman et al., 2016). These studies suggested that CoA might be limiting in the face of caloric excess (Musselman et al., 2016), biology that is perhaps consistent with the massive caloric intake during blood feeding in mosquitoes. Taken together, these data suggested the hypothesis that we have tested herein: direct manipulation of PanK-dependent CoA biosynthesis and the availability of Pan to control parasite infection in the mosquito host can be accomplished with few to no negative impacts on mosquito fitness.

To address this hypothesis, we quantified protein levels of PanK in both wild type A. stephensi and the two TG lines to establish a baseline for PanK-directed manipulation of Pan levels in these groups using small molecule modulators of PanK. Finally, we examined what impact these manipulations to the CoA biosynthesis pathway have on mosquito fitness to better assess the feasibility of using this novel approach to control malaria transmission.

2. Material and Methods:

2.1. Mosquito culture

Anopheles stephensi (Indian strain) wild type (WT) and transgenic CP-AstePTEN-HA (hereafter PTEN TG) and CP-AsteMKP4-HA (hereafter MKP4 TG) lines were reared and maintained as described previously (Simão-Gurge et al., 2021). In brief, mosquitoes were reared and maintained at 27°C and 80% humidity with 16h light/8 h dark cycling. A 10% sucrose solution was provided ad libitum to adult mosquitoes via soaked cotton pads and changed every two days. Colony propagation for adult A. stephensi WT or TG females was done by feeding on bovine blood (University of Arizona Food Products & Safety Laboratory, Tucson, AZ, USA) via glass membrane feeders or directly on CD-1 mice (Envigo, St. Louis, MO, USA) sedated with ketamine (50 mg/kg) and xylazine (5 mg/kg). All animal procedures were approved by the Institutional Animal Care and Use Committee of the University of Idaho (protocol number IACUC- 2020-10, approved 30 March 2020). For all experiments, 3–7 d old adult female mosquitoes were used.

2.2. Stocks, reagents, and chemicals

Mosquito midguts and fat bodies were dissected and stored in 5X protease inhibitor (PI) solution (Complete^™ Mini Protease Inhibitor Cocktail tablets; Sigma, St. Louis, MO, USA) prepared in 1X phosphate-buffered saline (PBS). Small molecule PanK modulators PZ-2891 (MedKoo Biosciences, Morrisville, NC, USA) and Compound 7 (Calbiochem, San Diego, CA, USA) (Sharma, et al., 2015) were resuspended in DMSO at 1.3 μM and 250.8 μM concentrations respectively to generate stock solutions. Alpha-tubulin monoclonal antibody (Developmental Studies Hybridoma Bank, Iowa City, IA, USA; 1:1,000) was commercially purchased and a custom polyclonal A. stephensi PanK antibody (1:1,000) was generated (Proteintech, Rosemont, IL, USA). Goat anti-rabbit 800 CW (1:10,000, LI-COR, Lincoln, NE, USA) and goat anti-mouse 680 RD (1:10,000, LI-COR, Lincoln, NE, USA) were used as a secondary antibodies for immunoblotting. For CoA and acetyl-CoA extraction and HPLC analysis, we used 5% perchloric acid with 50 mM dithiothreitol (DTT) per mg of sample, 100 mM monosodium phosphate and 75 mM sodium acetate at pH 4.6, to adjust pH, 85% phosphoric acid, and acetonitrile gradient grade for HPLC.

2.3. Real-time quantitative PCR of pank and ilp transcripts in A. stephensi transgenic lines

Pank expression was examined in the midgut epithelium of 3–5 day old non-transgenic (NTG) and PTEN TG female A. stephensi prior to blood feeding (non-blood fed or NBF) and 2, 6, 12, 24, 36, 48, and 72 h post-blood meal (PBM). Mosquito midguts were dissected in 1X phosphate buffered saline (PBS), immediately transferred to RNAlater and stored at −80°C. Total RNA was extracted using the RNeasy mini kit (Qiagen, Germantown, MD, USA) and quantified with a Nanodrop 2000 Spectrophotometer (Thermo Scientific, Grand Island, NY, USA). Genomic DNA contamination was eliminated by treating with TURBO DNase I (Invitrogen, Waltham, MA, USA) treatment, followed by cDNA synthesis using the High-Capacity cDNA Reverse Transcription kit (Thermo Scientific, Grand Island, NY, USA.) Quantitative real-time PCR (qPCR) assays were conducted using pank specific primers as described previously (Simão-Gurge et al., 2021). Amplification of A. stephensi ribosomal protein s7 (rps7) was used as an internal control to normalize target gene expression and results are reported as relative fold change. At least three biological replicates with unique cohorts of mosquitoes were performed for each treatment.

To evaluate the effect of PanK modulators on ilp expression, 3–5 day old A. stephensi female mosquitoes were fed on blood supplemented with 1.3 and 13 nM PZ-2891, 62 nM and 620 nM of Compound 7 or blood supplemented with an equivalent volume of DMSO as a control via the Hemotek blood feeding system. Engorged females were collected at 30 min, 2 h, 6 h and 24 h PBM for midgut and head dissection. Relative transcript levels were determined for A. stephensi ilp 1–5 using previously published primers (Taylor et al., 2020). Three pools of ten mosquito midguts or heads each were placed into 500 μL of Trizol (Thermo Scientific, Grand Island, NY, USA.), followed by total RNA extraction according to manufacturer’s instructions. cDNA was synthesized using QuantiTect reverse transcriptase kit (Qiagen, Germantown, MD, USA) according to manufacturer’s instructions. Target gene data were normalized to rps7 expression and reported as relative fold change. Biological replicates of transcript expression analyses were completed with eight separate cohorts of adult female A. stephensi derived from control and small molecule treated blood fed mosquitoes. Individual amplification reactions were performed in triplicate to confirm amplification consistency.

2.4. PanK protein expression analyses in the mosquito midgut and fat body

Proteins from mosquito midgut and fat body samples were size fractionated and screened for PanK protein expression as previously described (Simão-Gurge et al., 2021). Briefly, ten midguts or fat bodies were dissected in ice-cold PBS with complete protease inhibitor cocktail (PI-PBS, Sigma, St. Louis, MO, USA) at each time point (NBF, 2, 6, 12, 24, 36, 48, and 72 h PBM). Following dissection, midguts were washed with 1X PI-PBS until all residual blood was removed. Tissues were homogenized in cell lysis buffer (1X PBS, 1% Triton X-100, 12 mM sodium deoxycholate, 2% SDS) and Laemmli sample buffer (50 mM Tris-HCl pH 6.8, 10% glycerol, 2% SDS, 0.2 mg/mL bromophenol blue, 0.1 M dithiothreitol) and the isolated proteins fractionated onto 10% SDS-PAGE polyacrylamide gels. Proteins were then transferred to nitrocellulose membranes, blocked for 1 h at room temperature in 4% non-fat milk and 1X PBS (pH 7.4) and incubated overnight at 4°C with either A. stephensi PanK polyclonal antibody (1:1000; Proteintech) or alpha-tubulin monoclonal antibody (1:1000, Developmental Studies Hybridoma Bank, Iowa City, IA, USA). Membranes were washed 3X with 1X PBST followed by a 1 h incubation with goat anti-rabbit secondary antibody 4G (1:10,000, LI-COR, Lincoln, NE, USA) or goat anti-mouse 680 RD (1:10,000, LI-COR, Lincoln, NE, USA), visualized with LI-COR imaging system, and quantified with ImageJ (NIH) as previously described (Davarinejad, 2015). All immunoblots were replicated three times with unique cohorts of mosquitoes.

2.5. Quantification of Pan in the mosquito midgut and whole body

Pan was quantified using the official assay for Pan determination in foodstuffs by AOAC International (Cunniff, 1995), which we have optimized for mosquito tissues in our lab (Barton-Wright, 1972; Clarke, 1959; Fukuwatari and Shibata, 2012). Mosquito midguts (n = 30) or whole bodies (n = 10) were dissected from WT, MKP4 or PTEN A. stephensi in 1X PBS at timepoints prior to and after blood feeding or from WT mosquitoes that had been provisioned with the PanK activator PZ-2891 in either the bloodmeal or via water soaked cotton pads. Tissues were immediately transferred into a 1.5 ml centrifuge tube with 100 μl homogenization buffer [100 mM acetate buffer (pH 6.8) papain (20 mM), α-amylase (8.5 mM) and 2% chloroform], homogenized and incubated overnight at 37°C. Following overnight incubation the samples were autoclaved and stored at 4°C until Pan quantification was performed. The bacterium Lactobacillus plantarum (ATCC 8014), which requires Pan for growth, was cultured in MRS broth at 37°C for 24 h. The culture was then centrifuged at 4000 g for 10 min and the pellet resuspended in 5 mL of sterile saline (0.9% (w/v) sodium chloride). The resuspended bacteria (100 μl) were added to 5 ml of Difco assay medium that is deficient in Pan (BD, East Rutherford, NJ). The medium was supplemented with either calcium D (+) Pan to generate a standard curve (0, 5, 10, 20, 40, 80,160, and 320 ng/mL medium) or mosquito tissue extracts (15 midgut or 0.5 whole body equivalents as a source of Pan). The culture of L. plantarum was allowed to grow at 37°C for 20 h and turbidity measured at an optical density of 550 nm. The OD550 readings were compared against the Pan standard curve included with each assay to determine the absolute quantity of Pan per sample. All samples were replicated at least three times with unique cohorts of mosquitoes.

2.6. Quantification of total coenzyme A

Colorimetric quantification of total CoA levels (bound and free) in midguts and whole bodies of WT, PTEN TG and MKP4 TG A. stephensi females was performed using the CoA Assay Kit (Sigma-Aldrich, St. Louis, MO, USA) per the manufacturer’s instructions. Ten midguts or whole bodies were dissected for each timepoint (NBF to 48 h PBM). Midguts were washed with 1X PBS to remove the blood bolus. Samples were prepared using two midguts or 0.5 whole body equivalents of homogenate and the resulting reaction was quantified using an EPOCH/2 microplate reader (BioTek Instruments, Winooski, VT, USA) at 570 nm absorbance. Absorbance values were compared against a standard curve (0.1–10 nmol) generated for each plate to determine total CoA in the tissue sample. To establish baseline CoA levels in WT and TG mosquitoes three biological replicates using distinct cohorts of mosquitoes were performed for each sample. For total CoA assays with small molecule modulators of PanK we performed six biological replicates using unique cohorts of mosquitoes.

2.7. Sample preparation, extraction and quantification of free CoA and acetyl-CoA

Thirty 5–7-day-old female WT, MKP4 TG and PTEN TG A. stephensi were used per replicate to assess the levels of free CoA and AC. Females were starved for at least 45 min followed by access to whole human blood (MKP4 TG) or an artificial blood mixture (wild type and PTEN TG) for 30 min via a Hemotek Insect Feeding System (IFS; Discovery Workshops, Accrington, UK) or glass feeders (Chemglass). The artificial blood mixture consisted of 50% human red blood cells (RBCs) and 50% saline/ATP solution (5 mmol/L NaCl, 10 mmol/L NaHCO3, 1 mmol/L ATP, pH 7.0) or 50% serum (Hauck et al. 2013). Human RBCs were purchased from Interstate Blood Bank (Memphis, TN) and whole human blood was obtained from the American Red Cross (IBC Approval #2020–014) for these studies.

Mosquitoes were collected prior to blood feeding (NBF) and at 6 and 24 h after feeding to determine concentrations of free CoA and AC using methods adapted from published protocols (Shurubor et al. 2017). At each time point three technical replicates of 30 female mosquitoes each were prepared for each of three biological replicates. Mosquitoes were placed into 1.5 ml microcentrifuge tubes and flash frozen by immersion in liquid nitrogen. Following flash freezing, 10 μL of 5% perchloric acid with 50 mM DTT per mg of sample was added to the mosquitoes, followed by homogenization and sonication for 5 min in an ice water bath. Following homogenization, samples were centrifuged for 10 min at 10,000 g and the supernatant transferred to a fresh 1.5 ml microcentrifuge tube.

The method of Shurubor et al. (2017) was adapted with minor modifications. Briefly, the mosquito extracts and standards were injected into an Agilent 1100 series HPLC system equipped with an autosampler and a diode array detector (Agilent Technology, Germany). CoA and AC separation was conducted using an Acclaim 120 RP-C18 analytical column (150×3 mm, 3 μm, 120 Å; Thermofisher, Waltham, MA, USA) connected to a Phenomenex Security Guard column (Torrance, CA, USA) with a cartridge (C18, 4×2 mm). Isocratic elution was performed with 100 mM monosodium phosphate and 75 mM sodium acetate at pH 4.6 (adjusted with 85% phosphoric acid), which was premixed with 6% (v/v) acetonitrile. The mobile phase was freshly prepared before analysis and filtered through a 0.45 μm filter. The column temperature was maintained at 30°C, the mobile phase flow rate was 0.5 mL/min and 20 μL of the samples or standards was injected into the system. Total injection time was 13 min with 1 min post-run. Under the above-mentioned conditions, CoA and AC were eluted at 4.0 and 9.4 min after injection. Quantification was performed by plotting calibration curves using standard solutions of CoA and AC.

2.8. Metabolism assays

To assess the effect of PanK small molecule modulators on mosquito metabolism, adult female TG and WT A. stephensi mosquitoes were fed on blood supplemented with either PZ-2891 (13nM, 1.3nM), Compound 7 (620nM, 62nM), or blood supplemented with an equivalent volume of diluent (DMSO). Mosquitoes were collected prior to blood feeding (NBF), 2, 6 and 24 h PBM, snap frozen and stored until the metabolic assays were performed. Two whole female mosquito bodies were processed to extract glycogen, lipids and trehalose as previously described (Hun et al., 2021; Oringanje et al., 2021; Telang and Wells, 2004; Van Handel, 1965; Zhou et al., 2004). Total lipids were determined by modified vanillin reagent assay (Van Handel, 1965) and total glycogen and trehalose by a modified anthrone-based assay (Van Handel, 1985). All assays were replicated at least three times with unique cohorts of mosquitoes.

2.9. Reproductive assays

To study the effects of altered PanK activity on egg production of A. stephensi, we examined fecundity for the first and second gonotrophic cycles. Female A. stephensi (3–5 day-old) were provided an artificial blood meal supplemented with PZ-2891 (13nM, 1.3nM), Compound 7 (620nM, 62nM), or an artificial blood meal supplemented with an equivalent volume of diluent (DMSO). Non-blood fed females were discarded. At 48 h after the blood meal, female mosquitoes were transferred to individual cages and provided with an oviposition substrate. Female mosquitoes were allowed to oviposit for 48 h (until 96 h PBM) after which the eggs were collected, washed onto filter paper and manually counted under a microscope. Reproductive assays were replicated at least three times with unique cohorts of mosquitoes.

2.10. Lifespan analyses of WT and TG A. stephensi provisioned with small molecule modulators of PanK activity

To study the effects of altered PanK activity on the lifespan of female A. stephensi, we conducted survival assays with independent cohorts of WT, PTEN TG and MKP4 TG mosquitoes (n = 100–150 females per genotype per cohort). Blood meals were supplemented with PZ-2891 (13nM, 1.3nM), Compound 7 (620nM, 62nM), or an equivalent volume of diluent (DMSO) added to the blood meal as a control. Following the initial blood feeding, engorged female mosquitoes were retained for the survival assay, while non-blood fed individuals were discarded to ensure all mosquitoes had blood fed at the start of the lifespan assays. Blood fed mosquitoes were given small molecule- or DMSO-supplemented blood meals twice a week and 10% sucrose ad libitum. Dead mosquitoes were counted and removed daily until the final mosquito perished. Daily percent survival was determined for each cohort and an average of all biological replicates plotted. Lifespan assays were replicated 3–4 times with independent cohorts of A. stephensi.

2.11. Statistical analyses

Relative levels of A. stephensi pank transcript normalized to NTG controls and PanK protein levels normalized to alpha-tubulin were analyzed using one-way ANOVA with Tukey’s post hoc test. CoA concentrations were analyzed using ANOVA and Tukey’s post hoc test or Student’s t-test. ilp qPCR data were analyzed by ANOVA and Tukey’s post hoc test for each tissue and timepoint sample relative to the matched control. CoA and AC were analyzed using ANOVA and Tukey’s post hoc test. For CoA colorimetric assay differences between treatments and controls at various time points PBM were evaluated using one-way ANOVA with Dunnett’s multiple comparisons test. Three distinct biological cohorts of mosquitoes were evaluated for the first and second gonotrophic cycles and combined triplicate data were analyzed using one-way ANOVA with Dunnett’s multiple comparisons test. Lifespan data were analyzed using Wilcoxon test for sugar-fed and blood-fed mosquitoes, while metabolism data were analyzed using one-way ANOVA followed by Tukey’s post-hoc test relative to DMSO controls. All differences were considered significant at α = 0.05.

3. Results

3.1. Midgut pank transcript and PanK protein levels in TG A. stephensi

We previously reported increased pank transcript expression in the midgut of MKP4 TG A. stephensi and observed that pten transcript expression was also increased in this TG line (Souvannaseng et al., 2018). Accordingly, we sought to quantify midgut PanK protein levels in MKP4 TG A. stephensi and midgut pank transcript and protein levels in PTEN TG A. stephensi from 2–72 h after blood feeding. Relative to A. stephensi NTG sibling controls, PTEN TG mosquitoes had significantly increased pank transcript levels when transgene expression is maximal (36 h; Fig. 1A).

Figure 1: Pank midgut transcript and PanK protein levels in TG female A. stephensi.

A. pank transcript expression in PTEN TG A. stephensi relative to NTG sibling controls prior to blood feeding (NBF) and at various timepoints post-bloodmeal. pank expression was determined using qPCR and the 2^−ΔΔCt method for relative expression. B. Representative immunoblots showing midgut expression of specific PanK isoforms (67.6 kDa and 42.2 kDa; green bands) for MKP4 TG (left) and PTEN TG (right) A. stephensi for each time point (non-blood fed or NBF and 2, 6, 12, 24, 36, 48, and 72 h PBM) relative to alpha-tubulin as internal control (red band). Each lane represents two midgut equivalent from pools of ten mosquito midguts. Immunoblot analyses were conducted using three biological replicates with ten mosquitoes from unique cohorts of A. stephensi. Densitometry analyses of the 42.2 kDa (C) and 67.6 kDa (D) A. stephensi PanK isoforms for midguts collected from MKP4 TG and PTEN TG A. stephensi from triplicate immunoblots. Error bars represent the standard error of the mean. Differences among respective timepoints compared to NBF were evaluated using one-way ANOVA with Tukey’s post hoc test. *p < 0.05.

We previously demonstrated that our A. stephensi PanK antibody detected two midgut proteins (67.6 and 42.2 kDa), suggesting expression of two predicted pank splice variants in the midgut of wild type mosquitoes (Simão-Gurge et al., 2021). Immunoblot assays comparing midgut proteins from TG and NTG control mosquitoes did not show significant differences in the 42.2 kDa variant of MKP4 TG mosquitoes during a reproductive cycle (Figs. 1B). In contrast, we observed a significant decrease in 42.2 kDa PanK expression in PTEN TG mosquitoes at 2 h PBM relative to NBF controls (Figs. 1B), consistent with the reduced transcript expression during the first 12 h of the reproductive cycle (Fig. 1A). Although we observed an increase in 42.2 kDa PanK in MKP4 and PTEN TG mosquitoes late in the reproductive cycle (Fig. 1B) consistent with increased pank transcript expression at this time (Fig. 1A), due to replicate variability (Sup. Fig. 4) this increase was not significant. The larger 67.6 kDa PanK variant was significantly lower at 36 h PBM in TG MKP4 and PTEN mosquitoes relative to NBF controls (Figs. 1B). However, expression of the 67.6 kDa PanK variant was significantly increased in PTEN TG mosquitoes at 72 h PBM relative to NBF controls (Figs. 1B).

3.2. Effects of increased midgut PanK expression on midgut and whole body Pan and CoA levels in WT and TG A. stephensi

PanK is the rate limiting enzyme in the CoA biosynthesis pathway and is inhibited by increased CoA and AC levels, although the potency of inhibition by these cofactors varies substantially by species and tissue examined (Jackowski and Rock, 1981; Rock et al., 2003; Vallari et al., 1987). We hypothesized that increased midgut PanK levels in MKP4 and PTEN TG A. stephensi would be associated with increased conversion of Pan to CoA, resulting in a reduction in overall Pan levels and a corresponding increase in CoA levels in the mosquito. To test this assumption, we quantified Pan and total CoA in both the midgut and whole body of WT, MKP4 TG and PTEN TG mosquitoes. We observed a significant decrease in both midgut and whole body Pan in each TG line relative to wild type controls during the latter half of the reproductive cycle (36–48 h; Figs. 2A, B) when midgut PanK levels were significantly increased in both lines in response to increased MKP4 or PTEN expression (Fig. 1). Detectable Pan in human blood was low (Figs. 2A, B) and, therefore, assumed to contribute little to midgut and whole body Pan levels, particularly because the majority of the blood bolus was removed.

Figure 2: Pan and CoA levels in midgut or whole body of WT and TG female A. stephensi.

Pan levels were quantified in the midgut (A) or whole body (B) of WT, MKP4 TG and PTEN TG mosquitoes prior to blood feeding (NBF) and at 24, 36 and 48 h PBM. Pan levels in 0.5 μl of whole human blood (C) were also determined to account for any contribution of residual blood in the midgut. Similarly, CoA levels were assessed in the midguts (D) and whole bodies (E) in the same mosquito backgrounds and the same post-bloodmeal timepoints as for Pan. CoA levels were also assessed in 0.5 μl of whole human blood (F). Three separate biological cohorts of mosquitoes (15 midguts or 0.5 whole body per treatment) were analyzed. Differences between treatments and controls at various time points PBM were evaluated using One way ANOVA with Dunnett’s multiple comparisons test. *p < 0.05.

Using a colorimetric enzymatic assay that detects both free CoA and thioesters of CoA (hereafter, total CoA), we quantified midgut and whole body total CoA in WT, MKP4 TG and PTEN TG A. stephensi throughout the reproductive cycle. Total midgut CoA was significantly increased in PTEN transgenic mosquitoes from 24 to 48 h PBM relative to the NBF controls, but no significant change was observed in the MKP4 line (Fig. 2D). In whole body samples, total CoA was significantly increased in MKP4 TG mosquitoes from 24 to 36 h PBM relative to the NTG controls, while in PTEN TG mosquitoes, total CoA was significantly increased from 36 to 48 h PBM (Fig. 2E). Midgut and whole body data (Figs. 2D, E) suggest that CoA may have been mobilized from the midgut in MKP4 mosquitoes, but not in PTEN mosquitoes. Overall, data from both TG lines revealed reduced Pan levels corresponded to increased total CoA levels in whole bodies and/or midguts. As with Pan (Fig. 2C), total CoA in the human blood provisioned to the mosquitoes were minimal (Fig. 2F).

3.3. Effects of increased midgut PanK activity due to provisioning of the PanK small molecule activator PZ-2891 via the blood meal or water in WT A. stephensi

Previously we demonstrated that provisioning of the PanK small molecule activator PZ-2891 (1.3 or 13 nM) to female A. stephensi via a bloodmeal led to significantly increased midgut CoA levels at 6 and 24 h PBM (Simão-Gurge et al., 2021). In this study we assessed whether this increased CoA production led to significantly decreased Pan levels. When provisioned with 13 nM PZ-2891 in the bloodmeal, A. stephensi exhibited significantly decreased midgut Pan levels at 24 h PBM (Fig. 3A). In the whole body, we observed significantly decreased Pan levels from bloodmeal provisioning of either 1.3 nM or 13 nM PZ-2891 at both 6 h and 24 h PBM (Fig. 3B). When we provisioned PZ-2891 to the mosquitoes via water-soaked cotton pads for three days we again observed significantly decreased midgut Pan levels at a treatment concentration of 13 nM (Fig. 3C) and in the whole body following both 1.3 nM and 13 nM treatments (Fig. 3D).

Figure 3: Pan levels in midgut or whole body of WT female A. stephensi following oral provisioning of the small molecule PanK activator PZ-2891 via the bloodmeal or water soaked pads.

A, B. PZ-2891 provisioned in a bloodmeal significantly decreased Pan levels in midguts (MG; A) at 13 nM at 24h post-bloodmeal (PBM) and in whole body (WB; B) at both concentrations and timepoints. C, D. PZ-2891 provisioned via water pads for 3d significantly decreased Pan in both MG (C; 13 nM only) and WB (D). Differences in Pan levels were assessed by ANOVA followed by Tukey’s test. Different letters indicate significant differences. Three separate biological cohorts of mosquitoes were assayed.

3.4. Effect of PanK-directed small molecules on midgut CoA and AC and ilp transcript expression in TG A. stephensi

We previously reported that provisioning of the PanK-directed small molecules PZ-2891 or Compound 7 in the bloodmeal increased or decreased total midgut CoA, respectively, in WT A. stephensi (Simão-Gurge et al., 2021). Here, we examined the effects of provisioning these small molecules in the bloodmeal on midgut CoA in our two TG lines, both of which exhibited significantly increased total midgut CoA relative to NBF controls (Fig. 3A). Treatment of MKP4 TG A. stephensi with both concentrations of PZ-2891 was associated with significantly increased levels of total midgut CoA at both 2 and 6 h PBM relative to DMSO control (Fig. 4A). There was no effect of PZ-2891 on total midgut CoA levels in PTEN TG A. stephensi relative to DMSO control (Fig. 4B). Unexpectedly, we observed a modest, but significant increase relative to control at 24 h PBM in total midgut CoA following treatment of MKP4 TG A. stephensi with Compound 7 (Fig. 4C). Treatment of PTEN TG mosquitoes with Compound 7, however, revealed the expected effects of this compound. That is, PTEN TG mosquitoes treated with either 62 or 620 nM Compound 7 exhibited significantly decreased levels of total midgut CoA at 24 h PBM relative to DMSO controls (Fig. 4D).

Figure 4: Effect of PanK-directed small molecules on total midgut CoA in TG female A. stephensi.

MKP4 TG (A and C) and PTEN TG (B and D) A. stephensi were provisioned with two concentrations of PZ-2891 (A and B) or Compound 7 (C and D) and total midgut CoA was measured prior to blood feeding (NBF) and 2, 6 and 24 h PBM. Six unique biological cohorts of mosquitoes were assayed. Differences between treatments and controls at various time points PBM were evaluated using One way ANOVA with Dunnett’s multiple comparisons test. **p < 0.01 and *p < 0.05.

To detect effects of PZ-2891 and Compound 7 on whole body free CoA and AC in the same samples and to distinguish free versus total CoA, we used HPLC to assay extracts of WT and TG A. stephensi at 6 h and 24 h following small molecule provisioning. Relative to DMSO control, effects of PZ-2891 and Compound 7 on free CoA in WT and TG A. stephensi were modest, with decreased free CoA in PTEN TG mosquitoes treated with 620 nM Compound 7 and 1.3 nM PZ-2891 compared to those treated with 62 nM Compound 7 at 6 h PBM (Fig. 5A, top right panel). At 24 h PBM, AC levels were reduced in PTEN TG mosquitoes treated with 620 nM Compound 7 compared to controls and those treated with 62 nM Compound 7 (Fig. 5B, bottom right panel). AC levels were also reduced in PTEN TG mosquitoes treated with 13 nM PZ-2891 relative to control (Fig. 5B, bottom right panel). Together with analyses of total midgut CoA (Fig. 4), these data suggested that mosquito treatment with PanK-directed small molecules is associated to a greater extent with changes in levels of CoA thioesters rather than free CoA. This is perhaps not surprising given that the transfer of acetyl moieties from CoA to various biochemical targets is extremely rapid (< 0.5 sec even at 10°C), suggesting that the flux between CoA and AC is likely to be extremely variable under normal physiological conditions (Stern et al., 1982).

Figure 5: Effects of PanK-directed small molecules on free CoA and AC in whole body of WT and TG female A. stephensi.

A. Free CoA levels from wild type, MKP4 TG and PTEN TG A. stephensi treated with Compound 7 (C7) or PZ-2891 (PZ) at 6 h and 24 h post-blood meal (PBM). Free CoA levels were compared to DMSO controls (D) that were normalized to a value of 1. B. AC levels from wild type, MKP4 TG and PTEN TG A. stephensi treated with Compound 7 (C7) or PZ-2891 (PZ) at 6 h and 24 h post-feeding. Three technical replicates (n = 30 mosquitoes/technical replicate) were performed for each of three biologically unique cohorts of mosquitoes. The data were analyzed using ANOVA and Tukey’s post hoc test in which *p <0.005, **p <0.05.

As in both mice and C. elegans (Jackowski and Rock, 1981; Leonardi et al., 2014), the PanK pathway in mosquitoes is regulated in part through the IIS cascade as evidenced by increased midgut pank expression that we observed previously in PTEN TG A. stephensi (Souvannaseng et al., 2018). Thus, we were interested in whether manipulation of PanK activity by PZ-2891 and Compound 7 would impact the expression of insulin-like peptide (ilp) transcripts in A. stephensi midgut and head. However, neither PZ-2891 nor Compound 7 had any effect on ilp transcript expression relative to control in either the midgut (Supp. Fig. 1A) or head (Supp. Fig. 1B) at 0.5, 2, 6 or 24 h PBM. It should be noted that ilp transcript levels only indicate that ILP hormones are being replenished and changes in the storage or secretion of mature ILPs may still play a role in PanK manipulated mosquitoes.

3.5. Effect of PanK-directed small molecules on WT A. stephensi fecundity and lifespan

Previously we demonstrated that treatment of A. stephensi with PanK-directed small molecules significantly altered P. falciparum infection intensity (Simão-Gurge et al., 2021). Specifically, provisioning of 1.3 and 13 nM PZ-2891 significantly reduced P. falciparum oocyst load, whereas 620 nM Compound 7 significantly increased oocyst load relative to control (Simão-Gurge et al., 2021). Based on the data herein, we suggest that the previously observed reduction in oocyst intensity following PZ-2891 treatment is likely due the conversion of Pan to CoA, leading to reduced Pan levels in the mosquito and, therefore, reduced parasite development. In this context, it was important to determine whether treatment of A. stephensi with these PanK-directed small molecules, particularly PZ-2891, was associated with significant fitness costs to the mosquito host which could increase selection pressure for resistance. Thus, we assessed the impact of PZ-2891 and Compound 7 on WT A. stephensi egg production and adult survival. Provisioning PZ-2891 in the bloodmeal had no effect on the number of eggs produced per female during the first reproductive cycle (Fig. 6). In contrast, provisioning Compound 7 in the bloodmeal was associated with significant decreases in the number of eggs during the first reproductive cycle at both treatment concentrations relative to control (Fig. 6). No differences in egg production were observed during the second reproductive cycle with either PZ-2891 or Compound 7 relative to the DMSO control (Sup. Fig. 3). We also assessed survival of TG A. stephensi provided with twice weekly blood meals supplemented with the PZ-2891, Compound 7 or an equivalent volume of DMSO as a control. For these studies, we used MKP4 TG and PTEN TG A. stephensi – both with significantly increased midgut PanK, reduced midgut Pan, and increased midgut and whole body CoA relative to WT – to test whether exaggerating these effects with PZ-2891 in particular, beyond what would be observed in WT A. stephensi, might have a negative effect on lifespan. However, in replicated studies with separate cohorts of TG A. stephensi, there were no effects of either compound on TG A. stephensi lifespan (Fig. 7). Taken together, these data suggest that increasing PanK activity in A. stephensi to reduce parasite infection (i.e., through provisioning of PZ-2891) has no apparent effects on key parameters of mosquito fitness.

Figure 6: Effects of PanK-directed small molecules on fecundity of WT A. stephensi.

A. Graphs represent total egg production for first gonadotrophic cycle for WT A. stephensi provisioned with blood meals supplemented with PZ-2891 (PZ; 13 nM, 1.3 nM), Compound 7 (C7; 620 nM, 62 nM) or with and equivalent volume of DMSO (D) as control. B. Tabular summary of data from panel A. Significant differences are highlighted. Differences between treatments and DMSO controls were evaluated using ANOVA with Dunnett’s multiple comparisons test. **p < 0.01 and *p < 0.05. Three unique biological cohorts of mosquitoes were assayed and combined triplicate data is represented here.

Figure 7: Effect of PanK-directed small molecules on survival of TG female A. stephensi.

Representative Kaplan Meier survival curves are shown for each combination of TG A. stephensi (MKP4 or PTEN) and PanK-directed small molecules, each at two concentrations. Mosquitoes were provided a bloodmeal twice weekly supplemented with the appropriate PanK small molecule or with an equivalent volume of DMSO as a control. Dead mosquitoes were counted and removed daily, and assays were continued until the final mosquito perished. Survival assays were replicated three to four times with unique cohorts of TG mosquitoes. Summary data, including sample size, average age and significance are detailed in the table under the representative survival curves. Data was analyzed using the Wilcoxon test.

3.6. Effect of PanK-directed small molecules on WT and TG A. stephensi nutrient stores

Manipulation of PanK activity and, therefore, Pan, CoA and AC levels could impact fatty acid biosynthesis and oxidation as well as carbohydrate and amino acid metabolism (Hart et al., 2017; Leonardi et al., 2005; Lipmann, 1947). Therefore, we sought to determine whether these PanK small molecule modulators had any effects on nutrient stores in WT and TG A. stephensi. To do so, we quantified glycogen, trehalose and lipid levels in the whole bodies of 3–5 d old A. stephensi WT, PTEN TG and MKP4 TG mosquitoes at various time points PBM relative to NBF mosquitoes. There were no significant differences in trehalose or lipid levels in any A. stephensi line treated with small molecule modulators of PanK (Supp. Figs. 2A–C). We did observe a small but significant increase, however, in glycogen levels when WT mosquitoes were treated with 1.3 nM PZ-2891 compared to treatment with 13 nM PZ-2891; this effect was observed only at 24 h PBM (Supp. Fig. 2A). In MKP4 TG A. stephensi, we observed significant decreases in glycogen levels at 6 h PBM following treatment with either 62 nM or 620 nM Compound 7 relative to DMSO controls (Supp. Fig. 2C). There were no significant differences in any nutrient stores in the PTEN transgenic line (Supp. Fig. 2B).

4. Discussion

PanK is a ubiquitous enzyme that catalyzes the phosphorylation of Pan leading to CoA biosynthesis (Dansie et al., 2014). In turn, CoA supports a multitude of oxidative and synthetic metabolic processes such as fatty acid biosynthesis and oxidation, as well as carbohydrate and amino acid metabolism (Hart et al., 2017; Leonardi et al., 2005; Lipmann, 1947). We hypothesized that increasing PanK activity, either through overexpression of genes critical to PanK expression or through small molecule PanK modulators, would deplete Pan stores and deprive malaria parasites of this essential precursor. Previously we demonstrated that manipulation of PanK activity using small molecules could impact malaria parasite development in the highly invasive malaria mosquito A. stephensi (Simão-Gurge et al., 2021). While we observed that increased PanK activity and CoA levels decreased P. falciparum infection prevalence and intensity, we did not explore the impact of these manipulations directly on Pan stores in A. stephensi. In mice maintained on a standard diet, Pan stores are not typically a limiting resource (Brunetti et al., 2014; Kuo et al., 2007). However, it is possible to deplete endogenous Pan stores in mice by rearing them on a diet devoid of Pan or on a ketogenic diet (Brunetti et al., 2014; Kuo et al., 2007). This leads to phenotypes typically associated with pantothenate kinase-associated neurodegeneration (PKAN), including muscle contractions, poor balance, walking difficulties, and peripheral vision loss (Brunetti et al., 2014; Kuo et al., 2007). Providing Pan deficient mice with this vitamin rescues many of these symptoms, indicating that under extreme nutritional conditions mammalian Pan stores can be modified (Kuo et al., 2007). However, it is not clear if insects, particularly adult female mosquitoes, also possess robust stores of Pan or whether Pan is a limiting resource. Provisioning D. melanogaster with a high sucrose diet led to significant increases in many components of the CoA biosynthesis pathway and to obese flies with increased levels of triglycerides, but also decreased titers of CoA and Pan (Musselman et al., 2016). Providing these flies with additional Pan led to significantly increased triglyceride levels and decreased free fatty acids, demonstrating that Pan, and in turn CoA, is a limited resource in these flies (Musselman et al., 2016). These results are consistent with our finding in this work that manipulation of PanK activity can significantly reduce stores of Pan in both the midgut and whole body of female A. stephensi (Fig. 2). The fact that Pan is a limited resource in adult female mosquitoes is not surprising as their diet consists primarily of large, intermittent bloodmeals supplemented with occasional nectar feeding and neither resource is typically rich in Pan (Eissenstat et al., 1986; Kathman and Kies, 1984). This also provides a plausible explanation for our previous findings that increased PanK activity limits P. falciparum development (Simão-Gurge et al., 2021).

Here, we induced changes in A. stephensi CoA and Pan levels directly by provisioning PanK small molecule modulators to the mosquito and indirectly using TG mosquito lines that exhibit induced PanK expression. These observations suggest that there are multiple avenues to deplete Pan in the mosquito host to reduce P. falciparum infection levels, including direct activation of endogenous PanK via small molecules or increased expression of the PanK protein itself in a transgenic line. For the latter approach, we expect that direct overexpression of PanK in the mosquito midgut may provide for a more robust reduction in Pan, compared to the MKP4 and PTEN TG lines, and in turn could further increase the ability of the TG mosquitoes to resist P. falciparum infection. This is an avenue of research that we are actively exploring and could provide a complimentary control approach to the small molecule strategy.

Our data demonstrating the feasibility of reducing limited Pan stores in A. stephensi, coupled with our previous findings that enhancing PanK activity can suppress malaria parasite infection in the mosquito (Simão-Gurge et al., 2021), supports the idea that Pan depletion may be a feasible approach for malaria control. However, it is important to determine whether Pan depletion has any deleterious effects on mosquito fitness. Even modest fitness costs to the mosquito have the potential to limit the effectiveness of a proposed control strategy and raises the risk of physiological or behavioral resistance developing (Marrelli et al., 2006). We hypothesized that converting the Pan precursor to CoA would have either no or a minimal fitness effects on A. stephensi since the availability of CoA, a biologically relevant molecule for downstream metabolic processes, would be increased, not decreased in the mosquito. We did not observe any effects on adult lifespan when female A. stephensi reared under controlled, optimized conditions were treated with small molecule modulators of PanK, indicating that altered Pan levels did not affect adult survival (Fig. 7). Interestingly, both MKP4 and PTEN transgenic mosquitoes survive significantly longer than their NTG sibling controls (Hauck et al., 2013; Souvannaseng et al., 2018). Although there are several potential explanations for this lifespan extension, it is possible that enhanced CoA synthesis could be a contributing factor. This is particularly interesting considering the ability of CoA in D. melanogaster to mitigate the toxic effects of caloric overload (Musselman et al., 2016), a state exemplified by the massive intermittent bloodmeals consumed by female mosquitoes. We observed minimal effects on nutrient reserves when PanK activity was manipulated using small molecules, with significant changes only occurring with glycogen levels at restricted time points after blood feeding (Supp. Fig. 2a–c). This suggests that Pan depletion has only minimal and transient effects on mosquito metabolism. Further, we observed decreased egg production only when PanK activity was suppressed (Compound 7), suggesting that a lack of CoA limited the ability of the mosquito to provision her developing eggs (Fig. 6). When we increased PanK activity by providing bloodmeals supplemented with 1.3 or 13 nM of PZ-2891 – treatments that decreased Pan levels, increased CoA levels and decreased P. falciparum infection levels – we did not observe any differences in egg production relative to DMSO controls (Fig. 6). Furthermore, during the second reproductive cycle we did not observe any significant differences in egg production regardless of treatment (Supp. Fig. 3). These studies, conducted under controlled laboratory conditions, suggest that the impact of enhancing the CoA biosynthesis pathway, with the goal of reducing the pool of Pan available for malaria parasite growth and development, is unlikely to incur a major fitness cost in the mosquito or increase the risk of resistance developing against this control strategy. It should be noted however that the mosquitoes used in these studies were reared under optimal conditions, provided sugar ad libitum, and the adults maintained under static temperature and humidity conditions. It is possible that when PanK manipulated mosquitoes are starved or stressed, the reduced Pan levels may still impact fitness. Future studies will explore the impact suboptimal nutritional and environmental conditions have on the fitness of PanK manipulated mosquitoes.

The D. melanogaster genome encodes five PanK (also termed fumble) isoforms with various subcellular distributions and biological functions (Wu et al., 2009). In A. stephensi, we previously identified at least three putative PanK splice variants and confirmed two of them, 67.6 and 42.2 kDa variants, using PanK specific RNAi (Simão-Gurge et al., 2021). In WT A. stephensi, the 42.2 kDa isoform was modestly increased at the end of the reproductive cycle, whereas the 67.6 kDa isoform was more dramatically increased during this same period (Simão-Gurge et al., 2021). Interestingly, in our PTEN transgenic line we observed an even greater increase in PanK expression at 72 h PBM relative to WT controls (Fig. 1). This is consistent with the observed increase in pank transcript levels and dramatic increase in total midgut CoA we observed in this line. Interestingly, the 42.2 kDa PanK variant was differentially expressed between the transgenic lines, with an increase in this isoform in the MKP4 line from 36–48 h PBM, compared with a decrease in the PTEN line from 24–36 h PBM, followed by a modest increase after oviposition (i.e., 72 h PBM). The 67.6 kDa isoform is most similar to the D. melanogaster fumble-RE (long form) isoform, which is localized to the mitochondria and can rescue several phenotypes of fbl mutant flies, including survival, eclosion rates and motility (Wu et al., 2009). In contrast, fumble-RA, which most closely resembles the 42.2 kDa A. stephensi isoform, is cytosolic and was largely incapable of rescuing the fbl mutant phenotypes (Wu et al., 2009). This suggests that the larger 67.6 kDa isoform in A. stephensi is likely to be the more biologically relevant isoform in A. stephensi, although additional studies are necessary to confirm the exact role of these and other PanK isoforms in the mosquito.

In summary, we previously demonstrated that activation of PanK in the midgut of A. stephensi was capable of significantly reducing P. falciparum infection levels (Simão-Gurge et al., 2021). We hypothesized that these reduced parasite levels were due to depletion of Pan reserves in the mosquito, which the malaria parasite requires to successfully develop. Based on these finding we suggested that Pan depletion could be a novel approach to mitigating P. falciparum transmission and could be used in concert with similar approaches being explored for treating malaria infections in humans (i.e., parasite-targeted pantothenamides (Schalkwijk et al., 2019)), with the goal of developing a synergistic, two-hit approach to malaria control. In this work, we further explored the feasibility of this approach by demonstrating that Pan stores in A. stephensi are indeed a limited resource that can be experimentally reduced to inhibit malaria parasite growth. We also demonstrated that increased conversion of the Pan precursor into CoA did not have deleterious effects on mosquito lifespan or reproduction and had minimal effects on metabolism in adult female mosquitoes. Collectively, these findings expand our understanding of CoA biosynthesis in A. stephensi and lay the groundwork for developing a novel anti-malaria control strategy that can be used in concert with other control practices.

Supplementary Material

Sup. Figure 1A

Supplemental Figure 1. Expression levels of ilps 1–5 in A. stephensi treated with PZ-2891 or Compound 7. The expression levels of ilps 1–5 were not significantly different between the control group and the treated groups (13 nM or 1.3 nM PZ-2891 and 620 nM or 62 nM Compound 7) at 30 min, 2 h, 6 h and 24 h PBM and in both midgut (A) and head (B).

Sup. Figure 2A

Supplemental Figure 2. Nutrient stores in female A. stephensi treated with PZ-2891 or Compound 7. Wild type (A), PTEN TG (B) and MKP4 TG (C) A. stephensi were provisioned with PZ-2891 (1.3 nM or 13 nM) or Compound 7 (62 nM or 620 nM) in a bloodmeal. Glycogen, trehalose and lipids were extracted and quantified from mosquito pools and treatments were compared with DMSO controls. Significant differences between treatments were determined using a one-way ANOVA followed by a Tukey’s post-hoc test. *p < 0.05. All metabolic experiments were replicated with three unique cohorts of mosquitoes.

Sup. Figure 3

Supplemental Figure 3. Impact of PZ-2891 and Compound 7 on the second gonadotrophic cycle of WT A. stephensi. A. Graphs represent total egg production during the second gonadotrophic cycle for WT A. stephensi provisioned with blood meals supplemented with PZ-2891 (PZ; 13 nM, 1.3 nM), Compound 7 (C7; 620 nM, 62 nM) or with and equivalent volume of DMSO (D) as control. B. Tabular summary of data from panel A. No significant differences in egg production were observed for any treatments relative to the DMSO controls. Data was evaluated using ANOVA with Dunnett’s multiple comparisons test. ** p < 0.01 and * p < 0.05. Three unique biological cohorts of mosquitoes were assayed and combined data is represented here.

Sup. Figure 4

Supplemental Figure 4. Immunoblots of PanK expression in MKP4 and PTEN transgenic lines. The three replicate immunoblots (A-C) are presented with proteins detected using the AsPanK antibody (top), the α-tubulin loading control antibody (middle) and the merged antibodies (bottom). Replicate 2 was used as the representative immunoblot in Fig. 1 and densitometry data from all three were used for statistical analysis (Fig. 1B).

Highlights.

• Increased PanK activity reduces pantothenate stores in the A. stephensi midgut and whole body.

• Increased PanK levels in MKP4 and PTEN transgenic A. stephensi lead to increased CoA levels.

• Increased CoA biosynthesis and reduced pantothenate stores do not affect A. stephensi fitness.

Abbreviations

AC

Acetyl-CoA

C7

Compound 7

CoA

Coenzyme A

ILP

Insulin-Like Peptide

IIS

Insulin/Insulin Growth Factor 1 Signaling

JNK

c-Jun N-terminal kinase

MKP4

MAPK phosphatase 4

NBF

Non-bloodfed

NTG

Non-Transgenic

Pan

Pantothenate

PanK

Pantothenate kinase

PBM

Post-Bloodmeal

PKAN

Pantothenate kinase-associated neurodegeneration

PTEN

Phosphatase and tensin homolog

PZ

PZ-2891

TG

Transgenic

WT

Wild type