Synergistic role of CSPBP and SGS1 in sporozoite entry into Aedes aegypti salivary glands

Abstract

Vector-borne diseases, particularly malaria, continue to pose a major global health challenge, underscoring the urgent need for innovative and sustainable vector control strategies. Among emerging solutions, genetic control holds great promise for interrupting disease transmission at its source. Malaria transmission relies on the parasite’s ability to overcome multiple barriers within the mosquito and ultimately invade the salivary glands during the sporozoite stage. Using Plasmodium gallinaceum as a model parasite, this study investigates two Aedes aegypti proteins, the salivary gland surface protein 1 (SGS1) and the putative ortholog of circumsporozoite protein-binding protein (AaCSPBP), that are involved in this critical step. We identified AaCSPBP through in silico homology and structural analyses and achieved gene knockdown via RNA interference. Knockdown of AaCSPBP alone resulted in a ~ 62% reduction in salivary gland sporozoite numbers and increased accumulation in the hemolymph, indicating impaired salivary gland invasion. Dual knockdown of AaCSPBP and SGS1 resulted in a synergistic effect, reducing salivary gland sporozoites by ~ 94%, a reduction significantly greater than the sum of the individual effects. These findings suggest that AaCSPBP and SGS1 act cooperatively to mediate sporozoite invasion of mosquito salivary glands. Targeting this interaction offers a promising genetically based approach to disrupting malaria transmission.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-46444-7.

Introduction

Accounting for 17% of all infectious diseases and more than 700,000 deaths annually, vector-borne diseases are a significant detriment to human health¹. Among vectors, mosquitoes are the most significant, transmitting pathogens responsible for diseases such as malaria, dengue, yellow fever, and Zika¹. Malaria remains the deadliest, with an estimated 597,000 deaths in 2023, and recent trends indicate stagnation or slight increases in incidence despite prior progress². Malaria prevention will become increasingly difficult due to the effects of climate change³, the challenges associated with current malaria vaccines⁴, the growing mosquito resistance to insecticides⁵, and anti-malarial drug resistance in parasites⁶. Given these challenges, an investigation into alternative vector control methods is imperative.

While current approaches have reduced malaria mortality, they remain insufficient for elimination. Existing vaccines provide only partial and short-lived protection, particularly in endemic regions⁴. Widespread use of insecticides and antimalarial drugs has driven resistance in both mosquito vectors and Plasmodium parasites^5,6. These limitations highlight the need for alternative strategies. Population replacement through genetic control offers a promising approach by promoting the spread of genes that inhibit parasite transmission via gene drives (> 50% inheritance)⁷. The success of such strategies depends on identifying mosquito–parasite interactions that are critical for transmission yet impose minimal fitness costs on the vector.

One of the parasites responsible for avian malaria, Plasmodium gallinaceum, shares many structural and developmental characteristics with the human strain Plasmodium falciparum^8,9 and is therefore a suitable laboratory model for understanding parasitic propagation and migration throughout the mosquito¹⁰. Following ingestion of infected blood, parasites undergo gametogenesis in the mosquito midgut, leading to fertilization and formation of motile ookinetes. These traverse the midgut epithelium and develop into oocysts, which release sporozoites into the hemolymph. Transmission requires successful invasion of the salivary glands, a key bottleneck in the parasite life cycle¹¹.

Despite its importance, the molecular basis of sporozoite invasion remains incompletely understood. Salivary gland surface protein 1 (SGS1) in Aedes aegypti is one of these proteins, occupying regions associated with sporozoite salivary gland invasion and inducing an immune response in mosquitoes after secretion into their saliva¹². Furthermore, the introduction of anti-SGS1 immunoglobulin in P. gallinaceum-infected Ae. aegypti before the burst of oocysts substantially decreased salivary sporozoite numbers¹³, and the knockdown and knockout of SGS1 have been shown to significantly reduce the number of sporozoites in infected salivary glands by 67% and 64%, respectively¹⁴. SGS1 disruption may also affect earlier stages of parasite development, as reduced oocyst formation has been observed, while no significant fitness costs have been reported¹⁴.

Possibly functioning in addition to SGS1 is the circumsporozoite protein-binding protein (CSPBP). CSPBP is the associated receptor of the circumsporozoite protein (CSP), the most abundant surface protein in the Plasmodium parasite. CSP has long been known as essential for sporozoite invasion of the salivary glands, and its knockout in Plasmodium berghei has been shown to significantly inhibit sporozoite production^15,16. Subsequently, CSPBP antibodies have been shown to reduce sporozoite abundance by 25% and 90% in P. berghei-infected Anopheles gambiae after 14 and 18 days, respectively, with CSPBP knockdown preventing 75% of sporozoites in the hemolymph from entering the salivary glands of an infected mosquito¹⁶. The effects resulting from the suppression of the orthologous protein to CSPBP in Ae. aegypti (AaCSPBP) are uncertain, as is the extent to which its role parallels that of CSPBP in An. gambiae. The orthologous nature of CSPBP and Drosophila UPF3 domain implicates a role in nonsense-mediated mRNA decay (NMD), providing a possible mechanism for the prevention of sporozoite invasion through the dysregulation of NMD and the disruption of homeostatic processes in the salivary glands¹⁷. This, together with CSPBP expression outside the salivary glands¹⁶, suggests potential fitness costs upon knockout, although evidence remains limited.

In this study, we identified and knocked down the putative Ae. aegypti ortholog of CSPBP in P. gallinaceum-infected mosquitoes to assess its role in sporozoite invasion of the salivary glands. We also evaluated the combined knockdown of AaCSPBP and SGS1 to test for potential synergistic effects. Notably, our results show that dual gene silencing produces the greatest reduction in salivary gland sporozoite abundance, suggesting a synergistic effect. These findings highlight the potential of targeting multiple mosquito-parasite interactions through genetic engineering to disrupt malaria transmission.

Methods

In silico search and analyses

Using VectorBase (https://vectorbase.org/vectorbase/app)¹⁸, the known protein sequence of An. gambiae CSPBP (AGAP006649) was used to run a BLASTP^19,20 search against the proteome of Ae. aegypti. The results were used to determine the possible orthologs of CSPBP in Ae. aegypti (AAEL003735). Results with an E-value < 1e⁻¹⁰ were considered indicative of homology.

A structural comparison between the AlphaFold-predicted three-dimensional structures^18,21,22 of the highly conserved regions of AGAP006649 and AAEL003735 was performed using ChimeraX²³. The distance matrix alignment (Dali) tool²⁴, which can be accessed via the server at http://ekhidna.biocenter.helsinki.fi/dali, was used to analyze the similarity between the secondary structures of the two proteins. The AlphaFold-predicted structures were retrieved from VectorBase and uploaded to the pairwise structure comparison tool on the Dali server, and to the desktop version of ChimeraX. A superimposed picture of the highly conserved regions of the proteins was generated in ChimeraX, and the structural comparison tool was used to highlight regions of structural conservation. For the structural alignment, a Z-score of 2.4 or higher was considered significant. A low Root Mean Square Deviation (RMSD) (< 3 Å) with a low Z-score indicates high overall fold similarity, whereas a significant Z-score with a high RMSD indicates more investigation into the aligned structures is necessary²⁵.

Mosquitoes rearing

The Ae. aegypti Higgs white-eyed strain²⁶ was used in all experiments. Rearing and infection procedures were performed according to Kojin et al.¹⁴. Briefly, the Ae. aegypti Higgs white-eyed strain was maintained in the insectary at the Institute of Biomedical Sciences II, University of São Paulo, Brazil, at a temperature of 27 ± 2 °C and a relative humidity of 80%, along with a 12-h light:12-h dark cycle. The food provided to the adult mosquitoes was a 10% sucrose solution available ad libitum. Adult females had access to anesthetized mice for blood feeding and egg production, and larvae were fed on a solution of Tetramin and water.

Mosquito infection with Plasmodium gallinaceum

To infect chickens, an aliquot of P. gallinaceum (strain 8A, obtained from A. Krettli, René Rachou Institute of Research, FIOCRUZ, MG, Brazil) infected blood was transfused into 7-day-old Gallus gallus chicks (Granja Kunitomo, Mogi Das Cruzes, Brazil). To verify that the chickens were sufficiently infected, a drop of blood was collected from the foot of each inoculated chicken, which was then smeared with Giemsa and examined to determine parasitemia. Chickens deemed acceptable for blood feeding had parasitemia levels of 5–9%. All infected chickens were simultaneously exposed to both the control and the experimental groups. Five-to-seven-day-old female mosquitoes from both groups were deprived of the 10% sucrose solution for 16 h and then allowed to feed on the infected chickens for 15 min until satiation. Exclusively fully engorged mosquitoes were used for the following experiments, where each group had three biological replicates. For work with chickens at the Universidade de São Paulo, work was approved by Comissão de Etica no Uso de Animais (CEUA-ICB/USP) (equivalent to the Institutional Animal Care and Use Committee (IACUC) in the United States); 188/2012 extended until 03/12/2020 and done in accordance with CEUA-ICB/USP regulations.

Generation of double-stranded RNA and gene silencing assays

A 534 bp fragment from the AaCSPBP gene (AAEL003735) was amplified using a primer designed using E-RNAi Webservice (www.dkfz.de/signaling/e-rnai3/idseq.php) that incorporates T7 minimum promoter sequence at its 5′ end. The primer sequences were as follows: AaCSPBP—forward 5′TAATACGACTCACTATAGGGAGAGGAAAAAGACACGGCTGGTA3′ and reverse 5′TAATACGACTCACTATAGGGAGAGGTACCTCTCCTGAAGCTTTCTT3′. GFP was used as a control and was amplified using the following primers: T7-GFP-FWD (5′TAATACGACTCACTATAGGGAGAGAACTGTTCACCGGAGTGGT3′) and T7-GFP-REV (5′TAATACGACTCACTATAGGGAGATCACCAGGGTATCTCCTTCG3′). The PCR products were purified using the QIAquick PCR Purification Kit (QIAGEN), and double-stranded (ds) RNA was synthesized and purified using the MEGAscript T7 Transcription Kit (Ambion) according to the manufacturer’s protocol. The dsRNA for SGS1 was synthesized using the same methodology described for AaCSPBP. Both dsRNAs were generated and administered concurrently within the same experimental batch, in parallel with the experiments reported in the previously published study¹⁴. Primer design, in vitro transcription, and all knockdown procedures followed the protocol described in that study, ensuring consistency and reproducibility across both gene targets. Four-day-old adult females were allowed to feed on P. gallinaceum-infected chickens, and after 6 days, infected females were injected in the thorax with 3 µg of dsAaCSPBP or dsGFP, and 2 days after double-stranded RNA injection (8 days after P. gallinaceum blood meal), the salivary gland or hemolymph of individual females was dissected, and the number of sporozoites was determined. For the dual AaCSPBP/SGS1 genes knockdown experiment, dsAaCSPBP or dsGFP was injected on the 6th day post-infected blood meal. The injected females were allowed to rest for 24 h, followed by a second injection with dsSGS1 or dsGFP, and the number of sporozoites was determined in the salivary gland or hemolymph in the same way as in the single dsRNA injection experiment. Figure 1 illustrates a schematic representation of the experimental timeline.

Fig. 1.

Schematic representation of the experimental timeline for single and dual knockdown assays targeting AaCSPBP, SGS1, or both genes simultaneously. Aedes aegypti females were fed a Plasmodium gallinaceum-infected blood meal. For single knockdown, six days post-infection, females were injected in the thorax with either dsAaCSPBP or dsGFP (negative control). Two days after injection (i.e., 8 days post-infected blood meal), salivary glands or hemolymph from females were dissected or extracted, and the number of sporozoites was quantified. For the double knockdown experiment (dsAaCSPBP/dsSGS1), females were first injected with dsAaCSPBP or dsGFP on day 6 post-infection, allowed to rest for 24 h, and then injected a second time with dsSGS1 or dsGFP. Sporozoite numbers in the salivary gland were determined on day 8 using the same procedure as for the single knockdown experiments.

mRNA expression analyses

Reverse transcriptase PCR (RT-PCR) was performed using whole-body total RNA extracted from Ae. aegypti with TRIzol reagent (Invitrogen) and treated with DNase I (Invitrogen) to remove genomic DNA contamination. The amplification of diagnostic products was performed using the OneStep RT-PCR kit (Qiagen) and primers. The reaction mixture was incubated at 50 °C for 30 min and 95 °C for 15 min. Amplification conditions were 94 °C for 1 min followed by 30 cycles of 94 °C for 1 min, 60 °C for 1 min, and 72 °C for 1 min, and a final step of 10 min at 72 °C. The primer sequences were as follows: for AaCSPBP, the forward primer used was 5′AAGAACTTACGCCAGCTCCA3′, and the reverse was 5′GGACGATTTTTGTTGCGAAT3′. The forward actin primer used was 5′GAGCGTGGCTACTCCTTCAC3′, and the reverse was 5′AGTTTCGTGGATACCGCAAG3′. Amplification products were visualized on a 1% agarose gel stained with ethidium bromide, and bands were observed under UV light using a gel documentation system to assess the size and specificity of the PCR products.

Mosquito hemolymph extraction

To quantify sporozoites in mosquito hemolymph, one set of legs was removed from a single cold anesthetized female, and 1–5 μL of Phosphate Buffered Saline  (PBS) was intrathoracically injected from the opposite side. A small volume (0.4–1 μL) of diluted hemolymph was then collected from the leg stumps using a sterile pipette. Hemolymph from multiple mosquitoes was pooled until a total volume of 10 μL was obtained. The pooled sample was placed in a hemocytometer, and sporozoite counts were determined using phase-contrast microscopy. Hemolymph collection was performed 8 days after the infected blood meal, and sporozoite quantification was performed.

Mosquito salivary gland extraction

Adult females injected with dsAaCSPBP and/or dsSGS1, or dsGFP (negative control) were cold anesthetized, and their salivary gland pairs were individually dissected using forceps and a probe under a stereoscope equipped with bottom lighting to enhance gland visibility. Dissections were performed in a drop of 100 μL of PBS. Following dissection, each salivary gland pair was transferred to a fresh drop of PBS and gently rinsed to minimize tissue disruption while removing sporozoites adhering to the gland surface. The glands were then individually homogenized by pipetting up and down multiple times in PBS. Only intact salivary glands were used in the experiment.

Sporozoite quantification from mosquito salivary glands and hemolymph

Immediately following dissection, the recovered hemolymph or isolated salivary glands were transferred to a hemocytometer chamber, where sporozoite counts were conducted under phase-contrast microscopy. This approach enabled direct visualization and quantification of sporozoites for consistent assessment of parasite load across experimental groups.

Statistical analysis

The D’Agostino-Pearson omnibus normality test was applied to assess whether sporozoite counts followed a normal distribution. Based on the results, either an unpaired t-test or a Mann–Whitney test was used to evaluate statistical differences between control and experimental groups. The percentage reduction in sporozoite numbers was calculated using the formula:

The Wilcoxon one-sample signed-rank test was applied to assess whether the median observed reduction in sporozoite count for the dual knockdown of AaCSPBP and SGS1 differed from the theoretical reduction determined by the Bliss model of independence. All tests were performed using GraphPad Prism (version 10.5.0 for Windows, GraphPad Software, La Jolla, USA, www.graphpad.com). Statistical significance is indicated in figures as follows: p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****); ns denotes not significant.

Results

Gene transcript and protein structure of putative AaCSPBP

A Reciprocal BLASTP search was conducted against the whole genome of the Ae. aegypti Liverpool strain (AaegL3 assembly) to identify a putative ortholog. The protein alignment showed 42.77% similarity between the An. gambiae CSPBP gene sequence (AGAP006649) and the putative ortholog in Ae. aegypti (AAEL003735) over a 621 amino acid sequence, with an E-value of 5e⁻¹¹³, and a bit score of 355 (Fig. 2A). No other matches were found in the Ae. aegypti genome for the protein sequence. The retrieved gene comprises two coding exons within a 2365 bp genomic sequence, including 5′ and 3′ untranslated regions, introns, and exons. Its overall structure mirrors that of An. gambiae CSPBP, as depicted in the schematic in Fig. 2B. The corresponding transcript for Ae. aegypti CSPBP is 2,300 bp in length, with the protein comprising 685 amino acid residues. The transcript of An. gambiae CSPBP is 2,356 bp, with a protein sequence containing 661 amino acid residues (Fig. 2A). The secondary structures of Ae. aegypti and An. gambiae CSPBP predicted by AlphaFold were compared using the pairwise alignment tool on the Dali server, which returned a Z-score of 16.9, indicating that the structural similarities are very unlikely to be a random occurrence. The comparison of the secondary structures in which each amino acid residue participates is displayed in Fig. 2A. The structural comparison of the two sequences revealed a strong similarity between residues 71–159 of AgCSPBP and residues 86–173 of AaCSPBP. Further inquiry into AlphaFold-predicted²¹ domains revealed that these regions correspond to UPF3 domains, which are involved in NMD. We compared the three-dimensional structures of these residue regions to one another by generating a superimposed image using ChimeraX²³ as shown in Fig. 2C. This comparison yielded an RMSD of 0.649 Å, indicating orthology when considered in conjunction with the previously determined Z-score from the structural comparison.

Fig. 2.

Protein alignment and gene structure of the Aedes aegypti CSPBP ortholog. (A) Protein sequence alignment between Anopheles gambiae CSPBP (AGAP006649, labeled “Anoph”) and its putative Aedes aegypti ortholog (AAEL003735, labeled “Aedes”), with results from BLASTP analysis (VectorBase), with highlights to indicate different secondary structures. Gray highlights indicate low complexity regions, red highlights indicate alpha helices, light blue highlights indicate beta sheets, and the box indicates the UPF3 domain. (B) Schematic representation of the Aedes aegypti CSPBP and Anopheles gambiae gene structures. The two coding exons are represented by pink boxes for Aedes aegypti and dark green boxes for Anopheles gambiae; gray boxes represent the 5′ and 3′ untranslated regions (UTRs), and the green line denotes the intron. (C) Ribbon representation of the spatial superposition of amino acid residues in 3D space of the AaCSPBP and AgCSPBP, showing the predicted UPF3 domain (residues 71–159) of AgCSPBP (blue) with the predicted UPF3 domain (residues 86–173) of AaCSPBP (yellow) superimposed onto it.

AaCSPBP gene silencing is rapid in Aedes aegypti

To analyze the potential for silencing AaCSPBP and determine the duration of this silence, RT-PCR was performed on pools of 5 females, and transcript presence was verified in samples prepared 2, 3, 4, 5, and 6 days after thoracic injections of dsRNA. We observed a significant decrease in AaCSPBP transcript levels in samples prepared 2- and 3 days post-injection; however, the transcript levels seemingly returned to the same level as the control on the 4th day, as shown in Fig. 3. For clarity, images of the agarose gels were cropped to highlight the relevant bands, and the full, uncropped images are provided in Supplementary Fig. S1. The SGS1 knockdown was previously demonstrated in¹⁴, where a significant reduction in transcript levels was observed on day 2 post-injection. Although mRNA levels began to recover by day 3, they remained below control levels, with approximately full expression restored by day 4¹⁴. Together, these results highlighted a short period of knockdown and indicated the optimal window for performing gene knockdown injections in coordination with Plasmodium infection, as shown in Fig. 1 in the “Methods” section.

Fig. 3.

Downregulation of the AaCSPBP in Aedes aegypti by thoracic injection of double-stranded RNA. Total RNA was extracted from pools of 5 females at 2, 3, 4, 5, 6 days post-injection (dpi) of dsGFP or dsAaCSPBP, and used in an RT-PCR reaction with specific primers for the AaCSPBP sequence. The same RNA samples were also analyzed by actin-specific primers. The amplicons were visualized on a 1% agarose gel stained with ethidium bromide under UV illumination, and the displayed images are cropped images of the agarose gel.

Knockdown of AaCSPBP lowers the number of sporozoites invading the salivary gland

To evaluate whether AaCSPBP plays a role in P. gallinaceum invasion of Ae. aegypti salivary glands, similar to the function reported for its ortholog in An. gambiae during P. berghei infection, dsRNA targeting AaCSPBP and GFP (as a control) was initially synthesized at a concentration of 1500 ng/µL. These were intrathoracically injected into P. gallinaceum-infected Ae. aegypti females on the sixth day post-infection. Two days later, salivary glands were dissected, and sporozoite numbers were quantified. No significant difference was observed between the experimental and control groups using this initial concentration of dsRNA (Supplementary Fig. S2).

To enhance gene silencing efficacy, the experiment was adjusted and conducted using a doubled dsRNA concentration (3000 ng/µL of dsAaCSPBP and dsGFP), following the same injection timeline. The experimental design (Fig. 1) was guided by the developmental timeline of P. gallinaceum in Ae. aegypti²⁷ and by the rapid silencing kinetics previously observed for AaCSPBP (Fig. 3). In this optimized experiment, the number of salivary gland-associated sporozoites in dsAaCSPBP-injected mosquitoes was reduced by 68%, 71%, and 33% across three independent replicates, yielding an average reduction of 62% (Fig. 4). Notably, in the replicate with the lowest reduction (33%), the parasitemia level in the infected chicken host was the highest, suggesting that a high parasite load may partially overcome the effects of AaCSPBP knockdown.

Fig. 4.

Single knockdown of AaCSPBP reduces Plasmodium gallinaceum sporozoite numbers in the salivary glands of Aedes aegypti. Aedes aegypti mosquitoes injected with dsAaCSPBP or dsGFP infected with P. gallinaceum had their salivary glands dissected 8 days post-infection. The number of sporozoites contained in the salivary glands of dsAaCSPBP-injected mosquitoes was compared to the number in dsGFP-injected mosquitoes in three independent replicates. Median was used as the measure of average due to the data following a non-normal distribution.

To further investigate whether the reduced sporozoite invasion observed following AaCSPBP knockdown was due to a direct blocking effect rather than other factors, we quantified the number of sporozoites present in the hemolymph. These counts were performed using the same timeline as the salivary gland dissections, ensuring consistency with the window of parasite migration. Across three independent experiments, mosquitoes injected with dsAaCSPBP showed significantly higher numbers of sporozoites in the hemolymph, with 1.50-, 2.00-, and 3.83-fold increases compared to the control group (Fig. 5). These findings suggest that silencing AaCSPBP impairs the ability of sporozoites to invade the salivary glands, resulting in their accumulation in the hemolymph. The raw sporozoite counts for each replicate are presented in Table 1.

Fig. 5.

Effect of dsAaCSPBP knockdown on sporozoite abundance in the hemolymph. Plasmodium gallinaceum-infected Aedes aegypti mosquitoes injected with dsAaCSPBP or dsGFP had their hemolymph dissected 8 days post-infection. The number of sporozoites contained in the hemolymph of dsAaCSPBP-injected mosquitoes was compared to the number in dsGFP-injected mosquitoes in three independent replicates. A 2.00-, 1.50-, and 3.83-fold increase in the number of sporozoites contained in the hemolymph was observed for dsAaCSPBP-injected mosquitoes for each trial, respectively.

Table1.

Summary of sporozoite load in hemolymph and fold increase following dsAaCSPBP knockdown.

Experiment	Number of hemolymph sporozoites (n)		Number of mosquitoes in the pool (%)	Chicken parasitemia	Fold increase
	dsGFP	dsAaCSPBP
1	100	200	10	6	2.00
2	78	117	14	6	1.50
3	300	1150	13	9	3.83

Dual knockdown of AaCSPBP and SGS1 synergistically enhances the reduction of sporozoite invasion in Aedes aegypti salivary glands

We further proceeded to investigate if knocking down both AaCSPBP and SGS1 could potentially lead to an increase in the blocking effect of the sporozoite penetration in the salivary gland. This dual knockdown approach was designed to ensure that both AaCSPBP and SGS1 were effectively silenced at the time sporozoites began invading the salivary glands. The strategy also accounted for mosquito survival, as delivering the necessary concentrations of dsRNA required relatively high injection volumes. For this, we injected dsAaCSPBP on day 6 after mosquito infection with P.gallinaceum and allowed the females to recover for 24 h, and then proceeded with the injection of dsSGS1 on day 7 post-infection. Day 8, the salivary gland and hemolymph were dissected, and sporozoites were counted (Fig. 1). The SGS1 knockdown was performed concurrently with AaCSPBP and was previously published¹⁴; therefore, these data are not shown here. SGS1 knockdown is short, with effective silencing observed for only two days and transcript levels beginning to recover by the third day. This temporal profile guided our experimental design, leading us to inject dsCSPBP prior to dsSGS1, given the shorter duration of SGS1 silencing. Through the optimization of the injection time and combination of the gene targets, we aimed to achieve robust knockdown of both genes without compromising female viability, thereby enabling a reliable assessment of their synergistic role in mediating sporozoite invasion.

We observed a drastic reduction in the number of sporozoites in the salivary glands of P. gallinaceum-infected females following the dual knockdown of AaCSPBP and SGS1. Specifically, sporozoite counts were reduced by 93% in the first replicate, 92% in the second, and 92% in the third (Fig. 6). Across the three replicates, the median number of salivary gland sporozoites showed a consistent 94% reduction (Fig. 6), strongly indicating a synergistic effect of the dual knockdown on blocking sporozoite invasion. In Supplementary Fig. S3, we present the results of the single knockdown (dsCSPBP) and the double knockdown (dsCSPBP/dsSGS1) side by side to highlight the statistically significant difference between the AaCSPBP single knockdown dataset and the dsAaCSPBP/SGS1 double knockdown dataset.

Fig. 6.

Dual knockdown of AaCSPBP and SGS1 drastically reduces Plasmodium gallinaceum sporozoite invasion of Aedes aegypti salivary glands, indicating they act synergistically. Plasmodium gallinaceum-infected Aedes aegypti mosquitoes injected with dsAaCSPBP together with dsSGS1 or dsGFP (control) had their salivary glands dissected 8 days post-infection. The number of sporozoites contained in the salivary glands of dsAaCSPBP/dsSGS1-injected mosquitoes was compared to the number in dsGFP-injected mosquitoes in three independent replicates. The median was used as a measure of average because the data followed a non-normal distribution.

Discussion

Despite extensive efforts to characterize Plasmodium sporozoite invasion of mosquito salivary glands, key molecular mechanisms remain unresolved. Proteins such as CSPBP and SGS1 are implicated in this process, yet their precise roles and broader functions in mosquito biology remain poorly defined. This gap limits the rational design of genetic control strategies targeting parasite transmission. This study seeks to provide much-needed insight into sporozoite salivary gland invasion by investigating similarities among CSPBP proteins across different malaria vectors and the effects of AaCSPBP knockdown, alone and in combination with SGS1, in Ae. aegypti.

Sequence and structural similarity are commonly used to infer shared protein function, although exceptions exist^28–31. Our combined analyses support the classification of AaCSPBP as an ortholog of AgCSPBP, with significant amino acid sequence, secondary structure, and predicted three-dimensional conformational similarities^28,32. Furthermore, the predicted UPF3 domains in both proteins are highly similar in secondary and three-dimensional structure. AlphaFold²¹ analysis indicates that the only other domain present in AgCSPBP is a single alpha helix involved in helix-helix interactions from residues 219–257. This domain is not predicted for AaCSPBP, although a similar structure with a similar length appears from residues 231 to 266. One possible reason for this structure not being predicted as a domain is the uncertainty in the AlphaFold prediction in its latter half, although this domain may also not occur in AaCSPBP.

Limited similarity outside the conserved domain, together with AlphaFold predictions, suggests C-terminal divergence driven by low-complexity regions. In NMD-associated proteins, such regions may facilitate flexible protein interactions. Given the phylogenetic similarity between P. gallinaceum and P. falciparum³³ and CSP binding to the CSPBP N-terminus, the UPF3 domain likely represents a conserved CSP interaction site.

RT-PCR of mosquitoes injected with dsAaCSPBP and dsSGS1¹⁴ showed significant transcript reduction at 2–3 days post-injection, returning to control levels by day 4. Accordingly, the optimal injection window for Plasmodium infections is 6–7 days post-infection; earlier injections risk expression recovery, while later injections occur after salivary gland invasion. Sequential injections were required, as simultaneous delivery increased mortality. While CSPBP knockdown in An. gambiae used 1000 ng/µL dsRNA¹⁶ Ae. aegypti required 3000 ng/µL, consistent with species-specific differences in dsRNA efficacy^14,34.

Compared to our control, AaCSPBP knockdown reduced salivary gland sporozoites by 62% and increased hemolymph sporozoites by 2.44-fold. The reduction in salivary sporozoite count supports the idea that AaCSPBP specifically mediates salivary gland invasion, as the combined number of sporozoites in both tissues remains relatively constant between knockdown and control mosquitoes. In addition, since the dsRNA was injected after sporozoite release from oocysts, the observed impact could not have resulted from interactions with earlier stages of the parasite’s life cycle. If the mechanism of action involved killing, we would expect to observe an overall decrease in the total number of sporozoites. Furthermore, the limited duration of knockdown and the high concentration required for silencing could mean that AaCSPBP expression was not fully prevented. As a result of these factors, we may have obtained a smaller reduction than would be observed with the complete abolition of AaCSPBP.

Furthermore, knockdown of both AaCSPBP and SGS1 resulted in an even greater decrease in the number of sporozoites in the salivary glands, with an average of a 94% decrease across three replicates compared to the control. This combined trial could be subject to the same possible underestimation of percentage reduction as the trial with AaCSPBP silencing alone for the reasons identified above.

Due to the high degree of reduction in the dual knockdown experiment, we questioned whether AaCSPBP and SGS1 function independently in mediating sporozoite invasion of the salivary glands. We used the equation defined in the “Methods” section to determine the percentage reduction in sporozoite count for each mosquito in the experimental group for the AaCSPBP and SGS1 dual knockdown. By recontextualizing percentage reduction as the probability of sporozoite salivary gland entrance inhibition relative to the control, we can use the rules of probability to determine the expected reduction if the knockdowns were acting independently, using the equation:

where 100% is the percentage of sporozoites present in the corresponding control mosquitoes, X is the observed average percentage reduction for AaCSPBP knockdown, and Y is the observed average percentage reduction for SGS1 knockdown. Essentially, this equation is using the probability of sporozoite entry for dsCSPBP mosquitoes relative to the control (1-X) and the probability of sporozoite entry for dsSGS1 mosquitoes relative to the control (1-Y) to determine the probability of sporozoite entry in a mosquito where both proteins are knocked down and function independently, as the probability of entry across the two independent events occurring can be determined by the equation P(A ∩ B) = P(A)P(B). Then, this probability is subtracted from 100% (the likelihood of control sporozoite entry relative to itself) to determine the expected probability of inhibition for the dual knockdown trial, otherwise known as the percentage reduction. This equation is algebraically identical to the Bliss independence model³⁵.

Using the average percent reduction for the AaCSPBP and SGS1 trials, the corresponding percent reductions were input into this equation, resulting in a theoretical percent reduction of 100% − (1–0.62)*(1 − 0.67)¹⁴, or 87.46%. Taking the null hypothesis that there is no difference between the theoretical (independent) group and the observed group, we ran a Wilcoxon one-sample-sample signed-rank test using the observed group and this theoretical percentage reduction. It was determined that the average reduction in the observed group was greater than that of the theoretical prediction, and that this difference was statistically significant (p = 0.0130). This means we can reject the null hypothesis and conclude that the observed reduction is greater than the theoretical reduction. Since the theoretical reduction reflects the independent effects of AaCSPBP and SGS1 knockdown, and the observed reduction exceeds the theoretical reduction, we can conclude that the knockdowns are operating synergistically.

What accounts for the difference between the observed and theoretical groups? One possibility is higher knockdown efficacy in the combined trial, although there is no direct evidence for this. Alternatively, AaCSPBP and SGS1 may have overlapping roles in invasion, and dual knockdown may prevent functional compensation. They may also form a complex that facilitates sporozoite entry, resulting in a greater-than-additive effect. Overall, these findings support a synergistic role in invasion. However, sporozoite invasion was not eliminated, suggesting that additional receptors are involved, consistent with current understanding in the field¹⁷.

While this study focuses on Ae. aegypti salivary gland invasion by P. gallinaceum sporozoites, it may provide broader insight into parasite invasion in human malaria vectors. Parasite progression within mosquitoes involves multiple proteins that may act synergistically, although most have been studied individually¹⁷. Saglin, a well-characterized protein in Anopheles gambiae³⁶, plays a key role in midgut colonization and is implicated in salivary gland invasion^37–39 but lacks a clear ortholog in Ae. aegypti. SGS1 knockout in Ae. aegypti partially impairs parasite colonization in both tissues, suggesting a functional parallel with Saglin. Together, these findings indicate that parasite progression is mediated by multiple, potentially cooperative proteins rather than a single conserved factor.

The presence of an NMD-associated domain in AaCSPBP suggests that this protein may function as a post-transcriptional regulator by influencing nonsense-mediated mRNA decay pathways. AaCSPBP, through modulation of mRNA stability and turnover, could regulate the expression of multiple downstream factors that then act synergistically with SGS1, thereby indirectly contributing to parasite progression. This regulatory role is consistent with NMD functions that extend beyond transcript quality control to fine-tune gene expression networks^40,41, supporting a model in which AaCSPBP does not act as a direct facilitator of invasion but instead coordinates the availability of proteins involved in this process.

Although AaCSPBP contains an NMD-associated domain, it may not function exclusively in mRNA decay, as domains can be repurposed in host–parasite interactions. AaCSPBP may instead act as a host cofactor or entry mediator during salivary gland invasion. DeepLoc 2.1 analysis⁴¹ predicts cytoplasmic localization with a nuclear localization signal and membrane traversal capability, while SGS1 is predicted to be a transmembrane protein. This supports a model in which SGS1 mediates initial sporozoite interaction and entry, followed by intracellular interaction with CSPBP to facilitate trafficking and passage through secretory duct membranes. A direct role for CSPBP is further supported by its identification via CSP binding in a yeast surface display assay¹⁶.

This work supports a synergistic role for SGS1 and AaCSPBP in P. gallinaceum sporozoite invasion of Ae. aegypti salivary glands. Future gene knockout studies will clarify whether the differences between theoretical and observed dual knockdown effects are solely due to this synergy and whether these proteins have a greater impact than observed here. Overall, these findings advance our understanding of sporozoite invasion and support the development of genetic control strategies against malaria.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceived and designed the experiments: BB. Performed the experiments: AM, HC, BB. Analyzed the data: AM, BB. Contributed reagents/materials/analysis tools: BB, MC; Wrote the manuscript: AM, BB. All authors reviewed and approved the manuscript.

Data availability

All data generated or analyzed during this study are included in this published article (and its Supplementary Information files).

Declarations

Competing interests

The authors declare no competing interests.