Pyruvate kinase is post-translationally regulated by sirtuin 2 in Aedes aegypti mosquitoes

Abstract

We previously demonstrated that Aedes aegypti pyruvate kinase (AaPK) plays a key role in the regulation of both carbon and nitrogen metabolism in mosquitoes. To further elucidate whether AaPK can be post-translationally regulated by Ae. aegypti sirtuin 2 (AaSirt2), an NAD⁺-dependent deacetylase that catalyzes the removal of acetyl groups from acetylated lysine residues, we conducted a series of analysis in non-starved and starved female mosquitoes. Transcriptional and protein profiles of AaSirt2, analyzed by qPCR and western blots, indicated that the AaSirt2 is differentially modulated in response to sugar or blood feeding in mosquito tissues dissected at different times during the first gonotrophic cycle. We also found that AaSirt2 is localized in both cytosolic and mitochondrial cellular compartments of fat body and thorax. Multiple lysine-acetylated proteins were detected by western blotting in both cellular compartments. Furthermore, western blotting of immunoprecipitated proteins provided evidence that AaPK is lysine-acetylated and bound with AaSirt2 in the cytosolic fractions of fat body and thorax from non-starved and starved females. In correlation with these results, we also discovered that RNAi-mediated knockdown of AaSirt2 in the fat body of starved females significantly decreased AaPK protein abundance. Notably, survivorship of AaSirt2-deficient females maintained under four different nutritional regimens was not significantly affected. Taken together, our data reveal that AaPK is post-translationally regulated by AaSirt2.

1. Introduction

Sirtuins (Sirts), homologs of the Saccharomyces cerevisiae yeast silent information regulator 2, are known as NAD⁺-dependent deacetylases that play important roles in numerous biological processes, such as DNA repair, aging, stress resistance and nutrient-responsive pathways (Michan and Sinclair, 2007; Teixeira et al., 2020). In mammals, seven Sirts (SIRT1-7) were identified. SIRT1 mainly localizes in the nucleus, though SIRT1 has also been found in cytoplasm (Tanno et al., 2007). SIRT6 also localizes in nucleus (Mostoslavsky et al., 2006), while SIRT7 was reported to specifically reside in the nucleolus within nucleus (Ford et al., 2006). SIRT3, SIRT4 and SIRT5 are mitochondrial proteins (Ahuja et al., 2007; Hershberger et al., 2017; Schwer et al., 2002). SIRT2 is mainly a cytoplasmic enzyme (North et al., 2003), but it can also translocate into nucleus (Vaquero et al., 2006). Notably, it has also been shown that certain Sirts, such as SIRT2 can exhibit demyristoylase activity (Teng et al., 2015), SIRT4 and SIRT6 can act as mono-ADP-ribosyl transferases (Haigis et al., 2006; Liszt et al., 2005), whereas SIRT5 can display demalonylase and desuccinylase activities (Du et al., 2011).

Insect Sirts have not been extensively investigated. The reported number and types of Sirts vary in different insect species (Richards et al., 2008; Solignac et al., 2007; Zhang et al., 2022). In the fruit fly Drosophila melanogaster, five Sirts have been identified (dSirt1, 2, 4, 6 and 7). It has been postulated that dSirt2 is a homolog to both human SIRT2 and SIRT3, and that dSirt2 could regulate different physiological processes, including lifespan and mitochondrial function (Rahman et al., 2014; Rimal et al., 2023).

Growing evidence has indicated that the human SIRT2 participates in the regulation of carbohydrate pathways such as glycolysis, gluconeogenesis, and the pentose phosphate pathway (Cha et al., 2017; Wang et al., 2014; Zhang et al., 2017). Numerous enzymes, including phosphoenolpyruvate carboxykinase (PEPCK) 1, lactate dehydrogenase A, and phosphoglycerate mutase (PGAM), have been identified as substrates for SIRT2 (Jiang et al., 2011; Xu et al., 2014; Zhao et al., 2013). It was also postulated that in tumor cells maintained under low nutritional conditions, SIRT2 deacetylates and facilitates a tetramer formation and activation of the M2 isoform of human pyruvate kinase (PKM2), which leads to an increase of ATP synthesis via oxidative phosphorylation (Park et al., 2016).

The recombinant PK1 isoform from Aedes aegypti mosquitoes (AaPK1) has been characterized (Petchampai et al., 2019). It was found that AaPK1 is allosterically regulated by specific amino acids and phosphorylated sugars (Petchampai et al., 2019). It was also found that AaPK participates in ammonia detoxification, a process that involves multiple metabolic pathways in Ae. aegypti mosquitoes (Horvath et al., 2018, 2021; Isoe et al., 2017, 2022; Isoe and Scaraffia, 2013; Mazzalupo et al., 2016; Petchampai et al., 2020; Scaraffia et al., 2005, 2006, 2008, 2010; Scaraffia and Wells, 2003). In addition, it was reported that AaPK protein abundance and enzymatic activity increase in fat body but not in thorax of females deprived of sugar, indicating that AaPK in fat body tissues is modulated during starvation (Petchampai et al., 2020). Moreover, it was shown that the transient gene silencing of AaPK by RNA interference (RNAi) reduces glucose oxidation and metabolic pathways that support nitrogen disposal in Ae. aegypti (Petchampai et al., 2020). Although AaPK is critical for maintaining metabolic homeostasis in mosquitoes, mechanisms underlying the regulation of AaPK and the intersecting pathways are still unknown.

Despite the efforts in controlling Ae. aegypti mosquitoes, these vectors of arboviruses causing infectious diseases represent a serious public health burden worldwide (Dahmana and Mediannikov, 2020). The identification of the biochemical and molecular bases underlying the regulation of carbon and nitrogen metabolism in Ae. aegypti is crucial, so that novel targets can be discovered and effective metabolism-based strategies for mosquito control can be developed. In this report, we investigated whether AaPK can be post-translationally regulated by Ae. aegypti Sirt2 (AaSirt2). We found that AaSirt2 gene and protein profiles are distinct in tissues from sugar- and blood-fed mosquitoes. We also discovered that AaSirt2 is localized in both cytosolic and mitochondrial subcellular compartments of mosquito fat body and thorax, where numerous proteins acetylated on the lysine residues were also detected by western blotting. Immunoprecipitation-western blot analyses provided further evidence that AaPK is acetylated on the lysine residues, and that AaPK binds with AaSirt2 in the cytosolic fractions isolated from both non-starved and starved females. Furthermore, we discovered that genetic depletion of AaSirt2 by RNAi significantly impacts AaPK protein levels in fat body from mosquitoes deprived of sugar. In summary, our data demonstrate that AaPK is a target of AaSirt2.

2. Materials and methods

2.1. Reagents, chemicals, and antibodies

Tris, nicotinamide (NAM), trichostatin A (TSA), custom-designed primers, cOmplete^™ Mini EDTA-free Protease Inhibitor Cocktail (PIC) and other chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA). Oligo-(dT)₂₀ primer, reverse transcriptase, and GoTaq^® DNA Polymerase were purchased from Promega (Madison, WI, USA). PerfeCTa SYBR Green FastMix was acquired from Quanta BioSciences (Gaithersburg, MD, USA). HiScribe^™ T7 High Yield RNA Synthesis Kit and Monarch^® RNA Cleanup Kit were purchased from New England (Ipswich, MA, USA). TRIzol^® Reagent, Mitochondria Isolation Kit for Tissue, Pierce^™ Classic IP Kit, and Pierce^™ Microplate BCA Protein Assay Kit were obtained from Thermo Fisher Scientific Inc. (Waltham, MA, USA). Bovine blood was purchased from Pel-Freeze Biologicals (Rogers, AR, USA).

Custom-made rabbit polyclonal antibodies against AaPK and AaSirt2 were acquired from GenScript USA Inc. (Piscataway, NJ, USA). The recombinant AaPK1 protein was used as an antigen for AaPK antibody generation (Petchampai et al., 2019, 2020). The full-length bacterially overexpressed AaSirt2 was used as an antigen for AaSirt2 antibody production. An anti-α-tubulin mouse monoclonal antibody was purchased from Developmental Studies Hybridoma Bank (The University of Iowa, Iowa City, IA, USA). An anti-porin antibody and a normal Rabbit IgG antibody were obtained from Abcam (Cambridge, UK). Secondary antibodies, IRDye 800CW goat anti-rabbit, IRDye 800CW goat anti-mouse and IRDye 680RD donkey anti-mouse, were acquired from LI-COR Biosciences (Lincoln, NE, USA). An anti-acetyl lysine antibody and anti-acetyl lysine antibody agarose beads were purchased from ImmuneChem Pharmaceuticals Inc. (Burnaby, British Columbia, Canada).

2.2. Mosquito rearing and feeding

Ae. aegypti (NIH-Rockefeller strain) colony was maintained in Percival Intellus I-41VL incubators (Percival Scientific, Inc., Perry, IA, USA) and in a CARON 6015 Insect Growth Chamber connected to a CARON CRSY 102 condensate recirculating system (Caron Products and Services, Inc., Marietta, OH, USA) at 28 °C, and 75% relative humidity with a light: dark cycle of 16h: 8h as previously described (Isoe and Scaraffia, 2013; Mazzalupo et al., 2016). Newly-eclosed mosquitoes were continuously supplied with cotton balls soaked in 3% sucrose for the first 4 days. To assess the level of AaSirt2 mRNA and protein abundance, 4-day-old mosquitoes were fed with bovine blood supplemented with 5 mM ATP and maintained on 3% sucrose. Thorax, fat body, midgut, ovaries, and Malpighian tubules were dissected at different times post blood meal (PBM). The term thorax used in the manuscript refers to the thoracic segments, including the flight muscles and fat body within them, without appendages. The tissues were stored at −80 °C for quantitative real-time PCR (qPCR) and western blotting. For the rest of the experiments and unless otherwise stated, Ae. aegypti females (4-day-old) were provided with either 3% sucrose- or water-soaked cotton balls for 72 h. Fat body and thorax were subsequently collected and kept at −80 °C until used for protein extraction. Experiments were conducted using at least three different mosquito cohorts.

2.3. RNAi-mediated silencing of AaSirt2 by double-stranded RNA (dsRNA) microinjection

A gene encoding AaSirt2 (XP_001651467) was identified within the Ae. aegypti genome in the NCBI database. The dsRNAs targeting AaSirt2 gene or firefly luciferase (FL) gene (control) were synthesized as previously described for other genes (Isoe and Scaraffia, 2013). Oligonucleotide primer sequences for dsRNA synthesis are shown in Table 1. PCR was performed using GoTaq^® DNA Polymerase (Promega). In vitro dsRNA synthesis was performed using the HiScribe^™ T7 RNA Synthesis kit at 37 °C overnight following the manufacturer’s instructions. The purified dsRNA was resuspended in nuclease-free HPLC water at 7.5 μg/μl and stored at −80 °C until use. Ae. aegypti females were microinjected twice with 2.0 μg dsRNA using a Nanoject II micro-injector (Drummond Scientific Company, Broomall, PA, USA). The first injection was administered within 4 h after female eclosion, whereas the second injection was done at day 4 post-eclosion. Injected mosquitoes were maintained on 3% sucrose for 4 days, and then provided with either sucrose- or water-soaked cotton balls for 72 h. Fat body and thorax were then dissected to monitor the efficiency of the knockdown by western blotting. To access the impact of AaSirt2 deficiency on survival, four groups of 4-day-old dsRNA-injected females were fed four different diets: 1) 3% sucrose, 2) water, 3) a blood meal and maintained on 3% sucrose or 4) a blood meal and maintained on water after feeding. Survival was recorded daily until the last female died, as previously described (Petchampai et al., 2020).

Table 1.

Gene-specific primers for qPCR and dsRNA synthesis.

Genes, GenBank accession number		Primer sequence (5′ to 3′)	PCR, bp
Gene-specific primers used for qPCR
Sirtuin 2, XP_001651467	Forward	GCGCAGCACAAGCATTTCC
	Reverse	TGTTGGACATCGTGTATCAT	110
Ribosomal protein S7, XP_001660169	Forward	ACCGCCGTCTACGATGCCA
	Reverse	ATGGTGGTCTGCTGGTTCTT	131
Gene-specific primers used for RNAi
Sirtuin 2, XP_001651467	Forward	*GAGAAGATCGTGGAAGCTCA
	Reverse	*TGCCGTTTCGGTTGGGTT	564
Firefly luciferase, FL, U47295	Forward	*AGCACTCTGATTGACAAATACGA
pGL3-Basic Vector	Reverse	*AGTTCACCGGCGTCATCGTC	548

^*T7 promoter sequence (5′ TAATACGACTCACTATAGGGAGA 3′) was added in 5′ of each primer.

2.4. qPCR assays

Transcriptional expression of AaSirt2 in mosquito tissues in response to sugar and blood feeding was examined by qPCR as previously described for other genes (Isoe and Scaraffia, 2013). Total RNA was extracted using TRIzol^® Reagent, and cDNA was synthesized using oligo-(dT)₂₀ and reverse transcriptase. Ribosomal protein S7 mRNA was used as an internal normalization control. Primer sequences used in this study are shown in Table 1.

2.5. Protein extraction

Total proteins were extracted from mosquito tissues as previously described (Isoe et al., 2022; Isoe and Scaraffia, 2013; Mazzalupo et al., 2016). For midgut cellular protein extraction, bovine blood present in the midgut lumen, which may interfere with western blot analysis, was removed by washing the lumen with 1X phosphate-buffered saline containing 1X PIC. Mitochondrial and cytosolic proteins were extracted from fat body and thorax using Mitochondria Isolation Kit for Tissue according to the manufacturer’s protocol with minor modifications. A pool of approximately 200 mg of fat body or thorax was homogenized using Fisher Scientific^™ Laboratory Homogenizer (Model 125, Thermo Fisher Scientific Inc.). For downstream application of immunoprecipitation (IP) with an anti-acetyl lysine antibody, 10 mM NAM, 0.5 μM TSA, and 1X PIC were added to all buffers except Mitochondria Isolation Reagent B during mitochondria/cytosol isolation. For IP with an anti-AaSirt2 antibody, only 1X PIC was added to the buffers throughout the mitochondria/cytosol extraction process. The protein concentration was determined using Pierce^™ Microplate BCA Protein Assay Kit following manufacturer’s instructions.

2.6. Western blotting procedures

Western blotting was performed as previously described for other proteins (Isoe et al., 2022; Isoe and Scaraffia, 2013; Mazzalupo et al., 2016). Briefly, proteins were resolved on 4–15% gradient gels (Bio-Rad, Hercules, CA, USA), transferred to nitrocellulose membrane (LI-COR Biosciences), incubated with antibodies, and visualized with Odyssey Infrared Imaging System (LI-COR Biosciences) or Chemidoc MP Imaging System (Bio-Rad). Pre-stained protein markers were utilized as standards for estimating protein molecular weight (Bio-Rad). The dilutions of the primary antibodies were as follows: custom-made anti-AaSirt2 antibody (1:1000), anti-acetyl lysine antibody (1:1000), custom-made anti-AaPK antibody (1:1000), anti-α-tubulin antibody (1:2000), and anti-porin antibody (1:300). The secondary antibodies (IRDye 800CW goat anti-rabbit, IRDye 800CW goat anti-mouse, and IRDye 680RD donkey anti-mouse) were diluted at 1:10000. Protein band intensity was quantified by densitometry using ImageJ software (Schneider et al., 2012).

2.7. IP procedures

IP was performed using Pierce^™ Classic IP Kit according to the manufacturer’s protocols with minor modification. Cytosolic proteins (250 μg) were pre-cleared with Pierce Control Agarose Resin at 4 °C with a constant rotation for 1 h. For IP of lysine-acetylated proteins (250 μg), anti-acetyl lysine antibody agarose beads were incubated with the pre-cleared cytosolic proteins at 4 ^◦C with rotation overnight. IPs with normal rabbit IgG and no antibody were used as negative controls. The beads were washed with IP Lysis/Wash buffer supplemented with 10 mM NAM, 0.5 μM TSA, and 1X PIC, followed by washing with Pierce 1X Conditioning Buffer. The immunoprecipitated proteins were eluted with 50 μl elution buffer, neutralized with 1 M Tris, pH 9.5 (5 μl), and the samples (27.5 μl) were analyzed by western blotting with an anti-AaPK antibody.

For IP of AaSirt2, the cytosolic proteins (250 μg), which were pre-cleared with Pierce Control Agarose Resin, were incubated with an anti-AaSirt2 antibody at 4 °C with a constant rotation overnight, followed by incubating with Pierce Protein A/G Agarose at 4 °C for 1 h. IPs with normal rabbit IgG and no antibody were used as negative controls. The resin was washed with IP Lysis/Wash buffer supplemented with 1X PIC, followed by washing with Pierce 1X Conditioning Buffer. The immunoprecipitated proteins were eluted and analyzed by western blotting with an anti-AaPK antibody, as indicated above.

2.8. Statistical analysis

Statistical analyses were carried out using GraphPad Prism version 9.0 for Mac OS X (GraphPad Software, La Jolla, CA, USA). One-way ANOVA was used to examine the difference in AaSirt2 mRNA relative abundance during the first gonotrophic cycle (0–96 h PBM). The densitometry data were analyzed by an unpaired Student’s t-test. Survival analysis was performed by the Kaplan-Meier method and the differences were assessed by the Log-rank (Mantel-Cox) test. A value of p <0.05 was considered significant. Experiments were replicated at least three times with at least three independent biological samples.

3. Results

3.1. AaSirt2 is differentially regulated in response to blood feeding

A BLAST search of Ae. aegypti genome in the NCBI database identified four Sirt homologs, which are Sirt2, Sirt4, Sirt6, and Sirt7. Amino acid sequence alignments showed a highly conserved NAD⁺-dependent deacetylase domain in all four mosquito Sirts (Fig. 1). The deacetylase domain of AaSirt2 (XP_001651467) shares 91, 88, 62, 68, 65 and 59% sequence identity to this domain from Culex quinquefasciatus, Anopheles gambiae, D. melanogaster, Bombyx mori, Apis mellifera, and Homo sapiens, respectively (Fig. 1). In addition, AaSirt2 protein has 54% amino acid identity to human SIRT2 (AAK51133) and 52% identity to human SIRT3 (AAD40851). Accession numbers of Sirt proteins used in the sequence analysis are listed in Table S1.

Fig. 1. Protein sequence comparison of sirtuin 2, 4, 6, and 7 among insects and human.

The predicted amino acid sequence of each of the four sirtuins from Aedes aegypti was aligned with the corresponding protein sequence from Culex quinquefasciatus, Anopheles gambiae, Drosophila melanogaster, Bombyx mori, Apis mellifera, and Homo sapiens. The conserved catalytic domain shared by all sirtuins is highlighted in orange. The corresponding numbers refer to amino acid residues within the proteins. Sirt2, sirtuin 2; Sirt4, sirtuin 4; Sirt6, sirtuin 6; Sirt7, sirtuin 7. GenBank accession numbers for NAD⁺-dependent protein deacetylase sirtuins used in the alignments are shown in Table S1.

To determine how AaSirt2 is expressed, we evaluated gene and protein profiles of AaSirt2 in several mosquito tissues in response to sugar and blood feeding.

Transcriptional analyses of AaSirt2 in the fat body, thorax, midgut, Malpighian tubules, and ovaries of sugar- and blood-fed mosquitoes by qPCR revealed an overall stable constitutive gene expression pattern of AaSirt2 after blood feeding. However, the AaSirt2 mRNA level was up-regulated in the ovaries of the blood-fed females at 72 and 96 h PBM (Fig. S1).

AaSirt2 protein profiles were also analyzed in the tissues of sugar-and blood-fed Ae. aegypti by western blotting using a custom-made anti-AaSirt2 antibody. In sucrose-fed mosquitoes, the AaSirt2 protein was detected in all tissues analyzed. An up-regulation of the AaSirt2 protein was observed in the fat body, midgut, and ovaries at 3–24 h, 3–36 h, and 3–12 h PBM, respectively. In the fat body, AaSirt2 protein level had a peak at 12–24 h PBM and decreased at 36–72 h PBM. In the midgut, AaSirt2 was expressed at its highest level at 3–36 h PBM and slightly decreased thereafter. In the ovaries, AaSirt2 protein was up-regulated after blood feeding, reached the highest level at 12 h PBM and then decreased at 18–72 h PBM. In the thorax, AaSirt2 protein level remained high and stable throughout the time course analyzed. In the Malpighian tubules, the protein was constitutively expressed during the first gonotrophic cycle (Fig. 2).

Fig. 2. Aedes aegypti sirtuin 2 (AaSirt2) protein profiles in mosquito tissues.

Fat body, midgut, Malpighian tubules, thorax, and ovary tissues were collected from sucrose- and blood-fed mosquitoes at different times after feeding. Western blotting was performed using a custom-made anti-AaSirt2 antibody. For fat body and ovaries, each lane contains 1 tissue equivalent of the protein extracts. For thorax, midgut, and Malpighian tubules, each lane contains 0.3, 0.5, and 0.5 tissue equivalent of protein extracts, respectively. An anti-α-tubulin antibody was used as a loading control. Western blots are representative of 3 independent biological replicates. Each protein extract was prepared from a pool of 10 mosquito tissues. N = 270. M, protein markers; PBM, post blood meal; S, sucrose-fed.

3.2. AaSirt2 is localized in mitochondrial and cytosolic fractions isolated from mosquito tissues

As discussed above, AaSirt2 is evolutionarily closely related to human SIRT2 and SIRT3. To identify the cellular localization of AaSirt2 and its relative abundance in response to starvation, we isolated mitochondrial and cytosolic proteins from tissues of sucrose- and water-fed female mosquitoes. Subsequently, both protein fractions were subjected to western blot analysis using an anti-AaSirt2 antibody. As shown in Fig. 3A and B, AaSirt2 was detected in both mitochondrial and cytosolic fractions extracted from the fat body and thorax of non-starved and starved Ae. aegypti mosquitoes. The relative abundance of the AaSirt2 in each cellular compartment was similar, independent of how the females were fed (Fig. 3C and D).

Fig. 3. Aedes aegypti sirtuin 2 (AaSirt2) protein abundance in mitochondrial and cytosolic fractions isolated from fat body and thorax.

Newly-eclosed mosquitoes were maintained on 3% sucrose until day 4 post-eclosion, and then fed on either 3% sucrose or water for 72 h. Mitochondrial and cytosolic protein fractions were extracted from fat body (A) and thorax (B) of female mosquitoes fed on either 3% sucrose or water for 72 h. Western blotting was subsequently performed using an anti-AaSirt2 antibody. C and D. The AaSirt2 bands detected by western blots were quantified and normalized to porin or α-tubulin. Densitometric analysis was performed with NIH Image J software (Schneider et al., 2012). Each lane contains 20–30 μg of cytosolic proteins or 40–60 μg of mitochondrial proteins. Anti-porin and anti-α-tubulin antibodies were used as loading controls for mitochondrial and cytosolic proteins, respectively. Western blots are representative of at least 3 independent biological replicates. Each protein extract replicate was prepared from a pool of approximately 200 mg of tissues. For fat body tissues isolated from sucrose- and water-fed mosquitoes, N = 1,111 (n = 346, 404, and 361) and 1,920 (n = 745, 640, and 535), respectively. For thorax isolated from sucrose- and water-fed mosquitoes, N = 601 (n = 200, 201, and 200) and 626 (n = 207, 219, and 200), respectively. Data represent the mean fold change ± SEM of 3 independent western blots.

An unpaired Student’s t-test was employed. No significant difference (NS) was observed between treatments. M, protein markers; S, sucrose; W, water.

3.3. Multiple mitochondrial and cytosolic proteins of tissues from non-starved and starved mosquitoes are acetylated on lysine residues

To examine whether mitochondrial and cytosolic proteins are acetylated on the lysine residues, western blotting using an anti-acetyl lysine antibody was performed on the mitochondrial and cytosolic fractions isolated from tissues of sucrose- and water-fed mosquitoes. As shown in Fig. 4A and B, several lysine-acetylated proteins were detected in both fractions of fat body and thorax, and their acetylation profiles in each cellular compartment were similar in non-starved and starved females (Fig. 4C and D).

Fig. 4. Mitochondrial and cytosolic protein acetylation in Aedes aegypti fat body and thorax.

Newly-eclosed mosquitoes were maintained on 3% sucrose until day 4 post-eclosion, and then fed on either 3% sucrose or water for 72 h. Mitochondrial and cytosolic protein fractions were isolated from the fat body (A) and thorax (B) of mosquitoes fed on either 3% sucrose or water for 72 h. Western blotting was performed using an anti-acetyl lysine antibody. The acetylated protein bands detected by western blots were quantified and normalized to porin or α-tubulin (C and D). Densitometric analysis was performed with NIH Image J software (Schneider et al., 2012). Each lane contains 20 μg of proteins. Anti-porin and anti-α-tubulin antibodies were used as loading controls for mitochondrial and cytosolic proteins, respectively. Western blots are representative of 3 independent biological replicates. Each protein extract replicate was prepared from a pool of approximately 200 mg of tissues. For fat body tissues isolated from sucrose- and water-fed mosquitoes, N = 1,216 (n = 333, 444, and 439) and 2,143 (n = 746, 702, and 695), respectively. For thorax dissected from sucrose- and water-fed mosquitoes, N = 543 (n = 169, 201, and 173) and 579 (n = 233, 192, and 154), respectively. Data represent the mean fold change ± SEM of 3 independent western blots. An unpaired Student’s t-test was employed. No significant difference (NS) was observed between treatments. M, protein markers; S, sucrose; W, water.

3.4. AaPK physically interacts with AaSirt2 in mosquito fat body and thorax

To provide evidence that AaPK is acetylated on lysine residues, IP with an anti-acetyl lysine antibody was performed in the cytosolic fractions isolated from fat body and thorax of non-starved and starved females. The enriched lysine-acetylated proteins were then subject to western blotting using an anti-AaPK antibody. The IP/western blotting data demonstrate that AaPK is indeed a lysine-acetylated protein (Fig. 5A–D). To further demonstrate that AaSirt2 physically interacts with AaPK, IP with an anti-AaSirt2 antibody followed by western blotting probed with an anti-AaPK antibody was conducted in the cytosolic fractions from both tissues. As shown in Fig. 6A–D, AaPK binds to AaSirt2 independently of the nutritional status of the females.

Fig. 5. Aedes aegypti pyruvate kinase (AaPK) acetylation in fat body and thorax.

Newly-eclosed mosquitoes were maintained on 3% sucrose until day 4 post-eclosion, and then fed on either 3% sucrose or water for 72 h. Cytosolic proteins were extracted from fat body (A) and thorax (B) of mosquitoes fed on either 3% sucrose or water for 72 h. Immunoprecipitation was performed using an anti-acetyl lysine antibody followed by western blotting probed with an anti-AaPK antibody. Normal rabbit IgG and no antibody were used as negative controls. An anti-α-tubulin antibody was used as a loading control. The immunoprecipitated proteins detected by western blots were quantified and normalized to the input (C and D). Densitometric analysis was performed with NIH Image J software (Schneider et al., 2012). Western blots are representative of 3 independent biological replicates. Each protein extract replicate was prepared from a pool of approximately 200 mg tissue samples, and 250 μg of proteins was used for immunoprecipitation. For fat body tissues isolated from sucrose- and water-fed mosquitoes, N = 1,185 (n = 323, 433, and 429) and 2,038 (n = 711, 667, and 660), respectively. For thorax dissected from sucrose- and water-fed mosquitoes, N = 537 (n = 191, 192, and 154) and 565 (n = 191, 201, and 173), respectively. Data represent the mean fold change ± SEM of 3 independent western blots. An unpaired Student’s t-test was employed. No significant difference (NS) was observed between treatments. M, protein markers; S, sucrose; W, water.

Fig. 6. Interaction between Aedes aegypti sirtuin 2 (AaSirt2) and Ae. aegypti pyruvate kinase (AaPK) in fat body and thorax.

Newly-eclosed mosquitoes were maintained on 3% sucrose until day 4 post-eclosion, and then fed on either 3% sucrose or water for 72 h. Cytosolic proteins were extracted from fat body (A) and thorax (B) of female Ae. aegypti fed on either 3% sucrose or water for 72 h. Immunoprecipitation using an anti-AaSirt2 antibody was performed prior to western blotting probed with an anti-AaPK antibody. Normal rabbit IgG and no antibody were used as negative controls. An anti-α-tubulin antibody was used as a loading control. The immunoprecipitated proteins detected by western blots were quantified and normalized to the input (C and D). Densitometric analysis was performed with NIH Image J software (Schneider et al., 2012). Western blots are representative of 3 independent biological replicates. Each protein extract replicate was prepared from a pool of approximately 200 mg tissue samples, and 250 μg of proteins was used in immunoprecipitation. For fat body tissues isolated from sucroseand water-fed mosquitoes, N = 1,145 (n = 450, 324, and 371) and 2,332 (n = 830, 803, and 699), respectively. For thorax isolated from sucrose- and water-fed mosquitoes, N = 525 (n = 178, 177, and 170) and 560 (n = 195, 168, and 197), respectively. Data represent the mean fold change ± SEM of 3 independent western blots. An unpaired Student’s t-test was employed. No significant difference (NS) was observed between treatments. M, protein markers; S, sucrose; W, water.

3.5. RNAi-driven depletion of AaSirt2 significantly decreases AaPK protein abundance in fat body of starved females

To evaluate whether a presence of AaSirt2 directly affects AaPK protein abundance in tissues dissected from female mosquitoes maintained under different nutritional conditions, we employed RNAi against AaSirt2. As shown in Fig. 7A and B, representative western blots using single mosquito analysis indicate that AaSirt2 protein level was significantly reduced in fat body tissues of both sucrose- and water-fed AaSirt2-deficient mosquitoes when compared to RNAi-FL control mosquitoes. Next, we used the same protein extracts to determine AaPK protein abundance. As shown in Fig. 7C and D, AaPK relative abundance in the fat body tissues of non-starved AaSirt2-deficient mosquitoes was not different from those FL control mosquitoes, while AaPK protein level was significantly reduced in the fat body of starved AaSirt2-deficient mosquitoes (Fig. 7E and F). We also examined the same association of AaPK with AaSirt2 in the thorax. Although AaSirt2 protein abundance was strongly reduced by RNAi, AaPK protein profile was not affected in the thorax of AaSirt2-deficient mosquitoes under both nutritional conditions (Fig. S2).

Fig. 7. Effect of Aedes aegypti sirtuin 2 (AaSirt2) silencing via RNA interference on AaSirt2 and Ae. aegypti pyruvate kinase (AaPK) protein levels in fat body tissues.

Mosquitoes were microinjected with either dsRNA against firefly luciferase (FL) or AaSirt2 within 4 h after adult eclosion and at day 4 post-eclosion. The four-day-old sugar-fed microinjected mosquitoes were allowed to feed on either 3% sucrose or water for 72 h. Fat body protein extracts were individually examined by western blot analysis. AaSirt2 protein levels in fat body tissues of both sucrose- (A) and water-fed (B) AaSirt2-deficient mosquitoes. AaPK protein abundance in the fat body tissues of sucrose-fed AaSirt2-deficient mosquitoes (C and D). AaPK protein abundance in the fat body tissues of water-fed AaSirt2-deficient mosquitoes (E and F). Each lane was loaded with 0.25 tissue equivalent of fat body protein extracts. An anti-α-tubulin antibody was used as a loading control. Western blots are representative of 3 independent biological replicates. Each protein extract was prepared from a single mosquito tissue. For fat body tissues isolated from sucrose- and water-fed mosquitoes, 54 mosquitoes were used for each treatment. N = 108. The AaPK bands detected by western blots were quantified and normalized to α-tubulin. Densitometric analysis was performed with NIH Image J software (Schneider et al., 2012). Data represent the mean fold change ±SEM of 3 independent western blots. ***P < 0.001 when compared to dsRNA-FL by an unpaired Student’s t-test. M, protein markers; NS, not significant.

To examine the effect of genetic knockdown of AaSirt2 on mosquito survival and fitness, four-day-old AaSirt2-deficient females maintained on 3% sucrose were then provided four different diets. As shown in Fig. S3, disruption of AaSirt2 by RNAi did not significantly alter mosquito survival under four different nutrient regimens.

4. Discussion

In mammals, SIRT family proteins are known to have the ability to modify the acetylation status of specific proteins that play crucial roles in multiple cellular processes (Teixeira et al., 2020). In insects, the role of Sirts has not yet been fully addressed. Recent studies postulated that Sirt2 and Sirt5 from B. mori silkworms could have potential antiviral functions (Zhang et al., 2022). It was also shown that dSirt2 and dSirt4 from D. melanogaster are involved in regulating mitochondria function, energy homeostasis and longevity (Rahman et al., 2014; Rimal et al., 2023; Wood et al., 2018). In Helicoverpa armigera cotton bollworms, deacetylation of PK, PEPCK, and PGAM by Sirt2 was reported (Wang et al., 2018). It was suggested that low Sirt2 levels in H. armigera regulate pyruvate abundance and diapause initiation or lifespan extension (Wang et al., 2018).

In previous works, we provided evidence that the recombinant AaPK1 is allosterically activated or inhibited by specific sugars and amino acids or a combination of both effectors (Petchampai et al., 2019). Mass spectrometry-based stable-isotope tracing coupled with RNAi demonstrated that AaPK plays a crucial role in Ae. aegypti carbon and nitrogen metabolism (Horvath et al., 2021; Petchampai et al., 2020). We also observed that AaPK protein abundance and activity can be modulated in fat body tissues depending on nutritional status of the female mosquitoes (Petchampai et al., 2020). Here, we reveal a molecular association between AaPK and AaSirt2 in Ae. aegypti females. We demonstrate that AaPK is a lysine-acetylated protein. We also provide evidence that AaPK binds to AaSirt2, and that AaPK is a substrate of AaSirt2.

Our bioinformatic analysis indicated that AaSirt2 contains a highly conserved core domain that could confer an NAD⁺-dependent deacetylase activity, as observed in Sirts from other organisms (Frye, 1999). After sugar or blood feeding, AaSirt2 transcript remained stable in most mosquito tissues studied, except for the ovaries. Notably, AaSirt2 protein was differentially expressed in tissues, suggesting an occurrence of a tissue-specific control at the translational level in mosquito tissues in response to blood feeding. In mammals, the plethora of SIRT2 substrates identified reflect cell and tissue-specific functions of SIRT2. Regulation of SIRT2 expression is complex and exists at multiple levels, including at the translational level (Buler et al., 2016; Li et al., 2007; Maxwell et al., 2011).

Nematode and arthropod genomes do not encode mammalian SIRT3 homologs (Greiss and Gartner, 2009). It was reported that dSirt2 from D. melanogaster exhibits 59% identity to SIRT2 and 50% identity to SIRT3 (Rahman et al., 2014). Our bioinformatic data also indicated that AaSirt2 protein is a homolog to both SIRT2 and SIRT3. Even though AaSirt2 does not contain a predicted canonical mitochondrial targeting sequence, we empirically found that AaSirt2 localized in both mitochondrial and cytosolic fractions isolated from fat body and thorax of non-starved and starved female mosquitoes. It was also shown that dSirt2 from fruit flies localized to the mitochondria (Rahman et al., 2014). In addition, SIRT2 can be associated with inner mitochondrial membrane (Liu et al., 2017), while SIRT3 is known as a major deacetylase in mitochondria that regulates almost every aspect of mitochondrial functionality (Michishita et al., 2005; Sosulski et al., 2017). Our western blotting analysis also revealed that the AaSirt2 protein profiles in fat body and thorax of sugar- and water-fed mosquitoes remained similar in the same cellular compartments and tissues analyzed.

In dSirt2 mutant fruit flies, an increased lysine acetylation of mitochondrial proteins involved in pathways that generated energy and an increased starvation stress were reported (Rahman et al., 2014). Significantly, the wild-type flies lived longer than dSirt2 mutants (Rahman et al., 2014). Interestingly, it was shown that the mitochondrial dSirt4 can also regulate energy metabolism and longevity. dSirt4 knockout flies exhibited an increased sensitivity to starvation, an alteration of several pathways - including glycolysis, branched-chain amino acid metabolism, and certain fatty acids catabolism - and a shorter lifespan compared to control flies (Wood et al., 2018). In contrast, the lifespan of mosquitoes with AaSirt2 deficiency was not significantly altered compared to control mosquitoes. While genetic knockout in fruit flies leads to 100% loss of dSirt protein functions, the RNAi is not 100% effective and transient. It is also possible that feedback loops and other compensatory regulatory mechanisms, including activities of other members of the Sirt protein family and/or other deacetylases, reduced the potential adverse effects of AaSirt2 silencing.

The RNAi against AaSirt2 specifically degraded mRNA encoding AaSirt2, leading to a significant reduction of AaSirt2 protein levels for two possible effects, one catalytic and one structural. Since no adverse physiological effects on AaSirt2-deficient mosquitoes were observed, it is possible that AaSirt2 knockdown impacts AaSirt2 function in critical scaffolds for physiological effects in the reduction or absence of catalytic activity. A variety of small molecules targeting human SIRT2 catalytic activity have been identified (Penteado et al., 2023). The earliest SIRT2 selective inhibitors were developed from high-throughput screening of small-molecule libraries. New molecules have been developed to ensure selective inhibition of SIRT2, and SIRT2 inhibitors have been developed from natural sources (Hong et al., 2021; Ibrahim et al., 2023; Outeiro et al., 2007; Taylor et al., 2011). It would be worthwhile to verify if any of these small compounds also target the AaSirt2 catalytic site and then use them to differentiate between the effects of reduced catalytic activity and the negative impacts on protein scaffolds from reduced amounts of AaSirt2.

Nevertheless, our results clearly demonstrate that AaPK is a lysine-acetylated protein and that it physically interacts with AaSirt2, indicating that AaPK is a target of AaSirt2. A similar level of AaPK lysine acetylation in mosquitoes given sugar or water corresponded to a comparable level of the interaction between AaPK and AaSirt2 observed in fat body and thorax. These data also correlate with the unchanged level of AaSirt2 in tissues of mosquitoes maintained on sugar or water for 72 h, highlighting the complexity of the metabolic regulation of AaPK and AaSirt2 in different tissues.

The dynamic interplay between lysine acetylation and deacetylation can affect a variety of protein functions in cells. Previous studies in mammals have shown that the glycolytic enzyme PKM2 can be deacetylated and activated by SIRT2 in cancer cells (Park et al., 2016). It was proposed that under low nutritional conditions, SIRT2 and other SIRTs deacetylate downstream protein targets, including PKM2, leading to an increase of ATP synthesis via oxidative phosphorylation (Park et al., 2016). In the brains of the diapause-destined pupae of H. armigera cotton bollworm, it was shown that pyruvate levels are low (Wang et al., 2018). It was also observed that PK, PEPCK, and PGAM can be deacetylated by Sirt2 in vitro, and that deacetylation of PK by Sirt2 resulted in increased PK protein abundance, while PK acetylation by a Sirt inhibitor reduced the PK level in vitro (Wang et al., 2018). Thus, it was speculated that low Sirt2 expression causes a decrease of PK, PEPCK, and PGAM activities, and a reduction of pyruvate levels and metabolic activity, which induce onset of diapause in the cotton bollworm (Wang et al., 2018). In correlation with these results, we previously reported that AaPK protein level and activity increased in fat body tissues of starved females, and that silencing of AaPK by RNAi caused a reduction of glucose and nitrogen catabolism, and an enhanced survivorship in females maintained under specific nutritional conditions (Petchampai et al., 2020).

As discussed above, disruption of AaSirt2 by reverse genetics did not significantly impact survival of female mosquitoes subjected to different nutritional treatments, suggesting that AaSirt2 does not directly regulate lifespan in Ae. aegypti mosquitoes. However, RNAi-silencing of AaSirt2 significantly reduced AaPK protein level in fat body of starved females, which validates that AaPK is a target of AaSirt2 and that both AaPK and AaSirt2 modulate how mosquitoes respond to starvation. Acetylation/deacetylation of lysine residues can regulate other post-translational modifications such as SUMOylation and ubiquitination, leading to functional activation/inactivation of proteins, changes in protein abundance, protein stability and ubiquitin-proteasome degradation. Thus, we hypothesize that AaSirt2 silencing prevents AaSirt2-mediated deacetylation and activation of AaPK. Ultimately, it could stimulate SUMOylation and/or ubiquitination, and degradation of the acetylated AaPK, as previously reported for other acetylated enzymes including PPECK1 (Jiang et al., 2011) and PKM2 (Kim et al., 2015; Park et al., 2016; Xia et al., 2021), leading to the changes in AaPK protein abundance observed in fat body of starved females with AaSirt2 deficiency. While beyond the scope of this work, the complex regulatory mechanism of ubiquitin-mediated protein degradation, also known as the ubiquitin-proteasome system (Choy et al., 2013; Severo et al., 2013; Zhao et al., 2022), and its interaction with AaPK and AaSirt2 warrant further investigation. The profile of lysine-acetylated proteins in different cellular compartments in non-starved and starved Ae. aegypti mosquitoes should also be evaluated to potentially elucidate additional targets of AaSirt2 and protein interactions.

Taken together, our findings provide evidence that AaPK is post-translationally regulated by AaSirt2, and open new questions regarding how cytosolic and mitochondrial lysine-acetylated proteins interact with AaSirt2 and other deacetylases to maintain glucose and nitrogen homeostasis in Ae. aegypti mosquitoes. A deeper understanding of the the protein interactions and mechanistic regulations of Ae. aegypti is critically needed for the identification of selective metabolic targets or regulators and the further development of successful metabolism-based strategies for mosquito control.

Abbreviations:

AaPK

Aedes aegypti pyruvate kinase

AaSirt2

Ae. aegypti sirtuin 2

dsRNA

double-stranded RNA

dSirt

Drosophila sirtuin

FL

firefly luciferase

IP

immunoprecipitation

NAM

nicotinamide

PBM

post blood meal

PEPCK

phosphoenolpyruvate carboxykinase

PGAM

phosphoglycerate mutase

PIC

protease inhibitor cocktail

PK

pyruvate kinase

PKM2

human pyruvate kinase M2

qPCR

quantitative real-time PCR

RNAi

RNA interference

SIRT1-7

mammalian sirtuins 1–7

Sirts

sirtuins

TSA

trichostatin A