Description of a Brain Neuropeptidome of the Malaria Vector Nyssorhynchus Albimanus and Neuropeptide Differential Expression During Infection with Plasmodium Berghei

Background Insect neuropeptides, mainly synthesized in the brain, play a central role in the control of many physiological processes. A neuropeptidome of the mosquito Nyssorhynchus albimanus was described based in a comparative analysis of the mosquito genome complemented with high-throughput sequencing of brain transcriptomes. In addition, neuropeptides differentially expressed during Plasmodium infection were identied. Results We identied 3,811 transcripts associated to translation, oxidation-reduction process, protein binding, ATP binding, integral components of membrane and ribosome, among others. We identied 29 neuropeptide transcripts that predicted at least 60 biopeptides, including pyrokinin, glycoprotein hormone alpha (GPA2), prothoracicotropic hormone, neuropeptide-like precursor 1 (NPLP1), allatostatin C, orcokinin, corazonin, adipokinetic hormone I, SIFamide, pyrokinin capa-like, pigment-dispersing factor, adipokinetic hormone/corazonin-related peptide (ACP), tachykinin-related peptide, trissin, neuropeptide F, short neuropeptide F (sNPF), diuretic hormone 31, bursicon, crustacean cardioactive peptide (CCAP), allatotropin, allatostatin 1, partner of bursicon (PBURS), ecdysis triggering hormone (ETH), diuretic hormone 44 (Dh44), insulin-like peptides 5, 1, 3, 7, 2 (ILPs) and eclosion hormone (EH) and seven neuropeptide receptors. Transcript mapping to the Ny. albimanus genome provided evidence for the re-annotation of the myosuppressin gene. A quantitative analysis documented increased expression of adipokinetic hormone/corazonin-related peptide, pyrokinin and corazonin in the mosquito brain after Plasmodium berghei infection. Conclusion This work represents an initial effort to characterize the neuropeptide repertoire of Ny. albimanus and provides new information for understanding neuroregulation culturing gametocyte-infected mouse blood as described previously [54]. Groups of 300 female mosquitoes were fed for 1 h using articial membrane feeders with: (i) mouse blood + approximately 800 per µl GFP P. berghei ookinetes (infected group), or (ii) uninfected mouse blood (control group). Unfed mosquitoes were removed, and the engorged ones were incubated at 21 °C to allow for parasite invasion and interaction with the mosquito midgut. At 24 h post-blood feeding, mosquito midguts were analyzed under a 40 X uorescence microscope (Leica DM1000) to conrm the presence (P. berghei-infected group) or absence (control group) of parasites. We observed ookinetes and retorts forms. Only midguts containing more than 300 parasites were included in the infected group. Only brains from infected mosquitoes with conrmed infection from the infected group and all brains from the control group were collected. The expression of adipokinetic hormone/corazonin-related peptide, tachykinin related peptide, SIFamide, myosuppressin, pk/PBAN, corazonin, ILP-2, ILP-3, and IL-5 were investigated in the brain of P. berghei-infected and control mosquitoes through real-time PCR assays. For every neuropeptide gene, three biological replicates were conducted for both conditions. As an internal control, a fragment of actin was amplied using primers RTActU R: (CGA TCC ACT TGC AGA GCC AGT) and RTAct3.2 F: (5´-TAC GCC AAC ATT GTC ATG TCC − 3 ´) [82]. The real-time PCR program was conducted as described above. Generated qRT-PCR Ct values were analyzed using the 2- ∆∆ Ct method [83] and tested with one-way ANOVA, followed by a Kruskal-Wallis post-test (α = 0.05). and be multiple. Our results documented an increase in of ILP-5 transcription in at times when P. berghei parasites had initiated development on the mosquito midgut. Although this increase was consistent among mosquito samples, it was not statically signicant. On the other hand, no difference in ILP 3 transcription was observed between infected and uninfected mosquitoes. These results require further studies with a bigger sample size, but they may indicate that ILPs also participate in P. berghei infection, and that the induction of ILP is a generalized mechanism of Plasmodium parasites survival in mosquitoes. Our results on Tachykinin-related peptide transcription were also inconclusive.


Background
The Insect central nervous system (CNS), formed by the brain, thoracic and ventral segmental ganglia, control feeding behavior, muscle activity, and other relevant functions, that in mosquitoes are implicated in vectorial competence [1]. Neuropeptides are synthesized by the CNS neurons as pre-pro-neuropeptide. These include transmitter peptides and peptide hormones involved in neuron communication, neuroendocrine regulation, and many physiological and behavioral processes [2][3][4], such as reproduction, metamorphosis, growth and metabolism [5]. Pre-pro-peptides consist of an N-terminal signal peptide, one to several potentially active biopeptides anked by cleavage sites in basic residues like lysine-arginine and a conserved C-terminal amidated motif [6]. C-terminal alpha-amidation promotes peptide stability and bioactivity [7][8][9]. Some neuropeptides are secreted into the hemolymph and regulate several functions in various organs of the insect [10][11][12][13].
Anopheline mosquitoes transmit malaria. Although the genome sequence of more than 20 anopheline species are available [39], only the neuropeptidome of Anopheles gambiae has been described [40]. This neuropeptidome comprises at least 35 genes coding for tachykininrelated peptides, pyrokinins (Pk/PBAN, CAPA), myosuppresins, adipokinetic hormones, insulin-like peptides (ILP), sulfakinins, among others [40]. This information contributed to the understanding of the functional role of neuropeptides in diverse physiological and developmental processes of this mosquito [32,[41][42][43]. In addition, the role of ILP in mosquito growth and reproduction has been documented in An. stephensi, [40]. Various factors, including Plasmodium infection, alter the expression of these neuropeptides, indicating their participation in other mosquito functions [44]. However, the information about neuropeptides in other anophelines remains limited.
The subgenus Anopheles Nyssorhynchus was recently raised to the genus status Nyssorhynchus [45,46]. Nyssorhynchus albimanus is an important malaria vector in México, Central America, and northern South America [47]. In this work, we refer our results to Nyssorhynchus albimanus, but keep Anopheles albimanus, as registered, when referring to genome and proteome data sets. Some neuropeptides of Ny. albimanus have been described in the brain, thoracic, and abdominal tissues [48]; however, a comprehensive description of its neuropeptidome and the effect of Plasmodium infection is still lacking.
Novel approaches, such as genome/transcriptome mining, have been employed for discovery and characterization of diverse neuropeptides, and several peptides have been predicted for a variety of species [49,50]. Knowledge of neuropeptides and their role in physiological processes involved in malaria transmission could contribute to the development of novel control strategies [40]. We present herein the description of a neuropeptidome of Ny. albimanus constructed through a brain transcriptome and available genomic data analysis. As well as the analyses of the expression of various Ny. albimanus neuropeptides during ookinete Plasmodium betghei infection. Our results provide insights into the interactions between neuroregulation and immune response in this mosquito, and open new ways to characterize the regulatory network underlying these interactions.

Material And Methods insect rearing
White stripe strain Ny. albimanus females [51] were obtained from the insectary of the National Institute of Public Health (INSP) in Cuernavaca, Mexico. Mosquitoes were bred under a 12:12 photoperiod at 28 °C and 70-80% relative humidity and 8% sucrose in cotton pads ad libitum. During the 72 h before infection with P. berghei, mosquitoes were provided with PSN 1 × (5,000 U of Penicillin, Streptomycin at 5 mg/ml, and Neomycin at 10 mg/ml) and gentamicin (50 µg/ml) (Thermo Fisher Scienti c, Waltham, Massachusetts, USA). Cotton pads were changed daily. This antibiotic treatment eliminates almost all bacteria in mosquitoes' midguts [52].

Plasmodium berghei infection
Four days post-emergence mosquitoes were fed with ookinetes of P. berghei ANKA strain expressing the Green Fluorescent Protein (GFP) [53] (kindly donated by Robert E. Sinden, Imperial College, UK). Ookinetes were produced by culturing gametocyte-infected mouse blood as described previously [54]. Groups of 300 female mosquitoes were fed for 1 h using arti cial membrane feeders with: (i) mouse blood + approximately 800 per µl GFP P. berghei ookinetes (infected group), or (ii) uninfected mouse blood (control group). Unfed mosquitoes were removed, and the engorged ones were incubated at 21 °C to allow for parasite invasion and interaction with the mosquito midgut. At 24 h post-blood feeding, mosquito midguts were analyzed under a 40 X uorescence microscope (Leica DM1000) to con rm the presence (P. berghei-infected group) or absence (control group) of parasites. We observed ookinetes and retorts forms. Only midguts containing more than 300 parasites were included in the infected group. Only brains from infected mosquitoes with con rmed infection from the infected group and all brains from the control group were collected.

Neurotranscriptome preparation and sequencing
Mosquitoes were cold anesthetized for 10 min at 4 °C and maintained on ice. Three-hundred brains were obtained from each mosquito group (N = 300). Dissected tissues were kept in 200 µl of Trizol Reagent (Invitrogen Waltham, Massachusetts, USA) and stored at -70 °C until processing.
Total RNA from pooled P. berghei-infected and control mosquito brains was obtained using Trizol Reagent following the manufacturer's instructions (Thermo Scienti c). The RNA clean up kit (Zymoclean, Irvine, CA, USA) was used to eliminate the possible contamination with mosquito eye pigment. Total RNA concentration, integrity, and yield were determined using Agilent's 2100 Bioanalyzer with the RNA 6000 Pico kit, according to the manufacturer's instructions (Agilent Technologies, Santa Clara, CA, USA).
Full-length cDNA libraries were synthesized using the Mint-2 cDNA synthesis kit (Evrogen, Moscow, Russia), according to the manufacturer's instructions. Brie y, 1 µg of RNA from each sample group (previously digested with DNase I (Invitrogen) to remove contaminating DNA) was used for rst-strand cDNA synthesis with a dT oligo (CDS-GsuI : 5'-AAGCAGTGGTATCAACGCAGAGTACTGGAG(T)20VN-3') and a Plug oligo adapter (5'-AAGCAGTGGTATCAACGCAGAGTGGCCATTACGGCCGGGGG-3') (Evrogen). The rst-strand cDNA was used for second-strand cDNA synthesis by PCR ampli cation with the M1primer (5'-AAGCAGTGGTATCAACGCAGAGT-3') (Evrogen). Later, 3 µg of each double-strand cDNA was digested with GsuI (15U) for 6 h at 30 °C. The cDNA libraries were prepared using the GS FLX Titanium Rapid Library Preparation kit according to the manufacturer's instructions (Roche). The cDNA libraries were sequenced in a full Pico titer plate using the Genome Sequencer FLX Titanium platform (454-Roche).
Data ltering, trimming, and mapping The output raw sequences were ltered according to length (> 100 bp), sequence complexity, and quality. Primer adaptors were trimmed using the SeqClean software [55]. The ltered reads were mapped to the Anopheles albimanus transcriptome [56] using GS Reference Mapper software v.2.5.3, with default parameters, and genome (Strain: STECLA, version Gene set: AalbS2.6.), using Exonerate v.2.2 (Slater GS and Birney E, 2005) with the EST2 genome mode, and a threshold score of 300, and a maximum intron length of 20,000 bp. Output bam les were used to verify mappings to the An. albimanus genome to identify putative transcribed genes that were not annotated. The count of reads by gene was done with the HTseq count v0.11.1 [57] software. The dataset used in this work is available at [58].

Transcriptome brain analysis
To characterize the Ny. albimanus brain transcripts that we obtained, we retrieve gene ontology (GO) annotations and performed comparisons between identi ed GOs using BLAST2GO v4.0.2. Additionally, we conducted a large-scale data mining with our Ny. albimanus brain transcripts as queries to retrieve a set of 1:1 orthologs in the An. gambiae, Aedes aegypti and Drosophila melanogaster genomes with the web interface BioMart in VectorBase [59] and the D. melanogaster genome in Flybase [60]. The Ny. albimanus ortholog set was used to identi ed genes with recognized expression in the brain of the aforementioned insects [60][61][62]. A GO type enrichment analysis for biological processes, molecular functions and cellular components was performed using the R TopGO library [63]. Transcripts were translated into six frames and analyzed using InterProScan against the InterPro database for both mosquito groups. Transcripts were translated into six frames and analyzed using InterProScan against the InterPro database [64]. Transcripts containing neuropeptides and or hormone domains were selected and validated with tBLASTx and BLASTp using a cut-off e-value of 1.0e-5. BLAST outputs were retrieved, listed and compiled by descending sequence identity percentage and score, and ascending e-value. Each of them was used to verify mappings to our Ny. albimanus brain transcriptome.

Neuropeptide identi cation
To identify putative neuropeptides in both Ny. albimanus groups (P. berghei-infected and control), we compiled a dataset (referencedataset) of FASTA sequences from the Database for Insect Neuropeptide Research (DINeR) [65], as well as previously published neuropeptides and neuropeptide receptors of An. gambiae [61] and Ae. aegypti [62], and neuropeptide and brain expressed genes of D. melanogaster [60]. The reference-dataset was used to perform multiple BLAST searches (BLASTn, tBLASTx and BLASTp) against the genome of An. albimanus using a cut-off e-value of 1.0e-5. BLAST outputs were retrieved, listed and compiled in the order of descending sequence identity percentage and score, and ascending e-value. The putative neuropeptides and neuropeptide receptors of Ny. albimanus were identi ed through BLAST [66], pfam [67], prosite [68], superfamily [69], smart [70], panther [71], gene3D [72], Conserved Domain Database (CDD) [73], prints [74] and ProDom [75]. Recognized domain signatures were visually inspected and compared against the genes of the reference-dataset to corroborate their architecture similarities. Also, we retrieve from VectorBase 1:1 ortholog of An. albimanus, An. gambiae, Ae. aegypti and D. melanogaster to investigate their neuropeptide orthologue relationships. The set of neuropeptides and neuropeptide receptors was annotated with BLAST2GO v4.0.2 [76]. The signal peptide prediction of putative neuropeptides was conducted with SignalP v5.0 [77]. The presence of neuropeptide precursors were detected with the software NeuroPID [78]; the prediction of cleavage sites and neuropeptides were analyzed using the NeuroPred programm [79] and to predict transmembrane helices of neuropeptide receptors we used TMHMM Server v.2.0 [80].

Expression of neuropeptides in Ny. albimanus mosquitoes
Based on the available genomic data, we generated speci c oligonucleotides for eighteen Ny. albimanus identi ed neuropeptides (Additional le 1: Table S1) using Oligo Analyzer v3.1 [81]. We conducted real-time PCR assays and ampli cation of each potential biopeptide region by triplicate using an Applied Biosystems(ABI) Step One Plus Real-Time PCR System. The qRT-PCR program was 95 °C for 10 min, 95 °C for 15 seconds and 64 °C for 1 min, repeated for 40 cycles; then 95 °C for 15 sec, 64 °C for 15 sec, and 95 °C for 15 sec, for one cycle. The speci city of the SYBR green PCR signal was con rmed by melting curve analysis and 1.5% agarose gel electrophoresis. Differential expression of neuropeptides in P. berghei-infected and uninfected mosquitoes The expression of adipokinetic hormone/corazonin-related peptide, tachykinin related peptide, SIFamide, myosuppressin, pk/PBAN, corazonin, ILP-2, ILP-3, and IL-5 were investigated in the brain of P. berghei-infected and control mosquitoes through real-time PCR assays. For every neuropeptide gene, three biological replicates were conducted for both conditions. As an internal control, a fragment of actin was ampli ed using primers RTActU R: (CGA TCC ACT TGC AGA GCC AGT) and RTAct3.2 F: (5´-TAC GCC AAC ATT GTC ATG TCC − 3 ) [82]. The real-time PCR program was conducted as described above. Generated qRT-PCR Ct values were analyzed using the 2-∆∆Ct method [83] and tested with one-way ANOVA, followed by a Kruskal-Wallis post-test (α = 0.05).

Brain transcriptome analysis:
Sequencing yielded 101,520 raw reads from control group and 109,383 from the P. berghei-infected group. The average read lengths in both groups were of 145 bp. Almost half of the total reads of infected and control group mapped to the An. albimanus genome (45.1%, and 45.9%, respectively). The reference mapping identi ed 3,811 transcripts, 30.03% of the total transcripts set currently registered in the transcriptome of An. albimanus (12,687 transcripts)(AalbS2.6), 947 transcripts were located in control, 1,220 transcripts in infected, and 1,633 were shared by both groups (Fig. 1). Of these, 2,131 had been previously annotated and 1,671 lack of available metadata (Additional le 2: TableS2). Most transcripts were associated to translation, oxidation-reduction process, transmembrane transport and proteolysis (biological process), protein binding, ATP binding, structural constituent of ribosomes, nucleic acid binding (molecular function); integral component of membrane, ribosome, nucleus, cytoplasm and intracellular (cellular component). Molecular functions were the most representative sequences identi ed in the GO analysis. These included sequences associated to ATP synthesis coupled to proton transport (GO:0015986); SRP-dependent co-translational protein targeting to membranes (GO:0006614); retrograde vesiclemediated transport, Golgi to endoplasmic reticulum biological processes (GO:0006890); cytoplasm (GO:0005737); mediator complex (GO:0016592); integrator complex (GO:0032039), cellular component; NADH dehydrogenase (ubiquinone) activity (GO:0008137); transcription co-regulator activity (GO:0003712) and ATP binding (GO:0005524). (Additional le 3: Table S3).
The comparative in silico analysis using neuropeptide transcripts sequences of An. gambiae, Ae. aegypti and D, melanogaster yielded ten neuropeptides in addition to the above mentioned: partner of bursicon, insulin-like peptide-2, insulin-like peptide-5, ecdysis triggering hormone, eclosion hormone, crustacean cardioactive peptide, allatotropin, allatostatin A, allatostatin C and diuretic hormone 44. To con rm these identities, eighteen nucleotide sequences corresponding to these pre-pro-peptides were matched against An. gambiae genome in Vectorbase. Matching sequences were used to design oligonucleotides for ampli cation by RT-PCR of P. berghei-infected and uninfected mosquito brain samples (Additional le 5. Figure S1).
For brevity, we describe only neuropeptides with increased expression in the brains of the infected group compared to those of the control group. We also describe a myosuppressin, identi ed within the sequence of transcript AALB009255-RA. Details of the sequences of the other transcripts are presented in the supplementary material (Additional le 6 Figure S2).

Structural characterization of the myosuppressin gene:
The polypeptide AALB009255-PA is a predicted polypeptide of 1,166 amino acid residues long (Fig. 3A) and is coded by fourteen exons and thirteen introns. It is predicted to contain C2 domains (calcium-dependent membrane-targeting module involved in signal transduction or membrane tra cking), which are absent in myosuppressins [84]. However, Tblastn searches, mapping, and insect transcriptome data [85], previously published myosuppressins sequences [86,87] and a close sequence analysis, identi ed an open reading frame between the rst two exons and the second intron of AALB009255-RA (Fig. 3B). This codes for an identical peptide of An. gambiae, D. melanogaster, M. domestica and Ae. aegypti myosuppressins (Fig. 3F).

Adipokinetic hormone/corazonin-related peptide (ACP)
One transcript-encoding a precursor of ACP was identi ed (AALB008450-RA). This transcript encodes for a 106-residue product with a predicted signal peptide of 31 amino acid residues in the N-terminus, followed by a predicted neuropeptide-coding sequence for 11 amino acid residues (QVTFSRDWNAG). It has a G residue and KR cleavage sites in the C-terminus (Fig. 4A). ACP is identical to the ACP of D. melanogaster [88] and An. darlingi, An. gambiae, Ae. aegypti and C. pipiens [89] (Fig. 4B). Pk/PBAN One transcript encoding a precursor of pk/PBAN was identi ed (AALB008609-RA) with 199 amino acid residues. It has a predicted signal peptide of 22 amino acid residues in the N-terminus, and ve predicted neuropeptide-coding sequences (Fig. 5A) [90]. One such sequence codes for a peptide of 11 amino acid residues (AAAMWFGPRLG) that is identical to the An. darlingi, An. gambiae and Ae. aegypti peptides. The second one codes for a peptide with 14 amino acid residues (PQPLFYHTAAPRLG), identical to that of An. darlingi; The third one codes for a peptide with 16 amino acid residues (DSVGENHQRPPFAPRLG), identical to peptides of An. darlingi and An. gambiae; and the fourth codes for a peptide with nine amino acid residues (NLPFSPRLG), identical to peptides of An. darlingi, An. gambiae and Ae. aegypti. All these have a G residue and PRL-conserved motif in their C-terminus, and are anked by KR, RK, RR and R cleavage sites (Figs. 5A and 5B). The fth sequence codes for a peptide with 12 amino acid residues (EDDSGLEGNGVS), which is anked by RR and KR cleavage sites, indicating that this peptide may be processed, but does not have the conserved motif and G residue in its C-terminus (Fig. 5A, highlighted in gray).

Corazonin
One transcript encoding a precursor of corazonin was identi ed (AALB010867-RA). It has 156 amino acid residues, whit a predicted signal peptide of 20 amino acid residues in the N-terminus, followed by one predicted neuropeptide-coding sequence with 12 amino acid residues (QTFQYSRGWTNG). It has a G residue and KR cleavage sites in the C-terminus (Fig. 6A). This precursor is identical to the corazonin of, An. gambiae, D. melanogaster, Ae. aegypti, M. domestica, B. mori, D. pulex and N. vitripennis (Fig. 6B) [91]. Neuropeptide brain expression inP. berghei experimental infection.

Discussion
In this work, we report the expression of critical neuropeptides in Ny. albimanus, with increased expression during infection with a malaria parasite. Of the 29 putative neuropeptide transcripts, that predicted at least 60 possible biopeptides, 27 had coding sequence (CDS).
In addition, we describe transcripts mainly associated with translation, oxidation-reduction process, protein binding, ATP binding, integral components of membrane and ribosomes. Previously, we reported 19 brain proteins of this mosquito with differential expression after P. berghei infection, and we identi ed 17 transcripts of proteins previously reported [92] (Additional le 7. Figure S3). All transcripts reported herein have orthologues in other insect disease vectors [4,16,50,[93][94][95] and in relevant agricultural insect pest [5,96,97]. Although most of the encoded peptides are identical or similar to orthologs in other species, some of them exhibit amino acid sequence variations. Alike to other mosquito species (An. darlingi, An. gambiae, Ae. aegypti and Culex quinquefasciatus) [98]. Our results in Ny. albimanus indicate that the allatostatin-A transcript has the potential to generate seven biopeptides; similar to the allatostatin-A transcripts of D. melanogaster and M. domestica that can generate six and ve biopeptides, respectively. Whereas the tachykinin-related peptide transcript of Ny. albimanus is identical to that of Ny. darlingi and has the potential for generating four biopeptides, in Ae. aegypti this is limited to only one biopeptide. Other organisms, such as Procambarus clarkii (Decapoda: Cambaridae) and Phenacogrammus interruptus (Characiformes; Alestiidae) can generate three biopeptides. Insulin-like peptides presented the most remarkable amino acid sequence variation. This results could be related to the long size of these neuropeptides and the diversity of its functions [99][100][101].
The Ny. albimanus myosuppressin sequence is identical to that of other insects, including several anophelines orthologs [85] (Fig. 3F). A close analysis of the An. albimanus AALB009255-RA, transcript revealed that the rst two exons, coding for a signal peptide, are highly similar to that of An. gambiae myosuppressin precursor (Fig. 2B), but the downstream sequence lacked myosuppressin homology (Fig. 3A). Full reconstruction of the myosuppressin precursor, including the rst two exons of AALB009255-RA and the novel exon predicted a full ORF with the canonical structure of the myosuppressin precursor gene. This observation indicates that AALB009255-RA annotation was mischaracterized, since it codes for two or more different transcripts, one of which includes a myosuppressin (Fig. 3B-E). The predicted myosuppressin sequence was identi ed in the second intron of the genomic sequence ( Fig. 3B and E) and the Locus_31294_Length_691 of the annotated An. albimanus transcriptome version 2 [56] (Fig. 3C). This locus has an ORF of 97 amino acid residues; the rst 75 amino acid residues are identical to those of the peptide AALB009255-PA. This transcript encodes a signal peptide of 20 amino acid residues ( Fig. 3C and D, highlighted in red) and a predicted neuropeptide-coding sequence of 11 amino acid residues corresponding to myosuppressin (TDVDHVFLRFG) (Fig. 3D, highlighted in green).
Despite the importance of neuropeptides in adult insects, information about their function is limited, especially for anophelines. Neuropeptides are associated with most of the physiological processes in insects, including the immune response activation [102][103][104][105] and suppression [106] after infection by different microorganisms. Interestingly, we detected a signi cant increase in the transcription of ACP, pk/PBAN, and corazonin in Ny. albimanus infected with P. berghei.
The study of adipokinetic hormone/corazonin-related peptide (ACP) [88], initially characterized as adipokinetic hormone II (AKH II) [107][108][109], structurally intermediate between corazonin and adipokinetic hormone, has provided new evidence and questions on the role of neuropeptides in insects. The structural and sequential analysis described this neuropeptide in several insects, including anophelines. ACP is primarily expressed in the nervous system and, to a lesser extent, in other insect organs and tissues [110,111]. ACP transcripts were detected in the head and thorax of larvae, pupae, and adult of Ae. aegypti and An. gambiae [109]. Furthermore, ACP transcripts expression increases prominently in the brain and thoracic ganglia of Ae. aegypti after adult eclosion, suggest this neuropeptide may function in the regulation of post-ecdysis activities [112].
ACP transcripts in the brain of Ny. albimanus is consistent with previous results in An. gambiae where ACP (called AKHII) is expressed 72 h after feeding (blood or sugar) [109]. The differential expression of ACP between uninfected blood-fed and P. berghei-infected Ny. albimanus suggests that the midgut invasion by this parasite activates or modi es physiological pathways dependent on ACP. [14,15,17,24]. Thus, it is possible that the increased level of this neuropeptide in Plasmodium infected mosquitoes is part of the immune response activation against this parasite. Further studies are necessary to unravel the function of the ACP system; currently, no de nitive function for ACP has been determined and, functional studies in other insects revealed that ACP does not perform AKH and corazonin functions [112][113][114].
Pyrokinins have been identi ed in various insects [95,110,[115][116][117][118] and are generated from the capa and pk/pban genes. Capa produces at least two peptides with sequence XXFPRV and XXWFGPRL, while pk/pban produces at least one peptide with sequence XXWFGPRL. Two genes pk/pban-like (capa and hugin) encode pyrokinins PK1 (XXXXWFGPRL) and PK2 (XXXXXXRPPFAPRL) respectively in D. melanogaster. Several functions for these peptides were described, including stimulation of pheromone biosynthesis [119], induction of melanization [120], induction of embryonic diapause [121], stimulation of visceral muscle contraction [122] and termination of pupal diapause development [123]. But no evidence exists for the participation of pyrokinins in immune response mechanisms. Here, we show that a pk/pban and capa transcripts coding for PK, PVK-1, and PVK-2 peptides (Table 2) increase in Plasmodium-infected Ny. albimanus; suggesting that the pyrokinins PK-1, PK-2, and PK-3 could be involved in signaling or activation mechanisms of the immune response of mosquitoes to these parasites.
Corazonin is widely conserved across insect genera [124] with various functions including cuticular melanization [32]. Its increased expression in brains of P. berghei-infected Ny. albimanus could be related to these two functions, but other not yet identi ed roles of this neuropeptide in mosquito defenses await identi cation.
Insulin-like peptides are involved in several processes that are conserved among vertebrates and invertebrates, including immunity [125][126][127]. ILPs have been identi ed in Ae. aegypti [128], An. gambiae [129], and An. stephensi (42). In An. stephensi, the ingestion of P. falciparum-infected blood, increased ILPs expression (42), and the inhibition of this induction reduced parasite development, indicating the possibility that ILP induction is a parasite mechanism to avoid elimination by the mosquito.t [106]. However, while inhibition of ILP4 induced the expression of immune genes prior to parasite invasion of the mosquito midgut, the inhibition of ILP3 increased immune gene expression at 24 hours after infection, when parasites had already invaded the midgut, indicating that the relationship between ILP mediated mechanism and immunity could be multiple. Our results documented an increase in of ILP-5 transcription in Ny. albimanus at times when P. berghei parasites had initiated development on the mosquito midgut. Although this increase was consistent among mosquito samples, it was not statically signi cant. On the other hand, no difference in ILP 3 transcription was observed between infected and uninfected mosquitoes. These results require further studies with a bigger sample size, but they may indicate that ILPs also participate in P. berghei infection, and that the induction of ILP is a generalized mechanism of Plasmodium parasites survival in mosquitoes. Our results on Tachykinin-related peptide transcription were also inconclusive.
The neuropeptide receptors identi ed in this work, have characteristic domains, almost all are G-protein coupled receptors rhodopshin like (GPCRs) ( Table 1). However, the functional characterization of these receptors to understand the neuropeptide-receptor interaction and the role they play in the mosquito-parasite interaction awaits investigation.
In summary, we provide a brain neuropeptidome repertoire composed by 29 transcripts coding for at least 60 potential biopeptides in Ny. albimanus. Particularly we report the increased expression of the adipokinetic hormone/corazonin-related peptide (ACP), pk/pban pirokinin, and corazonin transcripts after P. berghei infection. At present, the functions of neuropeptides in insects during infection with parasites or viruses are partially understood. Further investigation is required as to whether signi cant changes in ACP and pk/PBAN transcription in the brain re ect an increase in the production or release of the respective biopeptides.
Most of the neuropeptides identi ed here showed high similarity with previously reported in other insects and other mosquito vectors. However, it is still necessary to explore the mechanisms triggered by neuropeptides up-regulated in the Ny. albimanus brain by a Plasmodium infection.

Conclusions
Most of the neuropeptides identi ed here showed high similarity with previously reported in other insects and other mosquito vectors. Our results indicate that P. berghei promoted a modi cation of transcripts neuropeptides expression in the mosquito brain at 24 hours postinfection. The pattern of differential expression in adipokinetic hormone/corazonin-related peptide, pk/PBAN, and corazonin indicates that the invasion of midgut tissue by Plasmodium triggered a brain response. However, it is still necessary to explore the mechanisms activated by neuropeptides up-regulated in the Ny. albimanus brain by a Plasmodium infection. Whether this change is due to stress or immune response remains unclear since the function of these neuropeptides is practically unknown in anophelines. Nevertheless, these ndings provide insights on the behavior and immune response of Anopheles during a Plasmodium invasion and contribute with initial leads for the understanding of its neuroregulation. Abbreviations NPLP1 neuropeptide-like precursor 1; sNPF:short neuropeptide F; CCAP; crustacean cardioactive peptide; PBURS:partner of bursicon; ETH:ecdysis triggering hormone, Dh44:diuretic hormone 44, ILP:insulin-like peptides; AKH:adipokinetic hormone; EH:eclosion hormone; PDH:pigment dispersing hormone; GPA2:glycoprotein 2; qPCR:quantitative polymerase chain reaction; CNS:central nervous system; K:lysine; R:arginine; GFP:green uorescent protein; GO:Gene ontology; BLAST:Basic Local Alignment Search Tool; CDS:Coding sequence.
Declarations design of tables and gures, analysis and interpretation results. MCR: Parasites culture. ATL: mosquito infection, mosquito tissue extraction. HLM: Experimental design and manuscript writing. All authors read and approved the nal version of the manuscript.

Competing interests
The authors declare that they have no competing interests.