Identification, presence, and possible multifunctional regulatory role of invertebrate gonadotropin-releasing hormone/corazonin molecule in the great pond snail (Lymnaea stagnalis)

In the last years, the interpretation of gonadotropin-releasing hormone (GnRH) neuropeptide superfamily has changed tremendously. One main driver is the investigation of functions and evolutionary lineage of previously identified molluscan GnRH molecules. Emerging evidence suggests not only reproductive, but also diverse biological effects of these molecules and proposes they should most likely be called corazonin (CRZ). Clearly, a more global understanding necessitates further exploration of species-specific functions and structure of invGnRH/CRZ peptides. Towards this goal, we have identified the full-length cDNA of invGnRH/CRZ peptide in an invertebrate model species, the great pond snail Lymnaea stagnalis, termed ly-GnRH/CRZ, and characterized the transcript and peptide distribution in the central nervous system (CNS) and peripheral organs. Our results are consistent with previous data that molluscan GnRHs are more related to CRZs and serve diverse functions. For this, our findings support the notion that peptides originally termed molluscan GnRH are multifunctional modulators and that nomenclature change should be taken into consideration.

stagnalis possesses unique features such as a well-defined central nervous system (CNS), clearly-defined behaviors, and accessible anatomy that can facilitate the characterization of invGnRH/CRZ. To accomplish our goal, we first identified the coding sequence of prepro-ly-GnRH/CRZ. Subsequently, reverse-transcription (RT) polymerase chain reaction (PCR) and immunohistochemistry (IHC) were performed to examine the anatomical distribution and cellular localization of ly-GnRH/CRZ transcript and peptide, respectively. Our data revealed the presence of an invGnRH/CRZ preprohormone in L. stagnalis. Further, the mRNA and peptide distribution patterns were consistent with the role of ly-GnRH/CRZ as a neuropeptide with diverse functions.

Experimental animals
Adult (5-6 months old) specimens of the pond snail, L. stagnalis, originating from our laboratory-bred stocks, were randomly selected for use in experiments. Snails were kept in large holding tanks containing oxygenated low-copper artificial snail water at a constant temperature of 20°C (±1.5 °C) and on light:dark regime of 12 h:12 h. Snails were fed lettuce ad libitum three times a week. All procedures on snails were performed according to the protocols approved by the Scientific Committee of Animal Experimentation of the Balaton Limnological Institute (VE-I-001/01890-10/2013). Efforts were made to minimize the number of animals used in the experiments.

In silico searches, cDNA sequencing, and bioinformatics analysis
In silico searches of L. stagnalis CNS transcriptome and genome were conducted with TBLASTN (NCBI). First, the coding sequence of ap-GnRH/CRZ (GnRH/CRZ of Aplysia californica; NM_001204553.1) was used as the search query and identified a 272 bp long L. stagnalis cDNA fragment in the genome consortium data (referred to as Lsta_scaffold164; genome consortium publication in preparation), indicating the presence of GnRH/CRZ in L.

stagnalis.
For RNA sample collection, the whole CNS was dissected from the snails (n=10) and homogenized using a TissueLyser LT (QIAGEN) in TRI reagent (#93289, Sigma-Aldrich). Consensus sequence was called and manually corrected also within CLC Genomics Workbench. The identified ly-GnRH/CRZ mRNA sequence was submitted to the GenBank Nucleotide database with the accession number MN385595.

Phylogenetic analysis
The alignment used to generate the maximum likelihood tree consisted of 56 amino acid (AA) sequences of GnRH superfamily preprohormones retrieved from GenBank and the deduced AA sequence of ly-GnRH/CRZ preprohormone (see Supplementary information). The alignment was performed using ClustalW with BLOSUM62 substitution matrix in Molecular Evolutionary Genetics Analysis v7 software (Kumar et al., 2016). Alignment was then analyzed to get the best fitting model, which was determined to be LG with gammadistributed rates. Bootstrapping support for the tree was conducted with 1000 bootstrap replicates.

RT-PCR analysis of ly-GnRH/CRZ expression in central and peripheral tissues
Total RNA from the CNS (buccal, cerebral, pedal, pleural, parietal, and visceral ganglia) and peripheral organs (heart, and ovotestis) was isolated as presented above. After the RNA preparation, an additional DNase treatment was performed using TURBO DNA-free™ Kit (#AM1907, Thermo Fisher Scientific) following manufacturer's instructions.
The RevertAid H Minus First Strand cDNA Synthesis Kit (#K1631, Thermo Fisher Scientific) was used for RT, applying random hexamer primers and 3 µL volume from the total RNA sample.
PCR primers were designed for the identified coding region of ly-GnRH/CRZ using SnapGene ® Viewer software (GSL Biotech, Chicago, IL, version 4.1.7). The applied primer set was as follows: the forward primer for ly-GnRH: 5'-CCT TGT CCT CCT GGC TGT AGT G -3'; the reverse primer for ly-GnRH: 5'-GAT GTG GCC GGT GTA CGA TGG -3' (Integrated DNA Technologies, Leuven, Belgium). As a control to ensure the quality of RT, actin was also amplified for all examined tissues. The applied primer set for actin was as follows: the forward primer: 5'-TCC CTT GAG AAG AGC TAC GAG C -3'; and the reverse primer: 5'-GAG TTG TAG GTG GTT TCG TGG -3' (Integrated DNA Technologies, Leuven, Belgium).
The PCR reaction was performed in 20 μ L reaction volume at 95 °C for 4 min (initial denaturation) followed by 35 cycles of 95 °C for 30 sec (denaturation), 55 °C for 30 sec (annealing) and 72 °C for 10 sec (extension) by a T1 Thermocycler PCR device (Biometra ® , Goettingen, Germany). PCR product was checked by agarose gel-electrophoresis using 2% gel (SeaKem LE agarose) and GeneRuler TM 100 bp Plus DNA ladder (#SM0328, Thermo Fisher Scientific). (1 µg/mL in DPBS; #62249, Thermo Scientific) for 10 min at room temperature. The slides were washed with DPBS and cover-slipped with fluorescent mounting medium (#S3023, Dako). The stained tissues were analyzed with a TCS SP8 DMI laser confocal scanning microscope (Leica Microsystems, Germany) equipped with appropriate wavelength-filter configuration settings and a transmitted light detector (BioMarker Ltd, Hungary). The necessary number of optical line sections (15-30) with 0.5-0.8 μ m step-size were made to capture of all visualized details. Image processing was performed by LasX software (Leica Microsystems, Germany). Because the primary antiserum was generated against ap-GnRH/CRZ conjugated with hemocyanin (KLH) from giant keyhole limpet (Megathura crenulata), to abolish cross-reaction to hemocyanin in tissues of L. stagnalis the antiserum was preadsorbed overnight against KLH (#H7017, Sigma). In some cases, the antibody was preadsorbed with 10 µM of synthetic ly-GnRH/CRZ (pQNYHFSNGWYA-amide) overnight at 4°C and applied to the sections.

Nucleotide sequence of ly-GnRH/CRZ preprohormone
Our in silico search in the L. stagnalis sequence data revealed a 272-bp positive hit (referred to as Lsta_scaffold164) containing a sequence of 32 nucleotides with a 100% match to the query, thus identifying the potential GnRH/CRZ coding sequence. We termed this sequence ly-GnRH/CRZ. Using nucleotide sequencing, we determined the complete coding sequence.
The full-length ly-GnRH/CRZ preprohormone spans 1225 bp (Fig. 1). Based on the data from scientific literature (Iwakoshi et al., 2002;Tsai and Zhang, 2008;Zhang et al., 2008), we identified the signal peptide, the mature peptide with the signature tribasic cleavage site and an alpha amidation signal (QNYHFSNGWYAGKKR), and the ly-GnRH/CRZ-associated peptide ( Fig. 1). As expected, the mature peptide sequence is highly conserved, while the signal peptide and GnRH/CRZ-associated peptide show high variability.

Phylogenetic analysis
To update the established evolutionary relationships of the GnRH superfamily with newly annotated sequences including ly-GnRH/CRZ, we performed a phylogenetic analysis. The results of this are presented in Fig. 2, and the segregated clusters from top to bottom are as follows: 1. Two clusters of AKHs from mollusks, arthropods, annelids, and nematodes. Based on data from the literature, all of molluscan, arthropod and nematode AKHs are true AKH molecules. Five manual corrections are required: Crassostrea gigas AKH belongs to molluscan AKH group, Bombyx mori ACP to the ACP group,

Branchiostoma floridae GnRH to the GnRH group, while Capitella teleta and
Helobdella robusta (annelids) AKHs are not true AKH molecules but referred to as AKH/GnRHs. Furthermore, one worm peptide, Priapulus caudatus AKH, should belong to AKH cluster.
2. An arthropod ACP cluster. Two manual corrections should also be performed here; Brachionus calyciflorus AKH has the consensus sequence for arthropod AKH but not ACP cluster, and Cionia intestinalis GnRH belongs to GnRH cluster.

A GnRH/CRZ cluster, including the newly determined ly-GnRH/CRZ (arrowhead)
and previously identified molluscan GnRH sequences, possesses both GnRH and CRZ properties.

A vertebrate
GnRH cluster containing all used sequences from chordates and echinoderms.
Actin (a housekeeping gene) amplification of RNA samples without RT was used as a negative control to test for genomic DNA contamination (Fig. 3, middle panel). The lower panel amplified actin (140 bp) to ensure RNA and RT quality. Negative controls including no RT (middle panel) and no template (NTC) controls did not yield positive signal.

Localization of ly-GnRH/CRZ peptide in central and peripheral tissues
IHC revealed the presence of ly-GnRH/CRZ peptide in all CNS ganglia (Fig. 4A). Large-(50-80 µm), medium-(30-50 µm) and small-sized (10-30 µm) ly-GnRH/CRZ-immunoreactive (ir) neurons were observed. The distribution of ly-GnRH/CRZ-ir neurons was highly variable in different ganglia (Fig. 4A). In the pleural (PlG) and parietal (PaG) ganglia, only a few ly-GnRH/CRZ-ir neurons were detected, but labeled neurons were more abundant in the buccal (BG), cerebral (CG), and pedal ganglia (PeG) (Fig. 4A). The BG contained several smalland medium-sized ly-GnRH/CRZ-ir neurons largely concentrated around the unlabeled B1 and B2, furthermore, around the immunopositive large-sized B3 and B4 feeding motoneurons Among peripheral tissues, we have investigated the heart and the ovotestis. Confocal projections showed abundant ly-GnRH/CRZ-ir fibers on the surface of the heart atria running perpendicularly to the longitudinal axis of the muscle fibers; however, the heart muscle fibers themselves were not immunopositive ( Fig. 5A1 and A2). In ovotestis, immunoreactivity was not observed (Fig. 5B). The immunostaining was completely abolished when the antiserum was preadsorbed with 10 µM ly-GnRH/CRZ (not shown).

DISCUSSION
Previous studies suggested that invGnRH/CRZ molecules, initially named invGnRH, may not assume a specialized reproductive function (Jung et al., 2014;Minakata et al., 2009;Tsai et al., 2010) and may need to be reclassified as CRZ (Hauser and Grimmelikhuijzen, 2014;Plachetzki et al., 2016;Tsai, 2018). The main goal of the present study was to further test this notion by identifying and investigating the invGnRH/CRZ molecule in L. stagnalis. The presence of a GnRH-like molecule in L. stagnalis has previously been hypothesized (Koene, 2010), but remained unconfirmed until the current report. The identified ly-GnRH/CRZ preprohormone sequence contains the mature peptide with the signature tribasic cleavage site and an alpha amidation signal (QNYHFSNGWYAGKKR). However, the predicted mature peptide sequence awaits validation by mass spectrometry. Based on the literature (Nagasawa et al., 2015), we hypothesize that two forms of ly-GnRH/CRZ may be derived from a single prohormone sequence: a mature amidated undecapeptide (pQNYHFSNGWYA-amide) and a non-amidated precursor dodecapeptide (pQNYHFSNGWYAG-OH).
Our phylogenetic analysis presents two challenges commonly seen in peptide hormone analysis. Firstly, the relatively short GnRH superfamily prepropeptide sequences (58-219 AA) contribute to the lower bootstrap values. Secondly, the conserved mature peptide sequence needed for biological activity (e.g., receptor binding) can cause incorrect clustering (Hauser and Grimmelikhuijzen, 2014). Nevertheless, the phylogenetic tree correlated well with previous studies (Hauser and Grimmelikhuijzen, 2014;Plachetzki et al., 2016;Tsai, 2018). The GnRH/CRZ branching suggests sequences originally termed molluscan GnRHs are no more related to vertebrate GnRH than to AKH, ACP or CRZ. This is the first study to demonstrate the occurrence and distribution of a snail GnRH/CRZ in the CNS and periphery. Our RT-PCR results show that ly-GnRH/CRZ is ubiquitously present in all central and peripheral tissues examined. However, IHC data show that the simultaneous presence of ly-GnRH/CRZ transcript and peptide occurs predominantly in the CNS. For example, although the ly-GnRH/CRZ transcript can be found in the heart and ovotestis (Fig.   3), the peptide was not observed within the heart muscle proper and absent in the ovotestis cluster of the VG are also involved in heart control and respiration, the peptide may also play a role in these processes. Supporting a cardio-respiratory role of ly-GnRH/CRZ, ly-GnRH/CRZ-ir fibers are present on the surface of heart atria muscles ( Fig. 5A1 and A2). In this respect, it is interesting to note that CRZ molecule was first discovered and named for its ability to stimulate the heart of the American cockroach (Periplaneta americana) (Veenstra, 1989). This effect was also documented by the investigation of CRZ of the blood-sucking bug, Rhodnius prolixus (Patel et al., 2014). Furthermore, the first described invGnRH/CRZ molecule, the oct-GnRH/CRZ, was also identified in the course of bioassay screenings of peptide fractions with cardio-acceleratory effects (Iwakoshi et al., 2002).
In summary, we conclude that ly-GnRH/CRZ, just like ap-GnRH/CRZ, is a multifunctional neuropeptide involved in diverse central functions including feeding, locomotion, heart control, and reproduction. These finding can be confirmed in the future with physiological studies using purified or synthetic ly-GnRH/CRZ peptide. Our results further suggest the term "invGnRH" (e.g., oct-GnRH, ap-GnRH) should be revised to avoid the connotation that these peptides serve specialized reproductive function similar to vertebrate GnRHs (Iwakoshi-Ukena et al., 2004;Jung et al., 2014;Kavanaugh and Tsai, 2016;Minakata et al., 2009;Nagasawa et al., 2015;Sun and Tsai, 2011;Treen et al., 2012;Tsai et al., 2010). Overall, our study reports the presence and peptide distribution of a second invGnRH/CRZ in a gastropod mollusk and adds to a growing body of literature supporting a multifunctional role of invertebrate GnRH/CRZ peptides.

Acknowledgement
Bioinformatic infrastructure was supported by ELIXIR Hungary.     and cross-sectional (triangles) heart atria muscles (A1). A2 represents a higher magnification detail of the atria muscle. Immunolabeled axon processes (arrows) running perpendicular the longer, longitudinal axis of the muscle fibers (m). Scale bar=25 µm on A1 and 10 µm on A2.