Relaxin

The sratoolkit (https://trace.ncbi.nlm.nih.gov/Traces/sra/sra.cgi?view=software) in combination with Trinity (Grabherr et al., 2011 (link)) was used in the search for transcripts encoding peptides that might be somewhat similar to insulin in insect gonad transcriptome short read archives (SRAs). The method consisted of using the tblastn_vdb command from the sratoolkit to recover individual reads from transcriptome SRAs that show possible sequence homology with insulin-like molecules. Since insulin-like peptides have highly variable sequences the command is run with the -seg no and -evalue 100 options. Reads that are identified are then collected using the vdb-dump command from the sratoolkit. The total number of reads recovered is much smaller than those typically present in an SRA and this allows one to use Trinity on a normal desktop computer to make a mini-transcriptome of those reads. This transcriptome is than searched using the BLAST+ program (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE_TYPE=BlastDocsDOC_TYPE=Download) for possible insulin transcripts. This first round usually yields numerous false positives and perhaps a few partial transcripts that look interesting. These promising but partial transcripts are then used as query using the blastn_vdb command from the sratoolkit on the same SRAs and reads are collected anew and Trinity is used to make another transcriptome that is again queried for the presence of insulin-like transcripts. In order to obtain complete transcript the blastn_vdb search may need to be repeated several times. Alternatively genes coding such transcripts were identified in genome assemblies using the BLAST+ program and Artemis (Rutherford et al., 2000 (link)). Once such transcripts had been found, it was often possible to find orthologs from related species. For example, once the honeybee gonadulin was found, it was much easier to find it in other Hymenoptera. The same methods were used to identify relaxin and C-terminally extended ilps, which have much better conserved primary amino acid sequences and consequently are more easily identified, as well as their putative receptors. Whenever possible all sequences were confirmed in both genome assemblies and in transcriptome SRAs. In many cases transcripts for the various ilps and receptors were already present in genbank, although they were not always correctly identified. All these sequences are listed in Spreadsheet S1.
Expression was estimated by counting how many RNAseq reads in each SRA contained coding sequence for each of the genes. In order to avoid untranslated sequences of the complete transcripts, that sometimes share homologous stretches with transcripts from other genes and can cause false positives, only the coding sequences were used as query in the blastn_vdb command from the sratoolkit. This yielded the blue numbers in Spreadsheet S2. In order to more easily compare the different SRAs these numbers were then expressed as per million spots in each particular SRA. These are the bold black numbers in Spreadsheet S2.
For the expression of alternative aIGF (arthropod insulin-like growth factor) splice forms reads for each splice variant were first separately identified. Unique identifiers in these two sets were determined to obtain the total number of reads for aIGF. Those identifiers that were present in the initial counts for both splice forms were counted separately and subtracted from the initial counts of the two splice variants to obtain the number of reads specific for each isoform.
The various SRAs that were used are listed in the supplementary pdf file and were downloaded from https://www.ncbi.nlm.nih.gov/sra/. The following genome assemblies were also analyzed: Aedes aegypti (Matthews et al., 2018 (link)), Blattella germanica (Harrison et al., 2018 (link)), Bombyx mori (Kawamoto et al., 2019 (link)), Galleria melonella (Lange et al., 2018 (link)), Glossina morsitans (Attardo et al., 2019 (link)), Hermetia illucens (Zhan et al., 2020 (link)), Latrodectus hesperus (https://www.ncbi.nlm.nih.gov/genome/?term=Latrodectus+hesperus), Mesobuthus martensii (Cao et al., 2013 (link)), Oncopeltus fasciatus (Panfilio et al., 2019 (link)), Parasteatoda tepidariorum (Schwager et al., 2017 (link)), Pardosa pseudoannulata (Yu et al., 2019 (link)), Periplaneta americana (Li et al., 2018 (link)), Stegodyphus dumicola (Liu et al., 2019 (link)), Tetranychus urticae (Grbic et al., 2011 (link)), Timema cristinae (Riesch et al., 2017 (link)), Tribolium castaneum (Herndon et al., 2020 (link)) and Zootermopsis nevadensis (Terrapon et al., 2014 (link)). All genomes were downloaded from https://www.ncbi.nlm.nih.gov/genome/.

Human embryonic kidney (HEK-293T) cells stably transfected with RXFP1 were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, 100 μg mL⁻¹ penicillin, 100 μg mL⁻¹ streptomycin, and 2 mM of l-glutamine and plated into 96-well plates pre-coated with poly-l-lysine for whole-cell binding assays (Merck, Darmstadt, Germany). Competition binding experiments using Eu³⁺-labeled H2 relaxin in the absence or presence of increasing concentrations of unlabeled H2 relaxin B-chain derivatives were conducted as previously described [13 (link)]. All data are presented as the mean ± S.E.M. of the % specific binding of triplicate wells, repeated in at least three separate experiments, and curves were fitted using one-site binding curves in GraphPad Prism 9 (GraphPad Inc., San Diego, CA, USA). Statistical differences in pKi values were analyzed using a one-way analysis of variance with uncorrected Fisher’s least significant difference (LSD) post hoc analysis in GraphPad Prism 9.

The ability of peptide analogs to activate cAMP in HEK-RXFP1 cells stably transfected with a pCRE β-gal reporter gene construct was tested as described in detail previously [13 (link)]. Cells were stimulated with increasing concentrations of peptide analogs in parallel to H2 relaxin (positive) or media (negative) controls, then incubated at 37 °C for 6 h, after which the media were aspirated, and the cells were frozen at −80 °C overnight. The following day, cAMP-driven β-gal expression was determined in cell lysates as described [32 (link)]. Experiments were performed in triplicate at least 3 times, and data were fitted to a four-parameter sigmoidal dose–response curve using GraphPad Prism 9 to determine ligand potency (pEC50).

Fluctuations in female sex hormones during pregnancy have also been linked to changes in IOP [7 (link), 10 (link), 14 (link), 28 (link)]. A meta-analysis by Wang et al. identified 15 studies that reported a significant reduction in IOP during the second and third trimesters of pregnancy compared with a non-gravid control group [7 (link)]. Subsequently, several additional studies have reported results in concordance with this meta-analysis. In a study comparing 165 pregnant with 105 non-pregnant women, the mean IOP was statistically lower in the pregnant cohort (13.2 ± 2.2 mmHg vs. 14.2 ± 2.7 mmHg, p = 0.001) [15 (link)]. In another study which included 235 women monitored across each of the trimesters of pregnancy and into the puerperal period, Tolunay et al. observed a significant, steady decline in IOP when comparing the first with the second and third trimesters of pregnancy, and a rebound in the postpartum period to levels indistinguishable from where they started in the first trimester of pregnancy [14 (link)]. Similarly, in a study by Efe et al. that performed a longitudinal evaluation of IOP in 25 women over the course of pregnancy found that IOPs decreased by an average of 9.5%, with the highest values in the first trimester and the lowest in the third trimester (13.8 ± 2.1 vs. 12.4 ± 2.08 mmHg, p < 0.001) [10 (link)]. IOP regressed to a level similar to that obtained in the first-trimester by 3-month postpartum. Finally, Naderan et al. reported a decrease in Goldmann and corneal-compensated IOP measurements during pregnancy in women with keratoconus, which persisted six months after pregnancy [28 (link)]. Bear in mind, the authors describe keratoconus to be a disorder of corneal elasticity, resulting in corneal thinning and thus suggesting a lower-pressure phenomenon.
As discussed by Wang et al., a widely proposed mechanism to account for the decrease in IOP observed during pregnancy hypothesizes a relationship between female sex hormones and an increase in aqueous humor outflow capacity, similarly to other body systems where venous capacity expands during pregnancy. Peak levels of progesterone, estrogen, and relaxin at the end of pregnancy correlate inversely with IOP [8 (link)]. Possible mechanisms to explain this change include the dilatory effects of progesterone and estrogen, resulting in decreased arterial pressure, subsequent reduction in humor production, and increased humor outflow facilitated by reduced episcleral venous pressure [7 (link), 14 (link)]. It is well described that estradiol increases nitric oxide-induced relaxation of the trabecular meshwork [14 (link), 16 (link)], while progesterone competes with endogenous glucocorticoid receptor sites, thus inhibiting the ocular hypertensive effect of glucocorticoid [7 (link), 14 (link), 15 (link), 28 (link)]. Another proposed mechanism attributes the decrease in IOP to increased relaxin levels during pregnancy, softening ligaments and thus reducing corneoscleral rigidity [7 (link), 10 (link), 14 (link)].