A novel peptide Phylloseptin‐PBu from Phyllomedusa burmeisteri possesses insulinotropic activity via potassium channel and GLP‐1 receptor signalling

Abstract Insulin, as one of the most important hormones regulating energy metabolism, plays an essential role in maintaining glucose and lipid homeostasis in vivo. Failure or insufficiency of insulin secretion from pancreatic beta‐cells increases glucose and free fatty acid level in circulation and subsequently contributes to the emergence of hyperglycaemia and dyslipidaemia. Therefore, stimulating the insulin release benefits the treatment of type 2 diabetes and obesity significantly. Frog skin peptides have been extensively studied for their biological functions, among which, Phylloseptin peptides discovered in Phyllomedusinae frogs have been found to exert antimicrobial, antiproliferative and insulinotropic activities, while the mechanism associated with Phylloseptin‐induced insulin secretion remains elusive. In this study, we reported a novel peptide named Phylloseptin‐PBu, isolated and identified from Phyllomedusa burmeisteri, exhibited dose‐dependent insulinotropic property in rat pancreatic beta BRIN‐BD11 cells without altering cell membrane integrity. Further mechanism investigations revealed that Phylloseptin‐PBu‐induced insulin output is predominantly modulated by KATP‐[K+] channel depolarization triggered extracellular calcium influx and GLP‐1 receptor initiated PKA signalling activation. Overall, our study highlighted that this novel Phylloseptin‐PBu peptide has clear potential to be developed as a potent antidiabetic agent with established function‐traced mechanism and low risk of cytotoxicity.

Clinically, the therapies for the treatment of T2DM are aimed at decreasing hepatic glucose production, improving insulin resistance and increasing insulin release. In particular, for the insulin secretion boosting therapy, two types of drugs are currently prescribed, one acts on closing the K ATP channels on the betacells for the calcium influx, and the other type targets on mimicking or enhancing the effect of gut hormones (also named as incretins), 7 a good example of the latter type is extendin-4, it was firstly isolated and characterized from the Heloderma suspectum venom, the glycaemic-control action of extendin-4 is via the similar way with glucagon-like peptide 1 (GLP-1), which mainly attributes to the potentiation of insulin secretion in a glucose-sensing manner. [7][8][9][10][11][12][13][14] Bioactive peptides from amphibian secretions and venoms with characteristics of diversity and specificity provide candidates for the development of novel drugs for combatting multiple diseases, including antibiotic-resistant pathogens caused infections, cancers and T2DM. 15 Previous elegant research has reported that some frog skin-derived peptides possess the ability to stimulate insulin release both in vitro and in vivo at relatively low concentration and dose without causing cellular toxicity. [16][17][18] Many have been extensively studied in the fields of amino acid isoform replacement, post-translational modification and primary structure modification, all of these focus on increasing the insulin release efficiency, preventing the degradations caused by endogenous enzymes and improving the cytotoxicity. 16,19,20 Nevertheless, the mechanism associated with how these bioactive peptides are involved in the insulin release action has rarely been investigated.
Thus, in this study, for the first time, we reported a novel insulinotropic peptide isolated from Phyllomedusa burmeisteri, in which species such peptide has never been identified, belongs to Phylloseptin family, named as Phylloseptin-PBu. At low concentration, it showed glucose-comparably insulin secretion stimulatory effects in pancreatic beta BRIN-BD11 cells. Cell signalling examinations revealed the involvements of K ATP -[K + ] channel-induced intracellular calcium elevation and partial GLP-1 receptor activation-induced PKA stimulation, which is firstly reported. Therefore, we not only provide a novel candidate for developing antidiabetic agent, but also elucidate the preliminary mechanism of its bioactivity, illuminating the future work for promoting this agent into clinical tests.

| Acquisition of frog skin secretions
The specimens of Burmeister's leaf frog, Phyllomedusa burmeisteri (n = 3, 6 and 8 cm snout-to-vent lengths), were sampled in the Atlantic Forest biome of Brazil. The skin secretions were obtained by mild transdermal electrical stimulation (5 V, 50 Hz, 4 ms pulse width). 21 The skin secretions were then collected by rinsing with distilled deionized water and subjected to snap frozen with liquid nitrogen, lyophilized and stored at À20°C prior to use.
2.2 | "Shotgun" cloning of a novel Phylloseptin-like peptide from skin secretion-derived cDNA library Five milligrams of lyophilized Phyllomedusa burmeisteri secretion powder was dissolved in 1 mL cell lysis/binding buffer to isolate polyadenylated mRNA by magnetic oligo-dT beads via using Dynabeads â mRNA DIRECT TM Kit (Dynal Biotech, Liverpool, UK). The reverse-transcription products were subjected to 3 0 -RACE PCR procedures to acquire the full length of preproprotein nucleic acid sequences using a SMART-RACE kit (Clontech, Palo alto, CA, USA).
Briefly, a nested universal primer (NUP) (supplied with the kit) and a degenerate primer (S1; 5 0 -ACTTTCYGAWTTRYAAGMCCAAABAT which was designed on the basis of the 5 0 -untranslated region of phylloxin cDNA from Phyllomedusa bicolor. 22    | 2805 molecular cloning, the solid-phase Fmoc strategy was adopted via Tribute TM automated solid-phase peptide synthesizer 4 (Protein Technologies, Tucson, AZ, USA) to synthesize the novel peptide as previously described. 23 The synthesized peptide replicates were purified by reverse phase HPLC with C18 column (Jupiter C18, 5 lm particle, 300 A pore, 250 9 10 mm, Phenomenex, UK), and both molecular masses and high-purity of the synthetic replicates were confirmed via MALDI-TOF.  All cell culture designs for the ELISA measurement were repeated with three independent studies, and every study contains three replicates for each treatment.

| Cell membrane integrity, haemolytic and antimicrobial evaluations
The lactate dehydrogenase (LDH) assay was employed to assess the cell membrane integrity after Phylloseptin-PBu treatment. The rate of LDH release from BRIN-BD11 cells was measured using Pierce LDH Cytotoxicity Assay Kit (Thermo scientific, USA) according to the manufacturer's instruction with the cell culture supernatant collected from acute insulin release assay.
Peptide concentrations range from 10 À7 to 10 À12 mol/L was prepared for assessing the haemolytic and antimicrobial properties of Phylloseptin-Pbu, and the detailed procedures were described previously. 28 2.9 | Western blotting BRIN-BD11 cells were seeded into 100-mm tissue culture dishes at a density of 2 9 10 6 cells/dish for 24 hours prior to 6-hours serumfree starvation, Phylloseptin-PBu (10 À7 mol/L) was then added for 16-hours incubation. The cells were rinsed twice with ice-cold PBS,

| Statistical analysis
All results are presented as mean AE SEM determined by two-tailed Student's t test or one-way ANOVA. Pairs comparisons of the means were made, and P < .05 was regarded as a significant difference.
The Bonferroni method was used to adjust the observed significance levels for the fact the multiple contrasts were being tested.   Phylloseptin-PBu peptide replicates were then synthesized by standard solid-phase Fmoc chemistry using Tribute TM automated solid-phase peptide synthesizer 4 (Protein Technologies, Tucson, AZ, USA). Following cleavage from the synthesis resin, impurities were removed from the synthetic replicates by rp-HPLC and the molecular masses of the purified major products were confirmed by MALDI-TOF mass spectrometry ( Figure S1).

| Secondary structure determination by circular dichroism analysis
The CD spectra ( Figure 2) indicated that Phylloseptin-PBu displayed diverse secondary structure conformations in different surroundings.
In aqueous solution, a mixed conformation of random coil (42%) and b-sheet (51%) was presented with a negative band at 200 nm. A similar conformation was observed in 50% methanol-water solution with a slightly higher presence of random coil. By contrast, a typical a-helical conformation was shown in 50% TFE-water solution with one positive band at 192 nm and two negative bands at 206 nm and 222 nm, respectively, and the predicted a-helical contents of Phylloseptin-PBu were 96% in 50%TFE-water solution. The percentage of secondary structure helicity of Phylloseptin-PBu in respective solutions are shown in Table 1.

BRIN-BD11 cells
To investigate whether Phylloseptin-PBu could stimulate insulin release, we treated rat pancreatic beta BRIN-BD11 cells with a concentration range from 10 À7 to 10 À12 mol/L, along with a 5.6 mmol/L glucose group, which mimics the glucose level in physiological conditions, a 16.7 mmol/L glucose group, which is a stimulatory glucose level (suggesting diabetic ketoacidosis), and a positive alanine group.
The results ( Figure 3A) showed that, upon Phylloseptin-PBu stimulation, BRIN-BD11 cells exhibited a significant higher, and a dosedependent, insulin release capacity compared with 5.6 mmol/L glucose group, this effect was even stronger than 16.7 mmol/L glucose group when administered the 10 À7 or 10 À8 mol/L concentration of the peptide. Although the insulinotropic effect of Phylloseptin-PBu is weaker compared to L-alanine, which mainly induces insulin output via Ca 2+ influx mechanism, 29 it still prompted us to explore the signal pathways that involved in this novel Phylloseptin-PBu peptide exhibited bioactivity.
As Phylloseptin-like peptides have been reported to exert antimicrobial activities against Gram-positive and Gram-negative bacterial and fungi by disrupting the cell membrane. 30 We therefore investigated whether Phylloseptin-PBu-triggered insulin release was caused by the damage to cell membrane. The minimal inhibitory concentration (MIC) assay implied that Phylloseptin-PBu possessed no antimicrobial activity against S. aureus, E. coli and C. albicans when the concentration was up to 10 À7 mol/L ( Figure S2). The lactate dehydrogenase (LDH) tests suggested that Phylloseptin-PBu treatment did not compromise cell membrane integrity in BRIN-BD11 cells at concentrations up to 10 À7 mol/L ( Figure 3B). Further haemolytic measurement using horse blood cell also did not show cytolytic effect of Phylloseptin-PBu at 10 À7 mol/L ( Figure 3C). Therefore, these results lead us to investigate the specific mechanism that related to insulin release effect of Phylloseptin-PBu.

| Potassium and calcium channel-modulated Ca 2+ influx examination
As indicated, elevated intracellular ATP level and increased ATP/ADP ratio leads to the closure of potassium channel, induction of cell membrane depolarization and opening of voltage-gated calcium channel.
These actions promote extracellular Ca 2+ influx, which plays a key role in mediating L-alanine-activated insulin release from BRIN-BD11 cells. 31 We employed diazoxide (activator of ATP-sensitive potassium channels), verapamil (inhibitor of voltage-gated calcium channels) and

| 2809
Ca 2+ -free medium to examine whether Phylloseptin-PBu stimulated insulin release is via Ca 2+ mobilization. As shown in Figure 4, preadministration of diazoxide and verapamil significantly attenuated Phylloseptin-PBu-induced insulin production, deprivation of Ca 2+ significantly attenuated the insulin producing activity in all tested groups, but significant enhancement of insulin release was still observed after Phylloseptin-PBu treatment even culturing the cells with Ca 2+ -free KRB buffer with or without diazoxide or verapamil.
Together, these results suggested that calcium influx, mediated by ATP-sensitive potassium channel induced depolarization and calcium channel opening, played the dominant role in the insulinotropic activity of Phylloseptin-PBu, nevertheless, this effect was not completely abrogated, it prompted us to hypothesize that other signalling might be involved.

| PKA signalling activation-modulated insulin secretion examination
Activation of cAMP/PKA was also reported to elicit insulin secretion in islet. 32 Thus, we interrogated the PKA signalling using the PKA activator (IBMX, 3-Isobutyl-1-methylxanthine) and inhibitor (H89) in the presence or absence of Phylloseptin-PBu. The results ( Figure 5A) showed that although H89 significantly blunted Phylloseptin-PBuinduced insulin secretion, compared with singular H89 treatment, Phylloseptin-PBu still has a trend to increase insulin release. While impair Phylloseptin-PBu's action. As the results shown in Figure 6, the insulin secretion effect was significantly decreased in both exendin-(9-39) and exendin-(9-39) with Phylloseptin-PBu co-incubation groups; nonetheless, the latter group also showed a prominent higher insulin output. While further consistently activating K ATP channels with diazoxide or blocking the calcium channels with verapamil showed an additive insulin secretion inhibition effect. In the meantime, the calcium deprivation patterns, again, exhibited significant lower insulin production, of note is that under such conditions, Phylloseptin-PBu still stimulated higher insulin production. As a whole, these results confirmed that Phylloseptin-PBu could interact with GLP-1 receptor as a partial agonist to active PKA signalling for modulating Ca 2+ channels, promoting insulin secretion. However, a Ca 2+ and PKA signalling-independent mechanism induced by Phylloseptin-PBu was likely to play a role in this action.

| Other mechanism examination for the induction of insulin release by Phylloseptin-PBu
In order to assess whether other signalling is existed in regulating Phylloseptin-PBu's insulinotropic activity, we treated the cells using diazoxide or verapamil together with H89, and measured the insulin release. As the Figure 7 shown, compared with singular H89 treatment, Phylloseptin-PBu still has moderately, but significantly higher insulin release activity even co-treated with H89 and/or with diazoxide/verapamil, but this effect was completely abolished in the Ca 2+free medium settings, which proposed a K ATP -and PKA-independent Ca 2+ influx mechanism.

| DISCUSSION
In the postprandial state, elevated blood glucose is sensed by pancreatic beta-cells which in response secrete insulin to maintain glucose homoeostasis. This process occurs under both physiological and pathological conditions, and thus, increasing insulin release has promising therapeutic potential. 5 The mechanism correlated with acute insulin secretion in islets has been well-studied, which is commonly summarized as two distinct pathways, one is related to the K ATP channel closure, and membrane depolarization-triggered intracellular calcium rise via Ca 2+ channels on both cell membrane and endoplasmic reticulum, the other one is through Ca 2+ -induced insulin secretory vesicle exocytosis. 31 In this study, we investigated the K ATP -channel, calcium We also confirmed that Phylloseptin-PBu partially stimulates GLP-1 receptor, resulting in PKA activation for the increased insulin secretion. Notwithstanding, it must be noticed that, in Figure 5, Ca 2+free medium remarkably decreased insulin output even under the condition of blocking calcium channel using verapamil, suggested a novel cytosol Ca 2+ import mechanism, which is further supported by the independent results of Figure 7, and we suppose that ionotropic receptors (also known as ligand-gated ion channels) might play the crucial role in it, which warrants further investigations.
Moreover, based on Figure 5 and part Figure 6 results, the novel peptide still enhanced insulin release significantly compared with no Phylloseptin-PBu treatment groups, which implied an intracellular Ca 2+ store (endoplasmic reticulum)-referred mechanism induced by Phylloseptin-PBu. In addition, another interesting issue is that, in Figure 5A, we observed the additive augment insulin release effect of Phylloseptin-PBu with IBMX, we presume that IBMX, as a phosphodiesterase inhibitor to raise intracellular cAMP, 36 has distinct PKA activation mechanism with Phylloseptin-PBu, which involves activating of GLP-1 receptor, and therefore, the hyperactive PKA by both Phylloseptin-PBu and IBMX contributed to the dramatically increased insulin secretion.
In consideration of the secondary structure of Phylloseptin-PBu, which could also be crucial for its biological relevance, we conducted the CD analysis, the results showed that in 50% TFE aqueous solution, this novel peptide formed a well-defined a-helical structure, while in polar solvent, much less of (7%) a-helicity was presented.
This result was similar to glucose-dependent insulinotropic polypeptide (GIP), GLP-1, extendin-4 and glucagon, 37-40 likewise, we assume that the full length of a-helical conformation might be important for Phylloseptin-PBu to exert the insulin induction activity. Nonetheless, further NMR or other structure-activity relationship studies are needed for determination.
In addition, we also measured the insulinotropic activity of Phylloseptin-PBu in 16.7 mmol/L glucose conditions to test whether it could further enhance insulin release, the results showed that no higher insulin release was observed ( Figure S3). We speculate that the high glucose level might induce the cells to be stressful, and in such condition, the capacity of insulin release from cells could be limited by the high baseline insulin release.
Overall, in this project, we isolated and identified a novel peptide from Phyllomedusa burmeisteri, the insulinotropic effect of this novel peptide Phylloseptin-PBu has been substantially studied, we confirmed the energy-mediated K ATP and Ca 2+ channels, and the GLP-1 receptor initiated PKA activation involvements in this effect, which has never been published from amphibian sources, our study continuously provide novel candidate and fundamental understanding for the treatment of T2DM in future.

ACKNOWLEDG EMENTS
This work was supported by the internal funding for Natural Drug Discovery Group from School of Pharmacy, Queen's University Belfast.

CONFLI CT OF INTEREST
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.