Insulinotropic activity of the host-defense peptide frenatin 2D: Conformational, structure-function and mechanistic studies

Accepted Manuscript Insulinotropic activity of the host-defense peptide frenatin 2D: conformational, structure-function and mechanistic studies Vishal Musale, Laure Guilhaudis, Yasser H.A. Abdel-Wahab, Peter R. Flatt, J. Michael Conlon PII:

S0300-9084(18)30268-2

DOI:

10.1016/j.biochi.2018.09.008

Reference:

BIOCHI 5512

To appear in:

Biochimie

Received Date: 20 July 2018 Accepted Date: 15 September 2018

Please cite this article as: V. Musale, L. Guilhaudis, Y.H.A. Abdel-Wahab, P.R. Flatt, J.M. Conlon, Insulinotropic activity of the host-defense peptide frenatin 2D: conformational, structure-function and mechanistic studies, Biochimie (2018), doi: https://doi.org/10.1016/j.biochi.2018.09.008. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Of four naturally occurring frenatin peptides tested, frenatin 2D (DLLGTLGNLPLPFI.NH2) from Discoglossus sardus was the most potent and effective in producing concentration-dependent stimulation of insulin release from BRIN-BD11 rat clonal β-cells without displaying cytotoxicity. The peptide also stimulated insulin release from 1.1B4 human-derived clonal β-cells and isolated mouse islets and improved glucose tolerance concomitant with increased circulating insulin concentrations in mice following intraperitoneal administration. The insulinotropic activity of frenatin 2D was not associated with membrane depolarization or an increase in intracellular [Ca2+] but incubation of the peptide (1µM) with BRIN-BD11 cells produced a modest, but significant (P < 0.05), increase in cAMP production. Stimulation of insulin release was abolished in protein kinase A-downregulated cells but maintained in protein kinase Cdownregulated cells. Circular dichroism studies showed that, in the presence of dodecylphosphocholine micelles, frenatin 2D exhibited a helical content of 35% and a turn content of 28%. Substitution of the Thr5, Asn8, Pro10, and Ile14 residues in frenatin2D by Trp and interchange of Pro12 and Phe13 led to loss of insulinotropic activity but the [D1W] and [G7W] analogues were as potent and effective as the native peptide. Frenatin 2D (1µM) also stimulated proliferation of BRIN-BD11 cells and provided significant protection of the cells against cytokine-induced apoptosis. It is concluded that the insulinotropic activity of frenatin 2D is mediated predominantly, if not exclusively, by the KATP channel-independent pathway.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Insulinotropic activity of the host-defense peptide frenatin 2D: conformational, structure-function and mechanistic studies

Vishal Musalea, Laure Guilhaudisb, Yasser H. A. Abdel-Wahaba, Peter R. Flatta,

a

RI PT

J. Michael Conlona*

SAAD Centre for Pharmacy and Diabetes, School of Biomedical Sciences, Ulster University,

M AN U

b

SC

Coleraine BT52 1SA, U.K.

Normandy University, COBRA, UMR 6014 & FR 3038, Université de Rouen, INSA Rouen, CNRS, 1 rue Tesnière 76821 Mont St Aignan, Cedex, France

*

TE D

Corresponding author. J. Michael Conlon, School of Biomedical Sciences, University of

AC C

EP

Ulster, Coleraine, BT52 1SA, Northern Ireland, UK. E-mail: [email protected]

1

ACCEPTED MANUSCRIPT ABSTRACT

Of four naturally occurring frenatin peptides tested, frenatin 2D (DLLGTLGNLPLPFI.NH2) from Discoglossus sardus was the most potent and effective in producing concentration-

RI PT

dependent stimulation of insulin release from BRIN-BD11 rat clonal β-cells without displaying cytotoxicity. The peptide also stimulated insulin release from 1.1B4 human-

derived clonal β-cells and isolated mouse islets and improved glucose tolerance concomitant

SC

with increased circulating insulin concentrations in mice following intraperitoneal

administration. The insulinotropic activity of frenatin 2D was not associated with membrane

M AN U

depolarization or an increase in intracellular [Ca2+] but incubation of the peptide (1µM) with BRIN-BD11 cells produced a modest, but significant (P < 0.05), increase in cAMP production. Stimulation of insulin release was abolished in protein kinase A-downregulated cells but maintained in protein kinase C-downregulated cells. Circular dichroism studies

TE D

showed that, in the presence of dodecylphosphocholine micelles, frenatin 2D exhibited a helical content of 35% and a turn content of 28%. Substitution of the Thr5, Asn8, Pro10, and Ile14 residues in frenatin-2D by Trp and interchange of Pro12 and Phe13 led to loss of

EP

insulinotropic activity but the [D1W] and [G7W] analogues were as potent and effective as the native peptide. Frenatin 2D (1µM) also stimulated proliferation of BRIN-BD11 cells and

AC C

provided significant protection of the cells against cytokine-induced apoptosis. It is concluded that the insulinotropic activity of frenatin 2D is mediated predominantly, if not exclusively, by the KATP channel-independent pathway.

2

ACCEPTED MANUSCRIPT Abbreviations:

CCK-8, Cholecystokinin-8 DPC, Dodecylphosphocholine

RI PT

GLP-1, Glucagon-like peptide 1 IBMX, 3-isobutyl-1-methylxanthine LDH, Lactate dehydrogenase

SC

MALDI-TOF, Matrix-assisted laser desorption/ionization-time of flight

PKC, Protein kinase C PMA, Phorbol 12-myristate 13-acetate T2DM, Type 2 diabetes mellitus TFE, Trifluoroethanol

AC C

EP

TE D

TNF-α, Tumour necrosis factor-α

M AN U

PKA, Protein kinase A

3

ACCEPTED MANUSCRIPT 1.

Introduction

The dramatic global increase in the incidence Type 2 diabetes mellitus (T2DM) in the past decade has necessitated the search for new naturally occurring therapeutic agents that

RI PT

regulate glucose concentrations and prevent the complications associated with the disease. The discovery of exendin-4 in the venom of a reptile, the Gila monster Heloderma

suspectum [1] highlights the importance of non-mammalian sources in the search for new

SC

antidiabetic peptides. Exendin-4 is an agonist at the glucagon-like peptide-1 (GLP-1)

receptor (GLP1R) that stimulates glucose-dependent insulin release and improves pancreatic

M AN U

β-cell function [2]. Exendin-4 is more potent and longer-acting in vivo than GLP-1 and is used in routine clinical practice in T2DM therapy [3]. The skin secretions of many frog species contain bioactive peptides that play an important contributory role in protecting the host from invasion by pathogenic microorganism in the environment and ingestion by

TE D

predators [4]. These peptides are multifunctional and may possess antimicrobial, antifungal, antiviral, anticancer and immunomodulatory activities (reviewed in [5,6]). Additionally, several host defence peptides that were first identified on the basis of their antimicrobial

EP

properties have been shown to stimulate insulin release from clonal β-cells and isolated

AC C

pancreatic islets and improve glucose tolerance following intraperitoneal administration in mice and so represent agents with therapeutic potential for treatment of patients with T2DM (reviewed in [7]).

The frenatins are a family of structurally related small peptides that were first

identified in skin secretions of the Australian treefrog Litoria infrafrenata (reclassified as Nyctimystes infrafrenatus) (Pelodryadidae) [8] Subsequently, frenatin 2D [DLLGTLGNLPLPFI.NH2] was isolated from skin secretions of the Tyrrhenian painted frog Discoglossus sardus (Alytidae) [9] and frenatin 2.1S [GLVGTLLGHIGKAILG.NH2],

4

ACCEPTED MANUSCRIPT frenatin 2.2S [GLVGTLLGHIGKAILS.NH2] and frenatin 2.3S [GLVGTLLGHIGKAILG] from the Orinoco lime frog Sphaenorhynchus lacteus (Hylidae) [10]. Frenatin 2D lacks antimicrobial activity but stimulated production of the proinflammatory cytokines TNF-α, and IL-1β by mouse peritoneal macrophages suggesting that the peptide may act on

RI PT

macrophages in frog skin to produce a cytokine-mediated stimulation of the adaptive

immune system in response to invasion by microorganisms [9]. In contrast, frenatin 2.1S and 2.2S show potent antimicrobial activity against Gram-negative bacteria and are cytotoxic to

SC

non-small cell lung adenocarcinoma A549 cells [10]. Frenatin 2.1S also stimulates production of pro-inflammatory cytokines by mouse peritoneal macrophages and

M AN U

downregulates production of the anti-inflammatory cytokine IL-10 by lipopolysaccharide stimulated cells. A single injection of frenatin 2.1S (100 µg) in BALB/c mice enhances the activation state and homing capacity of Th1 type lymphocytes [11] and led to a marked increase in the number and tumoricidal capacity of activated peritoneal natural killer (NK)

TE D

cells [12] in the peritoneal cavity suggesting that the peptide should be regarded as a candidate for antitumor immunotherapy. Activity against yellow fever virus has also been reported for the frenatin 2 peptides present in S. lacteus skin secretions [13].

EP

The aim of the present study was to investigate the therapeutic potential of frenatins

AC C

2D, 2.1S. 2.2S, and 2.3S for development into agents for the treatment of patients with T2DM. Their ability to stimulate insulin release in vitro was assessed using BRIN-BD11 rat clonal β-cells [14], 1.1 B4 human clonal β-cells [15] and isolated mouse islets and their ability to lower blood glucose concentration and stimulate insulin release in vivo was determined in overnight-fasted, male NIH Swiss TO mice. In addition to their beneficial effects on β-cell function, certain frog skin peptides, such as PGLa-AM1 from the African clawed frog Xenopus amieti [16], temporin A from the European common frog Rana temporaria [17], and peptide OA-A1 from the Asian frog Odorrana andersonii [18] have

5

ACCEPTED MANUSCRIPT been shown to mimic the effect of GLP-1 on promoting β-cell proliferation and survival. Consequently, the effects of frenatin 2D on β- cell proliferation and its ability to inhibit

Material and methods

2.1.

Peptide synthesis and purification

SC

2.

RI PT

cytokine-induced apoptosis was studied in BRIN-BD11 cells.

Frenatin 2D and its tryptophan-containing analogues were supplied in crude form by

M AN U

Synpeptide Ltd (Shanghai, China). Frenatin 2.1S, 2.2S, and 2.3S were supplied in crude form by GL Biochem Ltd (Shanghai, China). The peptides were purified to near homogeneity (>98 % purity) by reversed-phase HPLC as previously described [9,10]. The identity of all peptides was confirmed by MALDI-TOF mass spectrometry using a Voyager

Insulin release studies using clonal β-cells

EP

2.2.

TE D

DE PRO instrument (Applied Biosystems, Foster City, USA).

AC C

The culture of BRIN-BD11 rat clonal β-cells and 1.1B4 human-derived pancreatic βcells and the method for measuring the effects of peptides on the release of insulin have been described in detail in previous publications [16,19,20]. Incubations with the frenatin peptides (10-12 - 3 x 10-6 M; n = 8) were carried out for 20 min at 37 ˚C in Krebs-Ringer bicarbonate (KRB) buffer, pH 7.4 supplemented with 5.6 mM glucose. Control incubations were carried out in the presence of GLP-1 (10 nM), exendin-4 (10 nM) and alanine (10 mM). After incubation, aliquots of cell supernatant were removed for measurement of insulin concentrations by radioimmunoassay [21]. In order to determine cytotoxicity, the effects of

6

ACCEPTED MANUSCRIPT the frenatin peptides (10-7 M - 3 x 10-6 M; n = 4) on the rate of lactate dehydrogenase (LDH) release from BRIN-BD11 cells were measured using a CytoTox 96 non-radioactive cytotoxicity assay kit (Promega, Southampton, UK) according to the manufacturer’s instructions.

RI PT

In a second series of experiments designed to investigate mechanisms of action,

incubations of BRIN-BD11 cells with frenatin 2D (1 µM) were carried out in the presence of known modulators of insulin release [22]: the K+ channel activator diazoxide (300 µM), the

SC

L-type voltage-dependent Ca2+ channel blocker verapamil (50 µM), and a depolarizing

stimulus KCl (30 mM). To determine the role of extracellular calcium in mediating the

M AN U

insulinotropic activity of the peptide, cells were pre-incubated in calcium-free KRB buffer (pH 7.4) supplemented with 1.1 mM glucose and 1 mM EGTA for 1 h at 37 ˚C. After preincubation, cells were incubated for 20 min at 37 ˚C with frenatin 2D (1 µM) in calcium-free

2.3.

TE D

KRB buffer containing 5.6 mM glucose.

Insulin release studies using isolated mouse islets

EP

The preparation of isolated pancreatic islets from adult, male National Institutes of

AC C

Health NIH Swiss mice (Harlan Ltd, Bicester, UK) and the procedure for determining the effects of peptides on the rate of insulin release has been described previously [16,23]. The islets were incubated for 1 h at 37 °C with synthetic peptides (10-8 and 10-6 M) in KRB buffer supplemented with 16.7 mM glucose. Supernatants were removed for determination of insulin concentration by radioimmunoassay. The islet cells were retrieved and extracted with acid-ethanol to determine total insulin content as previously described [19].

7

ACCEPTED MANUSCRIPT 2.4.

Effect of frenatin 2D on membrane potential and intracellular calcium concentrations

Changes in membrane potential and intracellular calcium concentrations in response to incubation with frenatin 2D (1 µM) were determined fluorimetrically with monolayers of

RI PT

BRIN-BD11 cells using a FLIPR Membrane Potential Assay Kit and a FLIPR Calcium 5 Assay Kit (Molecular Devices, Sunnyvale, CA, USA) according to the manufacturer’s recommended protocols as previously described [24]. Cells were incubated in 5.6 mM

SC

glucose with frenatin 2D at 37 °C for 5 min and data were acquired using a FlexStation scanning fluorimeter with integrated fluid transfer workstation (Molecular Devices).

M AN U

Control incubations with 5.6 mM glucose alone, 5.6 mM glucose + 30 mM KCl, and 5.6 mM glucose + 10 mM alanine) were also carried out.

Effects of frenatin-2D on cyclic AMP production

TE D

2.5.

The procedure for determining the effects of 1 µM frenatin 2D on the production of cAMP in BRIN-BD11 cells has been described previously [19]. Incubations were carried out

EP

for 20 min in KRB buffer supplemented with 5.6 mM glucose and the phosphodiesterase

AC C

inhibitor, 3-isobutyl-1-methylxanthine (IBMX; 200µM). cAMP concentrations in the cell lysate were measured using a Parameter kit (R & D Systems, Abingdon, UK) following the manufacturer’s recommended protocol. Control incubations in the presence of 5.6 mM glucose alone and GLP-1 (10 nM) were also carried out.

8

ACCEPTED MANUSCRIPT 2.6.

Effects of down-regulation of the PKA and PKC pathways on insulin release

In order to investigate further the mechanism of insulinotropic action of frenatin 2D, BRIN-BD11 cells were incubated for 18 h at 37 °C in an atmosphere of 5 % CO2 and 95 %

RI PT

air with 25 µM forskolin (Sigma-Aldrich, UK) to downregulate the protein kinase A (PKA) pathway or with 10 nM phorbol 12-myristate 13-acetate (PMA; Sigma-Aldrich, UK) to

downregulate the protein kinase (PKC) pathway or with 25 µM forskolin + 10 nM PMA to

SC

downregulate both pathways. Details of the experimental procedure have been described previously [18]. Cells were preincubated for 40 min at 37 °C with KRB buffer, pH 7.4

M AN U

supplemented with 1.1 mM glucose and 0.1 % bovine serum albumin followed by a 20 min incubation with (a) frenatin 2D (1 µM), (b) GLP-1 (10 nM) and (c) CCK-8 (10 nM) in KRB buffer supplemented with 5. 6 mM glucose. Control incubations with forskolin alone (25 µM), PMA (10 nM) alone and forskolin (25 µM) + PMA (10 nM) alone were also carried

TE D

out. Aliquots of the cell supernatants were removed for measurement of insulin concentrations by radioimmunoassay

EP

Effects of frenatin 2D on cytokine-induced apoptosis and proliferation

AC C

2.7.

The ability of frenatin 2D to protect against cytokine-induced DNA damage was

analysed by incubating BRIN-BD11 cells, seeded at a density of 5 x 104 cells per well, for 18 h at 37 °C with a cytokine mixture (200 U/ml tumor-necrosis factor-a, 20 U/ml interferon-γ and 100 U/ml interleukin-1β), in the presence and absence of either frenatin 2D (1 µM) or GLP-1 (1 µM). Cells were rinsed with 0.9% phosphate-buffered saline (PBS) and fixed using 4 % paraformaldehyde. The cells were permeabilized with 0.1 M sodium citrate buffer, pH 6.0 at 94 °C for 20 min. To determine effects on apoptosis, the cells were incubated with

9

ACCEPTED MANUSCRIPT TUNEL reaction mixture (In situ Cell Death Detection Kit; Roche Diagnostics, Burgess Hill, UK) for 1 h at 37 °C following the manufacturer’s recommended procedure [16]. Slides were viewed using a fluorescent microscope with 488 nm filter (Olympus System Microscope, model BX51; Southend-on-Sea, UK) and photographed by a DP70 camera adapter system.

RI PT

In order to study effects on β-cell proliferation, BRIN-BD11 cells were incubated with either frenatin 2D (1 µM) or GLP-1 (1 µM) for 18 h at 37 0C as previously described [16]. The cells were fixed and permeabilized as above followed by treatment with rabbit anti-Ki-67

SC

primary antibody and subsequently with Alexa Fluor 594 secondary antibody, which stains proliferating cells in red (Abcam. Cambridge, UK). Proliferation frequency was determined

replicate were analysed.

In vivo insulin release studies

TE D

2.8.

M AN U

in a blinded fashion and expressed as % of total cells analysed. Approximately 150 cells per

All animal experiments were carried out in accordance with the UK Animals (Scientific Procedures) Act 1986 and EU Directive 2010/63EU for animal experiments and

EP

approved by Ulster University Animal Ethics Review Committee. All necessary steps were

AC C

taken to prevent any potential animal suffering. Eight-week-old male NIH Swiss TO mice (Harlan Ltd, Bicester, UK), were housed separately and maintained in an air-conditioned room (22 ± 2 °C) with a 12 h light: 12 h dark cycle. The procedure for determining the effects of intraperitoneal administration glucose alone (18 mmol/kg body weight) and in combination with frenatin 2D (75 nmol/kg body weight) or GLP-1 (25 nmol/kg body weight) has been described previously [16,25]. Blood samples were collected at the time points shown in Fig. 7. Blood glucose concentrations were measured using an Ascencia Contour

10

ACCEPTED MANUSCRIPT Blood Glucose Meter (Bayer, Newbury, UK) and plasma insulin concentrations by radioimmunoassay.

Circular dichroism (CD) studies

RI PT

2.9.

Spectra were recorded on a MOS-500 Circular Dichroism Spectrometer (Bio-Logic, Seyssinet Pariset, France). Data points were collected from 260 to 185 nm, with an integration

SC

time of 2 s per point and a step size of 1 nm, using a 1.0 mm path length rectangular quartz cell. Measurements were carried out at room temperature. Each CD spectrum represents the

M AN U

average of three scans which were compared before summation to detect possible alterations of the sample. The resulting spectrum was corrected by subtraction of a spectrum obtained for a solution lacking peptide but otherwise identical. The ellipticity is reported as mean residue molar ellipticity ([θ]MRE deg cm2 dmol−1). Peptides (0.085 mg/mL) were dissolved in

TE D

water or in 2,2,2-trifluoroethanol (TFE) aqueous solutions (25% v/v) or in methanol (MeOH) aqueous solutions ((25% v/v) or in 20 mM dodecylphosphocholine (DPC) aqueous solution.

the peptides.

EP

Higher concentrations of organic solvents could not be tested due to the limited solubility of

AC C

Peptide secondary structure was estimated using the online CD spectra deconvolution server Dichroweb [26,27]and the PEPFIT program [28]. For Dichroweb analysis, the secondary structure content was estimated by averaging the results given by CONTINLL and CDSSTR [29] deconvolution programs. α-helical content was also calculated by using the Forood formula [30].

11

ACCEPTED MANUSCRIPT 2.10.

Statistical Analysis

Data were compared using unpaired Student’s t-test (non-parametric, with two-tailed P values and 95% confidence interval) and one-way ANOVA with Bonferroni post-hoc test

RI PT

wherever applicable. Area under the curve (AUC) analysis was carried out using the

trapezoidal rule with baseline correction. Values are presented as mean ± SEM. Results are

3.1.

Results

M AN U

3.

SC

considered significant if p < 0.05.

Effects of frenatins on insulin release from BRIN-BD11 rat clonal β-cells

The rate of insulin release from BRIN-BD11 cells in the presence of 5.6 mM glucose

TE D

alone was 1.01 ± 0.04 ng/106 cells/20 min. Incubation with the established insulin secretagogue alanine (10 mM) increased the rate of insulin release to 5.42 ± 0.30 ng/106 cells/20 min and to 3.08 ± 0.10 ng/106 cells/20 min with GLP-1 (10 nM). The effects of

EP

incubation of BRIN-BD11 cells with increasing concentrations of frenatin 2D, 2.1S, 2.2S and

AC C

2.3S) are shown in Fig. 1. Frenatin 2D was the most potent peptide with a threshold concentration (the concentration producing a significant increase in the rate of insulin release compared with the rate in the presence of 5.6 mM glucose alone) of 0.1 nM and the most effective (a 2.3-fold increase in rate at 3µM concentration). At concentrations up to 3 µM, no significant increase in the rate of release of the cytosolic enzyme LDH from the BRIN-BD 11 cells was observed for any of the frenatin peptides indicating that the integrity of plasma membrane remained intact (data not shown).

12

ACCEPTED MANUSCRIPT 3.2.

Effects of frenatin 2D on insulin release from 1.1B4 human clonal β-cells and isolated

mouse islets

As shown in Fig. 2A, the stimulatory effect of frenatin 2D on insulin release was

RI PT

replicated in the glucose-responsive 1.1B4 human-derived cell line. The peptide produced a significant (P < 0.05) increase in the rate of insulin release at a concentration of 0.1 nM and an approximately 2-fold increase at 3 µM concentration. The magnitude of the response to 3

SC

µM frenatin 2D was less than the response to 10 nM GLP-1. Incubation of frenatin 2D with isolated mouse islets also produced a significant increase in the rate of insulin release at 10

M AN U

nM (P < 0.05) and at 1 µM (P < 0.01) compared with the rate in the presence of 16.7 mM glucose only (Fig. 2B). Again, the magnitudes of insulin responses to frenatin 2D were

AC C

EP

TE D

significantly less that the responses to GLP-1 at the same concentration.

13

ACCEPTED MANUSCRIPT A)

6

G lu c o s e ( 5 .6 m M )

***

G lu c o s e ( 5 .6 m M ) + A la n in e ( 1 0 m M ) G lu c o s e ( 5 .6 m M ) + G L P - 1 ( 1 0 - 8 M ) G lu c o s e ( 5 .6 m M ) + F r e n a t in 2 D 4

***

3

6

Insulin release

(ng/10 cells/20 min)

5

*** ***

2

***

***

***

***

***

***

***

**

**

1

3 x 1 0 -6

1 0 - 7 3 x 1 0 - 8 1 0 - 8 3 x 1 0 - 9 1 0 - 9 3 x 1 0 - 10 1 0 - 1 0 3 x 1 0 - 11 1 0 - 1 1 3 x 1 0 - 1 2 1 0 - 12

1 0 -6 3 x 1 0 -7

P e p ti d e c o n c e n tr a ti o n [ M ]

6

*** G l u c o s e ( 5 .6 m M ) G l u c o s e ( 5 .6 m M ) + A la n in e ( 1 0 m M ) G lu c o s e ( 5 . 6 m M ) + G L P - 1 ( 1 0 - 8 M )

4

G lu c o s e ( 5 . 6 m M ) + F r e n a t in 2 . 1 S

***

3

6

Insulin release

(ng/10 cells/20 min)

5

SC

B)

2

***

***

***

**

**

0 3 x 1 0 -6

1 0 -7 3 x 1 0 -8 1 0 -8

1 0 -6 3 x 1 0 -7

**

*

*

*

M AN U

*** 1

RI PT

0

3 x 1 0 -9 1 0 -9

3 x 1 0 - 1 0 1 0 - 10 3 x 1 0 - 1 1 1 0 - 1 1 3 x 1 0 - 1 2 1 0 - 1 2

P e p ti d e c o n c e n tr a ti o n [ M ]

C)

6

*** G lu c o s e (5 .6 m M )

G lu c o s e (5 .6 m M ) + Ala n in e (1 0 m M ) G lu c o s e (5 .6 m M ) + G L P -1 (1 0 - 8 M ) G lu c o s e ( 5 .6 m M ) + F r e n a t i n 2 . 2 S

4

***

3

2

***

***

1

0 3 x 1 0 -6

TE D

6

Insulin release

(ng/10 cells/20 min)

5

***

***

***

***

***

**

**

*

*

1 0 - 7 3 x 1 0 - 8 1 0 - 8 3 x 1 0 - 9 1 0 - 9 3 x 1 0 - 10 1 0 - 10 3 x 1 0 - 11 1 0 - 11 3 x 1 0 - 1 2 1 0 - 1 2

1 0 -6 3 x 1 0 - 7

D)

6

***

G lu c o s e ( 5 .6 m M )

G lu c o s e ( 5 .6 m M ) + Ala n in e ( 1 0 m M ) G lu c o s e ( 5 .6 m M ) + G L P - 1 ( 1 0 - 8 M )

4

G lu c o s e ( 5 .6 m M ) + F r e n a t in 2 .3 S

3

***

AC C

6

Insulin release

5

(ng/10 cells/20 min)

EP

P e p ti d e c o n c e n tr a ti o n [ M ]

2

***

***

***

***

***

***

***

***

1

**

*

0

3 x 1 0 -6

1 0 -6 3 x 1 0 -7

1 0 - 7 3 x 1 0 - 8 1 0 - 8 3 x 1 0 - 9 1 0 - 9 3 x 1 0 - 1 0 1 0 - 1 0 3 x 1 0 - 11 1 0 - 1 1 3 x 1 0 - 1 2 1 0 - 1 2

P e p tid e c o n c e n tr a tio n [ M ]

Fig. 1. Comparison of the effects of (A) frenatin 2D, (B) frenatin 2.1S, (C) frenatin 2.2S and (D) frenatin 2.3S on insulin release from BRIN-BD11 rat clonal β-cells. Values are mean ± SEM for n = 8. *P < 0.05, **P < 0.01, and ***P < 0.001 compared to 5.6 mM glucose alone.

14

ACCEPTED MANUSCRIPT A) 6

***

G lu c o s e ( 5 .6 m M ) G lu c o s e ( 5 .6 m M ) + Ala n in e ( 1 0 m M ) G lu c o s e ( 5 .6 m M ) + E x e n d in - 4 ( 1 0 -8 M ) G lu c o s e ( 5 .6 m M ) + F r e n a t in 2 D

4

*** 3

6

*** 2

***

***

***

***

*

1 0 -7

1 0 -8

1 0 -9

1

0 1 0 -6

*

1 0 -1 0

1 0 -1 1

1 0 -1 2

SC

3 x1 0 -6

RI PT

Insulin release

(ng/10 cells/20 min)

5

B)

1 .4 m M g lu c o s e 1 6 .7 m M g lu c o s e

M AN U

P e p tid e c o n c e n tr a tio n [M ]

1 6 .7 m M g lu c o s e + Ala n in e ( 1 0 m M ) 1 6 .7 m M g lu c o s e + G L P - 1

1 6 .7 m M g lu c o s e + F r e n a t in 2 D

***

25

***

20 15 10 5

** **

*

*

1 0 -6

1 0 -8

1 0 -6

1 0 -8

P e p tid e c o n c e n tr a tio n [ M ]

AC C

EP

0

***

TE D

content)

Insulin release

(%of total insulin

30

Fig. 2. Effects of frenatin 2D on insulin release from (A) 1.1B4 human clonal β-cells and (B) isolated mouse islets. In panel A, values are mean ± SEM for n = 8, *P < 0.05 and ***P < 0.001 compared to 5.6 mM glucose alone. In panel B, values are mean ± SEM for n = 4, *P < 0.05, **P < 0.01 and ***P < 0.001 compared to 16.7 mM glucose alone.

15

ACCEPTED MANUSCRIPT 3.3. Effects of peptides on membrane depolarization and intracellular calcium ([Ca2+]i)

Incubation of BRIN-BD11 cells with 30 mM KCl produced an immediate and sustained increase in membrane potential. In contrast, incubation with frenatin 2D (1 µM)

RI PT

had no significant effect on membrane depolarization (Fig. 3). Similarly, incubation of

BRIN-BD11 cells with 10 mM alanine produced an immediate and sustained increase in [Ca2+]i whereas incubation with frenatin 2D (1 µM) had no significant effect on this

Effect of established modulators of insulin release on the activity of frenatin 2D

M AN U

3.4.

SC

parameter (Fig. 4).

The effects of known modulators of insulin release on the insulinotropic activity of frenatin 2D are shown in Table 1. The ability of the peptide (1 µM) to stimulate insulin

TE D

release from BRIN BD11 cells was maintained in the presence of the K+ channel activator, diazoxide (300 µM), the L-type voltage-dependent Ca2+ channels blocker, verapamil (50 µM) and in calcium-free medium. The depolarizing stimulus 30 mM KCl produced a marked (3-

EP

fold) increase in the rate of insulin release and the rate was significantly (P < 0.01)

AC C

augmented when the 30 mM KCl solution was supplemented with frenatin 2D (1 µM).

16

ACCEPTED MANUSCRIPT A) Glucose (5.6 mM) Glucose (5.6 mM) + KCl (30 mM) Glucose (5.6 mM) + Frenatin 2D (10-6 M)

30 20 10 0 -10 0

100

200

300

B)

M AN U

SC

Time (s)

RI PT

Membrane potential (RFU)

40

G lu c o s e ( 5 .6 m M ) a lo n e

G lu c o s e ( 5 .6 m M ) + K C l ( 3 0 m M )

G lu c o s e ( 5 .6 m M ) + F r e n a t in 2 D ( 1 0 - 6 M )

***

(AUC)

5000

2500

0

TE D

Area under the curve

7500

KCl

F r e n a ti n 2 D

A d d it io n s

EP

Fig. 3. (A) Effects of frenatin 2D on membrane potential in BRIN-BD11 cells expressed as relative fluorescence units, RFU as a function of time and (B) the integrated response (area

AC C

under the curve). Values are mean ± SEM (n = 6). ***P < 0.001 compared with 5.6 mM glucose alone.

17

ACCEPTED MANUSCRIPT A)

G lu c o se ( 5.6 m M )

1 0 .0

Intracellular Calcium (RFU)

G lu co s e ( 5.6 m M ) + Alan in e ( 10 m M ) G lu co se (5 .6 m M ) + F r e n at in 2 D ( 10 -6 M )

7 .5 5 .0 2 .5

- 2 .5 0

100

200

B)

G lu c o s e (5 .6 m M ) a lo n e G lu c o s e (5 .6 m M ) + Ala n in e (1 0 m M )

300

SC

T im e ( s)

RI PT

0 .0

G lu c o s e (5 .6 m M ) + F r e n a tin -2 D (1 0 - 6 M )

M AN U

Area under the curve (AUC)

1500

***

1000

500

0

TE D

A la n in e

F r e n a tin 2 D

A d d it io n s

Fig. 4. (A) Effects of frenatin 2D on intracellular calcium ion concentration [Ca2+]i in BRIN-

EP

BD11 cells expressed as relative fluorescence units, RFU as a function of time and (B) the

AC C

integrated response (area under the curve). Values are mean ± SEM for n = 6. ***P < 0.001 compared with 5.6 mM glucose alone.

18

ACCEPTED MANUSCRIPT Table. 1. Insulin releasing activity of frenatin 2D (1 µM) on BRIN BD11 cells in the presence of established modulators of insulin secretion

RI PT

SC

M AN U

5.6 mM glucose 5.6 mM glucose + frenatin 2D 5.6 mM glucose + 300 µM diazoxide 5.6 mM glucose + 300 µM diazoxide + frenatin 2D 5.6 mM glucose + 50 µM verapamil 5.6 mM glucose + 50 µM verapamil + frenatin 2D 5.6 mM glucose (calcium free) 5.6 mM glucose (calcium free) + frenatin 2D 16.7 mM glucose 16.7 mM glucose + frenatin 2D 16.7 mM glucose + 30 mM KCl 16.7 mM glucose + 30 mM KCl + frenatin 2D

Insulin release (ng/106 cells/20 min) 0.88 ± 0.02 1.66 ± 0.09 *** 0.73 ± 0.07 1.66 ± 0.03 *** 0.55 ± 0.03 1.48 ± 0.04 *** 0.73 ± 0.02 1.31 ± 0.07 *** 1.22 ± 0.01 1.91 ± 0.13 *** 5.77 ± 0.21 6.82 ± 0.21 **

Values are mean ± SEM (n = 6). * P < 0.05, **P < 0.01, *** P < 0.001 compared with the appropriate glucose control.

Effects of frenatin 2D on intracellular concentrations of cyclic AMP

TE D

3.5.

EP

Incubation of BRIN-BD11 cells with GLP-1 (10 nM) in the presence of IBMX resulted in a 215% increase (P < 0.001) in cAMP concentration compared with cells

AC C

incubated with 5.6 mM glucose + IBMX alone. Incubation with frenatin 2D (1 µM) produced a smaller but still significant (P < 0.01) increase in cAMP suggesting an involvement of the PKA pathway (Fig 5A). In a second series of experiments, the effects on the insulinotropic activity of frenatin 2D on down-regulation of the PKA and PKC pathways were investigated by overnight incubation of BRIN-BD11 cells with forskolin and PMA respectively. When the activators were not present, the rates of insulin release produced by frenatin 2D, GLP-1, and CCK-8 were significantly (P < 0.001) greater than that produced by 5.6mM glucose alone (Fig. 5B). The insulin stimulatory activities of frenatin 2D and GLP-1, but not CCK-8, 19

ACCEPTED MANUSCRIPT were significantly reduced when the PKA pathway was down-regulated with 25µM forskolin. In contrast, down-regulation of the PKC pathway with 10 nM PMA was without significant effect on the stimulatory activity of frenatin 2D and GLP-1 but the effect of CCK-8 was abolished. Down-regulation of both the PKA and PKC pathways by preincubation with

RI PT

forskolin + PMA abolished the stimulatory responses of all peptides tested.

3.6. Effect of frenatin 2D on proliferation and cytokine-induced apoptosis in BRIN-BD11

SC

cells

M AN U

Incubation of BRIN-BD11 cells with frenatin 2D (1 µM) significantly (P < 0.05) increased proliferation by 18% compared with incubations in the presence of culture medium alone (Fig. 6A). This degree of proliferative stimulation was less than that provided by incubation with 1 µM GLP-1 (48% increase). Incubation of BRIN-BD11 cells with frenatin

TE D

2D (1 µM) or with GLP-1 (1 µM) did not affect the number of cells exhibiting DNA damage, as measured by TUNEL assay. Incubations of the cells with a mixture of proinflammatory cytokines resulted in a 272% increase (P < 0.001) in the number of cells displaying apoptosis.

EP

The number of the apoptotic cells was reduced by 48% (P < 0.001) when the BRIN-BD11

AC C

cells were co-incubated with GLP-1 (1 µM) ) and the cytokine mixture. A comparable (38%, P < 0.05) reduction in the number of apoptotic cells was observed when the cells were co-incubated with frenatin 2D (1 µM) and the cytokine mixture (Fig. 6B).

20

ACCEPTED MANUSCRIPT G lu c o s e a lo n e ( 5 . 6 m M )

A)

G lu c o s e a lo n e ( 5 . 6 m M ) + G L P - 1 ( 1 0 n M ) G lu c o s e a lo n e ( 5 . 6 m M ) + IB M X ( 2 0 0 µ M ) G lu c o s e a lo n e ( 5 . 6 m M ) + G L P - 1 ( 1 0 n M ) + IB M X ( 2 0 0 µ M ) G lu c o s e a lo n e ( 5 . 6 m M ) + IB M X ( 2 0 0 µ M ) + F r e n a t in 2 D ( 1 µ M ) 15

∆∆∆

(pmol/ml)

∆

RI PT

cAMP release

*** 10

***

5

**

*

0

B)

G lu c o s e (5 .6 m M ) a lo n e

SC

G lu c o s e (5 .6 m M ) + F o r s k o lin ( 2 5 µ M ) G lu c o s e ( 5 .6 m M ) + P M A ( 1 0 n M )

G lu c o s e ( 5 .6 m M ) + F o r s k o lin ( 2 5 µ M ) + P M A ( 1 0 n M ) G lu c o s e ( 5 .6 m M ) + C C K 8 ( 1 0 n M )

G lu c o s e ( 5 .6 m M ) + F r e n a t in 2 D (1 0 -6 M ) ***

∆

*** ∆∆∆

***

*** ***

*** ***

6

(ng/10 cellss/20min)

2

*** ∆∆∆

0 C o n tro l

***

***

***

***

∆∆∆

TE D

Insulin release

6

4

M AN U

G lu c o s e ( 5 .6 m M ) + G LP - 1 ( 1 0 n M )

∆∆∆

F o rsk o l i n

∆∆ ∆

PMA

∆∆∆

***

φ

φφ +++ φφφ+∆+∆ +∆ ∆ ∆ ∆ φφφ +

+ + + ∆∆ ∆ ∆∆ ∆

∆∆∆

++

∆∆∆

P M A + F o rsk o l in

EP

Fig. 5. (A) Effects of frenatin 2D on cAMP production in BRIN-BD11 cells. Values are mean ± SEM for n = 6, *P < 0.05, **P < 0.01, ***P < 0.001 compared to 5.6 mM glucose

AC C

alone. ∆P < 0.05, ∆∆∆P < 0.001 compared to 5.6 mM glucose + IBMX. (B) Effects of frenatin 2D on insulin release from BRIN-BD11 cells following down-regulation of the PKA and PKC pathways by overnight culture with 25 µM forskolin or 10 nM PMA, respectively. Values are mean ± SEM for n = 8. ***P < 0.001 compared with 5.6 mM glucose, ∆P < 0.05 ∆∆∆

P < 0.001, compared to standard culture conditions, +P < 0.05, ++P < 0.01, +++P < 0.001

compared to culture with PMA, φP < 0.05, φφP < 0.01, φφφP < 0.001, compared with culture with forskolin.

21

ACCEPTED MANUSCRIPT A) C o n tro l G L P - 1 ( 1 0 -6 M ) t r e a t e d c e lls F r e n a t in 2 D ( 1 0 -6 M ) t r e a t e d c e lls

% of Ki67 positive cells (total number of cells)

75

∆∆∆

***

*

RI PT

50

25

B )

C o n tro l t r e a t e d c e lls

G LP -1 (1 0

-6

M ) t r e a t e d c e lls

M ) t r e a t e d c e lls + C y t o k in e t r e a t e d c e lls

F r e n a t in 2 D ( 1 0

-6

M ) t r e a t e d c e lls

F r e n a t in 2 D ( 1 0

-6

M ) t r e a t e d c e lls + C y t o k in e t r e a t e d c e lls

***

∆ ∆ ∆

25

***

0

∆

**

TE D

cells)

cells (total number of

50

-6

M AN U

C y t o k in e G L P -1 (1 0

% of tunnel positive

SC

0

Fig. 6. Effects of frenatin 2D (1µM) on (A) proliferation and (B) cytokine-induced apoptosis

EP

of BRIN-BD11 cells compared with GLP-1 (1 µM). Values are mean ± SEM for n=3. *P < 0.05, **P < 0.01, ***P < 0.001 compared with incubation in culture medium alone, ∆P <

3.7.

AC C

0.05 and ∆∆∆P < 0.001 compared with incubation in cytokine-containing medium.

Effects of frenatin 2D on glucose tolerance and insulin concentrations in mice

No adverse effects were observed in the animals following administration of the peptides. Blood glucose concentrations in lean, male NIH swiss TO mice receiving intraperitoneal glucose plus frenatin 2D (75 nmol/kg body weight) were significantly (P < 0.05) lower at 15 min and 30 min after administration compared with animals receiving 22

ACCEPTED MANUSCRIPT glucose only (Fig. 7A). The integrated response of blood glucose (area under the curve) after frenatin 2D was significantly (P < 0.05) less after administration of vehicle only (Fig. 7B). Plasma insulin concentrations were significantly (P < 0.001) higher at 15 min after glucose administration in animals receiving frenatin 2D (Fig. 7C) and the integrated response (total

RI PT

amount of insulin released over 60 min) was significantly (P < 0.05) greater compared with animals receiving glucose only (Fig. 7D). However, the magnitude of the effects on blood glucose concentrations and insulin release produced by administration of 75 nmol/kg body

SC

weight frenatin 2D was less than the effects produced by administration of 25 nmol/kg body

A)

B)

Glucose alone

Glucose + Frenatin 2D (75 nmol/kg bw)

*

*

10

**

1250

Blood glucose (nmol/l.min)

Blood glucose (mmol/lit)

Glucose + GLP-1 (25 nmol/kg bw) 20

M AN U

weight of GLP-1 (Fig. 7 A-D).

**

Glucose + GLP-1 (25 nmol/kg bw) Glucose + Frenatin 2D (75 nmol/kg bw)

1000

* **

750

***

500 250

0 25

50

Time (minutes)

C) Glucose alone

75

0

100

TE D

0

Glucose alone

D)

Glucose + GLP-1 (25 nmol/kg bw)

2

EP

Glucose + GLP-1 (25 nmol/kg bw)

75

*** * ***

1

Plasma insulin AUC (ng/ml.min)

Glucose + Frenatin 2D (75 nmol/kg bw)

AC C

Plasma insulin (ng/ml)

Glucose alone

Glucose + Frenatin 2D (75 nmol/kg bw)

** *

50

25

0

0

25

50

75

0

100

Time (min)

Fig. 7. A comparison of the effects of intraperitoneal administration of frenatin 2D (75 nmol/kg body weight) and GLP-1 (25 nmol/kg body weight) on blood glucose (panels A and B) and plasma insulin (panels C and D) concentrations in lean mice after co-injection of glucose (18 mmol/ kg body weight). Values are mean ± SEM for n = 6). *P < 0.05, **P < 0.01 and ***P < 0.001 compared with glucose alone.

23

ACCEPTED MANUSCRIPT 3.8.

Structure-activity studies of frenatin 2D

The effect of replacing each amino acid in frenatin 2D with a tryptophan residue on

RI PT

the ability to stimulate insulin release from BRIN-BD11 cells is shown in Table 2. Substitution of the Thr5, Asn8, Pro10, and Ile14 residues in frenatin-2D by Trp and

interchange of Pro12 and Phe13 residues and led to a >30,000-fold loss in insulinotropic

SC

potency. In contrast, the [D1W] and [G7W] analogues were equipotent and equally effective as the native peptide. The potencies and maximum responses of the free acid forms of

M AN U

frenatin 2D and the [D1W] and [G7W] analogues were not significantly different from the C-terminally α-amidated forms of the peptides.

Secondary structure analysis of frenatin 2D and impact of the [P10W] substitution

TE D

3.9

In both water and 25% methanol-water, the CD spectra of frenatin 2D exhibited a negative minimum at 197 nm indicating absence, or low content, of ordered conformation

EP

(Fig. 8A). The presence of a slight negative shoulder between 210 and 240 nm suggested the presence of a small component of turns or β-sheets in these two media. The addition of 25%

AC C

methanol led to a small increase of the ordered conformation content as shown by the increase of the Cotton effect between 210 and 240 nm and the slight shift of the negative minimum from 197 to 198 nm. In the presence of 25% TFE and 20 mM DPC (Fig. 8B), the spectrum of the peptide exhibited helical features with a positive peak around 190 nm and double negative minima between 200 and 240 nm. The higher mean residue molar ellipticities values observed in the presence of DPC micelles showed that frenatin 2D was more structured in this eukaryotic cell membrane mimetic medium than in the presence of 25% TFE. In both media, the wavelengths of the positive and negative peaks were shifted 24

ACCEPTED MANUSCRIPT from the usual α-helix values (188 and 189 vs 192 nm; 202 and 205 vs 208 nm; 225 and 220 vs 222 nm) suggesting the presence of additional secondary structure elements. The inactive [P10W] analogue exhibited spectra that were qualitatively similar to those observed for frenatin 2D in the presence of organic solvents and DPC micelles (Fig. 8B). Nevertheless, the

RI PT

substitution produced modification of the ellipticities values as well as a shift to higher

wavelengths of all bands present in the CD spectra suggesting an impact of the mutation on the peptide on secondary structure.

SC

The percentage of secondary element content was first estimated using Dichroweb server [26,27]. As deconvolution of the CD spectra indicated that both peptides contained a

M AN U

significant amount of turn and β-sheet structure in all media, a second analysis with the PEPFIT program was performed [28]. This algorithm has been developed for the analysis of CD spectra of small linear peptides and takes into account several types of turns. The helical content was also estimated from the mean residue ellipticity at 222 nm [30]. The estimation

TE D

of secondary structure calculated by the three methods are presented in Supplementary Table 1. The different algorithms gave similar estimates of the percentage of helix in agreement with the fact that these deconvolution methods generally provide reasonable estimates of

EP

polypeptide helical content [31]. The maximum amount of secondary structure was observed

AC C

in the presence DPC micelles. In this medium, frenatin 2D exhibited an helical content of 35%, corresponding to approximately 5 residues out of 14, and a turn content of 28%. In all media, the replacement of Pro10 by Trp gave rise to an increase of the secondary structure of the peptide (10-15%). In water and organic solvents, this led to the presence of a greater amount of turn conformations. As found with frenatin 2D, the secondary structure of the [P10W] analogue was stabilized in the presence of DPC micelles. In this medium, the analogue exhibited a higher amount of helical (40%) and turn structures (38%) than the naturally occurring peptide.

25

ACCEPTED MANUSCRIPT Table 2. Effects of frenatin 2D and its analogues on insulin release from BRIN-BD 11 cells

0.1 0.1 10 3000 3000 >3000 1000 0.1 >3000 100 >3000 100 1000 3000 >3000

1.01 ± 0.04 2.30 ± 0.03 *** 2.61 ± 0.25 *** 1.81 ± 0.11 *** 1.21 ± 0.04 ** 1.26 ± 0.07 ** 1.17 ± 0.15 1.59 ± 0.05 *** 2.57 ± 0.12 *** 1.09 ± 0.04 1.63 ± 0.01 *** 1.10 ± 0.06 1.30 ± 0.02 *** 1.30 ± 0.05 *** 1.44 ± 0.02 *** 1.10 ± 0.04

0.1

2.14 ± 0.12***

0.1

2.00 ± 0.24***

>3000

SC

M AN U

TE D

0.1

RI PT

Insulin release at 3 µM (ng/106 cells/20 min)

2.01 ± 0.11***

1.00 ± 0.03

EP

5.6 mM Glucose Frenatin 2D [D1W] [L2W] [L3W] [G4W] [T5W] [L6W] [G7W] [N8W] [L9W] [P10W] [L11W] [P12W] [F13W] [114W] Frenatin 2D non-amidated [D1W] Non-amidated [G7W] Non-amidated [P12F, F13P] Frenatin 2D

Threshold Concentration (nM)

Values are mean ± SEM for n = 8. **P < 0.01 and ***P < 0.001 compared to 5.6 mM

AC C

glucose alone.

26

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 8. Circular dichroism spectra of frenatin 2D (solid line) and its inactive [P10W] analogue (dashed line) in A) water (black) and 25% methanol (grey) and (B) 20mM dodecylphosphocholine (DPC) (black) and 25% trifluoroethanol (grey).

27

ACCEPTED MANUSCRIPT 4.

Discussion

Although peptides with antimicrobial activity in frog skin secretions were once regarded as an essential component in the animal’s system of host-defense, it is becoming

RI PT

increasingly recognized that skin-associated commensal bacteria may play a more important role in protection against microbial pathogens [32]. Similarly, the initial enthusiasm which discovery of these peptides generated in terms of their ability to kill microorganisms that had

SC

become resistant to conventional antibiotics has dwindled and no frog skin peptide is

currently in clinical practice as an antimicrobial agent. More promising potential clinical

M AN U

applications of amphibian host-defense peptides lie in their use as immunomodulatory agents (reviewed in [33]), promotors of wound healing (reviewed in [34]) and as templates for the design of drugs to treat patients with T2DM (reviewed in [7]).

The present study has shown that naturally occurring frenatin peptides (frenatin 2D,

TE D

2.1S, 2.2S & 2.3S) demonstrate dose-dependent insulin-releasing activity from clonal β-cells at concentrations that are not cytotoxic to the cells. Frenatin 2D was the most potent peptide producing a significant increase in the rate of insulin release from BRIN-BD11 at a

EP

concentration of 100 pM and the most effective producing a 2.3-fold increase in rate at a

AC C

concentration of 3 µM (Fig. 1). The peptide was also active in vivo when administered to mice together with a glucose load, resulting in lower blood glucose and higher circulating insulin concentrations (Fig. 7). Consequently, frenatin 2D was selected for further studies aimed at designing analogues with increased insulinotropic activity and to elucidate the mechanism of action of the frenatins. Insulin release from pancreatic β-cells is regulated by KATP channel-dependent and KATP channel-independent pathways [35,36]. In the former pathway, an increase in intracellular ATP generated from glucose metabolism results in closure of ATP-sensitive

28

ACCEPTED MANUSCRIPT potassium channels and an influx of calcium ions through the opening of voltage-dependent Ca2+ channels leading to exocytosis. The insulinotropic frog skin peptides CPF-SE1 [37], tigerinin-1R [38], and PGLa-AM1 [19] depolarize the cell membrane of BRIN-BD11 and increase [Ca2+]i suggesting that they operate via the KATP channel-dependent pathway. In

RI PT

contrast, pseudin-2 [39], hymenochirin 1B [18], and temporins A, F, and G [17] stimulate insulin release without significant effects on membrane depolarization or changes in [Ca2+]i. Like this second group of peptides, incubation of BRIN-BD11 cells with frenatin 2D had no

SC

effect on membrane potential or [Ca2+]i (Figs. 3 and 4) suggesting an involvement of the KATP channel-independent pathway. Consistent with this hypothesis, the insulinotropic

M AN U

activity of peptide was preserved in calcium-free medium and in the presence of diazoxide, an agent that inhibits the secretion of insulin by opening ATP-sensitive potassium channels in β-cells, and verapamil, an agent that inhibits insulin release by blocking voltagedependent calcium channels (Table 1). In KATP channel-independent pathway, activation of

TE D

adenylate cyclase by an agonist such as GLP-1 results in accumulation of cAMP which in turn activates PKA to promote release of insulin from BRIN-BD11 cells [40,41]. Incubation of BRIN-BD11 cells with frenatin 2D produced a modest but significant increase in cAMP

EP

concentrations (Fig. 5A). Consistent with this observation, the stimulatory effect of frenatin

AC C

2D on insulin release was abolished in PKA-downregulated BRIN-BD11 cells whereas in PKC downregulated cells the stimulatory effect of peptides was maintained (Fig. 5B). It is concluded, therefore, that the insulinotropic activity of frenatin 2D in these cells is mediated predominantly, if not exclusively, by the KATP channel-independent pathway. The relative antimicrobial and cytotoxic activities of naturally occurring host-defense peptides are determined by complex interactions between molecular charge, conformation, hydrophobicity and, in the case of α-helical peptides, amphipathicity [42]. Studies with a wide range of such peptides have shown that increasing cationicity, while maintaining

29

ACCEPTED MANUSCRIPT amphipathicity, generally by substitution of appropriate neutral or acidic amino acids by Llysine, results in increased antimicrobial activity by promoting interaction with the negatively charged cell membrane of prokaryotes [43]. In the present study, each amino acid in frenatin 2D was replaced by a bulky, hydrophobic tryptophan residue in an attempt to

RI PT

promote interaction with the zwitterionic plasma membrane of BRIN-BD11 cells. The

strategy did not lead to the design of an analogue with increased insulinotropic activity but the study demonstrated that the frenatin 2D molecule was very sensitive to changes in amino

SC

acid composition. Replacement of Thr5, Asn8, Pro10, and Ile14 by Trp led to loss of activity at concentrations up to and including 3 µM and substitution at Phe13 also led to a marked

M AN U

decrease in activity (Table 2). In contrast, replacements at Asp1 and Gly7 was without effect on potency and the effect of the substitution Leu2 → Trp on the response to a 3 µM stimulus was relatively minor. This suggest that the C-terminal region of the peptide is of particular importance in interaction with the β-cell membrane although the presence of a C-terminal α-

TE D

amide group is not necessary for activity. Far UV circular dichroism spectroscopy indicated that the secondary structure of frenatin 2D in DPC micelles, a system that mimics a zwitterionic membrane environment, is a mix of helix (35%), turns (28%) and unordered

EP

conformation (36%) Somewhat disappointingly, CD measurements of the inactive [P10W]

AC C

did not provide insight into which structural feature was the more important in mediating insulinotropic activity as both helix and turn structures were stabilized in the analogue relative to the native peptide. The progression of T2DM is associated with a decrease in the ability of the pancreatic

β- cell to release insulin due to a decline in β-cell number and it has been shown that the GLP-1 stimulates proliferation of β-cells and protect the cells against apoptosis stimulated by cytokines, glucose and fatty acids [44,45]. Overnight incubation of BRIN-BD11 cells with frenatin 2D also stimulated β-cell proliferation but the effect was significantly less than

30

ACCEPTED MANUSCRIPT the effect of an equimolar concentration of GLP-1. Similarly, frenatin 2D provided significant protection of cells against cytokine-induced apoptosis and, in this case, the effect was comparable to that provided by GLP-1 (Fig. 6). Frenatin 2D shows relatively weak insulinotropic activity both in vitro and in vivo

RI PT

compared with GLP-1 and, in view of the fact that it stimulates the release of TNF-α [9], a cytokine implicated as a causative factor in obesity-associated insulin resistance and the pathogenesis of type 2 diabetes [46], the peptide does not show great potential as a template

SC

for the development into an agent for T2DM therapy. Nevertheless, frenatin 2D represents an interesting system in which to study the interaction of a peptide with the plasma

M AN U

membrane of a β-cell. It seems highly improbable that mouse islets and rat BRIN-BD11 and human 1.1B4 cell lines are associated with a specific receptor for a particular frog peptide and so further studies are warranted to investigate the precise nature of the ligand-membrane

Acknowledgements

TE D

interaction that results in stimulation of insulin release.

EP

Funding for this study was provided by a project grant from Diabetes UK (12/0004457) and

AC C

by the University of Ulster Research Strategy Funding. The authors also thank Labex Synorg (ANR-11-LABX-0029) for financial support.

31

ACCEPTED MANUSCRIPT References

[1] J. Eng, W.A. Kleinman, L. Singh, G. Singh, J.P. Raufman, Isolation and characterization of exendin-4, an exendin-3 analogue, from Heloderma suspectum

pancreas, J. Biol. Chem. 267 (1992) 7402-7405. [2]

RI PT

venom. Further evidence for an exendin receptor on dispersed acini from guinea pig

F. Fehse, M. Trautmann, J.J. Holst, A.E. Halseth, N. Nanayakkara, L.L. Nielsen, M.S.

SC

Fineman, D.D. Kim, M.A. Nauck, Exenatide augments first-and second-phase insulin secretion in response to intravenous glucose in subjects with type 2 diabetes, J. Clin.

[3]

M AN U

Endocrinol. Metab. 90 (2005) 5991-5997.

M.C. Bunck, A. Corner, B. Eliasson, R.J. Heine, RM., Shaginian, M.R., Taskinen, U. Smith, H. Yki-Jarvinen, M. Diamant, Effects of exenatide on measures of beta-cell function after 3 years in metformin-treated patients with type 2 diabetes, Diabetes

[4]

TE D

Care 34 (2011) 2041-2047.

C. Raaymakers, E. Verbrugghe, S. Hernot, T. Hellebuyck, C. Betti, C. Peleman, M. Claeys, W. Bert, V. Caveliers, S. Ballet, A. Martel, F. Pasmans, K. Roelants,

EP

Antimicrobial peptides in frog poisons constitute a molecular toxin delivery system

[5]

AC C

against predators, Nat. Commun. 8 (2017) 1495. J.M. Conlon, M. Mechkarska, M.L. Lukic, P.R. Flatt, Potential therapeutic applications of multifunctional host-defense peptides from frog skin as anti-cancer, anti-viral,

immunomodulatory, and anti-diabetic agents, Peptides 57 (2014) 67-77.

[6]

X. Xu, R. Lai, The chemistry and biological activities of peptides from amphibian skin secretions, Chem. Rev. 115 (2015) 1760-1846.

32

ACCEPTED MANUSCRIPT [7]

J.M. Conlon, M. Mechkarska, Y.H. Abdel-Wahab, P.R. Flatt, Peptides from frog skin with potential for development into agents for Type 2 diabetes therapy, Peptides 100 (2018) 275-281.

[8]

M.J. Raftery, R.J.Waugh, J.H. Bowie, J.C. Wallace, M.J. Tyler, The structures of the

RI PT

frenatin peptides from the skin secretion of the giant tree frog Litoria infrafrenata, J. Pept. Sci. 2 (1996) 117-24. [9]

J.M. Conlon, M. Mechkarska, J.M. Pantic, M.L. Lukic, L. Coquet, J. Leprince, P.F.

SC

Nielsen, A.C. Rinaldi, An immunomodulatory peptide related to frenatin 2 from skin secretions of the Tyrrhenian painted frog Discoglossus sardus (Alytidae), Peptides 40

M AN U

(2013) 65-71.

[10] J.M. Conlon, M. Mechkarska, G. Radosavljevic, S. Attoub, J.D. King, M.L. Lukic, S. McClean, A family of antimicrobial and immunomodulatory peptides related to the frenatins from skin secretions of the Orinoco lime frog Sphaenorhynchus lacteus

TE D

(Hylidae), Peptides 56 (2014) 132-140.

[11] J.M. Pantic, G.D. Radosavljevic, I.P. Jovanovic, N.N. Arsenijevic, J.M. Conlon, M.L. Lukic, In vivo administration of the frog skin peptide frenatin 2.1S induces

AC C

275.

EP

immunostimulatory phenotypes of mouse mononuclear cells, Peptides 71 (2015) 269-

[12] J.M. Pantic, I.P. Jovanovic, G.D. Radosavljevic, N.M. Gajovic, N.N. Arsenijevic, J.M. Conlon, M.L. Lukic,The frog skin host-defense peptide frenatin 2.1S enhances

recruitment, activation and tumoricidal capacity of NK cells, Peptides 93 (2017) 44-50.

[13] C. Muñoz-Camargo, M.C. Méndez, V. Salazar, J. Moscoso, D. Narváez, M.M. Torres, F.K. Florez, H. Groot, E. Mitrani, Frog skin cultures secrete anti-yellow fever compounds. J. Antibiot. (Tokyo) 69 (2016) 783-790.

33

ACCEPTED MANUSCRIPT [14]

N.H. McClenaghan, C.R. Barnett, E. Ah-Sing, Y.H.A. Abdel-Wahab, F.P. O’Harte, T.W. Yoon, S.K. Swanston-Flatt, P.R. Flatt, Characterization of a novel glucoseresponsive insulin-secreting cell line, BRIN-BD11, produced by electrofusion, Diabetes 45 (1996) 1132-1140.

RI PT

[15] J.T. McCluskey, M. Hamid, H. Guo-Parke, N.H. McClenaghan, R. Gomis, P.R. Flatt, Development and functional characterization of insulin-releasing human pancreatic beta cell lines produced by electrofusion, J. Biol. Chem. 286 (2011) 21982-21992. B.O. Owolabi, V. Musale, O.O. Ojo, R.C. Moffett, M.K. McGahon, T.M. Curtis, J.M.

SC

[16]

Conlon, P.R. Flatt, Y.H. Abdel-Wahab, Actions of PGLa-AM1 and its [A14K] and

M AN U

[A20K] analogues and their therapeutic potential as anti-diabetic agents, Biochimie 138 (2017) 1-12.

[17] V. Musale, B. Casciaro, M.L. Mangoni, Y.H.A. Abdel-Wahab, P.R. Flatt, J.M. Conlon, (2018). Assessment of the potential of temporin peptides from the frog Rana

TE D

temporaria (Ranidae) as anti-diabetic agents, J. Pept. Sci. 24 (2018) e3065. [18] W. Shang, X.Yang, X. Ju, Y. Xie, Y. Zhang, W.H. Lee, Characterization of an insulinotropic peptide from skin secretions of Odorrana andersonii, J. Pept. Sci. 23

B.O. Owolabi, O.O.Ojo, D.K. Srinivasan, J.M. Conlon, P.R. Flatt, Y.H. Abdel-Wahab,

AC C

[19]

EP

(2017) 707-715.

(2016) In vitro and in vivo insulinotropic properties of the multifunctional frog skin peptide hymenochirin-1B: a structure–activity study. Amino Acids, 48 (2016) 535547.

[20] G.V. Graham, J.M. Conlon, Y.H. Abdel-Wahab, V.A. Gault, P.R. Flatt, Evaluation of the insulinotropic and glucose-lowering actions of zebrafish GIP in mammalian systems: evidence for involvement of the GLP-1 receptor, Peptides 100 (2018) 182189.

34

ACCEPTED MANUSCRIPT [21] P.R. Flatt, C.J. Bailey, Abnormal plasma glucose and insulin responses in heterozygous lean (ob/ob) mice, Diabetologia, 20 (1981) 573-577. [22] M.E. Doyle, J.M. Egan, Pharmacological agents that directly modulate insulin secretion, Pharmacol. Rev. 55 (2003) 105-131. M. Goto, T. Maki, T. Kiyoizumi, S. Satomi, A.P. Monaco, An improved method for

RI PT

[23]

isolation of mouse pancreatic islets, Transplantation, 40 (1985) 437-438.

[24] J.C. Miguel, S. Patterson, Y.H. Abdel-Wahab, P.C. Mathias, P.R. Flatt, P.R. Time-

SC

correlation between membrane depolarization and intracellular calcium in insulin

secreting BRIN-BD11 cells: studies using FLIPR, Cell Calcium, 36 (2004) 43-50. D. Srinivasan, O.O. Ojo, Y.H. Abdel-Wahab, P.R. Flatt, L. Guilhaudis, J.M. Conlon,

M AN U

[25]

Insulin-releasing and cytotoxic properties of the frog skin peptide, tigerinin-1R: a structure–activity study, Peptides 55 (2014) 23-31.

[26] L. Whitmore, B.A. Wallace, DICHROWEB: an online server for protein secondary

(2004) W668-673.

TE D

structure analyses from circular dichroism spectroscopic data, Nucleic Acids Res. 32

[27] L. Whitmore, B.A. Wallace, Protein secondary structure analyses from circular

AC C

392-400.

EP

dichroism spectroscopy: Methods and reference databases. Biopolymers 89 (2008)

[28] J. Reed, T.A. Reed, A set of constructed type spectra for the practical estimation of peptide secondary structure from circular dichroism. Anal. Biochem. 254 (1997) 3640.

[29]

N. Sreerama, R.W. Woody, Estimation of protein secondary structure from CD spectra: Comparison of CONTIN, SELCON and CDSSTR methods with an expanded reference set. Anal. Biochem. 287 (2000) 252-260.

35

ACCEPTED MANUSCRIPT [30]

B. Farood, E.J. Filiciano, K.P. Niambiar, Stabilization of α-helical structures in short peptides via end capping, Proc. Natl. Acad. Sci. USA 90 (1993) 838-842.

[31] N.J. Greenfield, Using circular dichroism spectra to estimate protein secondary structure, Nat. Protoc. 1 (2006) 2876-2890.

wildlife diseases, Proc. Biol. Sci. 279 (2012) 1457-1465.

RI PT

[32] J.H. Daskin, R.A. Alford, Context-dependent symbioses and their potential roles in

[33] J.M. Pantic, I.P. Jovanovic, G.D. Radosavljevic, N.N. Arsenijevic, J.M. Conlon, M.L.

SC

Lukic, The potential of frog skin-derived peptides for development into

therapeutically-valuable immunomodulatory agents. Molecules 22 (2017) E2071.

M AN U

[34] M.L. Mangoni, A.M. McDermott, M. Zasloff, Antimicrobial peptides and wound healing: biological and therapeutic considerations. Exp. Dermatol. 25 (2016) 167-173. [35] J.C. Henquin, Triggering and amplifying pathways of regulation of insulin secretion by glucose, Diabetes, 49 (2000) 1751-1760.

TE D

[36] J.C. Henquin, Pathways in beta-cell stimulus-secretion coupling as targets for therapeutic insulin secretagogues, Diabetes, 53 Suppl 3 (2004) S48-58. [37] D. Srinivasan, M. Mechkarska, Y.H. Abdel-Wahab, P.R. Flatt, J.M. Conlon, Caerulein

EP

precursor fragment (CPF) peptides from the skin secretions of Xenopus laevis and

AC C

Silurana epitropicalis are potent insulin-releasing agents, Biochimie. 95 (2013)429435.

[38] O.O. Ojo, D.K. Srinivasan, B.O. Owolabi, M.K. McGahon, R.C. Moffett, T.M. Curtis, J. M. Conlon, P.R. Flatt, Y.H. Abdel-Wahab, Molecular mechanisms mediating the beneficial metabolic effects of [Arg4]tigerinin-1R in mice with diet-induced obesity and insulin resistance, Biol. Chem. 397 (2016) 753-764.

36

ACCEPTED MANUSCRIPT [39] Y.H. Abdel-Wahab, G.J. Power, M.T. Ng, P.R., Flatt, J.M. Conlon, Insulin-releasing properties of the frog skin peptide pseudin-2 and its [Lys18]-substituted analogue, Biol. Chem. 389 (2008) 143-148. [40] B. Green, V. A. Gault, P.R. Flatt, P, Harriott, B. Greer, F.P.M. O’Harte, Comparative

RI PT

effects of GLP-1 and GIP on cAMP production, insulin secretion, and in vivo

antidiabetic actions following substitution of Ala 8/Ala 2 with 2-aminobutyric acid. Arch. Biochem. Biophys., 428 (2004) 136-143.

B. Green, V.A. Gault, F.P.M. O' Harte, P.R. Flatt, A comparison of the cellular and

SC

[41]

biological properties of DPP‐IV‐resistant N‐glucitol analogues of glucagon‐like

M AN U

peptide‐1 and glucose‐dependent insulinotropic polypeptide. Diabet. Obes. Metab. 7 (2005) 595-604.

[42] B. Mojsoska, H. Jenssen, Peptides and peptidomimetics for antimicrobial drug design, Pharmaceuticals (Basel) 8 (2015) 366-415.

TE D

[43] P. Kumar, J.N. Kizhakkedathu. S.K. Straus, Antimicrobial peptides: Diversity, mechanism of action and strategies to improve the activity and biocompatibility in vivo, Biomolecules 8 (2018) E4.

EP

[44] M. Cornu, J.Y. Yang, E. Jaccard, C. Poussin, C. Widmann, B. Thorens,. Glucagon-like

AC C

peptide-1 protects beta-cells against apoptosis by increasing the activity of an IGF2/IGF-1 receptor autocrine loop, Diabetes 58 (2009) 1816-1825.

[45] Y.S. Lee, H.S. Jun, Anti-diabetic actions of glucagon-like peptide-1 on pancreatic betacells, Metabolism 63 (2014) 9-19.

[46] D.E. Moller, Potential role of TNF-α in the pathogenesis of insulin resistance and Type 2 diabetes, Trends Endocrinol. Metab. 11 (2000) 212-217.

37

ACCEPTED MANUSCRIPT Supplementary Table 1. Prediction of secondary structure content from CD spectra.

Water

Method Dichroweb PEPFIT [Θ]222

25% TFE

20 mM DPC

[Θ]222

Dichroweb PEPFIT [Θ]222

Dichroweb PEPFIT [Θ]222

Dichrowebb PEPFIT [Θ]222

Frenatin [P10W]

Dichrowebb PEPFIT [Θ]222

EP

25% MeOH

AC C

25% TFE

20 mM DPC

turnsa

random

3 0 4

15 23 -

12 12 -

70 65 -

5 2 9

9 21 -

6 17 -

80 60 -

11 18 -

14 24 -

55 47 -

30 35 36

12 1 -

14 28 -

44 36 -

3 0 5

29 23 -

16 23 -

52 54 -

9 3 7

17 12 -

6 37 -

68 48 -

16 14 15

26 9 -

20 37-

38 40 -

51 40 32

13 0 -

15 38 -

21 22 -

20 11 22

TE D

Water

Dichroweb PEPFIT

β sheet

M AN U

Frenatin 2D

25% MeOH

helix

RI PT

Medium

SC

Peptide

Dichroweb PEPFIT [Θ]222

Dichroweb PEPFIT [Θ]222

Values are given in %. a refers to the total turn component given by the deconvolution programs. b the indicated values correspond to the average of the results obtained with CDSSTR and CONTINLL.

38

ACCEPTED MANUSCRIPT

•

Frenatin 2D (DLLGTLGNLPLPFI.NH2) stimulates insulin release from rat and human clonal β-cells and improves glucose tolerance in mice The peptide acts via the KATP channel-independent pathway

•

In DPC micelles, frenatin 2D exhibits a helical content of 35% and a turn content of 28%.

Replacement of Thr5, Asn8, Pro10, and Ile14 by Trp and interchange of Pro12 and

AC C

EP

TE D

M AN U

Phe13 leads to loss of insulinotropic activity

SC

•

RI PT

•