Diabetes mellitus drug discovery: insights into targeting feline and human amylin with small molecules

Abstract

Background

Type 2 diabetes (T2D) is a health concern for both humans and cats, with cases rising over the past decade. Around 70% of patients from either species exhibit pancreatic aggregates of islet amyloid polypeptide (IAPP), a protein that proves toxic upon misfolding. These misfolded protein aggregates congregate in the islets of Langerhans of the pancreas, diminishing the capability of β-cells to produce insulin and further perpetuating disease.

Objective

Our team’s drug discovery program is investigating newly synthesized compounds that could diminish aggregates of both human and feline IAPP, potentially disrupting the progression of T2D.

Material and methods

We prepared 24 compounds derived from diaryl urea, as ureas have previously demonstrated great potential at reducing accumulations of misfolded proteins. Biophysical methods were employed to analyze the anti-aggregation activity of these compounds at inhibiting and/or disrupting IAPP fibril formation in vitro.

Results

The results demonstrate that compounds 12 and 24 were most effective at reducing the fibrillization and aggregation of both human and feline IAPP. When compared with the control for each experiment, samples treated with either compound 12 or 24 exhibited fewer accumulations of amyloid-like fibrils.

Conclusion

Urea-based compounds, such as compounds 12 and 24, may prove crucial in future pre-clinical studies in the search for therapeutics for T2D.

Keywords:

• Amylin
• cat
• drug discovery
• islet amyloid polypeptide
• pancreatic amyloidosis
• type 2 diabetes
• small molecule therapeutics
• veterinary medicine

1. Introduction

Amyloidosis refers to a wide range of protein misfolding disorders characterized by the accumulation of toxic protein aggregates (Muchtar et al. 2021; Buxbaum et al. 2022). Notable examples include Alzheimer’s disease (Busche and Hyman 2020), Parkinson’s disease (Bloem et al. 2021), and type 2 diabetes (Buxbaum et al. 2022). Diabetes mellitus, also known as type 2 diabetes (T2D) (Sun et al. 2022), is a disorder affecting both humans and other species, such as cats, in which many cases involve the accumulation of misfolded islet amyloid protein (IAPP) in the pancreas, affecting the islets of Langerhans (Buxbaum et al. 2022). As of 2021, 537 million people worldwide are living with T2D, with this number expected to rise to 783 million by the year 2045 (Sun et al. 2022). Diabetes mellitus is diagnosed in approximately 0.5% of domestic cats, with the highest predisposition observed in Burmese cats and cats older than six years of age (O’Neill et al. 2016). Both cats and humans share certain risk factors for T2D, such as obesity and the consumption of processed foods (Choi and Shi 2001; Prahl et al. 2007). Furthermore, pancreatic aggregates of islet amyloid polypeptide (IAPP) are reported in ≥ 70% of tested feline and human T2D patients (O’Brien 2002; Hull et al. 2004; Rand 2013).

In protein folding disorders, targeting the intermediate stages of protein aggregation is critical. Initially harmless, monomers can undergo structural distortions leading to their aggregation into larger, problematic entities called oligomers. These oligomers, potent drivers of cellular dysfunction and toxicity, initiate disruptive processes such as permeability of organelles, disturbances in cell membranes, alterations in ion channels, and oxidative stress (Iadanza et al. 2018). These events collectively contribute to cellular degeneration and other pathological effects. Protofilaments, forming early during protein misfolding, serve as pivotal precursors to mature fibrils. Markedly distinct from mature fibrils, protofilaments exhibit unique structural features and can even puncture membranes in a manner resembling nanotubes (Iadanza et al. 2018). As they mature into elongated aggregates composed of stacked β-strands, fibrils epitomize the characteristic amyloid accumulations observed in various disorders. Displaying polymorphism due to assembly from multiple protofilaments with diverse twists and alignments, fibrils also contribute to the development of amyloid plaques and intracellular inclusions. These cascading processes trigger tissue damage and dysfunction, notably evident in conditions such as Alzheimer’s, Parkinson’s, and prion diseases (Iadanza et al. 2018).

IAPP, also known as amylin, is typically co-secreted with insulin by pancreatic β-cells. In its normal state, IAPP functions as a hormone that signals satiety and primarily works to inhibit the secretion of insulin and glucagon (Lutz 2010; Westermark et al. 2011). However, when IAPP misfolds and aggregates in the islets of Langerhans, it leads to a decline in both β-cell mass and function (Sevcuka et al. 2022). Such hints that the intermediate species of IAPP (oligomers) may possess cytotoxic properties toward pancreatic β-cells (Lorenzo et al. 1994; Haataja et al. 2008; Zhang et al. 2017). In terms of its structure, IAPP is a protein comprised of 37 amino acid residues and is prone to aggregation. Of these, the hydrophobic core, which encompasses residues 20–29, has been suggested as the region most likely to promote aggregation, particularly in the human (Betsholtz et al. 1989; Cao et al. 2020; King et al. 2022). IAPP has been isolated and amino acid sequence has been determined in a multitude of mammals other than the human and the cat, including the Chinese hamster, cow, hare, degu, tamarin, mouse, raccoon, and baboon (Westermark et al. 2002; Guardado-Mendoza et al. 2009; Buxbaum et al. 2022). Moreover, the ability of IAPP to form fibrils has been shown to correlate with genetic variations in its encoding sequences (Betsholtz et al. 1989). For instance, rodent IAPP does not have a propensity to aggregate because their 20–29 residue core contains more proline, which can disassemble beta sheet conformations (Gut 2019). Human IAPP (hIAPP) and feline IAPP (fIAPP) share the most similarities in IAPP sequence (Figure 1), which is supported by the high propensity of aggregation and frequently results in the development of T2D in both species (Betsholtz et al. 1990). IAPP undergoes self-aggregation to form amyloid fibrils, a process linked to pancreatic dysfunction in T2D (Milardi et al. 2021). Specifically, residues 8–20 and 30–32 influence the propensity of IAPP to aggregate and form fibrils. Furthermore, the presence of an internal disulfide bond is found to impact the aggregation kinetics and stability of IAPP, highlighting its role in modulating amyloid formation (Milardi et al. 2021).

Figure 1. Representative figure displaying the four differing positions in the 37 amino acid peptide strands of human (hIAPP) and feline (fIAPP) islet amyloid polypeptide.

A multitude of factors have been demonstrated to increase IAPP aggregation, providing ties to other disorders and complications. For instance, Rodrguez Camargo et al. investigated the impact of disturbed plasma components on hIAPP aggregation using transgenic mouse models of T2D. LDL-cholesterol and sugars were found to influence hIAPP aggregation, revealing potential links between T2D and complications like cardiac dysfunction in obesity (Rodriguez Camargo et al. 2018). Additionally, Meleleo et al. examined how metal ions, particularly Hg²⁺ and Cd²⁺, impact hIAPP aggregation, influencing its secondary structure and aggregation propensity. The findings reveal that Hg²⁺ and Cd²⁺ exhibit distinct modulation effects on hIAPP, increasing its level of aggregation and subsequent toxicity (Meleleo et al. 2022).

As of now, there is no cure for T2D in humans nor felines. Furthermore, there are no established treatments for pancreatic amyloidosis. However, various therapeutics are currently under investigation. One group of researchers investigated the anti-amyloidogenic properties of flavonoid compounds derived from Scutellaria baicalensis, a popular Chinese herb with antioxidant potential (Wang et al. 2022). Several of these derivatives were able to prevent hIAPP from misfolding, which is thought to be attributed to the presence of the ortho-hydroxybenzene structure in many of these inhibitors. Natural products, particularly EGCG, curcumin, and resveratrol, have potential to mitigate hIAPP (human islet amyloid polypeptide) aggregation and associated toxicity. EGCG, sourced from green tea, exhibits promise in re-routing the aggregation path of hIAPP, forming safe, nonfibrillar aggregates, and reducing its toxicity (Pithadia et al. 2016). Resveratrol, typically found in grapes and red wine, disrupts hIAPP’s affinity for cell membranes, potentially averting harmful cellular impacts (Pithadia et al. 2016). Curcumin, found in turmeric, disrupts the helical structures within hIAPP intermediates (Pithadia et al. 2016). Cox et al. reiterates the interplay of curcumin within the process of amyloid aggregation; their findings on CurDAc, a curcumin derivative, demonstrate its capability to modulate aggregation of various amyloid peptides, as well as inhibit and disaggregate amyloid fibers (Cox et al. 2020). Another team of researchers in Vancouver, Canada was able to identify a small molecule that reduces hIAPP oligomer accumulation in mice by facilitating clearance through autophagy (Bhagat and Verchere 2023). Maity et al. demonstrated the potential of designed chemical tools, specifically macrocycles, to disrupt the noncovalent interactions responsible for amyloidogenic protein self-assembly. By encapsulating hydrophobic regions within their cavities, these macrocycles thwart the aggregation of amyloid proteins, providing promising therapeutic avenues for a range of amyloid-related ailments (Maity 2023). Finally, García-Viñuales et al. investigated the anti-aggregation potential of Silybin A and Silybin B, two components derived from the milk thistle plant. Silybin B was shown to have the greater aggregation activity of the two compounds, interacting heavily with the hIAPP core and reducing its toxicity (García-Viñuales et al. 2022). Even Human Serum Albumin (HSA), which serves as a chaperone for intrinsically disordered proteins (IDPs) like Aβ, alpha-synuclein, and insulin, has potential to thwart amyloid formation through four proposed mechanisms: direct binding to IDP monomers or oligomers, competitive inhibition, metal chelation, and membrane protection (Martinez Pomier et al. 2022).

Other investigators have focused on the potential of urea and urea derivatives in reducing misfolded protein accumulations. Urea has been suggested to weaken hydrophobic interactions and disaggregate fibril structures of IAPP, serum amyloid A, alpha-synuclein, and amyloid-beta proteins (Kim et al. 2004; Wang and Colon 2005; Fortin et al. 2016; Elkamhawy et al. 2017; Galamba 2022). This new class of compounds possesses the necessary attributes for further drug development due to their affinity for IAPP, with early in vitro studies suggesting them to be nontoxic (Petitclerc et al. 2004). Previous research from Fortin et al. identified two specific N-phenyl-N’-(2-ethyl) ureas as inhibitors of hIAPP fibril formation, leading to a reduction in its cytotoxicity (Fortin et al. 2016). In this regard, we have synthesized a novel class of small molecules containing the urea linker, which are highly effective in preventing the formation of IAPP fibrils and inhibiting the accumulation of toxic IAPP oligomers. These compounds were developed through a simple synthetic approach utilizing a diverse set of commercially available anilines and isocyanates. Compounds from three series (2-aminofluorenes, 4-morpholino anilines, and 4-aminoindoles) were coupled with various substituted isocyanates to generate the urea derivatives. The effects of synthesized compounds on hIAPP and fIAPP fibril formation were evaluated using Thioflavin T (ThT) to compare their anti-aggregation potential. The best-performing compounds were then assessed using transmission electron microscopy (TEM) to study their impact on the fibril structure and morphology. The objective of this work is to identify novel compounds with anti-fibrillary effects on both human and feline IAPP. This work represents a starting point in the search for compounds with improved bioavailability that can target the formation of IAPP cytotoxic oligomeric species and potentially serve as future treatments of T2D in human and feline patients.

2. Materials and methods

2.1. Chemicals

All compounds were dissolved in DMSO at 40 mM. Hexafluoroisopropanol (HFIP), DMSO, resazurin and ThT were obtained from Alfa Aesar (Ward Hill, MA).

2.2. Chemical synthesis: general considerations

All moisture-sensitive reactions were conducted in oven-dried glassware under an atmosphere of dry nitrogen. Reaction solvents (CH₂Cl₂, Et₂O, etc.) were purchased from Sigma Aldrich in anhydrous form. All other solvents and reagents were purchased from commercial suppliers and used as received unless otherwise specified. The diaryl urea derivatives 1–24 were synthesized by treating commercially available aromatic amines; namely, 2-aminofluorene (series 1, Table 1), 4-morpholinoaniline (series 2, Table 2), and 4-aminoindole (series 3, Table 3), with substituted aromatic isocyanates in anhydrous dichloromethane at room temperature for 12 h. The general synthetic route for the preparation of the diaryl derivatives has been shown in Scheme 1. Thin layer chromatography (TLC) was performed with glass plates precoated with silica 60 Å F254 (250 mm) and visualized by UV light. ¹H and ¹³C NMR spectra were recorded using a 500 MHz Bruker instrument working at a frequency of 500 MHz for ¹H and at 126 MHz for ¹³C. Chemical shifts are reported in ppm using residual solvent resonances as internal reference (d 7.26 and d 77.0 for ¹H and ¹³C in CDCl₃, DMSO-d₆ 2.50 and DMSO-d₆ 39.51 for ¹H and ¹³C in DMSO-d₆). ¹H NMR data are reported as follows: b = broad, s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet. Coupling constants are given in hertz. The purity of all compounds and synthetic intermediates was judged to be 95% or better based on ¹H NMR. Details of synthesis of each molecule are presented in Supplementary Material.

Scheme 1. Preparation of diaryl urea derivatives; dichloromethane (DCM), room temperature (rt), 8–12 h. Ar₁ (R₁) is a placeholder for one of the three series: 2-aminofluorene, 4-morpholinoaniline, or 4-aminoindole. Ar₂ (R₂) refers to one of the following substituents: o-methylphenyl, o-methoxyphenyl, o-fluorophenyl, o-nitrophenyl, p-(trifluoromethyl)phenyl, o,p-dimethoxylphenyl, o-methylchloro, m-(trifluoromethyl)phenyl, or o,p-dimethoxy.

2.3. Peptide synthesis

Synthetic hIAPP and fIAPP (1–37) was obtained from Anaspec (city, state, USA). For both peptides, the disulfide bonds were present between amino acids 2 and 7. For the feline IAPP, the peak area by HPLC was over 95% and no other peak major peak was observed in the chromatogram. Purity was considered at 95%. The mass obtained by mass spectrometry was 3912.6 ± 0.2%. For the human IAPP, the peak area by HPLC was over 95% and no other peak major peak was observed in the chromatogram. Purity was considered at 98%. The mass obtained by mass spectrometry was 3903.5 ± 0.2%.

2.4. Thioflavin T (ThT) fluorescence assay

hIAPP and fIAPP peptides from the stock solution of 1 mM were each added to 10 mM PBS buffer (pH 7.4) and transferred to a black 96-well microtiter plate with transparent bottom. Each well contained a final volume of 150 μL with a final peptide concentration of 10 μM. Experiments in the presence of hIAPP and fIAPP were performed with different compounds at a final concentration of 100 µM. For the dose response, final drug concentrations ranged from 1.56 to 100 μM. The background signal consisted of 0.25% DMSO without peptide and compound. The fluorescence emission experiments were performed as previously published (Fortin et al. 2016). The maximum fluorescence intensity in percentage reported in Table 1 was calculated as previously described (Fortin et al. 2016).

Table 1. Chemical structure, physicochemical descriptors, and the effect of the 2-aminofluorene series of compounds on hIAPP and fIAPP aggregation measured as fluorescence intensity using ThT assay.

ID#	Structure	MW	TPSA	cLogP	cLogD	LR5	hIAPP	fIAPP
1		318.3443	41.13	4.22	5.03	N	39.5 ± 3.3	41.4 ± 0.6
2		368.3518	41.13	4.94	5.76	N	59.3 ± 2.7	63.7 ± 1.4
3		368.3518	41.13	4.96	5.76	N	83.3 ± 2.5	70.9 ± 1.2
4		345.3514	86.95	3.42	4.82	Y	30.8 ± 31	25.5 ± 1.9
5		348.8255	41.13	4.35	5.47	N	31.8 ± 0.4	16.3 ± 3.2
6		314.3804	41.13	4.23	5.4	N	36.0 ± 1.9	61.0 ± 3.8
7		330.3798	50.36	3.84	4.73	Y	22.1 ± 2.6	24.0 ± 1.1
8		360.4058	59.59	3.88	4.57	Y	45.6 ± 0.8	45.8 ± 8.6

Notes: TPSA, Log P, and Log D were obtained with SwissADME or Chemaxon. The ThT (%) represent the maximum fluorescence intensity obtained at the plateau phase of the kinetics of fibrillization using 10 µM of IAPP. Compounds were tested at 100 µM (molar ratio 1:10). ThT % indicated are average with SEM. These results are representative of three independent experiments performed with technical triplicates. cLogD: calculated distribution coefficient at 7.4; cLogP: calculated partition coefficient; LR5: Lipinski’s rule of five; MW: molecular weight; ThT: Thioflavin T; TPSA: topological polar surface area.

2.5. Transmission electron microscopy (TEM)

hIAPP and fIAPP were each incubated in 10 mM PBS (pH 7.4, 25 °C) at 10 μM for 12 h with different compounds (12 or 24) at 100 µM. Samples were treated with one of the three conditions: DMSO (control), compound 12, or compound 24.

Following the time incubation , a volume of 10 µL was applied on a 400-mesh Formvar-carbon-coated copper grid (Electron Microscopy Sciences, Hatfield, PA). Grids were incubated for 1 min and washed once with distilled water. After air-drying, grids were incubated for 1 min in a fresh solution of 1% uranyl acetate. Grids were air-dried again and imaged via transmission electron microscopy (JEOL 1400 Flash, Japan). Pictures were acquired at an accelerating voltage of 100 kV and magnification of 20k and 40k.

2.6. Dynamic light scattering

The Malvern Zetasizer machine was utilized to conduct dynamic light scattering (DLS), in which the particle size of each sample is analyzed. In microcentrifuge tubes, hIAPP and fIAPP were each incubated in 10 mM PBS (pH 7.4, 25 °C) at 10 μM for 1 h with different compounds (12 or 24) at 100 µM. to attain a 1:10 molar ratio. From each sample, 50 μL was added to a low volume quartz cuvette (QS high precision cell, 105-251-85-40, Hellma analytics) and read at 25 °C using the fluorescence filter. The average of three repetitions of each sample were recorded.

2.7. Chemistry

The diaryl urea derivatives 1–24 were synthesized by treating commercially available aromatic amines; namely, 2-aminofluorene (series 1, Table 1), 4-morpholinoaniline (series 2, Table 2), and 4-aminoindole (series 3, Table 3), with substituted aromatic isocyanates in anhydrous dichloromethane at room temperature for 12 h. The general synthetic route for the preparation of the diaryl derivatives has been shown in Scheme 1. All compounds were obtained in moderate to high yields. The compounds were characterized using both proton (¹H) and carbon (¹³C) NMR spectroscopy. The melting point for each compound was also recorded. All spectral data have been reported in Supplementary Material.

3. Results

3.1. Structure-activity relationships of the 2-aminofluorene, 4-morpholinoaniline, 4-aminoindole series of compounds

The anti-aggregation activity of diaryl urea derivatives 1–24 were measured on both human and feline islet amyloid polypeptide (IAPP). The ThT fluorescence intensity was used to determine the compound effect on the IAPP fibril formation (Tables 1–3). In this test, lower fluorescence corresponds with a greater reduction of fibrils in the sample. Compounds with a measured fluorescence percentage around or less than 10% for either hIAPP or fIAPP were selected as the compounds with greatest potential for reducing IAPP fibril accumulation. Based on the data obtained, 4-aminoindolyl urea derivative 24 was found to be the most effective anti-fibrillary compounds (among the 3 series of compounds) with fluorescence intensities of 7.5 ± 0.4% respectively, for hIAAP; 6.9 ± 1.4%, respectively, for fIAPP. In the 2-aminoflurorene series, compounds 4, 5, and 7 exhibited the best anti-fibrillary activities for both hIAPP and fIAPP. Compound 6 was two-fold more effective inhibitor of aggregation over hIAPP than fIAPP while compound 5 was found to be more effective anti-aggregator toward fIAPP than hIAPP. Reduction in fluorescence intensity was improved with substitution on the phenyl ring with electron donating group, such as methoxy, at the ortho position. The presence of a methoxy confered an improvement in the anti-aggregation activity for this series. In the 4-morpholino urea series, compounds 9 to16 exhibited a better anti-fibrillar activity toward hIAPP than fIAPP. Compound 12 had the greatest anti-aggregation activity of the agents in this series with a fluorescence intensity of 9.8 ± 0.3 and 15.4 ± 0.4% on hIAPP and fIAPP, respectively. For the 4-aminoindole urea series, all compounds except compound 20 exhibited anti-fibrillary activities on both hIAPP and fIAPP. Substitution of two methoxy groups at ortho and para position on the phenyl ring resulted in one highly effective compound, 24, capable to inhibit the aggregation of both hIAPP and fIAPP. The overall range of fluorescence intensities for the 4-aminoindole series (Table 3) was 7–45%, which was far superior to the ranges of the 2-aminofluorene (16–71%, Table 1) and 4-morpholino urea (10–88%, Table 2) series.

Table 2. Chemical structure, physicochemical descriptors, and the effect of the 4-morpholinoaniline series of compounds on hIAPP and fIAPP aggregation measured as fluorescence intensity using ThT assay.

ID#	Structure	MW	TPSA	cLogP	cLogD	LR5	hIAPP	fIAPP
9		315.3421	53.60	2.57	3.15	Y	42.6 ± 2.7	72.1 ± 1.8
10		365.3496	53.60	3.29	3.89	Y	15.1 ± 0.7	33.4 ± 4.3
11		365.3496	53.60	3.27	3.89	Y	13.4 ± 0.3	20.9 ± 0.6
12		342.3492	99.42	1.68	2.95	Y	9.8 ± 0.3	15.4 ± 0.4
13		345.8233	53.60	2.75	3.6	Y	20.7 ± 0.5	27.8 ± 3.3
14		311.3782	53.60	2.61	3.52	Y	31.5 ± 1.1	88.8 ± 3.4
15		327.3776	62.83	2.19	2.85	Y	13.6 ± 1.3	27.9 ± 1.0
16		357.4036	72.06	2.22	2.69	Y	27.2 ± 2.0	50.4 ± 1.6

Notes: TPSA, Log P, and Log D were obtained with SwissADME or Chemaxon. The ThT (%) represent the maximum fluorescence intensity obtained at the plateau phase of the kinetics of fibrillization using 10 µM of IAPP. Compounds were tested at 100 µM (molar ratio 1:10). ThT % indicated are average with SEM. These results are representative of three independent experiments performed with technical triplicates. cLogD: calculated distribution coefficient at 7.4; cLogP: calculated partition coefficient; LR5: Lipinski’s rule of five; MW: molecular weight; ThT: Thioflavin T; TPSA: topological polar surface area.

Table 3. Chemical structure, physicochemical descriptors, and the effect of the 4-aminoindole series of compounds on hIAPP and fIAPP aggregation measured as fluorescence intensity using ThT assay.

ID#	Structure	MW	TPSA	cLogP	cLogD	LR5	hIAPP	fIAPP
17		269.2737	56.92	2.88	3.36	Y	22.5 ± 0.1	25.3 ± 1.9
18		319.2812	56.92	3.58	4.1	Y	14.0 ± 0.3	15.0 ± 2.1
19		319.2812	56.92	3.60	4.1	Y	16.7 ± 0.3	23.2 ± 0.9
20		296.2808	102.74	2.09	3.16	Y	26.9 ± 0.3	44.5 ± 1.9
21		299.7549	56.92	3.11	3.80	Y	31.1 ± 0.2	39.9 ± 0.9
22		265.3098	56.92	2.90	3.73	Y	22.6 ± 1.0	34.1 ± 2.8
23		281.3092	66.15	2.54	3.06	Y	11.6 ± 0.2	18.1 ± 0.6
24		311.3352	75.38	2.54	2.90	Y	7.5 ± 0.4	6.9 ± 1.4

Notes: TPSA, Log P, and Log D were obtained with SwissADME or Chemaxon. The ThT (%) represent the maximum fluorescence intensity obtained at the plateau phase of the kinetics of fibrillization using 10 µM of IAPP. Compounds were tested at 100 µM (molar ratio 1:10). ThT % indicated are average with SEM. These results are representative of three independent experiments performed with technical triplicates. cLogD: calculated distribution coefficient at 7.4; cLogP: calculated partition coefficient; LR5: Lipinski’s rule of five; MW: molecular weight; ThT: Thioflavin T; TPSA: topological polar surface area.

3.2. Evaluation of the synthesized compounds to inhibit the hIAPP and fIAPP aggregation by Thioflavin T (ThT) fluorescence assay

The ThT fluorescence assay is a biophysical method employed to study the kinetics of fibril formation in prone-to-aggregate proteins. In the assay, the diaryl derivatives were added followed by the addition of ThT. After this, hIAPP or fIAPP were mixed in the buffer. The kinetics of fibril formation begins when the protein is dissolved in the buffer and introduced onto the plate. In the ThT assay graph, the plateau phase represents the formation of mature fibrils, where an equilibrium is achieved between the aggregation and disaggregation process. The anti-fibrilization property of novel compounds is determined at the plateau phase, at which fluorescence intensity (%) of the tested compounds is compared for different incubations. The kinetic curves obtained for compounds 1–24 are depicted in Figure 2. As mentioned previously, the 4-indolyl series of compounds seem to have the greatest anti-aggregation activities, followed by the 4-morpholino series and the 2-aminofluorene series. Compounds that scored below or very near 10% in ThT assay for both conditions (hIAPP and fIAPP) were selected for further testing. These compounds were 12 and 24. Compounds 12, 23, and 24 were selected to be tested for their dose responsiveness as they had the lowest fluorescence (Figure 3).

Figure 2. Kinetics of IAPP fibril formation obtained with different diaryl derivatives of urea monitored with the thioflavin T (ThT) fluorescence assays. In this first-tier assay, compounds were tested at a final concentration of 100 μM in the presence of hIAPP and fIAPP at 10 μM. Molar ratio peptide: compound is 1:10. Series 1, 2, and 3 represent the 2-aminofluorene, morpholino, and 4-aminoindole, respectively.

Figure 3. Kinetics of IAPP fibril formation with increasing doses of three synthesized compounds 12, 23, and 24, assessed via thioflavin T (ThT) fluorescence assay. The original stock of the compound was tested at 100, 50, 25, 12.5, 6.25, 3.125, and 1.56 µM concentration. These compounds were tested in the presence of hIAPP and fIAPP at 10 μM.

Figure 4. Size of hIAPP based on light scattering. Control at 60 min showed one major fibril peak in the 1000 nm range. The two tested compounds (12 and 24) were individually added to aliquots of hIAPP and incubated for 60 min before analysis to determine the size of the particles contained within the sample.

Figure 5. TEM imaging of amylin fibrils incubated with two unique treatments. Samples of fIAPP were solubilized at 10 µM in PBS buffer. Samples were incubated with DMSO (0.25%), compound 12, or compound 24. After the end of fibril formation kinetics (≤ 12 h), samples were deposited on copper grids and stained for TEM visualization. All images are displayed at 20k magnification.

We performed the ThT assay using a Tris 20 mM buffer at a pH of 5.5 to closely mimic a very low physiological pH. In the human, the pancreatic interstitial tissue maintains a pH of 7.25 ± 0.04. In the feline, this pH rises to 7.41 ± 0.01 (Patel et al. 1995). Despite the acidic conditions, the proteins under investigation exhibited aggregation, and we observed that the compounds retained their efficacy in inhibiting this aggregation (Supplementary Material, Figure S1). This suggests that the inhibitory properties of the compounds are not significantly affected by the lower pH.

3.2.1. Dose response curves of the best anti-fibrillary compounds

The dose dependency of the anti-fibrillar effect of best compounds was then evaluated with different concentrations. The three best compounds were added at various concentrations, for the anti-aggregation effect on hIAPP and fIAPP (Figure 3). Specifically, compounds 12, 23, and 24 were selected to be tested as they had the lowest fluorescence. When tested, the anti-aggregation effect was found to be more prominent in hIAPP compared to fIAPP. However, for both peptides, compound 12 was the most effective at reducing the fibril formation even at low doses. Compound 24 also reduced fibril formation at low doses. Interestingly, low doses of compound 23 resulted in high fluorescence intensity (possibly increased fibril formation). Therefore, compound 23 was not included in further testing.

3.3. Impact of the best compounds on the particle size of hIAPP via dynamic light scattering assay

Dynamic light scattering (DLS) was used to determine the size and distribution of amylin proteins after 1 h of incubation with our two most promising compounds. Before testing, 10 µM of hIAPP incubated with 100 µM of one of the three treatments: DMSO (control), compound 12, or compound 24. The particle diameters were analyzed via DLS after 60 min. When measured at 0 min, the stock solution predominantly contained monomers, with some pre-fibrillar proteins present (Figure 4). The amylin misfolds quickly, with pre-fibrillar proteins forming within 15 min. After 60 min of incubation at 25 °C, the control contained fibrils, as indicated by the steep peak around 1000 nm. All compounds reduced the percentage of intensity of the peak at around 1000 nm with compounds 12 and 24 resulting in a slight left shift, confirming a slight reduction of particle size (Figure 4). Compound 12 exhibited peaks near 1000 nm, as well, indicating that neither of these compounds decreased the presence of aggregates in solution. Compound 24 proved more effective at reducing fibrils, but a peak around 660 nm indicates that this compound was not entirely effective at reducing large aggregates. Interestingly, small peaks around 1 nm and 35 nm suggest the presence of monomeric species in the sample treated with compound 24, indicating its partial success in preventing IAPP aggregates.

3.4. Direct visualization of fIAPP fibrils following compound treatment by transmission electron microscopy

TEM is one of the leading microscopy methods for imaging small particles, including proteinaceous fibrils, as the image is generated through the electron beam bombardment of the sample deposited onto a copper grid. fIAPP samples treated with DMSO exhibited large clusters of dense fibrils (Figure 5). When treated with compound 24, fIAPP fibrils were not significantly different from treatment with DMSO, but fIAPP fibrils were narrow and isolated, refusing to accumulate in large clusters (Figure 5). fIAPP samples treated with compound 12 exhibited a great reduction in the clustered accumulation of fibrils, as much of the field was empty except for occasional small, isolated packets.

4. Discussion

With the prevalence of T2D rising rapidly in both felines and humans, it is of great necessity to discover and develop small molecule therapeutics against the disease. Felines with diabetes exhibit reduced beta-cell mass and function, as well as decreased insulin sensitivity (McCann et al. 2007; O’Neill et al. 2016), resulting in a shortened lifespan and diminished quality of life for the affected cats. Cats who are overweight, inactive, and older are at highest risk for developing T2D, with neutered males comprising the largest subset of cats with T2D (McCann et al. 2007; Prahl et al. 2007; O’Neill et al. 2016). While sterilization itself is not a risk factor for developing T2D, cats who are altered are predisposed to weight gain and therefore more likely to develop T2D (McCann et al. 2007). Furthermore, certain breeds are genetically predisposed for acquiring diabetes; most notably the Burmese, but also the Tonkinese and Norwegian Forest breeds (McCann et al. 2007; O’Neill et al. 2016). The increased risk of T2D in Burmese cats is well published, with studies reporting that these cats are nearly four times as likely to develop diabetes mellitus (McCann et al. 2007). As such, it is imperative to provide a compound to prevent the pancreatic amyloid deposits in cats at risk for disease to lessen the severity of T2D. For patients already diagnosed with T2D, the selected compound(s) could be paired with current disease management techniques (i.e. insulin injection, dietary restrictions, and/or weight loss programs).

Our lab is at the early stage of drug discovery. While there are multiple methods for performing testing, we have opted for phenotypic drug screening. Phenotypic drug screening investigates the effect of compounds, including anti-fibrillar activity, anti-oligomer activity, and cellular protection. The FDA does not require the characterization of the mechanism of action, although this may be important for understanding the pharmacological effect of a test article.

Our team investigated the disaggregation potential of 24 newly synthesized compounds belonging to three distinct classes. Of these, compounds 12 and 24 exhibited the best overall potential. These two compounds prompted significant reduction of both fIAPP and hIAPP fibrils when analyzed via the biophysical methods described previously. Each of the promising compounds demonstrated a fluorescence intensity of less than 10% for either hIAPP or fIAPP when analyzed via ThT assay. Such indicates the scarce presence of large aggregates in samples treated with these compounds. When visualized via TEM, these same samples were observed to have few packets of fibrils. DLS results portrayed shorter peaks of diminished intensity for samples treated with compounds 12 and 24. Each of these results is indicative of the anti-fibrillar nature of these two compounds. Furthermore, the dose response kinetic curves yielded important information about the efficacy of various concentrations of each compound. Compound 12 is effective even at low concentrations, exhibiting promise for later pharmaceutical use. Future studies will be performed to understand the specific interactions between the compounds and their molecular targets, as well their sensitivity to the secondary structure of IAPP. For instance, Isothermal Titration Calorimetry (ITC) could be used to uncover a deeper understanding of the binding affinity between these molecules and IAPP.

Ideally, one or more of these compounds could serve as candidates for clinical trials in the future. These compounds respect Lipinski’s rule of five (LR5 in Tables 1 to 3), which establishes four guidelines to predict a compound’s bioavailability (Lipinski et al. 2001). Lipinski states that compounds with fewer than five hydrogen bond donors and ten hydrogen bond acceptors, as well as a molecular weight less than 500 g/mol and a calculated Log P less than 5, are most likely to have high bioavailability and therefore therapeutic potential (Lipinski et al. 2001). Compounds 12 and 24 both respect Lipinski’s rule of five, as do many other ureas. Urea derivatives consistently demonstrate conformation to the rule of five, exhibiting low calculated Log P values and a molecular weight less than 500 g/mol, in addition to being highly soluble and stable (Azam et al. 2012; Mannam et al. 2019). Mounetou et al. investigated the use of a urea derivative, chloroethylurea, as a therapeutic for colorectal cancer. Chloroethylurea follows the rule of five and remains stable in the tissue after administration, exhibiting good bioavailability (Mounetou et al. 2010). Furthermore, the investigators found that the compound primarily concentrates in organs of the digestive system upon intraperitoneal or intravenous injection (Mounetou et al. 2010). Work by Radhakrishnan et al. echoes the systematic distribution of urea derivatives; upon administration of a quinoxaline urea analog to mice via drinking water, high concentrations of the compound were observed in the pancreas and liver (Radhakrishnan et al. 2013). With the distribution of urea compounds overlapping with the target tissue for potential T2D therapeutics, ureas hold promise as an antiaggregant of IAPP.

To pursue further drug discovery efforts, the next step for this research would be a proof-of-concept study with a rodent model. The HIP rat, first reported by Butler et al. is a strain of rats prone to IAPP oligomerization and accumulation in the pancreas (Butler et al. 2004). At just five months of age, these rats exhibit a 60% reduction in ß-cell mass, a hallmark feature of diabetes. Furthermore, amyloid accumulations in the islets of Langerhans were detected in HIP rats as young as two months of age. Upon reaching 18 months of age, these rats experience a 90% reduction in ß-cell mass and a significant accumulation of islet amyloid accumulations in their pancreas (Butler et al. 2004). Therefore, the HIP rat model would serve as an indispensable tool to test the in vivo bioavailability and success of the chosen compound.

In summary, our work has led to the synthesis of two compounds that have exhibited potential to inhibit IAPP fibril accumulation. These findings are paramount in the discovery of anti-IAPP therapeutics for T2D in both human and feline patients. Such compounds could serve as preventatives for breeds most predisposed to T2D or lessen the severity of disease for current patients.

Acknowledgment

The authors would like to acknowledge the professional services of Alicia Withrow at the Michigan State University Center for Advanced Microscopy and Prasanth Saraswati for the technical expertise.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This project was funded by the EveryCat Health Foundation (W20-020). Kendall B. E. Moore and Natalie G. Horgan were part of the American Society for Investigative Pathology (ASIP) Summer Research Opportunity Program in Pathology (SROPP). Brooke Lenters was part of the Purdue Veterinary Scholar Summer Research Program supported by Boehringer Ingelheim.