Calcium elevation in mouse pancreatic beta cells evoked by extracellular human islet amyloid polypeptide involves activation of the mechanosensitive ion channel TRPV4
- 1.5k Downloads
To investigate the mechanism by which human islet amyloid polypeptide (hIAPP) fibril formation results in calcium influx across the plasma membrane of pancreatic beta cells, and its association with apoptosis.
Cytoplasmic intracellular calcium concentrations ([Ca2+]i) were monitored for 2 h as the 340/380 nm fluorescence ratio in fura-2 loaded cells of the MIN6 mouse pancreatic beta cell line. Cell morphology was evaluated by transmission electron microscopy, and viability by FACS.
hIAPP (10 μmol/l) increased [Ca2+]i in 21% of MIN6 cells in standard buffer, and in 8% of cells in Na+-free buffer. Transient receptor potential (TRP) channel inhibitors (gadolinium and ruthenium red) prevented the [Ca2+]i rise under both conditions, whilst nifedipine was only effective in the presence of Na+. hIAPP increased apoptosis in both insulinoma cells and islets in primary culture, and cell viability was partially rescued by ruthenium red (p < 0.001). By RT-PCR, we detected expression of the mechanosensitive TRP cation channel subfamily V member 4 (Trpv4) in MIN6 cells and mouse pancreas. Small interference RNA against Trpv4 prevented hIAPP-induced [Ca2+]i rises, decreased hIAPP-triggered expression of the endoplasmic reticulum (ER) stress response, and reduced hIAPP-triggered cell death by 50% (p < 0.05).
Alterations in [Ca2+]i play a key role in hIAPP-induced beta cell cytotoxicity. By electron microscopy, we detected extracellular hIAPP aggregates adjacent to irregular invaginated regions of the plasma membrane. We propose that TRPV4 channels may sense physical changes in the plasma membrane induced by hIAPP aggregation, enabling Ca2+ entry, membrane depolarisation and activation of L-type Ca2+ channels. Decreasing the activity of TRPV4 prevented hIAPP-induced [Ca2+]i changes, reduced hIAPP-triggered ER stress and improved cell viability.
KeywordsAmylin Amyloid Apoptosis Calcium IAPP Plasma membrane Trpv4
intracellular calcium concentration
fetal bovine serum
human islet amyloid polypeptide
islet amyloid polypeptide
rodent islet amyloid polypeptide
small interfering RNA
transient receptor potential
TRP cation channel subfamily V member 4
Islet amyloid polypeptide (IAPP), also known as amylin, is the main component of islet amyloid [1, 2], which is found within the pancreatic islets of Langerhans in patients with type 2 diabetes mellitus [3, 4]. Because deposits of islet amyloid colocalise with areas of cell degeneration, the process of amyloidosis has been strongly associated with the progressive loss of pancreatic beta cell mass that is one of the hallmarks of type 2 diabetes mellitus [5, 6, 7, 8]. A target region of human IAPP (hIAPP) between positions 20 and 29 is thought to be responsible for the formation of amyloid fibrils . Rodent IAPP (rIAPP), which differs from the human sequence at several residues within this critical region, is not amyloidogenic .
The process of ordered aggregation into amyloid fibrils is a common event in many cell-degenerative diseases, including Alzheimer’s disease, Parkinson’s disease, prion protein related encephalopathies and several other local and systemic amyloidoses . Although the peptides that form amyloid deposits show no apparent primary sequence homology, they all adopt a beta strand secondary structure after conversion to fibrils , suggesting that partially folded states serve as the precursors for nucleation [12, 13, 14]. However, the toxicity of amyloidogenic peptides appears to lie in the oligomeric intermediates rather than the mature fibrils [15, 16]. Thus, the conversion of a normally soluble protein into amyloid can give rise to a gain in toxic function, providing a common mechanism by which amyloid contributes to disease pathogenesis through cell toxicity and cell death .
Amyloid peptides have been reported to induce cell death by apoptosis [6, 18, 19, 20], which may result from the disruption of calcium homeostasis . Indeed, in the MIN6 mouse pancreatic beta cell line, we recently showed that hIAPP triggered calcium changes that were associated with the induction of apoptosis via a pathway involving activation of the endoplasmic reticulum (ER) stress response . The elevation of intracellular calcium in neurons and astrocytes has been reported to follow the formation and extracellular deposition of several amyloid peptides, including hIAPP [23, 24, 25, 26, 27], and a range of potential underlying mechanisms for the increased calcium permeability have been proposed. In cell-free systems such as phospholipid bilayers, hIAPP amyloid has been reported to form cation-permeable channels as a result of its insertion into membranes , but also to destabilise the membrane structure in parallel with the measured kinetics of hIAPP fibril growth  and to cause a non-specific increase in conductance [16, 30, 31, 32, 33]. In addition, however, activation of cell-surface receptors coupled to calcium influx has been reported in a neural cell line exposed to beta-amyloid peptide aggregation . Although diverse, these mechanisms are compatible with the idea that the cytotoxicity of hIAPP is initiated on the cell surface and requires close contact between the aggregation process and the beta cell plasma membrane . In regions where this occurs, the plasma membrane of pancreatic beta cells in amyloidogenic islets in vivo is usually characterised by ultrastructural morphological abnormalities [2, 35, 36, 37].
The aim of the present study was to investigate the mechanism by which hIAPP fibril formation results in calcium influx across the plasma membrane of pancreatic beta cells and its association with apoptosis.
The mouse pancreatic beta cell line MIN6  was cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS), 2 mmol/l l-glutamine, 5 μmol/l β-mercaptoethanol, 100 U/ml penicillin and 100 μg/ml streptomycin, at 37°C with 5% CO2.
Islet isolation and culture
Islets were isolated by collagenase (Roche Diagnostics, Indianapolis, IN, USA) digestion of the pancreas, and separated from exocrine tissue on a Histopaque (Sigma-Aldrich, Steinheim, Germany) density gradient. Hand-picked islets were transferred to RPMI-1640 medium (Gibco-BRL, Paisley, UK) containing 11.1 mmol/l glucose and supplemented with 10% FBS, 2 mmol/l l-glutamine, 100 U/ml penicillin and 100 μg/ml streptomycin, and cultured at 37°C with 5% CO2. All procedures were carried out with male mice on a C57BL/6 background and in accordance with the Institutional Animal Ethics Committee’s policies.
Solutions and chemicals
Synthetic human and rat IAPP peptides (Bachem, Bubendorf, Switzerland) were dissolved in sterile water at 500 μmol/l and incubated at room temperature for 10 min before use at a final concentration of 10 μmol/l. Drugs and chemicals were purchased from Sigma-Aldrich (Poole, UK) unless otherwise stated, and were added to the bath solutions as indicated. The bath solution for calcium imaging experiments was a Krebs–Ringer bicarbonate solution containing (mmol/l): 149.5 NaCl, 6.5 KCl, 31.2 NaHCO3, 2.7 MgCl2, 4.3 CaCl2 and 26 HEPES (pH 7.4), to which 1.3% (wt/vol.) BSA and 1 mmol/l glucose were added. The Na+-free buffer contained choline+ in place of Na+.
Transmission electron microscopy
For ultrastructural analysis of insoluble aggregates, a 4 μl sample of the same suspension was applied to a Formvar carbon-coated copper grid for 1 min, dried, stained with 2% (wt/vol.) uranyl acetate in water for 1 min and air dried. For cellular ultrastructural analysis, MIN6 cells in culture were fixed in 2.5% (wt/vol.) glutaraldehyde in 0.1 mol/l phosphate buffer (pH 7.4) at 4°C for 1.5 h, rinsed in 0.1 mol/l phosphate buffer and post-fixed in osmium tetroxide 1% (wt/vol.) in the same buffer for 1 h. The samples were dehydrated in an ascending series of acetone concentrations, embedded in Spurr resin and polymerised at 60°C for 48 h. Ultra-thin sections (60-90 nm) were cut using a Reichert-Jung Ultracut E ultramicrotome (Leica Instruments, Nussloch, Germany). They were then placed on 200 mesh copper grids and doubled stained with uranyl acetate (2% wt/vol.) and lead citrate. The grids were examined using a Jeol 1010 transmission electron microscope (Jeol, Tokyo, Japan) operating at an accelerating voltage of 80 kV.
Thioflavin T (ThT) fluorescence assay
The development of hIAPP fibrils was monitored by a ThT fluorescence assay. Each reaction was performed with 5 μmol/l ThT in assay buffer (50 mmol/l TRIS, 100 mmol/l KCl, 1 mmol/l EDTA, pH 8.0) containing 10 μmol/l freshly reconstituted hIAPP or rIAPP peptide. Measurements were recorded at 37°C by using a microplate spectrofluorometer (SpectraMax GeminiXS; Molecular Devices, Sunnyvale, CA, USA). Excitation and emission wavelengths were set at 440 and 482 nm, respectively.
Intracellular calcium concentration [Ca2+]i measurements
Cells were plated on glass-bottom dishes (Mattek, Ashland, MA, USA) 1–3 days prior to use and loaded with 2 μmol/l fura-2 as the acetoxymethylester (Molecular Probes, Leiden, the Netherlands) for 30 min in bath solution containing 1 mmol/l glucose at room temperature. Measurements were made on an inverted Fluorescence Microscope (Olympus IX71, Southall, UK) with a ×40 oil-immersion objective. Excitation at 340 and 380 nm was achieved using a combination of a 75 W Xenon arc lamp and a monochromator (Cairn Research, Faversham, UK) controlled by MetaFluor software (Molecular Devices, Wokingham, UK) and emission was recorded with an Orca ER CCD camera (Hamamatsu Photonics UK, Welwyn Garden City, UK) using a dichroic mirror and a 510 nm long pass filter. Minimal and maximal signals were recorded in the presence of 5 μmol/l ionomycin in 5 mmol/l EGTA/0 mmol/l Ca2+ and 5 mmol/l Ca2+, respectively, at the end of the experiment.
Small interfering RNA (siRNA) duplexes for mouse transient receptor potential cation channel subfamily V member 4 (Trpv4) and amylase 2 (Amy2) were obtained from Dharmacon (Lafayette, CO, USA). A total of 170 nmol/l of each siRNA duplexes was transfected into MIN6 cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Cells were processed and analysed after 48 h of transfection. Cells treated with siRNA against Amy2 (siAmy2) were used as control. Prior to siRNA experiments, transfection optimisation to ensure maximum siRNA uptake was performed with siGLO green fluorescent control reagent (Dharmacon). Transfection efficiency expressed as the percentage of green fluorescent-positive cells was 91 ± 6% (mean ± SEM from four independent experiments).
Total RNA was extracted using TRI reagent (Sigma-Aldrich) following the manufacturer’s instructions. Thereafter, cDNA was synthesised from 1 μg total RNA using SuperScript II RNase H reverse transcriptase (Invitrogen). PCR was carried out with 50 ng cDNA and specific oligonucleotide primers using Taq DNA polymerase (Promega, Madison, WI, USA). Primers for Trpv4 were 5′-ATCAACTCGCCCTTCAGAGA-3′ and 5′-GGTGTTCTCTCGGGTGTTGT-3′, for β-actin (Actb) were 5′-TGAGAGGGAAATCGTGCGTG-3′ and 5′-TGCTTGCTGATCCACATCTGC-3′, and for analysis of mouse X-box binding factor-1 (Xbp1) mRNA splicing were 5′-GAACCAGGAGTTAAGAACACG-3′ and 5′-AGGCAACAGTGTCAGAGTCC-3′. PCR products were separated by electrophoresis on a 1.5-2.5% (wt/vol.) agarose gel and visualised by ethidium bromide staining. Unspliced and spliced Xbp1 (sXbp1) gave products of 205 and 179 bp, respectively. The percentage of sXbp1 to total Xbp1 was determined by densitometry using Quantity One software (Bio-Rad Laboratories, Hercules, CA, USA).
Real-time quantitative RT-PCR
The heat shock protein 90 kDa beta member-1 (Hsp90b1) and Actb genes were amplified using SYBR Green PCR Core Reagents (Applied Biosystems, Warrington, UK) with primers 5′-CACACTAGGTCGTGGAACAACA-3′ and 5′-TCTGTCTTGCTACTCCACACGT-3′, and 5′-GATCTGGCACCACACCTTCTACA-3′ and 5′-CGTCACCGGAGTCCATCACAA-3′, respectively. PCR was run with 1 ng cDNA using an ABI Prism 7900HT Sequence Detection System (Applied Biosystems) following the manufacturer’s instructions. A standard curve of each primer set was generated from serial dilutions of cDNA. PCR products were verified using dissociation curve analysis after the run using SDS software (Applied Biosystems). Actb values were used to normalise Hsp90b1 expression.
MIN6 cells were harvested and pooled with their medium in order to evaluate also cells that had lost their adherent properties during apoptosis. The protocol for isolation of single islet cells was as published previously . Islets were digested in PBS containing 0.125 mg/ml trypsin and 0.05 mg/ml EDTA at 37°C. The cell suspension was cycled for 5 min on ice to allow islets to sediment. The supernatant fraction containing single cells was removed and placed in FBS. To obtain additional single islet cells, the digestion process was repeated three times. Thereafter, cells were stained with annexin V-FITC and propidium iodide (PI) using an Annexin V-FITC Apoptosis Detection Kit (Becton Dickinson, San Jose, CA, USA). Samples were analysed by FACS within 1 h on a FACS Calibur with Cell Quest software (Becton Dickinson). FACS gating based on forward and side scatter of 20,000 cells was included for analysis. Annexin V-FITC- and PI-negative cells were considered as viable, annexin V-FITC-positive cells as early apoptotic and those positive for both annexin V-FITC and PI as late apoptotic or dead cells.
Data are presented as means ± SEM. Comparisons between groups were performed by the Student’s t test. Differences in frequencies between groups were assessed by the χ2 test.
The results raised the possibility that a non-voltage-dependent Ca2+ channel, such as a transient receptor potential (TRP) channel, might be implicated in responses to hIAPP, and we speculated that a mechanosensitive member of the TRP channel family might play a role. To test this hypothesis, we first investigated the effect of the relatively non-specific TRP channel inhibitor, gadolinium (Gd3+, 100 μmol/l). In support of our hypothesis, addition of Gd3+, in the presence of Na+-free buffer and nifedipine, abolished the calcium responses to hIAPP (0.3% responders, 1/311 cells, p < 0.001 compared with hIAPP in Na+-free buffer without Gd3+; Fig. 3f).
We next examined our microarray database  for mechanosensitive TRP channels expressed in MIN6 cells, and the search identified the Ca2+-permeable channel Trpv4 as a possible candidate. To investigate the potential role of TRPV4, we tested the effect of ruthenium red (RR), a general inhibitor of TRPV channels. In both standard buffer and Na+-free buffer, RR (10 μmol/l) abolished the effect of hIAPP on [Ca2+]i (only 0.6% of cells responded in the standard bath and 1% responded in Na+-free buffer, p < 0.001 compared with hIAPP in standard or Na+-free buffer without RR, respectively; Fig. 3g,h).
To study the downstream effect of hIAPP cytotoxicity, we measured apoptosis by FACS analysis, assessing staining for both PI and annexin-V. Cells incubated for 24 h in the presence of hIAPP exhibited only 68% viability compared with control cells (compared with 93% viability using rIAPP; Fig. 4b). Addition of RR in the presence of hIAPP improved cell viability to 79% (p = 0.006), representing a 45% reduction in hIAPP-induced cell death (Fig. 4b). We next investigated whether we could confirm the key observed effects of RR on hIAPP cytotoxicity in primary cultures of mouse pancreatic islets (Fig. 4d). After 72 h treatment with or without 10 μmol/l hIAPP, islets were disaggregated in single cells before viability was analysed by FACS. Cell viability in treated islets decreased significantly to 22% compared with 73% in the controls (p < 0.001; Fig. 4c). Treatment with RR in the presence of hIAPP improved islet cell viability to 45% (p < 0.05), representing a 52% reduction in hIAPP-induced apoptosis (Fig. 4c).
We also investigated whether silencing the Trpv4 gene had a protective effect on hIAPP-induced apoptosis, by measuring cell viability and apoptosis after 24 h of hIAPP treatment. Cell viability was ~78% in control cells transfected with either siAmy2 or siTrpv4. Whilst hIAPP reduced viability in the control siAmy2 cells to 26%, it only decreased viability to 52% in the siTrpv4 cells (p = 0.002 for comparison between siAmy2 and siTrpv4 cells each treated with hIAPP). This represents a 50% reduction in hIAPP-triggered cell death in the cells with reduced expression of Trpv4 (Fig. 5e).
Recent findings revealed that the hIAPP-triggered calcium elevation and apoptosis in MIN6 cells were associated with activation of the ER stress response . To attempt to distinguish between the possibilities that hIAPP-induced ER stress might be either a trigger for, or a consequence of, the observed [Ca2+]i changes, we assessed the degree of ER stress in siTrpv4-treated cells, as these exhibited significantly reduced hIAPP-triggered [Ca2+]i changes. ER stress was quantified by measuring the activation of Xbp1. The spliced active form of Xbp1 encodes a key transcription factor for upregulation of ER chaperones during ER stress responses , and has been shown to be important for maintenance of pancreatic beta cell ER function . Splicing of Xbp1 mRNA in MIN6 cells under control conditions was 23 ± 2% of the total Xbp1 (Fig. 5f). Exposure of siAmy2-treated cells to 10 μmol/l hIAPP for 2 h significantly increased the levels of sXbp1 to 41 ± 3% (p < 0.05 vs basal condition), whereas no increment was observed in cultures treated with siTrpv4 (21 ± 2% of total Xbp1; Fig. 5f). Twenty-four hours after treatment, the increase of sXbp1 product had returned to near basal levels in siAmy2-and remained normal in siTrpv4-treated cells (30 ± 4% and 21 ± 0.4%, respectively; Fig. 5f). We also studied the expression level of Hsp90b1, which encodes a resident ER luminal molecular chaperone, known to be elevated by the activity of sXbp1 as a consequence of ER stress . siAmy2-treated cells incubated in the presence of hIAPP showed significantly increased expression of Hsp90b1 after both 2 and 24 h (p = 0.02 and p = 0.01 at 2 and 24 h, compared with untreated siAmy2 cells). In cells treated with siTrpv4 and incubated with hIAPP, expression of Hsp90b1 was normal at 2 and 24 h (Fig. 5g). These data represent a significant reduction in hIAPP-triggered ER stress in the cells with reduced expression of Trpv4 (Fig. 5f,g), which is in agreement with the observed protection against hIAPP-triggered cell death (Fig. 5e).
Our results show that hIAPP triggers calcium changes in a model of the pancreatic beta cell, the MIN6 cell line. Calcium elevation was progressive and severe, and frequently resulted in cell detachment. The calcium changes corresponded in time with the appearance of hIAPP aggregates in the solution, alterations in the surface membrane morphology, and a reduction in cell viability. Pharmacological and siRNA experiments suggested that activation of both TRPV4 and L-type calcium channels contributed to the hIAPP-triggered calcium rise.
The synthetic IAPP peptide concentration used in the present study is within the range of concentrations used in the literature to spontaneously form IAPP oligomers and fibrils in culture [6, 9, 16, 22]. In type 2 diabetes mellitus, IAPP fibrils are often found accumulated between beta cells and islet capillaries [2, 35, 36, 37], indicating that there may be insufficient clearance of secreted IAPP, and making it technically difficult to estimate in vivo the IAPP concentrations that accumulate locally throughout the years of disease progression. Nevertheless, we showed here that the ultrastructural morphology of MIN6 cells in the amyloidogenic conditions was characterised by plasma membrane alterations compared with the control cells. These alterations included cell membrane irregularities and invaginations filled with fibrils, similar to those described in human islets from type 2 diabetes mellitus patients [2, 35, 36, 37].
Amyloid oligomers (including hIAPP) have been reported to increase [Ca2+]i in neurons and astrocytes [23, 24, 25, 26, 27], although the effect of hIAPP on pancreatic beta cells is not well documented. We showed recently that increases in [Ca2+]i could be detected in hIAPP-treated MIN6 cells, with associated activation of the ER stress response and induction of apoptosis . An earlier study, by contrast, suggested that hIAPP does not trigger changes in [Ca2+]i in insulinoma cells , but the conclusions were derived from measurements of [Ca2+]i in cell populations, which might have excluded any non-attached cells. Our finding that individual MIN6 cells detached soon after experiencing an hIAPP-triggered Ca2+ rise might explain the previous failure to detect Ca2+ changes in insulinoma cells.
The mechanisms previously implicated in calcium responses to amyloid oligomers in other cell types and cell-free systems include direct formation of Ca2+-permeable pores by the amyloid oligomers themselves and the activation of voltage-gated calcium channels [28, 34]. Amyloid pores are unlikely to underlie our findings in MIN6 cells, as the Ca2+ changes were inhibited by nifedipine, Gd3+ and RR, none of which has been found to block amyloid channels. The inhibition of hIAPP-triggered Ca2+ responses by nifedipine in MIN6 cells indicated, however, that L-type Ca2+ channels were activated by hIAPP, either as a result of changes in the properties of the L-type Ca2+ channels themselves, or as a secondary consequence of membrane depolarisation. The latter idea is supported by the findings in Na+-free buffer, as an additional nifedipine-insensitive Ca2+-elevating pathway was identified under these conditions. This was found to be blocked by Gd3+ or RR, indicating the likely involvement of specific members of the TRP channel family. siRNA mediated knockdown of the mechanosensitive TRP channel, TRPV4, protected against the toxic effects of hIAPP on [Ca2+]i as well as on induction of ER stress and apoptosis. Our subsequent Trpv4 knockdown experiments, using loss of cell viability as a marker of hIAPP-induced apoptosis, supported the notion that activation of this channel may be an early response to hIAPP.
TRPV4 is a mechano- and osmosensitive channel located on the cell surface, with a Ca2+/Na+ permeability ratio of ~6 , which is inhibited by Gd3+ and RR, and is not blocked by 2-aminoethyl diphenylborinate (2-APB). The observed morphological deformations in the plasma membrane following hIAPP treatment suggest that the deposition of hIAPP insoluble aggregates, which can reach high molecular masses , might weaken the mechanical strength of the plasma membrane, making it susceptible to TRPV4 activation. We propose that it may provide a link between hIAPP-induced mechanical perturbations of the beta cell plasma membrane by hIAPP and the depolarising Ca2+ influx. The finding that siTrpv4 protected against both hIAPP-induced ER stress and [Ca2+]i changes suggests that this channel is involved at a step that precedes the ER stress response, perhaps at the level of [Ca2+]i modulation. However, as RR did not protect completely against amyloid cytotoxicity, our data cannot exclude the additional involvement of other early mechanisms in the induction of islet amyloid cytotoxicity. Activation of TRPV4 may play a more widespread role in mediating some of the effects of amyloid on [Ca2+]i that have been previously described in the literature. Amyloid-induced Ca2+ entry has, for example, been attributed to activation of channels with a relatively high selectivity for Ca2+ over Na+ , which are insensitive to L-type Ca2+ channel blockers . The potential involvement of TRPV4 and alternative mechanosensitive channels in mediating amyloid-induced Ca2+ changes, ER stress and apoptosis in other cell types will require further evaluation.
This study was sponsored by a Sardà Farriol Research Program, and supported by FIS (PI05/1215), REDIMET (RD06/0015) and CIBER de Diabetes y Enfermedades Metabólicas (CIBERDEM) initiatives of ISCIII from the Ministerio de Sanidad y Consumo, and by SGR2005/00019 grant from the Generalitat de Catalunya in Spain. S. Casas is a recipient of a Juan de la Cierva contract from the Ministerio de Educación y Ciencia in Spain and acknowledges the award of an Albert Renold Fellowship from the European Foundation for the Study of Diabetes. F. M. Gribble is supported by the Wellcome Trust, and F. Reimann by St John’s College, Cambridge.
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
- 1.Westermark P, Wernstedt C, Wilander E, Hyden DW, O’Brien TD, Johnson KH (1987) Amyloid fibrils in human insulinoma and islets of Langerhans of the diabetic cat are derived from a neuropeptide-like protein also present in normal islet cells. Proc Natl Acad Sci USA 84:3881–3885PubMedCrossRefGoogle Scholar
- 22.Casas S, Gomis R, Gribble FM, Altirriba J, Knuutila S, Novials A (2007) Impairment of the ubiquitin-proteasome pathway is a downstream endoplasmic reticulum stress response induced by extracellular human islet amyloid polypeptide and contributes to pancreatic beta-cell apoptosis. Diabetes 56:2284–2294PubMedCrossRefGoogle Scholar
- 31.Knight JD, Miranker AD (2004) Phospholipid catalysis of diabetic amyloid assembly. J Biol Chem 341:1175–1187Google Scholar
- 32.Green JD, Kreplak L, Goldsbury C et al (2004) Atomic force microscopy reveals defects within mica supported lipid bilayers induced by the amyloidogenic human amylin peptide. J Biol Chem 342:877–887Google Scholar
- 40.NCBIs Gene Expression Omnibus series accession number GSE2253. Available from http://www.ncbi.nlm.nih.gov/geo/, accessed 12 April 2005