Pro-islet amyloid polypeptide (proIAPP) is the precursor to islet amyloid polypeptide (IAPP), a normally soluble 37-residue peptide that is co-secreted with insulin from the β-cells of the pancreas. Pancreatic amyloid composed of IAPP and β-cell loss are features of Type 2 diabetics. The deposition of IAPP amyloid adjacent to β-cells and the observation that synthetic IAPP is cytotoxic to cultured β-cells, has implicated the formation of amyloid from IAPP with the loss of β-cells, which can result in an inability to produce insulin. Recently, proIAPP has also been shown to form amyloid in vitro, but is less amyloidogenic than IAPP. Moreover, although IAPP is the major constituent of pancreatic amyloid, an incompletely processed form of proIAPP with an uncleaved N-terminal has been detected in amyloid deposits of Type 2 diabetics. This suggested that proIAPP may have a role in the formation of amyloid in Type 2 diabetes.
Optimization of proIAPP purification
ProIAPP has been previously overexpressed in E. coli fused to a thioredoxin and histidine tag (trx-proIAPP) and a purification protocol had been developed. This purification protocol entailed the purification of trx-proIAPP using Ni-NTA chromatography under denaturing conditions in 6 M urea, and then RP-HPLC purification step, prior to enterokinase cleavage to remove the proIAPP from the fusion tag and then a further RP-HPLC purification step to separate proIAPP from the fusion tag.
A variety of different purification protocols were trialled to optimise the yield of proIAPP. The optimised purification protocol involved purifying the trx-proIAPP in 6 M guandinium hydrochloride, rather than 6 M urea, as guanidium hydrochloride was able to solubilize twice as much trx-proIAPP from the E. coli cell pellet compared to 6 M urea. The trx-proIAPP was then refolded with a glycerol-containing buffer while bound to the Ni-NTA column and eluted under native conditions. ProIAPP was cleaved from the fusion tag using enterokinase and the proIAPP separated from this tag using another Ni-NTA chromatography step under native conditions. This purification protocol resulted in purification of approximately 20 mg proIAPP from one litre of E. coli culture.
ProIAPP amyloid formation
IAPP has been studied extensively, however proIAPP, possibly due to the difficulty of synthesising and purifying it, has not. Therefore, the biophysical characteristics of proIAPP were investigated. The secondary structure of proIAPP was assessed by far-UV CD spectropolarimetry and this showed that proIAPP was a natively unfolded protein like IAPP. Analysis of the far-UV CD spectra of proIAPP indicated that proIAPP belonged to the pre-molten globule class of natively unfolded proteins rather than the coil-like group of unfolded proteins. Freshly redissolved proIAPP moderately enhanced the fluorescence of the hydrophobic probe, bis-ANS, indicating that it exhibited some exposed hydrophobic surface, which was consistent with its classification as a pre-molten globule type of natively unfolded protein. Fibrillar proIAPP enhanced the fluorescence of bis-ANS more than freshly redissolved proIAPP, indicating that proIAPP in a fibrillar conformation exhibited greater exposed hydrophobic surface than freshly redissolved proIAPP.
ProIAPP formed amyloid fibrils that were approximately 10 nm wide and hundreds of nm long, enhanced the fluorescence of Thioflavin T and bound to Congo Red, all typical characteristics of amyloid. ProIAPP formed amyloid over 350 h, measured using a Thioflavin T fluorescence assay. SDS-PAGE analysis of proIAPP during amyloid formation, showed that the amyloid was formed from full-length proIAPP and was not a result of proIAPP being degraded to IAPP. The slow rate of proIAPP amyloid formation confirmed previous findings that proIAPP was less amyloidogenic than IAPP. The reduced amyloidogenicity of proIAPP compared to IAPP, suggests that the pro-regions of proIAPP are important in enhancing the stability of the central IAPP-containing region.
The effect of clusterin on proIAPP amyloid formation and cytotoxicity
Serum clusterin levels are elevated in human Type 2 diabetes. Clusterin has been shown to prevent the stress-induced precipitation of proteins, allowing its classification as an extracellular molecular chaperone. Clusterin inhibited proIAPP amyloid formation, measured using a Thioflavin T fluorescence assay. The presence of clusterin did not alter the lag phase of amyloid formation, but inhibited the growth phase in a dose-dependent manner. The inhibition of amyloid formation by clusterin occurred at molar ratios of clusterin:proIAPP of 1 clusterin to 25-100 proIAPP. These substoichiometric ratios of clusterin to proIAPP suggest that clusterin may have been interacting with a small proportion of proIAPP, such as the nuclei of amyloid formation.
Zone electrophoresis of clusterin and proIAPP mixtures followed by capillary transfer to a polyvinylidene fluoride membrane was carried out in attempts to find putative complexes between clusterin and soluble proIAPP species however none could be detected by amido black staining or immunoblotting for clusterin. However, the nuclei are likely to be a very small proportion of the total proIAPP and may be therefore below the limit of detection using these techniques. Far-UV CD spectropolarimetry, showed that clusterin was unable to stabilise the natively unfolded structure of proIAPP, suggesting that if clusterin was binding to the nuclei of amyloid formation, clusterin was unable to allow the dissociation of these nuclei back to a natively unfolded conformation.
Clusterin co-localised with the insoluble fraction of proIAPP during amyloid formation, measured using a dot blot assay for clusterin. Immunoelectron microscopy of clusterin and proIAPP mixtures was carried out to determine the where on the proIAPP fibril clusterin was binding. Clusterin could be seen bound to the ends of the proIAPP fibrils, this may represent one way in which clusterin could inhibit the growth phase of amyloid formation, as fibrils grow by elongation at fibril ends.
The slow rate of proIAPP amyloid formation allowed detailed studies of the cytotoxicity of proIAPP, which was assessed using the pancreatic β-cell line RIN-m5f. The effects of clusterin were then investigated on the cytotoxicity of proIAPP. The cytotoxicity of proIAPP in the absence of clusterin varied depending on the stage of amyloid formation. ProIAPP became progressively more toxic as it passed down the amyloidogenic pathway during the lag phase and was most cytotoxic at the start of the growth phase. ProIAPP then became progressively less toxic during the growth phase and once amyloid formation had gone to completion was not cytotoxic. Clusterin at concentrations comparable to that in human serum was cytoprotective in a dose-dependent manner at every stage of amyloid formation. The cytoprotective effect of clusterin on proIAPP-mediated cytotoxicity suggests that clusterin may attenuate the destruction of pancreatic β-cells mediated by amyloid formation in Type 2 diabetes.
The effects of glycosaminoglycans on proIAPP amyloid formation
Glycosaminoglycans (GAGs) have been detected localised with amyloid deposits in a wide range of amyloid diseases, including Type 2 diabetes. Recently, a heparin/heparan sulfate-binding site has been identified in the N-terminal of proIAPP. As an incorrectly processed form of proIAPP with an uncleaved N-terminal has been detected in amyloid deposits of human Type 2 diabetics, the interaction between proIAPP and GAGs may be important in the formation of amyloid in Type 2 diabetes. The effects of the GAGs: heparin, heparan sulfate and chondroiting sulfate with proIAPP were investigated.
GAGs accelerated amyloid formation from proIAPP, by decreasing the duration of the lag phase, which indicated that nucleation occurred more rapidly in the presence of GAGs, and increasing the rate of amyloid formation during the growth phase. Heparin was the most potent enhancer of amyloid formation, followed by heparan sulfate and then chondroitin sulfate. Far-UV CD spectropolarimetry of proIAPP in the presence of GAGs showed that the GAGs induced partially folded conformations of proIAPP. These partially folded conformations of proIAPP may correspond to partially folded intermediates believed to be critical for the formation of a stable nucleus, which is necessary for amyloid formation to occur. The relative ability of the different GAGs to induce a more ordered secondary structure of proIAPP correlated with the GAG-specificity of the acceleration of proIAPP amyloid formation. Heparin induced the most ordered conformation of proIAPP, which contained approximately 20% α-helix, 20% β-sheet, 20% β-turns and 40% random coil.
The binding of proIAPP to heparin was weaker in a high ionic strength buffer (10 mM sodium phosphate, 150 mM NaCl) compared to a low ionic strength buffer (10 mM sodium phosphate, pH 7.4), assessed using a heparin-agarose pull-down assay. This may be because the increase in ionic strength, shields the electrostatic interactions between proIAPP and heparin. The weaker binding of heparin to proIAPP in high ionic strength conditions correlated with slower heparin-induced conformational changes assessed by far-UV CD in high ionic strength conditions (10 mM sodium phosphate, 150 mM NaF, pH 7.4) compared with that in low ionic strength conditions (10 mM sodium phosphate, pH 7.4).
The minimal subunit of heparin necessary for proIAPP binding, based on the abilities of different heparin species to accelerate proIAPP amyloid formation, was greater than a heparin disaccharide (664.4 Da) and less than a heparin nonosaccharide (3 kDa). The half-maximal effect of heparin on proIAPP amyloid formation was 3.1 ± 1.0 µM, 1.4 ± 0.6 µM and 0.3 ± 0.1 µM for 3 kDa, 4 kDa and 17-19 kDa heparin respectively. The decreasing molar half-maximal concentrations with increasing molecular weight indicated that the larger molecular weight heparins contain more proIAPP binding sites than the smaller heparins, this suggested that heparin may bind oligomers of proIAPP or by binding multiple proIAPP monomers and by bringing them in close proximity to one another, induce the formation of oligomers.
The GAG-induced acceleration of proIAPP amyloid formation, supports the hypothesis that amyloid formation in Type 2 diabetes occurs due to increased quantities of proIAPP being secreted from the β-cell in Type 2 diabetes; proIAPP then binds to GAGs in the extracellular matrix adjacent to the β-cells via its N-terminal heparin/heparan sulfate-binding site and forms a nidus for the subsequent accumulation of secreted IAPP, which forms the majority of the deposit.