• 2018-07
  • 2019-04
  • 2019-05
  • 2019-06
  • 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2019-12
  • 2020-01
  • 2020-02
  • 2020-03
  • 2020-04
  • 2020-05
  • 2020-06
  • 2020-07
  • 2020-08
  • 2020-09
  • Peptides undergoing only a single hydroxylation i


    Peptides undergoing only a single hydroxylation, i.e. PA-Hyp-KPAPK and PAPK-Hyp-APK, were then used to investigate the kinetics of vCPH co-factor and co-substrate dependences. The values for KMapp for Fe(II) and KM for 2OG for the two peptide substrates were within error (KMapp, Fe(II): 0.9 μM ± 0.1 μM for PA-Hyp-KPAPK and 1.1 μM ± 0.1 μM for PAPK-Hyp-APK. KM, 2OG: 29.5 μM ± 8.7 μM for PA-Hyp-KPAPK and 27.4 μM ± 6.1 μM for PAPK-Hyp-APK (Supplementary Fig. 9)). The KMapp for l-ascorbate differs substantially for the two peptides, with measured values of 1.7 μM ± 0.3 μM for PA-Hyp-KPAPK and 13.8 μM ± 4.4 μM for PAPK-Hyp-APK, implying a difference in l-ascorbate dependency for the two different hydroxylation sites. When vCPH was incubated with Fe(II), 2OG, and the peptide substrate, uncoupled turnover of 2OG to succinate occurred at a relatively slow rate, as observed by 1H NMR spectroscopy. The rate of uncoupled 2OG turnover was enhanced by addition of l-ascorbate. With Fe(II), 2OG and PAPKPAPK, 2OG turnover significantly increased and the reaction rapidly proceeded to completion under our standard conditions (Supplementary Fig. 10). In the presence of Fe(II), 2OG and l-ascorbate, PA-Hyp-KPAPK and PAPK-Hyp-APK had similar impacts on vCPH-catalysed 2OG turnover. With PA-Hyp-KPAPK, the absence of l-ascorbate had little impact on 2OG turnover. However, with PAPK-Hyp-APK in the absence of l-ascorbate, 2OG turnover was lower (but above uncoupled turnover levels) than with l-ascorbate. Consistent with MS-based kinetic studies (Supplementary Fig. 10D), these observations suggest that Fmoc-Trp-OH the l-ascorbate dependency of the hydroxylation of PAPK-Hyp-APK is greater than for PA-Hyp-KPAPK. Thus, at least under our assay conditions, the ascorbate requirement for vCPH-mediated hydroxylation appears to be sequence context dependent. The afore-described studies enabled the identification of conditions suitable for MS-based analysis of potential vCPH inhibitors, i.e. vCPH (1 µM), (NH₄)₂Fe(SO₄)₂ (50 µM), sodium l-ascorbate (100 µM), 2OG (disodium salt, 30 µM), PA-Hyp-KPAPK (10 µM), HEPES pH 7.5 (50 mM). Since the active sites of the prolyl hydroxylases for which structures are available are closely related (Supplementary Fig. 11),10, 13, 29 known PHD, C-P4H and broad-spectrum 2OG oxygenase inhibitors were tested as vCPH inhibitors (Fig. 3). Following initial screening at a single concentration by differential scanning fluorimetry (DSF) and MS (Supplementary Fig. 12), IC50 values were obtained for compounds showing activity at 100 µM (Supplementary Fig. 13). The most potent vCPH inhibitors, with single-digit micromolar IC50 values (with 1 µM vCPH), were Vadadustat30, 31 (IC50 < 1 µM), IOX4 (IC50 = 2.5 ± 0.2 µM), 2,4-pyridine dicarboxylic acid (2,4-PDCA, IC50 = 5.3 ± 0.3 µM), Roxadustat (IC50 = 5.0 ± 0.1 µM) and FG-2216 (IC50 = 4.5 ± 0.1 µM). All these compounds, with the exception of the broad-spectrum 2OG oxygenase inhibitor 2,4-PDCA,1, 3 were initially developed as HIF PHD inhibitors. Notably, despite being reported as selective for the HIF PHDs over human C-P4H, Roxadustat was found to be a strong inhibitor of vCPH. Roxadustat, FG-2216 and Vadadustat targeted vCPH more potently than FG-0041 (IC50 = 15.7 ± 1.4 µM), which was initially reported as a C-P4H inhibitor. IOX2 (IC50 = 8.5 ± 0.2 µM), a functional probe compound for the HIF PHDs, was also found to be a vCPH inhibitor. Two other reported PHD inhibitors, Daprodustat (IC50 = 13.4 ± 0.9 µM) and Molidustat (IC50 = 26.8 ± 2.4 µM), inhibited vCPH less potently. In the case of Molidustat (IC50 = 26.8 ± 2.4 µM) this is somewhat in contrast to the greater vCPH inhibition potency exhibited by the structurally related compound IOX4 (IC50 = 2.5 ± 0.2 µM), implying selectivity tuning for this series may be viable. These observations indicate that, although the clinical PHD inhibitors are vCPH inhibitors, there is not a direct correlation between vCPH and PHD2 potency, consistent with active site differences between the two enzymes (Supplementary Fig. 11).