JMJD5 is a human arginyl C-3 hydroxylase

Abstract

Oxygenase-catalysed post-translational modifications of basic protein residues, including lysyl hydroxylations and N^ε-methyl lysyl demethylations, have important cellular roles. Jumonji-C (JmjC) domain-containing protein 5 (JMJD5), which genetic studies reveal is essential in animal development, is reported as a histone N^ε-methyl lysine demethylase (KDM). Here we report how extensive screening with peptides based on JMJD5 interacting proteins led to the finding that JMJD5 catalyses stereoselective C-3 hydroxylation of arginine residues in sequences from human regulator of chromosome condensation domain-containing protein 1 (RCCD1) and ribosomal protein S6 (RPS6). High-resolution crystallographic analyses reveal overall fold, active site and substrate binding/product release features supporting the assignment of JMJD5 as an arginine hydroxylase rather than a KDM. The results will be useful in the development of selective oxygenase inhibitors for the treatment of cancer and genetic diseases.

Fig. 1

JMJD5 catalyses stereoselective C-3 hydroxylation of Arg137 of the 40S ribosomal protein S6 (RPS6). a MALDI-MS spectra for RPS6_129–144 showing the +16 m/z shift following treatment with JMJD5 and cofactors. b MS/MS analysis of the modified RPS6_129–144 indicating a +16 shift on y and b ion fragments y₈–y₁₄/b₉–b₁₅, which is not observed for ion fragments y₁–y₇/b₂–b₈, implying hydroxylation at RPS6 R137; observed/predicted masses for the fragment ions are given in Supplementary Fig. 5. c Plot of succinate formation monitored by ¹H-NMR. Succinate formation from a reaction with 10 μM JMJD5, 50 μM RPS6, 200 μM 2OG and 100 μM Fe(II) (black) is compared to that without RPS6 peptide (blue), with ascorbate (red), and with an inactive H321A variant substituting for wild type (yellow). Results are the mean ± s.e.m. (n = 3). d JMJD5-catalysed arginyl hydroxylation. e Extracted ion chromatograms (m/z = 345) from LC-MS analysis of synthetic (3R)- and (3S)-hydroxyarginine (HyR) standards and amino-acid hydrolysates from unmodified/hydroxylated RPS6_129–144. Peaks at 1.93 and 2.02 min correspond to (3R)- and (3S)-hydroxyarginine, respectively. f¹H–¹³C HSQC spectra of hydroxylated RPS6_129–144 (VPRRLGPKR^OHASRIRKL), labelled with aa assignments, supporting R137 C-3 hydroxylation. Red boxes highlight the hydroxyarginine (R^OH). Inset bottom right: ¹H–¹H COSY spectrum showing the correlation between C-2-R^OH and C-3-R^OH protons. ♯ Indicates residual HEPES buffer

Fig. 2

Overview of the JMJD5 structure. Topology (a) and ribbon representation (b) of the JMJD5.Mn.2OG complex and the active site close-up (c) (PDB: 6F4S). DSBH secondary structure elements are labelled I–VIII (green). The JMJD5 fold contains 7 α-helices, 13 β-strands and 3 3₁₀ helices. The N-terminal domain preceding the DSBH is in orange and the remainder of the structure, including the DSBH βIV–V insert, in light green. 2OG is shown by cyan sticks and the metal by a purple sphere

Fig. 3

Stereoviews of JMJD5.substrate complexes showing likely productive and non-productive arginine-binding modes. a JMJD5.Mn.NOG.RPS6 (Complex 1, PDB: 6F4P), b JMJD5(Q275C).Mn.NOG.RPS6(A138C) (Complex 2, PDB: 6F4Q), c JMJD5(N308C).Mn.NOG.RCCD1 (Complex 3, PDB: 6F4R), d JMJD5(N308C).Mn.2OG.RCCD1 (Complex 4, PDB: 6F4S) and e JMJD5(W414C).Mn.NOG.RCCD1 (Complex 5, PDB: 6F4T). The target arginine in both RPS6 (R137) and RCCD1 (R141) binds in a shallow channel on the JMJD5 surface forming hydrophobic interactions with Y243/W310 and hydrogen bonds with E238/ S318 in both 'productive' (Complex 1/A, Complex 3 and Complex 4) and 'non-productive' (Complex 1/B, Complex 2 and Complex 5) conformations. 2F_o−F_c (grey meshes) for JMJD5 and difference electron density (F_o−F_comit map, blue) for substrates are contoured to 1.2−1.5σ and 3σ, respectively. Red arrows indicate hydroxylated C–H bonds

Fig. 4

Conformational changes observed in complexes representing steps in JMJD5 catalysis. a Proposed mechanism of JMJD5-catalysed (3R)-Arg-hydroxylation (reactive intermediates in parentheses), b JMJD5.Mn.2OG, c JMJD5.Mn.NOG.RCCD1(139-142) (Complex 3), d JMJD5.Mn.succinate.RCCD1^(2S,3R-OH) (product) modelled complex, e JMJD5.Mn.succinate and f apo-JMJD5. The figure illustrates global and local changes occurring on substrate/cosubstrate binding and product release, which is a characteristic feature of JMJD5 catalysis. Two mobile regions are involved in JMJD5 catalysis: (i) the β3–β4 loop (N-terminal to the DSBH, aa 234–254, raspberry to cream), and (ii) the DSBH βIV–βV loop (aa 342–381, different blue shades). Analogous loops are often involved in substrate binding by other JmjC-hydroxylases and KDMs, including FIH, RIOX, KDM4A^, and PHF8. The β3-β4 loop, which is often relatively short as in some KDMs, is 'open' in the JMJD5.Mn.2OG-, 'partly open' in the substrate-, and disordered in the succinate-complexes. In all JMJD5 structures, except for apo-JMJD5, the βIV–βV loop adopts a similar conformation (i.e., partly helical); however, in the apo structure, part of βIV–βV loop (aa 351–361) adopts a hairpin conformation. Note the conformational changes in the sidechains of active site residues involved in catalysis. The JMJD5.Mn.succinate.RCCD1^(2S,3R-OH) complex was modelled using the JMJD5(N308C).RCCD1 (Complex 3) and JMJD5.succinate structures as templates

Fig. 5

Relationship of JMJD5 structure/topology with those of homologous JmjC-hydroxylases and JmjC KDMs. a–f Ribbon representations of JMJD5.Mn.NOG.RCCD1(139–142) (PDB: 6F4R) (a), TYW5.Ni (PDB: 3AL5) (b), FIH.Fe.NOG.HIF-1α (786–826) (PDB: 1H2K) (c), ycfD_RM.Mn.NOG.L16(72–91) (PDB: 4CUG) (d), PHF8.Fe.NOG.H3K4me₃K9me₂(2–25) (PDB: 3KV4) (e) and KDM4A.Ni.NOG.H3K9me₂(7–14) (PDB: 2OX0) (f). Middle panels show topologies as indicated, with DSBH core elements (βI–βVIII) in green, helices in cyan, additional β-strands in red, random coils in black, and the insert between the fourth and fifth DSBH β-strands in blue. JMJD5 catalysis involves substantial conformational changes in the β3–β4 and βIV–V loop regions (loops analogous to JMJD5 β3–β4 loop are labelled 'β3–β4 loop' in other JmjC-oxygenases); such movements have not been observed around the JmjC KDM active sites on N^ε-methylated lysine binding. In addition, JMJD5 lacks (i) the extended flexible loop region, immediately N-terminal to βI (yellow), which is involved in binding N^ε-methylated lysines (Kme_n) and (ii) the chromatin- and Zn-binding domains present in most JmjC KDMs (note, structures of PHF8 and KDM4A are only of catalytic domains). Thus, the overall JMJD5 fold together with its unique substrate binding features supports its assignment as a JmjC-hydroxylase and not a KDM

Fig. 6

Comparison of the active site substrate binding modes for JMJD5 and related JmjC enzymes. Red/blue arrows indicate hydroxylation/demethylation sites, respectively. Active-site metals (Fe or Fe surrogates, Mn/Ni) are colour-coded spheres. Values represent distances (Å) from the metal to the oxidised carbon. a ycfD_RM.Mn.NOG.L16(72–91) (PDB: 4CUG), b NO66.Mn.NOG.RPL8(205–224) (PDB: 4CCO), c PHF8.Fe.NOG.H3K4me₃K9me₂(2–25) (PDB: 3KV4), d KDM4A.Ni.NOG.H3K9me₂(7–14) (PDB: 2OX0), e FIH.Fe.NOG.HIF-1α (786–826) (PDB: 1H2K), f JMJD5.Mn.NOG.RCCD1(139–142) (PDB: 6F4R). Note variations in the N- to C-substrate directionality in the active site cleft (indicated by black arrows), and variations in the hydroxylation site relative to the metal (boxed). JMJD5 binds substrates (f) with the same substrate N/C-directionality, as for most KDMs (KDM4A, KDM6B and KDM6A, d), but differing from that for RIOX (a, b) and for one KDM, PHF8 (and likely other KDM2/7 subfamily members, c). Overall, the mode of JMJD5-catalysed hydroxylation more closely resembles that of FIH (e) compared to other JmjC-hydroxylases and KDMs, consistent with the proposal that the FIH/JMJD5 subgroup of JmjC-hydroxylases evolved from prokaryotic RIOX including by loss of the 'Winged helix' (WH) domains, which are located at the C-terminus of all RIOX^,