Selective inhibition of factor inhibiting hypoxia-inducible factor

Article abstract of DOI:10.1021/ja050841b

A set of four non-heme iron(II) and 2-oxoglutarate-dependent enzymes catalyze the post-translational modification of a transcription factor, hypoxia inducible factor (HIF), that mediates the hypoxic response in animals. Hydroxylation of HIF both causes it

Full text of DOI:10.1021/ja050841b

Published on Web 05/07/2005
Selective Inhibition of Factor Inhibiting Hypoxia-Inducible Factor
†
†
‡
‡
†
Michael A. McDonough, Luke A. McNeill, Melanie Tilliet, Cyril A. Papamica e¨ l, Qiu-Yun Chen,
†
†
,†
Biswadip Banerji, Kirsty S. Hewitson, and Christopher J. Schofield*
The Department of Chemistry and The Oxford Centre for Molecular Sciences, Chemistry Research Laboratory,
UniVersity of Oxford, Mansfield Road, Oxford, OX1 3TA, United Kingdom, INSA-IRCOF de Rouen,
Place E. Blondel, 76130 Mt-St-Aignan, France
Received February 9, 2005; E-mail: christopher.schofield@chem.ox.ac.uk
In many animals, cellular responses to reduced oxygen concen-
Scheme 1. Reactions Catalyzed by HIF Hydroxylases
tration (hypoxia) are mediated by an R/â-heterodimeric transcription
1
factor, hypoxia-inducible factor (HIF). Under hypoxia, HIF binds
to response elements linked to an array of genes associated with
the hypoxic response including, in humans, those associated with
angiogenesis (vascular endothelial growth factor, VEGF) and
erythropoiesis (erythropoietin, EPO). HIF-â is a constitutive nuclear
protein, but levels of HIF-R are low under normal oxygen
concentrations (normoxia) but increase in response to hypoxia.
Recent studies have identified the mechanism by which both the
levels and transcriptional activity of HIF-R are regulated by
molecular oxygen. Post-translational trans-4-hydroxylation of either
of two proline residues that form part of consensus motifs located
in the oxygen-dependent degradation domain of HIF-R is sufficient
the PHDs and FIH. Additionally, DMOG has been shown to activate
the HIF pathway in animals. However, 2 is also an inhibitor of
9
other 2OG oxygenases, including human enzymes such as pro-
collagen prolyl hydroxylase, for which it was first developed as an
inhibitor, and phytanoyl Co-A hydroxylase. Hence, it is unlikely
to be suitable for medicinal use.
Analysis of crystal structures of FIH-Fe(II)-2OG/NOG com-
plexed with fragments of the C-terminal transactivation domain of
HIF1-R suggested that it may be possible to inhibit FIH with
N-oxalyl amino acids derived from amino acids with hydrophobic
side chains.10 We therefore synthesized a set of N-oxalyl amino
acids via solution and/or solid-phase methodology (Schemes S1
and S2 in the Supporting Information).
The N-oxalyl amino acids were assayed as inhibitors of highly
purified recombinant FIH and an N-terminally truncated form of
PHD2 using fragments of human HIF-1R as substrates. Truncation
of the PHD2 was found to be necessary in order to produce soluble
highly active enzyme in Escherichia coli. The putative inhibitors
2
to target HIF-R to the von Hippel Lindau protein (pVHL), which
in turn recruits a ubiquitin ligase that mediates proteasomal
destruction. The transcriptional activity of HIF is more directly
inhibited by hydroxylation at the â-position of Asn803 in human
HIF-1R that blocks its interaction with the transcriptional coactivator
p300.
Four hydroxylases have been identified that catalyze the post-
translational modification of human HIF-R: three prolyl hydrox-
3
ylases (PHD1-3) and an asparaginyl hydroxylase, factor inhibiting
4
HIF (FIH). All belong to the family of ferrous iron and 2-oxo-
glutarate (2OG, 1)-dependent dioxygenases; most of these enzymes
bind their metal cofactor via a conserved 2His-1Asp/Glu triad of
residues; structural studies, however, reveal significant variation
in the 2OG binding site.5
(2-12) were initially screened in a fluorescence-based assay that
monitors consumption of 1 by its postincubation derivatization with
o-phenylenediamine.11 In the case of PHD2, the results indicated
that 2 was the most potent inhibitor, and that with the exception of
N-oxalyl-L-Ala (3), the degree of inhibition decreases significantly
as the size of the side chain increased. The increased potency of
the L- versus that of the D-N-oxalyl-Ala derivatives is consistent
with results obtained with crude extracts containing endogenous
Regulation of transcription by small molecules has the potential
for significant medical application, with respect to both the HIF
and other signaling pathways. One approach is via small molecules
6
or oligomers that interact directly with nucleic acids. Another is
to employ compounds that alter the activity of transcription factors
either directly or indirectly. We have been exploring HIF hydrox-
ylase inhibitors as a means of up-regulating hypoxically driven
transcripts.
2
PHD activity and with that of inhibition of pro-collagen prolyl
hydroxylases (see Supporting Information for reference).
The results for FIH were strikingly different. For all of the
enantiomeric forms of N-oxalyl amino acids tested, the D-enantiomer
was more potent (Figure 1). Further, although the N-oxalyl-D-Val
The substrate selectivity of the PHD enzymes for different HIF-R
7
isoforms has recently been investigated. Under normoxic condi-
tions, PHD2 appears to be important in regulating HIF-1R cellular
levels, but has a smaller action on HIF-2R levels. In contrast, PHD3
appears to more significantly regulate the effects of HIF-2R. Given
the differences in preference of target genes for the HIF-R isoforms,
inhibition of the individual HIF hydroxylases may be important in
modulating specific gene transcriptional activation to mediate a
therapeutic effect.8
(6) and N-oxalyl-D-Leu (8) derivatives were of a similar order of
potency to that of 2, the N-oxalyl-D-Phe (10) was significantly more
potent. The selectivity of 10 for FIH over PHD2 was confirmed
by assays measuring the release of 14CO
from labeled 2OG (1).
With increasing concentrations of 2, the activity of both FIH and
PHD2 could be entirely inhibited (Figure S1, apparent K for 2
2
i
N-Oxalylglycine (NOG, 2) and its pro-form dimethyl N-oxalyl-
glycine (DMOG) have been reported to inhibit the activity of both
versus FIH was 1.2 ( 0.3 mM). However, with 10, only FIH was
affected in the same manner; 10 was not an inhibitor of PHD2 at
1
i
mM. An apparent K of FIH of 83 ( 18 µM for 10 was obtained
†
University of Oxford.
INSA-IRCOF de Rouen.
‡
by an assay employing a HIF-1R786-826 peptide fused with
7680
9
J. AM. CHEM. SOC. 2005, 127, 7680-7681
10.1021/ja050841b CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
in a 2-oxo acid would act as an inhibitor, we then synthesized and
tested racemic 4-benzyl-2-oxo-glutarate (13). This compound was
a less potent inhibitor of FIH than was 2 but did not affect the
activity of PHD2 like 10, nor could it act as a substitute for 1 in
FIH assays where peptide hydroxylation was assayed by LC-MS.
The decreased potency of 13 can be attributed, in part, to its
increased conformational flexibility relative to 10 due to the absence
of the amide bond.
The results demonstrate that it is possible to selectively inhibit
one of the HIF-R hydroxylases, and so it may help to enable up-
regulation of specific genes controlled by the hypoxic response (e.g.,
EPO or VEGF). More generally, since ester derivatives of N-oxalyl
amino acids are active as pro-drugs in cell lines and animals,
N-oxalyl amino acids may be useful for functional assignments of
2OG oxygenases whose biochemical role is unknown, including
the JmjC family, some of whom are already known to be involved
in transcriptional regulation.12
Acknowledgment. We thank the Wellcome Trust, EU, BBRSC,
and a Glasstone Fellowship (K.S.H.) for funding. Thanks to the
CCLRC SRS Daresbury beamline 10.1, and PX group for beam-
time, and Profs. P. Ratcliffe and C. Pugh for encouragement and
discussion.
Figure 1. Effect of 1 mM N-oxalyl amino acid on the activity of (a) FIH
and (b) PHD2 as measured by 2OG (1) consumption. For full structures of
test compounds, see Supporting Information.
Supporting Information Available: Details of chemical synthesis,
enzymatic assays, crystallographic data collection and refinement.
Complete ref 3a. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
(
1) (a) Bruick, R. K. Genes DeV. 2003, 17, 2614-2623. (b) Safran, M.; Kaelin,
W. G. J. Clin. InVest. 2003, 111, 779-783. (c) Semenza, G. L. Physiology
(
Bethesda) 2004, 19, 176-182. (d) Schofield, C. J.; Ratcliffe, P. J. Nat.
ReV. Mol. Cell Biol. 2004, 5, 343-354.
Figure 2. View of the crystal structure of (A) FIH-Fe(II) complexed with
(
2) (a) Jaakkola, P.; Mole, D. R.; Tian, Y. M.; Wilson, M. I.; Gielbert, J.;
Gaskell, S. J.; von Kriegsheim, A.; Hebestreit, H. F.; Mukherji, M.;
Schofield, C. J.; Maxwell, P. H.; Pugh, C. W.; Ratcliffe, P. J. Science
1
0
1
0 (1YCI), (B) FIH-Fe(II) complexed with 1 (1H2N). 10 interacts with
iron (pink sphere) via bidentate coordination. Note the poor geometry of
the water molecule in (A) completing the hexacoordination of iron along
with two histidines, aspartic acid, and 10 due to its steric clash with the
phenyl ring of 10. |Fo - Fc| electron density contoured to 3σ displayed as
blue chicken wire representation (PYMOL). The refined structure has an R
value of 0.210 and an Rfree of 0.244.
2
001, 292, 468-472. (b) Ivan, M.; Kondo, K.; Yang, H. F.; Kim, W.;
Valiando, J.; Ohh, M.; Salic, A.; Asara, J. M.; Lane, W. S.; Kaelin, W.
G. Science 2001, 292, 464-468.
(
3) (a) Epstein, A. C. R. et al. Cell 2001, 107, 43-54. (b) Bruick, R. K.;
McKnight, S. L. Science 2001, 294, 1337-1340. (c) Ivan, M.; Haber-
berger, T.; Gervasi, D. C.; Michelson, K. S.; Gunzler, V.; Kondo, K.;
Yang, H.; Sorokina, I.; Conaway, R. C.; Conaway, J. W.; Kaelin, W. G.
Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 13459-13464.
glutathione-S-transferase as substrate; 10 was a competitive inhibitor
with respect to 1. To examine the mode of binding of 10 at the
active site, crystallographic analyses were performed. FIH crystals
were grown anaerobically in the presence of Fe(II) and 10 mM 10,
and data were collected to 2.7 Å resolution at a synchrotron
radiation source.
N-Oxalyl-D-phenylalanine 10 binds to FIH between the â-sheets
of its double-stranded â-helix fold in a manner similar to that
observed for 1 and 2.10 The oxoacid moiety of 10 binds to the iron-
(4) (a) Hewitson, K. S.; McNeill, L. A.; Riordan, M. V.; Tian, Y. M.; Bullock,
A. N.; Welford, R. W. D.; Elkins, J. M.; Oldham, N. J.; Battacharya, S.;
Gleadle, J.; Ratcliffe, P. J.; Pugh, C. W.; Schofield, C. J. J. Biol. Chem.
2002, 277, 26351-26355. (b) Lando, D.; Peet, D. J.; Gorman, J. J.;
Whelan, D. A.; Whitelaw, M. L.; Bruick, R. K. Genes DeV. 2002, 16,
1466-1471. (c) Mahon, P. C.; Hirota, K.; Semenza, G. L. Genes DeV.
2001, 15, 2675-2686.
(
5) (a) Hausinger, R. P. Crit. ReV. Biochem. Mol. 2004, 39, 21-68. (b) Costas,
M.; Mehn, M. P.; Jensen, M. P.; Que, L. Chem. ReV. 2004, 104, 939-
9
86. (c) Schofield, C. J.; Zhang, Z. H. Curr. Opin. Struct. Biol. 1999, 9,
722-731.
(
6) (a) Dickinson, L. A.; Burnett, R.; Melander, C.; Edelson, B. S.; Arora, P.
S.; Dervan, P. B.; Gottesfeld, J. M. Chem. Biol. 2004, 11, 1583-1594.
(b) Olenyuk, B. Z.; Zhang, G. J.; Klco, J. M.; Nickols, N. G.; Kaelin, W.
G., Jr.; Dervan, P. B.; Dickinson, L. A.; Burnett, R.; Melander, C.; Edelson,
B. S.; Arora, P. S.; Gottesfeld, J. M. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 16768-16773 (Epub 12004 Nov 16719).
(II) with the carbonyl oxygen O2′ trans to Asp201 and its
carboxylate oxygen O2 trans to His199 (Figure 2a). The 1′-car-
boxylate of 10 forms hydrogen bonds to the side chain Nδ2 of
Asn205, while its 5′-carboxylate hydrogen bonds to Lys214 Nú,
Tyr145 OH, and Thr196 Oγ1. The phenyl ring of 10 is accom-
modated into the active site of FIH by forming hydrophobic
interactions with Leu186, Leu188, Trp296, and Tyr102. The pre-
sence of the phenyl ring results in a slight distortion of the ligand
coordination to iron with respect to that of FIH-Fe(II) complexed
with 2OG or 2 (1H2L and 1H2N, respectively) (Figure 2b). The
vacant coordination site that presumably enables dioxygen binding
is partially occupied by the phenyl ring of 10, which likely also
interferes with productive peptide substrate binding. To investigate
whether the presence of the phenyl ring in an analogous position
(
7) Appelhoff, R. J.; Tian, Y. M.; Raval, R. R.; Turley, H.; Harris, A. L.;
Pugh, C. W.; Ratcliffe, P. J.; Gleadle, J. M. J. Biol. Chem. 2004, 279(37),
38458-38465.
(
8) Hewitson, K. S.; McNeill, L. A.; Schofield, C. J. Curr. Pharm. Des. 2004,
10, 821-833.
(
9) Milkiewicz, M.; Pugh, C. W.; Egginton, S. J. Physiol. 2004, 560, 21-26
(Epub 2004 Aug 2019).
(
10) Elkins, J. M.; Hewitson, K. S.; McNeill, L. A.; Seibel, J. F.; Schlemminger,
I.; Pugh, C. W.; Ratcliffe, P. J.; Schofield, C. J. J. Biol. Chem. 2003,
278, 1802-1806.
(
11) McNeill, L. A.; Bethge, L.; Hewitson, K. S.; Schofield, C. J. Anal.
Biochem. 2005, 336, 125-131.
(12) Ayoub, N.; Noma, K.; Isaac, S.; Kahan, T.; Grewal, S. I.; Cohen, A. Mol.
Cell. Biol. 2003, 23, 4356-4370.
JA050841B
J. AM. CHEM. SOC.
9
VOL. 127, NO. 21, 2005 7681