Complementary Steric Engineering at the Protein-Ligand Interface for Analogue-Sensitive TET Oxygenases

Article abstract of DOI:10.1021/jacs.8b05283

Ten-eleven translocation (TET) enzymes employ O₂, earth-abundant iron, and 2-ketoglutarate (2KG) to perform iterative C-H oxidation of 5-methylcytosine in DNA to control expression of the mammalian genome. Given that more than 60 such C-H oxyge

Full text of DOI:10.1021/jacs.8b05283

Subscriber access provided by STEPHEN F AUSTIN STATE UNIV
Article
Complementary Steric Engineering at the Protein-
Ligand Interface for Analogue-Sensitive TET Oxygenases
Babu Sudhamalla, Sinan Wang, Valerie Snyder, Sam Kavoosi, Simran Arora, and Kabirul Islam
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b05283 • Publication Date (Web): 20 Jul 2018
Downloaded from http://pubs.acs.org on July 20, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Complementary Steric Engineering at the Protein-Ligand Interface
for Analogue-Sensitive TET Oxygenases
Babu Sudhamalla,# Sinan Wang,# Valerie Snyder,#,§ Sam Kavoosi,#,§ Simran Arora,#,§ Kabirul Islam*,#
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
#Department of Chemistry, University of Pittsburgh, PA 15260, USA
ABSTRACT: Ten-eleven translocation (TET) enzymes employ O2, earth-abundant iron and 2-ketoglutarate (2KG) to perform iter-
ative C-H oxidation of 5-methylcytosine (5mC) in DNA to control expression of the mammalian genome. Given that more than
sixty such C-H oxygenases are present in humans, determining context-dependent functions of each of these enzymes is a pivotal
challenge. In an effort to tackle the problem, we developed analogue-sensitive TET enzymes to perturb the activity of a specific
member. We rationally engineered the TET2-2KG interface to develop TET2 variants with an expanded active site that can be spe-
cifically inhibited by the N-oxalylglycine (NOG) derivatives carrying complementary steric ‘bump’. Herein, we describe the identi-
fication and engineering of a bulky gatekeeper residue for TET proteins, characterize the orthogonal mutant-inhibitor pairs and
show generality of the approach. Employing cell-permeable NOG analogues, we show that TET2 mutant can be specifically inhib-
ited to conditionally modulate cytosine methylation in chromosomal DNA in intact human cells. Finally, we demonstrate applica-
tion of the orthogonal mutant-inhibitor pair to probe transcriptional activity of a specific TET member in cells. Our work provides a
general platform for developing analogue-sensitive 2KG-dependent oxygenases to unravel their functions in diverse signaling pro-
cesses.
INTRODUCTION
as compensatory functions in fluctuating cellular environments
constitutes a major challenge.11
Organismal multicellularity has been mechanistically linked to
a steady increase in oxygen pressure in the atmosphere.1 This
is likely due to the emergence of enzymes catalyzing oxygen-
dependent biochemical reactions essential to metabolic and
transcriptional regulations. An early class of such enzymes
employ simple co-substrates such as 2-ketoglutaric acid (2KG)
1 and iron for oxidative processing of proteins and nucleic
acids.2 In humans, more than sixty such oxygenases catalyze
C-H hydroxylation in diverse nuclear and cytoplasmic sub-
strates, and contribute significantly to cellular processes rang-
ing from DNA repair to hypoxic signaling.2 Within such
group, Ten-eleven translocation (TET) enzymes perform C-H
oxidation of 5-methylcytosine (5mC) to 5hydroxymethyl-
cytosine (5hmC) to control chromatin accessibility and there-
by expression potential of the human genome (Figure 1).3,4
Intriguing chemical transformations (iterative C-H function-
alization under ambient conditions) coupled with asymmetric
demethylation (paternal over maternal pronuclei) and causal
role in human disease have generated widespread interest in
studying how a specific TET protein reprograms our genome
in diverse physiological contexts.12 Diagramming TET-
specific gene regulatory network with genetic perturbation is
difficult owing to its slow acting nature and oftentimes loss of
the entire protein. Pharmacological perturbation involving
small molecules provides a powerful alternative to dissect
protein functions with high temporal control.13,14 However,
attempts to develop small-molecule probes for TET enzymes
with cellular activity have not met with much success as of
yet.15 Gaining specificity from such molecules would be chal-
lenging due to the similarity in active site fold, cofactors and
co-substrates among sixty 2KG-dependent human oxygenases.
TET-mediated successive oxidation of 5mC (once thought
to be a permanent mark because of the stable C-C connectivi-
ty) to 5hmC, 5-formylcytosine (5fC) and finally 5-
carboxycytidine (5caC) provides a biochemical basis for ac-
tive DNA demethylation.5 5caC could be recognized and ex-
cised by thymine DNA glycosylase (TDG) to effect base exci-
sion repair (BER)-mediated replacement of 5caC with deox-
ycytidine (dC).6 Direct decarboxylation of 5caC could be an-
other potential mechanism of active genome demethylation,
although more evidence is needed.7,8 Gene regulation by such
oxidative demethylation is critical for cellular differentiation
and embryogenesis. For example, TET-mediated demethyla-
tion on the paternal chromosome is essential to early devel-
opment of a fertilized egg.9 Aberrant gene regulation by TET
overexpression and somatic mutation are implicated in multi-
ple human conditions.10 Interrogating developmentally essen-
tial and clinically relevant TET members with distinct as well
O
O
OH
N
=
H
O
OH
O
O
NOG 2
O
O
OH
=
OH
O
OH
OH
2-KG 1
O2
CO2
NH2
NH2
N
H3C
N
Wild type TET
HO
N
O
N
O
Hole-modified TET
O2 CO2
5hmC
Genes on
5mC
Genes off
O
O
O
OH
O
O
OH
=
O
OH
O
OH
N
H
OH
=
2-KG 1
O
OH
Bumped NOG
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 8
Figure 1. The logic of analogue-sensitive (as) TET enzymes.
sized a series of NOG analogues 3-8 bearing alkyl and aryl
moieties with variegated steric bulk at C4 of NOG 2 (Figure
2c, Scheme S1-3).24 Next, we examined the inhibitory activity
of the NOG analogues towards TET2 and its mutants using the
MALDI-based hydroxylation assay (Figure 2c, S2,3). Wild
type TET2 and majority of its mutants were inhibited by NOG
2, reflecting their ability to bind and process 2KG 1 as cofac-
tor. However, with the exception of the V1900A and V1395A
mutants, which were moderately inhibited by the ethyl and
propyl NOGs 3 and 4, respectively, majority of the mutants
remained inert towards the bulky analogues (Figure 2c, S4).
Conserved bulky amino acids are replaced to introduce hole-
modified silent mutations in the active site. The mutants, although
act like wild type TETs, could be specifically inhibited by
‘bumped’ N-oxalylglycine (NOG) analogues that do not perturb
native 2KG-dependent enzymes.
1
2
3
4
5
6
7
8
To address these challenges, we envisioned developing or-
thogonal mutant-inhibitor pairs of TET enzymes by engineer-
ing a steric component of the TET-2KG interface. In this ap-
proach, known as analogue-sensitive chemical genetics or the
bump-and-hole tactic,16-23 the active site of TET enzymes is
modified by replacing bulky residues with alanine or glycine
(Figure 1). The catalytically competent TET variants with
expanded active site are then screened for selective inhibition
by 2KG-competitve molecules. The mutant-specific inhibitors
are obtained from N-oxalylglycine (NOG) 2, a non-specific
inhibitor of 2KG-utilizing enzymes,24 by appending comple-
mentary steric bulk at a suitable position (Figure 1). Some of
these NOG analogues are expected to be refractory to the wild
type enzymes due to a potential steric clash with the larger
amino acids in the active site, thus providing chemical probes
to interrogate a specific TET member. We selected NOG as
the inhibitor scaffold because of its ability to bind the TET
enzymes in a 2KG-competitive manner due to structural simi-
larity.25,26 The presence of amidoacid functionality in NOG (in
place of the ketoacid in 2KG) prevents its catalytic processing
by the enzymes, thus making it an effective inhibitor. In this
study, we describe our efforts towards development of the first
analogue-sensitive TET-NOG pairs and their application in
probing transcriptional activity of a specific TET member in
human cells.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a.
T1393
V1395
b. 120
Y1902
80
5mC
V1900
NOG
40
0
C1374
T1883
F1377
c.
R
O
H
O
OH
N
H
2
4
5
6
7
8
3
O
OH
WT
4
32.5581395
36.9565217
71.1111111
77.9411764
29.0322580
5
3
4
1
7
6
0
3.48837 2093
42.3913 0435
51.1111 1111
11.7647 0588
27.4193 5484
0
13. 95348837
5.4 34782609
57. 77777778
22. 05882353
0
0
0
0
V1395A
V1900A
0
0
0
0
13.043478 26
77.777777 78
41.176470 59
97.258064 52
10.86
28.88
17.64
61.29
9
5
6522
8889
5882
2258
8
2
8
1
.6 95652174
4. 44444444
.8 23529412
1. 29032258
5.43478
15.5555
29.4117
1.61290
2
6
09
56
71
26
V1395G
V1900G
8
8
5
5
7
0
6
4
T1393S/V1395G
61. 29032258
0
3
3
2
0
2
0
4
0
6
0
8
0
1
0
0
100
% Inhibition:
0
Figure 2. Development of analogue-sensitive (as) TET enzymes.
(a) Structure of TET2 bound with 5mC-DNA and NOG (PDB:
4NM6) showing conserved bulky hydrophobic amino acids. The
selected residues are targeted for mutagenesis. (b) Bar diagram
showing activity of selected alanine mutants of TET2 (% 5hmC
formation) as judged by MALDI MS analysis. (c) Heat-map rep-
resentation of % inhibition of selected hole-modified TET2 mu-
RESULTS AND DISCUSSION
Development of the orthogonal TET2-NOG pair. We se-
lected TET2 as representative DNA demethylase here because
of its pivotal role in global methyl erasure during embryogen-
esis, its established link to cancer and the availability of criti-
cal structural information.25 In order to develop TET2 variants
with an expanded active site, we analyzed the crystal structure
of TET2 bound with NOG and 5mC-containing DNA, and
identified several bulky residues that line up to hold 2KG in
the active site (Figure 2a). Given that the majority of these
amino acids are hydrophobic in nature and do not participate
in strong electrostatic interactions, either with 2KG or DNA,
we performed alanine scanning on these sites, one at a time.
Mutant proteins carrying the TET2 catalytic domain were ex-
pressed and subjected to a MALDI-based C-H hydroxylation
assay26 using a 5mC-containing short DNA and the cofactor
2KG 1 (Figure 2b). The wild type TET2 and most of the mu-
tants were capable of accepting 1 as their cofactor to generate
5hmC DNA albeit at varying degrees (Figure 2b, S1). F1377A
and Y1902A mutants were however completely inert, likely
due to the combined effect of active site misfolding and com-
promised DNA binding. Loss of H-bonding between 5mC and
Y1902 may contribute to the catalytic inability of the Y1902A
mutant.25
tants by
a set of NOG analogues. The best mutant
T1393S/V1395G is referred to as analogue-sensitive (as) TET2.
Careful analysis of the NOG-bound TET2 structure revealed
that the isopropyl side chains of V1395 and V1900 reside very
close (within 4.8Å) to the C4 methylene group carrying the
bulky substituents in NOG analogues. This led us to generate
the V1395G and V1900G mutants with a slightly larger active
site with intact demethylase activity (Figure S5). In the de-
methylation assay, multiple NOG analogues inhibited V1395G
mutant albeit to varying degrees (Figure 2c, S6). The smaller
NOGs such as 4 and 5 inhibited the mutant more effectively.
The analogues did not inhibit the V1900G mutant to a signifi-
cant degree (Figure 2c, S7). Given that removing steric block
at V1395 site sensitizes TET2 mutants towards multiple inhib-
itors, we reasoned that expanding the active site further might
help in gaining stronger inhibition by the bulkier NOGs. We
therefore introduced a second site mutation to V1395G to gen-
erate multiple double mutants. Majority of the mutants lost
their catalytic activity towards 5mC DNA. Remarkably, the
T1393S/V1395G mutant retained its DNA demethylase activi-
ty as demonstrated by hydroxylation assay (Figure S5). Fur-
thermore, the double mutant was strongly blocked by NOG
analogues 4-6 (Figure 2c, S8). Leucine NOG 5 appeared to be
This screening data provided us with a platform suitable for
developing analogue-sensitive TET2 enzyme, reminiscent of
the protein kinase mutants where engineering a bulky gate-
keeper residue sensitized the mutants towards ‘bumped’ inhib-
itors.27,28 To obtain such allele-specific inhibitors, we synthe-
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
particularly effective against the mutant, demonstrating the
importance of complementary steric engineering to develop
analogue-sensitive (as) TET2.
(Figure 3b,c). We further noticed that S1393 residue in the
mutant makes an extra H-bonding contact with the amidoacid
group in 5, potentially contributing to the improved inhibitor
efficacy against as TET2 (Figure 3d). Overall, the docking
study suggests that V1395G and T1393S mutations in as
TET2 may contribute differentially to the analogue-sensitivity,
with the former bringing more steric complementarity and the
latter more electrostatic affinity.
1
2
3
4
5
6
7
8
Characterization of the orthogonal TET2-NOG pair. We
next sought to assess the inhibitory efficacy of NOG ana-
logues (4-6) by determining their IC50 values against the se-
lected TET2 variants. The MALDI-based hydroxylation assay
is suitable for providing quantitative information on TET en-
zymatic activity because of the similar ionization potentials of
5mC and 5hmC containing DNAs.26 The propyl NOG 4
showed better IC50 against the double mutant T1393S/V1395G
(hereafter as TET2) compared to V1395G (120 ꢀM vs 473
ꢀM), consolidating the importance of expanded TET2 active
site for inhibitor potency of NOG analogues (Figure S9-11).
From our dose-response experiments, we found leucine NOG
5 to be the most effective inhibitor for as TET2 with an IC50 of
18 ꢀM (Figure 3a, S12-13). This compound did not show any
inhibition towards wild type TET2 in a wide range of concen-
trations (Figure 3a, S14), thus making it suitable for allele-
specific perturbation of TET2.
To further ensure that the ‘bumped’ NOGs do not inhibit
other human 2KG dependent enzymes, we examined a set of
structurally and functionally distinct demethylases: TET3
(DNA demethylase), FTO and ALKBH5 (RNA demethylas-
es), KDM4A, C and KDM6B (histone demethylases).29-35 The-
se wild type proteins efficiently acted on their respective sub-
strates in presence of cofactor 1 and were inhibited by canoni-
cal NOG 2 but not the ‘bumped’ NOG analogue 5 (Figure
S16-22). The fact that all the seven native demethylases, in-
cluding TET2, examined thus far remained refractory towards
5 lends further support for the orthogonality of the inhibitor
across human 2KG-dependent enzymes.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a.
a.
b.
S1393
•
•
•
•
Genomic DNA isolation
NH2
100
G1395
H3C
Genome-wide hydroxymethylation
Western blotting analysis
WT TET2
N
as TET2
75
50
25
N
O
Bioorthogonal functionalization
O
as TET2
IC50 = 18 1.3 ꢀM
O
OH
N
H
5mC
O
OH
d.
5
100
0
-7
80
60
40
20
0
NOG 5
-6
-5
Log [5], M
-4
c.
b.
c.
d.
No TET2
No as
TET2
V1395
G1395
- 2
- 5
+ 5
WT
TET2
+ 2
+ 5
as
as TET2
WT TET2
TET2
S1393
g.
e.
f.
No as
TET1
No as
TET3
NOG 5
NOG 5
hTET1: MNNGSTVVCTLTR
hTET2: MQNGSTLVCTLTR
hTET3: L YNGCTVVCTLTK
- 5
+ 5
- 5
as
as
TET1
TET3
Figure 3. Potency and specificity of the orthogonal pairs. (a)
NOG analogue 5 inhibits the steady-state activity of as TET2
variant with an IC50 of 18 ꢀM as determined by MALDI MS. The
compound did not inhibit wild type TET2. (b) Energy-minimized
structure showing expanded active site pocket in double mutant
(T1393S/V1395G) housing the bulky isobutyl group in NOG 5.
(c) Superimposed structures of WT-TET2 and as TET2 shows
steric clash (indicated by the red circle) between the isobutyl
group of NOG 5 and the isopropyl side chain of V1395 residue of
WT-TET2. (d) G1395 of as TET2 releases the steric clash with
NOG 5. Dashed blue line shows a unique H-bonding contact be-
tween the -OH group of S1393 of as TET2 and the amidoacid
moiety of NOG 5 in the energy-minimized structure.
+ 5
Figure 4. Activity of as TET-NOG 5 pairs on the genomic DNA.
(a) Schematic showing isolation of genomic DNA from human
cells followed by enzymatic hydroxylation of 5mC by as TET2
and its inhibition by 5 and subsequent biochemical analysis. (b)
Dot-blot assay showing activity of wild type TET2 on the ge-
nomic DNA. The activity was inhibited by NOG 2 but not by
‘bumped’ NOG 5. (c) Dot-blot assay showing activity of as TET2
on the genomic DNA, which was inhibited by the ‘bumped’ NOG
5. (d) Quantitative representation of the dot-blot results obtained
in (b) and (c). (e) Sequencing data showing T1393 and V1395 of
TET2 are fully conserved among other TET members. (f,g) Dot-
blot assays showing activity of as TET1 and TET3, respectively,
on the genomic DNA and their inhibition by NOG 5.
To rationalize the binding potency and specificity of 5 to as
TET2, we performed docking experiments using AutoDock
energy minimization software. For this purpose, we used the
coordinates from NOG-bound TET2 crystal structure (PDB
4NM6)25. In the most stable conformation, 5 occupied the site
of NOG 2 and maintained interactions with the as mutant in a
manner similar to that of 2 with wild type TET2 (Figure S15).
Most importantly, in the energy minimized model, the isobutyl
side chain of 5 occupied the hole created by V1395G mutation
in the as variant and induced a steric clash at the active site of
the wild type protein, which explains the observed specificity
Next, we examined the activity of the mutant-inhibitor pair
on complex substrates (Figure 4a). We collected genomic
DNA from human embryonic kidney 293T (HEK293T) cells.
In a dot-blot assay, using appropriate antibody, the isolated
genomic DNA showed robust level of methylated cytidine
(5mC) and almost undetectable amount of 5hmC, consistent
with minimal expression of the TET enzymes in these cells
(Figure 4b).36,37 Hydroquinone treatment, however, induced
5hmC formation in the genomic DNA of cultured cells, likely
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 8
via activation of the endogenous TET enzymes as reported
presence of ‘bumped’ NOG 5 (Figure 5b), thus correlating the
formation of biotinylated DNA to the activity of as TET2.
earlier (Figure S23).38 Upon incubating the genomic DNA
with wild type and as TET2 along with the cofactor 1, we
observed significant formation of 5hmC (Figure 4b,c,d) in
both cases, suggesting that the hole-modified mutant is capa-
ble of acting on complex substrate with efficiency comparable
to that of the wild type protein. More importantly, leucine
NOG 5 completely abrogated the formation of 5hmC by as
TET2 but not the wild type enzyme which was inhibited only
by NOG 2. Collectively, these results demonstrate that the
mutant-inhibitor pair is truly orthogonal and functional on
physiologically relevant substrate.
1
2
3
4
5
6
7
8
N3
N3
a.
O
O
NH2
NH2
N
HO
HO
HO
HO
NH2
N
H3C
N
HO
HO
HO
as TET2
NOG 5
UDP
O
N
O
N
O
β-GT
N
O
O
O
NH
O
O
NH
N
HN
N
H
N
N
S
•
Avidin enrichment
MS analysis
HN
N
H
9
S
9
•
•
O
HO
HO
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TET-specific
biotin-DNA
WB analysis
Cu(II), ascorbate, TBTA
HO
Development of analogue-sensitive TET1 and TET3 var-
iants. After demonstrating that as variant can be successfully
developed for TET2, we sought to expand the technology to
other members of the family. In mammals, TET family con-
sists of three members TET1-3, each with distinct functions.7
Sequence analyses of TET oxygenases revealed that the amino
acids corresponding to T1393 and V1395 in TET2 are highly
conserved among TET1-3 (Figure 4e).29 These bulky amino
acids are very likely acting as gatekeeper residues that pre-
clude binding of bulky NOG analogues as observed for TET2.
We generated T1683S/V1685G and T1088S/V1090G double
mutants of TET1 and 3, respectively, as their as variants.
When examined for activity, the TET1 and 3 mutants effi-
ciently produced 5hmC in genomic DNA as measured in the
dot-blot assay (Figure 4f,g), confirming their catalytic activity.
We subsequently showed that the as variants were selectively
inhibited by NOG analogue 5 (Figure 4f,g). Collectively, our
results provide strong evidence for the existence of a bulky
gatekeeper residue (an equivalent of V1395 in TET2) within
the TET subfamily. Mutations at these sites led to, for the first
time, analogue-sensitive DNA demethylases, suitable to de-
convolute their specific functions. We have shown earlier that
b.
c.
2907.3
100
100
+ NOG 5
- NOG 5
Biotin
DNA
No as
TET2
Biotin
DNA
50
50
0
- 5
as
TET2
2907.3
+ 5
0
2890
2910 2920
Mass (M/Z)
2900
2890 2900 2910 2920
Mass (M/Z)
Figure 5. Chemoenzymatic labeling of genome-wide TET activi-
ty. (a) Schematic showing steps in the β-GT/ ‘click’ chemistry
mediated biotinylation of 5hmC generated by as TET2 on ge-
nomic DNA. (b) MALDI MS showing successful azido-
glucosylated and ‘clicked’ product only in absence of 5. (c) Dot-
blot assay with anti-biotin antibody confirmed tandem azido-
glucosylation and biotinylation of 5hmC on genomic DNA. NOG
analogue 5 strongly inhibited such derivatization.
To further assess the ability of the engineered pair to modu-
late tandem hydroxymethylation-biotinylation process in ge-
nome-wide manner, the entire genomic DNA isolated from
HEK293T cells was treated with as TET2 variant with or
without 5. The modified DNA was subjected to β-GT-
mediated azidoglucosylation followed by ‘click’ chemistry
with 9. The ligated product was directly analyzed by dot-blot
assay using anti-biotin antibody (Figure 5c). Chemilumines-
cent signal from the non-inhibited sample confirmed catalytic
activity of as TET2. Significantly reduced signal was observed
when the TET assay was performed in presence of NOG 5.
These results provide a blueprint for the potential application
of as TET enzymes in probing genome-wide activity of a spe-
cific TET member when coupled to next-generation sequenc-
ing.
hole-modifying F185G mutation of KDM4A,
a 2KG-
dependent histone demethylase, sensitizes the mutant towards
a series of bulky 2KG cofactor analogues.23 F185 and V1395
residues in KDM4A and TET2, respectively, reside in close
proximity to the cofactor 2KG in the respective active sites.
These amino acids prevent binding of the ‘bumped’ cofactor/
inhibitor analogues to the wild type proteins by acting as gate-
keeper residues. Our studies disclose a potentially general
design principle towards engineering these gatekeeper sites to
achieve analogue-sensitivity for the 2KG-dependent human
enzymes.
Allele-specific perturbation of TET activity in cultured
human cells. The three human TET proteins, each with dis-
tinct structural features, have compensatory as well as non-
overlapping functions in cells. We sought to examine the ef-
fectiveness of the NOG-based inhibitors in modulating deme-
thylase activity of the analogue-sensitive TET enzymes in
intact human cells. Towards this end, we synthesized octyl
ester derivatives 10 and 11 of NOG 2 and leucine NOG 5,
respectively. It has been shown that esterification of 2KG and
2HG (2-hydroxyglutarate) improves their cell-permeability
(Figure 6a, Scheme S5).42,43,44
Functionalization of TET-specific activity on genomic
DNA. Chemoenzymatic functionalization of 5hmC followed
by affinity enrichment and sequencing has become a powerful
method to map genomic distribution of cytosine modifications
and hence TET activity.39 A particular approach includes β-
glucosyltransferase (β-GT) mediated glycosylation of 5hmC
present in genomic DNA with 6-azidoglucose-UDP followed
by ‘click’ ligation with appropriate reporter alkyne.39,40 We
sought to combine the as TET variants with the chemoen-
zymatic modification of 5hmC that would potentially allow
direct probing of genome-wide activity of a specific TET
member (Figure 5a). 5hmC generated on a short DNA sub-
strate by as TET2 mutant underwent smooth glycosylation to
6-azidoglucose-UDP using β-GT followed by Cu-catalyzed
‘click’ ligation41 with biotin-alkyne 9 as confirmed by MALDI
MS (Figure 5b, S24, Scheme S4). However, no such ligated
product was obtained when the TET2 assay was performed in
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
the compound from cell culture (Figure S26). These results
a.
O
R
O
O
b.
H
O
1
2
3
4
5
6
7
8
N
H
demonstrate that the demethylase activity of TET2 variants on
chromosomal DNA can be specifically and conditionally ma-
nipulated using cell-permeable and reversible ‘bumped’ NOG
analogue in intact cells.
or
O
10
OH
11
- 10
No TET2
- 10
O
O
OH
+ 100 ꢀM 10
+ 250 ꢀM 10
+ 250 ꢀM 11
N
H
WT TET2
WT
TET2
O
O
OH
OH
We also examined off-target activity of the NOG analogue
11 on 2KG-dependent endogenous histone demethylases. We
extracted nuclear histones from HEK293T cells treated with
100 ꢀM of 11 and analyzed for four key trimethylated lysine
residues in histone H3 (H3K4Me3, H3K9Me3, H3K27Me3 and
H3K36Me3) using corresponding antibodies. Upon compari-
son with untreated cells, we did not observe any significant
changes in histone methylation levels at these sites (Figure
S27). This result further confirms the in cellulo selectivity of
the ‘bumped’ NOG analogue, making it suitable to study only
the analogue-sensitive variant in human cells.
2
O
OH
as TET2
N
H
5
e.
100
80
9
d.
EpCAM
c.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
60
40
20
No as
TET2
- 11
- 11
GAPDH
+ 100 ꢀM 11
+ 200 ꢀM 11
+ 300 ꢀM 11
as
TET2
as
Gene regulatory function of TET enzymes is directly related
to their catalytic ability to generate 5hmC in loci-specific
manner. Overexpression of TET2 is commonly associated
with gene activation through the formation of 5hmC at pro-
moters and enhancers. We sought to examine whether the as
TET2 and the ‘bumped’ inhibitor pair could modulate expres-
sion of certain genes in cells. We focused on epithelial cellular
adhesion molecule (EpCAM) and transducin beta like 1 X-
linked (TBL1X) proteins as it has been shown that shRNA-
mediated knockdown of TET2 leads to the silencing of both
EPCAM and TBL1X genes primarily via the loss of 5hmC at
the promoter.45,46 Consistently, overexpression of TET2 leads
to re-activation of EpCAM and TBL1X proteins. Nuclear ex-
tracts from HEK293T cells expressing as TET2 variant and
treated with or without 100 ꢀM of 11 were collected and ana-
lyzed for the expression of both the proteins using appropriate
antibodies. Overexpression of as TET2 led to an increase in
EpCAM and TBL1X protein levels compared to the control
cells lacking as TET2 (Figure 6d,e, S28). These results con-
firm that as TET2 is functionally competent in cells and re-
quired for expression of certain genes much like its wild type
counterpart. We observed, however, more than three-fold re-
duction in EpCAM and TBL1X proteins when as TET2-
expressing cells grew in the presence of the ‘bumped’ inhibi-
tor 11 (Figure 6d,e, S28). Expression of housekeeping proteins
such as GAPDH remained unaffected in all the samples. Col-
lectively, our results show that bulky NOG inhibitor can mod-
ulate the gene regulatory functions of a specific TET member
in human cells. Our study further demonstrates that selective
inhibition of the catalytic activity of TET2 is sufficient to si-
lence expression of certain genes.
as TET2
TET2
Figure 6. Cellular activity of as TET2 and NOG analogues. (a)
Structures of cell-permeable NOG 10 and its bumped counterpart
11; in HEK293T cells, they are hydrolyzed to 2 and 5, respective-
ly. (b) Dot-blot assay showing activity of wild type TET2 on
chromosomal DNA in cells. Its activity was inhibited by NOG 2
(hydrolyzed from 10) but not by ‘bumped’ NOG 5 (hydrolyzed
from 11). Methylene blue (MB) showing DNA staining as loading
control. (c) Dot-blot assay showing dose-dependent inhibition of
as TET2 by 5 (hydrolyzed from 11) as measured by 5hmC for-
mation on chromosomal DNA in cells. Methylene blue (MB)
showing DNA staining as loading control. (d) Western blot analy-
sis showing increased expression of EpCAM, but not GAPDH, by
as TET2 enzyme. Selective inhibition of as TET2 by 5 (hydro-
lyzed from 11) resulted in reduction of EpCAM expression. (e)
Quantitative representation of western blot data in (d) showing
EpCAM expression decrease by 3 folds in presence of NOG ana-
logue 11 in cell.
Upon induced expression of the wild type (or hole-
modified) human TET2 catalytic domain in HEK293T cells,
as confirmed by western blot analysis (Figure S25), the ge-
nomic DNA was isolated and subjected to dot-blot assay. Ro-
bust level of 5hmC was observed only in cells expressing the
enzymes and not in non-transfected cells, clearly demonstrat-
ing that both the enzymes are catalytically active in cells (Fig-
ure 6b,c). We observed a reduction in 5hmC level, albeit to a
smaller extent, when cells expressing wild type TET2 were
treated with 100 and 250 ꢀM of 10 (Figure 6b). This is con-
sistent with the weaker inhibitory activity of NOG 2 (IC50
=
~150 ꢀM) towards wild type TET2.26 However, no change in
5hmC content was noticed when these cells were cultured in
the presence of 250 ꢀM of cell-permeable ‘bumped’ NOG
analogue 11 (Figure 6b), further confirming our screening
results that the ‘bumped’ NOG 5 does not inhibit the wild type
enzyme.
CONCLUSION
The diversity-based mechanism of protein evolution, in
which many protein products originate from relatively few
genes, is central to the formation of closely related protein
families. This has significantly hindered the development of
isoform/member-specific chemical probes. In this piece of
work, we employed structure-guided protein-ligand interface
engineering to overcome the barrier foisted by active site de-
generacy. TET mutants with a ‘hole’ at the active site while
still maintaining wild type catalytic activity are specifically
inhibited by designed NOG analogues carrying a complemen-
tary ‘bump’. These compounds do not interfere with the native
TETs and a range of other 2KG-dependent enzymes, and
hence can serve as chemical probes for functional analysis of
Importantly, 11 strongly inhibited as TET2 variant in cells
in a dose-dependent manner, as evident from the reduction in
5hmC formation (Figure 6c, S26). 100 ꢀM of the ester pro-
drug 11 was enough to reduce 5hmC level to a significant
degree (~75%) in these cells, consistent with the stronger in-
hibitory activity of 5 (IC50 = ~18 ꢀM) towards as TET2. The
lack of hydroxylase activity of as TET2 in these cells must
arise from allele-specific inhibition by the NOG analogue 11,
as full recovery of TET2 activity ensued upon withdrawal of
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 8
(2) Loenarz, C.; Schofield, C. J. Nat. Chem. Biol. 2008, 4, 152.
TET enzymes. Such specificity likely stems from a steric clash
between the hydrophobic ‘bump’ of the inhibitors and the
conserved bulky gatekeeper residue in TET enzymes, as we
have identified in the current study. Engineering at these gate-
keeper residues in TET proteins extended the approach to all
the three members of the subfamily. Following successful
probe development, we demonstrated their ability to condi-
tionally manipulate 5hmC formation on genomic DNA. Fur-
thermore, by combining the as technology with β-GT-
mediated azido-glycosylation and bioorthogonal ligation, we
developed a novel method for genome-wide mapping of TET
member-specific activity. Finally, cell-permeable ‘bumped’
NOG prodrug is shown to modulate the catalytic and tran-
scriptional activity of as TET2 variant in cultured human cells
in a reversible manner, thus providing a means for conditional
perturbation of DNA demethylases not attainable by either
shRNA-mediated knockdown or gene knockout. We anticipate
future experiments involving knock-in of as TET members
into the endogenous loci in somatic as well as mouse embry-
onic stem cells employing CRISPR-Cas9 genome-editing
technology for temporal manipulation of specific TET mem-
bers with cell-permeable NOG analogues.47,48 We foresee that
our engineering approach for developing analogue-sensitive
demethylation apparatus will be amenable to the other mem-
bers of the 2KG-dependent hydroxylase superfamily, particu-
larly ones for which context-dependent functions are largely
unknown.
(3) Tahiliani, M.; Koh, K. P.; Shen, Y.; Pastor, W. A.; Bandukwala,
H.; Brudno, Y.; Agarwal, S.; Iyer, L. M.; Liu, D. R.; Aravind,
L.; Rao, A. Science 2009, 324, 930.
(4) Lu, X.; Zhao, B. S.; He, C. Chem. Rev. 2015, 115, 2225.
(5) Ito, S.; Shen, L.; Dai, Q.; Wu, S. C.; Collins, L. B.; Swenberg, J.
A.; He, C.; Zhang, Y. Science 2011, 333, 1300.
(6) He, Y. F.; Li, B. Z.; Li, Z.; Liu, P.; Wang, Y.; Tang, Q.; Ding, J.;
Jia, Y.; Chen, Z.; Li, L.; Sun, Y.; Li, X.; Dai, Q.; Song, C. X.;
Zhang, K.; He, C.; Xu, G. L. Science 2011, 333, 1303.
(7) Schiesser, S.; Hackner, B.; Pfaffeneder, T.; Muller, M.;
Hagemeier, C.; Truss, M.; Carell, T. Angew. Chemie Int. Ed.
2012, 51, 6516.
(8) Wu, S. C.; Zhang, Y. Na. Rev. Mol. Cell Biol. 2010, 11, 607.
(9) Oswald, J.; Engemann, S.; Lane, N.; Mayer, W.; Olek, A.;
Fundele, R.; Dean, W.; Reik, W.; Walter, J. Curr. Biol. 2000, 10,
475.
(10) Ko, M.; Huang, Y.; Jankowska, A. M.; Pape, U. J.; Tahiliani,
M.; Bandukwala, H. S.; An, J.; Lamperti, E. D.; Koh, K. P.;
Ganetzky, R.; Liu, X. S.; Aravind, L.; Agarwal, S.; Maciejewski,
J. P.; Rao, A. Nature 2010, 468, 839.
(11) Huang, Y.; Chavez, L.; Chang, X.; Wang, X.; Pastor, W. A.;
Kang, J.; Zepeda-Martinez, J. A.; Pape, U. J.; Jacobsen, S. E.;
Peters, B.; Rao, A. Proceed. Natl. Acad. Sci. USA 2014, 111,
1361.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(12) Booth, M. J.; Raiber, E. A.; Balasubramanian, S. Chem. Rev.
2015, 115, 2240.
(13) Schreiber, S. L.; et al.. Cell 2015, 161, 1252.
(14) Knight, Z. A.; Shokat, K. M. Cell 2007, 128, 425.
(15) Nishio, K.; Belle, R.; Katoh, T.; Kawamura, A.; Sengoku, T.;
Hanada, K.; Ohsawa, N.; Shirouzu, M.; Yokoyama, S.; Suga, H.
ChemBioChem 2018.
(16) Lopez, M. S.; Choy, J. W.; Peters, U.; Sos, M. L.; Morgan, D.
O.; Shokat, K. M. J. Am. Chem. Soc. 2013, 135, 18153.
(17) Bishop, A. C.; Ubersax, J. A.; Petsch, D. T.; Matheos, D. P.;
Gray, N. S.; Blethrow, J.; Shimizu, E.; Tsien, J. Z.; Schultz, P.
G.; Rose, M. D.; Wood, J. L.; Morgan, D. O.; Shokat, K. M.
Nature 2000, 407, 395.
ASSOCIATED CONTENT
Supporting Information. Synthetic procedure and charac-
terization of all the relevant compounds, methods for pro-
tein expression and purification, and biochemical assay,
supplementary figures and tables, and NMR spectra. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(18) Carter-O'Connell, I.; Jin, H.; Morgan, R. K.; David, L. L.;
Cohen, M. S. J. Am. Chem. Soc. 2014, 136, 5201.
(19) Gibson, B. A.; Zhang, Y.; Jiang, H.; Hussey, K. M.; Shrimp, J.
H.; Lin, H.; Schwede, F.; Yu, Y.; Kraus, W. L. Science 2016,
353, 45.
(20) Belshaw, P. J.; Schoepfer, J. G.; Liu, K.-Q.; Morrison, K. L.;
Schreiber, S. L. Angew. Chemie Int. Ed. 1995, 34, 2129.
(21) Lin, Q.; Jiang, F.; Schultz, P. G.; Gray, N. S. J. Am. Chem. Soc.
2001, 123, 11608.
(22) Baud, M. G.; Lin-Shiao, E.; Cardote, T.; Tallant, C.; Pschibul,
A.; Chan, K. H.; Zengerle, M.; Garcia, J. R.; Kwan, T. T.;
Ferguson, F. M.; Ciulli, A. Science 2014, 346, 638.
(23) Breski, M.; Dey, D.; Obringer, S.; Sudhamalla, B.; Islam, K. J.
Am. Chem. Soc.y 2016, 138, 13505.
(24) McDonough, M. A.; McNeill, L. A.; Tilliet, M.; Papamicael, C.
A.; Chen, Q. Y.; Banerji, B.; Hewitson, K. S.; Schofield, C. J. J.
Am. Chem. Soc. 2005, 127, 7680.
(25) Hu, L.; Li, Z.; Cheng, J.; Rao, Q.; Gong, W.; Liu, M.; Shi, Y. G.;
Zhu, J.; Wang, P.; Xu, Y. Cell 2013, 155, 1545.
(26) Sudhamalla, B.; Dey, D.; Breski, M.; Islam, K. Anal. Biochem.
2017, 534, 28.
(27) Bishop, A. C.; Buzko, O.; Shokat, K. M. Trends Cell Biol. 2001,
11, 167.
(28) Islam, K. ACS Chem. Biol. 2015, 10, 343.
AUTHOR INFORMATION
Corresponding Author
kai27@pitt.edu
Author Contributions
§ These authors contributed equally
Notes
The authors declare no competing financial interest
ACKNOWLEDGMENT
We thank the University of Pittsburgh and the NIH
(R01GM123234) for financial support; Profs. Y. Xu, R.
Trievel, C. Schofield, D. Fujimori, C. He, H. Leonhardt for the
expression constructs used in the current study; Dr. D.
Chakraborty and members of our laboratory for critical read-
ing and editing of the manuscript. Support for MALDI-TOF
MS instrumentation was provided by a grant from the National
Science Foundation (CHE-1625002).
(29) Hu, L.; Lu, J.; Cheng, J.; Rao, Q.; Li, Z.; Hou, H.; Lou, Z.;
Zhang, L.; Li, W.; Gong, W.; Liu, M.; Sun, C.; Yin, X.; Li, J.;
Tan, X.; Wang, P.; Wang, Y.; Fang, D.; Cui, Q.; Yang, P.; He,
C.; Jiang, H.; Luo, C.; Xu, Y. Nature 2015, 527, 118.
(30) Fu, Y.; Jia, G.; Pang, X.; Wang, R. N.; Wang, X.; Li, C. J.;
Smemo, S.; Dai, Q.; Bailey, K. A.; Nobrega, M. A.; Han, K. L.;
Cui, Q.; He, C. Nat. Commun. 2013, 4, 1798.
REFERENCES
(31) Chen, W.; Zhang, L.; Zheng, G.; Fu, Y.; Ji, Q.; Liu, F.; Chen,
(1) Jeltsch, A. Trends Biochem. Sci. 2013, 38, 172.
H.; He, C. FEBS Letts. 2014, 588, 892.
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
(32) Williams, S. T.; Walport, L. J.; Hopkinson, R. J.; Madden, S. K.;
(41) Demko, Z. P.; Sharpless, K. B. Angew. Chemie Int. Ed. 2002,
Chowdhury, R.; Schofield, C. J.; Kawamura, A. Epigenetics
2014, 9, 1596.
(33) Krishnan, S.; Trievel, R. C. Structure 2013, 21, 98.
(34) Pack, L. R.; Yamamoto, K. R.; Fujimori, D. G. The Journal of
biological chemistry 2016, 291, 6060.
(35) Gerken, T.; Girard, C. A.; Tung, Y. C.; Webby, C. J.; Saudek,
V.; Hewitson, K. S.; Yeo, G. S.; McDonough, M. A.; Cunliffe,
S.; McNeill, L. A.; Galvanovskis, J.; Rorsman, P.; Robins, P.;
Prieur, X.; Coll, A. P.; Ma, M.; Jovanovic, Z.; Farooqi, I. S.;
Sedgwick, B.; Barroso, I.; Lindahl, T.; Ponting, C. P.; Ashcroft,
F. M.; O'Rahilly, S.; Schofield, C. J. Science 2007, 318, 1469.
(36) Liu, M. Y.; Torabifard, H.; Crawford, D. J.; DeNizio, J. E.; Cao,
X. J.; Garcia, B. A.; Cisneros, G. A.; Kohli, R. M. Nat. Chem.
Biol. 2017, 13, 181.
41, 2110.
1
2
3
4
5
6
7
8
(42) MacKenzie, E. D.; Selak, M. A.; Tennant, D. A.; Payne, L. J.;
Crosby, S.; Frederiksen, C. M.; Watson, D. G.; Gottlieb, E. Mol.
Cell. Biol. 2007, 27, 3282.
(43) Xu, W.; Yang, H.; Liu, Y.; Yang, Y.; Wang, P.; Kim, S. H.; Ito,
S.; Yang, C.; Wang, P.; Xiao, M. T.; Liu, L. X.; Jiang, W. Q.;
Liu, J.; Zhang, J. Y.; Wang, B.; Frye, S.; Zhang, Y.; Xu, Y. H.;
Lei, Q. Y.; Guan, K. L.; Zhao, S. M.; Xiong, Y. Cancer Cell
2011, 19, 17.
(44) Zengeya, T. T.; Kulkarni, R. A.; Meier, J. L. Org. Letts. 2015,
17, 2326.
(45) Chen, J. Y.; Luo, C. W.; Lai, Y. S.; Wu, C. C.; Hung, W. C.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Oncogenesis 2017, 6, e369.
(46) Wang, Y.; Xiao, M.; Chen, X.; Chen, L.; Xu, Y.; Lv, L.; Wang,
P.; Yang, H.; Ma, S.; Lin, H.; Jiao, B.; Ren, R.; Ye, D.; Guan, K.
L.; Xiong, Y. Mol. Cell 2015, 57, 662.
(47) Bartkowiak, B.; Yan, C.; Greenleaf, A. L. Biochim. Biophys.
Acta 2015, 1849, 1179.
(37) Lee, M.; Li, J.; Liang, Y.; Ma, G.; Zhang, J.; He, L.; Liu, Y.; Li,
Q.; Li, M.; Sun, D.; Zhou, Y.; Huang, Y. J. Am. Chem. Soc.
2017, 139, 4659.
(38) Coulter, J. B.; O'Driscoll, C. M.; Bressler, J. P. The J. Biol.
Chem. 2013, 288, 28792.
(48) Moyer, T. C.; Holland, A. J. Methods Cell Biol. 2015, 129, 19.
(39) Song, C. X.; Clark, T. A.; Lu, X. Y.; Kislyuk, A.; Dai, Q.;
Turner, S. W.; He, C.; Korlach, J. Nat. Methods 2011, 9, 75.
(40) Szwagierczak, A.; Bultmann, S.; Schmidt, C. S.; Spada, F.;
Leonhardt, H. Nucleic Acids Res. 2010, 38, e181.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
1
2
3
4
5
6
7
8
9
NH2
N
Inhibition of ~60
oxygenases
NH2
H3C
N
HO
O
Wild type TET
N
O
O
OH
N
O
N
H
O
OH
NOG
O
O
O
O
OH
+
O
2
CO2
+
OH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
OH
2KG
OH
Inhibition of a
hole-modified
oxygenase
O
C-H
Oxidation
Gene
Regulation
Hole-modified
TET
O
OH
N
H
O
OH
‘Bumped’ NOG
TOC Figure
8
ACS Paragon Plus Environment