Understanding Flavin-Dependent Halogenase Reactivity via Substrate Activity Profiling

Article abstract of DOI:10.1021/acscatal.6b02707

The activity of four native FDHs and four engineered FDH variants on 93 low-molecular-weight arenes was used to generate FDH substrate activity profiles. These profiles provided insights into how substrate class, functional group substitution, electronic activation, and binding affect FDH activity and selectivity. The enzymes studied could halogenate a far greater range of substrates than have been previously recognized, but significant differences in their substrate specificity and selectivity were observed. Trends between the electronic activation of each site on a substrate and halogenation conversion at that site were established, and these data, combined with docking simulations, suggest that substrate binding can override electronic activation even on compounds differing appreciably from native substrates. These findings provide a useful framework for understanding and exploiting FDH reactivity for organic synthesis.

Full text of DOI:10.1021/acscatal.6b02707

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES
Article
Understanding Flavin-Dependent Halogenase
Reactivity via Substrate Activity Profiling
Mary C. Andorfer, Jonathan E. Grob, Christine E. Hajdin, Julia R.
Chael, Piro Siuti, Jeremiah Lilly, Kian L. Tan, and Jared C. Lewis
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b02707 • Publication Date (Web): 31 Jan 2017
Downloaded from http://pubs.acs.org on February 5, 2017
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 free 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 accessible to all
readers and 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.
ACS Catalysis 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
ACS Catalysis
1
2
3
4
5
6
7
8
9
Understanding Flavin-Dependent Halogenase Reactivity via Sub-
strate Activity Profiling
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
Mary C. Andorfer,1 Jonathan E. Grob,2 Christine E. Hajdin,2 Julia R. Chael,1 Piro Siuti,2 Jeremiah
Lilly,2 Kian L. Tan,2,* Jared C. Lewis1,*
2
1Department of Chemistry, University of Chicago, Chicago, IL 60637. Global Discovery Chemistry, Novartis Insti-
tutes for Biomedical Research, 250 Massachusetts Ave, Cambridge, MA 02139.
ABSTRACT: The activity of four native FDHs and four engineered FDH variants on 93 low molecular weight arenes was
used to generate FDH substrate activity profiles. These profiles provided insights into how substrate class, functional
group substitution, electronic activation, and binding impact FDH activity and selectivity. The enzymes studied could
halogenate a far greater range of substrates than previously recognized, but significant differences in their substrate speci-
ficity and selectivity were observed. Trends between the electronic activation of each site on a substrate and halogenation
conversion at that site were established, and these data, combined with docking simulations, suggest that substrate bind-
ing can override electronic activation even on compounds differing appreciably from native substrates. These findings
provide a useful framework for understanding and exploiting FDH reactivity for organic synthesis.
KEYWORDS: halogenase, flavin, C-H functionalization, biocatalysis, site selective
and PrnA suggest that these enzymes promote electro-
philic aromatic substitution of enzyme-bound tryptophan
by a formal halenium ion (X+) donor, which is proposed
to be either a lysine-derived haloamine (K79 in RebH)8 or
bound HOX9 (Scheme 1B). This mechanism is consistent
with the fact that relatively electron rich substrates are
halogenated. Other factors are clearly important, howev-
er, as less electronically activated sites can be halogenated
in the presence of more electronically activated sites.10
Introduction
The importance of halogenation for the synthesis and
function of organic compounds has driven extensive ef-
forts to identify1,2 and engineer3 halogenases to selectively
install halogen substituents. Among the several classes of
halogenases identified to date, flavin-dependent halogen-
ases (FDHs) have proven particularly promising in this
regard.4 FDHs for which in vitro activity has been estab-
lished halogenate electron rich arenes (Scheme 1A). While
some FDHs require substrates linked to carrier proteins,
many have activity on free substrates, which greatly sim-
plifies biocatalysis and engineering efforts.1
A number of fungal halogenases with activity on phe-
nols and anisoles have also been identified.1 For example,
Rdc2, an FDH from C. chiversii, was found to halogenate a
variety of resorcylic acid lactone substrates, the acyclic
natural product curcumin, and two hydroxyisoquino-
lines.11,12 GsfI, a homologue of Rdc2 from P. aethiopicum,
catalyzes a selective chlorination in the biosynthesis of
griseofulvin.13 No information on how the substrate scope
of these enzymes compare with Trp-FDHs has been re-
ported.
Despite the sequence and structural homology of Trp-
FDHs, we14,15 and others16,17 have observed significant dif-
ferences in their activity toward non-native substrates.
Likewise, the conservation of key sequence motifs in
FDHs in general belies the apparent specificity of Trp,
fungal, and likely other FDHs for distinct substrate classes
based on the limited exploration of substrates reported to
date.1 We therefore became interested in establishing de-
tailed profiles18,19 of FDH activity toward a diverse panel of
substrates (substrate activity profiles). These profiles pro-
vided insights into how substrate class, functional group
Scheme 1. (A) General scheme for FDH-catalyzed halo-
genation. (B) Mechanism for generation of proposed ha-
lenium for electrophilic aromatic substitution (EAS) with-
in FDHs.
Tryptophan halogenases (Trp-FDHs), including RebH,5
PrnA,6 and PyrH,7 are perhaps the best studied FDHs with
activity on free substrates. Mechanistic studies on RebH
ACS Paragon Plus Environment

ACS Catalysis
substitution, electronic activation, and binding impact
FDH activity and selectivity. These data, in turn, provide a
useful framework for understanding and exploiting FDH
reactivity in organic synthesis.
Page 2 of 8
compounds are representative of those commonly used
for fragment based drug design,24 so understanding their
reactivity toward FDHs could provide information on
FDH activity toward motifs found in pharmaceuticals.
1
2
3
4
5
6
FDH Substrate Activity Profiles
Results
FDHs and Substrates Selected for Activity Profiles
The eight FDH genes were co-expressed with the Gro-
EL/ES genes, and the resulting FDHs were purified as
previously described for RebH.20 This protocol produced
high yields of all eights FDHs, including the fungal en-
zymes GsfI (~35 mg/L) and Rdc2 (~60 mg/L). Purified
enzyme was then use to conduct two analytical biocon-
versions on each substrate (Scheme 2). One set of reac-
tions was allowed to proceed overnight while the other
was quenched after 1 hour to ensure that monohalogena-
tion could be observed on more reactive substrates for
which dihalogenation is possible.15,25 Despite differences
in optimal reaction conditions for the FDHs, conditions
were standardized to directly compare conversions. Reac-
tions were analyzed by LCMS, and conversion data were
used to generate substrate activity profiles (Fig. 2, Table
S1).
FDHs were selected to evaluate differences in the activi-
ty and selectivity between wild-type Trp-FDHs (RebH,
Thal),20,21 engineered RebH variants (0K, 4V, 6TL, and
10S),10,15 and fungal halogenases (Rdc2 and GsfI)11,13. The
scope of RebH has been examined to a limited ex-
tent,15,20,22 but far less is known about the activity of Thal,
a Trp-6-FDH, other than its ability to halogenate D- and
L-tryptophan21. Similarly, while Rdc2 activity has been
confirmed on a handful of non-native phenol-containing
substrates,11,12 there are no previous reports of GsfI activity
on non-native substrates. RebH variant 0K was created by
replacing active site residue Glu461 with Lys based on the
success of the analogous mutation in PrnA to increase
activity on benzoic acids.16 Variant 4V was developed to
accept larger, biologically active compounds,15 while vari-
ants 6TL and 10S were engineered during an effort10 to
alter the selectivity of tryptamine halogenation.
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
Scheme 2. General scheme for FDH bioconversions.
Trp-FDH Activity on Anilines, Indoles, Pyrroles, and Azoles
The Trp-FDHs exhibited broad scope toward panel 1
substrates; 67% of these electron-rich arenes were halo-
genated to some extent (Fig. 2). Notably, activity on a
wide range of substituted anilines was observed, indicat-
ing that the narrow range of anilines evaluated to date16
could be significantly expanded. The Trp-FDHs also
demonstrated remarkable functional group tolerance by
accepting substrates containing amines, alcohols, esters,
amides, sulfonamides, nitriles, thioethers, pyridines,
quinolones, azoles, and pyrroles. Qualitatively similar
scope was observed for Thal, 6TL, and RebH, which each
halogenated 40-47% of substrates examined with >1%
conversion. In contrast, the scope and activity for variants
10S and 0K are more limited (>1% conversion for 23-24%
of substrates). In the case of 0K, this could result from the
fact that this enzyme was designed for higher activity on
benzoic acid substrates,16 but few panel 1 substrates con-
tain acidic functionality. Variant 10S was evolved to halo-
genate the 5-position of tryptamine with high selectivity,
while 6TL, an intermediate in the 10S lineage, halogenat-
ed the 5-, 6-, and 7-positions of tryptamine with similar
efficiency.10 The improved scope of 6TL relative to 10S is
thus consistent with other examples in which evolution-
ary intermediates serve as more general catalysts than
variants with high activity/selectivity on target sub-
strates.26,27 Interestingly, the FDH with the broadest scope
toward panel 1 substrates was 4V. This variant was engi-
neered to accept large indole-containing compounds,15
and this appears to have enabled activity on a wider range
of substrates than was used in the original engineering
effort.
Figure 1. Phylogenetic tree of FDHs examined. Branch labels
display amino acid substitutions per site.
When the sequence homology of these 8 FDHs is as-
sessed, they fall into two categories according to organism
of origin (Figure 1). RebH and Thal, both of which are
bacterial enzymes, have much higher sequence identities
to one another (61% identity, 99% query coverage) than
RebH and the fungal FDH, Rdc2 (43% identity, 48% query
coverage). We hypothesized this dramatic difference in
primary sequence could lead to significant differences in
scope, even beyond that previously observed between
wild-type and engineered Trp-FDHs.
A substrate panel18,23 comprised of 86 N-containing
compounds (panel 1, Table S1) was used to evaluate FDH
activity. Because halogenation with FDHs is believed to
proceed via electrophilic aromatic substitution,6 sub-
strates in this panel contained at least one electronically-
activated site. In addition, substrates were selected to
probe significant steric and electronic variation, as well as
functional group substitution, across a range of substrate
classes, including anilines, indoles, azoles or pyrroles. A
second panel (panel 2) of seven phenols and anisoles was
also constructed, taking into consideration the native
substrates of the fungal FDHs.1 The 93 compounds con-
tained within these two panels constitute the largest sur-
vey of FDH substrate scope to date. Moreover, these
ACS Paragon Plus Environment

Page 3 of 8
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure 2. Substrate activity profiles in heat map form for eight FDHs on substrates in panels 1 (indoles, pyrroles, azoles,
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
anilines, anilides) and 2 (phenols). Maximum conversion is shown for each enzyme-substrate pair (for complete data sets
at
both
time
points,
see
Table
S1.).
13-21). This trace activity, confirmed by LCMS using chlo-
rine isotope patterns, provides a starting point for evolv-
ing enzymes with good activity on these different com-
pound classes.15 We have previously demonstrated the
feasibility of this with FDHs.10,15 Moreover, this finding
clearly shows that even minor changes in halogenase se-
quence can significantly impact specificity and scope,
which highlights the importance of evaluating multiple
enzymes on a given substrate(s) to find active catalysts,
be it for preparative reactions or for starting points for
evolution.26,28,29
Comparison of Trp-FDH and Fungal FDH Activity
While many panel 1 substrates were accepted as sub-
strates by the Trp-FDHs, very few (7%) were halogenated
by the fungal FDHs (Fig. 2). For this reason, the activity of
these enzymes toward panel 2 substrates was evaluated.
Many of these less nucleophilic substrates30 were not hal-
ogenated to a significant extent by the Trp-FDHs; howev-
er, all were halogenated by at least one of the fungal
FDHs (Fig. 4). Rdc2 and GsfI, in contrast to Trp-FDHs,
have a preference for the less electron rich substrates30,
which indicates selectivity is more complex than a simple
electronic activation. The practical consequence is that
expanded substrate scope can be achieved by investigat-
ing diverse enzyme sets.
Figure 3. A) Conversion data from initial activity profile for
selected substrates and B) Structures depicted in heat map.
Compounds for which >5% conversion was observed are
highlighted in green; those halogenated to a trace extent are
highlighted in yellow.
Figure 4. Conversion data from initial activity profile and
structures for substrates 22-28.
Closer examination of the activity profiles revealed sig-
nificant differences in the specific substrates halogenated
by each Trp-FDH (Fig. 3A). Many instances arose in
which significant conversion (>5%) was observed for Thal
or the engineered variants on substrates where trace or no
activity was observed for RebH (Fig. 3B, 1-12). A number
of cases were also observed in which neither RebH nor
Thal displayed any conversion on a substrate, but trace
activity was seen for the engineered halogenases (Fig. 3B,
Selectivity of FDH Halogenation
The conversion data for substrates from panels 1 and 2
provide valuable insights into the scope and specificity of
the eight FDHs in this study. To determine whether site
selective halogenation occurred in reactions of non-native
substrates, however, RebH- and Rdc2-catalyzed reactions
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 8
that provided >5% conversion under high throughput
screening conditions (36 total) were conducted on a 1-10
mg scale. roducts for 29 of these reactions were obtained
from preparative LCMS. Comparable conversion to the
initial substrate panel evaluation was generally observed
(Fig. 2). Characterization of these products by NMR spec-
troscopy and HRMS established that most of these sub-
strates (22) were halogenated at a single position with
>95% selectivity (Fig.5, representative examples; Table S2,
all products).
ic parameters were used to rationalize the regioselectivity
1
2
3
4
5
6
7
8
of electrophilic aromatic substitution reactions involving
small molecule reagents.30,32-34 HalA values for each sp2
hybridized carbon on each substrate for which halogena-
tion selectivity was unambiguously assigned were calcu-
lated (29 substrates, Table S2), and a plot of conversion
versus HalA at each site was constructed (Fig. 6).
P
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
Figure 6. Conversion versus halenium affinity (HalA) for
each sp2 carbon on each substrate in Table S2. HalA ranges
for panel 1 substrates halogenated by RebH and panel 2 sub-
strates halogenated by Rdc2 are highlighted in blue and
green, respectively. The inset shows representative HalA
values for common arenes (not substrates).
Several useful trends can be obtained from this plot.
Good conversion of panel 1 substrates by RebH (Fig. 6,
blue diamonds), occurs only at sites with HalA values
>160 kcal/mol. Poor conversion of panel 2 substrates,
which all possess HalA values <164 kcal/mol, was ob-
served for RebH (Fig. 6, red squares), but Rdc2 provides
significant conversion of panel 2 substrates at sites with
HalA values as low as 152 kcal/mol (Fig. 6, green circles).
Similarly, engineered RebH variants were able to halo-
genate sites on panel 2 substrates with HalA values as low
as 154 kcal/mol (Fig. 6, orange triangles). Sites on sub-
strates with HalA values above the minimum for each
enzyme, however, were often not halogenated, so other
factors clearly impact FDH selectivity. In short, our initial
data suggest that there is a minimum electronic activa-
tion (HalA~154 kJ/mol) necessary for halogenation by the
FDHs examined, and this can serve as a coarse filter for
identifying potential new substrates for these enzymes.
To better understand the selectivity of the reactions sub-
strate binding must be considered (vida infra).
Figure 5. Conversions for substrates with multiple electroni-
cally activated sites that provided a single halogenated prod-
uct. Reactions were conducted on 1-10 mg scale using either
RebH (29a-43a) or Rdc2a (25a, 27a). See Table S2 for full
product list. Conversion and selectivity for 27a determined
by comparison with authentic material.
Quantitative Evaluation of Substrate Activation
As previously noted, FDH catalyzed halogenation is be-
lieved to proceed via electrophilic aromatic substitution.6
Substrates bound within the FDH active site are presum-
ably bound such that a single site is located proximal to
an electrophilic halogen species (a Lys79-chloramine and
hypohalous acid have been proposed)8,9 that formally
transfers a halenium ion (X+) to the substrate. This ena-
bles Trp-FDHs to halogenate the benzo ring of trypto-
phan over the more reactive pyrrolo ring. The selectivity
of the reactions whose products are shown in Figure 5 are
similarly remarkable given that more than one site on
each of substrates might reasonably be considered suffi-
ciently activated for electrophilic aromatic substitution.30
Importantly, however, this level of catalyst control was
achieved on diverse substrates without the need for mul-
tiple halogenase engineering efforts.10
Substrate Docking
Deviation of FDH selectivity from outcomes expected
based on electronic effects alone likely results in part
from the unique binding of different substrates within
FDH active sites.7,10,16 Sixteen substrates for which the
selectivity of RebH- was established (Fig. 5) were there-
fore docked into a multi-conformer model of RebH to
determine if computationally inexpensive methods35,36
could improve on a purely electronic view of RebH selec-
tivity. Docking was performed using ICM Molsoft dock-
ing37 and ROCS pharmacophore overlay38. The former
We previously used calculated halenium affinity (Ha-
lA)31 values to quantitate the extent to which the observed
selectivity of RebH variants overrides the electronic pref-
erence of tryptamine toward (hypothetical) halogenation
by Cl+ at different positions.10 This approach is analogous
to several previous examples in which calculated electron-
ACS Paragon Plus Environment

Page 5 of 8
ACS Catalysis
and charge,38 was also used. As in our previous work, cata-
employs a Monte Carlo minimization algorithm com-
1
2
3
4
5
6
7
8
bined with empirical energy terms to model van der
Waals forces, hydrophobic properties, and electrostatics.39
This procedure returns multiple energetically favorable
poses of a ligand within a target, but it is blind to known
binding interactions. Given the existence of several crystal
structures of RebH-tryptophan complexes, the ROCS
pharmacophore overlay approach, which aligns query
molecules to a given template based on molecular shape
lytically relevant poses were taken to be those in which an
sp2 hybridized carbon was positioned proximal to Lys79
in the RebH active site.10
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
Figure 7. Poses observed for ROCS pharmacophore overlay. Tryptophan in the RebH-tryptophan complex (PDB 2OA1) is
shown in pink. Lys79 is labeled in each model, and the site of halogenation for each substrate is indicated with a red ar-
row. Structures for each substrate are shown with HalA values for each site and the site halogenated in green.
In general, neither method provided significant predic-
tive ability regarding substrate specificity (i.e. whether or
not a given substrate would be halogenated). Many unre-
active substrates were nonetheless predicted to bind in
the RebH active site. In addition, while ICM Molsoft
docking returned several poses for each substrate within
the RebH active site, very few of these appeared catalyti-
cally relevant. This was likely due to both the relatively
large size of the binding pocket and the relatively small
size of the compounds, allowing for multiple high scoring
(energetically feasible) docking poses.36 ROCS pharmaco-
phore overlay, on the other hand, provided a number of
interesting results regarding the site selectivity of sub-
strates that were halogenated (Fig. 7, representative ex-
amples). Perhaps most notably, panels A-D show cases
where the site halogenated does not have the highest Ha-
lA, but is predicted to bind proximal to Lys79. This selec-
tivity is inconsistent with purely electronic effects (sub-
strate control) but is consistent with predicted binding
within the active site (catalyst control), suggesting that
ACS Paragon Plus Environment

ACS Catalysis
FDHs could be useful for obtaining novel selectivity on
Page 6 of 8
1
2
3
4
5
6
7
8
diverse structures. Moreover, the pharmacophore overlay
predicts that this selectivity results from different polar
functional groups distal to the site of halogenation bind-
ing to the pocket occupied by the substrate amino acid
moiety in the crystallized RebH-tryptophan complex.
Similar anchoring effects have been exploited for other
enzymes,40 but these results suggest that a range of differ-
ent functional groups and even heterocycles can fulfill
this role in RebH.
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
Several substrates, including the one shown in panel E,
were predicted to bind such that the most electronically
activated site was proximal to Lys79, but cases in which
electronic effects appear to outweigh any predicted bind-
ing effects were also found. In panels F-H for example,
substrates were halogenated at the most electronically
activated site even when this site was predicted to bind
distal to Lys79. Presumably, in these cases, alternate sub-
strate orientations within the RebH active site can be
sampled, leading to halogenation at the most activated
site(s). While ICM Molsoft docking tended to underrepre-
sent binding modes similar to native Trp binding, phar-
macophore overlay emphasized such binding modes, both
of which reduce predictive utility. When combined with
HalA values, however, computationally inexpensive dock-
ing simulations appear to provide insight into the ob-
served selectivity of FDH-catalyzed halogenation on non-
native substrates in many cases.
Controlling Site Selectivity Using Different FDHs
The catalyst control exhibited by RebH toward various
substrates also suggests that different FDHs could be used
to produce different halogenated products. To explore
this possibility, halogenation of several substrates with
multiple electronically activated sites that were selectively
halogenated by RebH or Rdc2 was also examined using
4V, 6TL, and 10S (Table 1, entries 1-6). Two reactions in
which RebH provided mixtures of compounds were also
examined to determine if other enzymes could improve
selectivity (Table 1, entries 7 and 8). Finally, reactions of
two phenols were also examined (Table 1, entries 9 and
10). To compare the product profiles of different halogen-
ases with those produced by a common stoichiometric
chlorinating reagent, these reactions were also conducted
with N-chlorosuccinimide (NCS)41.
Substrate selectivity profiles for the different halogenat-
ing species are presented in Table 1, with results for en-
zymes that provide similar profiles removed for clarity.
Altered site-selectivity was observed not only between
FDHs, but also between the FDHs and NCS. A complete
switch in site-selectivity is observed using different FDHs
for substrates such as 27, 41 and 46. A remarkable exam-
ple of this is phenol (27), for which Rdc2 halogenates ex-
clusively ortho to the alcohol, while 4V only halogenates
para. This particular example highlights how effective
enzymatic catalyst control can be even on simple non-
native substrates. For many substrates within this subset,
an FDH gave a single isomer when mixtures were seen
with NCS (Table 1; entries 2-6, 10). In addition, NCS often
begins to di-halogenate substrates to significant amounts
before complete consumption of starting material. This
problem is circumvented by using FDHs, which typically
have decreased activity on halogenated products (Table 1;
entries 2-4, 7). FDHs are also catalytic and use chloride as
a halide sources and air as a terminal oxidant.2
Table 1. Relative product distributions for halogenation of
representative substrates using FDHs and NCS. Sites of halo-
genation (P1-3) are indicated for select substrates.
For certain substrates, such as 46 (Table 1; entry 8), dif-
ferent ratios of two mono-chlorinated products can be
observed for NCS, RebH and 4V. The use of FDHs pre-
sents the unique opportunity in such cases to tune exist-
ing selectivity through directed evolution.10 For example,
RebH halogenates primarily para to the methyl group, but
is not highly selective for this site (70%). RebH could be
tuned to give better selectivity at this position. On the
6
ACS Paragon Plus Environment

Page 7 of 8
ACS Catalysis
These results constitute the largest examination of FDH
other hand, 4V halogenates primarily ortho to the methyl
1
2
3
4
5
6
7
8
group (77%) and could be used as a starting point to se-
lectivity halogenate this position is desired.
substrate scope and selectivity to date and significantly
improve our understanding of the types of substrates that
can be halogenated using these enzymes. The range of
substrates that could be halogenated clearly attests to the
potential for halogenases evolved for one application to
find utility toward another. Moreover, this broad sub-
strate scope should facilitate the identification of novel
FDHs, since native substrates are not necessarily required
to confirm FDH activity.18 Given the unique scope and
selectivity of the enzymes examined in this study relative
to one another and to NCS, the use of probe substrates for
functional characterization of putative FDHs holds great
promise for expanding the utility of enzymatic halogena-
tion.1
Discussion
In this work, the substrate scope of several FDHs was
probed by measuring their activity toward diverse panels
of low molecular weight aromatic compounds. The result-
ing substrate activity profiles contain information that
cannot be obtained by studying enzyme activity on native
substrates or simple derivatives of these substrates.18,19,23
While substitution of compounds distal to sites of reac-
tion is commonly used to probe electronic effects in or-
ganic reactions (e.g. LFERs),42 this can perturb substrate
binding as well as electronic effects within an enzyme
active site,43 leading to spurious correlations between
substitution and activity. The points in Fig. 6 correspond-
ing to substrate sites with sufficient electronic activation
(as indicated by calculated HalA values) but for which
halogenation was not observed likely exemplify this phe-
nomenon. On the other hand, sampling a sufficient range
of substrate classes, substituents, and substitution pat-
terns provides general information regarding the halo-
genating ability of different enzymes. Analogous efforts
on other enzymes could prove similarly enlightening.
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
AUTHOR INFORMATION
Corresponding Author
*Department of Chemistry, University of Chicago, Chicago,
IL 60637
E-mail: jaredlewis@uchicago.edu
*Global Discovery Chemistry, Novartis Institutes for Biomed-
ical Research, 250 Massachusetts Ave, Cambridge, MA 02139
E-mail: kian.tan@novartis.com
Author Contributions
The small set of FDHs examined provided remarkably
broad substrate scope considering the substantial varia-
tion of the substrates evaluated relative to the native sub-
strates of these enzymes. A wide range of functional
groups were tolerated, but it is worth noting that this
would not necessarily be apparent from evaluating indi-
vidual substrates. For example, the simple ethylsulfona-
mide-substituted aniline 51 (Table S1) was not halogenat-
ed, but sulfonamides 2 and 10 were both halogenated,
showing that minor changes in substituents distal to reac-
tion sites can significantly impact substrate reactivity.
Similarly, compounds 21 (Fig. 3) and 44 (Table 1, entry 5)
are regioisomeric methyl (aminohydroxybenzoates), but
while 44 is halogenated in high conversion by multiple
FDHs (100% with 4V, 56% with 6TL, 55% with RebH), 21
was halogenated only by 0K to a trace amount. Examining
activity on diverse substrates23 can account for these dif-
ferences and provide an accurate picture of the types of
functional groups and fragments that can potentially be
halogenated. The molecular context in which these moie-
ties appear, however, will clearly influence their impact
on FDH catalysis.
All authors have given approval to the final version of the
manuscript.
ASSOCIATED CONTENT
Supporting Information
A full description of materials, a complete substrate list, ex-
perimental details, and characterization of all compounds
and enzymes are provided in the Supporting Information.
The Supporting Information is available free of charge on the
ACS Publications website.
ACKNOWLEDGMENT
This work was supported by the NIH (1R01GM115665), The
Novartis Institutes for Biomedical Research, and the NSF
under the CCI Center for Selective C–H Functionalization
(CHE-1205646). MCA was supported by an NIH Chemistry
and Biology Interface training grant (T32 GM008720), an NSF
predoctoral fellowship (DGE-1144082), a University of Chica-
go Department of Chemistry Helen Sellei-Beretvas Fellow-
ship, and an ARCS Scholar Award. This work was completed
in part with resources provided by the University of Chicago
Research Computing Center with assistance from Dr. Jona-
than Skone. We thank Prof. Karl-Heinz van Pée for providing
a plasmid encoding Thal, Clayton Springer for building mod-
els of halenium adducts for all substrates studied, and Laurel
Heckman and Randy Richter for housing MCA.
Subsequent analysis of substrates that undergo halo-
genation can then be used to rationalize how substrate
properties impact FDH selectivity. Each sp2 carbon on
each substrate constitutes a potential reaction site, so
conversion values for each site can be plotted against ste-
ric or electronic descriptors. The use of calculated HalA
values31 in this work indicated clear cutoffs in electronic
activation that are required for substrate halogenation by
FDHs, but electronics alone could not explain the ob-
served selectivity in most cases. Docking simulations pro-
vided further insight into selectivity when combined with
HalA values, and more rigorous computational methods
(e.g. MD simulations)44 could significantly improve the
predictive models for FDH activity and selectivity.
ABBREVIATIONS
FDH, flavin-dependent halogenase; Trp-FDH, tryptophan-
flavin-dependent halogenase; EAS, electrophilic aromatic
substitution;
HalA,
halenium
affinity;
NCS,
N-
chlorosuccinamide; LFER, linear free energy relationship
REFERENCES
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 8
(1)
Weichold, V.; Milbredt, D.; van Pée, K.-H. Angew.
(30)
(31)
Pratihar, S.; Roy, S. J. Org. Chem. 2010, 75 (15), 4957–
4963.
Ashtekar, K. D.; Marzijarani, N. S.; Jaganathan, A.;
Holmes, D.; Jackson, J. E.; Borhan, B. J. Am. Chem.
Soc. 2014, 136 (38), 13355–13362.
1
2
3
Chem. Int. Ed. 2016, 55 (22), 6374–6389.
Vaillancourt, F. H.; Yeh, E.; Vosburg, D. A.; Garneau-
Tsodikova, S.; Walsh, C. T. Chem. Rev. 2006, 106 (8),
3364–3378.
(2)
4
5
6
7
8
9
(3)
Payne, J. T.; Andorfer, M. C.; Lewis, J. C. Engineering
Flavin-Dependent Halogenases, 1st ed.; Methods in
Enzymology; Elsevier Inc., 2016; Vol. 575, pp 93–126.
van Pée, K.-H.; Patallo, E. Appl. Microbiol. Biot. 2006,
70 (6), 631–641.
Yeh, E.; Garneau, S.; Walsh, C. T. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102 (11), 3960–3965.
Dong, C.; Flecks, S.; Unversucht, S.; Haupt, C.; van
Pée, K.-H.; Naismith, J. H. Science 2005, 309 (5744),
2216–2219.
Zhu, X.; Laurentis, W. D.; Leang, K.; Herrmann, J.;
Ihlefeld, K.; van Pée, K.-H.; Naismith, J. H. J. Mol. Biol.
2009, 391 (1), 74–85.
Yeh, E.; Blasiak, L. C.; Koglin, A.; Drennan, C. L.;
Walsh, C. T. Biochemistry 2007, 46 (5), 1284–1292.
Flecks, S.; Patallo, E. P.; Zhu, X.; Ernyei, A. J.; Seifert,
G.; Schneider, A.; Dong, C.; Naismith, J. H.; van Pée,
K.-H. Angew. Chem. Int. Ed. 2008, 47 (49), 9533–9536.
Andorfer, M. C.; Park, H.-J.; Vergara-Coll, J.; Lewis, J.
C. Chem. Sci. 2016, 7 (6), 3720–3729.
Zeng, J.; Zhan, J. ChemBioChem 2010, 11 (15), 2119–2123.
Zeng, J.; Lytle, A. K.; Gage, D.; Johnson, S. J.; Zhan, J.
Bioorg. Med. Chem. Lett. 2013, 23 (4), 1001–1003.
Chooi, Y.-H.; Cacho, R.; Tang, Y. Chem. Biol. 2010, 17
(5), 483–494.
Durak, L. J.; Payne, J. T.; Lewis, J. C. ACS Catal. 2016, 6
(3), 1451–1454.
Payne, J. T.; Poor, C. B.; Lewis, J. C. Angew. Chem. Int.
Ed. 2015, 54 (14), 4226–4230.
Shepherd, S. A.; Karthikeyan, C.; Latham, J.; Struck,
A.-W.; Thompson, M. L.; Menon, B. R. K.; Styles, M.
Q.; Levy, C.; Leys, D.; Micklefield, J. Chem. Sci. 2015, 6
(6), 3454–3460.
Latham, J.; Henry, J.-M.; Sharif, H. H.; Menon, B. R.
K.; Shepherd, S. A.; Greaney, M. F.; Micklefield, J. Nat.
Commun. 2016, 7, 1–8.
Landwehr, M.; Carbone, M.; Otey, C. R.; Li, Y.; Ar-
nold, F. H. Chem. Biol. 2007, 14 (3), 269–278.
Zhang, K.; Damaty, El, S.; Fasan, R. J. Am. Chem. Soc.
2011, 133 (10), 3242–3245.
Payne, J. T.; Andorfer, M. C.; Lewis, J. C. Angew.
Chem. Int. Ed. 2013, 52 (20), 5271–5274.
Seibold, C.; Schnerr, H.; Rumpf, J.; Kunzendorf, A.;
Hatscher, C.; Wage, T.; Ernyei, A. J.; Dong, C.; Nai-
smith, J. H.; van Pée, K.-H. Biocatal. Biotransfor-
mation 2006, 24 (6), 401–408.
Frese, M.; Sewald, N. Angew. Chem. Int. Ed. 2014, 54
(1), 298–301.
Carr, R.; Alexeeva, M.; Enright, A.; Eve, T. S. C.; Daw-
son, M. J.; Turner, N. J. Angew. Chem. Int. Ed. 2003, 42
(39), 4807–4810.
(32)
Tishchenko, O.; Pham-Tran, N.-N.; Kryachko, E. S.;
Nguyen, M. T. J. Phys. Chem. A 2001, 105 (38), 8709–
8717.
Schwabe, T.; Grimme, S. Eur. J. Org. Chem. 2008, 2008
(35), 5928–5935.
Aoki, S.; Matsuo, Y.; Ogura, S.; Ohwada, H.;
Hisamatsu, Y.; Moromizato, S.; Shiro, M.; Kitamura,
M. Inorg. Chem. 2011, 50 (3), 806–818.
Stiefl, N.; Gedeck, P.; Chin, D.; Hunt, P.; Lindvall, M.;
Spiegel, K.; Springer, C.; Biller, S.; Buenemann, C.;
Kanazawa, T.; Kato, M.; Lewis, R.; Martin, E.; Polya-
kov, V.; Tommasi, R.; van Drie, J.; Vash, B.; White-
head, L.; Xu, Y.; Abagyan, R.; Raush, E.; Totrov, M. J.
Chem. Inf. Model. 2015, 55 (4), 896–908.
Cross, J. B.; Thompson, D. C.; Rai, B. K.; Baber, J. C.;
Fan, K. Y.; Hu, Y.; Humblet, C. J. Chem. Inf. Model.
2009, 49 (6), 1455–1474.
Abagyan, R.; Totrov, M.; Kuznetsov, D. J. Comput.
Chem. 1994, 15 (5), 488–506.
Grant, J. A.; Gallardo, M. A.; Pickup, B. T. J. Comput.
Chem. 1996, 17 (14), 1653–1666.
Schapira, M.; Totrov, M.; Abagyan, R. J. Mol. Recog.
1999, 12 (3), 177–190.
Negretti, S.; Narayan, A. R. H.; Chiou, K. C.; Kells, P.
M.; Stachowski, J. L.; Hansen, D. A.; Podust, L. M.;
Montgomery, J.; Sherman, D. H. J. Am. Chem. Soc.
2014, 136 (13), 4901–4904.
Ponce, M. A.; Tarzi, O. I.; Erra-Balsells, R. J. Heterocy-
clic Chem. 2003, 40 (3), 419–426.
Anslyn, E. V.; Dougherty, D. A. Modern Physical Or-
ganic Chemistry; University Science Books, 2006; pp
441-471.
Weise, N. J.; Parmeggiani, F.; Ahmed, S. T.; Turner, N.
J. J. Am. Chem. Soc. 2015, 137 (40), 12977–12983.
Narayan, A. R. H.; Jiménez-Osés, G.; Liu, P.; Negretti,
S.; Zhao, W.; Gilbert, M. M.; Ramabhadran, R. O.;
Yang, Y.-F.; Furan, L. R.; Li, Z.; Podust, L. M.; Mont-
gomery, J.; Houk, K. N.; Sherman, D. H. Nat. Chem.
2015, 7 (8), 653–660.
(4)
(5)
(6)
(33)
(34)
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
(35)
(7)
(8)
(9)
(36)
(37)
(38)
(39)
(40)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(41)
(42)
(43)
(44)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
(25)
(26)
(27)
(28)
(29)
Erlanson, D. A.; McDowell, R. S.; O'Brien, T. J. Med.
Chem. 2004, 47 (14), 3463–3482.
Heemstra, J. R.; Walsh, C. T. J. Am. Chem. Soc. 2008,
130 (43), 14024–14025.
Lewis, J. C.; Arnold, F. H. CHIMIA 2009, 63 (6), 309–
312.
Matsumura, I.; Ellington, A. D. J. Mol. Biol. 2001, 305
(2), 331–339.
Rentmeister, A.; Arnold, F. H.; Fasan, R. Nat. Chem.
Biol. 2009, 5 (1), 26–28.
Truppo, M. D.; Escalettes, F.; Turner, N. J. Angew.
Chem. Int. Ed. 2008, 47 (14), 2639–2641.
8
ACS Paragon Plus Environment