Combating Enhanced Intracellular Survival (Eis)-Mediated Kanamycin Resistance of Mycobacterium tuberculosis by Novel Pyrrolo[1,5-a]pyrazine-Based Eis Inhibitors

Article abstract of DOI:10.1021/acsinfecdis.6b00193

Tuberculosis (TB) remains one of the leading causes of mortality worldwide. Hence, the identification of highly effective antitubercular drugs with novel modes of action is crucial. In this paper, we report the discovery and development of pyrrolo[1,5-a]p

Full text of DOI:10.1021/acsinfecdis.6b00193

Article
pubs.acs.org/journal/aidcbc
Combating Enhanced Intracellular Survival (Eis)-Mediated
Kanamycin Resistance of Mycobacterium tuberculosis by Novel
Pyrrolo[1,5‑a]pyrazine-Based Eis Inhibitors
Atefeh Garzan,† Melisa J. Willby,‡ Huy X. Ngo,† Chathurada S. Gajadeera,† Keith D. Green,†
Selina Y. L. Holbrook,† Caixia Hou,† James E. Posey,*,‡ Oleg V. Tsodikov,*,†
and Sylvie Garneau-Tsodikova*,†
†Department of Pharmaceutical Sciences, College of Pharmacy, University of Kentucky, 789 South Limestone Street, Lexington,
Kentucky 40536-0596, United States
‡Laboratory Branch, Division of Tuberculosis Elimination, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention,
Centers for Disease Control and Prevention, Atlanta, Georgia 30329, United States
S
* Supporting Information
ABSTRACT: Tuberculosis (TB) remains one of the leading causes of mortality worldwide.
Hence, the identification of highly effective antitubercular drugs with novel modes of action is
crucial. In this paper, we report the discovery and development of pyrrolo[1,5-a]pyrazine-based
analogues as highly potent inhibitors of the Mycobacterium tuberculosis (Mtb) acetyltransferase
enhanced intracellular survival (Eis), whose up-regulation causes clinically observed resistance to
the aminoglycoside (AG) antibiotic kanamycin A (KAN). We performed a structure−activity
relationship (SAR) study to optimize these compounds as potent Eis inhibitors both against
purified enzyme and in mycobacterial cells. A crystal structure of Eis in complex with one of the
most potent inhibitors reveals that the compound is bound to Eis in the AG binding pocket,
serving as the structural basis for the SAR. These Eis inhibitors have no observed cytotoxicity to
mammalian cells and are promising leads for the development of innovative AG adjuvant
therapies against drug-resistant TB.
KEYWORDS: aminoglycoside acetyltransferase, bacterial resistance, enzyme inactivation, drug combination,
structure−activity-relationship analysis
uberculosis (TB), caused by the pathogenic Mycobacte-
rium tuberculosis (Mtb), is the deadliest global bacterial
not pharmacologically relevant.14 A better approach would be
to use inhibitors of Eis as adjuvant therapies in combination
with KAN to combat or forestall the emergence of KAN
resistance through Eis up-regulation. We recently reported the
discovery and development of Eis inhibitors with an isothiazole
S,S-dioxide heterocyclic core,15 and sulfonamide-based,16
methyl 4H-furo[3,2-b]pyrrole-5-carboxylate,17 and 3-(1,3-diox-
olano)-2-indolinone17 scaffolds, which yielded compounds that
potently inhibited Eis in vitro and abolished KAN resistance of
the Mtb mutant strain K204 that is KAN-resistant due to Eis
up-regulation. We previously reported 25 hit compounds
identified by high-throughput screening (HTS) of a library
composed of ∼23,000 small molecules that displayed Eis
inhibitory activities.18 Here, we pursue one of these preliminary
hits (compound 1a*, Scheme 1A) and report the chemical
synthesis of this compound and that of 47 analogues (Scheme
1B), along with their biochemical and biological studies.
Among compounds in this series, we have generated novel and
promising Eis inhibitors that not only efficiently inhibit the
purified enzyme but also restore KAN sensitivity of KAN-
T
infection. The number of people infected by Mtb is currently
estimated to be 2−3 billion worldwide.1 The situation is greatly
aggravated by the emergence and spread of difficult or
impossible to treat multidrug-resistant (MDR) and extensively
drug-resistant (XDR) TB.2 Aminoglycoside (AG) antibiotics
such as kanamycin A (KAN) and amikacin (AMK) are used for
the treatment of MDR and XDR-TB. However, successful
outcomes of the treatment of these MDR- and XDR-Mtb
strains can be impeded as a result of KAN resistance.3 We
previously discovered that up-regulation of the enhanced
intracellular survival (eis) gene, due to point mutations in its
promoter, is a cause of resistance to KAN in one-third of KAN-
resistant Mtb infections in the clinic.4,5 The development of
new AGs or use of enhanced intracellular survival (Eis)
inhibitors are two potential solutions for overcoming the effect
of Eis in Mtb. We recently demonstrated that Eis is an AG
acetyltransferase (AAC) found in a variety of bacterial strains
that can inactivate AGs via a multiacetylation mechanism.6−11
As Eis has been shown to multiacetylate a wide variety of
molecules,12,13 including many AG scaffolds, the development
of new AGs to combat its action is not likely to be successful.
Some metal salts are inhibitory to Eis, but this strategy alone is
Received: November 11, 2016
Published: February 13, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX

ACS Infectious Diseases
Article
oxidative aromatization instead of hydrogenation. More details
about this heteroaromatic aromatization were summarized in an
excellent review.19 Compounds 11 and 12 were further reacted
with the different substituted 2-bromoacetophenones to furnish
the desired fully aromatized analogues 1a*−k* and 2a*−k*.
These compounds were evaluated for Eis inhibition using the
clinically relevant KAN as the AG substrate (IC50 values in
Table 1).
Scheme 1. (A) Structures of All Compounds Generated in
This Study; (B) Synthetic Scheme Used to Prepare the
Compounds in Panel A
We first tested whether the freshly synthesized parent
compound 1a* was indeed a potent Eis inhibitor. Expectedly,
the freshly synthesized compound 1a* was found to display
potent inhibition of Eis (IC50 = 0.064 0.008 μM), which was
∼6-fold better than the IC50 value of the commercially available
compound 1a* (IC50 = 0.36 0.03 μM) from our previous
HTS. (Freshly synthesized powders are often more active than
HTS library compounds, which may degrade upon storage.18)
The hit scaffold 1a* contains a pyrrolo[1,5-a]pyrazine core, a
phenyl ring adjacent to the pyrrolo[1,5-a]pyrazine core
(containing R1), and an acetophenone moiety (containing
R2). A comparison of the chemical structure of compound 1a*
with those of the previously published isothiazole S,S-dioxide-
based Eis inhibitors cocrystallized with Eis (Figure S160)15
suggested that 1a* binds to Eis at the AG binding pocket,20
similarly to the isothiazole S,S-dioxides. Examination of the Eis
crystal structure bound to the AG tobramycin (TOB) (PDB:
4JD620) indicated that the positively charged pyrrolo[1,5-
a]pyrazine core is presumably essential for binding to the
negatively charged AG binding pocket and, thus, should not be
modified. On the basis of our previous survey of the hits of this
HTS,18 we also determined that the phenyl ring adjacent to the
pyrrolo[1,5-a]pyrazine core (containing R1) is likely important
for Eis inhibition. In fact, replacing the phenyl ring of 1a* with
an ethyl group resulted in a 25-fold reduction in the inhibitory
activity (IC50 = 9.25
1.50 μM).18 Also, from the crystal
structure of Eis in complex with the isothiazole S,S-dioxide-
based Eis inhibitors, we have rationalized that this phenyl ring is
important due to its snug fit in a hydrophobic binding pocket in
the AG binding cavity. On the other hand, the π-electron-rich
acetophenone moiety (containing R2) and the fully aromatic
pyrrolo[1,5-a]pyrazine core were predicted to be crucial for
binding due to potential π−π interactions with aromatic
residues within the Eis binding pocket. However, it remains
unexplored whether and which substitutions at R1 and R2
positions would be beneficial. We hypothesized that (i) tailor
fitting the Eis binding pocket by introducing subtle
modifications at R1 and R2 would lead to the discovery of
novel optimized inhibitors from our hit scaffold 1a* and (ii)
disruption of the aromaticity of the pyrrolo[1,5-a]pyrazine core
would be detrimental to the binding affinity of the molecule to
the Eis binding pocket. In our biochemical analysis, we will first
examine the aromatic compounds and then explore their
nonaromatic counterparts. Both the aromatic and nonaromatic
molecules are divided into two series. In series 1, R1 was kept
constant (R1 = H), and various substituted acetophenones were
installed onto the pyrrolo[1,5-a]pyrazine core (changing R2).
Similarly, in series 2, R1 was kept constant (R1 = p-F), and the
same various substituted acetophenones were installed onto the
pyrrolo[1,5-a]pyrazine core (changing R2). For the non-
aromatic compounds, four additional members were added to
a third series (series 3) in which R1 was m,p-di-F.
resistant Mtb bacteria. We also present a crystal structure of Eis
in complex with CoA and one potent inhibitor (compound
2k*), which explains the structure−activity relationship (SAR).
Compound 1a* and 47 additional analogues 1a−3k with
different R1 and R2 substituents on the two phenyl rings and
either a fully aromatized (indicated by an asterisk after the
compound number) or a nonaromatized pyrrolo[1,5-a]-
pyrazine core were generated for a thorough SAR analysis of
Eis inhibition (Scheme 1B). The synthesis of all compounds
started with a reaction between the commercially available
pyrrole and 2-chloroethylamine, which afforded compound 4 in
quantitative yield. Compound 4 was reacted with different
substituted benzoyl chlorides to obtain amides 5−7 in 66−71%
yields. The resulting amides were mixed with phosphorus(V)
oxychloride to generate cyclized products 8−10. Then,
compounds 8−10 were reacted with various substituted 2-
bromoacetophenones to obtain the desired nonaromatized
products 1a−k, 2a−k, and 3a,d,h,k. To generate the aromatized
counterparts of these products, compounds 8 and 9 were first
aromatized in the presence of Pd/C to generate molecules 11
and 12. Conventionally, Pd/C is a hydrogenation catalyst. In
the absence of hydrogen gas, Pd/C is known to catalyze an
We began our analysis of the aromatic compounds by
investigating series 1. To probe the ortho position of the
acetophenone moiety, compound 1b* (R1 = H, R2 = o-F) was
B
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX

ACS Infectious Diseases
Article
C
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX

ACS Infectious Diseases
Article
generated and found to display a ∼3-fold reduction in Eis
inhibitory activity (IC50 = 0.21 0.05 μM) when compared to
the freshly synthesized parent 1a* (From here on, when
comparing to 1a*, we refer to the freshly synthesized 1a*),
indicating that ortho substitution was not beneficial. We then
explored substitutions at the meta position with compounds
1c*−1g*. The meta-substituted compound 1c* (R1 = H, R2 =
m-F) was found to have Eis inhibitory activity (IC50 = 0.06
0.01 μM) comparable to that of the parent compound 1a*. We
systematically increased the size of the halogen substituents in
compounds 1d* (R1 = H, R2 = m-Cl) and 1e* (R1 = H, R2 = m-
Br). Interestingly, compound 1d* displayed an IC50 value of
0.025 0.006 μM, which was ∼3-fold smaller than that of the
parent compound 1a*. On the other hand, compound 1e*
(IC50 = 0.34 0.10 μM) was not as potent as compound 1d*,
suggesting that the Br substituent was possibly too sterically
hindered and, thus, not well tolerated in the Eis binding pocket.
Compound 1f* (R1 = H, R2 = m-OH) (IC50 = 1.8 0.5 μM)
was less potent than the parent compound, which suggested
that having a highly polar substituent could be disfavored.
Alternatively, replacing the hydroxyl group by a methoxy
(compound 1g* (R1 = H, R2 = m-OMe)) yielded a molecule
with improved Eis inhibition (IC50 = 0.21 0.07 μM) when
compared to 1f*. Overall, we found that Cl was the best
substituent at the meta position. We pondered whether this
trend would translate to the para position, which prompted us
to evaluate compounds 1h* (R1 = H, R2 = p-F), 1i* (R1 = H,
R2 = p-Cl), and 1j* (R1 = H, R2 = p-Br). Intriguingly, the
smaller F substituent was optimal at the para position with an
IC50 value of 0.029 0.007 μM, contrary to what we observed
at the meta position. Also, inhibitor 1h* displayed an IC50 of
= p-F, R2 = m-Br; IC50 = 0.06 0.02 μM) was much more
potent, which pointed to the possibility that changing R1 from
H to p-F could lead to a slight variation in the binding
orientation of the molecule, especially near the meta position of
the acetophenone ring. Additionally, the m-hydroxy and m-
methoxy substitutions, as in the cases of 2f* (R1 = p-F, R2 = m-
OH) and 2g* (R1 = p-F, R2 = m-OMe) either completely
abolished Eis inhibitory activity or resulted in moderate
inhibition of the enzyme (IC50 = 8.4 2.8 and 0.53 0.11
μM, respectively). This was consistent with the observations
made with compounds 1f* (R1 = H, R2 = m-OH) and 1g* (R1
= H, R2 = m-OMe). For the para-substituted analogues (2h*
(R1 = p-F, R2 = p-F), 2i* (R1 = p-F, R2 = p-Cl), and 2j* (R1 =
p-F, R2 = p-Br)), similarly to what was observed with series 1,
the larger halogen substituents were generally less favorable
with activities varying in the range of 0.13−0.50 μM. We also
evaluated the para-methylated analogue 2k* (R1 = p-F, R2 = p-
Me) and found that 2k* displayed excellent activity with an
IC50 value of 0.08 0.03 μM, which was 2-fold better than that
of 1k* (IC50 = 0.19 0.02 μM).
Having established the general SAR trends for the aromatic
analogues, we next aimed to determine whether their
nonaromatic counterparts (1a−k and 2a−k) would exhibit
decreased activity due to potential disruption of the π−π
interactions with Eis aromatic amino acid residues. Indeed, we
found that most of the nonaromatic analogues generally
displayed less potent Eis inhibition than their aromatic
counterparts did. In 4 of 22 cases, the aromatic and
nonaromatic compounds displayed nearly equipotent inhibition
of Eis. In the case of compounds 1c and 1c* (R1 = H, R2 = m-
F), the IC50 values were virtually the same (IC50 = 0.05 0.01
and 0.06 0.01 μM, respectively). Additionally, compounds 2e
and 2e* (R1 = p-F, R2 = m-Br) were also practically equipotent
in terms of Eis inhibitory activities (IC50 = 0.07 0.02 and 0.06
0.02 μM, respectively). Compounds 1g and 1g* (R1 = H, R2
0.029
0.007 μM, which, similarly to the IC50 values for
several other inhibitors, was smaller than half of the enzyme
concentration used in our assay (0.25/2 = 0.125 μM). This
effect has at least three potential explanations: (1) If inhibition
of one monomer per Eis hexamer by one inhibitor molecule
leads to the loss of activity of the entire hexamer, then the IC50
can, in principle, be as low as 0.125/6 = 0.02 μM. (2) If only a
fraction of Eis protein is active in acetylating KAN and binding
inhibitors (e.g., due to protein aggregation), then the
concentration of active Eis in the assay is an overestimate of
the concentration of active enzyme. (3) If inhibitor binding to
Eis causes Eis aggregation and inactivates multiple hexamers,
then IC50 is also a fraction of half of the enzyme concentration
in the assay. Mechanism 3 is unlikely, because we do not
observe global conformational changes in Eis upon inhibitor
binding in crystal structures. Distinguishing among these
mechanisms of these highly potent analogues is a goal of
ongoing work in our group.
Subsequently, we investigated the effect of R1 substitutions
on the phenyl ring adjacent to the pyrrolo[1,5-a]pyrazine core.
As the p-F substitution was one of the best in terms of activity
when we varied R2 in series 1, we first decided to install the p-F
substituent at the R1 position and generated series 2 analogues.
Analogue 2a* (R1 = p-F, R2 = H) displayed weaker Eis
inhibition (IC50 = 0.35 0.07 μM) than did 1a*. Similarly to
1b* (R1 = H, R2 = o-F), the ortho-substituted analogue 2b* (R1
= p-F, R2 = o-F) was also not optimal. When we increased the
size of the R2 substituents at the meta position, we observed
that the larger halogens led to more potent Eis inhibition (Br =
Cl > F) and yielded compounds with IC50 values varying from
0.06 to 0.22 μM. Unlike compound 1e* (R1 = H, R2 = m-Br;
IC50 = 0.34 0.10 μM), the m-Br substituted analogue 2e* (R1
= m-OMe) (IC50 = 0.15
respectively) were also similar. For the pair 1i and 1i* (R1 = H,
R2 = p-Cl) (IC50 = 0.31 0.09 and 0.56 0.20 μM,
0.04 and 0.21
0.07 μM,
respectively), the nonaromatic counterpart 1i was marginally
better. Regardless of whether the analogue in series 2 was
aromatic or nonaromatic, it was conclusive that at the meta
position of the acetophenone moiety, bigger halogen
substituents such as Cl and Br were generally better suited,
and at the para position of the acetophenone, the smaller F
substituent was the best.
Once our pyrrolo[1,5-a]pyrazine derivatives were optimized
for inhibition of the purified Eis enzyme in vitro, we set out to
confirm whether these compounds could display Eis inhibitory
activity in the Mtb culture by measuring the effect of the
compounds on KAN MIC (MICKAN). Compounds were tested
in combination with KAN against the KAN-sensitive H37Rv
Mtb strain as a control and against the KAN-resistant Mtb
K204, which is H37Rv Mtb bearing a clinically occurring point
mutation in the eis promoter leading to overexpression of Eis.4
Mtb H37Rv has an MICKAN of 1.25 μg/mL, whereas KAN-
resistant Mtb K204 has an MICKAN of ≥10 μg/mL. Active
compounds were expected to resensitize Mtb K204 to KAN.
The compounds were generally tested at concentrations that
were 100-fold higher than their respective IC50 values in the
enzymatic assays, to correct for the variation in the potency of
Eis inhibition. Weakly potent compounds (IC50 > 1 μM) were
tested at 100 μM in the MIC assays. Mtb is notorious for its
highly lipophilic and complex cell envelope, which provides
D
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX

ACS Infectious Diseases
Article
Figure 1. Mammalian cytotoxicity of selected compounds (1e*, 1i*, 2c*, 2h*, and 2i*) alone (represented as dark color columns) or in the
presence of 50 μg/mL (equivalent to 86 μM) KAN (represented as light color columns immediately to the right of the dark color column of the
corresponding compound in the absence of KAN) against (A) A549, (B) HEK-293, and (C) J774A.1 cells. The non-normalized data are presented
in Figure S158.
intrinsic resistance to many antibacterial compounds and
presents an immense challenge for antitubercular drug
discovery. Indeed, as shown in our previous Eis inhibitors
studies,15 some of the most potent in vitro compounds were
not active in Mtb cultures. We also cannot exclude low
solubility or aggregation of the compounds in the culture media
as a reason for poor activity. Herein, we determined the MIC
values for KAN (MICKAN) against Mtb K204 in the absence or
presence of our Eis inhibitors and compared them to the
MICKAN of the drug-sensitive Mtb H37Rv strain. As anticipated,
most compounds caused a reduction in the MICKAN for Mtb
K204, overcoming KAN resistance. Poor Eis inhibitors such as
compounds 1f* and 2f* with relatively high IC50 values were
unable to resensitize Mtb K204 to the action of KAN. These
observations, together with the lack of toxicity of these
inhibitors when used without KAN for either Mtb strains,
validate Eis inhibition as the principal mechanism of MICKAN
reduction by these compounds. Mtb H37Rv, for which MICKAN
is virtually unaffected by the inhibitors, serves as an important
negative control in this regard. Some of the good Eis inhibitors,
such as compounds 1c−1e, 1g, 1h, 2d, 2e, 2h, 3a, 3d, 3h, and
3k, did not resensitize Mtb K204 to the action of KAN despite
their nanomolar IC50 values (MICKAN ≥ 10 μg/mL), indicating
that these molecules may not go through the cell envelope.
Compounds 1a*−1e*, 1g*, 1h*, 1i, 1j*, 1k*, 2a*−2e*, 2g*,
displayed better potency compared to the analogues in series 3
in Mtb culture. Although the charged nature of these
compounds may contribute adversely to the permeability of
the compounds through the greasy mycobacterial cell envelope,
their better solubility in aqueous solution when compared to
other uncharged Eis inhibitors may offset this potential issue.
Therefore, these compounds serve as valuable alternatives to
our previously reported uncharged Eis inhibitors of other
scaffolds in further preclinical development of Eis inhibitors as
KAN adjuvants in TB therapy. Indeed, two compounds, 1i*
and 2h*, are highly promising for future development, as they
completely restore the potency of KAN, fully overcoming Eis
up-regulation.
Cytotoxicity to three different mammalian cell lines was
tested for five representative potent Eis inhibitors (Figure 1) in
the absence and presence of KAN at the concentration of 50
μg/mL (equivalent to 86 μM), greatly exceeding the MICKAN
of Mtb. The negative control corresponded to no inhibitor
treatment and was standardized as 100% cell survival. The
positive control was a treatment with Triton X-100, where we
observed most of the cells killed. The normalized and non-
normalized cytotoxicity data are presented in Figure 1 and
Figure S158, respectively. We observed that at sub-IC50
concentrations, our Eis inhibitors induced cell proliferation,
thereby displaying >100% cell survival. With the exception of
compound 1e*, which at 50 μM exhibited significant
cytotoxicity against one of the three cell lines, none of the
and 2i*−2k* partially restored the activity of KAN (MICKAN
=
2.5−5 μg/mL). Generally, the analogues in series 1 and 2
E
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX

ACS Infectious Diseases
Article
compounds were cytotoxic against any of the three cell lines up
to 50 μM. Three compounds (1i*, 2c*, and 2h*) had no
significant cytotoxicity up to 100 μM, without or with KAN.
The lack of cytotoxicity indicates the absence of toxic off-target
effects in the mammalian cells, strengthening the promise of
these compounds as candidates for animal and clinical studies.
Eis inhibitors appear to be less toxic when they are co-
administered with KAN. Upon assessing the cytotoxicity of
KAN alone (Figure S159) and Eis inhibitors alone (Figure 1
and Figure S158), one can observe that exposure to KAN or to
Eis inhibitors alone at sub-IC50 concentrations promotes cell
growth. This phenomenon of increased growth in the presence
of small quantities of xenobiotics has been previously
observed.21−25 Due to this effect, cotreatment with KAN and
Eis inhibitors may result in a faster cell growth than that for
KAN alone.
To rationalize our SAR study and understand the binding
mode of our inhibitors to Eis, we determined a crystal structure
of Eis in complex with CoA and inhibitor 2k* (one of our best
inhibitors; IC50 = 0.08 0.03 μM) at the resolution of 2.4 Å
(Figure 2 and Table S1). The crystal structure demonstrates
that inhibitor 2k* is indeed bound in the AG binding site
(established by our reported Eis-CoA-TOB crystal structure;20
Figure S160) and is stabilized in the bound state by numerous
hydrophobic interactions with Eis. The pyrazine ring is stacked
between the side chain of Glu401 of Eis and its C-terminal
carboxyl group, with mutually orthogonal π−π interactions of
the pyrazine ring with the indole ring of Trp36. This
observation supports our initial hypothesis that the aromaticity
of the pyrrolo[1,5-a]pyrazine core is crucial for activity and
explains why aromatized compounds resulted in better activities
during SAR analysis. Attached to the pyrrolo[1,5-a]pyrazine
core, the acetophenone ring displayed parallel π−π interaction
with Phe84 and showed that our prediction about the
importance of the aromatic acetophenone was correct.
Additionally, the R2-substituted acetophenone fits in a mainly
hydrophobic environment of Leu63, Trp36, and Arg37. Hence,
adding a polar hydroxyl group at R2, such as in compounds 1f,
1f*, 2f, and 2f*, destabilizes binding within the hydrophobic
environment and results in weaker Eis inhibitory activity.
Furthermore, the para position of the acetophenone ring is
sterically restrained on all sides, being sandwiched between
Phe84 and Trp36 as well as abutting Trp13 and Met65,
explaining why the larger groups in para position, such as in
compounds 1i/i*, 1j/j*, 2i/i*, and 2j/j*, resulted in decreased
Eis inhibitory activities. This R2 binding pocket of Eis
accommodates structurally similar substitutions to that of our
previously published inhibitors with different core scaffolds
(Figure S160). Otherwise, these previously reported inhibitors
are bound in distinct orientations in the large AG binding
cavity. As shown in Figure 2 and in our SAR analysis, near the
two meta positions of the acetophenone ring, there are spacious
pockets allowing incorporation of larger substituents at the
meta position without compromising activity. Finally, the
phenyl ring containing R1 is located in a spacious binding
pocket lined by the terminus of the phosphopantetheinyl tail of
the CoA molecule, Asp26, the C-terminal carboxyl group,
Ser83, and Phe24, explaining why a phenyl group is preferred
over an ethyl group at the R1 position. In summary, the crystal
structure of the EisC204A-CoA-inhibitor 2k* complex allowed
us to explain our biological data and provides a basis for future
additional structure-based development of Eis inhibitors.
Figure 2. (A) Crystal structure of EisC204A-CoA-inhibitor 2k*
complex (PDB ID 5TVJ). CoA is colored yellow. Compound 2k* is
colored green. (B) Zoom-in view of the binding pocket of compound
2k*. Amino acid residues that are interactive with compound 2k* are
highlighted in red. The strong omit Fo − Fc electron density map
contoured at 3σ generated without the inhibitor is shown by the mesh,
demonstrating that the inhibitor molecule is unambiguously defined by
the X-ray diffraction data.
In conclusion, via a SAR study, we tailor-fitted Eis inhibitors
possessing the pyrrolo[1,5-a]pyrazine core to its Eis binding
pocket and identified multiple novel nanomolar potency
inhibitors. We validated our hypothesis that the aromaticity
of the pyrrolo[1,5-a]pyrazine core was important for activity
and that aromatic analogues were overall better inhibitors than
their nonaromatic counterparts. For the aromatic analogues,
our study indicated that the SAR strongly correlates with the
size of the halogen substituent(s). At the meta position of the
acetophenone, bigger halogens such as Cl and Br were
generally well tolerated. On the other hand, at the para
position, substitutions of a smaller F atom and a methyl group
produced analogues with substantially improved activities. The
SAR analysis also revealed that the substitution of a polar
functional group such as the hydroxyl group greatly perturbed
the hydrophobic environment, leading to decreased activity.
These SAR observations were explained by the crystal structure
analysis, which will greatly facilitate future medicinal chemistry
studies. Most significantly, by in vitro Mtb culture assays, we
F
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX

ACS Infectious Diseases
Article
confirmed that our Eis inhibitors were capable of penetrating
the Mtb cell wall and canceling the KAN resistance of Mtb
K204, which overexpresses Eis. As exemplified by a clinically
used combination of a β-lactamase inhibitor, clavulanic acid,
and penicillin, these Eis inhibitors may become similarly
significant as adjuvant molecules in a combination therapy with
KAN to prevent the emergence of and combat KAN resistance
in MDR- and XDR-TB.
ABBREVIATIONS
■
AAC, aminoglycoside N-acetyltransferase; AG, aminoglycoside;
AMK, amikacin; CoA, coenzyme A; Eis, enhanced intracellular
survival; HTS, high-throughput screening; KAN, kanamycin A;
MDR, multidrug-resistant; MIC, minimum inhibitory concen-
tration; Mtb, Mycobacterium tuberculosis; SAR, structure−
activity relationship; TB, tuberculosis; TOB, tobramycin;
XDR, extensively drug-resistant
ASSOCIATED CONTENT
■
REFERENCES
■
S
* Supporting Information
(1) World Health Organization. (2016) Global Tuberculosis Report;
Geneva, Switzerland.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acsinfec-
dis.6b00193.
(2) Gandhi, N. R., Nunn, P., Dheda, K., Schaaf, H. S., Zignol, M., van
Soolingen, D., Jensen, P., and Bayona, J. (2010) Multidrug-resistant
and extensively drug-resistant tuberculosis: a threat to global control of
tuberculosis. Lancet 375, 1830−1843.
(3) Yew, W. W., Lange, C., and Leung, C. C. (2011) Treatment of
tuberculosis: update 2010. Eur. Respir. J. 37, 441−462.
(4) Zaunbrecher, M. A., Sikes, R. D., Jr., Metchock, B., Shinnick, T.
M., and Posey, J. E. (2009) Overexpression of the chromosomally
encoded aminoglycoside acetyltransferase eis confers kanamycin
resistance in Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. U. S. A.
106, 20004−20009.
(5) Campbell, P. J., Morlock, G. P., Sikes, R. D., Dalton, T. L.,
Metchock, B., Starks, A. M., Hooks, D. P., Cowan, L. S., Plikaytis, B. B.,
and Posey, J. E. (2011) Molecular detection of mutations associated
with first- and second-line drug resistance compared with conventional
drug susceptibility testing of Mycobacterium tuberculosis. Antimicrob.
Agents Chemother. 55, 2032−2041.
All experimental procedures and characterization data of
all new compounds synthesized as well as their H and
1
13C NMR spectra and HPLC traces (Figures S1−S156).
Representative IC50 curves (Figure S157). Plots showing
the non-normalized cytotoxicity data (the normalized
data in Figure 1) are provided in Figure S158. Plots
showing that KAN alone is not toxic to three mammalian
cell lines in the concentration range tested (Figure
S159). Finally, superimposition of the crystal structure of
Eis in complex with inhibitors and TOB (Figure S160).
A table of X-ray diffraction data collection and
refinement statistics for the EisC204A-CoA-inhibitor
2k* complex (Table S1) is also provided (PDF)
(6) Chen, W., Biswas, T., Porter, V. R., Tsodikov, O. V., and
Garneau-Tsodikova, S. (2011) Unusual regioversatility of acetyltrans-
ferase Eis, a cause of drug resistance in XDR-TB. Proc. Natl. Acad. Sci.
U. S. A. 108, 9804−9808.
(7) Tsodikov, O. V., Green, K. D., and Garneau-Tsodikova, S. (2014)
A random sequential mechanism of aminoglycoside acetylation by
Mycobacterium tuberculosis Eis protein. PLoS One 9, e92370.
(8) Green, K. D., Pricer, R. E., Stewart, M. N., and Garneau-
Tsodikova, S. (2015) Comparative study of Eis-like enzymes from
pathogenic and non-pathogenic bacteria. ACS Infect. Dis. 1, 272−283.
(9) Chen, W., Green, K. D., Tsodikov, O. V., and Garneau-
Tsodikova, S. (2012) Aminoglycoside multiacetylating activity of the
enhanced intracellular survival protein from Mycobacterium smegmatis
and its inhibition. Biochemistry 51, 4959−4967.
(10) Green, K. D., Biswas, T., Chang, C., Wu, R., Chen, W., Janes, B.
K., Chalupska, D., Gornicki, P., Hanna, P. C., Tsodikov, O. V.,
Joachimiak, A., and Garneau-Tsodikova, S. (2015) Biochemical and
structural analysis of an Eis family aminoglycoside acetyltransferase
from Bacillus anthracis. Biochemistry 54, 3197−3206.
(11) Pricer, R. E., Houghton, J. L., Green, K. D., Mayhoub, A. S., and
Garneau-Tsodikova, S. (2012) Biochemical and structural analysis of
aminoglycoside acetyltransferase Eis from Anabaena variabilis. Mol.
BioSyst. 8, 3305−3313.
(12) Houghton, J. L., Green, K. D., Pricer, R. E., Mayhoub, A. S., and
Garneau-Tsodikova, S. (2013) Unexpected N-acetylation of capreo-
mycin by mycobacterial Eis enzymes. J. Antimicrob. Chemother. 68,
800−805.
(13) Kim, K. H., An, D. R., Song, J., Yoon, J. Y., Kim, H. S., Yoon, H.
J., Im, H. N., Kim, J., Kim, D. J., Lee, S. J., Kim, K. H., Lee, H. M., Kim,
H. J., Jo, E. K., Lee, J. Y., and Suh, S. W. (2012) Mycobacterium
tuberculosis Eis protein initiates suppression of host immune responses
by acetylation of DUSP16/MKP-7. Proc. Natl. Acad. Sci. U. S. A. 109,
7729−7734.
(14) Li, Y., Green, K. D., Johnson, B. R., and Garneau-Tsodikova, S.
(2015) Inhibition of aminoglycoside acetyltransferase resistance
enzymes by metal salts. Antimicrob. Agents Chemother. 59, 4148−4156.
(15) Willby, M. J., Green, K. D., Gajadeera, C. S., Hou, C., Tsodikov,
O. V., Posey, J. E., and Garneau-Tsodikova, S. (2016) Potent inhibitors
AUTHOR INFORMATION
Corresponding Authors
*(S.G.-T.) E-mail: sylviegtsodikova@uky.edu. Phone: (859)
218-1686. Fax: (859) 257-7585.
*(J.E.P.) E-mail: jposey@cdc.gov.
*(O.V.T.) E-mail: oleg.tsodikov@uky.edu.
ORCID
Sylvie Garneau-Tsodikova: 0000-0002-7961-5555
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
■
Funding
This study was funded by a National Institutes of Health
(NIH) Grant AI090048 (S.G.-T.), a grant from the Firland
Foundation (S.G.-T.), a grant from the Center for Chemical
Genomics (CCG) at the University of Michigan (S.G.-T), and
startup funds from the College of Pharmacy at the University of
Kentucky (S.G.-T. and O.V.T.). S.Y.L.H. is partially supported
by a University of Kentucky Presidential Fellowship.
Notes
Use of trade names is for identification only and does not
constitute endorsement by the U.S. Department of Health and
Human Services, the U.S. Public Health Service, or the CDC.
The findings and conclusions expressed by the authors do not
necessarily reflect the official opinion of the Centers for Disease
Control and Prevention or the authors’ affiliated institutions.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Martha Larsen, Steve Vander Roest, and Paul
Kirchhoff (from the CCG) for their help with HTS.
G
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX

ACS Infectious Diseases
Article
of acetyltransferase Eis overcome kanamycin resistance in Mycobacte-
rium tuberculosis. ACS Chem. Biol. 11, 1639−1646.
(16) Garzan, A., Willby, M. J., Green, K. D., Gajadeera, C. S., Hou,
C., Tsodikov, O. V., Posey, J. E., and Garneau-Tsodikova, S. (2016)
Sulfonamide-based inhibitors of aminoglycoside acetyltransferase Eis
abolish resistance to kanamycin in Mycobacterium tuberculosis. J. Med.
Chem. 59, 10619−10628.
(17) Garzan, A., Willby, M. J., Green, K. D., Tsodikov, O. V., Posey, J.
E., and Garneau-Tsodikova, S. (2016) Discovery and optimization of
two Eis inhibitor families as kanamycin adjuvants against drug-resistant
M. tuberculosis. ACS Med. Chem. Lett. 7, 1219.
(18) Green, K. D., Chen, W., and Garneau-Tsodikova, S. (2012)
Identification and characterization of inhibitors of the aminoglycoside
resistance acetyltransferase Eis from Mycobacterium tuberculosis.
ChemMedChem 7, 73−77.
(19) Kawashita, Y., and Hayashi, M. (2009) Synthesis of
heteroaromatic compounds by oxidative aromatization using an
activated carbon/molecular oxygen system. Molecules 14, 3073−3093.
(20) Houghton, J. L., Biswas, T., Chen, W., Tsodikov, O. V., and
Garneau-Tsodikova, S. (2013) Chemical and structural insights into
the regioversatility of the aminoglycoside acetyltransferase Eis.
ChemBioChem 14, 2127−2135.
(21) Quave, C. L., Estevez-Carmona, M., Compadre, C. M., Hobby,
G., Hendrickson, H., Beenken, K. E., and Smeltzer, M. S. (2012)
Ellagic acid derivatives from Rubus ulmifolius inhibit Staphylococcus
aureus biofilm formation and improve response to antibiotics. PLoS
One 7, e28737.
(22) Hall, B. S., Bot, C., and Wilkinson, S. R. (2011) Nifurtimox
activation by trypanosomal type I nitroreductases generates cytotoxic
nitrile metabolites. J. Biol. Chem. 286, 13088−13095.
(23) Xu, W., Zhu, X., Tan, T., Li, W., and Shan, A. (2014) Design of
embedded-hybrid antimicrobial peptides with enhanced cell selectivity
and anti-biofilm activity. PLoS One 9, e98935.
(24) Shrestha, S. K., Fosso, M. Y., Green, K. D., and Garneau-
Tsodikova, S. (2015) Amphiphilic tobramycin analogues as anti-
bacterial and antifungal agents. Antimicrob. Agents Chemother. 59,
4861−4869.
(25) Fosso, M. Y., Shrestha, S. K., Green, K. D., and Garneau-
Tsodikova, S. (2015) Synthesis and bioactivities of kanamycin B-
derived cationic amphiphiles. J. Med. Chem. 58, 9124−9132.
H
DOI: 10.1021/acsinfecdis.6b00193
ACS Infect. Dis. XXXX, XXX, XXX−XXX