New methods to assess 6-thiopurine toxicity and expanding its therapeutic application to pancreatic cancer via small molecule potentiators

Article abstract of DOI:10.1039/c9md00010k

6-Thiopurine (6TP) is a potent cytotoxic agent that is a clinically prescribed anti-metabolite employed in the treatment of numerous blood cancers since 1952. However, its reported severe toxicities limit its general usage in the clinic. We previously have undertaken investigations into identifying the mode of toxicity for 6TP, and have found that the oxidative metabolites of 6TP, specifically 6-thiouric acid (6TU), is responsible for the in vitro inhibition of UDP-glucose dehydrogenase (UDPGDH) in a UV-vis method. In this method, inhibition was quantified through the quantification of NADH production, however, purines absorb at the same wavelength and thereby can interfere with the NADH detection. Herein, we report a HPLC method that allows for direct quantification of UDP-glucuronic acid, product from UDPGDH, for the assessment of inhibition towards UDPGDH with no interference from purines. In this method it was revealed that 6TP possesses a greater inhibitory properties than previously observed; 111 vs. 288 μM. Building upon the data collected from a previously performed rat hepatocyte study, which correlated our in vitro to in vivo inhibition theories about UDPGDH, we have developed a bio-mimic in vitro assay to aid in the inhibitory assessment of 6TP and analogs. In our efforts to expand the use of 6TP, and analogs constructed, our laboratory has undertaken a screening campaign to identify small molecule potentiators that work in synergy with 6TP in other types of cancers. Three chalcone-based compounds have been discovered through our total synthesis campaign of uvaretin, and it has been found that 11c has strong synergism with 6TP in the pancreatic cancer cell line MIA PaCa-2. Through the work presented herein, we reveal new methods to assess toxicity of 6TP and future analogs and new small molecules that work in synergy to expand the therapeutic applications of this neglected cytotoxic agent.

Full text of DOI:10.1039/c9md00010k

MedChemComm
View Article Online
View Journal
RESEARCH ARTICLE
New methods to assess 6-thiopurine toxicity and
expanding its therapeutic application to
pancreatic cancer via small molecule
potentiators†
Cite this: DOI: 10.1039/c9md00010k
Chamitha Weeramange,‡ Ashabha Lansakara,‡ Johnathan Dallman,‡ Thi Nguyen,
Wasundara Hulangamuwa and Ryan J. Rafferty
*
6
-Thiopurine (6TP) is a potent cytotoxic agent that is a clinically prescribed anti-metabolite employed in
the treatment of numerous blood cancers since 1952. However, its reported severe toxicities limit its gen-
eral usage in the clinic. We previously have undertaken investigations into identifying the mode of toxicity
for 6TP, and have found that the oxidative metabolites of 6TP, specifically 6-thiouric acid (6TU), is respon-
sible for the in vitro inhibition of UDP-glucose dehydrogenase (UDPGDH) in a UV-vis method. In this
method, inhibition was quantified through the quantification of NADH production, however, purines absorb
at the same wavelength and thereby can interfere with the NADH detection. Herein, we report a HPLC
method that allows for direct quantification of UDP-glucuronic acid, product from UDPGDH, for the as-
sessment of inhibition towards UDPGDH with no interference from purines. In this method it was revealed
that 6TP possesses a greater inhibitory properties than previously observed; 111 vs. 288 μM. Building upon
the data collected from a previously performed rat hepatocyte study, which correlated our in vitro to
in vivo inhibition theories about UDPGDH, we have developed a bio-mimic in vitro assay to aid in the inhib-
itory assessment of 6TP and analogs. In our efforts to expand the use of 6TP, and analogs constructed, our
laboratory has undertaken a screening campaign to identify small molecule potentiators that work in syn-
ergy with 6TP in other types of cancers. Three chalcone-based compounds have been discovered through
our total synthesis campaign of uvaretin, and it has been found that 11c has strong synergism with 6TP in
the pancreatic cancer cell line MIA PaCa-2. Through the work presented herein, we reveal new methods
to assess toxicity of 6TP and future analogs and new small molecules that work in synergy to expand the
therapeutic applications of this neglected cytotoxic agent.
Received 8th January 2019,
Accepted 14th March 2019
DOI: 10.1039/c9md00010k
rsc.li/medchemcomm
as a dGMP mimic that results in the cell undergoing apopto-
1
. Introduction
1
1–14
sis upon checkpoint activation.
Numerous enzymatic
6
-Thiopurine (6TP, 1, Fig. 1) has been a continually prescribed
transformations are required to convert 6TP into 2, some of
anticancer agent since 1952, two years following its discovery
by Gertrude Elion and George Hitchings at Burroughs
Wellcome Laboratories in 1950. This agent serves as a treat-
the thiol methylated intermediate species serves as check-
1
5–18
point activator.
In addition to 6TPs therapeutic pathway,
1
an excretion pathway is occurring simultaneously. Xanthine
oxidase (XO) oxidizes both the C2 and C8 positions of 6TP
giving rise to 6-thiouric acid (6TU, 3), an excretable species,
however 3 is retained by the body beyond 24 hours post-6TP
ment option for numerous diseases, such as, but not limited
2
,3
to: acute lymphocytic leukemia (ALL), non-Hodgkin's leuke-
4
,5
6–8
mia,
Crohn's disease,
and inflammatory bowel dis-
9
,10
19
ease.
The therapeutic mode of action for 6TP comes from
treatment. Many of the intermediates within the therapeutic
its biochemical transformation into deoxythioguanosine tri-
phosphate (2), which is subsequently incorporated into DNA
pathway are also susceptible to enzymatic transformations/ox-
idations leading to the formation of additional 6TU. As such,
the prescribed dosage of 6TP is increased to compensate for
the loss from such transformations.
While 6TP has been used in the treatment of leukemia for
over 60 years, its general application has been on the decline
due to toxicities associated with its use. The most predomi-
nate toxic side effects are jaundice and hepatotoxicity,
Department of Chemistry, Kansas State University, 1212 Mid-Campus Drive
North, Manhattan, KS 66506, USA
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9md00010k
These authors contributed equally to the work.
‡
This journal is © The Royal Society of Chemistry 2019
Med. Chem. Commun.

View Article Online
Research Article
MedChemComm
inhibition towards UDPGDH with a K of 288 μM, however,
i
its main excretion metabolite 6TU has a 41-fold increase in
inhibition with a K of 7 μM. The method that was developed
i
to assess inhibition of UDPGDH by these excretion metabo-
lites is a UV/vis based procedure in which the rate of enzy-
matic turn-over was measured by NADH production.
Based upon our mode of toxicity onset investigations, in
which the C8 position was shown to play a vital role in
UDPGDH inhibition, C8-halogen bearing analogs of 6TP were
2
4
synthesized. Inhibition towards UDPGDH was greatly re-
duced with these C8-analogs, and furthermore, they were
shown to retain cytotoxicity towards the acute lymphocytic
leukemia cell line Reh. In our continuing efforts to redesign
6TP into a form with reduced toxicity, while retaining its
therapeutic mode of action, further enhanced methods to as-
sess 6TP and any analogs constructed is required. In these ef-
forts, a more accurate correlation of the inhibitory properties
of 6TP, excretion metabolites, and analogs constructed is nec-
essary. However, our previous method for inhibition assess-
ment has a critical flaw, purines interfere with the NADH sig-
nal quantification. As such, this requires the development of
a new method of inhibition assessment that does not depend
upon NADH quantification. Herein is disclosed our HPLC
method that was developed to assess the inhibitory proper-
ties of these species via direct quantification of UDPGA, in
which purines possess no interfering signals. Additionally,
the construction of a bio-mimic system, based upon previ-
ously findings from our rat hepatocyte study, has been ac-
complished. From these new models that assess toxicity of
6TP, new avenues of investigation for newly synthesized ana-
logs can be performed. Lastly, research has also been under-
taken, and disclosed in this work, in the discovery of small
molecules that work in synergy with 6TP to expands is thera-
peutic usefulness to pancreatic cancer.
Fig.
1 Illustration of the therapeutic and excretion, leading to
reported off-target toxicities, pathways of 6-thiopurine. Atom number
has been included on 6TP and 6TU for reference.
corresponding to the reported increase in bilirubin (6) levels
within patients.
the excretion of bilirubin, post red-blood cell senescence.
Excretion of this non-polar species is achieved through the
conjugation of two UDP-glucuronic acids (UDPGA, 5) by UDP-
glucuronosyltransferase (UGT-1A), forming an excretable po-
lar species. UDPGA is formed from the oxidation of UDP-
glucose (UDPG, 4) by UDP-glucose dehydrogenase
1
3,16
The bilirubin pathway is responsible for
2
0
2
1,22
(UDPGDH).
While 6TP is a currently prescribed antileuke-
mic agent, its toxic side effects are so potent and prevalent
that the administration of 6TP is given in an on/off strategy,
allowing time for the toxic species to be cleared by the body.
Unfortunately, this greatly reduces its therapeutic efficacy, re-
stricts the quality of life of the patients taking the drug, and
has ultimately limited its use as an anticancer therapy.
2
. Results and discussion
2.1 Development of an HPLC method for direct UDPGA
quantification for UDPGDH inhibition assessment
Our laboratory is interested in investigating drugs that ei-
ther have been removed from clinical use due to fatal toxic-
ities or ones currently used in limited applications due to as-
sociated toxicities, such as 6TP. To accomplish this, we
investigate to identify the target that results in the reported
toxicity from the agent and then explore for possible sub-
strate to target correlations. Analogs are constructed based
upon these observations in the aims of reducing the reported
toxicity, so that the agent can be returned to clinical use, and
if possible, in new expanded therapeutic capacities. For 6TP,
we have previously disclosed our investigations into the role
that the excretion metabolites, oxidized at either or both the
C2/C8 positions, have upon UDPGDH. From this, we have
hypothesized that the hepatotoxicity, and its related side ef-
fects, caused by 6TP administration arises from inhibition of
UDPGDH that prevent proper bilirubin conjugation and ex-
cretion. Our laboratory has shown that 6TP has no significant
Our laboratory has previously reported the inhibitory effects
of the excretion metabolites of 6TP, in which 6TU was found
to be the most inhibitory, towards UDPGDH through a UV-vis
based method through NADH quantification. While a mini-
mal concentration of the purines were employed in this
method, there was interfering signals between the purine
and NADH via absorption quantification. The development of
a new method to assess inhibition, in which the interfering
signal of the purines cannot bias absorption, was undertaken
to further corroborate our initial findings of the effects of the
hydroxylated C2/8 positions towards UDPGDH. To assess the
inhibitory effects of the excretion metabolites upon
UDPGDH, a HPLC method was developed that allows for
2
3
+
baseline separation of the substrates (UDPG and NAD ) and
products (UDPGA and NADH). Baseline separation and col-
umn regeneration, allowing for reproducible back-to-back
runs, is accomplished over a 23 minute period with a dual-
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2019

View Article Online
MedChemComm
Research Article
solvent gradient method (HPLC chromatogram and mobile
phase gradient profile, and the UDPGA standard curve
constructed is presented in the ESI†). This HPLC method re-
quires no derivatization of UDPGA, nor multiple mobile
phases as previously employed in other HPLC
mains unaffected and that inhibition of UDPGDH by 6TP is
not significantly impairing for multiple reasons, such as: 1)
the oxidation of 6TP to 6TU by xanthine oxidase is rapid, 2)
the inhibition by 6TP is 21-folds lower than 6TU, and 3) the
biotransformation into its therapeutic nucleotide mimic spe-
cies occurs concurrently to oxidation by xanthine oxidase
thereby leaving a lower available pool of 6TP to elicit any in-
hibition towards UDPGDH.
2
5–27
methodologies.
With an enhanced HPLC method developed for the quan-
tification of UDPGDH activity by monitoring UDPGA forma-
tion established, efforts were directed towards assessing inhi-
bition of 6TP and 6TU. Following the protocol previously
2
3
2.2 Generation of an in vitro bio-mimic UDPGDH/UGT-1A
evaluation system
employed by our laboratory, kinetic studies were performed
against UDPGDH with four varying concentrations of 6TP
and 6TU employed, and the determination of K values was
accomplished by plotting the slope of from each independent
analysis set verses the concentration of the purine; calcula-
Recently, we reported a hepatocyte study that allowed for
in vivo corroboration of our in vitro data regarding 6TP and
6TU inhibition of UDPGDH and UGT-1A. Highlights of the
overall findings from this hepatocyte investigation are
i
tions of the K was performed by taking the negative-inverse
i
2
4
of the x-intercept. The K for 6TU was found to be 5 μM via di-
presented in Fig. 3. The treatment of 6TP resulted in a mar-
ginally lower level of BDG (bilirubin diglucuronide) forma-
tion, which we can now correlate to the recently observed
greater inhibitory property of 6TP towards UDPGDH. This
also results in a lower liable pool of UDPGA and the observed
BMG (bilirubin monoglucuronide) depression. 6TU was
shown to have a larger impact upon the formation levels of
BDG, through the proposed inhibition of UDPGDH. Valida-
tion of this UDPGDH inhibition claim was revealed when ex-
ternal UDPGA was added to a system in the presence of 6TU,
in which levels of BMG formation returned to those of the
control group.
i
rect analysis, showing only a marginal difference of the 7 μM
observed in our in-direct method of assessment (Fig. 2B).
However, 6TP was shown to have considerably more inhibi-
tory properties towards UDPGDH when assessed in this direct
method. The K for 6TP via direct assessment was found to
i
be 111 μM, where previously we reported 288 μM through
our in-direct NADH detection method (Fig. 2A).
Through the development of a non-interfering method to
assess the inhibition of UDPGDH by 6TP and 6TU, it was re-
vealed that 6TP has a two-fold greater inhibitory effect than
previously reported. No significant change in the K for 6TU
i
was observed in this method. We have previously reported
our hypothesis that 6TU, as well as 8-OH-6TP are the main
excretion metabolites of 6TP responsible for UDPGDH inhibi-
tion, which correlates to the reported toxic side effects from
In alignment with the overall objective of decreasing the
toxic response from 6TP, while retaining its known therapeu-
tic mode of action, efforts towards analog construction are
underway. Evaluation of inhibition towards UDPGDH and
UGT-1A by all synthesized analogs will be required. While we
have previously successfully employed a hepatocyte model to
confirm our working hypothesis of inhibition, the need for a
more readily accessible toxicity assessment model is
6
TP administration. While the current investigation has re-
vealed a K of 111 μM for 6TP towards UDPGDH, levels that
i
could be potentially impact enzymatic activity, we believe that
our hypothesis of inhibition by C8-hydroxylated species re-
Fig. 2 Assessment of inhibition of UDP-glucose dehydrogenase by 6TP and 6TU via direct quantification of UDP-glucuronic acid by HPLC separa-
tion and quantification. A) Concentration of 6TP, varying UDPG, screened were 0, 50, 75, 100 μM with the calculated slopes of each line equaling
0
.0353, 0.0549, 0.0612, 0.0680, respectively. Plotting slopes versus concentration gave a regression line of y = 0.000328x + 0.0364 with a 0.9881
R-square. B) Concentration of 6TU, varying UDPG, screened were 0, 5, 7.5, 10 μM with the calculated slopes of each line equaling 0.0147, 0.0254,
.0328, 0.0394, respectively. Plotting slopes versus concentration gave a regression line of y = 0.002474x + 0.01416 with a 0.9946 R-square. Each
0
data point for Lineweaver Burk plots are an average of three independent runs with standard error bars shown.
This journal is © The Royal Society of Chemistry 2019
Med. Chem. Commun.

View Article Online
Research Article
MedChemComm
of UDPGA will complete with regorafenib resulting in biliru-
bin conjugation. If the levels are UDPGA are too low, this will
result in a near complete inhibition of BDG and BMG forma-
tion post regorafenib treatment. Employing a 1 : 1 ratio of
UDPGDH to UGT-1A resulted in a saturating amount of
UDPGA formation, as did a 1 : 400, 1 : 800, and 1 : 1200 ratio
(
results are presented in the ESI†). Presented in Fig. 4 is the
control group set from the hepatocyte study (Fig. 3).
Employing a 1 : 2200 ratio of UDPGDH to UGT-1A afford a
similar distribution of UCB, BMG, and BDG to the control
set. When 10 μM of regorafenib was employed, the suppres-
sion of both BMG and BDG formation observed matched sat-
isfactorily to the same investigational parameters within the
hepatocyte study. This supports that the in vitro model
mimics the hepatocyte model towards UGT-1A. Employing 5
μM of 6TU showed a decrease in the production of BMG and
BDG; however not to the same extent as regorafenib. This re-
sult was expected, since 6TU is inhibiting UDPGDH and
thereby limiting the amount of UDPGA formed, which corre-
sponds to the decreased production of both BMG and BDG.
To fully confirm that UDPGDH is inhibited in this in vitro
model, subjecting the system to 5 μM of 6TU in the presence
of 15 μM of UDPGA showed recovery of the system back to
control levels. Employing a 1 : 2200 ratio of UDPGDH to
microsomes has shown to provide data consistent with our
hepatocyte model, and corrects correlates 6TU inhibition of
UDPGDH and not UGT-1A.
Fig. 3 Overview of the performed hepatocyte studies to correlate 6TP
and 6TU inhibition of UDPGDH. Each group is an average of three
independent runs with standard error bars shown. UCB: unconjugated
bilirubin (free), BMG: bilirubin diglucuronide (one UDPGA attached),
BDG: bilirubin diglucuronide.
essential. Cellular levels of UDPGDH vary, and as such the
2
8,29
development of a physiological model can be difficult.
As
such, the in vitro model was chosen to mimic the result from
our hepatocyte in vivo investigations. Our bio-mimic in vitro
model system was based upon the buffer composition
employed in our UGT-1A assay, a phosphate buffer system,
containing bilirubin and alamethicin (inhibitor of glucuroni-
dase, present within the microsomes employed in this
2
3
+
study). In addition, both NAD and UDPG were included to
allow for the formation of UDPGA, which can be used to as-
sess BDG formation.
Efforts in the development of a bio-mimic in vitro system,
to mimic the observed trends of the hepatocyte study (Fig. 3),
commenced with investigations into determining the ratio of
UDPGDH (6.8 units per mL) and microsomes (containing 2.2
units per mL of UGT-1A) was then undertaken. We decided
to use 4.4 units per mL of UGT-1A in the presence of 2 mM
2.3 Expansion of 6-thiopurine therapeutic applications
With the development of new methods to assess the inhibi-
tory properties of any 6TP analogs constructed, three of
+
24
bilirubin along with 10 mM NAD and 5 mM UDPG in a final
which have already been reported by our laboratory, we
volume of 500 μL as this mimicked our reported UGT-1A as-
turned our attention to expanding the therapeutic application
of 6TP, and analogs. To accomplish this, we propose to inves-
tigate the effects of small molecules upon 6TP when used in
combination, also referred to as small molecule potentiation.
The action in which small molecules work in synergy with
chemotherapeutic agents is variable. In some cases they help
attenuate activity by modulating a key biochemical pathway
in which the agents elicits is therapeutic action, and in
others they aid in the transport of the chemotherapeutic
2
3
say. The addition of a UDPGDH solution marked the start
of the reaction. Employing 10 μM regorafenib, a UGT-1A in-
hibitor, 45 minutes into the enzymatic incubation will allow
assessment to determine if the ratio of the two enzymes
matches the hepatocyte model. The same concentration of
regorafenib resulted in depression of BMG and BDG forma-
tion, not complete inhibition. As such, if the ratio of
UDPGDH to microsomes is too high, the enhanced formation
Fig. 4 Development of an in vitro model to assess UDPGDH and UGT-1A inhibition. Each group is an average of three independent runs with stan-
dard error bars shown.
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2019

View Article Online
MedChemComm
Research Article
agent, to mention a few possible modes of action. Discover-
ing small molecules with the ability to potentiate the activity
of chemotherapeutic agents is a difficult endeavor, most
times their discovery is quite serendipitous. Our efforts to
discover such small molecules lies within the intermediates
accessed from total synthesis routes. We proposed that inter-
mediates, whether fruitful or not towards the targeted natural
product, can be used in general screens to determine if they
have sole cytotoxic properties or possibly possess potentiating
properties towards current chemotherapeutic agents, such as
Evaluation of 11a–c was then performed in a panel of can-
cerous cell lines: HeLa (cervical), A549 (lung), MIA PaCa-2
(pancreatic), U937 (lymphoma), and Reh (acute lymphocytic
leukemia, ALL). Of these three chalcones, it was found that
11c possessed the most potent cytotoxicity towards the cell
lines investigated as a sole agent (Fig. 5B). While 11a–b were
observed to have lower potency in comparison to 11c, poten-
tiating studies was performed with 6-thiopurine with all three
chalcones. The assessment of the combinational properties
of 11a–c with 6TP in various cancerous cell lines was
performed in a matrix format (3 × 5 matrix) against five can-
cer types (HeLa-cervical, A549-lung, MIA PaCa-pancreatic,
U937-lymphoma, and Reh-acute lymphocytic leukemia). The
matrix was designed so that two concentrations of 11a–c
would induce cell death between 10 and 20% within a 72
hour treatment, and the concentrations of 6TP would result
in up to 50% cell death over the same time period. To quan-
tity the effects of the two drugs in combination, as well as
the combination index (CI) the Chou–Talalay method was
6
TP. Our efforts towards the total synthesis of uvaretin,
shown in the insert of Fig. 5A, has allowed for the access of
numerous small molecules possessing interesting and diverse
biological activity. Illustrated in Fig. 5A is our basic synthetic
route employed to gain access to the core structure of
uvaretin. Starting from phloroglucinol (8), acetylation was
performed under Lewis-acid conditions and acetic anhydride
to afford the corresponding aromatic ketone (9). Details on
the elaboration of 9 onto its three corresponding phenol
protected counterparts 10a–c is presented within the ESI.†
While numerous groups were attempted in route that
accessed uvaretin, the methoxymethyl (MOM) group is
highlighted in this work for the unique properties associated
with the final chalcone product. Subjecting 10a–c to aldol
condensation condition with benzaldehyde gave access to
chalcones 11a–c.
3
0
employed. CI values were calculated for each combination,
this resulted in 10 CI values per cell line investigated. Quanti-
fication of the combination experiments is shown in Fig. 5C,
with the median CI values shown for each of the 6TP and
11a–c combinations. CI values of <1 indicate a strong syn-
ergy, with lower values possessing stronger synergy. Synergy
was observed between 11c and 6TP in the pancreatic cell line
Fig. 5 Synthesis of chalcone small molecules and investigations into possible synergy with 6-thiopurine for either increased cytotoxicity or
expansion of therapeutic application scope. A) General synthesis route employed to gain access to the three chalcones 11a–c. Synthetic
procedures: a. MOMCl (2.2 eq.), DIPEA, CH Cl , b. MOMCl (3.3 eq.), DIPEA, CH Cl , and c. Me SO , K CO , acetone. B) General cytotoxicity of
2 2 2 2 2 4 2 3
chalcones 11a–c in HeLa (cervical), A549 (lung), MIA PaCa-2 (pancreatic), U937 (lymphoma), and Reh (acute lymphocyte leukemia) cancerous cell
lines. C) Results of the average cell death induced by 11c & 6-thiopurine combinations evaluated in the 3 × 5 matrix and corresponding quantifica-
tion of synergy with CI values (n = three biologic replicates).
This journal is © The Royal Society of Chemistry 2019
Med. Chem. Commun.

View Article Online
Research Article
MedChemComm
MIA PaCa-2, giving low CI values. Given that 6TP has no
reported toxicity within pancreatic cancer cell lines, this find-
ing illustrates how the expansion of 6TP application can be
accomplished via small molecule potentiation. Unexpectedly,
no synergism was observed between 11a–b and 6TP in MIA
PaCa-2 as observed with high CI values. Evaluation of poten-
tial synergism between 11c and 6TP in Reh, the cancer type
that 6TP is prescribed for, showed mild synergy. No synergy
was observed for 11a–b in Reh as well as for 11a–c in U937,
A549, and HeLa cancerous cell lines.
Of the three chalcones synthesized, it was found that 11c
possesses the best synergy with 6TP within the pancreatic
cancer cell line MIA PaCa-2. Investigations are currently un-
derway to elucidated the mode in which 11c potentiates the
activity of 6TP. It is anticipated that these studies will give in-
sight into how 11c works with 6TP in MIA PaCa-2, and also
why 11c does not possess the same synergism with 6TP
within the acute lymphocytic leukemia cell line Reh, the pri-
mary cancer type that 6TP is prescribed to treat. The discov-
ery of small molecule potentiators is difficult, as the activity
can differ between cancer types, as observed here with MIA
PaCa-2 and Reh cell lines. We have shown that through the
use of intermediates accessed within total synthesis cam-
paigns that the discovery of such small molecules is possible.
Furthermore, it has allowed for the discovery of 11c that ex-
pands the use of 6TP beyond acute lymphocytic leukemia to
pancreatic cancers.
ing of chalcones 11a–c with 6TP in the pancreatic (MIA PaCa-
2) and acute lymphocytic leukemia (Reh) cancerous cell lines.
Chalcone 11c was found to possess strong synergism with
6TP in the MIA PaCa-2, but not ReH. The enhanced cytotoxic-
ity observed with the combination treatment of 6TP and 11c
in MIA PaCa-2 relative to 6TP alone in the same cell line re-
veals an expansion of the application scope of 6TP to pancre-
atic cancer. Efforts are currently underway to explore the
mode of action between 11c and 6TP and how it translated to
this new synergist therapeutic effect. In summary, the devel-
opment of new methods to assess inhibition about UDPGDH,
as well as UGT-1A, has been performed and shown to be com-
parable to in vivo studies previously conducted. Furthermore,
investigations into the discovery of small molecules to either
potentiate the activity of 6TP or expand its therapeutic useful-
ness was performed and 11c was found to expand 6TP appli-
cation into the pancreatic cancerous cell line MIA PaCa-2.
4. Experimental section
4.1 Reagents and equipment
All standard chemicals used in this study were the highest
grades available and were purchased through Sigma-Aldrich
(Saint Louis, MO, USA), VWR (Radnor, PA, USA), or Fisher
Scientific (Denver, CO, USA). Specialized reagents were pur-
chased through specific vendors. Glycylglycine (gly-gly:
+
G1127), ß-nicotinamide adenine dinucleotide (NAD : N1636),
uridine 5′-diphosphoglucose (UDPG disodium salt: U4625),
uridine 5′-diphosophoglucuronic acid (UDPGA: U6751), uri-
dine 5′-diphosphoglucose dehydrogenases (UDPGDH: U6885),
6-thiopurine monohydrate (6TP), bilirubin (including three
mixed isomers, B4126), 4,5-diamino-6-hydroxypyrimidine
hemisulfate salt (D19303), alamethicin (A5361), and pooled
rat liver microsomes (M9066) were purchased from Sigma-
Aldrich. 6-Mercaptopurine-2-ol (6-TX, QA-6668) was purchased
from Combi-Blocks, and 6-thiouric acid (6-TU, SC-213040)
from Santa Cruz Biotechnology. HPLC-grade water was
obtained by passing distilled water through a reverse osmosis
system followed by treatment with a Thermo Scientific
Barnstead Smart2Pure 3UV purification system (Fisher, 10-
451-045), herein referred to as nanopure water.
3
. Conclusion
6
-Thiopurine has been an approved chemotherapeutic drug
since 1952, and is currently primarily used in the treatment
of acute lymphocytic leukemia, in addition to other diseases.
Unfortunately, its clinical application is limited due to
reported toxicities associated with its application. Our labora-
tory has undertaken investigations into elucidating the mode
of toxicity of 6TP in the aims of constructing new analogs
with decreased, if not eliminated, toxicities. In addition,
through screening campaigns of small molecules synthesized
through our total synthesis campaigns, we seek to discover
small molecules that work in synergy with 6TP to either en-
hance its activity or to expand the scope of therapeutic appli-
cation. Through the new HPLC method developed for the as-
sessment of UDPGDH inhibition, 6TP was shown to possess
a two-fold greater inhibitory profile in compared to our previ-
ous method. Building upon this new method to assess the in-
hibition of UDPGDH, the found target of 6TP and its oxida-
tive metabolites, and based upon a previously employed
hepatocyte study, the construction of an in vitro bio-mimic
inhibition model was undertaken. In this method, we are
able to assess the inhibition of 6TP and 6TU towards
UDPGDH and/or UGT-1A. Inhibitory assessment of 6TP ana-
logs synthesized can now be done in a rapid and efficient
process through this bio-mimic model system developed. In-
vestigations into expanding the scope of 6TP application was
successfully accomplished through the combinational screen-
The cancerous cell line used in this investigation were
purchased directly from the American Type Culture Collec-
tion: HeLa, U-937, A549, MIA PaCa-2, and Reh. Cells were
grown in media supplemented with fetal bovine serum (FBS)
−
1
−1
and antibiotics (100 μg mL penicillin and 100 U mL strep-
tomycin). Cells were incubated at 37 °C in a 5% CO , 95%
2
humidity atmosphere. Cellular viability was determined by
quantification via Alamar blue.
All standard consumable supplies used in this study were
purchased from VWR or Fisher Scientific. Specific equipment
utilized in this work are: 1) Hewlett-Packard 8452 diode array
UV/vis spectrophotometer (Palo Alto, CA, USA) equipped with
a Lauda Brinkmann Ecoline RE 106 E100 circulating water
bath purchased from VWR, 2) HPLC system consisting of an
CBM-20A/20Alite system controller, SIL-20AHT auto sampler,
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2019

View Article Online
MedChemComm
Research Article
SPD-20A, SPD-20AV UV-vis detector, LC-20AT Solvent delivery
module, CTO-20A column oven, DGU-20A3R degassing unit
and LC-20AD/20AT Gradient Valve Kit purchased from
Shimadzu Scientific Instruments (Kyoto, Japan), and 3) all in-
cubated reactions were performed with a Labcare America
PRECISION water bath model 25 purchased from Fisher Sci-
entific. All HPLC separations were performed on a Discovery
C18 analytical column, 4.6 mm × 100 mm, 5 μm particle size
solutions) and added to a 96-well plate over a range of con-
centrations (31.6 nM to 200 μM) with media, and 40 μL was
added to the 384-well plate in triplicate for each concentra-
tion of compound. Cells were then added to the plate (2000
cells per well for adherent and 3500 cells per well for suspen-
sion cells) in 10 μL of media. After 69 h of continuous expo-
sure, 5 μL of Alamar Blue was added to each well, and the
cells were allowed to incubate for an additional 3 h. The
plates were then read for fluorescence intensity with an exci-
tation of 560 nm and emission of 590 nm on a BioTek Syn-
ergy H1 plate reader. Doxorubicin and etoposide were both
used as positive death controls, and wells with no com-
pounds added as negative death controls. IC50 values were
determined from three or more independent experiments
using GraphPad Prism 7.0. (LaJolla, CA, USA).
Combination studies. All combinational cell death
experiments were performed in 96-well plates with a total vol-
ume of 100 μL. To each well was added 49 μL of cell media,
either 2.5 or 5 μM of 6TP (from a DMSO stock solution), and
compounds 11a–c, independently. To each well on the plate
was added two different concentration of 11a–c and five con-
centrations of 6TP (both prepared from 10 mM stock solu-
tions in DMSO), cell media (adjusted to reach 50 μL volume)
and 0.5 μL of DMSO to achieve a 1% DMSO concentration.
To each well was then added 50 μL of a suspension of cells
to obtain a final cell density of 4000 cells per well (adherent
cells) and 7000 cells per well (suspension cells). Doxorubicin
and etoposide were both used as positive death controls, and
wells with no compounds added as negative death controls.
(504955-30) along with the respective guard column (59576)
purchased from Sigma-Aldrich. Data was processed and all
figures and tables constructed via the program Prism 7.02 for
Mac, GraphPad Software (La Jolla, CA, USA). All chemical
structures were prepared with ChemDraw Professional 16.0
by PerkinElmer (Waltham, MA). All statistical calculations
within this body of work was performed by the treatment of
two-way factorials (positive and negative controls, design
structure of RCBD, and T-tests) with Statistical Analysis Sys-
tem (SAS) software for Windows (Cary, NC, USA).
4
.2 Biology
.2.1 HPLC UDPGDH method for UDPGA detection and
4
+
quantification. The substrates (UDP-glucose and NAD ) and
products (UDPGA and NADH) of the enzymatic reaction car-
ried out by UDPGDH was separated on a Discovery C18 ana-
lytical column, 4.5 mm × 100 mm, 5 μM particle size with
guard column. A dual mobile phase was employed; the aque-
ous phase consisted of an 8 mM imidazole & 2.5 mM tetra-
butylammonium hydrogen sulfate (TBAHS) buffer at a pH of
6
.5 in nanopure water and methanol as the organic phase. A
2
Plates were incubated at 37 °C with 5% CO for 72 h, at
gradient elution profile was employed for full separation at a
which time they were assessed by Alamar blue. IC50 values
were determined from three or more independent experi-
ments using GraphPad Prism 7.0.
−
1
flow rate of 0.5 mL min , the method starts with a 90 : 10
aqueous–organic percentage) mobile phase composition that
(
is held for three-minutes, changed to 40 : 60 (aqueous–or-
ganic percentage) over one-minute, held for six-minutes, and
then returned to initial composition over 2.7 minutes and
held for 10 minutes to allow for column regeneration. The
detection wavelength was 262 nm with a sample injection vol-
ume of 5 μL. Various concentrations of UDGPA were ana-
lyzed, and peak areas obtained were plotted relative to said
concentrations to generate a working standard curve that was
used to assess inhibition of UDPGDH by 6TP and 6TU as a
function of UDPGA production.
4
.3 Chemistry
All reagents were commercially available and used without
purification unless otherwise stated. NMR spectra were
recorded with a Varian 400 MHz instrument. The chemical
shifts are given in parts per million (ppm) relative to residual
CHCl at δ 7.26 ppm or DMSO δ 2.50 ppm for proton spectra
3
and relative to CDCl at δ 77.23 ppm or DMSO δ 39.52 ppm
3
for carbon spectra, unless otherwise noted. Flash column
4
.2.2 UDP-glucose dehydrogenase incubation. All protocols
chromatography was performed with silica gel grade 60 (230–
used for the formation of UDPGA via the enzymatic conver-
2 2
400 mesh). Dichloromethane (CH Cl ), and acetone were
sion by UDPGDH, and for purine inhibition, was performed
degassed with argon and passed through a solvent purifica-
tion system containing alumina or molecular sieves. All com-
mercially available reagents were used as received. All proce-
dures including anhydrous solvents were performed with
rigorously dried glassware under inert atmosphere.
4.3.1 1-(2,4,6-Trihydroxyphenyl)ethan-1-one (9). To a flame
dried round bottom (RBF) under an argon atmosphere was
added boron trifluoride etherate (8.81 mL, 71.37 mmol, 3
eq.), followed by phloroglucinol (8) (3.0 g, 23.79 mmol, 1 eq.),
and acetic anhydride (2.25 mL, 23.79 mmol, 1 eq.). The solu-
tion was allowed to stir at room temperature for 16 h, at
2
3
by previously published procedures by our laboratory.
.2.3 UDP-glucuronosyltransferase activity assay. All proto-
4
cols used for the quantification of UCB, BMG, BDG via the
enzymatic conversion by UGT-1A, and for purine inhibition,
was performed by previously published procedures by our
2
3,24
laboratory.
.2.4 General cell death assessment via Alamar blue
4
quantification
General procedure for single agent IC50 determination.
Compounds were solubilized in DMSO (10 μM stock
This journal is © The Royal Society of Chemistry 2019
Med. Chem. Commun.

View Article Online
Research Article
MedChemComm
which time the reaction was quenched with the addition of a
155.59, 116.07, 96.14, 95.01, 94.70, 94.08, 56.55, 56.41, 56.05,
32.72. HRMS calcd. for C H O m/z [M + H] 271.1182;
13 19 6
found 271.1192.
+
+
1
0% sodium acetate solution (10× volume) and allowed to stir
for an additional 24 h. The solution was filtered, the precipi-
tate washed with washed to obtain 2,4,6-
trihydroxyacetophenone (9) as a yellow solid in 67% yield
4.3.5 (E)-1-(2-Hydroxy-4,6-bisIJmethoxymethoxy)phenyl)-3-
phenylprop-2-en-1-one (11a). To a RBF containing ethanol (10
mL) was added benzaldehyde (0.14 mL, 1.4 mmol, 1 eq.) and
10a (360 mg, 1.4 mmol, 1 eq.) and cooled to 0 °C. Once at
temperature, KOH (1.34 g, 24 mmol, 17 eq.) dissolved in wa-
ter was added to the reaction mixture and left for 16 h. The
reaction was quenched with the addition of a 10% HCl solu-
tion (one volume eq.) and extracted with EtOAc (×3). The or-
ganic layers were combined, dried with sodium sulfate, and
concentrated. The crude material was purified via flash silica
gel chromatography (4 : 1 Hex : EtOAc) affording 11a in 93%
1
(2.68 g). H NMR (400 MHz, methanol-d
4
) δ 5.79 (s, 2H), 2.59
1
3
(m, 3H). C NMR (101 MHz, methanol-d ) δ 204.56, 166.32,
4
1
65.90, 105.60, 95.59, 32.71.
.3.2 1-(2-Hydroxy-4,6-bisIJmethoxymethoxy)phenyl)ethan-
4
1
9
-one (10a). To a RBF under an argon atmosphere was added
(500 mg, 2.97 mmol, 1 eq.) and 15 mL of CH Cl , which
2
2
was then cooled to 0 °C and then diisopropylethylamine (1.6
mL, 8.91 mmol, 3 eq.) and MOMCl (0.5 mL, 6.5 mmol, 2.2
eq.) was added. The solution was then stirred overnight,
allowed to warm to room temperature, and then quenched
with methanol and washed with brine (×2). The combined or-
ganic layers were dried over anhydrous sodium sulfate, con-
1
yield (449 mg). H NMR (400 MHz, chloroform-d) δ 13.81 (s,
1H), 7.93 (d, J = 15.6 Hz, 1H), 7.79 (d, J = 15.7 Hz, 1H), 7.68–
7.57 (m, 3H), 7.46–7.35 (m, 4H), 6.26 (d, J = 2.4 Hz, 1H), 5.29
1
3
centrated, and purified via flash silica gel chromatography
(s, 2H), 5.19 (s, 2H), 3.51 (s, 3H), 3.50 (s, 3H). C NMR (101
1
(4 : 1, Hex : EtOAc) to afford 10a in 53% yield (406 mg).
H
MHz, chloroform-d) δ 193.15, 167.57, 163.74, 160.12, 142.72,
NMR (400 MHz, chloroform-d) δ 13.70 (s, 1H), 6.24 (1H), 6.22
135.69, 130.40, 129.16, 128.54, 127.61, 107.76, 97.73, 95.38,
+
(
s, 1H), 5.24 (s, 2H), 5.15 (s, 2H), 3.50 (s, 3H), 3.45 (s, 3H),
.63 (s, 3H). C NMR (101 MHz, chloroform-d) δ 203.40,
94.99, 94.30, 57.11, 56.70. HRMS calcd. for C19
[M + Na] 345.1338; found 345.1354.
21 6
H O Na m/z
1
3
+
2
1
5
2
67.02, 163.66, 160.57, 107.12, 97.32, 94.69, 94.21, 56.91,
4.3.6 (E)-3-Phenyl-1-(2,4,6-trisIJmethoxymethoxy)phenyl)-
prop-2-en-1-one (11b). To a RBF containing ethanol (4 mL)
was added benzaldehyde (0.05 mL, 0.5 mmol, 1 eq.) and 10b
(170 mg, 0.5 mmol, 1 eq.) and cooled to 0 °C. Once at tem-
perature, KOH (480 mg, 8.48 mmol, 17 eq.) dissolved in water
was added to the reaction mixture and left for 16 h. The reac-
tion was quenched with the addition of a 10% HCl solution
(one volume eq.) and extracted with EtOAc (×3). The organic
layers were combined, dried with sodium sulfate, and con-
centrated. The crude material was purified via flash silica gel
chromatography (4 : 1 Hex : EtOAc) affording 11b in a quanti-
+
+
6.64, 33.21. HRMS calcd. for C12
H16ONa m/z [M + Na]
79.0845; found 279.0842.
4
.3.3 1-(2,4,6-TrisIJmethoxymethoxy)phenyl)ethan-1-one
(10b). To a RBF under an argon atmosphere was added 9
(
250 mg, 1.49 mmol, 1 eq.) and 8 mL of CH Cl , which was
2
2
then cooled to 0 °C and then diisopropylethylamine (1.3 mL,
.95 mmol, 5 eq.) and MOMCl (0.37 mL, 4.91 mmol, 3.3 eq.)
7
was added. The solution was then stirred overnight, allowed
to warm to room temperature, and then quenched with
methanol and washed with brine (×2). The combined organic
layers were dried over anhydrous sodium sulfate, concen-
trated, and purified via flash silica gel chromatography (4 : 1,
1
tative yield (195 mg). H NMR (400 MHz, chloroform-d) δ
7.54–7.49 (m, 2H), 7.39–7.33 (m, 4H), 6.97 (d, J = 16.2 Hz,
1
Hex : EtOAc) to afford 10b in 39% yield (173 mg). H NMR
1H), 6.57 (s, 2H), 5.18 (s, 3H), 5.11 (s, 4H), 3.50 (s, 3H), 3.38
1
3
(
4
400 MHz, chloroform-d) δ 6.50 (s, 2H), 5.13 (s, 2H), 5.13 (s,
H), 3.46 (s, 3H), 3.45 (s, 6H), 2.48 (s, 3H). C NMR (101
(s, 6H). C NMR (101 MHz, chloroform-d) δ 194.50, 159.88,
156.06, 145.22, 135.02, 130.61, 129.22, 129.12, 128.55, 114.99,
97.36, 94.99, 94.81, 94.79, 56.51. HRMS calcd. for
1
3
MHz, chloroform-d) δ 201.65, 159.68, 155.41, 117.16, 97.34,
+
+
9
4.99, 94.69, 56.55, 56.44, 56.34, 32.76. HRMS calcd. for
C H O Na m/z [M + Na] 411.1420; found 411.1423.
21 24 7
+
+
C
H O
14 20 7
Na m/z [M + Na] 323.1107; found 323.1112.
.3.4 1-(2-Methoxy-4,6-bisIJmethoxymethoxy)phenyl)ethan-
-one (10c). To a dried RBF under an argon atmosphere was
added dry acetone (15 mL) to which was added 10a (400 mg,
.56 mmol, 1 eq.) under vigorous stirring. Once a homoge-
4.3.7 (E)-1-(2-Methoxy-4,6-bisIJmethoxymethoxy)phenyl)-3-
4
phenylprop-2-en-1-one (11c). To a RBF containing ethanol (15
mL) was added benzaldehyde (0.16 mL, 1.55 mmol, 1 eq.)
and 10b (350 mg, 1.29 mmol, 1 eq.) and cooled to 0 °C. Once
at temperature, KOH (1.23 g, 21.9 mmol, 17 eq.) dissolved in
water was added to the reaction mixture and left for 16 h.
The reaction was quenched with the addition of a 10% HCl
solution (one volume eq.) and extracted with EtOAc (×3). The
organic layers were combined, dried with sodium sulfate,
and concentrated. The crude material was purified via flash
silica gel chromatography (7 : 3 Hex : EtOAc) affording 11c in
1
1
nous solution was obtained, potassium carbonate (1.07 g, 7.8
mmol, 5 eq.) and dimethyl sulfate (0.74 mL, 7.8 mmol, 5 eq.)
were added. The reaction mixture was refluxed for 4 h,
quenched with saturated ammonium chloride, and extracted
with EtOAc (×2). The organic layers were combined, dried
over sodium sulfate, concentrated, and purified via flash sil-
1
ica gel chromatography (7 : 3 Hex : EtOAc) to obtain 10c in
a 80% yield (370 mg). H NMR (400 MHz, chloroform-d) δ
1
8
5% yield (358 mg). H NMR (400 MHz, chloroform-d) δ 6.45
7.52 (dq, J = 6.5, 4.0, 3.0 Hz, 2H), 7.38–7.34 (m, 4H), 6.96 (d, J
= 16.1 Hz, 1H), 6.50 (dd, J = 6.2, 2.0 Hz, 1H), 6.37 (d, J = 2.0
Hz, 1H), 5.20 (s, 1H), 5.11 (s, 1H), 3.76 (s, 3H), 3.51 (s, 3H),
(
(
d, J = 2.0 Hz, 1H), 6.30 (d, J = 2.0 Hz, 1H), 5.15 (s, 2H), 5.13
s, 2H), 3.78 (s, 3H), 3.47 (s, 3H), 3.45 (s, 3H), 2.47 (s, 3H).
1
3
13
C NMR (101 MHz, chloroform-d) δ 201.84, 159.92, 158.11,
3.39 (s, 3H). C NMR (101 MHz, chloroform-d) δ 194.56,
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2019

View Article Online
MedChemComm
Research Article
1
1
60.10, 158.77, 156.16, 144.91, 135.13, 130.52, 129.19, 129.07,
28.58, 96.15, 94.84, 94.80, 94.21, 56.53, 56.48, 56.16. HRMS
12 S. Haglund, J. Taipalensuu, C. Peterson and S. Almer, Br. J.
Clin. Pharmacol., 2008, 65, 69–77.
13 K. Rowland, L. Lennard and J. S. Lilleyman, Xenobiotica,
+
+
calcd. for C20
H O
23 6
m/z [M + H] 359.1494; found 359.1495.
1
999, 29, 615–628.
14 D. M. Tidd and A. R. P. Paterson, Cancer Res., 1974, 34,
38–746.
Conflicts of interest
7
The authors declare no conflict of interest.
1
1
1
5 P. Karran, Br. Med. Bull., 2006, 79–80, 153–170.
6 P. Karran and N. Attard, Nat. Rev. Cancer, 2008, 8, 24–36.
7 L. Chouchana, A. A. Fernández-Ramos, F. Dumont, C.
Marchetti, I. Ceballos-Picot, P. Beaune, D. Gurwitz and M.-A.
Loriot, Genome Med., 2015, 7, 37.
Acknowledgements
This work could not have been undertaken without the gra-
cious financial support from the Johnson Cancer Research
Center of Kansas State University and Startup Capital from
Kansas State University. Funding for this research, and for
funding for Johnathan Dallman, was provided by the NSF
REU program under grant number CHE-1460898.
1
8 T. Dervieux, T. L. Brenner, Y. Y. Hon, Y. Zhou, M. L.
Hancock, J. T. Sandlund, G. K. Rivera, R. C. Ribeiro, J. M.
Boyett, C.-H. Pui, M. V. Relling and W. E. Evans, Blood,
2
002, 100, 1240–1247.
1
9 J. M. Carethers, M. T. Hawn, D. P. Chauhan, M. C. Luce, G.
Marra, M. Koi and C. R. Boland, J. Clin. Invest., 1996, 98,
References
1
99–206.
1
2
3
G. H. Hitchings and G. B. Elion, Pharmacol. Rev., 1963, 15,
65–405.
J. M. Torpy, C. Lynm and R. M. Glass, JAMA, J. Am. Med.
Assoc., 2009, 301, 452.
T. Moriyama, R. Nishii, T.-N. Lin, K. Kihira, H. Toyoda, N.
Jacob, M. Kato, K. Koh, H. Inaba and A. Manabe,
Pharmacogenet. Genomics, 2017, 27, 236–239.
C. Vendrik, J. Bergers, W. De Jong and P. Steerenberg,
Cancer Chemother. Pharmacol., 1992, 29, 413–429.
O. Eden, J. Clin. Pathol., 2000, 53, 55–59.
20 R. Troxler and S. Brown, in Bilirubin, CRC Press, 2018, pp.
1–38.
3
21 E. A. Hullah, P. A. Blaker, A. M. Marinaki, M. P. Escudier
and J. D. Sanderson, J. Oral Pathol. Med., 2015, 44, 761–768.
22 P. P. Siva, K. M. Murali, A. Deepak, S. Murali, S. Michael and
M. Sandhya, Drug Metab. Lett., 2016, 10, 264–269.
23 C. J. Weeramange, C. M. Binns, C. Chen and R. J. Rafferty,
J. Pharm. Biomed. Anal., 2018, 151, 106–115.
24 A. X. Torres Hernandez, C. Weeramange, P. Desman, A.
Fatino, O. Haney and R. J. Rafferty, MedChemComm,
2019, 10, 169–179.
25 R. B. Burrows and C. Cintron, Anal. Biochem., 1983, 130,
376–378.
26 P. C. Gaffney and W. M. Cooper, Ann. N. Y. Acad. Sci.,
1954, 60, 478–482.
4
5
6
C. Cuffari, E. Seidman, S. Latour and Y. Theoret, Can. J.
Physiol. Pharmacol., 1996, 74, 580–585.
7
8
D. S. Rampton, BMJ [Br. Med. J.], 1999, 319, 1480.
M. L. Seinen, D. P. van Asseldonk, N. K. de Boer, N.
Losekoot, K. Smid, C. J. Mulder, G. Bouma, G. J. Peters and
A. A. van Bodegraven, Optimalisation of conventional therapies
in inflammatory bowel disease, 2016, vol. 7, pp. 812–819.
X. Roblin, L. P. Biroulet, J. M. Phelip, S. Nancey and B.
Flourie, Am. J. Gastroenterol., 2008, 103, 3115.
27 J. Singh, L. Schwarz and F. Wiebel, Biochem. J., 1980, 189,
369–372.
28 A. S. Hyde, A. M. Thelen, J. J. Barycki and M. A. Simpson,
J. Biol. Chem., 2013, 288, 35049–35057.
9
1
1
0 R. B. Gearry and M. L. Barclay, J. Gastroenterol. Hepatol.,
005, 20, 1149–1157.
1 J. A. Nelson, J. W. Carpenter, L. M. Rose and D. J. Adamson,
29 B. J. Sommer, J. J. Barycki and M. A. Simpson, J. Biol. Chem.,
2004, 279, 23590–23596.
30 T.-C. Chou and P. Talalay, Adv. Enzyme Regul., 1984, 22,
27–55.
2
Cancer Res., 1975, 35, 2872–2878.
This journal is © The Royal Society of Chemistry 2019
Med. Chem. Commun.