Design and synthesis of triple inhibitors of janus kinase (JAK), histone deacetylase (HDAC) and Heat Shock Protein 90 (HSP90)

Article abstract of DOI:10.1016/j.bmcl.2018.03.009

Inhibition of multiple signaling pathways in a cancer cell with a single molecule could result in better therapies that are simpler to administer. Efficacy may be achieved with reduced potency against individual targets if there is synergy through multiple pathway inhibition. To achieve this, it is necessary to be able to build multi-component ligands by joining together key pharmacophores in a way which maintains sufficient activity against the individual pathways. In this work, designed triple inhibiting ligands are explored aiming to block three completely different target types: a kinase (JAK2), an epigenetic target (HDAC) and a chaperone (HSP90). Although these enzymes have totally different functions they are related through inter-dependent pathways in the developing cancer cell. Synthesis of several complex multi-inhibiting ligands are presented along with initial enzyme inhibition data against 3 biological target classes of interest. A lead compound, 47, was discovered which had low micromolar activity for all 3 targets. Further development of these complex trispecific designed multiple ligands could result in a ‘transient drug’ an alternative combination therapy for treating cancer mediated via a single molecule.

Full text of DOI:10.1016/j.bmcl.2018.03.009

Accepted Manuscript
Design and Synthesis of Triple Inhibitors of Janus Kinase (JAK), Histone De-
acetylase (HDAC) and Heat Shock Protein 90 (HSP90)
Lianbin Yao, Sten Ohlson, Brian W. Dymock
PII:
S0960-894X(18)30188-4
https://doi.org/10.1016/j.bmcl.2018.03.009
BMCL 25663
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
31 January 2018
1 March 2018
2 March 2018
Please cite this article as: Yao, L., Ohlson, S., Dymock, B.W., Design and Synthesis of Triple Inhibitors of Janus
Kinase (JAK), Histone Deacetylase (HDAC) and Heat Shock Protein 90 (HSP90), Bioorganic & Medicinal
Chemistry Letters (2018), doi: https://doi.org/10.1016/j.bmcl.2018.03.009
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
journal homepage: www.els evier.com
Design and Synthesis of Triple Inhibitors of Janus Kinase (JAK), Histone
Deacetylase (HDAC) and Heat Shock Protein 90 (HSP90)
Lianbin Yao,a Sten Ohlson,b,* Brian W. Dymocka,c
*
a Department of Pharmacy, National University of Singapore, 18 Science Drive 4, Singapore 117543
b School of Biological Sciences, Nanyang Technological University (NTU), Singapore 637551
c Current address: QEDDI, Uniquest Pty Ltd, General Purpose South Bldg, Staff House Road, The University of Queensland, Brisbane QLD 4072 Australia
ARTICLE INFO
ABSTRACT
Article history:
Received
Inhibition of multiple signaling pathways in a cancer cell with a single molecule could result in
better therapies that are simpler to administer. Efficacy may be achieved with reduced potency
against individual targets if there is synergy through multiple pathway inhibition. To achieve
this, it is necessary to be able to build multi-component ligands by joining together key
pharmacophores in a way which maintains sufficient activity against the individual pathways. In
this work, designed triple inhibiting ligands are explored aiming to block three completely
different target types: a kinase (JAK2), an epigenetic target (HDAC) and a chaperone (HSP90).
Although these enzymes have totally different functions they are related through inter-dependent
pathways in the developing cancer cell. Synthesis of several complex multi-inhibiting ligands
are presented along with initial enzyme inhibition data against 3 biological target classes of
interest. A lead compound, 47, was discovered which had low micromolar activity for all 3
targets. Further development of these complex trispecific designed multiple ligands could result
in a ‘transient drug’, an alternative combination therapy for treating cancer mediated via a single
molecule.
Revised
Accepted
Available online
Keywords:
Multiple ligand
Janus kinase
HDAC
HSP90
Transient Drug
2009 Elsevier Ltd. All rights reserved
.
Combination drug treatment is the mainstay of modern cancer therapy.1 However even with the welcome progress made with
immune-oncology therapeutics,2 there remains a high unmet need in treatment of many cancers and especially drug resistant
disease.3 Combination therapies are typically given as two or more separate drugs which need to be scheduled to ensure maximal
efficacy with the lowest possible toxicity. However this is not always successful and drug-drug interactions can be a particular
problem.4 New ways of delivering combination therapy are therefore urgently needed. The new concept of the transient drug, where
less potency at the target is balanced with multiple target inhibition creating a synergistic effect which is efficacious but without
overt toxicity, could be one such new approach.5 To approach such a challenge molecules able to inhibit multiple targets will need
to be designed and synthesized. The choice of target is crucial and must include targets important in disease pathways, or known to
be important when combined with inhibition of certain other pathways.6 In this regard, we have been interested in inhibition of
kinase signaling pathways, epigenetic targets and chaperone proteins.
Our previous work has led to dual inhibitors of JAK (Janus) kinases and HDACs (histone deacetylases), EY3238 (1) and
YLB343B (2), with low nanomolar potency against members of both target families.7, 8 In this present work, we aimed to stretch the
possibilities of small molecule chemistry to assess the feasibility of preparing single molecules which could have potential to be
inhibitors of 3 different pathways by inhibition of 3 different types of enzyme target.
OH
HN
O
O
N
N
O
N
O
NH
OH
N
O
N
H
N
N
N
H
2
)
YLB343B (
1
EY3238 (
)
JAK1 IC50 = 11 nM
HDAC1 IC50 = 6 nM
HDAC6 IC50 = 1 nM
JAK2 IC50 = 1.4 nM
HDAC6 IC50 = 2.1 nM
Figure 1. Structures of the JAK2-HDAC dual inhibitors EY3238 (1) and YLB343B (2).

We aimed to build upon our previous work in JAK-HDAC dual inhibitors by adding on a third activity against the chaperone
Heat Shock Protein 90 (HSP90). HSP90 is massively upregulated in developing cancer cells due to its importance in ensuring the
correct folding of a wide array of client proteins many of which are oncoproteins.9 HSP90 promotes activation of the oncogenic
inflammatory mediator STAT-3.10 JAK kinases are direct activators of STATs11 hence inhibition of HSP90 and JAK could be
synergistic. Inhibition of HSP90 has been shown to be effective in cells with an activated JAK-STAT pathway.12 HDACs, in
particular HDAC6, are known to be important in modulating HSP90 complexes.13, 14 Hence exploring a triple inhibitor of all three
enzymes could have value in treating cancer with JAK-STAT and/or HSP90 dependency. We selected the marketed drugs
ruxolitinib (3),15 vorinostat (4)16 and the published HSP90 inhibitor BEP800 (5)17 as the base molecules for initial exploratory work
towards a triple inhibitor of JAK-HDAC-HSP90.
Analysis of published crystal structures of 418 and 519 indicates clearly which regions of the structures are solvent exposed. There
are no published crystal structures of 3 in complex with JAK kinases but structures of similar molecules such as tofacitnib are
available.20 These structures indicate that there are two regions of the structure of 3 which are solvent exposed, or with a direct path
out to solvent: the nitrile group and the central carbon of the pyrimidine. These were therefore considered as viable attachment
points for either an HDAC or HSP90 binding moiety. Compound 4 has a solvent exposed terminal phenyl group which we have
established as an attachment point for a JAK inhibiting moiety. HSP90 binds small molecules in its ATP site which is a deep pocket
but the structure of molecules related to 5 indicates two solvent exposed regions: the amide alkyl group and the meta position of the
phenyl.19 Using this information our initial designs focused on using a simplified core based on 3 with attachment of a HSP90
binding moiety, based on 5, on the pyrimidine ring and an HDAC binding moiety, based on 4, on the pyrazole (compare with 28)
(Figure 2a). Alternatively we planned to explore the HDAC binding group attached to the pyrimidine and the HSP90 binding group
attached to the pyrazole (Figure 2b).
Figure 2. Known SAR of reference compounds ruxolitinib (3), vorinostat (4) and BEP800 (5) enables design of multiple ligands. Solvent exposed areas
(blue colour) of each inhibitor can be used to identify connection points for combination compounds. Key hydrogen bonding groups (red colour) are
preserved to ensure maximum chance of binding at each target. A: a design based on the compact structure of 28 with attachment of the HSP90 binder to the
pyrimidine ring of the JAK binding motif, used for synthesis of compound 39 ; B: HDAC binder is attached through the pyrimidine of the JAK inhibitor
while the HSP90 binding component can be attached to the pyrazole via a linker, used for synthesis of compounds 17-19, 26, 27, 46, 47.
In our previous work we had developed a suitable template for JAK-HDAC dual inhibitors based on 38 but we had not developed
methods for dual HSP90 inhibitors therefore before embarking on the challenging synthesis of a triple inhibitor we decided to firstly
explore JAK-HSP90 dual inhibitor chemistry.
Synthesis of a prototype JAK-HSP90 dual inhibitor (based on the designs of Figure 2a), started by building the HSP90 portion
from aldehyde 6 through a SNAr reaction using 2-mercaptoethylacetate and cyclisation of the intermediate to form thienyl
pyrimidine 7 in good yield (Scheme 1). Suzuki coupling with hydroxyphenyl boronic acid gave 8 followed by ethyl amide
formation gave phenol 9 in moderate overall yield. Installation of linkers with 4, 5 and 6 carbons was achieved with the respective
dibromo alkanes giving bromides 10-12. The previously reported8 tosyl-protected JAK binding fragment, 13, could then be attached
via N-alkylation giving 14-16. Finally deprotection with sodium methoxide gave the desired products 17-19.

These three prototype JAK-HSP90 inhibitors (17-19) were tested in a JAK2 ATPase kinase assay using radio-labelled ATP and
in an HSP90 fluorescence polarisation assay using a geldanamycin-based probe. Although 18 and 19 had JAK2 IC50s of around
1 µM unfortunately 17 was not active against JAK2 and none of the compounds registered any activity against HSP90 at
concentrations up to 10 µM. We reasoned that this could be due to a weak HSP90 pharmacophore. In published studies leading to
5, halogen substituents on the phenyl ring played an important role in achieving good potency. Hence we prepared two analogues,
26 and 27, with a 2,4-dichlorophenyl in the HSP90 portion using similar chemistry with the expectation that this change would
enhance HSP90 potency (Scheme 2). Following enzyme assay testing we were pleased to observe that both 26 and 27 inhibited
HSP90 with IC50s of 0.4 and 0.11 µM, respectively (Table 1).
Scheme 1. Preparation of prototype JAK-HSP90 inhibitorsa
aReagents and Conditions: (a) HSCH2CO2Et, K2CO3, MeCN, reflux, O/N, 70%; (b) 3-hydroxyphenyl boronic acid, Pd(PPh3)4, K3PO4, dioxane, 100 °C, O/N,
68%; (c) EtNH2, EtOH, MW, 130 °C, 3 h, 30%; (d) 1,4-dibromobutane (n=1), 1,4-dibromopentane (n=2), 1,4-dibromohexane (n=3), DMF, RT, 18 h, 46%;
(e) 13, DMF, K2CO3, RT, 96% (n=1); (f) NaOMe, MeOH, THF, 70 °C, 3 h, 62-76%.
Having validated the connection strategy for the BEP800-based HSP90 fragment, we then progressed towards synthesis of the
first triple inhibitors (Schemes 3 and 4). Pyrazole boronate ester 28 was protected with ethyl vinyl ether and coupled to
dichloropyrrolopyrimidine 30 in good yield (Scheme 3). This building block offers a functional group handle via the chloro attached
to the pyrimidine which is later used to install the HSP90 binding fragment.

Scheme 2. Preparation of 2,4-dichlorophenyl productsa
Cl
N
O
N
H2
N
N
H2
N
N
N N
X
HN
O
S
S
n
N
N
Cl
S
a
c
d
O
7
O
Cl
N
Cl
N
N
R
HN
Br
H2
N
O
OH
n
Cl
Cl
24
25
(n=2) R=Ts
(n=3)
20
21
X = O
X = NH
22
23
(n=2)
(n=3)
b
e
26
(n=2)
27 (n=3)
R=H
aReagents and Conditions: (a) 2,4-dichloro-5-hydroxyphenyl boronic acid, Pd(PPh3)4, K3PO4, dioxane, 100 °C, O/N, 57%; (b) EtNH2, EtOH, MW, 130 °C,
3 h, 36%; (c) 1,4-dibromopentane (n=2), 1,4-dibromohexane (n=3), DMF, RT, 18 h, 59-68%; (d) 13, DMF, K2CO3, RT; (e) NaOMe, MeOH, THF, 70 °C,
3 h, 34-44% (over 2 steps).
Table 1. JAK2 and HSP90 enzyme data for JAK-HSP90 dual inhibitors
IC50 (µM)a
Compound
JAK2
-
HSP90
0.016b
Gd
1
1.4c
>10
1.09
1.42
3.45
1.71
17
18
19
26
27
>10
>10
>10
0.40
0.11
a
10 dose IC50 determination.7 b Reported value (http://www.reactionbiology.com). c reported value from same assay7
Protection of the pyrrole nitrogen was necessary to enable control of the subsequent reaction and was achieved with a tosyl group
giving 32. Removal of the ethoxy ethyl protecting group under acidic conditions revealed the pyrazole 33 which could then be
alkylated with the linker using methyl bromoheptanoate furnishing ester 34. Aminophenol displacement of the chloro to give 35
was followed by alkylation with the HSP90 fragment 23 (see Scheme 1) using potassium carbonate in DMF. Having all the carbons
of the target compound in place all that remained was to install the hydroxamate, via THP-protected hydroxylamine, and then
deprotect the tosyl and THP groups (37, 38) to give the first triple target, 39.
Following the design of Figure 2b, a similar strategy was used to synthesise 46 (Scheme 4). In this case, the carboxylic acid 47
was also isolated probably from some incomplete hydroxamate formation. This acid was considered worthy of biological testing
given that acids have been described as HDAC inhibitors, albeit weaker than hydroxamates. The acid was synthesized directly from
43 via saponification.
The three target compounds were tested against HDAC1, HDAC6, JAK2 and HSP90 (Table 2). Surprisingly, 39 was very potent
against HDAC6 with an IC50 of 8 nM. This is not surprising since it is based on compounds very similar to 2 with the preferred 6-
carbon linker connecting the hydroxamate to the heterocycle.8 However it was considerably less active against HDAC1 and no
significant activity was detected against either JAK2 or HSP90 at the top concentration tested (10 µM). It is generally the case with
dual inhibitors that HDAC1 potency is more challenging to maintain due to the size and shape of the binding site at the immediate
surface next to the HDAC substrate pocket.7, 8 This indicates that attaching the HSP90 binder to the pyrimidine of the JAK binding
core is not a suitable combination. Hence the proven strategy explored with the prototype JAK-HSP90 inhibitors (Schemes1 and 2,
Table 1) was expected to provide better results (Table 2). Indeed, 46 did have activity for JAK2 and HDAC6 but unfortunately not
for HSP90. However, the carboxylic acid 47 did fully inhibit HSP90 with an IC50 of 20.2 µM as well as JAK2 (3.76 µM) and
HDAC6 (6.3 µM).
Scheme 3. Preparation of JAK-HSP90-HDAC triple inhibitor based on design shown in figure 2aa

aReagents and Conditions: (a) toluene, 35°C,13 h; (b) Pd(PPh3)4, Na2CO3, dioxane/H2O, 100°C, overnight, 70%; (c) TsCl, Et3N, DMAP, DCM, rt, 3 h, 68%;
(d) HCl (aq.), THF, rt, ON; (e) methyl-7-bromoheptanoate, K2CO3, DMF, RT, ON, 86%; (f) p-nitrophenol, Pd2(dba)3, X-phos, K2CO3, tBuOH, 100°C, 24 h,
43%; (g) 23, K2CO3, DMF, rt, ON, 65%; (h) (1) THPONH2, LiOH, THF, H2O; (2) HATU, TEA, DMSO, rt, ON, 74%; (i) MeONa, MeOH, THF, 75°C, 5 h
70 %; (j) HCl, dioxane, rt, 18h, 46%.
This broad micromolar activity demonstrated with 47 shows that it is possible to inhibit three different biological targets with a
single small molecule. Although the inhibition is in the micromolar range , if there is synergy between the inhibition of these
targets, this could be the first step towards a so-called ‘transient’ drug, i.e. a molecule which inhibits a group of specific targets
weakly, but strongly enough to elicit desired biological effects through the combination of its inhibition profile without causing
overt toxicity.5 Further work will be required to optimize these initial structures and to evaluate cell permeable trispecific inhibitors
for the optimal balance of target potency for best effects in cancer cells and tumors.

Scheme 4. Preparation of JAK-HSP90-HDAC triple inhibitor based on design shown in figure 2ba
aReagents and Conditions: (a) Pd2(dba)3, XPhos, K2CO3, tBuOH, 100°C, overnight, 77%; (b) HCl (aq), THF, rt, 24h; (c) 23, K2CO3, DMF, 60°C, 24 h;
(d) (1) THPONH2, LiOH, THF, H2O; (2) HATU, TEA, DMSO, RT, ON; (e) MeONa, MeOH, THF, 70°C, 5 h; (f) HCl (aq.), 1,4-dioxane rt, 18h; (g)
NaOMe, THF, MeOH, 75°C, 5 h, 59%.
Table 2. Enzyme inhibition for designed triple inhibitors
IC50 (µM)a
Compound
HDAC1b
HDAC6b
JAK2c
HSP90αd
39
46
47
2.47
0.008
2.82
6.30
>100
14.1
3.76
>100
>100
20.2
>100
>100
a
10 dose IC50 determinations.7 b Vorinostat used as a positive control (HDAC1 IC50 = 306 nM; HDAC6 IC50 = 20 nM). c Ruxolitinib used as a positive
control (IC50 <1nM). d Geldanamycin used as a positive control (IC50 = 27 nM).
In conclusion, we have shown in this work that it is possible to develop small molecules which have specific inhibition of three
different enzyme classes, all of which are strongly implicated in oncogenesis. The relatively weak inhibition exhibited by the triple
inhibitor 47 may be the first step towards a transient drug.5 Further work will be required to understand the scope of these designs
and to develop cell permeable inhibitors.
Acknowledgments
This research was funded by a generous grant from the National Medical Research Council of Singapore (NMRC CBRG Grant
NMRC/CBRG12Nov114 (B.W.D./S.O.))
References and notes
1.
nci/budget/plan/treating/combining-therapies
2. Sharma, P.; Allison, J. P. Immune checkpoint targeting in cancer therapy: toward combination strategies with curative potential.
Cell 2015, 161, 205-214.
https://www.cancer.gov/about-nci/budget/plan/treating/combining-therapies.
https://www.cancer.gov/about-

3. https://www.cancer.gov/about-cancer/treatment/research/drug-combo-resistance.
cancer/treatment/research/drug-combo-resistance
https://www.cancer.gov/about-
4.
5.
6.
Scripture, C. D.; Figg, W. D. Drug interactions in cancer therapy. Nature Reviews Cancer 2006, 6, 546-558.
Ohlson, S. Designing transient binding drugs: a new concept for drug discovery. Drug Discovery Today 2008, 13, 433-439.
Anighoro, A.; Bajorath, J.; Rastelli, G. Polypharmacology: challenges and opportunities in drug discovery. J. Med. Chem. 2014,
57, 7874-7887.
7. Yang, E. G.; Mustafa, N.; Tan, E. C.; Poulsen, A.; Ramanujulu, P. M.; Chng, W. J.; Yen, J. J.; Dymock, B. W. Design and
synthesis of janus kinase 2 (JAK2) and histone deacetlyase (HDAC) bispecific inhibitors based on pacritinib and evidence of dual pathway
inhibition in hematological cell lines. J. Med. Chem. 2016, 59, 8233-8262.
8. Yao, L.; Mustafa, N.; Tan, E. C.; Poulsen, A.; Singh, P.; Duong-Thi, M.-D.; Lee, J. X.; Ramanujulu, P. M.; Chng, W. J.; Yen, J.
J.; Dymock, B. W. Design and Synthesis of Ligand Efficient Dual Inhibitors of Janus Kinase (JAK) and Histone Deacetylase (HDAC)
Based on Ruxolitinib and Vorinostat. J. Med. Chem. 2017, 60, 8336-8357.
9.
Whitesell, L.; Lindquist, S. L. HSP90 and the chaperoning of cancer. Nature Reviews Cancer 2005, 5, 761-772.
10. Bohonowych, J.; Hance, M.; Nolan, K.; Defee, M.; Parsons, C.; Isaacs, J. Extracellular Hsp90 mediates an NF-κB dependent
inflammatory stromal program: Implications for the prostate tumor microenvironment. The Prostate 2014, 74, 395-407.
11.
12.
Rawlings, J. S.; Rosler, K. M.; Harrison, D. A. The JAK/STAT signaling pathway. J. Cell Sci. 2004, 117, 1281-1283.
Proia, D. A.; Foley, K. P.; Korbut, T.; Sang, J.; Smith, D.; Bates, R. C.; Liu, Y.; Rosenberg, A. F.; Zhou, D.; Koya, K.
Multifaceted intervention by the Hsp90 inhibitor ganetespib (STA-9090) in cancer cells with activated JAK/STAT signaling. PLoS One
2011, 6, e18552.
13.
aryl hydrocarbon receptor signaling. J. Biol. Chem. 2009, 284, 7436-7445.
14. Kovacs, J. J.; Murphy, P. J.; Gaillard, S.; Zhao, X.; Wu, J.-T.; Nicchitta, C. V.; Yoshida, M.; Toft, D. O.; Pratt, W. B.; Yao, T.-P.
HDAC6 regulates Hsp90 acetylation and chaperone-dependent activation of glucocorticoid receptor. Mol. Cell 2005, 18, 601-607.
Kekatpure, V. D.; Dannenberg, A. J.; Subbaramaiah, K. HDAC6 modulates Hsp90 chaperone activity and regulates activation of
15.
16.
Mesa, R. A.; Yasothan, U.; Kirkpatrick, P. Ruxolitinib. Nat. Rev. Drug Discovery 2012, 11, 103-104.
Marks, P. A.; Breslow, R. Dimethyl sulfoxide to vorinostat: development of this histone deacetylase inhibitor as an anticancer
drug. Nat. Biotechnol. 2007, 25, 84-90.
17.
Massey, A. J.; Schoepfer, J.; Brough, P. A.; Brueggen, J.; Chène, P.; Drysdale, M. J.; Pfaar, U.; Radimerski, T.; Ruetz, S.;
Schweitzer, A. Preclinical antitumor activity of the orally available heat shock protein 90 inhibitor NVP-BEP800. Mol. Cancer Therap.
2010, 9, 906-919.
18.
snapshots of human HDAC8 provide insights into the class I histone deacetylases. Structure 2004, 12, 1325-1334.
19. Brough, P. A.; Barril, X.; Borgognoni, J.; Chene, P.; Davies, N. G.; Davis, B.; Drysdale, M. J.; Dymock, B.; Eccles, S. A.;
Somoza, J. R.; Skene, R. J.; Katz, B. A.; Mol, C.; Ho, J. D.; Jennings, A. J.; Luong, C.; Arvai, A.; Buggy, J. J.; Chi, E. Structural
Garcia-Echeverria, C. Combining hit identification strategies: fragment-based and in silico approaches to orally active 2-aminothieno [2, 3-
d] pyrimidine inhibitors of the Hsp90 molecular chaperone. J. Med. Chem. 2009, 52, 4794-4809.
20. Alicea-Velazquez, N. L.; Boggon, T. J. The use of structural biology in Janus kinase targeted drug discovery. Curr. Drug Targets
2011, 12, 546-555.
Supplementary Material
Supplementary material describing methods of synthesis and characterization of compounds is available.
* Correspondance
BWD: bwdnus@gmail.com
SO: SOhlson@ntu.edu.sg

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
Design and Synthesis of Triple Inhibitors of
Janus Kinase (JAK), Histone Deacetylase
(HDAC) and Heat Shock Protein 90 (HSP90)
Leave this area blank for abstract info.
Lianbin Yao, Sten Ohlson, Brian W. Dymock