Design and synthesis of protein kinase Cα activators based on 'out of pocket' interactions

Article abstract of DOI:10.1016/j.bmcl.2011.04.108

Novel types of PKCα activators based on isobenzofuranone bearing a myo-inositol moiety were designed and synthesized. The derivatives with bulky substituents on the myo-inositol moiety significantly activated PKCα, but their binding sites were not the same as that of phorbol ester.

Full text of DOI:10.1016/j.bmcl.2011.04.108

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3587–3590
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and synthesis of protein kinase C
interactions
a activators based on ‘out of pocket’
Go Hirai a, Megumi Ohkubo a, Yuki Tamura a,b, Mikiko Sodeoka a,b,
⇑
a RIKEN Advanced Science Institute, Hirosawa, Wako 351-0198, Japan
b Sodeoka Live Cell Chemistry Project, ERATO, Japan Science and Technology Agency, Wako 351-0198, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Novel types of PKC
a
activators based on isobenzofuranone bearing a myo-inositol moiety were designed
Received 11 March 2011
Revised 21 April 2011
Accepted 25 April 2011
Available online 28 April 2011
and synthesized. The derivatives with bulky substituents on the myo-inositol moiety significantly acti-
vated PKC , but their binding sites were not the same as that of phorbol ester.
a
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Protein kinase C
a
Isobenzofuranone derivatives
Enzymatic activity
myo-Inositol
Binding model
Protein kinase C (PKC) isozymes are known to play important
roles in intracellular signal transduction pathways associated with
proliferation, differentiation, and apoptosis.1 The mammalian PKC
family comprises 11 isozymes grouped into three classes, known
also show anti-cancer activity and have recently been suggested to
be candidate therapeutic agents for Alzheimer’s disease.4 There-
fore, various analogues of these complex natural PKC activators
have been synthesized.5 As simpler compounds, DAG-based C1 do-
main ligands were developed by Marquez and co-workers.6 We
also reported conformationally constrained DAG-based isobenzof-
uranone (IB) derivatives such as 1 (Fig. 1A) as C1 domain ligands of
PKCa7 and revealed the importance of the phenolic alkyl chain at
as conventional PKCs (cPKCs:
a, bI/bII,
c), novel PKCs (nPKCs: d,
e
,
g
, h), and atypical PKCs (aPKCs: f, s/k). PKCs have a kinase do-
main that phosphorylates their substrate proteins and regulatory
domains that control the kinase activity. Regulatory domains of
cPKCs are composed of the membrane targeting domains, two C1
domains (C1A and C1B) and a C2 domain, and an autoinhibitory
pseudo-substrate sequence. The activity of cPKCs is regulated by
activating ligands, such as 1,2-diacylglycerol (DAG), phosphatidyl-
serine (PS), and calcium ion. Although the molecular mechanism of
cPKCs activation has not been fully clarified, it has been proposed
that cPKCs initially binds to the membrane via Ca2+-dependent
interaction of PS with the C2 domain. Further binding of activating
ligands to the C1 domain on the lipid bilayer allows release of the
pseudo-substrate sequence from the active site in the kinase do-
main and stabilizes the active conformer of cPKCs.
It is well-known that various natural products, such as phorbol
esters (PMA and PDBu, Fig. 1A), bryostatins, aplysiatoxin, and tele-
ocidin, also bind to the C1 domain and activate cPKCs and nPKCs.2
Since many of these compounds have strong tumor-promoting
activity, PKC inhibitors have attracted attention as anti-cancer drug
candidates.3 On the other hand, PKC activators, such as bryostatins,
the C7 position for PKC
in development of the more potent PKC
During the course of SAR studies, we found that PKC
a
activation.8 Further SAR studies resulted
activator 2.9
a
a
activator
3 having a straight alkyl chain on the acyl group of IB showed weak
competitive binding against [3H]PDBu, but exhibited significant
PKC
bol esters bind preferably to the C1B domain and DAGs bind to the
C1A domain in PKC . These results suggested that the DAG-like
molecule 3 is likely to bind to the C1A domain to activate PKC , be-
a
activating ability.9 According to Cho and co-workers,10 phor-
a
a
cause 3 showed weak binding affinity toward the phorbol ester
binding domain (Phorbol BD). On the other hand, Irie and Blum-
berg reported that phorbol esters binds to their chemically synthe-
sized or GST-fusion C1A and C1B domains with similar affinities.11
Therefore the possibility that 3 might bind to an unknown binding
site on PKC
either case, to further analyze the molecular mechanism of PKC
activation, a series of potent PKC activators with different selec-
tivity to each ligand-binding site would be required. We have al-
a other than C1 domains could not be ruled out. In
a
a
ready reported a highly potent PKC
a activator with high affinity
for the Phorbol BD.9 Here, we report a novel class of PKC
tors having negligible affinity for the Phorbol BD.
a activa-
⇑
Corresponding author. Fax: +81 48 462 4666.
E-mail address: sodeoka@riken.jp (M. Sodeoka).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.04.108

3588
G. Hirai et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3587–3590
Scheme 1. Synthesis of IB derivatives 4a–4f.
Figure 1. (A) Structures of phorbol esters and isobenzofuranone (IB) derivatives 1–
3. (B) Binding model of 1 (short chain analogue: R4 = C3H7) with PKC
C1B domain.
a
Essential amino acids for binding to the C1B domain are indicated in brown. Outer
amino acids around the binding site are indicated in green. Amino acids in
parentheses are the corresponding amino acids of the C1A domain. (C) Structure of
designed PKCa activator 4. (D) Schematic illustration of the concept of ‘out of
pocket’ interaction.
To design new C1A domain ligand molecules, we postulated
that the core structure of IB derivatives is able to bind to both
C1A and C1B domains, and modification of the side chains could
change the affinity towards each domain. Since the much weaker
Figure 2. (A) Ki values of compounds estimated from competitive inhibition of
PKCa activating ability of 3 than that of 2 suggested insufficient
[3H]PDBu binding to PKC
a
. (B) Relative activity of PKC
indicated compounds (0% = phosphorylation level without activating ligand;
100% = phosphorylation level with 10 M PMA).
a in the presence of the
binding affinity of 3 to its binding site, we selected the high affinity
phorbol BD ligand 2 as a parent compound. The isobenzofuranone
structure is expected to interact with the binding pocket by form-
ing hydrogen bonds with key inner amino acids, which are quite
similar between the two domains (for C1A: Thr48, Ile57, and
Gly59; for C1B: Thr113, Leu122, and Gly124). In contrast, the outer
amino acids around the binding sites of the two C1 domains are
significantly different (Fig. 1B). Therefore, to develop a C1A-selec-
tive ligand, we focused on the outer amino acids, and designed
compound 4 (Fig. 1C). Although 2 has high affinity for the C1B do-
main, we anticipated that the ‘out of pocket’ interaction between
the substituents R5 on the inositol moiety and the outer amino acid
residues would discriminate the two C1 domains and influence the
l
long alkyl chain at the C7 position would be important for the
interaction of ligand-C1 domain complex with PS-containing lipid
bilayer to stabilize the active conformer of PKCa, and thus such a
chain was introduced at the axial hydroxyl group of the inositol
moiety.
Synthesis of 4 is illustrated in Scheme 1. Since we have already
established methodology for asymmetric synthesis of 5,9 transfor-
mation of 5 to the corresponding triflate 6 was performed via a
routine three-step sequence of trifluoromethanesulfonylation,
hydrogenolysis of benzyl ether, and re-protection of the resulting
primary alcohol with a TBS group. Synthesis of the recognition
moiety began from the known C2-symmetric myo-inositol deriva-
tive 7.12 Alkylation of the axial secondary alcohol of 7 with a
straight C12 chain, followed by etherification with propargyl bro-
mide, gave fully protected inositol derivative 8 in good yield. After
selectivity (Fig. 1D). Furthermore, in case of whole PKCa, interac-
tion of the substitutents R5 with the other domains such as C2 do-
main may influence the binding to each C1 domains. myo-Inositol
structure was chosen as the recognition moiety for ‘out of pocket’
interaction because of its stability and C2-symmetric structure, and
was introduced at the C7 position of 2 via a rigid alkyne linker. A

G. Hirai et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3587–3590
3589
removal of the cyclic acetal of 8, introduction of various groups
into tetraol 9 by acetylation (10a) or alkylation with MeI (10b),
prenyl bromide (10c), or benzyl bromides (10d–10f) afforded
10a–10f. Connection of alkynes 10a–10f with triflate 6 were inves-
tigated utilizing Hagiwara–Sonogashira conditions.13 Although
only 29% yield of the desired coupling product was obtained in
the case of 10c, probably due to the presence of the prenyl ether
moiety, all other coupling reactions proceeded smoothly in 60%
to quantitative yield. Finally, removal of the TBS group gave IB
derivatives 4a–4f.
4b have binding ability to the C1B domain but 4c–4f do not. To
speculate the origin of this selectivity, a binding model of 4b with
the C1B domain of PKCa was constructed based on the binding
model of 1 shown in Figure 1B, in which we postulated the IB skel-
eton of 4b would bind to the ligand binding site and form hydrogen
bonds similar to compound 1. The substituent at C7 and the t-Bu
group of R3 of 1 were replaced by alkyne-inositol moiety 10b and
CH(t-Bu)2 group, and molecular dynamics and energy minimiza-
tion calculations were performed by Discovery Studio 2.5. As
shown in Figure 3A, introduction of the bulky hydrophobic CH(t-
Bu)2 group would result in the stabilization of the complex of IB
derivatives with the C1B domain, in which the CH(t-Bu)2 group
effectively interacted with Leu121 and Leu125 (Fig. 3A right).
Moreover, one of the R5 substituents on the inositol moiety would
be located around Pro112 and appear to contribute to ‘out of pock-
et’ interaction with the C1B domain (Fig. 3A). However, in case of
bulky R5 substituents such as the benzyl group, this ‘out of pocket’
interaction would be unfavorable. Namely, the steric repulsion be-
tween bulky R5 substituents and Pro112 might prevent these com-
pounds from interacting with the C1B domain (Fig. 3B). On the
other hand, the C1B domain would exist between the C1A and
the C2 domain, and tight interaction with these domains was pro-
posed.15–17 These reports suggested that neighboring C1A and C2
domains would be located around the binding site of phorbol ester
on the C1B domain. Thus there is another possibility that IB deriv-
atives having bulky R5 substituents could not bind to C1B domain
due to the steric repulsion with neighboring domains (Fig. 3B).
The binding affinity of these compounds was evaluated by assay
of competitive inhibition of the binding of [3H]PDBu to whole PKC
a
in the presence of PS vesicles.14 As summarized in Fig. 2A, variation
of substituent R5 on the inositol moiety resulted in different bind-
ing profiles toward the phorbol BD. As reported before, 2 showed
very strong inhibition of [3H]PDBu binding (Ki = 43 nM), while
compounds 4a and 4b with small substituents R5 on the inositol
moiety showed modest inhibition, with Ki values of 2.34 lM and
598 nM, respectively. These results indicated that the existence
of the myo-inositol moiety affected the binding affinity for the
phorbol BD, nevertheless 4a and 4b still showed significant bind-
ing. In contrast, 4c–4f showed weaker inhibition even at high con-
centration (Ki >20 lM), indicating that these derivatives possessed
negligible binding affinity for the phorbol BD. Because the binding
ability of prenyl derivative 4c and benzyl derivatives 4d–4f was
dramatically reduced compared to that of 4a and 4b, the size of
R5 seems to be critical for the inhibition of [3H]PDBu binding, as
we expected.
Finally, PKCa activator potency was assessed by comparison of
As mentioned above, at least one of the phorbol BD is the C1B
domain. Thus these results of binding assay suggest that 4a and
the phosphorylation levels in the absence and presence of IB deriv-
atives 4a–4f, using
a
fluorescence-labeled peptide substrate
A
Inositol
Inositol
L125
IB
IB
L125
P112
P112
L121
T113
Gly124
L122
IB
PKCαC1B
Inositol
B
PS Membrane
Ca2+
Steric
Repulsion
C1A
C1B
C1B
Steric
Repulsion
C2
C2
C1A
Catalytic
Catalytic
Active
Pseudo-substrate
Figure 3. (A) Binding model of compound 4b with PKC
a (left: ribbon model; right: surface model). (B) Hypothetical model for the decrease of binding affinity of IB derivatives
with bulky substituents to phorbol ester binding site and for origin of their activation potency.

3590
G. Hirai et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3587–3590
(Fig. 2B).8 As expected, all the new IB derivatives were found to in-
duce PKC activation, suggesting that the inositol moiety of IB
derivatives does not disturb the stabilization of the active con-
former of PKC induced by IB derivatives. Importantly, although
compounds 4c–4f showed negligible inhibition of [3H]PDBu bind-
ing to PKC , these derivatives were able to activate PKC . These
facts indicate that 4c–4f would stabilize the active conformer by
the binding to a particular site on PKC other than phorbol BD.
Although further investigation should be required, we speculate
that these molecules would bind to C1A domain and activate PKC
in view of the DAG-based structure of these molecules and Cho’s
report (Fig. 3B).18 The PKC
activation levels in the presence of
M 4c–4f were comparable or slightly lower than that with the
2. (a) Kishi, Y.; Rando, R. R. Acc. Chem. Res. 1998, 31, 163; (b) Castagna, M.; Takai,
Y.; Kaibuchi, K.; Sano, K.; Kikkawa, U.; Nishizuka, Y. J. Biol. Chem. 1982, 257,
7847; (c) Fujiki, H.; Tanaka, Y.; Miyake, R.; Kikkawa, U.; Nishizuka, Y.;
Sugimura, T. Biochem. Biophys. Res. Commun. 1984, 120, 339; (d) Smith, J. B.;
Smith, L.; Pettit, G. R. Biochem. Biophys. Res. Commun. 1985, 132, 939; (e)
Berkow, R. L.; Kraft, A. S. Biochem. Biophys. Res. Commun. 1985, 131, 1109.
3. Mackay, H. J.; Twelves, C. J. Nat. Rev. Cancer 2007, 7, 554.
4. (a) Etcheberrigaray, R.; Tan, M.; Dewachter, I.; Kuipéri, C.; Van der Auwera, I. V.;
Wera, S.; Qiao, L.; Bank, B.; Nelson, T. J.; Kozikowski, A. P.; Van Leuven, F.;
Alkon, D. L. PNAS 2004, 101, 11141; (b) Khan, T. K.; Nelson, T. J.; Verma, V. A.;
Wender, P. A.; Alkon, D. L. Neurobiol. Dis. 2009, 34, 332.
5. Selected recent reports: (a) Nakagawa, Y.; Yanagita, R. C.; Hamada, N.;
Murakami, A.; Takahashi, H.; Saito, N.; Nagai, H.; Irie, K. J. Am. Chem. Soc.
2009, 131, 7573; (b) Wender, P. A.; Verma, V. A. Org. Lett. 2008, 10, 3331; (c)
Keck, G. E.; Poudel, Y. B.; Rudra, A.; Stephens, J. C.; Kedei, N.; Lewin, N. E.; Peach,
M. L.; Blumberg, P. M. Angew. Chem., Int. Ed. 2010, 49, 4580; (d) Mach, U. R.;
Lewin, N. E.; Blumberg, P. M.; Kozikowski, A. P. ChemMedChem 2006, 1, 307; (e)
Bertolini, T. M.; Giorgione, J.; Harvey, D. F.; Newton, A. C. J. Org. Chem. 2003, 68,
5028.
6. Selected recent reports: (a) Raifman, O.; Kolusheva, S.; Comin, M. J.; Kedei, N.;
Lewin, N. E.; Blumberg, P. M.; Marquez, V. E.; Jelinek, R. FEBS J. 2010, 277, 233;
(b) Comin, M. J.; Czifra, G.; Kedei, N.; Telek, A.; Lewin, N. E.; Kolusheva, S.;
Velasquez, J. F.; Kobylarz, R.; Jelinek, R.; Blumberg, P. M.; Marquez, V. E. J. Med.
Chem. 2009, 52, 3274.
7. (a) Baba, Y.; Ogoshi, Y.; Hirai, G.; Yanagisawa, T.; Nagamatsu, K.; Mayumi, S.;
Hashimoto, Y.; Sodeoka, M. Bioorg. Med. Chem. Lett. 2004, 14, 2963; (b) Baba, Y.;
Mayumi, S.; Hirai, G.; Kawasaki, H.; Ogoshi, Y.; Yanagisawa, T.; Hashimoto, Y.;
Sodeoka, M. Bioorg. Med. Chem. Lett. 2004, 14, 2969.
8. Hirai, G.; Shimizu, T.; Watanabe, T.; Ogoshi, Y.; Ohkubo, M.; Sodeoka, M.
ChemMedChem 2007, 2, 1006.
9. Hirai, G.; Ogoshi, Y.; Ohkubo, M.; Tamura, Y.; Watanabe, T.; Shimizu, T.;
Sodeoka, M. Tetrahedron Lett. 2009, 50, 3609.
10. Ananthanarayanan, B.; Stahelin, R. V.; Digman, M. A.; Cho, W. J. Biol. Chem.
2003, 278, 46886.
a
a
a
a
a
a
a
1
l
strong phorbol BD binder 2. However, they were much higher than
that with 3, and 4e was found to be the most potent among the
synthesized compounds. Thus, compound 4e should be useful as
a probe molecule to elucidate the mechanism of PKC
In conclusion, we have developed a new class of PKC
which do not show binding affinity for the phorbol BD. Few PKC
a
a
activation.
activators
a
activators with selective binding affinity for the C1A domain are
known,11,19 and therefore these new activators are expected to
be useful tools in PKCa research if these molecules selectively bind
to C1A domain. In particular, because there is still controversial
over the binding selectivity of the physiological ligand DAG and
the strong tumor promotor phorbol ester to the C1A and C1B do-
mains, it is of interest to elucidate the binding site of 4e to under-
stand the physiological and non-physiological activation modes of
11. (a) Irie, K.; Nakagawa, Y.; Ohigashi, H. Chem. Rec. 2005, 5, 185; (b) Pu, Y.; Perry,
N. A.; Yang, D.; Lewin, N. E.; Kedei, N.; Braun, D. C.; Choi, S. H.; Blumberg, P. M.;
Garfield, S. H.; Stone, J. C.; Duan, D.; Marquez, V. E. J. Biol. Chem. 2005, 280,
27329.
PKCa. Efforts along this line are in progress.
12. Riley, A. M.; Jenkins, D. J.; Potter, B. V. L. Carbohydr. Res. 1998, 314, 277.
13. Murty, K. V. S. N.; Vasella, A. Helv. Chim. Acta 2001, 84, 939.
14. Lewin, N. E.; Blumberg, P. M. Methods Mol. Biol. 2003, 233, 129.
15. (a) Conesa-Zamora, P.; Gómez-Fernández, J. C.; Corbalán-García, S. Biochim.
Biophys. Acta 2000, 1487, 246; (b) Bittova, L.; Stahelin, R. V.; Cho, W. J. Biol.
Chem. 2001, 276, 4218; (c) Stahelin, R. V.; Wang, J.; Blatner, N. R.; Rafter, J. D.;
Murray, D.; Cho, W. J. Biol. Chem. 2005, 280, 36452; (d) Slater, S. J.; Seiz, J. L.;
Cook, A. C.; Buzas, C. J.; Malinowski, S. A.; Kershner, J. L.; Stagliano, B. A.;
Stubbs, C. D. J. Biol. Chem. 2002, 277, 15277.
Acknowledgments
This work was supported in part by RIKEN project funding and a
Grant-in-Aid for Scientific Research (C).
Supplementary data
16. The structure of PKCa has been proposed. See: Srinivasan, N.; Bax, B.; Blundell,
Supplementary data (1H NMR and HRMS data of compounds
4a–f and methods for the evaluation of the biochemical activity
of compounds) associated with this article can be found, in the on-
line version, at doi:10.1016/j.bmcl.2011.04.108.
T. L.; Parker, P. J. Proteins: Struct. Funct. Genet. 1996, 26, 217.
17. Recently crystal structure of large part of PKC bII (C1B, C2, and catalytic
domain) was reported, and interesting activation mechanism was also
proposed. In this structure, it was indicated that C1B domain would be
_
surrounded by C1A, C2, and catalytic domain. See: Leonard, T. A.; Rózycki, B.;
Saidi, L. F.; Hummer, G.; Hurley, J. H. Cell 2011, 144, 55–66.
18. As far as we know, there is no direct method to estimate the binding affinity
References and notes
toward C1A domain using whole PKCa.
19. Selected Paper: Irie, K.; Masuda, A.; Shindo, M.; Nakagawa, Y.; Ohigashi, H.
1. (a) Nishizuka, Y. Science 1992, 258, 607; (b) Yang, C.; Kazanietz, M. G. Trends
Pharmacol. Sci. 2003, 24, 602; (c) Newton, A. C. Chem. Rev. 2001, 101, 2353.
Bioorg. Med. Chem. 2004, 12, 4575.