The EPAC (exchange protein directly activated by cAMP) proteins are GEFs (guanine nucleotide-exchange factors) that activate Rap GTPases upon binding to cAMP. The involvement of these proteins in a number of diseases, neurodegenerative, inflammatory and metabolic, has started to show how they may prove to be important targets for therapeutic intervention. We first became interested in EPAC when we discovered that the expression levels of both EPAC1 and EPAC2 were altered in those regions of the brain associated with Alzheimer's disease [McPhee, Breslin, Kewney, MacKenzie, Cooreman, Gibson and Hammond (2004) International Patent number WO 2004/096199 A2]. It was known that compounds could be designed to be selective for EPAC over PKA (protein kinase A); however, these compounds were all based around the core structure of cAMP. We decided to screen a small compound library (10000 compounds) to investigate the possibility of developing a compound series outside of the cAMP structure. We subsequently developed a novel, high-throughput screen based on the displacement of [3H]cAMP from the EPAC cAMP-binding site and identified small molecule hits from the Scottish Biomedical Lead Generation Library. These compounds selectively bind to the cAMP-binding sites of EPAC1 and EPAC2 and are structurally dissimilar to cAMP. They have similar affinities for both EPAC1 and EPAC2 and have a high degree of specificity for EPAC over PKA. We believe that these compounds provide a valuable starting point for a drug optimization programme.

Introduction

Although the effects of cAMP-mediated signalling are wide ranging, there are only two general cAMP effectors, i.e. proteins that are activated by binding to cAMP directly [1,2]. The ion channels that are activated by binding to cAMP are restricted to certain specialized cells [3]; only PKA (protein kinase A) and EPACs (exchange proteins directly activated by cAMP) have wide ranging expression patterns and, between them, are responsible for mediating the vast majority of cAMP responses. PKA and its targeting through association with AKAPs (A-kinase anchoring proteins) has been extensively studied and the links with important cell functions well established [4]. EPACs were discovered more recently [1,2] and, like PKA, have been found to be important cAMP effector molecules, implicated in a number of cell functions [5,6], some of which had previously been attributed solely to PKA. There are two EPAC proteins, EPAC1 and EPAC2, both of which function as cAMP-mediated GEFs (guanine nucleotide-exchange factors), activating Rap proteins (Rap1 and Rap2) [7]. At present, these small GTPases are the only known mechanism of mediating the EPAC response.

EPACs and disease

As research into EPACs progresses, their pertinence to certain diseases such as cancer, diabetes and central nervous system disorders is becoming increasingly apparent [811]. We discovered that the expression levels of both EPAC1 and EPAC2 were altered in both the frontal cortex (Figure 1) and hippocampus regions of brains showing Alzheimer's pathology compared with non-diseased control brains. These changes were restricted to those regions of the brain associated with Alzheimer's disease and were not in the cerebellum, a region resistant to this pathology [12]. Although it had been known that altered cAMP signalling plays a role in the progression of the disease [13], no other direct link between Alzheimer's disease and EPAC had been reported at the time. From our preliminary results, it was not clear whether EPAC1 elevation or EPAC2 decrease was an important factor in Alzheimer's pathology. To investigate this, and the potential of EPAC as a drug target, we decided to screen part of our in-house small molecule compound library for EPAC activators and inhibitors. Subsequent studies have discovered additional evidence for this link, showing that specific EPAC activation can affect the processing of the APP (amyloid precursor protein) in neuronal cell lines and primary neurons [10,11]. These studies demonstrated that the amount of soluble cleavage product (APPα) could be enhanced by stimulation of EPAC. However, these studies were unable to eliminate the role of EPAC inhibitors in this process since no EPAC-specific inhibitors were available and RP-cAMP, an inhibitor of both PKA and EPAC, also elevated APPα levels [10].

Northern blots for EPAC1 and EPAC2 on the frontal cortex total RNA obtained from control brains (Cont) (n=3) and Alzheimer's brains (Alz) (n=4)

Figure 1
Northern blots for EPAC1 and EPAC2 on the frontal cortex total RNA obtained from control brains (Cont) (n=3) and Alzheimer's brains (Alz) (n=4)

An EPAC1-positive control (+ve) kidney total RNA was also included.

Figure 1
Northern blots for EPAC1 and EPAC2 on the frontal cortex total RNA obtained from control brains (Cont) (n=3) and Alzheimer's brains (Alz) (n=4)

An EPAC1-positive control (+ve) kidney total RNA was also included.

EPAC drug discovery

There are sufficient differences between the PKA and EPAC cAMP-binding domains to obtain EPAC-specific activators, all of which are cAMP-based structures [14,15]. To investigate the feasibility of developing EPAC-specific compounds without a cAMP core and develop SAR (structure–activity relationship) around novel, non-nucleoside-based hits, we screened approx. 10000 compounds from the Scottish Biomedical Lead Generation Library. We developed a competition-based high-throughput screen using the cAMP-binding region of EPAC. In this way, any compound capable of binding to the EPAC cAMP-binding site would prevent [3H]cAMP from binding and could be identified by a reduction in the radioactivity of the EPAC fraction. In this way, the compounds identified as hits had the potential to be either activators or inhibitors of EPAC. This assay identified molecule hits that selectively bound to the cAMP-binding site of EPAC but were structurally dissimilar to cAMP. When tested for specificity, we found that they have similar affinities for both EPAC1 and EPAC2 and have a high degree of specificity for EPAC over PKA (Table 1). Although the compounds showed very little specificity for one or the other EPAC, SB 40308 was approx. 5-fold more selective for EPAC1 over EPAC2 and SB 50510 was approx. 3-fold more selective for EPAC2 over EPAC1. This suggests the potential for developing drugs specifically targeting each EPAC.

Table 1
Concentrations of cAMP, 8pCPT-2Me and lead compounds isolated from the Scottish Biomedical Library required to displace 50% of the [3H]cAMP from the EPAC cAMP-binding domain or the PKA domain

NI, no inhibition; ND, not done; 8pCPT-2Me, 8-(4-chlorophenylthio)-2′-O-methyladenosine 3′,5′-cAMP.

Concentration required to displace 50% [3H]cAMP (μM)
EPAC1 (n=2)EPAC2 (n=2)PKA (n=3)
SB 40308 4.3 19.3 NI 
SB 43001 1.5 0.9 70 
SB 43234 16.2 9.35 100 
SB 43636 41 66 NI 
SB 45217 16 24.8 202 
SB 45221 12.3 12.1 NI 
SB 45682 162 284 NI 
SB 45900 16 14 NI 
SB 46193 23 30.5 113 
SB 46779 134 
SB 47592 15 11.9 NI 
SB 48329 25 28.5 NI 
SB 49821 20 26.9 NI 
SB 51050 14 5.25 NI 
SB 51162 53 26.5 NI 
8pCPT-2Me 3.5 3.4 ND 
cAMP 4.4 
Concentration required to displace 50% [3H]cAMP (μM)
EPAC1 (n=2)EPAC2 (n=2)PKA (n=3)
SB 40308 4.3 19.3 NI 
SB 43001 1.5 0.9 70 
SB 43234 16.2 9.35 100 
SB 43636 41 66 NI 
SB 45217 16 24.8 202 
SB 45221 12.3 12.1 NI 
SB 45682 162 284 NI 
SB 45900 16 14 NI 
SB 46193 23 30.5 113 
SB 46779 134 
SB 47592 15 11.9 NI 
SB 48329 25 28.5 NI 
SB 49821 20 26.9 NI 
SB 51050 14 5.25 NI 
SB 51162 53 26.5 NI 
8pCPT-2Me 3.5 3.4 ND 
cAMP 4.4 

Conclusion

As described above, our strategy towards identifying compounds that bind to EPAC and modulate the EPAC-specific activation of Rap has successfully provided a number of novel small molecule hits. While we have discounted a number of ‘false-positive’ hits (structures with non-selective/promiscuous motifs), we were pleased to recognize a small number of compounds that, we believe, are true hits (based on SAR and prior knowledge of EPAC/cAMP-binding domains). Within these compounds, we found functional activators and inhibitors of EPAC-modulated Rap activation.

From our initial investigation of the EPAC-positive modulation hit compounds, we have been able to develop an understanding of the ligand structural requirements for EPAC activation and have identified a pharmacophoric model for EPAC agonism. The current hit structures have good potential for medicinal chemistry development, obey Lipinski's rule of 5 and are non-cAMP (structure)-based EPAC activators, which we believe is key to developing selective and efficacious EPAC ligands as potential clinical candidates. These compounds are very useful as starting points for a chemical optimization programme and potentially useful as pharmacological tools to elucidate the roles of EPAC modulation further. Similarly, studies of the SAR around the EPAC inhibitors discovered by Scottish Biomedical have helped to illuminate a binding mode for EPAC inhibitors. These non-cAMP (structure)-based compounds are also useful as pharmacological tools and particularly as starting points for a chemical optimization programme.

Cellular Information Processing: A Focus Topic at BioScience2005, held at SECC Glasgow, U.K., 17–21 July 2005. Edited by F. Antoni (Edinburgh, U.K.), C. Cooper (Essex, U.K.), M. Cousin (Edinburgh, U.K.), A. Morgan (Liverpool, U.K.), M. Murphy (Cambridge, U.K.), S. Pyne (Strathclyde, U.K.) and M. Wakelam (Birmingham, U.K.).

Abbreviations

     
  • APP

    amyloid precursor protein

  •  
  • EPAC

    exchange protein directly activated by cAMP

  •  
  • PKA

    protein kinase A

  •  
  • SAR

    structure–activity relationship

References

References
1
de Rooij
 
J.
Zwartkruis
 
F.J.
Verheijen
 
M.H.
Cool
 
R.H.
Nijman
 
S.M.
Wittinghofer
 
A.
Bos
 
J.L.
 
Nature (London)
1998
, vol. 
396
 (pg. 
474
-
477
)
2
Kawasaki
 
H.
Springett
 
G.M.
Mochizuki
 
N.
Toki
 
S.
Nakaya
 
M.
Matsuda
 
M.
Housman
 
D.E.
Graybiel
 
A.M.
 
Science
1998
, vol. 
282
 (pg. 
2275
-
2279
)
3
Kaupp
 
U.B.
Seifert
 
R.
 
Physiol. Rev.
2002
, vol. 
82
 (pg. 
769
-
824
)
4
Wong
 
W.
Scott
 
J.D.
 
Nat. Rev. Mol. Cell Biol.
2004
, vol. 
5
 (pg. 
959
-
970
)
5
Bos
 
J.L.
 
Nat. Rev. Mol. Cell Biol.
2003
, vol. 
4
 (pg. 
733
-
738
)
6
Springett
 
G.M.
Kawasaki
 
H.
Spriggs
 
D.R.
 
BioEssays
2004
, vol. 
26
 (pg. 
730
-
738
)
7
de Rooij
 
J.
Rehmann
 
H.
van Triest
 
M.
Cool
 
R.H.
Wittinghofer
 
A.
Bos
 
J.L.
 
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
20829
-
20836
)
8
Smith
 
P.G.
Wang
 
F.
Wilkinson
 
K.N.
Savage
 
K.J.
Klein
 
U.
Neuberg
 
D.S.
Bollag
 
G.
Shipp
 
M.A.
Aguiar
 
R.C.
 
Blood
2005
, vol. 
105
 (pg. 
308
-
316
)
9
Holz
 
G.G.
 
Diabetes
2004
, vol. 
53
 (pg. 
5
-
13
)
10
Maillet
 
M.
Robert
 
S.J.
Cacquevel
 
M.
Gastineau
 
M.
Vivien
 
D.
Bertoglio
 
J.
Zugaza
 
J.L.
Fischmeister
 
R.
Lezoualc'h
 
F.
 
Nat. Cell Biol.
2003
, vol. 
5
 (pg. 
633
-
639
)
11
Robert
 
S.
Maillet
 
M.
Morel
 
E.
Launay
 
J.M.
Fischmeister
 
R.
Mercken
 
L.
Lezoualc'h
 
F.
 
FEBS Lett.
2005
, vol. 
579
 (pg. 
1136
-
1142
)
12
McPhee
 
I.
Breslin
 
C.
Kewney
 
J.P.
MacKenzie
 
S.J.
Cooreman
 
A.
Gibson
 
L.C.D.
Hammond
 
S.
 
International Patent number
WO 2004/096199 A2, 
2004
13
Fowler
 
C.J.
Cowburn
 
R.F.
Garlind
 
A.
Winblad
 
B.
O'Neill
 
C.
 
Mol. Cell. Biochem.
1995
, vol. 
149–150
 (pg. 
287
-
292
)
14
Christensen
 
A.E.
Selheim
 
F.
de Rooij
 
J.
Dremier
 
S.
Schwede
 
F.
Dao
 
K.K.
Martinez
 
A.
Maenhaut
 
C.
Bos
 
J.L.
Genieser
 
H.G.
, et al 
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
35394
-
35402
)
15
Rehmann
 
H.
Schwede
 
F.
Doskeland
 
S.O.
Wittinghofer
 
A.
Bos
 
J.L.
 
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
38548
-
38556
)