The identification of a novel lead class for phosphodiesterase 2 inhibition by fragment-based drug design

Article abstract of DOI:10.1016/j.bmcl.2017.10.054

We have identified a novel PDE2 inhibitor series using fragment-based screening. Pyrazolopyrimidine fragment 1, while possessing weak potency (K_i = 22.4 μM), exhibited good binding efficiencies (LBE = 0.49, LLE = 4.48) to serve as a start for structure-based drug design. With the assistance of molecular modeling and X-ray crystallography, this fragment was developed into a series of potent PDE2 inhibitors with good physicochemical properties. Compound 16, a PDE2 selective inhibitor, was identified that exhibited favorable rat pharmacokinetic properties.

Full text of DOI:10.1016/j.bmcl.2017.10.054

Accepted Manuscript
The Identification of a Novel Lead Class for Phosphodiesterase 2 Inhibition by
Fragment-Based Drug Design
Ashley B. Forster, Pravien Abeywickrema, Jaime Bunda, Christopher D. Cox,
Tamara D. Cabalu, Melissa Egbertson, John Fay, Krista Getty, Dawn Hall,
Maria Kornienko, Jun Lu, Gopal Parthasarathy, John Reid, Sujata Sharma,
William D. Shipe, Sean M. Smith, Stephen Soisson, Shawn J. Stachel, Hua-Poo
Su, Deping Wang, Richard Berger
PII:
S0960-894X(17)31053-3
https://doi.org/10.1016/j.bmcl.2017.10.054
BMCL 25383
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
25 July 2017
18 October 2017
22 October 2017
Please cite this article as: Forster, A.B., Abeywickrema, P., Bunda, J., Cox, C.D., Cabalu, T.D., Egbertson, M., Fay,
J., Getty, K., Hall, D., Kornienko, M., Lu, J., Parthasarathy, G., Reid, J., Sharma, S., Shipe, W.D., Smith, S.M.,
Soisson, S., Stachel, S.J., Su, H-P., Wang, D., Berger, R., The Identification of a Novel Lead Class for
Phosphodiesterase 2 Inhibition by Fragment-Based Drug Design, Bioorganic & Medicinal Chemistry Letters (2017),
doi: https://doi.org/10.1016/j.bmcl.2017.10.054
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.

The Identification of a Novel Lead Class for Phosphodiesterase 2 Inhibition by
Fragment-Based Drug Design
Ashley B. Forster,a* Pravien Abeywickrema,b Jaime Bunda,a Christopher D. Cox,a Tamara D.
Cabalu,c Melissa Egbertson,a John Fay,d Krista Getty,e Dawn Hall,b Maria Kornienko,b Jun Lu,f
Gopal Parthasarathy,f John Reid,f Sujata Sharma,b William D. Shipe,a Sean M. Smith,g Stephen
Soisson,f Shawn J. Stachel,a Hua-Poo Su,f Deping Wang,h and Richard Bergera
aDiscovery Chemistry, MRL, 770 Sumneytown Pike, West Point, PA 19486, USA
bTarget Protein Design, MRL, 770 Sumneytown Pike, West Point, PA 19486, USA
cPPDM Preclinical ADME, MRL, 770 Sumneytown Pike, West Point, PA 19486, USA
d Pharmacology, MRL, 770 Sumneytown Pike, West Point, PA 19486, USA
e Assay Development, Protein Science, 770 Sumneytown Pike, MRL, West Point, PA 19486, USA
f Structural Chemistry, MRL, 770 Sumneytown Pike, West Point, PA 19486, USA
g Neuroscience, MRL, 770 Sumneytown Pike, West Point, PA 19486, USA
h Chemical Modeling & Informatics, MRL, 770 Sumneytown Pike, West Point, PA 19486, USA
*Corresponding Author
E-mail: Ashley_nomland@merck.com
Phone: 215-652-6572
KEY WORDS: PDE2, Phosphodiesterase 2, fragment-based screening, structure-based drug design
ABSTRACT: We have identified a novel PDE2 inhibitor series using fragment-based screening. Pyrazolopyrimidine
fragment 1, while possessing weak potency (Ki = 22.4 µM), exhibited good binding efficiencies (LBE = 0.49, LLE = 4.48)
to serve as a start for structure-based drug design. With the assistance of molecular modeling and X-ray crystallography,
this fragment was developed into a series of potent PDE2 inhibitors with good physicochemical properties. Compound 16,
a PDE2 selective inhibitor, was identified that exhibited favorable rat pharmacokinetic properties.
CF3
NH
NH
N
N
N
N
N
N
H
Cl
N
Cl
N
1
16
PDE2A Ki = 22400 nM
LBE, LLE = 0.49, 4.48
PDE2A Ki = 14.3 nM
LBE, LLE = 0.45, 4.02
Alzheimer’s Disease (AD) is a progressive, neurodegenerative disorder affecting an estimated 5.2 million people
nationwide and is the sixth leading cause of death in the US.1,2 Symptoms of AD include decline in cognitive functions
such as memory, speech and learning, and eventually lead to the patient’s inability to perform routine daily tasks. Current
approved therapies provide only temporary symptomatic treatment of AD. No agents are available that cure, prevent or
slow disease progression.1 It is estimated that by 2025 approximately 13.8 million people will be afflicted with
Alzheimer’s Disease, indicating there is a severe unmet medical need.1,2

The cyclic nucleotides, cAMP and cGMP, are secondary messengers which mediate the release of neurotransmitters
implicated in the regulation of normal cognitive function.3-6 Signaling within these neurotransmitter pathways, which
heighten synaptic plasticity, learning, and memory is impaired in patients with Alzheimer’s Disease.1,4,7,8 The
phosphodiesterases (PDEs) are a superfamily of phosphohydrolases that catalyze the hydrolysis of the 3’-cyclic phosphate
bonds of cAMP and cGMP, which is the primary mechanism for their inactivation.3,4,7 To date, 11 different PDE gene
families have been identified which differ broadly in cellular function, location and affinity for cAMP and cGMP.3-5 The
PDE2 isoform is a dual substrate that hydrolyzes both cAMP (Km = 30 µM) and cGMP (Km = 10 µM),3,4,9 and is highly
expressed in the hippocampus and frontal cortex,4 regions of the brain which are associated with cognition, learning and
memory. 6-8 It has been demonstrated that pharmacological inhibition of PDE2 enhances cGMP and cAMP signaling
within the neurotransmitter pathways involved in cognition.4,6 In addition, preclinical studies have shown that PDE2
inhibition improves cognition in a variety of rodent behavioral models relevant to AD.4,6-8,10 Due to the low homology
that exists between the catalytic domain of the different PDE isoforms, ranging from 20 – 50%,4 there is a strong
possibility for the identification of a PDE2 selective inhibitor. For this reason PDE2 has been targeted as a treatment for
AD-related cognitive deficits.
Figure 1. BAY60-75507,11 and PF-05180999.7,12
PDE2 inhibitors such as BAY60-7550 and PF-05180999 have previously been disclosed in the literature (Fig. 1). 6-8,11,12
In an effort to identify novel chemical matter differentiated from the current landscape, a fragment-based screening (FBS)
campaign was initiated in parallel with a traditional high throughput screen (HTS). The concept of fragment-based
screening is to identify small molecules that, while displaying weak binding, exhibit this through high binding efficiencies
(LBE and LLE).13 These small fragments also possess more drug-like physicochemical properties than larger and more
potent HTS leads.5,14-16 Such attributes provide an improved starting point for the lead optimization process and increase
the probability that these properties could be incorporated into more potent molecules if carefully maintained through the
drug discovery process.15-18
Biochemical (High Concentration Screening [HCS]) and biophysical (Surface Plasmon Resonance [SPR]) methods
were used to screen our proprietary fragment library19 for PDE2 inhibitors.14-16 A total of 54 structurally distinct hits with
PDE2 Ki and Kd < 200 µΜ were identified and importantly were found to exhibit competitive binding with known PDE2
inhibitors by SPR. The pyrazolopyrimidine fragment 1 was selected for SAR development based on favorable
physicochemical properties and high binding efficiencies (Table 1).
In order to efficiently progress a fragment lead, it is crucial that crystallographic data be available to optimize the
compound using structure-based drug design. Initially, a representative inhibitor-bound structure was solved for BAY60-
7550 (Fig. 2).7,10 Analysis of internal and published7,10 structures observed key ligand interactions and pockets of the
PDE2 active site. The heterocyclic core of BAY60-7550 is anchored via π-stacking with F862 and F830 residues (A and
B)7,10,20 and the pyrimidone functionality interacts through hydrogen bonds with Q859 and Q812, the latter of which is

unique to PDE2 (C), while Q859 is conserved among PDEs.10,20 Interestingly, the propylphenyl group is accommodated
through a novel hydrophobic induced binding pocket (D) located under L7707,10 BAY60-7550 also interacts with the
surface hydrophobic patch (E).10 It has been previously demonstrated that PDE2 inhibitory potency can be enhanced
through all of these interactions, however selective inhibitors have relied on interactions with PDE2 specific Q812 and
also through occupying the hydrophobic induced binding pocket.10,20
Figure 2. Two views of crystal structure with reference compound BAY60-7550, exemplifying key interactions with the
active site. (4HTX PDB)
Figure 3. Crystal structure overlay of fragment 1 (orange) with BAY60-6550 (yellow). (6B96 PDB, 4HTX PDB)
Similar to BAY60-6550, the ligand bound PDE2 crystal structure of 1 shows the pyrazolopyrimidine ring anchored in
the cAMP binding pocket through a π-stacking interaction with F862 and F830. The structure also showed compound 1

has a water-mediated hydrogen bond to residue Q812 but does not directly or indirectly interact with Q859 (C) (Fig. 3).
Examination of the X-ray overlay of 1 and BAY60-6550 also indicated a potential for extending from the aminomethyl
moiety into the unoccupied hydrophobic pocket (D). To investigate this possibility, a para-chloro benzyl group was added
(2). This modification resulted in a 3-fold increase in potency in support of this hypothesis (Table 1). Based on this
result, further SAR development of the fragment hit 1 was explored. An efficient route was devised for rapid analogues of
the 4-position to investigate binding from the commercially available 4,6-dichloropyrazolopyrimidines (Scheme 1)
Primary and secondary amines were introduced via SNAR reaction to afford compounds 1 – 12 and 16. Modifications to
the Cl substitution pattern of the R1 group identified the para substitution (2) as optimal. Further SAR investigations
demonstrated that the p-trifluoromethyl substituent (5), afforded an additional 3-fold improvement in potency over 2 with
comparable LBE, LLE and solubility. Incorporation of the 3-pyridyl moiety (6), by analogy to PF-05180999, was less
tolerated and led to a 7-fold decrease in potency compared to 5 with concomitantly reduced LBE. To explore the depth of
the induced binding pocket, homologation of the linker to a more flexible phenethyl derivative (7) resulted in a decrease in
potency and significant reduction in LLE.
Scheme 1. (a) RR1NH, TEA, THF, 50°C
Molecular modeling suggested the trajectory of the aromatic ring to the PDE2 hydrophobic induced binding pocket (D)
could be improved through structural modifications to the benzylic methylene. Incorporation of the gem-dimethyl
substitution as in compound 8 resulted in a 10-fold boost in potency over compound 2. Further optimization identified the
(R)-stereoisomer of the mono-methyl analog 9 as the preferred enantiomer, affording a 30-fold improvement in PDE2
inhibitory activity over 2, while the (S)-stereoisomer was equipotent to 2. With this boost in potency, compound 9 shows
enhanced LBE and LLE values despite an increasing AlogP observed with the incorporation of the CF3 group.
Table 1. Benzylic Optimization
PDE2 Ki
(µM)
Solubility
pH6.5 (µM)
Compound
R1
LBE/LLE
0.49/4.48
AlogP98
1.16
Me
1
22.4
6.90
206
145
2
3
0.35/1.75
0.35/1.70
3.41
3.41
7.81
78

4
5
11.7
2.34
0.34/1.52
0.34/1.94
44
3.41
3.69
164
6
7
8
9
15.0
0.29/2.07
0.28/0.99
0.34/1.92
0.38/2.56
176
150
183
159
2.75
4.01
4.27
4.07
10.1
0.642
0.237
Figure 4. Crystal structure overlay of inhibitor 9 (green) with fragment 1 (orange). The heterocycle of inhibitor 9 has
flipped in the hydrophobic clamp relative to 1 and does not interact with either Q859 or Q812. (6B96 PDB, 6B97 PDB)
Having succeeded in identifying a sub-micromolar inhibitor, the co-crystal structure was solved in the PDE2 active site
to confirm the design hypothesis. Quite surprisingly, 9 was found to bind in a 180° orientation relative to fragment hit 1
(Fig. 4). These types of observations are common in fragment design, supported by reports that compounds derived from
fragments often do not maintain the original binding mode,14 underscoring the importance of X-ray crystallography for
structural confirmation and compound design in fragment advancement. In the new structure orientation, while the π-
stacking interaction with F862 and F830 are maintained and the hydrophobic pocket (D) is occupied, there is a lack of any

interaction between 9 and Q859 and Q812, (C) yielding a non-selective inhibitor (Fig. 4, Table 4). While it is speculated
that the interactions in the hydrophobic induced binding pocket contribute to PDE2 inhibitory potency,10 it is a
combination of occupying this pocket and interactions with the glutamine residues that are necessary for PDE isoform
selectivity. Therefore, it was hypothesized that the selectivity profile for 9 could be improved by reestablishing these key
hydrogen bonds along with further optimization of binding in the hydrophobic pocket.
New molecular modeling predictions based on the ring flip suggested that optimization of the trajectory might be
achieved through incorporation of either a cyclic amine (10 and 11)21 or a benzylic cyclopropane (12) (Table 2). Cyclized
derivatives 10 and 11, while potent, possess suboptimal LLE’s due to the significant increases in their AlogP values. The
cyclopropane derivative 12 was found to be superior to 9, affording a 50-fold improvement in potency while maintaining
the favorable LBE, LLE and solubility of fragment hit 1. Based on crystal structures of 9 and 12, conformational analysis
using Freeform22 suggested that the bioactive conformation of 12 is more stable than that of 9 relative to their
global minimum. The difference is largely due to the preferred torsion angle between cyclopropyl and distal phenyl ring
in 12 and supports the observed potency enhancement seen with this analog. The PDE isoform selectivity profile of 12
was also now significantly improved over 9, likely an effect of the optimized interactions in the hydrophobic induced
binding pocket (Table 4).
Table 2. Hydrophobic Pocket Optimization
Solubilit
y (µM)
pH6.5
PDE2 Ki
(µM)
Compound R2
LBE/LLE
0.35/1.83
0.37/2.48
AlogP98
4.99
10
11
0.152
138
0.0972
139
156
4.53
CF3
12
0.00430
0.46/4.33
4.03
NH
While we were able to realize large increases in potency, these were driven mainly by lipophilic interactions as noted
by low LLEs, hence increases in compound AlogP were observed (Table 1, 2). Because high AlogP values have been
linked to compound attrition in the clinic,16-18 we sought to modulate the lipophilic interactions through increasing
compound polarity. Two areas amenable for structural modifications not yet explored were the 6-chloro and 1-methyl
substituents. Modifications to the 6-position of the pyrazolopyrimidines were derived from compound 12 and are
described in Scheme 2. Replacement of the 6-chloro substituent with a hydroxyl (13) or a cyclopropane (14) modestly
reduced the AlogP which led to an overall decrease in LLE compared to 12. Their decrease in potencies however, led to
reduced PDE-selectivity profiles (Table 4). Conversely, the 1,6-dimethyl (15) and the 6-chloro-1-des-methyl (16) analogs

displayed reduced AlogP values, which afforded improved LLE’s and their enhanced potencies contributed to their
respectable PDE off-target selectivity profiles (Table 4).
Scheme 2. (13, R4 = OH) 1M NaOH, Dioxane, 150 °C (14, R4 = cyclopropyl) CyPrB(OH)2, K3PO4, Pd(dppf)Cl2, 95:5
Toluene:H2O, 100 °C (15, R4 = Me) KBF3Me, Cs2CO3, Pd(dppf)Cl2, 10:1 THF:H2O, 160 °C
Table 3. Physical Property Optimization
PDE2 Ki
(µM)
Solubility
(µM) pH6.5
Compound
R3
R4
LBE/LLE
AlogP98
1323
14
OH
Me
Me
0.195
0.37/3.25
0.36/3.25
175
125
3.46
3.89
0.0196
15
16
Me
Cl
Me
H
0.00549
0.0143
0.45/5.12
0.45/4.02
11024
105
3.14
3.82
Table 4. PDE Isoform Selectivity
Compound
PDE Selectivity Ratiosi
9
12
13
14

15
16
i. Selectivity ratio = (PDEx Ki)/ (PDE2 Ki) Red = 0 – 100-fold selective for PDE2, Orange = 101 – 500-fold selective for PDE2, Green >
500-fold selective for PDE2
Due to their promising activity, selectivity profiles and physical properties, lead molecules 15 and 16 were evaluated
for their pharmacokinetic properties (Table 5). While compound 15 displayed sub-optimal PK, compound 16, lacking both
metabolically labile methyl groups, showed a robust profile exemplified by high bioavailability and low in vivo clearance.
Consistent with the properties of the pyrazolopyrimidine class, compounds 15 and 16 demonstrated good cell permeability
and were not substrates for Pgp efflux, and are therefore anticipated to be highly brain penetrant.
Table 5. Rat Pharmokinetics Data
Rat
Pgp/Papp
(10-6m/s)ii
PDE2 Ki
(nM)
Solubility
(µM) pH6.5
Cl
(mL/min/kg)
AUCN
µM*h*kg/mg
Vd t
(L/kg) (h)
Rat
PPB%
½
Compound
R3
R4
%F
15
16
Me
Cl
Me 5.49
11024
105
44.35
8.44
0.24
6.59
3.78
4.64
1.34
16.9
146
92.6
97.4
0.82/26.7
1.63/26.5
H
14.3
6.33
ii. Rat LLC-MDR1a cells
Based on the promising overall properties of compound 16, a crystal structure of the inhibitor was solved (Fig. 5). As
designed, the benzylic amine reaches into the hydrophobic induced binding pocket (D) and the heterocycle core π-stacks
between F862 (A) and F830 (B). Compound 16 differentiates itself from other inhibitors through the extensive water
network formed with the 1- and 7-nitrogens of the inhibitor with both Q812 and Q859 (C). Further investigations into
improving upon the selectivity profile while balancing the interactions in the hydrophobic pocket of the series will be the
topic of future publications.

Figure 5. Crystal structure of optimized inhibitor 16. (6B98 PDB)
In conclusion, with the aid of molecular modeling and X-ray crystallography, fragment 1 was rapidly developed
into a viable starting point for a lead optimization chemistry effort. Potency and selectivity were achieved through a small
number of analogs (25 singles) by accessing three key interactions with the PDE2 enzyme (A/B, C, D). It has been
demonstrated that X-ray confirmation of structures is important for fragment optimization. In advancing fragment 1, a
large improvement in potency was achieved (>5000-fold, compound 12), while maintaining favorable LLE, LBE,
solubility and AlogP values, as shown with compounds 12, 15 and 16. In addition, compound 16 has shown good oral
pharmacokinetics in rat, further validating the series as a good starting point for lead optimization efforts. Subsequent
efforts to further optimize this structural class will be the subject of future publications.
Acknowlegements: This work was funded by Merck & Co., Inc., Kenilworth, NJ USA. The authors thank the West
Point NMR group for structure elucidation support and Jessica Frie for helpful reviews and insights.
References and Notes
1. Fargo, Keith and Bleiler, Laura. Alzheimer’s Association Report: 2014 Alzheimer’s disease facts and figures.
Alzheimer’s & Dementia 2014, 10, e47-e92.
2. Takizawa, Claire; Thompson, Paul; van Walsem, Anneloes; Faure, Celine; Maier, William C. Epidemiological and
Economic Burden of Alzheimer’s Disease: A Systematic Literature Review of Data across Europe and the United States of
America. Journal of Alzheimer’s Disease 2015, 43, 1271-1284.
3. Maurice, Donald H.; Ke, Hengming; Ahmad, Faiyax; Wang, Yousheng; Chung, Jay; Manganiello, Vincent C. Advances
in targeting cyclic nucleotide phosphodiesterases. Nature Reviews Drug Discovery 2014, 13, 290-314.
4. Schmidt, Christopher J. Phosphodiesterase Inhibitors as Potential Cognition Enhancing Agents. Current Topics in
Medicinal Chemistry 2010, 10, 222-230.
5. Card, Graeme L.; Blasdel, Landy; England, Bruce P.; Zhang, Chao; Suzuki, Yoshihisa; Gillette, Sam; Fong, Daniel;
Ibrahim, Prabha N.; Artis, Dean R.; Bollag, Gideon; Milburn, Michael V.; Kim, Sung-Hou; Schlessinger, Joseph; Zhang,

Kam Y.J. A family of phosphodiesterase inhibitors discovered by cocrystallography and scaffold-based drug design.
Nature Biotechnology 2005, 23, 201-207.
6. Trabanco, Andres A.; Buijnsters, Peter and Rombouts, J.R. Towards selective phosphodiesterase 2A (PDE2A)
inhibitors: a patent review (2010 – present). Expert Opinion of Therapeutic Patents, 2016, 26, 933-946.
7. Gomez, Laurent; Breitenbucher, J. Guy; PDE2 inhibition: Potential for the treatment of cognitive disorders. Bioorganic
& Medicinal Chemistry Letters 2013, 23, 6522-6527.
8. Heckman, P.R.A; Wouters, C.; Prickaerts, J. Phosphodiesterase Inhibitors as a Target for Cognition Enhancement in
Aging and Alzheimer’s Disease: A Translational Overview. Current Pharmaceutical Design, 2015, 21,1-15.
9. Zaccolo, Manuela and Movsesian, Matthew A. cAMP and cGMP Signaling Cross-Talk Circulation Research 2007,
100, 1569-1578.
10. Zhu, Jian; Yang, Qiqi; Dai, Dandan; Huang, Qiang. X-ray Crystal Structure of Phosphodiesterase 2 in Complex with a
Highly Selective, Nanomolar Inhibitor Reveals a binding-Induced Pocket Important for Selectivity. Journal of American
Chemical Society, 2013, 135, 11708-11711.
11. Boess, F.-J.; Grosser, R.; Hendrix, M.; Koenig, G.; Niewoehner, U.; Schauss, D.; Schlemmer, K.-H.; Schreiber, R.;
Van Der Staay F.-J. WO 2,002,050,078
12. Helal, C. J.; Chappie, T. A.; Humphrey, J. M.; Verhoest, P. R.; Yang, E. US 20,120,214,791.
13. LBE = ∆G/N or -RT(ln Ki or ln IC50) / N, where N = number on non-hydrogen atoms. LLE = pIC50 – LogP]
14. Chen, Haijun; Zhou, Xiaobin; Wang, Ailan; Zheng, Yunquan; Gao, Yu; Zhou, Jia. Evolutions in fragment-based drug
design: the deconstruction-reconstruction approach. Drug Discovery Today, 2015, 20, 105-113.
15. Baker, Monya. Fragment-based lead discovery grows up. Nature Reviews Drug Discovery 2013, 12, 5-7.
16. Hartshorn, Michael J.; Murray, Christopher W.; Cleasby, Anne; Frederickson, Martyn; Tickle, Ian J.; Jhoti, Harren.
Fragment-Based Lead Discovery Using X-ray Crystallography. Journal of Medicinal Chemistry, 2005, 48, 403-413.
17. Murray, Christopher W.; Verdonk, Marcel L.; Rees, David C. Experiences in fragment-based drug discovery. Trends
in Pharmacological Sciences, 2012, 33, 224-232.
18. Leeson, Paul D. and St-Gallay, Stephen A. The influence of the ‘organizational factor’ on compound quality in drug
discovery. Nature Reviews Drug Discovery, 2011, 10, 749-765.
19. Key features of this 1940 compound library are low molecular weight (< 250), an AlogP < 3, H-bond count < 3 and
the rotatable bond count < 3.
20. Buijnsters, Peter; De Angelis, Meri; Langlois, Xavier; Rombouts, Frederik J.R.; Sanderson, Wendy; Tresadern, Gary;
Ritchie, Alison; Trabanco, Andres A.; VanHoof, Greet; Van Roosbroeck, Yves; Andres, Jose-Ignacio. Structure-Based

Design of a Potent, Selective and Brain Penetrating PDE2 Inhibitor with Demonstrated Target Engagement. ACS
Medicinal Chemistry Letters, 2014, 5, 1049-1053.
21. Compounds 10 and 11 are racemic mixtures.
22. SZYBKI 1.8.0.1: OpenEye Scientific Software, Santa Fe, NM
23. Inhibitory data from previous analogs with this substitution at the 6-position and modeling studies suggest this
compound may have flipped back to the original orientation as seen in the fragment. This theory would support the 45-
fold decrease in potency compared to 12 and lack of PDE selectivity observed.
24. A predicted solubility value was used as solubility at this pH was unable to be measured. Predictive solubility values
for this series trend accurately with observed data