Structure-based design of isoquinoline-5-sulfonamide inhibitors of protein kinase B

Article abstract of DOI:10.1016/j.bmc.2005.09.055

Structure-based drug design of novel isoquinoline-5-sulfonamide inhibitors of PKB as potential antitumour agents was investigated. Constrained pyrrolidine analogues that mimicked the bound conformation of linear prototypes were identified and investigated by co-crystal structure determinations with the related protein PKA. Detailed variation in the binding modes between inhibitors with similar overall conformations was observed. Potent PKB inhibitors from this series inhibited GSK3β phosphorylation in cellular assays, consistent with inhibition of PKB kinase activity in cells.

Full text of DOI:10.1016/j.bmc.2005.09.055

Bioorganic & Medicinal Chemistry 14 (2006) 1255–1273
Structure-based design of isoquinoline-5-sulfonamide inhibitors
of protein kinase B
Ian Collins,a,* John Caldwell,a Tatiana Fonseca,a Alastair Donald,a Vassilios Bavetsias,a
Lisa-Jane K. Hunter,a Michelle D. Garrett,a Martin G. Rowlands,a G. Wynne Aherne,a
Thomas G. Davies,b Valerio Berdini,b Steven J. Woodhead,b Deborah Davis,b
Lisa C. A. Seavers,b Paul G. Wyatt,b Paul Workmana and Edward McDonald a,*
aCancer Research,UK Centre for Cancer Therapeutics, The Institute of Cancer Research, 15 Cotswold Road,
Sutton, Surrey SM2 5NG, UK
bAstex Technology Ltd, Science Park, Milton Road, Cambridge CB4 0QA, UK
Received 10 June 2005; revised 8 September 2005; accepted 23 September 2005
Available online 24 October 2005
Abstract—Structure-based drug design of novel isoquinoline-5-sulfonamide inhibitors of PKB as potential antitumour agents was
investigated. Constrained pyrrolidine analogues that mimicked the bound conformation of linear prototypes were identified and
investigated by co-crystal structure determinations with the related protein PKA. Detailed variation in the binding modes between
inhibitors with similar overall conformations was observed. Potent PKB inhibitors from this series inhibited GSK3b phosphoryla-
tion in cellular assays, consistent with inhibition of PKB kinase activity in cells.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
PKB-b may be more effective than inhibitors of either
isoform alone.6,7
Protein kinase B (PKB, or Akt) is an important protein
involved in the control of cell proliferation and survival,
and is a key downstream component of the phosphati-
dylinositol-3 kinase (PI3K) signalling pathway.1–5 In
response to receptor tyrosine kinase activation by extra-
cellular hormones, PI3K activation in turn produces
phosphatidylinositols anchored to the inner side of the
plasma membrane. Cytosolic PKB binds to these
through a pleckstrin homology (PH) domain, transloca-
ting the protein to the plasma membrane, where it is
activated by phosphorylations on Ser473 and Thr308.
Activated PKB phosphorylates a variety of protein sub-
strates, including, for example, GSK-3b, FKHRL1,
BAD and mTOR, and signals to pathways controlling
cell proliferation, protein translation, cell survival and
apoptosis, and progression through the cell cycle. There
are 3 known isoforms of PKB (a, b and c). Recent in vi-
tro research suggests that dual inhibitors of PKB-a and
There is substantial evidence to suggest that deregula-
tion of signalling on the PI3K pathway, and in particu-
lar through PKB, is intimate to the development of
some human cancers.2,3 Loss of function of the lipid
phosphatase PTEN, which dephosphorylates 30-phos-
phatidylinositols, results in overactivation of signalling
through PKB. PTEN is frequently deleted or mutated
in human tumours, notably glioblastoma, endometrial
and prostate cancers. Amplification and mutation of
the gene encoding the PI3K 110a catalytic subunit are
also frequently observed in human cancers, particularly
ovarian and cervical tumours, resulting in overactiva-
tion of the pathway. The overexpression of Erb2 also
leads to constitutive PKB activity in breast and other tu-
mours. For these reasons, inhibitors that prevent the
phosphorylation of substrates by PKB offer potential
as therapies for human malignancies.8
Blockade of signalling via PKB might be achieved by pre-
venting the phosphorylation, and hence the activation, of
the enzyme by inhibition of upstream kinases such as
PDK1 or through inhibition of the association of the
PH domain of the enzyme with PIs.6,7 Alternatively,
Keywords: PKB; Inhibitors; Cancer; Structure-based drug design.
*
Corresponding authors. Fax: +44 208 722 4205; e-mail addresses:
Ian.Collins@icr.ac.uk; Ted.McDonald@icr.ac.uk
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.09.055

1256
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
the number of unacceptable docking solutions, a phar-
macophoric restraint was constructed26 based on the
available crystal structures of isoquinoline sulfonamides
(H-series) complexed with PKA27 (PDB 1YDR, 1YDS
and 1YDT). The pharmacophoric restraint contained
an acceptor term representing the binding of the iso-
quinoline nitrogen to the hinge region of the protein, a
hydrophobic term centred on the isoquinoline bicycle
and a term, composed of two acceptors and one donor,
to represent the position of the sulfonamide group. The
radius used for individual atoms in the pharmacophore
was 0.7 A. To define the most favourable rotational
conformation of the sulfonamide, an extensive search
of crystal structures from the PDB and CCDC revealed
the conformation observed for H-8 (PDB 1YDS) to be
the global minimum, in which one S@O bond lies paral-
lel to the plane of the aromatic group. The alternative
arrangement seen for H-7 [1-(5-isoquinolinylsulfonyl)-
2-methylpiperazine] (PDB 1YDR), in which the sulfonyl
group is twisted out of the plane of the aromatic ring,
was determined as a local minimum. Once identified,
the global minimum conformation was used to select
preferred docking solutions and for the design of new
molecules.
Figure 1. Known isoquinoline-5-sulfonamide inhibitors of PKA/PKB.
26
˚
direct inhibition of the kinase activity of PKB is possible,
and a number of ATP-competitive inhibitors of PKB have
been identified.9–14 Most recently, preliminary reports
of in vivo efficacy in rodent models have been made for
a series of pyridine-isoquinolines.12,15
As part of a programme to discover and optimise new,
small-molecule, ATP-competitive inhibitors of PKB as
potential cancer therapeutics,16 we have explored com-
pounds based on the isoquinoline-5-sulfonamide kinase
inhibitor template.17 Prior to this study, a combinatorial
chemistry approach was adopted by Levitski et al.11 to
convert the PKA-selective inhibitor H-89 (Fig. 1) to
the PKB/PKA inhibitor NL-71-101. We anticipated that
alternative modifications to the structure of H-89 would
reveal other scaffolds capable of generating PKB inhib-
itors with improved physicochemical properties. In par-
ticular, we sought replacements for the potentially
metabolically labile alkenyl functionalities of H-89 and
NL-71-101, and to reduce the lipophilicities of these li-
gands (ClogP 4.4 and 5.0, respectively). Selectivity for
PKB over other AGC kinases was also a consideration,
although this was anticipated to be a challenge for PKA
since the enzymes have a close sequence (ꢀ68%) homol-
ogy in the kinase domain,18 and within the adenine bind-
ing site itself only three residues differ between PKBb
and PKAa.19 The availability of the X-ray crystal struc-
ture of PKBb20 offered the opportunity to use structure-
based design techniques in pursuit of novel inhibitors.
Such an approach has been taken previously using a
PKA-PKB chimeric protein, to gain insight into the dif-
ferences between the two enzymes.19 As a result, ATP-
competitive inhibitors with selectivity for PKB over
PKA have been developed from templates derived from
the natural product balanol, that exploit selectivity fea-
tures outside the adenine and ribose binding regions.9,10
A structure-based approach, again using balanol deriva-
tives as the starting point, was used to delineate the
determinants of selectivity between PKA and PKC.21
The PKB structure has also been used in a virtual screen
to identify inhibitors of the enzyme.22
A proprietary virtual screening platform28 was used to
generate suggested elaborations of the isoquinoline sul-
fonamide template. Virtual libraries were constructed
from the coupling of isoquinolin-5-yl sulfonyl chloride
and commercially available amines, and the libraries
were virtually screened using the docking methodology
described above with ÔDefault 4Õ Genetic Algorithm set-
tings.24 The selection of compounds for synthesis was
made by analysing the binding mode, the conformation,
the number of hydrogen bonds formed to key residues
of PKB and the value of the scoring function.
3. Synthetic chemistry
Direct mono-sulfonylation of ethane-1,2-diamine with
freshly prepared isoquinoline-5-sulfonyl chloride 129
gave the sulfonamide 2,29,30 which was further elaborat-
ed by reductive amination with 2-benzyloxyacetalde-
hyde to provide the benzyloxyethyl substitution
product 3 (Scheme 1).31 To prepare aryl-substituted ana-
logues 4-13, ethylene glycol was mono-alkylated with
the appropriate benzyl halides 14 and the resulting alco-
hols 15 were oxidised to the aldehydes 16 using Dess–
Martin reagent.32,33 Reductive amination of the crude
aldehydes 16 with 2 furnished the analogues 4–13.
The unsubstituted analogues 1729 and 18 were prepared
straightforwardly by mono-sulfonylation of symmetrical
diamines. Various cyclic diamines were also mono-sulf-
onylated with 1 (Scheme 2). As previously demonstrat-
ed,34 the reaction of unprotected 4-aminopiperidine led
exclusively to sulfonylation of the primary amine to give
19. Similar regioselectivity was seen in the sulfonylations
giving 21,34 2334 and 24.34 The regioisomeric piperidine
sulfonamide 20 was prepared from commercially
available Boc-protected 4-aminopiperidine followed
by deprotection. The unambiguous synthesis of 20
2. PKB modelling and docking studies
Compounds were docked into a binding pocket consist-
˚
ing of all the protein atoms within a 6 A shell around the
AMP–PNP ligand of the available PKB structure20
(PDB 106K). Docking was performed with the program
GOLD,23 using the ÔDefault 3Õ Genetic Algorithm set-
tings,24 and docked conformations were scored using
the Chemscore scoring function.25 In order to reduce

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1257
Scheme 3. Reagents and conditions: (i) PhCHO, NaB(OAc)3H,
˚
ClCH2CH2Cl, AcOH, 4 A molecular sieves, rt; (ii) TFA, CH2Cl2, rt;
(iii) 1, Et3N, CH2Cl2, rt; (iv) DIBAL-H, CH2Cl2, ꢁ78 ꢁC; (v)
1-bromomethyl-4-chlorobenzene, NaH, DMF, rt; (vi) 1-chloroethyl
chloroformate, ClCH2CH2Cl, reflux.
Scheme 1. Reagents and conditions: (i) ethylenediamine, CH2Cl2, rt;
˚
(ii) 17, NaB(OAc)3H, ClCH2CH2Cl, AcOH, 4 A molecular sieves, rt;
(iii) ethylene glycol, NaH, THF, reflux; (iv) Dess–Martin periodinane,
CH2Cl2, rt. R1 3 H, 4 2-Me, 5 3-Me, 6 4-Me, 7 2-Cl, 8 3-Cl, 9 4-Cl, 10
3,4-Cl2, 11 2-OMe, 12 3-OMe, 13 4-OMe.
ly available (2S,4R)- and (2S,4S)-N-Boc-4-aminopyrr-
olidine-2-carboxylate methyl esters 30 and 31 (Scheme
4). Sulfonylation was followed by sulfonamide protec-
tion with the SEM group,37 to permit selective O-benzy-
lation of the alcohols 36 and 37 later in the sequence.
Acidic deprotection of the N-SEM and N-Boc-protect-
ing groups was carried out in a single step to yield 40
and 41.
Scheme 2. Reagents and conditions: (i) diamine, Et3N, CH2Cl2, rt;
(ii) (a) mono-Boc-diamine, Et3N, CH2Cl2, rt; (b) TFA, CH2Cl2, rt.
confirmed the regiochemistry of the sulfonylation lead-
ing to 19. The trans-cyclohexyl-1,4-diamine analogue
22 was also synthesised from the mono-Boc-protected
diamine since attempted mono-sulfonylation of the
unprotected diamine gave an intractable mixture of
products.
The most efficient route for elaboration of the analogue
22 was found to be reductive amination of benzaldehyde
with the mono-Boc diamine, followed by deprotection
and sulfonylation to give 25 (Scheme 3). The prepara-
tion of the 3-substituted piperazine 29 required a new
approach from the known N-benzyl-protected pipera-
zine ester 26.35 Sulfonylation of 26 and complete reduc-
tion of the ester 27 were followed by alkylation of the
alcohol to give 28. Selective removal of the N-benzyl-
protecting group of 28 in the presence of the 4-chlorob-
enzyloxy substituent was achieved through N-activation
using chloroethyl chloroformate.36
Scheme 4. Reagents and conditions: (i) 1, Et3N, DMAP, CH2Cl2, rt;
(ii) NaH, SEMCl, DMF, 0 ꢁC; (iii) LiAlH4, THF, 0 ꢁC; (iv) NaH,
4-chlorobenzyl bromide, DMF, 0 ꢁC; (v) 6 M HCl, MeOH, reflux.
Single diastereoisomer, substituted aminopyrrolidines
related to 23 were prepared separately from commercial-

1258
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
For substituted analogues of the aminomethylpyrroli-
dine 24, the single diastereoisomer hydroxyproline
esters38 42 and 43 were separately O-alkylated and
reduced to the alcohols 44 and 45, respectively (Scheme
5). The alcohols were activated as the mesylates and
displaced with sodium azide to give the azides 46 and
47. Hydrogenation of 46, sulfonylation and deprotec-
tion were carried out without isolation of intermediates
to give the analogue 48. When this sequence was per-
formed on the azide 47, however, some reduction of
the 4-chlorobenzyl substituent was seen in the hydroge-
nation step. The mixture of 4-chloro and des-chloro
compounds was carried through the sulfonylation and
deprotection, and was separated by preparative HPLC
to give compounds 49 and 50.
Scheme 6. Reagents and conditions: (i) (Boc)2O, Et3N, DMAP,
CH2Cl2, rt; (ii) NaBH4, MeOH, 0 ꢁC; (iii) NaH, 4-chlorobenzyl
bromide, DMF, 0 ꢁC; (iv) 1 M HCl in Et2O, MeOH, rt; (v) 1, Et3N,
CH2Cl2, rt.
To prepare a related 2,5-syn-pyrrolidine 55, the known
racemic bicyclic amide 5139 was doubly protected with
the N-tert-butyloxycarbonyl group (Scheme 6). On
treatment of 52 with sodium hydroxide, the activated
bicyclic amide was ring-opened to generate the racemic
2,5-syn-substituted pyrrolidine scaffold.39 Without isola-
tion, the liberated carboxylic acid was reduced with
sodium borohydride to the alcohol 53, which was alkyl-
ated on oxygen to give 54. Complete N-deprotection of
54 gave a diamine that was sulfonylated selectively on
the primary amine to yield the required analogue 55.
To allow the investigation of replacements of the par-
ent isoquinoline-5-sulfonamide in the analogue 41, the
relevant single diastereoisomer diamine side-chain 59
was made (Scheme 7). The protected proline deriva-
tive 56 was prepared from (2S,4R)-4-hydroxyproline
following literature procedures.40 Alkylation of 56
was followed by selective O-deprotection using mild
acid. The secondary alcohol 58 was activated as the
mesylate and displaced, with inversion, by sodium
azide.41 Reduction of the intermediate azide gave
Scheme 7. Reagents and conditions: (i) KH, 4-chlorobenzyl chloride,
THF, 0 ꢁC; (ii) PPTS, EtOH, 55 ꢁC; (iii) methanesulfonic anhydride,
Et3N, DCM, 0 ꢁC; (iv) NaN3, DMF, 80 ꢁC; (v) PPh3, H2O, THF, rt;
(vi) 4-methylisoquinoline-5-sulfonyl chloride, Et3N, DMAP, CH2Cl2,
rt; (vii) 4M HCl, dioxane, rt; (viii) 1-chloroisoquinoline-5-sulfonyl
chloride, Et3N, DMAP, CH2Cl2, rt; (ix) NaH, SEMCl, THF, 0 ꢁC;
(x) KOSiMe3, THF, 80 ꢁC; (xi) 4 M HCl, dioxane, rt.
the desired amine 59, which was subsequently coupled
to 4-methylisoquinoline-5-sulfonyl chloride42 and
deprotected to give 61. Sulfonylation of 59 with 1-
chloroisoquinoline-5-sulfonyl chloride43 and sulfon-
amide N–H protection with the SEM group gave
the sulfonamide 63. Displacement of the iminoyl chlo-
ride from 63 through SNAr reaction with potassium
trimethylsilanolate44 and acidic deprotection of the
pyrrolidine and sulfonamide led to the 1-hydroxyiso-
quinoline analogue 65.
Scheme 5. Reagents and conditions: (i) NaH, 4-chlorobenzyl bromide,
DMF, 0 ꢁC; (ii) LiAlH4, THF, 0 ꢁC; (iii) Ms2O, Et3N, CH2Cl2, 0 ꢁC;
(iv) NaN3, DMF, 80 ꢁC; (v) 10% Pd/C, 1 atm. H2, EtOAc, rt; (vi) 1,
Et3N, DMAP, CH2Cl2, rt; (vii) 1 M HCl in Et2O, MeOH, rt.

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1259
Table 1. Kinase inhibitory activity and cell growth inhibition of 3–13
and 17–24
4. PKA—ligand co-crystallography
Compounda
R1a
PKB IC50 (lM)b
SRB IC50(lM)c
As part of the structure-based design approach, the co-
crystal structures of selected inhibitors with PKA were
determined. The a catalytic subunit of bovine PKA
was expressed and purified with reference to previously
described protocols.27,45 Hanging drops containing
17 mg/ml of tri-phospho PKA, 25 mM Mes–BisTris,
pH 6.5, 75 mM LiCl, 0.1 mM EDTA, 1 mM DTT,
1.5 mM octanoyl-N-methylglucamide and 1 mM
PKI(5–24) were equilibrated at 4 ꢁC against 15% (v/v)
methanol. Crystals of apo PKA appeared overnight
and were soaked for approximately 18 h at 4 ꢁC in a
solution containing 10% PEG 400, 25 mM Mes–BisTris,
pH 6.5, 0.1 mM EDTA, 1 mM DTT, and 10 mM of
inhibitor. The crystals were briefly immersed in a cryo-
protectant containing 22.5% MPD, 25 mM Mes–Bis-
Tris, pH 6.5, 0.1 mM EDTA, and 1 mM DTT, before
plunge-freezing into liquid nitrogen.
3
4
H
1.17 ( 0.18)
0.57
>50d
37 ( 3)
2-Me
3-Me
4-Me
2-Cl
3-Cl
4-Cl
3,4-Cl2
2-MeO
3-MeO
4-MeO
—
e
5
1.1
1.53
e
e
e
6
7
0.96
0.37
8
9
0.26 ( 0.02)
0.21
25 ( 1)
13
10
11
12
13
17
18
19
20
21
22
23
24
H-8
H-89
e
6.6
2.2
>50
>50
2.9
4.8
e
e
e
e
e
e
e
e
e
—
—
23
8.7
—
—
>30
57
—
—
8.1
3.9
—
39
7.6 ( 0.40)
0.59 ( 0.07)
X-ray diffraction data were collected using a Rigaku-
MSC Jupiter CCD mounted on an RU-H3R rotating
anode generator. Data were integrated using MOS-
FLM46 before input into a proprietary procedure for
automated scaling, refinement and ligand fitting.47 The
starting model used for refinement was the previously
solved PKA structure PDB 1YDS,48 with ligand and
waters removed. After ligand fitting, subsequent cycles
of model adjustment and refinement were carried out
using QUANTA (Accelrys Inc.) and REFMAC5.49
Data processing and refinement statistics are given in
the Supplementary Material.
18 ( 1.4)
a See structures in Schemes 1 and 2.
b Mean of two independent IC50 determinations or mean ( SEM) for
n > 2 determinations. Standards H-8 and H-89 gave SD 24%
(n = 20) and 37% (n = 10), respectively, in this assay.
c Mean of two independent IC50 determinations or mean ( SEM) for
n > 2 determinations. Standard H-89 gave SD 34% (n = 20) in this
assay.
d Single determination.
e Not determined
4–6, was tolerated at all positions, and the most active
2-methyl isomer 4 now showed measurable activity in
cells. Chloro substitution, as in compounds 7–9, was
beneficial for both in vitro PKB inhibition and cellular
antiproliferative activity. In contrast, electron-donating
methoxy substituents 11–13 were not advantageous.
5. Results and discussion
Following the observation that the benzhydryl group of
NL-71-101 imparted a degree of selectivity for inhibition
of PKB over PKA,11 we initially attempted to prepare
benzhydryl and close derivatives where the isoquino-
line-5-sulfonamide was replaced by various aromatic
and heteroaromatic groups linked through an amide.
However, this generally led to compounds with poor sol-
ubility that confounded accurate testing in vitro (data
not shown). We sought to improve the solubility of
the ligands by restricting the terminal substituent to a
single aromatic group, as in H-89, while incorporating
a polar heteroatom into the linker. As well as potentially
improving the physicochemical properties, this strategy
also addressed the replacement of the potential metabol-
ic site of the alkene in the linker.
The selectivity of the more active compounds 7–9
against PKA and PKCa was measured (Table 3). The
compounds were 10- to 20-fold selective against PKCa,
but showed equipotent inhibition of PKA. This com-
pares with the 8-fold selectivity for PKA shown by H-
89, the change from the rigid all-carbon 3-atom linker
to the more polar, and more flexible, 4-atom linker hav-
ing overturned this.
Although the 4-chlorophenyl analogue 9 had sufficient
potency to permit further investigations in cellular as-
says (vide infra), improvements in the activity and selec-
tivity of the ligands were sought. The large number of
rotatable bonds (nRot = 10) in these structures was con-
sidered undesirable, since inverse correlations of flexibil-
ity and membrane permeability have been reported,52–54
and this feature might account in part for the low cellu-
lar activity of the compounds relative to their in vitro
enzyme inhibition. In addition, conformational restric-
tion of linear chains can lead to increased potency by
lowering the entropic penalty on binding and may also
provide a potential means of increasing ligand selectivi-
ty.55 To be successful in this series of compounds,
Thus, a novel series of 2-benzyloxyethyl-substituted
analogues 3–13 were prepared16 (Table 1). The parent
phenyl analogue 3 showed good in vitro PKB inhibitory
activity in the radiometric enzyme inhibition assay50 but
no activity in the SRB (sulforhodamine B) cellular anti-
proliferation assay.51 The effects of simple aromatic sub-
stitution were examined and, in general, lipophilic
electron-withdrawing or neutral substituents were
optimal, a structure–activity relationship that parallels
previous observations with analogues leading to
NL-71-101.11 Methyl substitution, as in compounds

1260
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
conformational restriction of the diamine linker must
optimally deliver the terminal lipophilic aryl group,
whilst not disrupting the important hydrogen-bonding
interaction of the central amine. There is evidence from
X-ray structural determinations of isoquinoline ligands
bound to PKA that this interaction has some flexibili-
ty,27 and it is reasonable to expect that PKB might
exhibit similar tolerance in the positioning of the amine.
tion pattern on this scaffold, or on the more
synthetically demanding 4-aminopiperidine 19. Instead,
attention was focussed on the pyrrolidines 23 and 24.
For the 3-aminopyrrolidine constraint, PKB docking
studies and comparison to the structure of H-89 docked
to PKA (PDB code 1YDT) indicated that the 5-position
would provide the best vector for further substitution
(Fig. 2). The calculated free energy difference between
cis and trans isomers using the Chemscore scoring func-
tion was <0.5 kcal molꢁ1, too low to allow any predic-
tion to be drawn. In the event, substitution at this
position was favourable for the (2S,4S) diastereomer
41, which proved significantly more active than either
the unsubstituted parent 23 or the (2S,4R) isomer 40.
Simple cyclic diamines were introduced onto the iso-
quinoline-5-sulfonamide in an attempt to find conform-
ationally restricted scaffolds of equal, or better, potency
than the parent ethylenediamine (H-8) or the propylen-
ediamine homologue 17 (Table 1). Virtual docking stud-
ies were undertaken to further select replacements for
the linear diamines, and additionally to determine the
positions on the cyclic diamine scaffolds from which
elaboration with a 4-chlorobenzyloxy group, or similar,
would be most likely to allow occupation of appropriate
space in the enzyme.
For the (S)-2-aminomethylpyrrolidine constraint 24,
docking indicated the 4-position as suitable for substi-
tution (Fig. 3). Alternative orientations of the pyrroli-
dine ring were identified that suggested a (2R,5R)
substitution pattern could also be accommodated, pro-
vided adjustments were made in the length of the linker
at the 5-position. Some indication of preferred relative
stereochemistry was obtained from the modelling of
the 4-substituted compounds, in favour of the (S,S)
diastereomer 50 over the (2S,4R) diastereisomer 48
(Chemscore DDG 4.8 kcal molꢁ1). This observation
was not used in a proscriptive sense, however, and
both isomers were prepared. In practice, a large
improvement in activity over the unsubsituted parent
24 was observed upon attachment of the side chain
at either the 4-position of the (S)-2-aminomethylpyrr-
olidine, 48–50, or at the 5-position in the alternative
enantiomeric series 55. For the 4-substituted com-
pounds, the (2S,4S) isomer 50 showed 3.5-fold more
potent enzyme inhibition than the (2S,4R) isomer 48.
The importance of the lipophilic chlorine substituent
for high activity, as observed in the linear diamines
(cf. 9 vs 3), was emphasised by the 7-fold drop in activ-
ity between 50 and the des-chloro analogue 49.
Of the unsubstituted analogues prepared, the 4-aminopi-
peridine 19, trans-1,4-diaminocyclohexane 22 and 3-
aminopyrrolidine 23 showed activity equivalent to the
linear scaffolds of H-8 and 17 (Table 1). There was some
divergence between the observed activity of the unsub-
stituted diamines, and the scaffolds predicted by model-
ling to be more promising for substituent addition. For
example, although the parent amine had good activity
against PKB, substitution of the trans-1,4-diaminocyclo-
hexane 22 from the distal amine was predicted to occur
along unfavourable vectors for reaching the lipophilic
binding pocket. A similar problem was anticipated for
the 4-aminopiperidine 20. For the piperizine 18 and 4-
aminopiperidine 19, an apparent need to preserve a free
N–H for optimum hydrogen bonding to the enzyme
in the docking studies suggested that substitution from
the a-carbon would be more productive. In contrast,
the 2-(S)-aminomethylpyrrolidine 24, although of rela-
tively low activity, was predicted to be a competent scaf-
fold for delivering the terminal lipophilic group when
appropriately substituted. 3-Aminopyrrolidine was
identified as a highly suitable scaffold, both from the
modelling and the observation of the inhibitory activity
of compound 23. For the pyrrolidine templates, accurate
prediction of the best relative stereochemistry of substi-
tution was difficult, and both cis and trans-substitution
patterns were considered for synthesis. Although in gen-
eral no distinct preference for the absolute stereochemis-
try of the pyrrolidine fragments was discernible from the
modelling, for the aminomethylpyrrolidine 24 the 2-(S)
enantiomer gave consistently better docking predictions
for both the parent and elaborated derivatives. Only the
(S)-enantiomer of 24 was synthesised and tested.
Inhibition of cell growth was investigated for the more
active pyrrolidine derivatives. Arguably, a limited gen-
eral improvement in cellular penetration was achieved
through the reduction in the number of rotatable
bonds in these analogues (nRot = 7–8), since com-
pounds with biochemical activities of 1–5 lM (40, 41,
48, 50, and 55) showed measurable cellular activity.
This contrasts with the ethylenediamine analogues 3–
13 (Table 1), where biochemical activity below 1 lM
was apparently necessary before measurable cellular
activity was achieved. A caveat to this interpretation
is that the cellular assay is not necessarily a simple inte-
gration of PKB/PKA activity and membrane perme-
ability, and could reflect any other activities of the
compounds. The selectivities of selected constrained
analogues versus PKA and/or PKCa were determined
(Table 3). In general, no significant selectivity was seen
between PKB and PKA activities regardless of the
presence or absence of the lipophilic side chain. In con-
trast, selectivity (ca. 10- to 100-fold) was observed for
PKB/PKA versus PKCa for the fully elaborated pyrr-
olidines 41 and 50. In these respects, constraint of
For the selected scaffolds, addition of appropriately
linked 4-chlorophenyl groups was made (Table 2). As
anticipated from the modelling studies, N-substitution
of the trans-1,4-diaminocyclohexane 22 to give com-
pound 25 did not improve the activity. Substitution a
to nitrogen of the piperazine 18 was investigated to give
compound 29. The marginal increase in activity ob-
served was not encouraging for pursuing this substitu-

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
Table 2. Kinase inhibitory activity and cell growth inhibition of constrained analogues
1261
Compound
R1
R2
PKB IC50 (lM)a
SRB IC50 (lM)b
O2S
R1
N
R2
H
N
c
c
25
29
40
41
48
49
50
55
61
65
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
OH
6.6
HN
Cl
NH
12
N
O
H
N
5.2 ( 0.41)
25
18
O
Cl
HN
H
N
0.78 ( 0.12)
O
Cl
Cl
HN
HN
HN
O
1.7
23
HN
HN
c
O
O
3.4
HN
HN
0.48
1.7
28
21
9.8
40
Cl
H
N
O
HN
Cl
H
N
2.6
O
HN
HN
H
N
Cl
Cl
0.52
O
a Mean of two independent IC50 determinations or mean ( SEM) for n > 2 determinations, see footnotes to Table 1.
b Single determination.
c Not determined.
the diamine linker did not provide a biochemical
advantage over the linear scaffolds, for which similar
selectivity profiles were observed.
oxidation at C-1 of the isoquinoline is known to lead
to a more potent in vivo inhibitor of the rho-kinase en-
zyme.57 However, the 1-hydroxyisoquinoline analogue
65 showed equivalent activity against PKB to the parent
41. No significant selectivity was observed for PKB or
PKA with compound 65.
For one of the most active pyrrolidines 41, a limited
study of conservative changes to the isoquinoline group
was made, inspired by the known structure–activity rela-
tionships of isoquinoline-5-sulfonamide inhibitors of
rho-kinase56 (Table 2). Methylation of the rho-kinase
inhibitor fasudil at C-4 of the isoquinoline has been
shown to increase the affinity of the analogue for rho-ki-
nase, while leaving PKA activity unchanged. For the
compound 61, introduction of the C-4 methyl group
was detrimental to PKB activity, while PKA activity
was unaffected, leading to 8-fold selectivity for PKA
over PKB (Table 3). The metabolism of fasudil by
The binding mode of selected ligands (9 and 48) was
evaluated crystallographically with PKA. PKA provides
a useful surrogate for PKB as the active sites are
approximately 80% identical,18,19 and PKA is a robust
crystallographic system amenable to high-throughput
data collection. Given the general equipotency of this
series of compounds for PKB and PKA, a correspon-
dence between the binding modes to the two proteins
would be expected. The constrained ligands 41, 48 and

1262
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
Table 3. Kinase inhibitory activites of selected compounds for PKB,
PKA and PKCa
Compound
PKB IC50
(lM)a
PKA IC50
(lM)b
PKCa IC50
(lM)c
7
8
0.96
0.37
1.0
16
0.41
0.17
3.9
8.3
1.9
9
0.26 ( 0.02)
8.7
8.1
d
19
22
23
41
48
50
61
65
H-8
H-89
5.7
d
54
19
3.9
0.78 ( 0.12)
1.7
0.48
0.42
0.40
9.9
d
0.17
0.32
28
d
2.6
0.52
d
1.3
5.3 ( 0.53)
0.073
7.6 ( 0.40)
0.59 ( 0.07)
29 ( 4.9)
2.7
a Mean of two independent IC50 determinations or mean ( SEM) for
n > 2 determinations, see footnotes to Table 1.
b Single determinations from duplicate points. Standard H8 gave
SD 38% (n = 14) in this assay.
c Single determinations from duplicate points. Standard H8 gave
SD 41% (n = 6) in this assay.
d Not determined.
Figure 2. Overlay of H-89–PKA complex (1YDT)26 and the modelled
structure of 23 docked to PKB (1O6K)19 (cyan, H-89–PKA; magenta,
23; green, PKB).
Figure 4. Crystal structure of the PKA–9 complex (PDB 2C1A).
between the hydrophobic side chains of Val57 and
Leu173, and forming a single hydrogen bond between
its heteroaromatic nitrogen and the backbone amide of
Val123 on the kinase hinge. A ÔCHOÕ type interaction
between the isoquinoline C-1 hydrogen and the back-
bone carbonyl of Glu121 is also observed, and the fused
phenyl ring of the heterocycle is orientated towards the
ÔgatekeeperÕ Met120. No close protein contacts are ob-
served with the sulfonamide group of the ligand, which
adopts a rotational conformation placing the S–N bond
orthogonal to the plane of the aromatic ring.
Figure 3. Overlay of H-89–PKA complex (1YDT)26 and the modelled
structure of 24 docked to PKB (1O6K)19 (cyan, H-89–PKA; magenta,
24; green, PKB).
50 showed equivalent inhibition of PKA activity to the
linear inhibitor 9 (Table 3).
The flexible, 7-atom chain between the sulfonamide and
the terminal aromatic ring adopts a pronounced U-
shaped fold which is similar to that observed for H-89
complexed with PKA.27 The protonated nitrogen occu-
pies the PKA ribose binding site and is involved in
salt-bridging and hydrogen-bonding interactions with
The co-crystal structure of 9 bound to PKA (PDB
2C1A) reveals an overall ligand orientation as expected
from previous crystallographic studies with isoquino-
line-5-sulfonamide ligands27,58 (Fig. 4). The isoquinoline
binds in the adenine site of the ATP cleft, sandwiched

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1263
the carboxylate of Asp184 and the side-chain carboxam-
ide of Asn171. The other potential salt-bridging partner
in this region of the protein, Glu127, does not partici-
pate in direct interactions to the ligand. The ligandÕs
terminal 4-chlorophenyl ring is directed towards the
glycine-rich loop and occupies a hydrophobic pocket
defined primarily by the side chains of Thr51, Val53,
Lys72, Leu74 and the backbone carbonyls of Gly52
and Gly55.
The structure of PKA with the pyrrolidine ligand 48
(PDB 2C1B) allows a comparison with the linear com-
pound 9 and shows that the isoquinoline sulfonamide
moieties adopt identical positions and conformations
within the adenine site and form essentially equivalent
interactions with the hinge (Fig. 5). Likewise, the termi-
nal 4-chlorophenyl groups occupy very similar regions
of space, although there are subtle differences in the ex-
act positioning and orientation of the chlorophenyl due
to additional constraints on the cyclised ligand. The
largest difference, however, is observed for the pyrroli-
dine linker, the basic amine being displaced by ca.
Figure 6. Crystal structure of the PKA–48 complex (PDB 2C1B).
region of the PKA (and by homology, PKB) kinase do-
main. The redundancy of potential interactions to the
basic amine and flexibility of the protein in this region
would imply that the other constrained inhibitors 40,
41, 49, 50 and 55 may adopt slightly different contacts
to the protein, whilst maintaining a high degree of over-
all conformational similarity. These changes in detailed
binding mode, and the associated small changes in pro-
tein conformation and water structure, may make the ef-
fects of ligand conformational restriction on binding
affinity harder to predict exactly.
˚
2.4 A, relative to the position in the unconstrained mol-
ecule, and having a significant effect on the binding
interactions of this functional group (Fig. 6). The salt
bridge between the basic group and Asp184 is lost and
is replaced by an interaction with Glu127. Asp184,
which has been observed to be flexible in other struc-
tures, now adopts an alternative rotamer in which it
forms a salt bridge with the basic amine of Lys72. In
addition, a water mediated hydrogen bond is formed be-
tween the basic pyrrolidine amine and the phenol of
Tyr330. The positions of other residues in the ATP-site
of PKA for both inhibitor complexes are essentially
equivalent.
A comparison of the co-crystal structure of 48 and PKA
with the structure predicted from docking of 48 to PKB
(PDB Code 1O6K) (Fig. 7) shows that there is a good
correspondence in the overall conformation of the li-
gand. The glycine-rich loop of PKB is displaced by ca.
˚
Thus, although the constrained pyrrolidine 48 is a very
effective mimic of the overall fold of the linear molecule
9, there is a change in the detailed binding mode be-
tween the two inhibitors. This may reflect the presence
of multiple potential binding sites (principally Asp184
and Glu127) for a basic group within the ribose binding
1 A relative to its position in PKA, and this is reflected
Figure 5. Overlay of the ligands 9 (cyan) and 48 (magenta) bound to
PKA (not shown), as determined by crystallography.
Figure 7. Overlay of the crystal structure of the PKA–48 complex
(cyan) with the model of 48 (magenta) docked to PKB (green).

1264
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
by a shift in position of the 4-chlorophenyl ring by a
similar amount. This loop is known to be a highly flex-
ible region of kinases and can adopt a variety of confor-
mations depending upon the ligand bound and the
protein crystal system.59,60 The contacts and positions
of the isoquinoline ring are identical in the PKB model
and the PKA determined structure. The model also cor-
rectly predicts the interaction of the basic amine of the
ligand 48 with Glu127 rather than Asp184.
Figure 8. Western blot showing titration of inhibitors 9 and 41 for
inhibition of pSer9 GSK3b phosphorylation in PC-3 cells following 1 h
treatment, D, DMSO vehicle control.
Although suitable conformational constraints of the lin-
ear inhibitor 9 were identified by modelling, and shown
by ligand–protein co-crystallography to adopt favour-
able binding conformations, no improvement in the
in vitro inhibition of PKB was seen. The expected entro-
pic gain in affinity through rotatable bond restriction is
difficult to quantify, but has been estimated to be
approximately 2- to 3-fold/rotor.61 For these com-
pounds, it is possible that any relatively small affinity
gain achieved by constraining the molecule is offset by
less favourable protein–ligand interactions in the new
complex. Although Glu127 and Asp184 may be equiva-
lent sites for binding to ligand basic groups, predicting
the precise energetic consequences associated with
changing these interactions is difficult, particularly when
protein movements and changes in hydration are in-
volved. In addition, it should be noted that the subtle
changes in the presentation of the chlorophenyl group
to the hydrophobic pocket in the glycine loop may affect
the free-energy of binding.
the inhibitors was achieved using a Ser9 GSK3b cellular
ELISA protocol to measure the extent of GSK3b phos-
phorylation in PC-3 cells.65 The inhibitors 9 and 41 inhib-
ited Ser9 GSK3b phosphorylation in cells with IC50
values of 15 and 17 lM, respectively, consistent with both
the Western blot data, and their cell growth inhibition in
the SRB assay. Since ligands from the isoquinolinesulf-
onamide class have been shown to be ATP-competitive
inhibitors of a range of kinases, particularly serine/threo-
nine kinases,11,66 the correlation of PKB enzyme inhibi-
tion with antiproliferative effects on cells would be
enhanced by further kinase counterscreening data.67
6. Conclusions
A series of isoquinoline-5-sulfonamide inhibitors of
PKB was developed, with sub-micromolar activity and
structurally distinct from previous ligands in this class.
Structure-based design using the PKB crystal structure
was successful in guiding the preparation of constrained
analogues that mimic the bound conformation of the
linear prototypes, which was confirmed by co-crystal
structure determinations using PKA. Detailed variation
in the binding mode between inhibitors with very similar
overall conformations was observed. Although effective
in this manner, the conformational constraints did not
significantly change the in vitro activity of the com-
pounds. Selectivity for in vitro inhibition of PKB over
PDK1 and PKCa was observed in this series of ligands,
while the compounds were equipotent at PKA. In cellu-
lar assays, the most potent inhibitors in this series
showed the characteristic inhibition of GSK3b phos-
phorylation expected for PKB inhibitors. These mole-
cules should prove useful tools for further exploration
of the biological effects of inhibition of PKB.
The constrained analogues offered no advantage in
selectivity for PKB versus PKA, although it was noted
that PKB/PKA selectivity did show some response to
changes in the spacer between the basic amine and
terminal lipophilic group (cf. H-89 vs 9, 41). Recent
studies have shown that high selectivity for PKB versus
PKA can be achieved when ATP-competitive inhibitors
access a region of the enzyme beyond the glycine-rich
loop.9,10 Significant structural extension of the inhibitors
described here might be needed to occupy similar sites in
the enzyme, and although the current compounds fall
within the accepted ranges of Ôdrug-likeÕ physicochemi-
cal properties62 (e.g., 9 MW 420, ClogP 4.0; 41 MW
432, ClogP 3.9), significant additional functionality
might lead to a less favourable physicochemical profile.
The effects of the inhibitors 9 and 41 on the PI3K-PKB
pathway in cells were studied in more detail. In particu-
lar, the ability of the inhibitors to block GSK3b phos-
phorylation on Ser9 by PKB was examined by
Western blotting in PC-3 cells (Fig. 8). This cell line is
expected to show increased signalling activity through
PI3K-PKB due to the deletion of the negative regulator
PTEN.5,63 At concentrations equivalent to or greater
than the IC50 for cell growth inhibition, both com-
pounds showed inhibition of GSK3b phosphorylation,
indicating that the compounds inhibit PKB kinase activ-
ity towards cells. No effects on total levels of GSK3b or
PKB protein were observed. Additionally, 9 and 41 were
assayed for their ability to inhibit PDK1 kinase activity
in vitro, and found to have no activity at concentrations
up to 30 lM.64 Quantification of the cellular effects of
7. Experimental
7.1. PKB in vitro enzyme assay
Recombinant, purified PKBb enzyme was provided
by Professor David Barford, Structural Biology, Ins-
titute of Cancer Research, Chester Beatty Labs,
Fulham Road. Calbiochem, Merck Biosciences, UK,
supplied the peptide substrate, AKTide-2T (ARK-
RERTYSFGHHA), and assays were run in polypropyl-
ene 96-well plates. Assay volume was 25 lL, with a

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1265
buffer consisting of 20 mM Mops, pH 7.2, 25 mM b-
7.3. Compound preparation and characterisation
glycerophosphate, 5 mM EDTA, 15 mM MgCl2 and
1 mM DTT. The final peptide concentration was
25 lM, with 30 lM ATP and 0.35 lCi [33P-c]ATP. In
the enzyme reaction, the 33P-c-labelled phosphate from
ATP was transferred to the serine residue of the peptide.
Incubations were stopped with 30 lL of 2% orthophos-
phoric acid. The reaction mixture was transferred to a
phosphocellulose filter plate (MAPHNOB from Milli-
pore, UK), where the labelled peptide binds, and the un-
used ATP was washed away with 0.5% orthophosphoric
acid. The filter plate was dried, scintillant was added to
the wells and the incorporated activity was measured on
a TopCountTM (Perkin-Elmer). Under these conditions,
enzyme activity was linear with time (up to 30 min) and
enzyme concentration (up to 1.5 nM). Compounds were
diluted in 12.5% DMSO and added to the assay such
that the final DMSO concentration was 2.5%. Inhibitor
IC50 values were determined from an eight-point con-
centration range run in duplicate and calculated in
GraphPad Prism 4.00. The standard H-89 gave IC50
0.59 ( 0.07) lM under these conditions (lit. value11
IC50 2.5 lM, assay [ATP] 10 lM).
7.3.1. Isoquinoline-5-sulfonic acid (2-(2-(4-chlorobenzyl-
oxy)ethylamino)ethyl)amide (9)
7.3.1.1. General method A. Ethylene glycol (1.67 mL,
30 mmol) was added over 10 min to a slurry of NaH
(60% dispersion in oil, 0.246 g, 10.3 mmol) and dry
THF (15 mL) at rt. The mixture was warmed to 50 ꢁC,
and a solution of 4-chlorobenzyl bromide (2.0 g,
9.8 mmol) in dry THF (10 mL) was added during
45 min. The mixture was refluxed for 18 h, cooled to rt
and partitioned between Et2O (20 mL) and water
(20 mL). The organic layer was washed with brine,
dried, filtered and concentrated. Flash silica chromatog-
raphy (10% MeOH–CH2Cl2) gave the alcohol 15 (R = 4-
Cl) as a pale yellow oil (1.5 g, 8.0 mmol, 82%). 1H NMR
(CDCl3, 250 MHz) d 2.02 (1H, br s), 3.60–3.63 (2H, m),
3.77–3.80 (2H, m), 4.55 (2H, s), 7.31 (2H, dd, J = 6,
3 Hz), 7.35 (2H, dd, J = 6, 3 Hz).
Dess–Martin periodinane (0.670 g, 1.60 mmol) was add-
ed in one portion to a solution of 15 (R = 4-Cl) (0.273 g,
1.46 mmol) in wet CH2Cl2 (15 mL). After stirring at rt
for 2 h, the mixture was diluted with saturated aq
NaHCO3 and sodium thiosulfate (50 mL). After the dis-
appearance of the precipitate (10 min), the mixture was
extracted with CH2Cl2 and the organic phase was
washed with brine, dried and concentrated to give the
aldehyde 16 (R = 4-Cl) as a colourless oil (0.253 g,
1.38 mmol, 94%) that was used without further purifica-
7.2. General synthetic chemistry experimental
Reaction was carried out under N2. Organic solutions
were dried over MgSO4 or Na2SO4. Starting materials
and solvents were purchased from commercial suppli-
ers and were used without further purification. Flash
silica chromatography was performed using Merck sil-
ica gel 60 (0.025–0.04 mm). Ion exchange chromatogra-
phy was performed using Isolute Flash SCX-II (acidic)
1
tion. H NMR (CDCl3, 250 MHz) d 4.14 (2H, s), 4.62
(2H, s), 7.28–7.39 (4H, m), 9.75 (1H, s).
1
or Flash NH2 (basic) resin cartridges. H NMR spec-
A solution of 16 (R = 4-Cl) (0.116 g, 0.71 mmol) in 1,2-
dichloroethane (6 mL) was added to a stirred mixture of
˚
the amine 2 (0.136 g, 0.57 mmol), activated 4 A molecu-
tra were recorded on a Bruker AC250 instrument and
chemical shifts (d) are reported relative to TMS and/
or referenced to the solvent in which they were mea-
sured. Combined HPLC–MS analyses were recorded
using a Waters Alliance 2795 Separations Module
and Waters/Micromass LCT mass detector with elec-
trospray ionisation (+ve or ꢁve ion mode as indicated).
Analytical HPLC was performed using a Supelco DIS-
COVERY C18 5 cm · 4.6 mm i.d., 5 lm column, with a
gradient elution of 10–90% MeOH/0.1% aqueous for-
mic acid at a flow rate of 1 mL minꢁ1 and a running
time of 10 min. Compounds were detected at 254 nm
using a Waters 2487 Dual k Absorbance Detector.
High resolution mass spectra were measured at the
Department of Chemistry, Cambridge University,
UK, on a Bruker Bio Apex II 4.7e FTICR instrument,
or a Waters LCT Premier ESI-ToF instrument, and are
within 5 ppm.
lar sieves (155 mg) and 1,2-dichloroethane (10 mL) at
room temperature. The reaction mixture was stirred
for 20 min followed by addition of NaB(OAc)3H
(0.259 g, 1.18 mmol) in one portion and then AcOH
(10 drops). The reaction mixture was stirred at rt until
the reactants were consumed as determined by the
TLC analysis. The reaction was quenched by addition
of saturated aq NaHCO3 and filtered through a layer
of Celite. The product was extracted with EtOAc, and
the combined organic extracts were dried and concen-
trated. Flash silica chromatography (10% MeOH–
1
CH2Cl2) gave 9 (0.073 g, 0.17 mmol, 30%). H NMR
(CD3OD, 250 MHz) d 2.69–2.71 (4H, m), 3.03 (2H, t,
J = 6 Hz), 3.51 (2H, t, J = 6 Hz), 4.49 (2H, s), 7.40 (4H,
br s), 7.82 (1H, t, J = 8 Hz), 8.43 (1H, d, J = 8 Hz), 8.51
(1H, d, J = 8 Hz), 8.59 (1H, d, J = 7 Hz), 8.65 (1H, d,
J = 6 Hz), 9.42 (1H, s); MS (ESI) m/z 420 [(M+H)+].
HRMS M+H+ calcd for C20H23ClN3O3S420.1143; found
420.1135. HPLC tR 5.98 min; purity (AUC) > 99%.
1
Synthetic intermediates were characterised by H NMR
spectra and MS (where electrospray ionisation permit-
ted observation of a relevant ion) for identity and were
assessed by 1H NMR for homogeneity. Compounds
for biological testing were additionally characterised
by HRMS for identity and were assessed by HPLC
(UV detection) for homogeneity.
7.3.2. Isoquinoline-5-sulfonic acid (2-(2-benzyloxyethyl-
amino)ethyl)amide (3). Prepared from 2-benzyloxyacetal-
1
dehyde using General Method A. H NMR (CD3OD,
250 MHz) d 2.72–2.81 (4H, m), 3.08 (2H, t, J = 6 Hz),
3.57 (2H, t, J = 6 Hz), 4.56 (2H, s), 7.36–7.45 (5H, m),
7.89 (1H, t, J = 8 Hz), 8.47 (1H, d, J = 7 Hz), 8.55 (1H,
d, J = 7 Hz), 8.61 (1H, d, J = 7 Hz), 8.69 (1H, d,
Compounds 211,29,30 and 1729 were prepared according
to the literature procedures.

1266
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
J = 6 Hz), 9.46 (1H, s); MS (ESI) m/z 386 [(M+H)+].
HRMS M+H+ calcd for C20H24N3O3S 386.1533; found
386.1537. HPLC tR 5.62 min; purity (AUC) 97%.
7.3.8. Isoquinoline-5-sulfonic acid (2-(2-(3,4-dichloroben-
zyloxy)ethylamino)ethyl)amide (10). Prepared from 3,4-
dichlorobenzyl alcohol using General Method A. 1H
NMR (CD3OD, 250 MHz) d 2.70–2.78 (4H, m), 3.09
(2H, t, J = 6 Hz), 3.58 (2H, t, J = 5 Hz), 4.55 (2H, s),
7.36 (1H, dd, J = 10, 2 Hz), 7.57 (1H, d, J = 8 Hz),
7.59 (1H, s), 7.91 (1H, t, J = 8 Hz), 8.48 (1H, d,
J = 8 Hz), 8.57 (1H, dd, J = 8, 2 Hz), 8.64 (1H, d,
J = 8 Hz), 8.71 (1H, d, J = 5 Hz), 9.48 (1H, s). MS
(ESI) m/z 454 [(M+H)+]. HRMS M+H+ calcd for
C20H22Cl2N3O3S 454.0753; found 454.0745. HPLC tR
4.67 min; purity (AUC) >99%.
7.3.3. Isoquinoline-5-sulfonic acid (2-(2-(2-methylbenzyl-
oxy)ethylamino)ethyl)amide (4). Prepared from 2-meth-
1
ylbenzyl alcohol using General Method A. H NMR
(CD3OD, 250 MHz) d 2.34 (3H, s), 2.61–2.69 (4H, m),
3.00 (2H, t, J = 6 Hz), 3.51 (2H, t, J = 6 Hz), 4.51 (2H,
s), 7.15–7.31 (4H, m), 7.83 (1H, t, J = 8 Hz), 8.41 (1H,
d, J = 8 Hz), 8.49 (1H, d, J = 7 Hz), 8.56 (1H, d,
J = 6 Hz), 8.62 (1H, d, J = 6 Hz), 9.40 (1H, s); MS
(ESI) m/z 400 [(M+H)+]. HRMS M+H+ calcd for
C21H26N3O3S 400.1689; found 400.1690. HPLC tR
5.80 min; purity (AUC) 95%.
7.3.9. Isoquinoline-5-sulfonic acid (2-(2-(2-methoxyben-
zyloxy)ethylamino)ethyl)amide (11). Prepared from 2-
methoxybenzyl alcohol using General Method A. 1H
NMR (CDCl3, 250 MHz) d 2.61–2.67 (4H, m), 3.00
(2H, t, J = 5 Hz), 3.50 (2H, t, J = 5 Hz), 3.83 (3H, s),
4.51 (2H, s), 6.92 (1H, d, J = 5 Hz), 6.98 (1H, t,
J = 7 Hz) 7.29–7.35 (2H, m), 7.73 (1H, t, J = 7 Hz),
8.23 (1H, J = 7 Hz), 8.46–8.50 (2H, m), 8.71 (1H, d,
J = 5 Hz), 9.39 (1H, s). MS (ESI) m/z 416 [(M+H)+].
HRMS M+H+ calcd for C21H26N3O4S 416.1644; found
416.1649. HPLC tR 3.77 min; purity (AUC) 97%.
7.3.4. Isoquinoline-5-sulfonic acid (2-(2-(3-methylbenzyl-
oxy)ethylamino)ethyl)amide (5). Prepared from 3-meth-
1
ylbenzyl alcohol using General Method A. H NMR
(CD3OD, 250 MHz) d 2.36 (3H, s), 2.61–2.69 (4H, m),
3.01 (2H, t, J = 6 Hz), 3.48 (2H, t, J = 6 Hz), 4.46 (2H,
s), 7.14–7.27 (4H, m), 7.84 (1H, t, J = 8 Hz), 8.41 (1H,
d, J = 8 Hz), 8.50 (1H, d, J = 7, 1 Hz), 8.57 (1H, d,
J = 6 Hz), 8.63 (1H, d, J = 6 Hz), 9.41 (1H, s); MS
(ESI) m/z 400 [(M+H)+]. HRMS M+H+ calcd for
C21H26N3O3S 400.1689; found 400.1691. HPLC tR
4.22 min; purity (AUC) > 99%.
7.3.10. Isoquinoline-5-sulfonic acid (2-(2-(3-methoxyben-
zyloxy)ethylamino)ethyl)amide (12). Prepared from 3-
methoxybenzyl alcohol using General Method A. 1H
NMR (CDCl3, 250 MHz) d 2.55–2.75 (2H, m), 2.89
(2H, t, J = 4 Hz), 2.97 (2H, t, J = 4 Hz), 3.30–3.50
(2H, m), 3.67 (3H, d, J = 5Hz), 4.36 (2H, s), 6.73 (1H,
t, J = 8 Hz), 6.81 (1H, t, J = 6 Hz) 7.10–7.25 (2H, m),
7.52 (1H, dt, J = 4, 7 Hz), 8.02 (1H, dd, J = 2, 7 Hz),
8.24–8.40 (2H, m), 8.48 (1H, d, J = 5 Hz), 9.18 (1H, s).
MS (ESI) m/z 416 [(M+H)+]. HRMS M+H+ calcd for
C21H26N3O4S 416.1644; found 416.1653. HPLC tR
3.69 min; purity (AUC) 96%.
7.3.5. Isoquinoline-5-sulfonic acid (2-(2-(4-methylbenzyl-
oxy)ethylamino)ethyl)amide (6). Prepared from 4-meth-
1
ylbenzyl alcohol using General Method A. H NMR
(CD3OD, 250 MHz) d 2.35 (3H, s), 2.61–2.66 (4H, m),
3.01 (2H, t, J = 6 Hz), 3.46 (2H, t, J = 6 Hz), 4.44 (2H,
s), 7.15–7.25 (4H, m), 7.84 (1H, t, J = 8 Hz), 8.41 (1H,
d, J = 8 Hz), 8.49 (1H, d, J = 7, 1 Hz), 8.57 (1H, d,
J = 6 Hz), 8.63 (1H, d, J = 6 Hz), 9.41 (1H, s); MS
(ESI) m/z 400 [(M+H)+]. HRMS M+H+ calcd for
C21H26N3O3S 400.1689; found 400.1685. HPLC tR
5.99 min; purity (AUC) 96%.
7.3.11. Isoquinoline-5-sulfonic acid (2-(2-(4-methoxyben-
zyloxy)ethylamino)ethyl)amide (13). Prepared from 4-
methoxybenzyl alcohol using General Method A. 1H
NMR (CDCl3, 250 MHz) d 2.37 (2H, t, J = 5 Hz), 2.47
(2H, t, J = 4 Hz), 2.81 (2H, t, J = 5 Hz), 3.23 (2H, t,
J = 5 Hz), 3.66 (3H, s), 4.24 (2H, s), 6.72–6.77 (2H, m)
7.06–7.25 (2H, m), 7.55 (1H, t, J = 7 Hz), 8.05 (1H, d,
J = 8 Hz), 8.28–8.31 (2H, m), 8.54 (1H, d, J = 6 Hz),
9.21 (1H, s). MS (ESI) m/z 416 [(M+H)+]. HRMS
M+H+ calcd for C21H26N3O4S 416.1644; found
416.1649. HPLC tR 3.64 min; purity (AUC) 93%.
7.3.6. Isoquinoline-5-sulfonic acid (2-(2-(2-chlorobenzyl-
oxy)ethylamino)ethyl)amide (7). Prepared from 2-chlo-
1
robenzyl alcohol using General Method A. H NMR
(CD3OD, 250 MHz) d 2.62–2.72 (4H, m), 3.02 (2H, t,
J = 6 Hz), 3.56 (2H, t, J = 6 Hz), 4.60 (2H, s), 7.30–
7.53 (4H, m), 7.84 (1H, t, J = 8 Hz), 8.41 (1H, d,
J = 8 Hz), 8.50 (1H, d, J = 6 Hz), 8.57 (1H, d,
J = 6 Hz), 8.63 (1H, d, J = 6 Hz), 9.41 (1H, s); MS
(ESI) m/z 420 [(M+H)+]. HRMS M+H+ calcd for
C20H23ClN3O3S 420.1143; found 420.1138. HPLC tR
5.79 min; purity (AUC) 95%.
7.3.12. Isoquinoline-5-sulfonic acid piperazin-1 ylamide11
(18)
7.3.7. Isoquinoline-5-sulfonic acid (2-(2-(3-chlorobenzyl-
oxy)ethylamino)ethyl)amide (8). Prepared from 3-chlo-
7.3.12.1. General Method B. A solution of 1 (0.228 g,
1 mmol) in CH2Cl2 (10 mL) was added dropwise to a
stirred solution of piperazine (0.172 g, 2 mmol) and
Et3N (0.275 mL, 2 mmol) in CH2Cl2 (10 mL) cooled in
an ice bath. The solution was warmed to rt. After 3 h,
water (20 mL) was added and the two phases were sep-
arated. The organic layer was dried and concentrated to
give 1811 (0.090 g, 0.324 mmol, 32%) as an oil, which
crystallised on storage at ꢁ10 ꢁC. 1H NMR (CDCl3,
250 MHz) d 1.87 (1H, br s), 2.87–3.16 (4H, m), 3.12–
3.16 (4H, m), 7.72 (1H, t, J = 8 Hz), 8.21–8.24 (1H,
1
robenzyl alcohol using General Method A. H NMR
(CD3OD, 250 MHz) d 2.61–2.69 (4H, m), 3.02 (2H, t,
J = 6 Hz), 3.49 (2H, t, J = 6 Hz), 4.49 (2H, s), 7.29–
7.37 (4H, m), 7.84 (1H, t, J = 8 Hz), 8.41 (1H, d,
J = 8 Hz), 8.50 (1H, dd, J = 7, 1 Hz), 8.58 (1H, d,
J = 6 Hz), 8.64 (1H, d, J = 6 Hz), 9.41 (1H, s); MS
(ESI) m/z 420 [(M+H)+]. HRMS M+H+ calcd for
C20H23ClN3O3S 420.1143; found 420.1140. HPLC tR
5.98 min; purity (AUC) > 99%.

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1267
m), 8.36–8.39 (1H, m), 8.52–8.55 (1H, m), 8.68 (1H, d,
J = 8 Hz), 9.35 (1H, s). MS (ESI) m/z 278 [(M+H)+].
HRMS M+H+ calcd for C13H16N3O2S 278.0958; found
278.0961. HPLC tR 1.60 min; purity (AUC) 90%.
8.42 (1H, d, J = 8 Hz), 8.52 (1H, d, J = 8 Hz), 8.57
(1H, d, J = 5 Hz), 8.64 (1H, d, J = 8 Hz), 9.42 (1H, s).
MS (ESI) m/z 306 [(M+H)+]. HRMS M+H+ calcd for
C15H20N3O2S 306.1271; found 306.1273. HPLC tR
1.13 min; purity (AUC) 93%.
7.3.13. Isoquinoline-5-sulfonic acid piperidin-4-ylamide
(19). Prepared from 4-aminopiperidine using General
Method B. H NMR (CD3OD, 250 MHz) d 1.30–1.46
(2H, m), 1.88 (2H, br d, J = 12 Hz), 2.63–2.74 (3H, m),
3.86 (2H, br d, J = 12 Hz), 7.88 (1H, t, J = 8 Hz), 8.46
(1H, d, J = 8 Hz), 8.46 (1H, dd, J = 5, 1 Hz), 8.60–8.66
(2H, m), 9.42 (1H, s). MS (ESI) m/z 292 [(M+H)+].
HRMS M+H+ calcd for C14H18N3O2S 292.1114; found
292.1119. HPLC tR 2.71 min; purity (AUC) 95%.
7.3.17. Isoquinoline-5-sulfonic acid pyrrolidin-3-ylamide
(23). Prepared from 3-aminopyrrolidine using General
Method C. H NMR (CD3OD, 250 MHz) d 1.60–1.73
(1H, m), 2.01–2.14 (1H, m), 3.04 (1H, dd, J = 10,
5 Hz), 3.33–3.58 (4H, m), 7.85 (1H, dd, J = 8, 8 Hz),
8.40–8.49 (2H, m), 8.62–8.68 (2H, m), 9.39 (1H, s);
MS (ESI) m/z 278 [(M+H)+]. HRMS M+H+ calcd for
C13H16N3O2S 278.0963; found 278.0961. HPLC tR
1.34 min; purity (AUC) 94%.
1
1
7.3.14. Isoquinoline-5-sulfonic acid (piperidin-4-ylmeth-
yl)amide (21). Prepared from 4-aminomethyl-piperidine
using General Method B. 1H NMR (CD3OD,
250 MHz) d 0.75–0.93 (2H, m), 1.24–1.51 (3H, m),
2.22–2.36 (2H, m), 2.62 (2H, d, J = 7 Hz), 2.77–2.81
(2H, m), 7.66–7.74 (1H, m), 8.22–8.36 (2H, m), 8.42–
8.53 (2H, m), 9.27 (1H, s). MS (ESI) m/z 306 [(M+H)+].
HRMS M+H+ calcd for C15H20N3O2S 306.1276; found
306.1267. HPLC tR 1.32 min; purity (AUC) 95%.
7.3.18. Isoquinoline-5-sulfonic acid ((2S)-pyrrolidin-2-
ylmethyl)amide (24). Prepared from (2S)-2-aminometh-
ylpyrrolidine using General Method C. 1H NMR
(CD3OD, 250 MHz) d 1.49–1.84 (4H, m), 3.02–3.09
(2H, m), 3.39–3.52 (2H, m), 4.00–4.10 (1H, m), 7.91
(1H, dd, J = 8, 8 Hz), 8.50 (1H, d, J = 8 Hz), 8.55 (1H,
d, J = 8 Hz), 8.67 (1H, d, J = 7 Hz), 8.83 (1H,
J = 7 Hz), 9.44 (1H, s); MS (ESI) m/z 292 [(M+H)+].
HRMS M+H+ calcd for C14H18N3O2S 292.1120; found
292.1109. HPLC tR 1.47 min; purity (AUC) 97%.
7.3.15. Isoquinoline-5-sulfonic acid (4-aminopiperidin-1-
yl)amide (20)
7.3.15.1. General Method C. 4-(N-Boc-amino)piperi-
dine (891 mg, 4.45 mmol) was treated as described in
General Method A followed by recrystallisation from
CH2Cl2–hexane to afford [1-(isoquinoline-5-sulfo-
nyl)piperidin-4-yl]carbamic acid tert-butyl ester as a
white solid (398 mg, 1.02 mmol, 47%). 1H NMR
(CDCl3, 250 MHz) d 1.40 (9H, s), 1.86–1.98 (4H, m),
2.67–2.77 (2H, m), 3.40–3.49 (1H, br s), 3.75–3.80 (2H,
m), 4.37–4.42 (1H, br s), 7.72 (1H, t, J = 8 Hz), 8.23
(1H, d, J = 8 Hz), 8.38 (1H, d, J = 8 Hz), 8.46 (1H, d,
J = 7 Hz), 8.68 (1H, d, J = 6 Hz), 9.36 (1H, s). MS
(ESI) m/z 336 [(M+H-tBu)+].
7.3.19.
Isoquinoline-5-sulfonic
acid
trans-(4-ben-
zylaminocyclohexyl)amide (25). A solution of benzalde-
hyde (0.044 mL, 0.43 mmol) in 1,2-dichloroethane
(5 mL) was added to a vigorously stirred mixture of
(trans-4-aminocyclohexyl)carbamic acid tert-butyl ester
˚
(98 mg, 0.45 mmol), activated 4 A molecular sieves
(164 mg) and 1,2-dichloroethane (20 mL) at 0 ꢁC. After
20 min, Na(OAc)3H (179 mg, 0.8 mmol) was added in
one portion, followed by AcOH (0.5 mL). The reaction
mixture was stirred at rt until the reactants were con-
sumed as determined by TLC. The reaction was
quenched by addition of a saturated solution of NaH-
CO3, followed by filtration through a layer of Celite.
The filtrate was extracted with EtOAc and the organic
extract was dried and concentrated. Flash silica chroma-
tography (1% NH3–9% MeOH–90% CH2Cl2) gave
trans-(4-benzylaminocyclohexyl)carbamic acid tert-bu-
Trifluoroacetic acid (2 mL) was added dropwise to a
stirred solution of [1-(isoquinoline-5-sulfonyl)piperidin-
4-yl]carbamic acid tert-butyl ester (309 mg, 0.79 mmol)
in CH2Cl2 (5 mL) cooled in an ice bath. After 3 h, the
solvents were concentrated in vacuo. The crude mixture
was dissolved in water (20 mL), basified to pH 10 by
addition of 1 M aq NaOH and extracted with EtOAc.
The organic layer was dried and concentrated. Flash sil-
ica chromatography (20% MeOH–CH2Cl2) gave 20
(61 mg, 0.21 mmol, 26%). 1H NMR (CD3OD,
250 MHz) d 1.53–1.66 (2H, m), 2.01–2.06 (2H, m),
2.71–2.81 (2H, m), 3.05–3.14 (1H, m), 3.97–4.03 (2H,
m), 7.91 (1H, t, J = 8 Hz), 8.48–8.52 (2H, m), 8.60
(1H, d, J = 8 Hz), 8.68 (1H, d, J = 5 Hz), 9.45 (1H, s).
MS (ESI) m/z 292 [(M+H)+]. HRMS M+H+ calcd for
C14H18N3O2S 292.1114; found 292.1116. HPLC tR
2.72 min; purity (AUC) 95%.
1
tyl ester (62 mg, 0.20 mmol, 45%). H NMR (CDCl3,
250 MHz) d 1.06–1.31 (4H, m), 1.43 (9H, s), 1.99–2.02
(5H, m), 2.45–2.53 (1H, m), 3.39 (1H, br s), 3.81 (2H,
s) 4.31 (1H, br s), 7.28–7.36 (5H, m). MS (ESI) m/z
306 [(M+H)+].
CF3CO2H (1 mL) was added dropwise at 0 ꢁC to a stir-
red solution of trans-(4-benzylaminocyclohexyl)carbam-
ic acid tert-butyl ester (62 mg, 0.2 mmol) in CH2Cl2
(1 mL). After 3 h, the solvents were concentrated. The
crude product was purified by ion exchange on basic res-
in (MeOH) to afford N-benzyl-trans-cyclohexane-1,4-di-
amine (37 mg, 0.18 mmol, 91%). 1H NMR (CD3OD,
250 MHz) d 1.12–1.29 (4H, m), 1.92–2.07 (4H, m),
2.46–2.56 (1H, m), 2.65–2.75 (1H, m), 3.82 (2H, s)
7.28–7.39 (5H, m). MS (ESI) m/z 205 [(M+H)+].
7.3.16. Isoquinoline-5-sulfonic acid trans-(4-amino-
cyclohexyl)amide (22). Prepared from trans-1,4-diamino-
cyclohexane using General Method C. 1H NMR
(CD3OD, 250 MHz) d 0.94–1.91 (8H, m), 2.52–2.61
(1H, m), 3.01–3.07 (1H, m), 7.84 (1H, t, J = 8 Hz),
N-Benzyl-trans-cyclohexane-1,4-diamine (37 mg, 0.18 mmol)
was treated as described in General Method A to give

1268
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1
25 (11 mg, 0.03 mmol, 15%) as a white solid. H NMR
(CD3OD, 250 MHz) d 1.05–1.33 (4H, m), 1.62–1.93
(4H, m), 2.42 (1H, m), 3.05 (1H, m), 3.75 (2H, s) 7.28–
7.39 (5H, m), 7.86 (1H, t, J = 8 Hz), 8.42 (1H, d,
J = 8 Hz), 8.53 (1H, d, J = 8 Hz), 8.58 (1H, d,
J = 8 Hz), 8.66 (1H, d, J = 8 Hz), 9.43 (1H, s). MS
(ESI) m/z 396 [(M+H)+]. MS (ESI) m/z 292 [(M+H)+].
HRMS M+H+ calcd for C22H26N3O2S 396.1740; found
396.1740. HPLC tR 2.72 min; purity (AUC) 95%.
dropwise over 10 min to a stirred solution of 28
(84 mg, 0.16 mmol) in 1,2-dichloroethane (4 mL) at
0 ꢁC. Stirring was continued at 0 ꢁC for 10 min, after
which the mixture was refluxed for 3 days. The mixture
was cooled to rt and concentrated. The resulting oil was
dissolved in MeOH (5 mL) and refluxed for 2 h. The sol-
vent was evaporated and the residue was dissolved in
water (10 mL). The solution was washed sequentially
with Et2O and CH2Cl2, then basified to pH 9 with NaH-
CO3 (s) and extracted with CH2Cl2. The organic extract
was dried and concentrated. Flash silica chromatogra-
phy (10% MeOH–CH2Cl2) gave 29 (29 mg, 0.07 mmol,
7.3.20. Isoquinoline-5-sulfonic acid (3-(4-chlorobenzyl-
oxymethyl)piperazin-1-yl)amide (29). 1-Benzyl-pipera-
zine-2-carboxylic acid ethyl ester35 (26) (60 mg,
0.24 mmol) was treated as described in General Method
A, followed by flash silica chromatography, (4%
MeOH–CH2Cl2), to give 1-benzyl-4-(isoquinoline-5-sul-
fonyl)piperazine-2-carboxylic acid ethyl ester (27)
(61 mg, 0.14 mmol, 57%). 1H NMR (CDCl3,
250 MHz) d 1.20 (3H, t, J = 8 Hz), 2.48–2.56 (1H, m),
2.99–3.45 (5H, m), 3.61–3.70 (2H, m), 3.88–4.11 (3H,
m), 7.23–7.32 (5H, m), 7.74 (1H, t, J = 8 Hz), 8.26
(1H, d, J = 8 Hz), 8.40 (1H, dd, J = 8, 1 Hz), 8.47 (1H,
d, J = 6 Hz), 8.70 (1H, d, J = 6 Hz), 9.38 (1H, s). MS
(ESI) m/z 440 [(M+H)+].
1
42%). H NMR (CDCl3, 250 MHz) d 2.35–2.46 (1H,
m, 1H), 2.67–3.63 (9H, m), 4.41 (2H, s), 7.20 (2H, d,
J = 8 Hz), 7.30 (2H, d, J = 8 Hz), 7.72 (1H, t,
J = 8 Hz), 8.23 (1H, d, J = 7 Hz), 8.36 (1H, d, J = 8
Hz), 8.50 (1H, d, J = 6 Hz), 8.68 (1H, d, J = 6 Hz),
9.35 (1H, s). MS (ESI) m/z 432 [(M+H)+]. HRMS
M+H+ calcd for C21H23ClN3O3S 432.1143; found
432.1147. HPLC tR 4.49 min; purity (AUC) 99%.
7.3.21. Isoquinoline-5-sulfonic acid (2S,4R)-[2-(4-chloro-
benzyloxymethyl)pyrrolidin-4-yl]amide (40). (2S,4R)-4-
Aminopyrrolidine-1,2-dicarboxylic acid 1-tert-butyl
ester 2-methyl ester hydrochloride (30) (265 mg,
0.94 mmol) was treated according to General Method
A solution of 27 (193 mg, 0.44 mmol) in THF (2 mL) was
added to a stirred suspension of LiAlH4 (1 M in THF,
0.45 mL, 0.45 mmol) at 0 ꢁC. After 1.5 h, water (30 lL)
was added, followed by 1 M aq NaOH (40 lL). After
stirring at rt for 18 h, further water (30 lL) was added
and the suspension was filtered through a layer of Celite,
washing the filter cake with Et2O. The organic filtrate
was concentrated. Flash silica chromatography (10%
MeOH–CH2Cl2) gave [1-benzyl-4-(isoquinoline-5-sulfo-
B
followed by flash silica chromatography (5%
MeOH–CH2Cl2) to give (2S,4R)-4-(isoquinoline-5-sul-
fonylamino)pyrrolidine-1,2-dicarboxylic acid 1-tert-bu-
tyl ester 2-methyl ester (32) as an oil (309 mg,
1
0.595 mmol, 63%). H NMR (CDCl3) d 1.27 (9H, s),
1.81–2.24 (2H, m), 2.90–3.12 (1H, m), 3.37–3.50 (1H,
m), 3.62 (3H, s), 3.77–3.94 (1H, m), 4.10–4.27 (1H, m),
5.91–6.27 (1H, m), 7.60–7.68 (1H, m), 8.15–8.18 (1H,
m), 8.31–8.41 (2H, m), 8.55–8.62 (1H, m), 9.30 (1H, s).
MS (ESI) m/z 436 [(M+H)+].
nyl)piperazin-2-yl]-methanol
((116 mg,
0.29 mmol,
66%). 1H NMR (CDCl3, 250 MHz) d 2.20 (1H, s),
2.41–2.50 (1H, m), 2.72–2.73 (1H, m), 2.86–2.96 (2H,
m), 3.04–3.12 (1H, m), 3.37–3.65 (3H, m), 3.63–3.84
(1H, m), 3.92–4.06 (2H, m), 7.24–7.35 (5H, m), 7.75
(1H, t, J = 8 Hz), 8.26 (1H, d, J = 8 Hz), 8.40 (1H, dd,
J = 8, 1 Hz), 8.53 (1H, d, J = 6 Hz), 8.70 (1H, d,
J = 8 Hz), 9.39 (1H, s). MS (ESI) m/z 398 [(M+H)+].
NaH (60% dispersion in oil, 43 mg, 1.07 mmol) was add-
ed in one portion to a stirred solution of 32 (357 mg,
0.82 mmol) in DMF (8 mL) at 0 ꢁC. After 5 min, SEM-
Cl (152 lL, 0.859 mmol) was added and stirring was
continued for 5 min. The mixture was diluted with water
(30 mL) and extracted with EtOAc (3 · 30 mL). The
organic layers were dried and concentrated. Flash silica
chromatography (70% EtOAc–petroleum ether) gave
(2S,4R)-4-[isoquinoline-5-sulfonyl(2-trimethylsilanyleth-
oxymethyl)amino]pyrrolidine-1,2-dicarboxylic acid 1-
tert-butyl ester 2-methyl ester (34) as an oil (372 mg,
7.3.20.1. General Method D. NaH (60% dispersion in
oil, 27 mg, 0.67 mmol) was added at rt to a suspension
of [1-benzyl-4-(isoquinoline-5-sulfonyl)piperazin-2-yl]-
methanol (130 mg, 0.33 mmol) in DMF (5 mL). After
15 min, 4-chlorobenzyl bromide (64 mg, 0.31 mmol)
was added. After 4.5 h, the solution was diluted with
water (20 mL) and extracted with EtOAc. The organic
layer was washed with water, dried and concentrated.
Flash silica chromatography (10% MeOH–CH2Cl2)
gave 5-[4-benzyl-3-(4-chloro-benzyloxymethyl)pipera-
zine-1-sulfonyl]-isoquinoline (28) (84 mg, 0.16 mmol,
1
0.658 mmol, 80%). H NMR (CDCl3) d 0.00 (9H, s),
0.80–0.88 (2H, m), 1.37, 1.42 (9H, 2· s), 1.85–2.55
(2H, m), 3.26–3.37 (1H, m), 3.52–3.62 (3H, m), 3.69
(3H, s), 4.28–4.49 (2H, m), 4.87–5.04 (2H, m), 7.68–
7.74 (1H, m), 8.22–8.46 (3H, m), 8.69–8.71 (1H, m),
9.37 (1H, s). MS (ESI) m/z 566 [(M+H)+].
1
52%). H NMR (CDCl3, 250 MHz) d 2.32–2.46 (1H,
m), 2.67–2.84 (2H, m), 3.06–3.93 (8H, m), 4.31 (2H,
br s), 7.16–7.33 (9H, m), 7.70 (1H, t, J = 8 Hz), 8.20
(1H, d, J = 8 Hz), 8.36 (1H, d, J = 7 Hz), 8.52 (1H,
br s), 8.67 (1H, d, J = 6 Hz), 9.34 (1H, s). MS (ESI)
m/z 522 [(M+H)+].
7.3.21.1. General Method E. To a solution of 34
(330 mg, 0.58 mmol) in THF (10 mL) at 0 ꢁC was added
LiAlH4 (584 lL of a 1 M sol in THF, 0.58 mmol). After
20 min, further LiAlH4 (300 lL, 0.30 mmol) was added.
After 20 min, a third portion of LiAlH4 (200 lL,
0.20 mmol) was added. The reaction mixture was diluted
with Et2O (20 mL) and quenched by dropwise addition
of water (40 lL), followed by 10% aq NaOH (40 lL).
A solution of 1-chloroethyl chloroformate (0.040 mL,
0.38 mmol) in 1,2-dichloroethane (2 mL) was added

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1269
After stirring for 18 h, water (120 lL) was added
dropwise and the resulting white precipitate was filtered
through a pad of Celite. The filtrate was concentrated.
Flash silica chromatography (EtOAc) gave (2S,4R)-2-
hydroxymethyl-4-[isoquinoline-5-sulfonyl(2-trimethylsi-
lanylethoxymethyl)amino]pyrrolidine-1-carboxylic acid
tert-butyl ester (36) as an oil (227 mg, 0.422 mmol,
matography (Et2O) to give (2S,4R)-4-(4-chlorobenzyl-
oxy)-2-hydroxymethylpyrrolidine-1-carboxylic acid
tert-butyl ester (44) (1.23 g, 3.60 mmol, 44%) as an oil.
1H NMR (CDCl3) d 1.50 (9H, s), 1.6–2.3 (2H, m),
3.40–4.20 (6H, m), 4.45–4.55 (2H, m), 4.7–4.9 (1H, m),
7.25–7.36 (4H, m). MS (ESI) m/z 364 [(M+Na)+].
1
72%). H NMR (CDCl3) d 0.00 (9H, s), 0.81–0.87 (2H,
7.3.23.1. General Method F. Ms2O (2 M in CH2Cl2,
932 lL, 1.87 mmol) was added dropwise to a solution
of 44 (425 mg, 1.24 mmol) and Et3N (520 lL,
3.73 mmol) in CH2Cl2 (4 mL) at 0 ꢁC and the resulting
solution was stirred for 15 min. CH2Cl2 (10 mL) was
added and the solution was washed with 1 M HCl
(10 mL), saturated aq NaHCO3 (10 mL) and brine
(10 mL). The organic layer was dried and concentrated
to give the crude mesylate as an oil (483 mg), which
was used immediately without further purification. The
oil (483 mg) was dissolved in DMF (6 mL) and NaN3
was added (162 mg, 2.49 mmol). After heating at 80 ꢁC
for 3 h, the solution was cooled, water (30 mL) was add-
ed and the mixture was extracted with EtOAc (30 mL).
The organic layer was washed with brine (30 mL), dried
and concentrated to give (2S,4R)-2-azidomethyl-4-(4-
chlorobenzyloxy)pyrrolidine-1-carboxylic acid tert-butyl
m), 1.44 (9H, s), 1.85–2.30 (2H, m), 3.30–4.50 (8H, m),
4.87–5.04 (2H, m), 7.69–7.75 (1H, m), 8.22–8.26 (1H,
m), 8.35–8.37 (1H, m), 8.44–8.47 (1H, m), 8.70–8.72
(1H, m), 9.38 (1H, s). MS (ESI) m/z 538 [(M+H)+],
560 [(M+Na)+].
Compound 36 (113 mg, 0.21 mmol) was treated accord-
ing to General Method D with flash silica chromatogra-
phy (70% EtOAc–petroleum ether) to give (2S,4R)-2-(4-
chlorobenzyloxymethyl)-4-[isoquinoline-5-sulfonyl(2-tri-
methylsilanylethoxymethyl)amino]pyrrolidine-1-carbox-
ylic acid tert-butyl ester (38) as an oil (92 mg,
1
0.139 mmol, 66%). H NMR (CDCl3) d 0.00 (9H, s),
0.82–0.88 (2H, m), 1.41 (9H, s), 1.80–2.35 (2H, m),
3.21–4.05 (7H, m), 4.38 (2H, s), 4.52–4.63 (1H, m),
4.90–5.05 (2H, m), 7.16–7.32 (4H, m), 7.52–7.58 (1H,
m), 8.16–8.19 (1H, m), 8.30–8.36 (2H, m), 8.65–8.68
(1H, m), 9.35 (1H, s). MS (ESI) m/z 662 [(M+H)+].
1
ester (46) as an oil (400 mg, 1.09 mmol, 88%). H NMR
(CDCl3) d 1.50 (9H, s), 3.2–4.2 (8H, m), 4.50 (2H, br s),
7.25–7.37 (4H, m). MS (ESI) m/z 389 [(M+Na)+].
Six molar HCl (1 mL) was added to a solution of 38
(22 mg, 0.046 mmol) in MeOH (1 mL) at rt. After
refluxing for 4 h, the solution was cooled, concentrated
and purified by ion exchange on basic resin (MeOH)
to give 40 as an oil (15 mg, 0.035 mmol, 76%). 1H
NMR (CDCl3) d 1.48–1.57 (2H, m), 2.51 (1H, dd,
J = 11, 5 Hz), 2.91 (1H, dd, J = 11, 6 Hz), 3.11–3.34
(3H, m), 3.66–3.75 (1H, m), 4.33 (2H, s), 7.08–7.25
(4H, m), 7.44–7.47 (1H, m), 8.13–8.16 (1H, m), 8.31–
8.40 (2H, m), 8.59–8.61 (1H, m), 9.30 (1H, s). MS
(ESI) m/z 432 [(M+H)+]. HRMS M+H+ calcd for
C21H23ClN3O3S 432.1143; found 432.1146. HPLC tR
6.00 min; purity (AUC) >99%.
A solution of 46 (50 mg, 0.14 mmol) in EtOAc (3 mL)
with 10% Pd/C (10 mg) was hydrogenated at rt, 1 atm
for 2 h. The mixture was filtered and concentrated. Ion
exchange on acidic resin (MeOH, then 1 M NH3 in
MeOH) gave the crude amine as an oil (27 mg), which
was treated according to General Method B. The crude
sulfonamide thus obtained was dissolved in MeOH
(1 mL) and 1 M HCl in Et2O (2 mL). After 3.5 h, the
mixture was concentrated. Ion exchange on acidic resin
(MeOH, then 1 M NH3 in MeOH) gave 48 (31 mg,
1
0.072 mmol, 53%) as an oil. H NMR (CDCl3) d 1.44–
1.52 (1H, m), 1.92–2.06 (1H, m), 2.67–2.75 (2H, m),
2.88–3.05 (2H, m), 3.45–3.52 (1H, m), 3.98–4.03 (1H,
m), 4.37 (2H, s), 7.13–7.35 (4H, m), 7.67–7.74 (1H, m),
8.17–8.23 (1H, m), 8.43–8.46 (2H, m), 8.69–8.73 (1H,
m), 9.37 (1H, s). MS (ESI) m/z 432 [(M+H)+]. HRMS
M+H+ calcd for C21H23ClN3O3S 432.1143; found
432.1146. HPLC tR 4.87 min; purity (AUC) 96%.
7.3.22. Isoquinoline-5-sulfonic acid (2S,4S)-(2-(4-chloro-
benzyloxymethyl)pyrrolidin-4-yl)amide (41). Prepared as
described for 40 starting from (2S,4S)-4-aminopyrroli-
dine-1,2-dicarboxylic acid 1-tert-butyl ester 2-methyl es-
1
ter hydrochloride (31). H NMR (CDCl3) d 1.25–1.43
(1H, m), 1.97–2.10 (1H, m), 2.67–2.80 (2H, m), 3.27–
3.56 (3H, m), 3.79–3.84 (1H, m), 4.51 (1H, d,
J = 12 Hz), 4.58 (1H, d, J = 12 Hz), 7.29–7.40 (4H, m),
7.66–7.73 (1H, m), 8.15–8.22 (2H, m), 8.40–8.43 (1H,
m), 8.58–8.60 (1H, m), 9.36 (1H, d, J = 1 Hz). MS
(ESI) m/z 432 [(M+H)+]. HRMS M+H+ calcd for
C21H23ClN3O3S 432.1143; found 432.1143. HPLC tR
4.57 min; purity (AUC) 95%.
7.3.24. Isoquinoline-5-sulfonic acid (2S,4S)-(4-(benzyl-
oxy)pyrrolidin-2-yl-methyl)amide (49) and Isoquinoline-
5-sulfonic acid (2S,4S)-(4-(4-chlorobenzyloxy)pyrrolidin-
2-yl-methyl)amide (50). Prepared following the proce-
dure described for 48 to give a mixture of 49 and 50.
A portion of the mixture was separated by preparative
1
HPLC to give 49: H NMR (CDCl3) d 1.84–1.88 (1H,
m), 2.26–2.38 (1H, m), 3.15–3.22 (1H, m), 3.33–3.45
(2H, m), 3.62–3.67 (1H, m), 4.05–4.12 (1H, m), 4.26
(1H, br s), 4.45 (1H, d, J = 12 Hz), 4.56 (1H, d,
J = 12 Hz), 7.25–7.38 (4H, m), 7.61–7.68 (1H, m),
8.16–8.20 (1H, m), 8.35–8.38 (1H, m), 8.48–8.50 (1H,
m), 8.63–8.66 (1H, m), 9.34 (1H, s). MS (ESI) m/z 398
[(M+H)+]. HRMS M+H+ calcd for C21H24N3O3S
398.1533; found 398.1535. HPLC tR 4.45 min; purity
7.3.23. Isoquinoline-5-sulfonic acid (2S,4R)-(4-(4-chlo-
robenzyloxy)pyrrolidin-2-yl-methyl)amide (48). (2S,4R)-
4-Hydroxypyrrolidine-1,2-dicarboxylic acid 1-tert-butyl
ester 2-methyl ester38 (42) (2.00 g, 8.15 mmol) was treat-
ed according to General Method D with flash silica
chromatography (40% EtOAc–petroleum ether) to give
an oil (1.81 g, impure). The crude material was treated
according to General Method E with flash silica chro-
1
(AUC) >99%. 50: H NMR (CDCl3) d 1.83–1.89 (1H,

1270
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
m), 2.26–2.38 (1H, m), 3.16–3.20 (1H, m), 3.32–3.41
(2H, m), 3.61–3.66 (1H, m), 3.95–4.15 (1H, m), 4.25
(1H, br s), 4.40 (1H, d, J = 12 Hz), 4.53 (1H, d,
J = 12 Hz), 7.21–7.30 (4H, m), 7.63–7.69 (1H, m),
8.18–8.21 (1H, m), 8.34–8.37 (1H, m), 8.47–8.50 (1H,
m), 8.64–8.68 (1H, m), 9.36 (1H, s). MS (ESI) m/z 432
[(M+H)+]. HRMS M+H+ calcd for C21H23N3O3ClS
432.1143; found 432.1146. HPLC tR 4.94 min; purity
(AUC) >99%.
[(M+H)+]. HRMS M+H+ calcd for C22H25ClN3O3S
446.1300; found 446.1297. HPLC tR 5.30 min; purity
(AUC) >99%.
7.3.26. 4-Methylisoquinoline-5-sulfonic acid (2S,4S)-(2-
(4-chlorobenzyloxymethyl)pyrrolidin-2-yl)amide
(61).
KH (30% dispersion in oil, 8.19 g, 61 mmol) was washed
with hexane (30 mL) and suspended in THF (50 mL).
The suspension was cooled in an ice bath and a solution
of (2S,4R)-2-hydroxymethyl-4-(tetrahydropyran-2-yl-
oxy)-pyrrolidine-1-carboxylic acid tert-butyl ester40
(56) (11.3 g, 37.5 mmol) in THF (50 mL) was added
dropwise via cannula. The mixture was stirred for
5 min, followed by addition of a solution of 4-chlorob-
enzyl chloride (7.83 g, 48.6 mmol) in THF (20 mL) via
cannula. The mixture was stirred for 48 h at rt and then
poured into EtOAc (300 mL). The solution was washed
with water (100 mL), saturated aq NaHCO3 (50 mL)
and brine (50 mL). The organic fraction was dried and
concentrated. Flash silica chromatography (20%
EtOAc–hexane) gave (2S,4R)-2-(4-chlorobenzyloxy-
methyl)-4-(tetrahydropyran-2-yloxy)-pyrrolidine-1-carb-
oxylic acid tert-butyl ester (57) (8.66 g, 20.4 mmol, 54%).
1H NMR (250 MHz, CDCl3) d 1.40–2.00 (8H, m), 1.47
(9H, s), 3.35–3.75 (4H, m), 3.75–4.25 (2H, m), 4.40–
4.85 (4H, m), 7.23–7.40 (4H, m). MS (ESI) m/z 448
[(M+Na)+].
7.3.25. Isoquinoline-5-sulfonic acid (2R*,5S*)-(5-(4-chlo-
robenzyloxymethyl)pyrrolidin-2-yl-methyl)amide
(55).
Et3N (3.5 mL, 24.8 mmol), DMAP (303 mg, 2.48 mmol)
and Boc2O (3.25 g, 14.9 mmol) were added to a stirred
solution of (1S*,5R*)-3,8-diazabicyclo[3.2.1]octan-2-
one39 (51) (626 mg, 4.96 mmol) in CH2Cl2 (20 mL) at
rt. After 15 h, the mixture was absorbed onto silica
and purified by flash silica chromatography (50–80%
Et2O–hexane) to give (1S*,5R*)-2-oxo-3,8-diazabicy-
clo[3.2.1]octane-3,8-dicarboxylic acid di-tert-butyl ester
(692 mg, 2.13 mmol, 43%) (52) as an oil. 1H NMR
(CDCl3) d 1.48 (9H, s), 1.54 (9H, s), 1.72–1.83 (1H,
m), 2.05–2.23 (3H, m), 3.43 (1H, dd, J = 1 Hz), 3.87
(1H, dd, J = 1 Hz), 4.56 (2H, br s). MS (ESI) m/z 349
[(M+Na)+].
NaBH4 (76 mg, 2.00 mmol) was added in one portion to
a solution of 52 (327 mg, 1.00 mmol) in MeOH (10 mL)
at 0 ꢁC. After 1 h, water (20 mL) was added and the
mixture was extracted with Et2O (3· 20 mL). The organ-
ic extracts were dried and concentrated to give
(2R*,5S*)-2-(tert-butoxycarbonylaminomethyl)-5-hydro
xymethylpyrrolidine-1-carboxylic acid tert-butyl ester
(53) (331 mg, 2.00 mmol, 100%) as an oil. 1H NMR
(CDCl3) d 1.36 (9H, s), 1.42 (9H, s), 1.6–2.0 (4H, m),
2.9–4.0 (6H, m). MS (ESI) m/z 353 [(M+Na)+].
PPTS (0.59 g, 2.3 mmol) was added to a stirred solution
of 57 (8.66 g, 20.3 mmol) in EtOH (50 mL). The solution
was heated at 55 ꢁC for 4 h. The cooled solution was
diluted with EtOAc (250 mL) and then washed with sat-
urated aq NaHCO3 (50 mL) and brine (50 mL). The
organic fraction was dried and concentrated. Flash
silica chromatography (30–60% EtOAc–hexane) gave
(2S,4R)-2-(4-chlorobenzyloxymethyl)-4-hydroxy-pyrrol-
idine-1-carboxylic acid tert-butyl ester (58) (5.13 g,
1
Compound 53 (205 mg, 0.62 mmol) was treated accord-
ing to General Method D followed by flash silica chro-
matography (50–100% EtOAc–hexane) to give
(2R*,5S*)-2-(tert-butoxycarbonylaminomethyl)-5-(4-chl
orobenzyloxymethyl)pyrrolidine-1-carboxylic acid tert-
15.0 mmol, 74%). H NMR (250 MHz, CDCl3) d 1.47
(9H, s), 2.00–2.30 (2H, m), 3.40–3.80 (4H, m), 4.05–
4.25 (1H, m), 4.45–4.60 (3H, m), 7.25 (2H, d,
J = 10 Hz), 7.34 (2H, d, J = 10 Hz). MS (ESI) m/z 364
[(M+Na)+].
1
butyl ester (54) (200 mg, 0.44 mmol, 71%) as an oil. H
NMR (CDCl3) d 1.35 (18H, s), 1.6–2.0 (4H, m), 3.07–
3.42 (4H, m), 3.87 (2H, br s), 4.36–4.51 (2H, m), 7.16–
7.26 (4H, m). MS (ESI) m/z 477 [(M+Na)+].
Compound 58 (1.86 g, 5.40 mmol) was treated accord-
ing to General Method F to give the crude azide. The
azide was dissolved in THF (20 mL) and PPh3 (3.56 g,
13.6 mmol) and water (2 mL, 111 mmol) were added.
The mixture was stirred at rt for 18 h. Ion exchange
on acidic resin (MeOH, then 2 M NH3 in MeOH) and
flash silica chromatography (5–10% MeOH–CH2Cl2)
gave (2S,4S)-4-amino-2-(4-chlorobenzyloxymethyl)-pyr-
rolidine-1-carboxylic acid tert-butyl ester (59) (1.29 g,
3.78 mmol, 70%). 1H NMR (250 MHz, CD3OD) d
1.45 (9H, s), 1.82 (1H, m), 2.42 (1H, m), 3.03 (1H, dd,
J = 6.3, 11.0 Hz), 3.42 (1H, t, J = 6.6 Hz), 3.60 (1H,
dd, J = 2.5, 9.4 Hz), 3.75 (1H, dd, J = 6.9, 10.6 Hz),
3.95 (1H, br s), 4.53 (1H, d, J = 12 Hz), 4.60 (1H, d,
J = 12 Hz), 7.36 (4H, s). MS (ESI) m/z 341 [(M+H)+].
One molar HCl in Et2O (3 mL) was added to a stirred
solution of 54 (116 mg, 0.255 mmol) in MeOH (3 mL)
at rt After 22 h, the solution was concentrated and the
residues were purified by ion exchange on acidic resin
(MeOH, then 1 M NH3 in MeOH). The basic fractions
were concentrated to give the crude deprotected amine
as an oil (59 mg). Treatment of the crude amine accord-
ing to General Method B was followed by ion exchange
on basic resin (MeOH). Flash silica chromatography
(10% MeOH–CH2Cl2) gave 55 (45 mg, 0.101 mmol,
40%) as an oil. 1H NMR (CDCl3) d 1.42–1.86 (4H,
m), 2.76 (2H, d, J = 6 Hz), 3.41–3.54 (2H, m), 3.69–
3.79 (1H, m), 3.85–3.94 (1H, m), 4.38 (1H, d,
J = 12 Hz), 4.43 (1H, d, J = 12 Hz), 7.12–7.27 (4H, m),
7.60–7.66 (1H, m), 8.12–8.16 (1H, m), 8.35–8.38 (1H,
m), 8.60–8.62 (2H, m), 9.27 (1H, s). MS (ESI) m/z 446
Et3N (50 lL, 0.36 mmol), 4-methylisoquinoline-5-sulfo-
nylchloride42 (30 mg, 0.12 mmol) and DMAP (2 mg,
0.016 mmol) were added to 59 (40 mg, 0.12 mmol) in
CH2Cl2 (2 mL). The mixture was stirred for 72 h at rt.

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1271
Flash silica chromatography (5% MeOH–CH2Cl2) gave
(2S,4S)-2-(4-chlorobenzyloxymethyl)-4-(4-methylisoqui
noline-5-sulfonylamino)-pyrrolidine-1-carboxylic acid
tert-butyl ester (60) (67 mg, 0.12 mmol, 100%). 1H
NMR (250 MHz, CDCl3) d 1.45 (9H, s), 1.60–1.80
(1H, m), 1.95–2.20 (1H, m), 2.52 (1H, br s), 3.05 (3H,
s), 3.47–3.60 (3H, m), 3.90–4.30 (3H, m), 4.60–4.75
(2H, m), 6.80–7.50 (4H, m), 7.57 (1H, dd, J = 7.5,
7.5 Hz), 8.17 (1H, d, J = 10 Hz), 8.44 (1H, d,
J = 7.5 Hz), 8.55 (1H, s), 9.16 (1H, s). MS (ESI ꢁve
ion) m/z 544 [(MꢁH)ꢁ].
(1 mL). The mixture was stirred for 5 min and then
SEM-Cl (200 lL, 1.1 mmol) was added. The mixture
was stirred for 90 min and then poured into CH2Cl2
(75 mL). The solution was washed with saturated aq
NH4Cl (25 mL). The organic layer was dried and
concentrated. Flash silica chromatography (30%
EtOAc–hexane) gave (2S,4S)-2-(4-chlorobenzyloxym-
ethyl)-4-[(1-chloroisoquinoline-5-sulfonyl)-(2-trimethylsi-
lanyleth-oxymethyl)amino]-pyrrolidine-1-carboxylic acid
tert-butyl ester (63) (0.44 g, quantitative). 1H NMR
(250 MHz, CDCl3) d 0.00 (9H, s), 0.82 (2H, t, J = 8 Hz),
1.44 (9H, s), 2.05–2.30 (2H, m), 3.20 (1H, dd, J = 10.4,
10.6 Hz), 3.40–4.00 (5H, m), 4.16 (1H, m), 4.48 (2H, s),
4.93 (1H, d, J = 11 Hz), 4.99 (1H, d, J = 11 Hz), 7.24
(2H, d, J = 8 Hz), 7.34 (2H, d, J = 8 Hz), 7.81 (1H, dd,
J = 8, 7.5 Hz), 8.42 (1H, d, J = 6 Hz), 8.43 (1H, d,
J = 6 Hz), 8.53 (1H, d, J = 7.5 Hz), 8.70 (1H, d,
J = 8 Hz). MS (ESI) m/z 718 [(M+Na)+].
Four molar HCl in dioxane (1 mL) was added to 60
(58 mg, 0.106 mmol). The mixture was stirred at rt for
1 h and then diluted with Et2O (4 mL). The liquid was
decanted and the solid residue was triturated and
washed with a further portion of Et2O (4 mL). The solid
was dissolved in MeOH and purified by ion exchange on
acidic resin (MeOH, then 2 M NH3 in MeOH) to give 61
(38 mg, 0.085 mmol, 81%). 1H NMR (250 MHz, CDCl3)
1.72–1.85 (1H, m), 2.20–2.35 (1H, m), 2.93 (1H, dd,
J = 5, 11 Hz), 3.06 (3H, s), 3.20 (1H, d, J = 11 Hz),
3.41–3.60 (2H, m), 3.65 (1H, dd, J = 3, 9 Hz), 3.90–
3.97 (1H, m), 4.55 (1H, d, J = 12 Hz), 4.63 (1H,
J = 12 Hz), 7.29–7.31 (4H, m), 7.60 (1H, dd, J = 8,
8 Hz), 8.17 (1H, dd, J = 1, 8 Hz), 8.51 (1H, dd, J = 1,
8 Hz), 8.56 (1H, d, J = 1 Hz), 9.17 (1H, s). MS (ESI)
m/z 446 [(M+H)+]. HRMS M+H+ calcd for
C22H25ClN3O3S 446.1300; found 446.1298. HPLC tR
5.04 min; purity (AUC) 95%.
KOSiMe3 (0.16 g, 1.3 mmol) was added to a solution of
63 (0.102 g, 0.146 mmol) in THF (2 mL). The solution
was stirred at 80 ꢁC for 2 h, then cooled to rt and parti-
tioned between CH2Cl2 (25 mL) and saturated aq
NH4Cl (250 mL). The aqueous layer was extracted with
CH2Cl2 (2 · 25 mL) and the combined organic extracts
were dried and concentrated. Flash silica chromatogra-
phy (50–100% EtOAc–hexane) gave (2S,4S)-2-(4-chlo-
robenzyloxymethyl)-4-[(1-hydroxyisoquinoline-5-sulfon-
yl)-(2-trimethylsilanylethoxymethyl)amino]-pyrrolidine-
1-carboxylic acid tert-butyl ester (64) (62 mg,
1
0.092 mmol, 63%). H NMR (250 MHz, CDCl3) d 0.00
7.3.27. 1-Hydroxyisoquinoline-5-sulfonic acid (2S,4S)-(2-
(65).
(9H, s), 0.85 (2H, t, J = 8 Hz), 1.42 (9H, s), 2.04–2.35
(1H, m), 3.21 (1H, t, J = 11 Hz), 3.40–3.95 (6H, m),
4.47 (2H, s), 4.90 (1H, d, J = 11 Hz), 4.96 (1H, d,
J = 11 Hz), 7.20–7.35 (6H, m), 7.60 (1H, t, J = 8 Hz),
8.39 (1H, dd, J = 1, 8 Hz), 8.70 (1H, dd, J = 1, 8 Hz),
11.32 (1H, br s). MS (ESI) m/z 700 [(M+Na)+].
(4-chlorobenzyloxymethyl)pyrrolidin-4-yl)amide
Chlorosulfonic acid (9 mL, 135 mmol) was added to 1-
chloroisoquinoline (3.38 g, 20 mmol) and the solution
was heated at 125 ꢁC for 72 h. After cooling to rt, the
solution was added dropwise to saturated aq NaHCO3
(500 mL) and the resulting precipitate was collected by
filtration. The solid was washed with water (100 mL)
and dried to give 1-chloroisoquinoline-5-sulfonyl chlo-
ride (4.21 g, 16 mmol, 80%). 1H NMR (250 MHz, d6-ac-
etone) d 8.11 (1H, dd, J = 10, 8 Hz), 8.51 (1H, dd, J = 6,
1 Hz), 8.65 (1H, d, J = 6 Hz), 8.76 (1H, dd, J = 8, 1 Hz),
8.91 (1H, d, J = 10 Hz).
HCl (1 mL, 4 M solution in dioxane) was added to a
solution of 64 (65 mg, 0.096 mmol) in MeOH (1 mL).
The solution was stirred at rt for 24 h and then diluted
with Et2O (10 mL). The supernatant was decanted and
the solid was washed with Et2O (10 mL). The solid
was purified by ion exchange on acidic resin (MeOH,
then 0.5 M NH3 in MeOH). Flash silica chromatogra-
phy (12.5% MeOH–CH2Cl2) gave 65 (10 mg,
0.022 mmol, 24%). 1H NMR (250 MHz, CD3OD) d
1.21 (1H, ddd, J = 8, 8, 13 Hz), 1.87 (1H, ddd, J = 8,
8, 13 Hz), 2.53 (1H, dd, J = 6, 11 Hz), 2.77 (1H, dd,
J = 7, 11 Hz), 3.21 (1H, m), 3.31 (1H, dd, J = 2, 5 Hz),
3.60 (1H, m), 4.36 (2H, s), 7.15–7.27 (6H, m), 7.54
(1H, dd, J = 8, 8 Hz), 8.23 (1H, dd, J = 1, 8 Hz), 8.49
(1H, dd, J = 1, 8 Hz). MS (ESI) m/z 448 [(M+H)+].
HRMS M+H+ calcd for C21H23ClN3O4S 448.1092;
found 448.1097. HPLC tR 4.70 min; purity (AUC) 95%.
To a solution of 59 (0.32 g, 0.94 mmol) in CH2Cl2 (5 mL)
were added Et3N (300 lL, 2.1 mmol) and then 1-chloro-
isoquinoline-5-sulfonyl chloride (0.25 g, 0.94 mmol) and
DMAP (11 mg, 0.09 mmol). The solution was stirred at
rt for 72 h. Flash silica chromatography (5% MeOH–
CH2Cl2) gave (2S,4S)-2-(4-chlorobenzyloxymethyl)-4-
(1-chloroisoquinoline-5-sulfonylamino)-pyrrolidine-1-
carboxylic acid tert-butyl ester (62) (0.44 g, 0.78 mmol,
1
83%). H NMR (250 MHz, CDCl3) d 1.29 (9H, s), 1.52
(1H, m), 2.24 (1H, m), 3.03 (1H, d, J = 12 Hz), 3.15–
3.40 (2H, m), 3.70–4.10 (3H, m), 4.51 (2H, s), 7.20–7.33
(4H, m), 7.65 (1H, dd, J = 8.5, 8 Hz), 7.96 (1H, d,
J = 6 Hz), 8.21 (1H, d, J = 6 Hz), 8.32 (1H, dd, J = 1,
8 Hz), 8.54 (1H, d, J = 8.5 Hz). MS (ESI) m/z 588
[(M+Na)+].
Acknowledgments
This work was supported by Cancer Research UK
[CUK] Grant No. C309/A2187. Paul Workman is a
Cancer Research UK Life Fellow. The authors thank
Christopher Bond for preliminary synthetic studies on
NaH (60% dispersion in oil, 0.128 g, 3.2 mmol) was
added to a solution of 62 (0.34 g, 0.60 mmol) in THF

1272
I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
13–17, 2005; American Chemical Society: Washington,
DC, 2005; MEDI-143.
the preparation of compound 65, Wendy Blakemore for
assistance with X-ray data collection, Marcel Verdonk
for assistance with molecular modelling studies, Brent
Graham for PKA purification, and Amin Mirza and
Meirion Richards for assistance with compound
characterisation.
14. Gong, J.; Woods, K. W.; Li, T.; Fisher, J.; Packard, G.;
Gandhi, V. B.; Claibone, A.; Luo, Y.; Shi, Y.; Liu, X.;
Klinghofer, V.; Bouska, J.; Shoemaker, A.; Oleksijew, A.;
Jarvis, K.; Stoll, V. S.; Hutchins, C.; De Jong, R.;
Oltersdorf, T.; Li, Q.; Rosenberg, S. H.; Giranda, V. L.;
Zhu, G.-D. Abstracts of Papers, 229th National Meeting
of the American Chemical Society, San Diego, CA, March
13–17, 2005; American Chemical Society: Washington,
DC, 2005; MEDI-142.
15. Giranda, V.; Luo, Y.; Li, Q.; Shoemaker, A.; de Jong,
R.; Liu, X.; Han, E.; Woods, K.; Thomas, S.; Rosen-
berg, S. Eur. J. Cancer Suppl. 2004, 8, 76, Abstracts of
the 16th EORTC-NCI-AACR Symposium on Molecular
Targets and Cancer Therapeutics, Geneva, Switzerland,
2004.
Supplementary material
Supplementary data associated with this article can be
found in the online version at doi:10.1016/
j.bmc.2005.09.055.
References and notes
16. McDonald, E.; Faria De Fonseca, T.; Bavetsias, V.;
Caldwell, J.; Wyatt, P. G.; Berdini, V. Patent Appl.
WO2005011697, 2005; Chem. Abstr. 2005, 142, 219156.
17. Ono-Saito, N.; Niki, I.; Hidaka, H. Pharmacol. Ther.
1999, 82, 123.
18. Jones, P. F.; Jakubowicz, T.; Pitossi, F. J.; Maurer, F.;
Hemmings, B. A. Proc. Natl. Acad. Sci. U.S.A. 1991, 88,
4171.
1. Brazil, D. P.; Hemmings, B. A. Trends Biochem. Sci. 2001,
26, 657.
2. Hill, M. M.; Hemmings, B. A. Pharmacol. Ther. 2002, 93,
243.
3. Li, Q.; Zhu, G. Curr. Topics Med. Chem. 2002, 2, 939.
4. Vara, J. A. F.; Casado, E.; de Castro, J.; Cejas, P.; Belda-
Inestia, C.; Gonazalez-Baron, M. Cancer Treatment Rev.
2004, 30, 193.
5. Brazil, D. P.; Yang, Z.-Z.; Hemmings, B. A. Trends
Biochem. Sci. 2004, 29, 233.
19. Gassel, M.; Breitenlechner, C. B.; Ruger, P.; Jucknischke,
U.; Schneider, T.; Huber, R.; Bossemeyer, D.; Engh, R. A.
J. Mol. Biol. 2003, 329, 1021.
6. Lindsley, C. W.; Zhao, Z.; Leister, W. H.; Robinson, R.
G.; Barnett, S. F.; Defeo-Jones, D.; Jones, R. E.;
Hartman, G. D.; Huff, J. R.; Huber, H. E.; Duggan, M.
E. Bioorg. Med. Chem Lett. 2005, 15, 761.
7. Barnett, S. F.; Defeo-Jones, D.; Fu, S.; Hancock, P. J.;
Haskell, K. M.; Jones, R. E.; Kahana, J. A.; Kral, A. M.;
Leander, K.; Lee, L. L.; Malinowski, J.; McAvoy, E. M.;
Nahas, D. D.; Robonson, R. G.; Huber, H. E. Biochem. J.
2005, 385, 399.
20. Yang, J.; Cron, P.; Good, V. M.; Thompson, V.; Hem-
mings, B. A.; Barford, D. B. Nat. Struct. Biol. 2002, 9, 940.
21. Akamine, P.; Madhusudan; Brunton, L. L.; Ou, H. D.;
Canaves, J. M.; Xuong, N.; Taylor, S. S. Biochemistry
2004, 43, 85.
22. Forino, M.; Jung, D.; Easton, J. B.; Houghton, P. J.;
Pellecchia, M. J. Med. Chem. 2005, 48, 2278.
23. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor,
R. J. Mol. Biol. 1997, 267, 727.
8. Yang, L.; Dan, H. C.; Sun, M.; Liu, Q.; Sun, X.; Feldman,
R. I.; Hamilton, A. D.; Polokoff, M.; Nicosia, S. V.;
Herlyn, M.; Sebti, S. M.; Cheng, J. Q. Cancer Res. 2004,
64, 4394.
9. Breitenlechner, C. B.; Wegge, T.; Berillon, L.; Graul, K.;
Marzenell, K.; Friebe, W.; Thomas, U.; Schumacher, R.;
Huber, R.; Engh, R. A.; Masjost, B. J. Med. Chem. 2004,
47, 1375.
10. Breitenlechner, C. B.; Friebe, W.; Brunet, E.; Werner, G.;
Graul, K.; Thomas, U.; Kunkele, K.; Schafer, W.; Gassel,
M.; Bossemeyer, D.; Huber, R.; Engh, R. A.; Masjost, B.
J. Med. Chem. 2005, 48, 163.
11. Reuveni, H.; Livnah, N.; Geiger, T.; Klein, S.; Ohne, O.;
Cohen, I.; Benhar, M.; Gellerman, G.; Levitzki, A.
Biochemistry 2002, 41, 10304.
24. Verdonk, M. L.; Cole, J. C.; Hartshorn, M. J.; Murray, C.
W.; Taylor, R. D. Proteins 2003, 52, 609.
25. Eldridge, M.; Murray, C. W.; Auton, T. A.; Paolini, G. V.;
Lee, R. P. J. Comput-Aided Mol. Design 1997, 11, 425.
26. Verdonk, M. L.; Berdini, V.; Hartshorn, M. J.; Mooij, W.
T.; Murray, C. W.; Taylor, R. D.; Watson, P. J. Chem.
Inf. Comput. Sci. 2004, 44, 793.
27. Engh, R. A.; Girod, A.; Kinzel, V.; Huber, R.; Bosse-
meyer, D. J. Biol. Chem. 1996, 271, 26157.
28. Watson, P.; Verdonk, M.; Hartshorn, M. J. Mol. Graph
Model. 2003, 22, 71.
29. Morikawa, A.; Sone, T.; Asano, T. J. Med. Chem. 1989,
32, 42.
30. Morikawa, A.; Sone, T.; Asano, T. J. Med. Chem. 1989,
32, 46.
12. Gong, J.; Woods, K. W.; Li, T.; Fisher, J.; Packard, G.;
Gandhi, V. B.; Claibone, A.; Luo, Y.; Shi, Y.; Liu, X.;
Klinghofer, V.; Bouska, J.; Shoemaker, A.; Oleksijew, A.;
Jarvis, K.; Stoll, V. S.; Hutchins, C.; De Jong, R.;
Oltersdorf, T.; Li, Q.; Rosenberg, S. H.; Giranda, V. L.;
Zhu, G.-D. Abstracts of Papers, 229th National Meeting
of the American Chemical Society, San Diego, CA, March
13–17, 2005; American Chemical Society: Washington,
DC, 2005; MEDI-142.
13. Gandhi, V. B.; Gong, J.; Li, T.; Woods, K. W.; Fisher, J.;
Packard, G.; Song, X.; Luo, Y.; Shi, Y.; Liu, X.;
Klinghofer, V.; Bouska, J.; Shoemaker, A.; Oleksijew,
A.; Jarvis, K.; Stoll, V. S.; Park, C.; De Jong, R.;
Oltersdorf, T.; Li, Q.; Rosenberg, S. H.; Giranda, V.; Zhu,
G.-D. Abstracts of Papers, 229th National Meeting of
the American Chemical Society, San Diego, CA, March
31. Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am. Chem
Soc. 1973, 93, 2897.
32. Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4156.
33. Takano, D.; Nagamitsu, T.; Hideaki, U.; Shiomi, K.;
Yamaguchi, Y.; Masuma, R.; Kuwajima, I.; Omura, S.
Tetrahedron Lett. 2001, 42, 3017.
34. Czarniecki, M. F.; Bauer, B. E. US Patent 4, 857, 301, 198;
. Chem. Abstr 1990, 112, 165007.
35. Gubert, S.; Braojos, C.; Sacristan, A.; Ortiz, J. A.
Synthesis 1991, 318.
36. Yang, B. V.; OÕRourke, D.; Li, J. Synlett 1993, 195.
37. Davis, F. A.; Srirajan, V. J. Org. Chem. 2000, 65, 3248.
38. Zhang, X.; Schmitt, A. C.; Jiang, W. Tetrahedron Lett.
2001, 42, 5335.
39. Jain, S.; Sujatha, K.; Krishna, K. V. R.; Roy, R.; Singh, J.;
Anand, N. Tetrahedron 1992, 48, 4985.

I. Collins et al. / Bioorg. Med. Chem. 14 (2006) 1255–1273
1273
40. Shigeyasu, M.; Kuwahara, M.; Sisido, M.; Ishikawa, T.
Chem. Lett. 2001, 634.
41. DÕCosta, M.; Kumar, V.; Ganesh, K. N. Tetrahedron Lett.
2002, 43, 883.
42. Hidaka, H.; Matsuura, A.; Matsuzaki, T. US Patent
6,153,608, 200; . Chem. Abstr. 1997, 127, 190760.
43. Hidaka, H.; Sone, T. EP Patent 187371, 1986; . Chem.
Abstr. 1987, 106, 32864.
44. Krapcho, P. A.; Waterhouse, D. Synth. Commun. 1998,
3415.
Amore, B. M.; Wilson, A. G. E.; Stouten, P. F. W.;
Burton, P. S. J. Med. Chem. 2004, 47, 6104.
55. King, F. D. In Medicinal Chemistry Principles and
Practice; King, F. D., Ed., 2nd ed.; Royal Society of
Chemistry: London, 2002; pp 342–347.
56. Sasaki, Y.; Suzuki, M.; Hidaka, H. Pharmacol Ther. 2002,
93, 225.
57. Utsunomiya, T.; Satoh, S.; Ikegagi, I.; Toshima, Y.; Asano,
T.; Shimokawa, H. Br. J. Pharmacol. 2001, 134, 1724.
58. Breitenlechner, C.; Gassel, M.; Hidaka, H.; Kinzel, V.;
Huber, R.; Engh, R. A.; Dossemeyer, D. Structure 2003,
11, 1595.
45. Herberg, F. W.; Bell, S. M.; Taylor, S. S. Protein Eng.
1993, 6, 771.
46. Leslie, A. G. W.; Brick, P.; Wonacott, A. Daresbury Lab.
Inf. Quart. Protein Crystallogr. 2004, 18, 33.
47. Hartshorn, M. J.; Murray, C. W.; Cleasby, A.; Freder-
ickson, M.; Tickle, I. J.; Jhoti, H. J. Med. Chem. 2005, 48,
403.
59. Taylor, S. S.; Radzio-Andzelm, E.; Madhusudan Cheng,
X.; Ten Eyck, L.; Narayana, N. Pharmacol. Ther. 1999,
82, 133.
60. Engh, R. A.; Bossemeyer, D. Pharmacol. Ther. 2002, 93,
99.
48. Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.;
Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E.
Nucleic Acids Res. 2000, 28, 235.
61. Doig, A. J.; Sternberg, M. J. E. Protein Sci. 1995, 4, 2247.
62. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P.
J. Adv. Drug Delivery Res. 1997, 23, 3.
49. Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Acta
Crystallogr. D 2004, 53, 240.
63. Vlietstra, R. J.; van Alewijk, D. C.; Hermans, K. G.; van
Steenbrugge, G. J.; Trapman, J. Cancer. Res. 1998, 58,
2720.
50. Andjelkovic, M.; Maira, S.-M.; Cron, P.; Parker, P. J.;
Hemmings, B. A. Mol. Cell Biol. 1999, 19, 5061.
51. Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMa-
hon, J.; Vistica, D.; Warren, J. T.; Bokesch, H.; Kenney, S.;
Boyd, M. R. J. Natl. Cancer Inst. 1990, 82, 1107.
52. Veber, D. F.; Johnson, S. R.; Cheng, H.-Y.; Smith, B. R.;
Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 45, 2615.
53. Vieth, M.; Siegel, M. G.; Higgs, R. E.; Watson, I. A.;
Robertson, D. H.; Savin, K. A.; Durst, G. L.; Hipskind, P.
A. J. Med. Chem. 2004, 47, 224.
64. IC50 Profiler ExpressTM, Upstate Co.
65. Garrett, M. D.; Hunter, L. J.; Fonseca, T.; Caldwell, J.;
Rowlands, M. G.; Hardcastle, A.; Collins, I.; McDonald,
E.; Thompson, N.; Workman, P. Eur. J. Cancer. Suppl.
2004, 8, 98, Abstracts of the 16th EORTC-NCI-AACR
Symposium on Molecular Targets and Cancer Therapeu-
tics, Geneva, Switzerland, 2004.
66. Davies, S. P.; Reddy, H.; Caivano, M.; Cohen, P.
Biochem. J. 2000, 351, 95–105.
54. Lu, J. J.; Crimin, K.; Goodwin, J. T.; Crivori, P.;
Orrenius, C.; Xing, L.; Tandler, P. J.; Vidmar, T. J.;
67. Knight, Z. A.; Shokat, K. M. Chem. Biol. 2005, 12, 621–
637.