Effect of linker length and composition on heterobivalent ligand-mediated receptor cross-talk between the A1 adenosine and β2 adrenergic receptors

Article abstract of DOI:10.1002/cmdc.201300286

Heterobivalent ligands that possess pharmacophores designed to interact with both the A₁ adenosine receptor (A₁AR) and the β₂ adrenergic receptor (β₂AR) were prepared. More specifically, these ligands contain an adenosine moiety that is linked via its N⁶-position to the amino group of the saligenin-substituted ethanolamine moiety present in the well-known β₂AR agonist, salbutamol. The affinities of these ligands were determined at both receptors and found to vary with linker length and composition. With all compounds, affinity and functional potencies were found to have selectivity for the A ₁AR over the β₂AR. In all cases, cAMP accumulation (a β₂AR-mediated response) was mainly observed when the A ₁AR was blocked or its function decreased by pertussis toxin or chronic agonist treatment. This suggests that heterobivalent compounds for receptors that mediate opposite responses might be useful for elucidating the mechanisms of receptor cross-talk and how this interaction, in terms of responsiveness, may change under pathophysiological conditions. A molecular double date: Heterobivalent ligands containing pharmacophores designed to interact with both the A₁ adenosine receptor (A₁AR) and the β₂ adrenergic receptor (β₂AR) were prepared. The affinity and potency of these ligands at both receptors were found to be dependent upon the linker length and composition. The data suggest that heterobivalent ligands for receptors mediating opposite responses might be useful for investigating the regulation of receptor cross-talk in health and disease. Copyright

Full text of DOI:10.1002/cmdc.201300286

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201300286
Effect of Linker Length and Composition on
Heterobivalent Ligand-Mediated Receptor Cross-Talk
between the A1 Adenosine and b2 Adrenergic Receptors
Nicholas Barlow,[a] Stephen P. Baker,*[b] and Peter J. Scammells*[a]
Heterobivalent ligands that possess pharmacophores designed
to interact with both the A1 adenosine receptor (A1AR) and the
b2 adrenergic receptor (b2AR) were prepared. More specifically,
these ligands contain an adenosine moiety that is linked via its
N6-position to the amino group of the saligenin-substituted
ethanolamine moiety present in the well-known b2AR agonist,
salbutamol. The affinities of these ligands were determined at
both receptors and found to vary with linker length and com-
position. With all compounds, affinity and functional potencies
were found to have selectivity for the A1AR over the b2AR. In
all cases, cAMP accumulation (a b2AR-mediated response) was
mainly observed when the A1AR was blocked or its function
decreased by pertussis toxin or chronic agonist treatment. This
suggests that heterobivalent compounds for receptors that
mediate opposite responses might be useful for elucidating
the mechanisms of receptor cross-talk and how this interac-
tion, in terms of responsiveness, may change under pathophy-
siological conditions.
Introduction
Activation of membrane-bound receptors, such as G protein-
coupled receptors (GPCRs), ligand-gated ion channels and cat-
alytic receptors, results in a diverse range of outcomes de-
pending on the nature of the ligand, as well as the associated
effector domains or protein signalling partners. All play a vital
role in cell regulation and the coordination of intracellular
events with broader biological function, and their ubiquitous
nature has made them the targets for over 40% of marketed
pharmaceuticals.[1,2] Endogenous ligands for these receptors in-
clude small molecules, peptides and metal ions.
phores joined via a linker or spacer moiety, and might be able
to bind in a bridging or top–tail fashion, providing simultane-
ous binding in a manner that might not be possible by coad-
ministration of two independent drugs. Simultaneous binding
at two different receptors has the potential to result in unique
intracellular signalling responses and consequently unique
downstream effects through altered recruitment of shared in-
tracellular signalling partners.
Our research focuses on targeting the b2 adrenergic receptor
(b2AR) and the A1 adenosine receptor (A1AR). On activation,
both of these GPCRs cause differential effects on the trans-
membrane enzyme, adenylyl cyclase, which is responsible for
catalysing the conversion of ATP to the secondary messenger
cAMP. Agonist binding at these receptors increases the rate of
trimeric G-protein dissociation, with the Ga subunit going on
to interact with adenylyl cyclase. The nature of the particular
Ga subunit isoform determines the subsequent signalling cas-
cade that is recruited. In the case of the b2AR, signalling via
the Gas subunit causes activation of adenylyl cyclase, whereas
the A1AR signals via the Gai subunit, which is inhibitory to-
wards adenylyl cyclase. Heterobivalent ligands that can simul-
taneously deliver an agonist moiety at each receptor can po-
tentially modify the opposing effects on adenylyl cyclase to
achieve a desired response. For example, the specific targeting
of the bAR/A1AR system arises from the observation in the car-
diovascular system, where enhanced A1AR activation might
overly attenuate normal b adrenergic responsiveness or pre-
vent some adverse effects from exogenously administered cat-
echolamines.[9,10]
The complex interplay between intracellular signalling cas-
cades that can share downstream signalling partners allows for
potential receptor cross-talk. In this manner, activation of one
receptor might result in the altered signalling of others, allow-
ing unique downstream signalling outcomes dependent on
the cell environment and activating ligand. However, whilst si-
multaneous activation of more than one membrane receptor is
a common endogenous event, drugs targeting membrane re-
ceptors are often designed to be as selective as possible, in
order to minimise off-target effects.
More recently, heterobivalent ligands have been investigated
as a means of providing novel pharmacological outcomes
through an ability to bind concomitantly to two different
target receptors.[3–8] Such ligands incorporate two pharmaco-
[a] N. Barlow, Prof. P. J. Scammells
Medicinal Chemistry, Monash Institute of Pharmaceutical Sciences
Monash University, 381 Royal Parade, Parkville VIC 3052 (Australia)
E-mail: peter.scammells@monash.edu
[b] Prof. S. P. Baker
Previous work reported by our group has generated hetero-
bivalent b2AR–A1AR ligands with full agonist profiles at both
receptors, compared with known classical ligands.[3] Such li-
gands comprised of an adenosine moiety linked through the
Department of Pharmacology & Therapeutics
University of Florida College of Medicine
Box 100267, Gainesville, FL 32610 (USA)
E-mail: spbaker@ufl.edu
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
1
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
N6-position via a short alkyl linker of 2, 4 or 6 carbon atoms to
the amino component of
a
catecholamine derivative
(Figure 1).[3] A biphasic cAMP response was observed when
Figure 1. Heterobivalent ligands based on adenosine and fromoterol.
these compounds were presented to DDT1 MF-2 cells express-
ing both receptors. At low ligand concentrations, b2AR-mediat-
ed stimulation of cAMP accumulation dominated, due to the
greater potency imparted by this pharmacophore. In contrast,
A1AR inhibition of cAMP accumulation dominated at higher
ligand concentrations, due to higher population of A1ARs and
dominance of Gi over Gs.
The effect of linker lengths in bivalent ligands that allow
bridging of two receptors has been discussed by numerous au-
thors.[6,7,11–14] Reports by Portoghese and co-workers of an opti-
mal linker length that allows bridging active sites in opioid
dimers[4,5] led us to consider similar linker lengths in our own
bivalent designs. Furthermore, the effect of linker type had not
been explored in our previous work. The current investigation
aims to look at the effect of linker-related structural considera-
tions on potential b2AR and A1AR cross-talk.
Figure 2. Heterobivalent target compounds with linked A1R and b2AR
agonist pharmacophores.
Results and Discussion
Ligand design
proved solubility, but also resulted in greatly decreased A1AR
affinity and potency.[3] Accordingly, in the current study linkers
were designed with an alkyl chain attached to the N6-position
of the adenosine moiety (to maintain A1AR potency), but also
included a PEG component distal to the adenosine moiety to
promote better solubility. A “monovalent control” comprised
of the b2AR pharmacophore joined to only linker was also pre-
pared for pharmacology studies.
Bivalent ligands were designed to incorporate an A1AR agonist
pharmacophore and a b2AR agonist pharmacophore, separated
by a linker moiety (Figure 2). Adenosine (A) was chosen as the
A1AR agonist and appended to a linker moiety at the N6-posi-
tion. The incorporation of an alkyl group in the N6-position is
known to enhance A1AR affinity and selectivity,[15] and alkyl
linkers in this position have been used to append a wide
range of moieties to adenosine.[16] The saligenin-based ethanol-
amine moiety (B) present in the well-known b2AR agonist, sal-
butamol, was chosen as the b2AR agonist pharmacophore and
appended to the linker moiety through the side chain amine.
Linkers were designed to be between 19 and 38 atoms in
chain length to provide for the possibility of cooperative bind-
ing. As diaminoalkanes are only commercially available in
lengths of up to 12 methylene units, longer linker lengths
were achieved by coupling diaminoalkanes to amino acids via
amide bond formation. In our previous study,[3] we found that
bivalent b2AR–A1AR ligands with lengthy alkyl linkers pos-
sessed poor solubility properties. Replacement of the alkyl
chain with the corresponding polyethylene glycol (PEG) linker
attached to N6 atom of the adenosine pharmacophore (A) im-
Chemistry
Bivalent ligands were prepared via a convergent approach
where the b2AR agonist moiety and A1AR agonist moiety were
prepared separately, then combined in the penultimate step of
the synthesis.
The b2AR agonist component of these ligands was prepared
in six steps, starting with commercially available 1-aminodode-
canoic acid (2) (Scheme 1). Reductive alkylation of 2 with ben-
zaldehyde under basic conditions afforded the benzylamine 3.
An ethanolic solution of 3 was then treated with oxalyl chlo-
ride to give the corresponding ethyl ester (4). Alkylation of the
secondary amine moiety of 4 with 1-(4-(benzyloxy)-3-(hydroxy-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
2
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
8a and 8b were prepared, using previously reported method-
ology, by treating 7 with the appropriate diaminoalkane.[18] An-
alogues with longer linkers were prepared by attachment of
either 1-aminohexanoic acid or 12-aminododecanoic acid to
the purine riboside core through an SNAr displacement of the
6-chloro group to give 10a and 10b, respectively. Compound
10b was extended via BOP-mediated coupling to either 1,6-
aminohexane or 1,12-diaminododecane to provide 11 a and
11 b, respectively. Analogues with PEG chains were prepared
similarly by BOP-mediated coupling of 2,2’-(ethylenedioxy)die-
thylamine to 10a and 10b, affording 13a and 13b.
The A1AR agonist and b2AR agonist components were then
combined via a BOP-mediated amide coupling reaction. Subse-
quent cleavage of the N-benzyl protecting group by hydroge-
nolysis provided the desired dual pharmacophore compounds
(Scheme 2). This reaction sequence was repeated, combining
different A1AR agonist units to the b2AR agonist component
(6) to provide the target compounds 9a, 9b, 12a, 12b, 14a
and 14b.
A “monovalent” compound, comprised of b2AR agonist com-
ponent 6 and attached “linker” (but no adenosine moiety), was
also prepared as a control. This compound (16) was obtained
by BOP-mediated coupling of hexylamine with 6, followed by
N-debenzylation, as outlined in Scheme 3.
Scheme 1. Reagents and conditions: a) NaOH, benzaldehyde, MeOH, RT, 2 h;
b) NaBH4, EtOH, 08C!RT, 15 h; c) oxalyl chloride, EtOH, ꢀ778C, then DMF,
RT, 15 h; d) 1-(4-(benzyloxy)-3-(hydroxymethyl)phenyl)-2-bromoethanone,
DIPEA, CHCl3, 15 h, RT; e) NaBH4, EtOH, 08C!RT, 1 h; f) NaOH, MeOH, CH2Cl2,
RT, 15 h.
methyl)phenyl)-2-bromoethanone[17] with direct sodium boro-
hydride reduction of the resulting a-amino ketone provided
amino alcohol 5. Saponification of the ester moiety of 5 with
sodium hydroxide in a solution of water, methanol and di-
chloromethane provided the desired b2AR agonist moiety (6).
A1AR agonist pharmacophores with a linker moiety contain-
ing a distal primary amine were prepared from commercially
available 6-chloropurine riboside (7) (Scheme 2). Compounds
Scheme 2. Reagents and conditions: a) 1,6-diaminohexane or 1,12-diaminododecane, tBuOH, reflux, 15 h; b) 6, BOP, Et3N, DMF, RT, 15 h; c) H2, Pd(OH)2, NH4OH,
MeOH, RT, 15 h; d) 6-aminohexanoic acid or 12-aminododecanoic acid, DIPEA, tBuOH, reflux, 24 h; e) 1,6-diaminohexane or 1,12-diaminododecane, Et3N, BOP,
DMF, RT, 15 h; f) 2,2’-(ethylenedioxy)diethylamine, BOP, RT, 15 h.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
3
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
At the A1AR, the bivalent compounds with 19 and 25 linker
atoms (9a and 9b, respectively) showed a 1.6- and 1.8-fold de-
crease in affinity, as compared with CPA, whereas the bivalents
with 32 and 37 linker atoms (12a and 12b, respectively)
showed only a very weak interaction with the receptor. In
terms of the bivalent ligands with a PEG component in the
linker (i.e., 14a and 14b), the compound with a 28-atom linker
had a 27-fold lower affinity compared with CPA, whereas the
compound with a 34-atom linker showed only a 3-fold de-
crease in affinity. All of the bivalent compounds, where deter-
minations were made, showed a higher affinity for the A1AR
compared with the b2AR. These ranged from a 2.9-fold higher
affinity at the A1AR for 9b to 62-fold for 9a.
The bivalent compounds were tested for their ability to in-
hibit and stimulate cAMP accumulation through the A1AR and
b2AR, respectively. As indicated in Table 1, all of these com-
pounds except 12b inhibited forskolin-stimulated cAMP accu-
mulation in a concentration-dependent manner. In these
assays, propranolol was added to prevent any ligand stimula-
tion of cAMP accumulation through the b2AR. The intrinsic ac-
tivities of these compounds are equal to or greater than 92%
of the maximal response produced by CPA indicating that they
are approaching full agonist activity at the A1AR. The bivalent
compound with a 38-atom linker (12b) did not inhibit cAMP
accumulation.
Scheme 3. Reagents and conditions: a) hexylamine, BOP, Et3N, DMF, RT, 15 h;
b) H2, Pd(OH)2, NH4OH, MeOH, RT, 15 h.
Pharmacology
Binding and cAMP functional assays were performed for both
bivalent and monovalent control compounds. The affinity of
each compound was determined by displacement binding
assays using [125I]-(ꢀ)iodopindolol for the b2AR and [3H]-8-cy-
clopentyl-1,3-dipropylxanthine (DPCPX) for the A1AR. The clas-
sical bAR agonist (ꢀ)-isoproterenol and A1AR agonist N6-cyclo-
pentyladenosine (CPA) were used for comparison. As shown in
Table 1, the monovalent b2AR-only ligand control (16) exhibit-
ed a decreased affinity (6.3-fold) for the receptor as compared
with (ꢀ)-isoproterenol. This decreased affinity might be attrib-
uted to the linker component, although a direct comparison of
the N-alkyl substituent is not possible due to the different phe-
nolic substituents. Bivalent compound 9a, with a 19-atom
linker, showed a 20.6-fold decrease in affinity compared with
the monovalent control (16). Increasing the linker to 25 atoms
(9b) resulted in an affinity similar to 16, whereas bivalent com-
pounds 12a and 12b with longer linkers of 32 and 38 atoms,
respectively, had little or no interaction with the b2AR at the
highest concentration used (100 mm). A different linker effect
on affinity was observed for the compounds that possessed
a PEG component in their linker (14a and 14b). Compound
14a, which possessed a 28-atom linker, displayed mid-micro-
molar affinity, which was increased by 16.6-fold when the
linker was extended to 34 atoms in compound 14b.
As shown in Figure 3, (ꢀ)-isoproterenol, salbutamol and
monovalent control 16 stimulated cAMP accumulation in a con-
centration-dependent manner. All of the bivalent compounds
alone produced little (<15% of the (ꢀ)-isoproterenol maxi-
mum) or no increase in cAMP accumulation. However, in the
presence of DPCPX, added to block the A1AR, a robust concen-
tration-dependent stimulation of cAMP was observed except
for 12b, which was inactive either in the presence or absence
of DPCPX (not shown on graph). The intrinsic activity of the
monovalent control (16) and bivalent compounds 9a, 9b, 12a
and 14b were less than that of (ꢀ)-isoproterenol (0.28–0.56)
Table 1. Intrinsic activities (IA), Ki and EC50 values of standard agonists and bivalent derivatives at the b2 adrenergic (b2A) and A1 adenosine (A1A) receptors.
Compd
l[a]
b2A receptor
EC50 [nm][c]
A1A receptor
Ki [nm][b]
IA[d]
1.00
Ki [nm][b]
EC50 [nm][c]
IA[d]
(ꢀ)-Isoproterenol
66ꢁ8
ND[e]
ND[e]
12ꢁ3
93ꢁ20
ND[e]
ND[e]
ND[e]
ND[e]
ND[e]
ND[e]
1.00
ND[e]
1.01ꢁ0.01
0.98ꢁ0.02
0.92ꢁ0.02
–
Salbutamol
CPA
0.97ꢁ0.09
ND[e]
ND[e]
85ꢁ29
ND[e]
137ꢁ39
154ꢁ39
>100000[f]
>100000[f]
2323ꢁ12
256ꢁ31
1.6ꢁ0.1
16
9a
9b
12a
12b
14a
14b
–
413ꢁ114
8514ꢁ1228
454ꢁ171
712ꢁ89
7789ꢁ1713
2827ꢁ620
7670ꢁ1698
inactive
NC[g]
4843ꢁ1052
0.56ꢁ0.03
0.43ꢁ0.05
0.47ꢁ0.05
0.28ꢁ0.03
–
ND[e]
19
25
32
38
28
34
168ꢁ20
624ꢁ88
1885ꢁ219
inactive
614ꢁ89
2226ꢁ371
>100000[f]
inactive
28820ꢁ9774
–
0.98ꢁ0.01
1739ꢁ226
0.48ꢁ0.07
0.98ꢁ0.03
[a] Linker length (l). [b] Ki values were calculated from the ligand concentration required to inhibit specific binding of [125I]-(ꢀ)iodopindolol to the b2AR or
[3H]-DPCPX to the A1AR by 50%. Values are the meanꢁSEM of n=3–7 separate determinations. [c] b2AR EC50 values are the ligand concentration required
to stimulate cAMP accumulation by 50% in the presence of DPCPX (10 mm), while A1AR EC50 values are the ligand concentration required to inhibit forsko-
lin (1.0 mm)-stimulated cAMP accumulation by 50% in the presence of propranolol (1.0 mm). [d] Intrinsic activity (IA) is the maximal stimulation (b2AR) or in-
hibition (A1AR) of cAMP accumulation as compared with (ꢀ)-isoproterenol or N6-cyclopentyladenosine (CPA), respectively, which are set at 1.00. Data are
the meanꢁSEM of n=3–7 separate determinations. [e] ND: not determined. [f] Less than 50% inhibition of radioligand binding at concentrations up to
100 mm. [g] NC: EC50 could not be determined as the stimulation did not approach a plateau at the highest concentration used (100 mm).
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
4
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
To further investigate a paradigm where bivalent ligand
stimulation of cAMP accumulation can be unmasked, the
effect of partial desensitization of the A1AR was investigated.
Compound 9b was chosen as its functional potency for the
two receptors was relatively close (4.5-fold), suggesting that
changes in agonist potency for the A1AR might unmask b2AR
stimulation. Pretreatment of DDT cells with 200 nm CPA for
18 h followed by six cell wash cycles to remove the CPA result-
ed in an 11.6-fold decrease in the EC50 value for CPA to inhibit
forskolin-stimulated cAMP accumulation without a change in
the maximal inhibition (control: EC50 =1.4ꢁ0.2; CPA pretreat-
ed: EC50 =16.3ꢁ4.1 nm, n=3). As depicted in Figure 4, com-
Figure 3. Effect of isoproterenol, salbutamol, compound 16 and the bivalent
derivatives on cAMP accumulation in DDT1 MF-2 cells. Cells were incubated
with rolipram (20 mm) and the indicated concentration of test compound for
10 min at 368C. In parallel assays with the bivalent compounds, DPCPX
(10 mm) was included. Each data point on the graphs is the meanꢁSEM of
n=3–7 separate experiments. The isoproterenol data is the same on both
graphs.
Figure 4. Effect of N6-cyclopentyladenosine (CPA) pretreatment on 9b-stimu-
lated cAMP accumulation in DDT1 MF-2 cells. Cells were incubated without
(open symbols) or with (filled symbols) CPA (200 nm) for 18 h and then
washed six times to remove the CPA. Cells were then incubated with roli-
pram (20 mm) and the indicated concentration of 9b without or with DPCPX
(1 mm) for 10 min at 368C. Each point on the graph is the meanꢁSEM n=4
separate experiments.
indicating that they are partial b2AR agonists (Table 1). Al-
though bivalent compound 14a also showed stimulation of
cAMP accumulation in the presence of DPCPX, no EC50 or in-
trinsic activity value could not be determined as the maximal
stimulation was not reached at the highest concentration used
(100 mm). The stimulation of cAMP accumulation by 50 mm of
each of the active bivalent compounds in the presence of
DPCPX was blocked by 10 mm propranolol, which is consistent
with this activity being mediated by the b2AR. In addition, pre-
treatment of DDT cells with pertussis toxin (100 ngmLꢀ1 for
18 h) to uncouple the A1AR from Gi and prevent the cAMP in-
hibitory effect of this receptor resulted in stimulation of cAMP
accumulation by the bivalent compounds alone, similar to that
observed in untreated cells in the presence of DPCPX (data not
shown). This finding is consistent with the inhibitory effect on
cAMP accumulation being mediated through the A1AR. Similar
to the affinity selectivity, the potency of the bivalent com-
pounds (EC50 values) was greater for the A1AR inhibition of
cAMP accumulation compared with the b2AR stimulatory re-
sponse.
pound 9b produced less than 20% stimulation of cAMP accu-
mulation alone, and this stimulation was greatly increased in
the presence of DPCPX. In CPA pretreated cells, 9b alone ini-
tially stimulated cAMP accumulation to about 75% of the max-
imum stimulation in the presence of DPCPX followed by a de-
crease at higher 9b concentrations. The stimulation of cAMP
accumulation in the presence of DPCPX was the same in both
control and CPA-pretreated cells. This data show that partial
desensitization of the A1AR can result in a substantial unmask-
ing of a bivalent b2AR response.
The current study extends our previous work showing that
covalent linking of pharmacophores for the b2AR and the A1AR
into a single compound can retain activity at both receptors.[3]
Using a limited series of compounds, the effects of extended
linker length and composition on affinity and agonist proper-
ties were investigated. In general, the bivalent compounds had
decreased affinities and functional potencies for each receptor
as compared with the classical bAR and A1AR agonists, with
greater decreases for the b2AR. Varying the linker length had
some interesting effects on the affinity of the bivalent com-
pounds for the b2AR. In terms of the bivalent ligands with link-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
5
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
ers consisting predominantly of methylene units (i.e., 9a, 9b,
12a and 12b), the compound that contains a 25-atom linker
(9b) showed optimal affinity, which was decreased when the
linker was either shortened (19 atoms) or lengthened (32 and
38 atoms). A similar effect related to linker length was ob-
served for functional potency at the b2AR. A different pattern
emerged with the A1AR, where both affinity and potency of
compounds 9a, 9b, 12a and 12b decreased with increasing
linker length. Previous studies have provided evidence for si-
multaneous bridging across two receptors by a single bivalent
molecule, partly based upon increases in ligand affinity for
their targets.[19–21] In our present work, the observation that
there was no parallel increase in affinity for the two receptors
with the bivalents 9a, 9b, 12a and 12b suggests that simulta-
neous bridging across two receptors did not occur.
GPCRs are under constant regulation, with their expression
and sensitivity being altered under changing conditions
through a variety of mechanisms. We sought to investigate if
one potential alteration can affect the response profile of a bi-
valent compound. In order explore this possibility the A1AR
system was desensitized by chronic treatment with an A1AR
agonist and a bivalent response subsequently determined. As
expected, desensitization not only decreased the potency with
which CPA inhibits cAMP accumulation, but also led to an un-
masking of the ability of a bivalent compound alone to stimu-
late cAMP accumulation. The unmasking of cAMP accumula-
tion occurred at the lower bivalent concentrations and was fol-
lowed by a decrease in activity at the higher concentrations
likely due to A1AR activation.
In a previous report, heterobivalent b2AR and A1AR agonists
with much shorter linkers had higher affinities for the b2AR or
similar affinities for the two receptors.[3] These compounds
alone produced biphasic cAMP responses with stimulation at
lower concentrations followed by inhibition as the concentra-
tions were increased. Although these compounds were full ag-
onists at both receptors, for several, the A1AR-mediated inhibi-
tory component imparted partial b2AR agonist properties. In
the present work, using a different b2AR pharmacophore and
extending the linker length resulted in A1AR selectivity with
a predictable masking of b-stimulation unless the A1AR was
blocked, uncoupled from its G protein or partly desensitized.
Furthermore, these bivalents were direct partial b2AR agonists.
Taken together, the data from these studies suggest that the
potency and response characteristics of heterobivalent com-
pounds for receptors producing opposite responses can be de-
signed for selective effect by altering the spacing and/or linker
composition between the two pharmacophore head groups.
Furthermore, these compounds might be useful for investiga-
tive studies on how receptor cross-talk for opposing responses
is regulated during health and disease.
In the bivalent ligands that possessed a PEG component in
the linker (14a and 14b), a 9- to 16-fold increase in affinity for
both receptors was observed when the linker length was in-
creased from 28 to 34 atoms. This could be consistent with
bridging across two receptors, though additional experimental
approaches would be needed to further establish if this is ac-
tually the case. Although the affinity increased with increasing
linker length, the potency for inhibiting cAMP accumulation
was decreased. This difference between the changes in affinity
and potency might be due to the composite nature of poten-
cy, which is dependent upon both affinity and efficacy (rela-
tionship between receptor occupancy and response) with effi-
cacy perhaps being a more dominant factor in the potency
change with increasing linker length. Ligand efficacy might
also contribute to the observed differences between the affini-
ty (Ki) and potency (EC50) values for the other bivalent com-
pounds tested (Table 1) and has been reported previously for
another series of bivalent ligands for the b2AR/A1AR.[3] Howev-
er, other factors might have contributed towards the observed
differences in affinity and potency. For example, the assay con-
ditions for measuring affinity (isolated membranes) and poten-
cy (intact cells) were different.
With one exception, the bivalent compounds alone had little
or no effect on cAMP accumulation in DDT cells. However,
when the A1AR was blocked with an antagonist or uncoupled
from its Gi protein with pertussis toxin, a robust stimulation of
cAMP accumulation was observed with the majority of com-
pounds. This stimulation appears to be mediated by activation
of the b2AR, as presence of the bAR antagonist, propranolol,
prevented the effect. Previous studies have shown that A1AR-
mediated inhibition of cAMP accumulation in DDT cells is dom-
inant over b2AR stimulation.[3,22] As these bivalent compounds
have higher affinity and functional potencies for the A1AR over
the b2AR, the lack of b2AR stimulation in the absence of DPCPX
or with G protein-uncoupled A1AR is likely due to their prefer-
ential binding to and activation of the inhibitory A1AR. Interest-
ingly, all of the active bivalent compounds were full agonists
at the A1AR, but partial agonists at the b2AR. This partial ago-
nist activity appears to involve the linker moiety, as salbuta-
mol-derived monovalent control 16 was also found to be a par-
tial agonist, whereas salbutamol is a full b-agonist.
Experimental Section
General procedures: Melting points (mp) were determined on
a Mettler Toledo MP50 melting point system and are uncorrected.
1H and 13C NMR spectra were recorded on Bruker Ultrashield 400
Plus at 400 MHz and 101 MHz, respectively. Analytical HPLC was
performed on a Waters Alliance 2690 fitted with a Waters 5996
photodiode array (PDA) detector and a Phenomenex Luna C8
column (5 mm, 100 ꢁ, 150ꢂ4.60 mm). Analyses were conducted
using a gradient of 0!64% MeCN in H2O over 10 min with 0.1%
TFA throughout. All reagents were of >95% purity. Preparatory
HPLC was performed on a Waters Prep LC 4000 system fitted with
a Waters 486 Tuneable Absorbance Detector and either a Phenom-
enex Luna C8 (10 mm, 100 ꢁ, 250ꢂ30 mm) column or a Phenomen-
ex Luna C18 (10 mm, 100 ꢁ, 50ꢂ21.2 mm) column. Low-resolution
mass spectrometry (LRMS) was performed on an Agilent 6120
single quadrapole LCMS system using electron spray ionization
(ESI). High-resolution mass spectrometry (HRMS) was performed on
a Waters Premier XE time-of-flight (ToF) mass spectrometer using
ESI.
12-(Benzylamino)dodecanoic acid (3): A suspension of 12-amino-
dodecanoic acid (2) (1 g, 4.6 mmol) in MeOH (25 mL) was treated
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
6
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
with freshly crushed NaOH (220 mg, 5.6 mmol). After the suspen-
sion dissolved, benzaldehyde (560 mL, 5.6 mmol) was added and
the solution stirred for 2 h at RT. The reaction was then cooled on
ice, and NaBH4 (200 mg, 5.3 mmol) was added batchwise. The reac-
tion was allowed to warm to RT and then stirred for an additional
15 h. The solution was adjusted to pH 6 with concd HCl(aq), and the
solvent was removed in vacuo. Acetone (50 mL) was added and
the suspension sonicated. The precipitate was isolated by filtration,
washing with water (2.0 mL) then acetone (2.0 mL), to provide the
title compound as a white solid (1.6 g, 94%): mp: 151–1538C;
1H NMR (400 MHz, CD3OD): d=7.47–7.37 (m, 5H), 4.08 (s, 2H), 2.87
(t, J=8.0 Hz, 2H), 2.21 (t, J=7.3 Hz, 2H), 1.72–1.19 ppm (m, 18H);
13C NMR (101 MHz, CD3OD): d=179.9, 131.9, 130.3, 130.0, 129.7,
51.7, 47.8, 36.4, 29.4, 29.3, 29.3, 29.2, 29.1, 28.9, 26.7, 26.4,
26.1 ppm; LRMS (ESI+): m/z (%): 306 [M+H]+ (100).
4.71–4.61 (m, 3H), 3.71 (d, J=13.4 Hz, 1H), 3.61 (d, J=13.4 Hz, 1H),
2.75–2.39 (m, 4H), 2.28 (t, J=7.4 Hz, 2H), 1.65–1.52 (m, 2H), 1.50–
1.38 (m, 2H), 1.33–1.12 ppm (m, 14H); 13C NMR (101 MHz, CD3OD):
d=178.5, 184.8, 168.7, 167.0, 164.8, 159.0, 158.4, 157.7, 157.4,
157.0, 156.5, 156.2, 155.5, 155.4, 140.7, 100.2, 99.2, 91.4, 89.5, 88.6,
83.5, 63.3, 58.8 (2C), 58.7 (2C), 58.5, 58.3, 56.5, 56.1, 54.2 ppm;
LRMS (ESI+): m/z (%): 562.5 [M+H]+ (100).
6-Chloropurine riboside (7) was commercially available. N6-(6-Ami-
nohexyl)adenosine (8a) and N6-(12-aminododecanyl)adenosine
(8b) were prepared using a previously reported methodology, by
treating 7 with the appropriate diaminoalkane.[18] Characterisation
data were as reported in the literature.
N6-(6-(12-((2-Hydroxy-2-(4-hydroxy-3-(hydroxymethyl)phenyl)e-
thyl)amino)dodecanamido)hexyl)adenosine (9a): A solution of
8a (231 mg, 0.63 mmol) in DMF (1 mL) was treated with 6
(154 mg, 0.27 mmol), DIPEA (430 mL, 2.4 mmol) and BOP (280 mg,
0.63 mmol). The reaction was stirred for 15 h at RT. The solution
was then diluted with EtOAc (100 mL) and washed with water (5ꢂ
50 mL), brine (50 mL), dried over MgSO4, filtered and concentrated
in vacuo. The resulting crude was then purified by flash chroma-
tography (MeOH/CH2Cl2, 1:20) to provide N6-(6-(12-((2-hydroxy-2-(4-
benzyloxy-3-(hydroxymethyl)phenyl)ethyl)benzylamino) dodecana-
mido)hexyl)adenosine as a colourless oil (66 mg, 26%): 1H NMR
(400 MHz, CD3OD): d=8.22 (s, 1H), 8.19 (s, 1H), 7.46–7.26 (m, 11H),
7.16 (dd, J=8.4, 2.2 Hz, 1H), 6.95 (d, J=8.4 Hz, 1H), 5.94 (d, J=
6.5 Hz, 1H), 5.11 (s, 2H), 4.76–4.69 (m, 4H), 4.68 (s, 2H), 4.32 (dd,
J=5.1, 2.5 Hz, 1H), 4.22–4.11 (m, 1H), 3.92 (d, J=13.4 Hz, 1H), 3.88
(dd, J=12.6, 2.5 Hz, 1H), 3.81 (d, J=13.4 Hz, 1H), 3.73 (dd, J=12.6,
2.5 Hz, 1H), 3.57 (brs, 2H), 3.16 (t, J=6.9 Hz, 2H), 2.86–2.59 (m,
4H), 2.15 (t, J=7.4 Hz, 2H), 1.75–1.16 ppm (m, 26H); 13C NMR
(101 MHz, CD3OD): d=176.3, 157.1, 156.8, 153.5, 149.7, 141.5,
138.8, 138.0, 135.9, 131.1, 130.9, 129.6, 129.5, 128.9, 128.8, 128.4,
127.3, 127.1, 121.3, 112.7, 91.3, 88.3, 75.5, 72.7, 71.2, 71.1, 63.6,
62.4, 60.4, 59.8, 55.3, 41.5, 40.3, 37.2, 30.6 (3C) 30.4 (3C) 30.2, 28.2,
27.7, 27.7, 27.1, 27.0 ppm; LRMS (ESI+): m/z (%): 910.5 [M+H]+
(100).
Ethyl 12-(benzylamino)dodecanoate (4): A solution of 3 (500 mg,
1.6 mmol) in abs EtOH at ꢀ778C was treated with oxalyl chloride
(500 mL, 5.8 mmol). The solution was stirred and allowed to come
to RT. DMF (1 drop) was added and the solution stirred at RT for
a further 15 h. The solvent was removed in vacuo, saturated aq
NaHCO3 (50 mL) was added, and the solution extracted with CH2Cl2
(2ꢂ50 mL). The combined organic extracts were washed with
brine (50 mL), dried over MgSO4, filtered and concentrated in va-
cuo to provide the title product as a white solid (486 mg, 91%):
mp: 85–918C; 1H NMR (400 MHz, CD3OD): d=7.28–7.13 (m, 5H),
4.04 (q, J=7.1 Hz, 2H), 3.72 (brs, 2H), 2.55 (t, J=7.2 Hz, 1H), 2.20
(t, J=7.6 Hz, 1H), 2.10 (brs, 1H), 1.60–1.16 ppm (m, 13H); 13C NMR
(101 MHz, CDCl3): d=173.9, 140.1, 128.4, 128.2, 127.0, 77.4, 77.1,
76.8, 60.1, 53.9, 49.4, 34.4, 29.9, 29.6, 29.5 (3C), 29.4, 29.3, 29.1,
27.3, 25.0, 14.3 ppm; LRMS (ESI+): m/z (%): 334.4 [M+H]+ (100).
Ethyl 12-(benzyl(2-(4-(benzyloxy)-3-(hydroxymethyl)phenyl)-2-
hydroxyethyl)amino)dodecanoate (5): A stirred solution of 4
(185 mg, 0.56 mmol) and DIPEA (300 mL, 1.7 mmol) in CHCl3 (5 mL)
was treated with 1-(4-(benzyloxy)-3-(hydroxymethyl)phenyl)-2-bro-
moethanone (185 mg, 0.25 mmol). The solution was stirred for
15 h at RT. The solvent was removed in vacuo at 258C, and the res-
idue was redissolved in cold (08C) EtOH (10 mL). NaBH4 (100 mg,
2.6 mmol) was added, and after the initial effervescence, the reac-
tion was allowed to come to RT. The solution was stirred for 1 h,
and the solvent was removed in vacuo at 258C. The residue was
redissolved in CHCl3 (50 mL), and the solution was washed with
water (3ꢂ50 mL), brine (50 mL), dried over MgSO4, filtered and
concentrated in vacuo. The resulting residue was purified by flash
chromatography (EtOAc/petroleum spirits) to provide the title
compound as a colourless oil (150 mg, 46%): 1H NMR (400 MHz,
CD3OD): d=7.49–7.10 (m, 12H), 6.93 (d, J=8.4 Hz, 1H), 5.10 (s,
2H), 4.73–4.62 (m, 3H), 4.10 (q, J=7.1 Hz, 2H), 3.71 (d, J=13.4 Hz,
1H), 3.60 (d, J=13.4 Hz, 1H), 2.74–2.39 (m, 4H), 2.27 (t, J=7.4 Hz,
2H), 1.66–1.13 ppm (m, 21H); 13C NMR (101 MHz, CD3OD): d=
175.6, 156.7, 140.5, 138.8, 136.6, 130.9, 130.3, 129.5, 129.2, 128.8,
128.3, 128.1, 127.4, 127.2, 112.6, 72.1, 71.1, 63.3, 61.4, 60.5, 60.2,
55.3, 35.1, 30.7, 30.6 (3C), 30.5, 30.4, 30.2, 28.3, 27.9, 26.0,
14.6 ppm; LRMS (ESI+): m/z (%): 589.4 [M+H]+ (100).
A
solution of N6-(6-(12-((2-benzyloxy-2-(4-benzyloxy-3-hydroxy-
methyl)-phenyl)ethyl)amino)dodecanamido)hexyl)adenosine
(55 mg, 0.061 mmol) and 18m aq NH4OH (20 mL, 0.36 mmol) in
MeOH (5 mL) was treated with Pd(OH)2 (20% on carbon, 2.1 mg,
0.003 mmol). This suspension was stirred for 15 h under a H2 at-
mosphere at RT. The mixture was then diluted with CH2Cl2 (5 mL),
filtered through a nylon membrane filter (0.45 mm pore size), and
concentrated in vacuo to yield a colourless oil. This oil was purified
by preparatory HPLC on a C18 column (0–45% MeCN in H2O with
0.1% AcOH over 40 min) to give the title compound as a white
1
foam (17 mg, 39%); H NMR (400 MHz, CD3OD/CDCl3, 1:1): d=8.21
(s, 1H), 8.07 (s, 1H), 7.21 (s, 1H), 7.09 (d, J=8.3 Hz, 1H), 6.76 (d, J=
8.3 Hz, 1H), 5.87 (d, J=6.7 Hz, 1H, 1H), 4.76–4.69 (m, 4H), 4.65 (s,
2H), 4.32 (dd, J=5.1, 1.9 Hz, 1H), 4.21 (d, J=2.0 Hz, 1H), 3.91 (dd,
J=12.7, 2.0 Hz, 1H), 3.73 (dd, J=12.7, 2.0 Hz, 1H), 3.55 (brs, 1H),
3.14 (t, J=7.0 Hz, 1H), 2.88–2.59 (m, 4H), 2.13 (t, J=7.6 Hz, 2H),
1.91 (s, 3H), 1.74–1.19 ppm (m, 26H); 13C NMR (101 MHz, CD3OD/
CDCl3, 1:1): d=176.1, 156.1, 155.8, 153.4, 148.7, 141.2, 134.1, 128.5,
127.0, 126.9, 121.3, 115.8, 91.3, 88.1, 75.3, 72.6, 71.9, 63.4, 61.0,
56.9, 49.8, 41.4, 40.2, 37.1, 30.4 (3C), 30.3 (2C), 30.2, 30.1, 29.2, 27.9,
27.6, 27.5, 26.9 ppm; HPLC: 98% pure (tR =8.0 min); HRMS (ESI+):
m/z [M+H]+ calcd for C37H60N7O8: 730.4498, found: 730.4479.
12-(Benzyl(2-(4-(benzyloxy)-3-(hydroxymethyl)phenyl)-2-hydrox-
yethyl)amino)dodecanoic acid (6): A solution of
5 (600 mg,
1.0 mmol) in CH2Cl2 (10 mL) was treated with 3m methanolic NaOH
(2.00 mL, 6.0 mmol). The solution was stirred for 15 h at RT, neutral-
ized with 3m methanolic HCl, and the solvent was removed in
vacuo. The resulting crude product was purified by flash chroma-
tography (CH2Cl2/MeOH+Et3N, 1:4:0.005) to provide the title prod-
N6-(12-(12-((2-Hydroxy-2-(4-hydroxy-3-(hydroxymethyl)phenyl)e-
thyl)amino)dodecanamido)dodecyl)adenosine (9b): A solution of
8b (285 mg, 0.63 mmol) in DMF (1 mL) was treated with 6
1
uct as a colourless oil (350 mg, 61%): H NMR (400 MHz, CD3OD):
d=7.50–7.13 (m, 12H), 6.94 (d, J=8.4 Hz, 1H), 5.14–5.11 (m, 2H),
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
7
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
(178 mg, 0.32 mmol), DIPEA (430 mL, 2.5 mmol) and BOP (280 mg,
0.63 mmol). The solution was stirred for 15 h at RT. The solution
was then diluted with EtOAc (100 mL) and washed with water (5ꢂ
50 mL) and brine (50 mL), dried over MgSO4, filtered and concen-
trated in vacuo. The resulting crude was purified by flash chroma-
tography (MeOH/CH2Cl2, 1:20) to provide N6-(12-(12-((2-hydroxy-2-
(4-benzyloxy-3-(hydroxymethyl)phenyl)ethyl)benzylamino)dodeca-
lution was heated at reflux for 24 h and then concentrated in
vacuo. The residue was partitioned between 1m HCl (100 mL) and
CH2Cl2 (100 mL) and filtered. The precipitate was retained, and the
organic portion of the filtrate was separated, washed with brine
(20 mL), dried over MgSO4, filtered, and concentrated to dryness in
vacuo. The resulting residue along with the retained precipitate
were combined and purified by flash chromatography (MeOH/
CH2Cl2, 1:20) to provide the title compound as a white amorphous
solid (704 mg, 84%): mp: 145–1498C; 1H NMR (400 MHz, CD3OD):
d=8.22 (s, 1H), 8.07 (s, 1H), 5.86 (d, J=6.7 Hz, 1H), 4.74 (dd, J=
6.7, 5.1 Hz, 1H), 4.31 (dd, J=5.1, 1.9 Hz, 1H), 4.21 (dd, J=2.1,
1.9 Hz, 1H), 3.91 (dd, J=12.8, 2.1 Hz, 1H), 3.73 (dd, J=12.8, 2.1 Hz,
1H), 3.55 (brs, 2H), 2.25 (t, J=7.5 Hz, 2H), 1.73–1.21 ppm (m,
18H); 13C NMR (101 MHz, CD3OD): d=177.3, 155.4, 152.8, 147.7,
140.4, 120.9, 91.3, 87.8, 74.5, 72.1, 63.1, 41.1, 34.5, 29.8, 29.8, 29.7,
29.7, 29.6, 29.4, 27.2, 25.3 ppm; LRMS (ESI+): m/z (%): 466.3 [M+
H]+ (100).
namido)dodecyl)adenosine as
a colourless oil (130 mg, 41%):
1H NMR (400 MHz, CD3OD): d=8.20 (s, 1H), 8.16 (s, 1H), 7.43–7.18
(m, 11H), 7.14 (dd, J=8.4, 2.2 Hz, 1H), 6.90 (d, J=8.4 Hz, 1H), 5.91
(d, J=6.5 Hz, 1H), 5.08 (s, 2H), 4.73 (dd, J=6.5, 5.1 Hz, 1H), 4.69 (s,
2H), 4.67–4.62 (m, 1H), 4.32 (dd, J=5.1, 2.3 Hz, 1H), 4.23–4.16 (m,
1H), 3.89 (dd, J=12.7, 2.3 Hz, 1H), 3.78 (d, J=13.4 Hz, 1H), 3.73
(dd, J=12.7, 2.3 Hz, 1H), 3.60 (d, J=13.4 Hz, 1H), 3.54 (s, 2H), 3.13
(t, J=7.0 Hz, 2H), 2.73–2.43 (m, 4H), 2.14 (t, J=7.4 Hz, 2H), 1.73–
1.14 ppm (m, 38H); 13C NMR (101 MHz, CD3OD): d=175.9, 156.4,
156.0, 153.2, 148.5, 141.0, 139.4, 138.3, 135.8, 130.6, 130.1, 129.3,
129.1, 128.6, 128.1, 128.0, 127.0, 126.9, 121.2, 112.4, 91.3, 88.1, 75.2,
72.5, 71.3, 70.9, 63.3, 62.8, 60.4, 59.7, 55.0, 41.5, 40.2, 37.0, 30.4
(6C), 30.3, 30.2 (6C), 30.0, 28.1, 27.8, 27.7, 27.5, 26.9 ppm; LRMS
(ESI+): m/z (%): 994.5 [M+H]+ (100).
N6-(11-((6-Aminohexylamino)formyl)undecyl)adenosine (11a): A
solution of 1,6-diaminohexane (73 mg, 0.63 mmol) in DMF (500 mL)
was treated with 10b (98 mg, 0.21 mmol), Et3N (88 mL, 0.63 mmol)
and BOP (186 mg, 0.42 mmol). The solution was stirred for 15 h at
RT then diluted with water (5 mL) and purified by preparatory
HPLC on a C18 column (0–20% MeCN in H2O with 0.1% TFA over
40 min) to provide the title compound as a colourless oil (30 mg,
A solution of N6-(12-(12-((2-benzyloxy-2-(4-hydroxy-3-(benzyloxy-
methyl)phenyl)ethyl)
amino)dodecanamido)dodecyl)adenosine
(85 mg, 0.091 mmol) and 18m aq NH4OH (20 mL, 0.36 mmol) in
MeOH (5 mL) was treated with Pd(OH)2 (20% on carbon, 3.2 mg,
0.006 mmol). The suspension was stirred for 15 h under a H2 at-
mosphere at RT. The mixture was then diluted with CH2Cl2 (5 mL),
filtered through a nylon membrane filter (0.45 mm pore size), and
concentrated in vacuo to yield a colourless oil. This oil was purified
by preparatory HPLC on a C18 column (0–45% MeCN in H2O with
0.1% AcOH over 40 min) to give the title compound as a white
1
25%): H NMR (400 MHz, CD3OD): d=8.20 (s, 1H), 8.02 (s, 1H), 5.90
(d, J=5.3 Hz, 1H), 4.58 (brs, 1H), 4.37–4.23 (m, 1H), 4.14 (brs, 1H),
3.84 (d, J=12.4 Hz, 1H), 3.71 (d, J=12.4 Hz, 1H), 3.50 (brs, 1H),
3.09 (t, J=7.0 Hz, 2H), 2.80 (t, J=7.6 Hz, 2H), 2.08 (t, J=7.4 Hz,
2H), 1.81–1.10 ppm (m, 22H); 13C NMR (101 MHz, CD3OD/CDCl3,
1:1): d=175.7, 158.1, 153.3, 148.5, 141.5, 120.5, 90.9, 87.5, 75.4,
71.8, 62.7, 42.0, 40.0, 39.5, 36.8, 30.0 (4C), 29.9, 29.8 (3C), 29.5, 27.7,
26.6, 26.5, 26.3 ppm; LRMS (ESI+): m/z (%): 664.3 [M+H]+ (100).
1
foam (52 mg, 70%): H NMR (400 MHz, CD3OD/CDCl3, 1:1): d=8.21
(s, 1H), 8.06 (s, 1H), 7.21 (s, 1H), 7.09 (d, J=8.3 Hz, 1H), 6.76 (d, J=
8.3 Hz, 1H), 5.87 (d, J=6.7 Hz, 1H), 4.79–4.71 (m, 2H), 4.66 (s, 2H),
4.32 (dd, J=5.1, 1.9 Hz, 1H), 4.21 (d, J=2.0 Hz, 1H), 3.91 (dd, J=
12.7, 2.0 Hz, 1H), 3.72 (dd, J=12.7, 2.0 Hz, 1H), 3.54 (brs, 2H), 3.13
(t, J=7.2 Hz, 2H), 2.92–2.70 (m, 4H), 2.15–2.08 (m, 2H), 1.92 (s,
3H), 1.72–1.21 ppm (m, 38H); 13C NMR (101 MHz, CD3OD/CDCl3,
1:1): d=176.2, 156.3, 156.2, 153.7, 148.7, 141.4, 134.1, 128.5, 127.4,
127.2, 121.7, 116.4, 91.9, 88.6, 75.5, 72.9, 71.8, 63.9, 61.9, 56.9, 50.0,
42.0, 40.7, 37.6, 30.8 (4C), 30.7, 30.6 (4C), 30.5, 29.2, 28.2 (3C), 28.2,
27.3 ppm; HPLC: 98% pure (tR =9.5 min); HRMS (ESI+): m/z [M+
H]+ calcd for C43H72N7O8: 814.5437, found: 814.5441.
N6-(11-((12-Aminododecanylamino)formyl)undecyl)adenosine
(11b): A solution of 1,12-diaminohexane (126 mg, 0.63 mmol) in
DMF (500 mL) was treated with 10b (98 mg, 0.21 mmol), Et3N
(88 mL, 0.63 mmol) and BOP (186 mg, 0.42 mmol). The solution was
stirred for 15 h at RT then diluted with water (5 mL) and purified
by preparatory HPLC on a C18 column (0–30% MeCN in H2O with
0.1% TFA over 40 min) to provide the title compound as a colour-
1
less oil (45 mg, 32%): H NMR (400 MHz, CD3OD): d=8.25 (s, 1H),
8.00 (s, 1H), 5.81 (d, J=6.9 Hz, 1H), 4.57 (brs, 1H), 4.20 (brs, 1H),
4.06 (s, 1H), 3.90 (d, J=12.2 Hz, 1H), 3.76 (d, J=12.2 Hz, 1H), 3.56
(brs, 2H), 3.13 (t, J=7.2 Hz, 2H), 2.85 (t, J=7.4 Hz, 2H), 2.13 (d, J=
7.6 Hz, 2H), 1.83–1.16 ppm (m, 38H); 13C NMR (101 MHz, CD3OD):
d=175.4, 155.5, 153.3, 149.7, 141.3, 120.4, 90.9, 87.5, 75.2, 71.7,
62.6, 40.1, 39.9, 36.8, 29.9 (3C), 29.8, 29.7 (3C), 29.5, 27.9, 27.3, 26.8,
26.4 ppm; LRMS (ESI+): m/z (%): 648.5 [M+H]+ (100).
N6-(5-Carboxypentyl)adenosine (10a): A solution of 7 (500 mg,
1.8 mmol) and DIPEA (620 mL, 7.0 mmol) in tBuOH (50 mL) was
treated with 6-aminohexanoic acid (660 mg, 3.5 mmol). The solu-
tion was heated at reflux for 24 h and then concentrated in vacuo.
When an attempt was made to partition the residue between 1m
aq HCl (100 mL) and CH2Cl2 (100 mL), a precipitate formed. This
precipitate was collected by suction filtration to provide the title
compound as an amorphous solid (280 mg, 41%): mp: 151–1558C;
1H NMR (400 MHz, CD3OD): d=8.21 (s, 1H), 8.01 (s, 1H), 5.84 (d, J=
6.8 Hz, 1H), 4.75 (dd, J=6.8, 5.1 Hz, 1H), 4.31 (dd, J=5.1, 1.7 Hz,
1H), 4.22 (dd, J=2.0, 1.7 Hz, 1H), 3.91 (dd, J=12.8, 2.0 Hz, 1H),
3.72 (dd, J=12.8, 2.0 Hz, 1H), 3.56 (brs, 2H), 2.29 (t, J=7.4 Hz, 2H),
1.80–1.41 ppm (m, 6H); 13C NMR (101 MHz, CD3OD): d=177.1,
155.4, 152.8, 147.8, 140.5, 120.9, 91.2, 87.8, 74.5, 72.1, 63.1, 40.9,
34.5, 29.4, 26.8, 25.1 ppm; LRMS (ESI+): m/z (%): 382.2 [M+H]+
(100).
N6-(11-((6-(12-((2-Hydroxy-2-(4-hydroxy-3-(hydroxymethyl)phe-
nyl)ethyl)amino)dodecanamido)hexylamino)formyl)undecyl)ade-
nosine acetate (12a): A solution of 11 a (60 mg, 0.11 mmol) in
DMF (1 mL) was treated with 6 (120 mg, 0.21 mmol), Et3N (74 mL,
0.53 mmol) and BOP (49 mg, 0.11 mmol). The solution was stirred
for 15 h at RT then diluted with EtOAc (100 mL) and washed with
water (5ꢂ50 mL) and brine (50 mL), dried over MgSO4, filtered and
concentrated in vacuo. The resulting crude was purified by flash
chromatography (MeOH/CH2Cl2, 1:20) to provide N6-(11-(1-(6-(12-
((2-hydroxy-2-(4-benzyloxy-3-(hydroxymethyl)phenyl)ethyl)-benzyla-
mino)dodecanamido)hexananamino)formyl)undecyl)adenosine as
a colourless oil (70 mg, 59%): 1H NMR (400 MHz, CD3OD/CDCl3,
1:1): d=8.21 (s, 1H), 7.97 (s, 1H), 7.43–7.23 (m, 11H), 7.12 (dd, J=
8.4, 2.1 Hz, 1H), 6.86 (d, J=8.4 Hz, 1H), 5.82 (d, J=6.8 Hz, 1H), 5.05
(s, 2H), 4.75 (dd, J=6.8, 5.2 Hz, 1H), 4.68 (s, 2H), 4.65–4.57 (m, 2H),
N6-(11-Carboxyundecyl)adenosine (10b): A solution of 7 (500 mg,
1.8 mmol) and DIPEA (620 mL, 7.0 mmol) in tBuOH (50 mL) was
treated with 12-aminododecanoic acid (750 mg, 3.5 mmol). The so-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
8
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
4.24–4.21 (m, 1H), 3.91 (dd, J=12.8, 2.8 Hz, 1H), 3.89 (d, J=
13.4 Hz, 1H), 3.71 (dd, J=12.8, 2.8 Hz, 1H), 3.60 (d, J=13.4 Hz, 1H),
3.53 (s, 2H), 3.12 (t, J=7.0 Hz, 4H), 2.46–2.70 (m, 4H), 2.12 (t, J=
7.6 Hz, 4H), 1.74–1.17 ppm (m, 44H); 13C NMR (101 MHz, CD3OD/
CDCl3, 1:1): d=176.0, 175.7, 157.0, 156.4, 153.8, 148.6, 141.3, 139.9,
138.3, 135.5, 131.0, 130.4, 129.9, 129.8, 129.3, 128.6, 128.6, 127.5,
127.3, 121.9, 112.9, 92.4, 88.8, 75.3, 73.1, 71.5, 70.8, 64.1, 63.5, 61.8,
59.9, 55.3, 42.0, 40.4, 37.9, 37.9, 36.8, 34.8, 30.9 (4C), 30.8 (4C) 30.7
(4C), 30.6 (4C), 30.4, 28.6 (2C), 28.2, 28.1, 27.5, 27.3, 26.6 ppm;
LRMS (ESI+): m/z (%): 554.5 [M+2H]2+/2 (100); 1107.9 [M+H]+
(15).
A mixture of N6-(11-(1-(12-(12-((2-hydroxy-2-(4-benzyloxy-3-(benzy-
loxymethyl)phenyl)ethyl)amino)dodecanamido)dodecanamino)for-
myl)undecyl)adenosine (150 mg, 0.13 mmol) and 18m aq NH4OH
(20 mL, 0.36 mmol) in MeOH (5 mL) was heated at reflux until disso-
lution was observed. Pd(OH)2 (20% on carbon, 5.4 mg,
0.006 mmol) was added, and the solution was further heated at
reflux under a hydrogen atmosphere for 2 h. The solution was
cooled and diluted with CH2Cl2 (5 mL) before being filtered
through a nylon membrane filter (0.45 mm) and concentrated in
vacuo to yield a colourless oil. This oil was purified by preparatory
HPLC on a C18 column (0–45% MeCN in H2O with 0.1% AcOH over
40 min) to provide the title compound as a white foam (71 mg,
56%): 1H NMR (400 MHz, CD3OD/CDCl3, 1:1): d=8.21 (s, 1H), 8.07
(s, 1H), 7.21 (s, 1H), 7.10 (d, J=8.3 Hz, 1H), 6.76 (d, J=8.3 Hz, 1H),
5.87 (d, J=6.7 Hz, 1H), 4.74 (dd, J=6.7, 5.2 Hz, 1H), 4.71–4.73 (m,
partially obscured by H2O signal, 1H), 4.66 (s, 2H), 4.32 (dd, J=5.2,
1.9 Hz, 1H), 4.21 (dd, J=2.0, 1.9 Hz, 1H), 3.91 (dd, J=12.7, 2.0 Hz,
1H), 3.73 (dd, J=12.7, 2.0 Hz, 1H), 3.54 (brs, 2H), 3.13 (t, J=7.2 Hz,
4H), 2.90–2.64 (m, 4H), 2.13 (t, J=7.5 Hz, 4H), 1.73–1.16 ppm (m,
56H); 13C NMR (101 MHz, CD3OD/CDCl3, 1:1): d=176.2, 156.4,
156.2, 153.7, 148.7, 141.4, 134.2, 128.4, 127.4, 127.2, 121.8, 116.5,
92.0, 88.6, 75.5, 72.9, 72.0, 63.9, 62.1, 57.1, 50.6, 42.0, 40.7, 37.6,
30.8 (11C), 30.6 (10C), 30.5, 29.4, 28.3, 28.2 (2C), 27.3 ppm (2C);
HPLC: 99% pure (tR =10.5 min); HRMS (ESI+): m/z [M+H]+ calcd
for C55H95N8O9: 1011.7217, found: 1011.7200.
A
solution of N6-(11-(1-(6-(12-((2-benzyloxy-2-(4-benzyloxy-3-(hy-
droxymethyl)phenyl)ethyl)amino)dodecanamido)hexylamino)formy-
l)undecyl)adenosine (75 mg, 0.068 mmol) and 18m aq NH4OH
(20 mL, 0.36 mmol) in MeOH (5 mL) was treated with Pd(OH)2 (20%
on carbon, 2.4 mg, 0.003 mmol). The suspension was shaken in
a Parr hydrogenator under a H2 atmosphere (350 kPa) for 2 d at RT.
The mixture was then diluted with CH2Cl2 (5 mL), filtered through
a nylon membrane filter (0.45 mm pore size) and concentrated in
vacuo to yield a colourless oil. This oil was purified by preparatory
HPLC on a C18 column (0–45% MeCN in H2O with 0.1% AcOH over
40 min) to provide the title compound as a white foam (20 mg,
32%): 1H NMR (400 MHz, CD3OD/CDCl3, 1:1): d=8.21 (s, 1H), 8.14
(s, 1H), 7.26 (s, 1H), 7.12 (d, J=8.3 Hz, 1H), 6.77 (d, J=8.3 Hz, 1H),
5.90 (d, J=6.6 Hz, 1H), 4.83–4.79 (m, partially obscured by H2O
signal, 1H), 4.75–4.70 (m, partially obscured by H2O signal, 1H),
4.66 (s, 2H), 4.32 (dd, J=5.1, 2.2 Hz, 1H), 4.20 (d, J=2.2 Hz, 1H),
3.90 (dd, J=12.7, 2.2 Hz, 1H), 3.74 (dd, J=12.7, 2.2 Hz, 1H), 3.55
(brs, 2H), 3.14 (t, J=7.0 Hz, 4H), 2.95 (d, J=7.3 Hz, 2H), 2.84 (t, J=
7.6 Hz, 2H), 2.14 (t, J=7.5 Hz, 4H), 1.91 (s, 3H), 1.73–1.19 ppm (m,
48H); 13C NMR (101 MHz, CD3OD/CDCl3, 1:1): d=175.8 (2C), 155.9,
155.7, 153.1, 148.2, 140.9, 133.0, 128.2, 126.7, 126.5, 115.8, 91.3,
88.0, 75.0, 72.3, 70.7, 63.3, 61.0, 55.8, 41.4, 39.9, 37.0, 30.3, 30.2
(4C), 30.1 (3C), 30.0 (2C), 29.9 (2C), 27.9, 27.6, 27.5, 27.1, 26.7 (2C),
23.0 ppm; HPLC: 99% pure (tR =9.3 min); HRMS (ESI+): m/z [M+
2H]2+/2 calcd for C49H84N8O9: 464.3175, found: 464.3167.
N6-(5-((2-(2-(2-Aminoethyloxy)ethoxy)ethylamino)formyl)penty-
l)adenosine (13a): A solution of 10a (130 mg, 0.34 mmol) in DMF
(200 mL) was treated with 2,2’-(ethylenedioxy)diethylamine
(219 mg, 0.36 mmol) and BOP (315 mg, 0.71 mmol). The solution
was stirred for 15 h at RT then diluted with water (1 mL) and puri-
fied by preparatory HPLC on a C8 column (0–25% MeCN in H2O
with 0.1% TFA over 40 min) to provide the title compound as
1
a white foam (76 mg, 44%): H NMR (400 MHz, CD3OD): d=8.21 (s,
1H), 8.16 (s, 1H), 5.91 (d, J=6.6 Hz, 1H), 4.74 (dd, J=6.6, 5.2 Hz,
1H), 4.33 (dd, J=5.2, 2.1 Hz, 1H), 4.20 (dd, J=2.1 Hz, 1H), 3.90 (dd,
J=12.7, 2.3 Hz, 1H), 3.74 (dd, J=12.7, 2.3 Hz, 1H), 3.69–3.64 (m,
2H), 3.64–3.61 (m, J=1.0 Hz, 4H), 3.58 (brs, 2H), 3.53 (t, J=5.7 Hz,
2H), 3.35 (t, J=5.7 Hz, 2H), 3.05 (t, J=5.0 Hz, 2H), 2.21 (t, J=
7.5 Hz, 2H), 1.77–1.59 (m, 4H), 1.52–1.37 ppm (m, 2H); 13C NMR
(101 MHz, CD3OD): d=175.8, 155.9, 153.1, 148.4, 141.0, 121.2, 91.2,
88.0, 75.1, 72.4, 71.0, 70.9, 70.4, 68.1, 63.3, 40.3, 40.2, 39.8, 36.7,
27.1, 26.3, 23.9 ppm; LRMS (ESI+): m/z (%): 512.3 [M+H]+ (100).
N6-(11-((12-(12-((2-Hydroxy-2-(4-hydroxy-3-(hydroxymethyl)phe-
nyl)ethyl)amino)dodecanamido)dodecanamino)formyl)undecy-
l)adenosine (12b): A solution of 11 b (200 mg, 0.31 mmol) in DMF
(1 mL) was treated with 7 (200 mg, 0.36 mmol), Et3N (215 mL,
1.5 mmol) and BOP (138 mg, 0.31 mmol). The solution was stirred
for 15 h at RT then diluted with EtOAc (100 mL), washed with
water (5ꢂ50 mL) and brine (50 mL), dried over MgSO4, filtered and
concentrated in vacuo. The resulting crude was then purified by
flash chromatography (MeOH/CH2Cl2, 1:20) to provide N6-(11-(1-(12-
(12-((2-hydroxy-2-(4-benzyloxy-3-(hydroxymethyl)phenyl)ethyl)ben-
zylamino)dodecanamido)dodecanamino)formyl)undecyl)adenosine
N6-(11-((2-(2-(2-Aminoethyloxy)ethoxy)ethylamino)formyl)unde-
cyl)adenosine (13b): 2,2’-(Ethylenedioxy)diethylamine (1.6 mL,
11 mmol) at 08C was treated with 10b (1.00 g, 2.1 mmol) followed
by BOP (1.9 g, 4.3 mmol). The reaction was stirred for 15 h at RT.
The solution was then diluted with water (5 mL) and purified by
preparatory HPLC on a C18 column (0–40% MeCN in H2O over
40 min with 0.1% NH4OH) to provide the title compound as
a white amorphous solid (1.2 g, 94%): mp: 147–1508C; 1H NMR
(400 MHz, [D6]DMSO): d=8.33 (s, 1H), 8.19 (s, 1H), 7.87 (brs, 1H),
7.82 (d, J=5.7 Hz, 1H), 5.87 (d, J=6.2 Hz, 1H), 5.44 (brs, 2H), 5.18
(brs, 1H), 4.60 (dd, J=6.2, 4.6 Hz, 1H), 4.14 (dd, J=4.6, 3.1 Hz, 1H),
3.96 (dd, J=3.1, 3.4 Hz, 1H), 3.67 (dd, J=12.0, 3.4 Hz, 1H), 3.54 (d,
J=12.0, 3.4 Hz, 1H), 3.51–3.41 (m, 8H), 3.38 (t, J=5.3 Hz, 2H), 3.34
(t, J=5.8 Hz, 2H), 3.17 (dt, J=5.7, 5.3 Hz, 2H), 2.62 (t, J=5.8 Hz,
1H), 2.03 (t, J=7.4 Hz, 1H), 1.64–1.19 ppm (m, 18H); 13C NMR
(101 MHz, CD3OD/CDCl3, 1:1): d=175.5, 154.8, 152.0, 149.0, 141.7,
121.0, 90.4, 87.8, 75.6, 72.2, 71.2, 71.0, 70.5, 67.8, 63.1, 41.7, 40.3,
39.9, 36.9, 30.5 (6C), 30.3, 30.1, 27.8, 26.8 ppm; HRMS (ESI+): m/z
[M+H]+ calcd for C28H48N7O7: 596.3766, found: 596.3785.
1
as a colourless oil (150 mg, 70%): H NMR (400 MHz, CD3OD/CDCl3,
1:1): d=8.16 (s, 1H), 8.00 (s, 1H), 7.36–7.19 (m, 11H), 7.10 (dd, J=
8.5, 2.1 Hz, 1H), 6.84 (d, J=8.5 Hz, 1H), 5.83 (d, J=6.7 Hz, 1H), 5.02
(s, 2H), 4.71 (dd, J=6.7, 5.0 Hz, 1H), 4.69–4.63 (m, 2H), 4.29 (dd,
J=5.0, 1.9 Hz, 1H), 4.18 (dd, J=1.9, 12.7 Hz, 1H), 3.96 (d, J=
13.4 Hz, 1H), 3.87 (dd, J=12.7, 1.9 Hz, 1H), 3.72 (d, J=13.3 Hz, 1H),
3.68 (dd, J=12.7, 1.9 Hz, 1H), 3.48 (brs, 2H), 3.09 (t, J=7.2 Hz, 4H),
2.79–2.51 (m, 4H), 2.09 (t, J=7.5 Hz, 4H), 1.67–1.10 ppm (m, 56H);
13C NMR (101 MHz, CD3OD/CDCl3, 1:1): d=175.3, 175.3, 156.1,
155.4, 152.8, 147.7, 140.5, 137.5, 136.2, 134.2, 130.3, 130.1, 129.1,
128.9, 128.5, 128.3, 127.6, 126.5, 126.1, 120.9, 112.0, 91.2, 87.7, 74.6,
72.1, 70.4, 69.6, 63.1, 61.5, 60.2, 58.8, 54.4, 41.1, 40.0, 39.9, 36.8,
36.8, 30.0, 29.9, 29.9, 29.9, 29.8, 29.7, 29.7, 29.6, 27.5, 27.3, 27.3,
26.4, 26.2 ppm; LRMS (ESI+): m/z (%): 596.5 [M+2H]2+/2 (100).
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
9
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
N6-(5-((2-(2-(2-(12-((2-Hydroxy-2-(4-hydroxy-3-(hydroxymethyl)-
phenyl)ethyl)amino)dodecanamido)ethyloxy)ethoxy)ethylami-
no)formyl)pentyl)adenosine acetate (14a): A solution of 13a
(180 mg, 0.35 mmol) in DMF (1 mL) was treated with 7 (200 mg,
0.36 mmol), Et3N (150 mL, 1.1 mmol) and BOP (315 mg, 0.71 mmol).
The solution was stirred for 15 h at RT then diluted with water
(1 mL) and purified directly by preparatory HPLC on a C18 column
(0–30% MeCN in H2O with 0.1% TFA over 40 min) to provide the
title compound N6-(5-((2-(2-(2-(12-((2-benzyloxy-2-(4-benzyloxy-3-
(hydroxymethyl)phenyl)ethyl)amino)dodecanamido)ethyloxy)ethox-
y)ethylamino)formyl)pentyl)adenosine as a white foam (96 mg,
26%): 1H NMR (400 MHz, CD3OD): d=8.24 (s, 1H), 8.20 (s, 1H),
7.47–7.22 (m, 12H), 7.18 (dd, J=8.4, 1.9 Hz, 1H), 6.94 (d, J=8.4 Hz,
1H), 5.98 (d, J=6.5 Hz, 1H), 5.08 (s, 2H), 4.83–4.74 (m, 2H), 4.72 (s,
2H), 4.36 (dd, J=5.0, 2.4 Hz, 1H), 4.20 (d, J=2.3 Hz, 1H), 4.06 (d,
J=13.4 Hz, 1H), 3.96 (d, J=13.3 Hz, 1H), 3.90 (dd, J=12.5, 2.3 Hz,
1H), 3.75 (dd, J=12.5, 2.3 Hz, 1H), 3.58 (s, 4H), 3.56–3.54 (brs, 2H),
3.54–3.50 (m, 4H), 3.38–3.32 (m, 4H), 2.99–2.70 (m, 4H), 2.25–2.13
(m, 4H), 1.74–1.19 ppm (m, 24H); 13C NMR (101 MHz, CD3OD): d=
176.2, 176.0, 156.7, 156.1, 153.5, 148.8, 141.3, 138.6, 136.1, 135.4,
131.1, 129.7, 129.4, 129.4, 128.8, 128.2, 127.0, 126.8, 121.3, 112.6,
91.2, 88.1, 75.5, 72.6, 71.2, 70.9, 70.5, 63.4, 61.7, 60.3, 60.1, 59.4,
55.0, 41.4, 40.2, 37.0, 36.8, 30.5 (2C), 30.4 (2C), 30.3, 30.2, 30.1, 27.9,
27.4, 26.9, 26.6, 26.3 ppm; LRMS (ESI+): m/z (%): 528.3 [M+2H]2+
/2 (100); 1055.6 [M+H]+ (10).
6.8 Hz, 1H), 5.07 (s, 2H), 4.88 (brs, 1H), 4.74 (dd, J=6.8, 5.2 Hz,
1H), 4.70 (s, 2H), 4.48 (d, J=13.2 Hz, 1H), 4.37 (d, J=13.2 Hz, 1H),
4.32 (dd, J=5.2, 1.9 Hz, 1H), 4.21 (d, J=1.9 Hz, 1H), 3.90 (dd, J=
12.7, 2.1 Hz, 1H), 3.73 (dd, J=12.7, 2.1 Hz, 1H), 3.60 (brs, 2H), 3.59
(s, 4H), 3.52 (t, J=5.5 Hz, 4H), 3.36 (d, J=5.5 Hz, 4H), 3.24–3.07 (m,
4H), 2.26–2.10 (m, 4H), 1.85–1.14 ppm (m, 36H); 13C NMR
(101 MHz, CD3OD): d=175.6, 156.6, 155.6, 152.9, 147.9, 140.7,
137.5, 132.8, 131.6, 130.9, 130.8, 130.0, 129.7, 129.0, 128.4, 127.7,
126.4, 125.8, 121.0, 112.3, 91.2, 87.8, 74.7, 72.2, 70.6, 70.5, 70.1,
68.0, 63.1, 60.0, 59.2, 58.3, 54.5, 41.2, 39.6, 36.8, 30.1 (3C), 30.0 (3C),
29.9, 29.8, 29.7, 29.5, 27.4, 27.0, 26.4, 26.3, 24.1 ppm; LRMS (ESI+):
m/z (%): 554.5 [M+2H]2+/2 (100); 1107.9 [M+H]+ (20).
A solution of N6-(11-(1-(2-(2-(2-(12-((2-benzyloxy-2-(4-benzyloxy-3-
(hydroxymethyl)phenyl)ethyl)amino)dodecanamido)ethyloxy)ethox-
y)ethylamino)formyl)undecyl)adenosine (85 mg, 0.075 mmol) and
18m aq NH4OH (20 mL, 0.36 mmol) in MeOH (5 mL) was treated
with Pd(OH)2 (20% on carbon, 2.6 mg, 0.003 mmol). The suspen-
sion was stirred for 15 h under a H2 atmosphere at at RT. The mix-
ture was then diluted with CH2Cl2 (5 mL), filtered through a nylon
membrane filter (0.45 mm pore size) and concentrated in vacuo to
yield a colourless oil. This oil was purified by preparatory HPLC on
a C18 column (0–20% MeCN in H2O with 0.1% AcOH over 40 min)
to provide the title compound as a colourless oil (39 mg, 55%):
1H NMR (400 MHz, CD3OD/CDCl3, 1:1): d=8.21 (s, 1H), 8.03 (s, 1H),
7.20 (s, 1H), 7.09 (d, J=8.2 Hz, 1H), 6.76 (d, J=8.2 Hz, 1H), 5.85 (d,
J=6.6 Hz, 1H), 4.82 (dd, J=6.6, 4.5 Hz, 1H), 4.77–4.72 (m, 1H), 4.66
(s, 2H), 4.32 (dd, J=4.5, 1.4 Hz, 1H), 4.22 (d, J=1.4 Hz, 1H), 3.91
(dd, J=12.7, 1.6 Hz, 1H), 3.72 (dd, J=12.7, 1.6 Hz, 1H), 3.59 (brs,
6H), 3.52 (t, J=5.3 Hz, 4H), 3.36 (t, J=5.3 Hz, 4H), 3.02–2.78 (m,
4H), 2.15 (t, J=7.6 Hz, 4H), 1.92 (s, 3H), 1.71–1.19 ppm (m, 36H);
13C NMR (101 MHz, CD3OD/CDCl3, 1:1): d=179.6, 175.4, 155.5,
152.9, 147.7, 140.5, 132.3, 127.6, 126.5, 126.2, 120.9, 115.7, 91.2,
87.8, 74.6, 72.1, 70.5, 70.2, 69.6, 63.1, 61.3, 55.0, 48.5, 41.1, 39.5,
36.7, 30.0 (3C), 29.9 (3C) 29.8 (3C), 29.7, 29.6, 27.3, 27.1, 26.9, 26.3,
26.2 (2C), 23.9 ppm; HPLC: 99% pure (tR =7.3 min); HRMS (ESI+):
m/z [M+H]+ calcd for C49H83N8O11: 959.6176, found: 959.6179.
A solution of N6-(5-((2-(2-(2-(12-((2-benzyloxy-2-(4-benzyloxy-3-(hy-
droxymethyl)phenyl)ethyl)amino)dodecanamido)ethyloxy)ethoxy)e-
thylamino)formyl)pentyl)adenosine (90 mg, 0.085 mmol) and 18m
aq NH4OH (20 mL, 0.36 mmol) in MeOH (5 mL) was treated with
Pd(OH)2 (20% on carbon, 3.0 mg, 0.004 mmol). The suspension was
stirred for 15 h under a H2 atmosphere at RT. The mixture was then
diluted with CH2Cl2 (5 mL), filtered through a nylon membrane
filter (0.45 mm pore size) and concentrated in vacuo to yield a col-
ourless oil. This oil was purified by preparatory HPLC on a C18
column (0–45% MeCN in H2O with 0.1% AcOH over 40 min) to
1
give the title compound as a colourless oil (65 mg, 87%): H NMR
(400 MHz, CD3OD/CDCl3, 1:1): d=8.20 (s, 1H), 8.11 (s, 1H), 7.25 (s,
1H), 7.11 (d, J=8.2 Hz, 1H), 6.78 (d, J=8.2 Hz, 1H), 5.90 (d, J=
6.6 Hz, 1H), 4.88–4.80 (m, 1H), 4.73–7.74 (m, 1H), 4.66 (s, 2H), 4.34
(dd, J=5.0, 2.0 Hz, 1H), 4.21 (d, J=2.0 Hz, 1H), 3.90 (dd, J=12.7,
2.0 Hz, 1H), 3.73 (dd, J=12.6, 2.0 Hz, 1H), 3.59 (brs, 6H), 3.52 (t,
J=5.4 Hz, 4H), 3.35 (t, J=5.1 Hz, 4H), 3.12–2.85 (m, 4H), 2.25–2.11
(m, 4H), 1.92 (s, 3H), 1.74–1.21 ppm (m, 24H); 13C NMR (101 MHz,
CD3OD/CDCl3, 1:1): d=179.7, 175.6 (2C), 175.4, 140.7, 132.2, 128.1,
126.6, 126.4, 121.0, 115.8, 91.1, 87.8, 78.7, 78.4, 78.0, 74.9, 72.2,
70.7, 70.2, 69.5, 63.1, 61.0, 54.8, 48.5, 41.0, 39.7, 36.8, 36.6, 30.0
(2C), 29.9 (2C), 29.8, 29.7, 27.2, 27.0, 26.7, 26.4, 26.1, 23.9 ppm;
HPLC: 99% pure (tR =8.5 min); HRMS (ESI+): m/z [M+H]+ calcd
for C43H71N8O11: 875.5237, found: 875.5267.
N6-Hexyl-12-((2-hydroxy-2-(4-benzoxy-3-(hydroxymethyl)pheny-
l)ethyl)benzylamino)dodecanamide (15): A solution of hexylamine
(70 mg, 0.53 mmol) in DMF (500 mL) was treated with 6 (200 mg,
0.36 mmol), Et3N (150 mL, 1.1 mmol) and BOP (115 mg, 0.71 mmol).
The reaction was stirred for 15 h at RT. The solution was then dilut-
ed with EtOAc (100 mL) and washed with water (5ꢂ50 mL) and
brine (50 mL), dried over MgSO4, filtered and concentrated in
vacuo to provide the title compound as a colourless oil (120 mg,
52%): 1H NMR (400 MHz, CD3OD): d=7.39–7.07 (m, 12H), 6.81 (d,
J=8.4 Hz, 1H), 5.01 (s, 2H), 4.63 (s, 2H), 4.56–4.49 (m, 1H), 3.82 (d,
J=13.4 Hz, 1H), 3.41 (d, J=13.4 Hz, 1H), 3.17–3.10 (m, 2H), 2.60–
2.28 (m, 4H), 2.05 (d, J=2.9 Hz, 2H), 1.67–1.08 (m, 26H), 0.80 ppm
(t, J=6.8 Hz, 3H); 13C NMR (101 MHz, CD3OD): d=174.9, 157.2,
138.4, 137.3, 135.1, 131.2, 131.2, 130.0, 129.6, 129.6, 129.0, 128.5,
127.3, 127.2, 62.4, 61.0, 59.6, 55.3, 41.5, 40.3, 36.8, 30.8 (3C), 30.7
(3C), 28.4, 27.2, 26.9, 26.7, 23.1, 14.5 ppm; LRMS (ESI+): m/z (%):
545.5 [M+H]+ (100).
N6-(11-((2-(2-(2-(12-((2-Hydroxy-2-(4-hydroxy-3-(hydroxymethyl)-
phenyl)ethyl)amino)dodecanamido)ethyloxy)ethoxy)ethylami-
no)formyl)undecyl)adenosine acetate (14b): A solution of 13b
(215 mg, 0.36 mmol) in DMF (500 mL) was treated with 6 (200 mg,
0.36 mmol), Et3N (150 mL, 1.1 mmol) and BOP (31 mg, 0.71 mmol).
The solution was stirred for 15 h at RT then diluted with water
(1 mL) and purified directly by preparatory HPLC on a C18 column
(0–30% MeCN in H2O with 0.1% TFA over 40 min) to provide N6-
(11-((2-(2-(2-(12-((2-hydroxy-2-(4-benzyloxy-3-(hydroxymethyl)phe-
nyl)ethyl)benzylamino)dodecanamido)ethyloxy)ethoxy)ethylamino)-
formyl)undecyl)adenosine as a white foam (95 mg, 23%): 1H NMR
(400 MHz, CD3OD): d=8.20 (s, 1H), 8.08 (s, 1H), 7.62–7.21 (m, 12H),
7.16 (dd, J=8.5, 2.1 Hz, 1H), 6.91 (d, J=8.5 Hz, 1H), 5.88 (d, J=
N6-Hexyl-12-((2-Hydroxy-2-(4-hydroxy-3-(hydroxymethyl)pheny-
l)ethyl)amino)dodecanamide (16): A solution of 15 (100 mg,
0.18 mmol) and 18m aq NH4OH (20 mL, 0.36 mmol) in MeOH (5 mL)
was treated with Pd(OH)2 (20% on carbon, 6.4 mg, 0.009 mmol).
The suspension was stirred for 15 h under a H2 atmosphere at RT.
The mixture was then diluted with CH2Cl2 (5 mL), filtered through
a nylon membrane filter (0.45 mm pore size) and concentrated in
vacuo to yield a colourless oil. This oil was purified by preparatory
HPLC on a C18 column (0–45% MeCN in H2O with 0.1% AcOH over
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&10
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
40 min) to give the title compound as a white foam (17 mg, 39%):
1H NMR (400 MHz, CD3OD/CDCl3, 1:1): d=7.21 (s, 1H), 7.09 (d, J=
8.3 Hz, 1H), 6.75 (d, J=8.3 Hz, 1H), 4.69–4.64 (m, 3H), 3.13 (t, J=
7.1 Hz, 2H), 2.83–2.51 (m, 4H), 2.14 (t, J=7.5 Hz, 2H), 1.64–1.21 (m,
26H), 0.87 ppm (d, J=6.9 Hz, 3H); 13C NMR (101 MHz, 101 MHz,
CD3OD/CDCl3, 1:1): d=175.6, 155.4, 134.4, 127.7, 126.7, 126.6,
115.7, 72.4, 61.4, 57.4, 50.0, 40.0, 36.9, 32.1, 30.1, 30.1, 30.1 (3C),
30.0 (3C), 29.9, 29.8, 27.9, 27.2, 26.6, 23.2, 14.3 ppm; HPLC: 95%
pure (tR =9.9 min); HRMS (ESI+): m/z [M+H]+ calcd for
C27H49N2O4: 465.3687, found: 465.3687.
Abbreviations
(Benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluoro-
phosphate (BOP); N,N-diisopropylethylamine (DIPEA); N,N-dime-
thylformamide (DMF); ethylenediaminetetraacetic acid (EDTA); tri-
fluoroacetic acid (TFA).
Acknowledgements
The authors wish to thank the Australian Research Council (ARC)
Centre for Free Radical Chemistry & Biotechnology for financial
support. The authors also thank Dr. Jason Dang (Monash Insti-
tute of Pharmaceutical Sciences, Monash University) for the ac-
quisition of MS data.
Pharmacology
Cell culture and cAMP assay: DDT1 MF-2 cells were grown in
150 mm plastic dishes using Dulbecco’s modified Eagle’s medium
(DMEM) containing 5% fetal bovine serum (FBS), streptomycin
(0.1 mgmLꢀ1), penicillin G (100 UmLꢀ1
)
and amphotericin B
(2.5 mgmLꢀ1) in a humidified atmosphere of 95% air and 5% CO2.
Cells were subcultured twice weekly and used in experiments at
one-day preconfluence. Experiments were started by aspirating the
DMEM and rinsing the cells with warm Hank’s balanced salt solu-
tion (HBSS). The cells were then gently detached by using HBSS
containing EDTA (1 mm) and washed twice in HBSS by centrifuga-
tion (1000 rpm for 5 min) and gentle resuspension. The final cell
Keywords: A1 adenosine receptors · b2 adrenergic receptors ·
combined pharmacophores
structure–activity relationships
·
heterobivalent ligands
·
[1] E. Jacoby, R. Bouhelal, M. Gerspacher, K. Seuwen, ChemMedChem 2006,
1, 760–782.
[2] R. Schrçder, N. Janssen, J. Schmidt, A. Kebig, N. Merten, S. Hennen, A.
Muller, S. Blattermann, M. Mohr-Andrꢃ, S. Zahn, J. Wenzel, N. J. Smith, J.
Gomeza, C. Drewke, G. Milligan, K. Mohr, E. Kostenis, Nat. Biotechnol.
2010, 28, 943–949.
[3] P. Karellas, M. McNaughton, S. P. Baker, P. J. Scammells, J. Med. Chem.
2008, 51, 6128–6137.
[4] D. J. Daniels, A. Kulkarni, Z. Xie, R. G. Bhushan, P. S. Portoghese, J. Med.
Chem. 2005, 48, 1713–1716.
[5] D. J. Daniels, N. R. Lenard, C. L. Etienne, P.-Y. Law, S. C. Roerig, P. S. Por-
toghese, Proc. Natl. Acad. Sci. USA 2005, 102, 19208–19213.
[6] D. Huber, H. Hꢄbner, P. Gmeiner, J. Med. Chem. 2009, 52, 6860–6870.
[7] D. Huber, S. Lçber, H. Hꢄbner, P. Gmeiner, Bioorg. Med. Chem. 2012, 20,
455–466.
[8] R. H. Kramer, J. W. Karpen, Nature 1998, 395, 710–713.
[9] J. G. Dobson, R. A. Fenton, F. D. Romano, Circ. Res. 1990, 66, 1381–
1390.
[10] Y. Song, J. C. Shryock, H. J. Knot, L. Belardinelli, Am. J. Physiol. Heart Circ.
Physiol. 2001, 281, H2789–H2795.
[11] M. Keller, S. Teng, G. Bernhardt, G. Buschauer, ChemMedChem 2009, 4,
1733–1745.
[12] J. Kꢄhhorn, H. Hꢄbner, P. Gmeiner, J. Med. Chem. 2011, 54, 4896–4903.
[13] F. M. McRobb, I. T. Crosby, E. Yuriev, J. R. Lane, B. Capuano, J. Med. Chem.
2012, 55, 1622–1634.
[14] J. Shonberg, P. J. Scammells, B. Capuano, ChemMedChem 2011, 6, 963–
974.
[15] S. A. Hutchinson, P. J. Scammells, Curr. Pharm. Des. 2004, 10, 2021–
2039.
[16] K. A. Jacobson, Bioconjugate Chem. 2009, 20, 1816–1835.
[17] Y. Rong, A. E. Ruoho, Synth. Commun. 1999, 29, 2155–2162.
[18] P. J. van Galen, A. P. IJzerman, W. Soudijn, FEBS Lett. 1987, 223, 197–
201.
pellet was resuspended in HBSS and diluted to 1ꢂ106 cellsmLꢀ1
.
The cAMP content of the cells was determined using a HitHunter
cAMP XS+ assay kit (EFC chemiluminescence; DiscovRx Corp., Fre-
mont, USA) according to the manufacturer’s instructions with
slight modifications. HBSS containing rolipram (5 mm), adenosine
deaminase (0.5 UmLꢀ1) and with and without varying concentra-
tions of test compounds were added to 384-well plates followed
by cells (5000/well) and cAMP antibody, and then the plates were
incubated for 10 min at 378C. At the end of the incubation, lysis
buffer and luminescence reagents were added according to the
manufacturer’s instructions, and the samples were incubated over-
night at room temperature. The luminescence was determined
using a BioTek Synergy 2 plate reader. The amount of cAMP per
well was measured from a standard curve determined in parallel
incubations containing known amounts of cAMP (1 nm to 10 mm).
All test compounds were dissolved in DMSO with an equal volume
of solvent added to control incubations. The final DSMO concen-
tration was 5%.
Receptor binding assays: The preparation of DDT1 MF-2 cell mem-
branes and the ability of the test compounds to displace [3H]-8-cy-
clopentyl-1,3-dipropylxanthine from the A1AR were determined as
previously described.[22] The ability of the test compounds to dis-
place [125I]-(ꢀ)-iodopindolol from the b2AR in the DDT cell mem-
branes was determined as previously described.[3] For both recep-
tor binding assays, 5’-guanylyl-imidodiphosphate (10 mm) was in-
cluded to maintain the receptors in the agonist low-affinity state.
[19] P. S. Portoghese, G. Ronsisvalle, D. L. Larson, C. B. Yim, L. M. Sayre, A. E.
Takemori, Life Sci. 1982, 31, 1283–1286.
[20] P. S. Portoghese, D. L. Larson, C. B. Yim, L. M. Sayre, G. Ronsisvalle, A. W.
Lipkowski, A. E. Takemori, K. C. Rice, S. W. Tam, J. Med. Chem. 1985, 28,
1140–1141.
[21] V. Ramkumar, W. W. Barrington, K. A. Jacobson, G. L. Stiles, Mol. Pharma-
col. 1990, 37, 149–156.
[22] J. Zhang, L. Belardinelli, K. A. Jacobson, D. H. Otero, S. P. Baker, Mol.
Pharmacol. 1997, 52, 491–498.
Data analysis: The concentration of test compounds that stimulat-
ed or inhibited cAMP accumulation (EC50) was determined by non-
linear regression analysis of the concentration–response curves
using the GraphPad Prism 3.0 program. The intrinsic activity (IA) of
the compounds for stimulating or inhibiting cAMP accumulation
was calculated as a fraction of the maximal stimulation produced
by (ꢀ)-isoproterenol or maximal inhibition produced by CPA. The
concentration of compounds that inhibited radioligand binding by
50% (IC50) was also determined by the GraphPad program. The dis-
sociation constant (Ki) of the compounds for each receptor was
then calculated using the IC50 values as described by Cheng and
Prusoff.[23]
[23] Y. Cheng, W. H. Prusoff, Biochem. Pharmacol. 1973, 22, 3099–3108.
Received: June 28, 2013
Published online on && &&, 0000
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&
11
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
N. Barlow, S. P. Baker,* P. J. Scammells*
A molecular double date: Heterobiva-
lent ligands containing pharmacophores
designed to interact with both the A1
adenosine receptor (A1AR) and the b2
adrenergic receptor (b2AR) were pre-
pared. The affinity and potency of these
ligands at both receptors were found to
be dependent upon the linker length
and composition. The data suggest that
heterobivalent ligands for receptors
mediating opposite responses might be
useful for investigating the regulation of
receptor cross-talk in health and
&& – &&
Effect of Linker Length and
Composition on Heterobivalent
Ligand-Mediated Receptor Cross-Talk
between the A1 Adenosine and b2
Adrenergic Receptors
disease.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 12
&12
&
ÞÞ
These are not the final page numbers!