Synthesis and Evaluation of Dimeric Derivatives of Diacylglycerol-Lactones as Protein Kinase C Ligands

Article abstract of DOI:10.1021/acs.bioconjchem.7b00299

Protein kinase C (PKC) mediates a central cellular signal transduction pathway involved in disorders such as cancer and Alzheimer's disease. PKC is regulated by binding of the second messenger sn-1,2-diacylglycerol (DAG) to its tandem C1 domains, designated C1a and C1b, leading both to PKC activation and to its translocation to the plasma membrane and to internal organelles. Depending on the isoform, there may be differences in the ligand selectivity of the C1a and C1b domains, and there is different spacing between the C1 domains of the conventional and novel PKCs. Bivalent ligands have the potential to exploit these differences between isoforms, yielding isoform selectivity. In the present study, we describe the synthesis of a series of dimeric derivatives of conformationally constrained diacylglycerol (DAG) analogs (DAG-lactones). We characterize the derivatives in vitro for their binding affinities, both to a single C1 domain (the C1b domain of PKCδ) as well as to the conventional PKCα isoform and the novel PKCδ isoform, and we measure their abilities to cause translocation of PKCδ and PKC? in intact cells. The dimeric compound with the 10-carbon linker was modestly more effective for the isolated PKCδ C1b domain than was the monomeric compound. For the intact PKCα and PKCδ, the shortest DAG-lactone dimer had similar affinity to the monomer and affinity decreased progressively up to the 16-carbon linker. The dimeric derivatives did not cause the Golgi accumulation of PKCδ. The present results provide important insights into the development of new chemical tools for biological studies on PKC.

Full text of DOI:10.1021/acs.bioconjchem.7b00299

Subscriber access provided by UNIVERSITY OF CONNECTICUT
Article
Synthesis and evaluation of dimeric derivatives
of diacylglycerol-lactones as PKC ligands
Nami Ohashi, Ryosuke Kobayashi, Wataru Nomura, Takuya Kobayakawa, Agnes
Czikora, Brienna Herold, Nancy Lewin, Peter M. Blumberg, and Hirokazu Tamamura
Bioconjugate Chem., Just Accepted Manuscript • DOI: 10.1021/acs.bioconjchem.7b00299 • Publication Date (Web): 03 Jul 2017
Downloaded from http://pubs.acs.org on July 6, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Bioconjugate Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis and Evaluation of Dimeric Derivatives of
DiacylglycerolꢀLactones as PKC Ligands
†
†
†
†
‡
Nami Ohashi , Ryosuke Kobayashi , Wataru Nomura , Takuya Kobayakawa , Agnes Czikora ,
‡
‡
‡
†
Brienna K Herold , Nancy E. Lewin , Peter M. Blumberg, and Hirokazu Tamamura
†
Department of Medicinal Chemistry, Institute of Biomaterials and Bioengineering, Tokyo
Medical and Dental University, 2ꢀ3ꢀ10 Kandasurugadai, Chiyodaꢀku, Tokyo 101ꢀ0062, Japan
‡
Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer
Institute, National Institutes of Health, Bethesda, Maryland 20892, USA
Key words: C1 domains, protein kinase C, phorbol ester, drug design
Correspondence should be addressed to tamamura.mr@tmd.ac.jp.
ACS Paragon Plus Environment
1

Bioconjugate Chemistry
Page 2 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A
BSTRACT: Protein kinase C (PKC) mediates a central cellular signal transduction pathway
involved in disorders such as cancer and Alzheimer’s disease. PKC is regulated by binding of
the second messenger snꢀ1,2ꢀdiacylglycerol (DAG) to its tandem C1 domains, designated C1a
and C1b, leading both to PKC activation and to its translocation to the plasma membrane and to
internal organelles. Depending on the isoform, there may be differences in the ligand selectivity
of the C1a and C1b domains and there is different spacing between the C1 domains of the
conventional and novel PKCs. Bivalent ligands have the potential to exploit these differences
between isoforms, yielding isoform selectivity. In the present study, we describe the synthesis
of a series of dimeric derivatives of conformationally constrained diacylglycerol (DAG) analogs
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
DAGꢀlactones). We characterize the derivatives in vitro for their binding affinities both to a
single C1 domain (the C1b domain of PKCδ) as well as to the conventional PKCα isoform and
the novel PKCδ isoform and we measure their abilities to cause translocation of PKCδ and
PKCε in intact cells. The dimeric compound with the 10 carbon linker was modestly more
effective for the isolated PKCδ C1b domain than was the monomeric compound. For the intact
PKCα and PKCδ, the shortest DAGꢀlactone dimer had similar affinity to the monomer and
affinity decreased progressively up to the 16 carbon linker. The dimeric derivatives did not
cause Golgi accumulation of PKCδ. The present results provide important insights into
development of new chemical tools for biological studies on PKC.
ACS Paragon Plus Environment
2

Page 3 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
I
NTRODUCTION
Protein kinase C (PKC) is a serine/threonine protein kinase that is involved in cellular signal
1
2
3
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
transduction leading to proliferation, differentiation, migration and apoptosis. PKC is an
important therapeutic target because of its involvement in disorders such as diabetes, cancer,
5ꢀ8
heart diseases, autoimmune diseases and Alzheimer’s disease. The isozymes in the PKC
family are divided into conventional PKCs (cPKC) (α, βI/II, γ), novel PKCs (nPKC) (δ, ε, η, θ)
9
and atypical PKCs (aPKC) (ζ, ι/λ). With the exception of the aPKCs, the cPKCs and nPKCs are
regulated by ligand binding through their tandem C1 (C1a, C1b) domains. In the case of the
2
+
cPKCs, the additional binding of Ca to the C2 domain is required. The inactive form of PKC,
which has the substrate binding domain capped by its own pseudosubstrate, is located in cytosol.
Upon ligand binding, PKC shows translocation from cytosol to the plasma membrane and
internal membranes. The endogenous ligand for PKC is snꢀ1,2ꢀdiacylglycerol (DAG), which is
a second messenger generated downstream of receptor tyrosine kinases and Gꢀprotein coupled
10
receptors. DAG is produced at the inner face of the plasma membranes and its binding to the
11,12
C1 domain causes a conformational change of PKC into the active form and translocation
13ꢀ15
followed by signaling through multiple downstream pathways.
The mechanism of action of diacylglycerol and of the phorbol esters, their ultrapotent
16
analogs, is well understood. The ligand fits into a hydrophilic cleft at the top of the C1 domain,
completing a hydrophobic surface and contributing additional hydrophobicity through its acyl
side chains, thereby promoting membrane insertion of the C1 domain. This ternary complex of
diacylglycerol – C1 domain – phospholipid bilayer is driven by hydrogen bonds between the
peptide backbone of the binding cleft with the primary hydroxyl group of the diacylglycerol and
one of its carbonyls, by hydrogen bonding between the second carbonyl of the diacylglycerol
1
7
with the phospholipid head groups, and by hydrophobic interactions of the lipid acyl chains
ACS Paragon Plus Environment
3

Bioconjugate Chemistry
Page 4 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
8,19
with both the acyl chains of the diacylglycerol and the face of the C1 domain.
The influence
of the side chains of diacylglycerols and diacylglycerol lactones on binding potency and
20,21
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
biological outcome has been extensively studied.
Diversity among C1 domains provides multiple potential opportunities for establishing
selectivity among the isoforms of PKC as well as among members of the seven families of
signaling proteins with DAGꢀresponsive C1 domains. Some, such as PKC and PKD, have two
C1 domains. Others, such as RasGRP or the chimaerins, have only a single C1 domain. Among
those with tandem C1 domains, the distance between them differs between the conventional
PKCs, the novel PKCs and the PKD isoforms. The separations for the conventional and novel
PKCs are 14 and 21ꢀ22 amino acids respectively. For PKD, the separation is considerably
greater, 73 amino acids. Unfortunately, the lack of complete crystallographic or cryoꢀEM
solution structures, even less of such structures in complex with ligand and membranes, means
that the optimal linker length for a dimeric ligand remains unknown. For those targets with two
C1 domains, viz. PKC and PKD, the affinities for phorbol ester may differ between the C1a and
C1b domains and the structure activity relations may be different. Furthermore, as suggested by
mutational analysis, the role of the C1a and C1b domains may play different roles in the PKC
2
2ꢀ24
translocation induced by phorbol 12ꢀmyristate 13ꢀacetate (PMA).
have the potential to be exploited through appropriate bivalent ligands.
To date, several bivalent PKC ligands based on naphthylpyrrolidone, phorbol ester
All of these differences
2
5
26,27
2
8
and isobenzofuranone have been synthesized and evaluated for their PKC binding and
membrane interaction. In no case has there been substantial capture of enhanced affinity
suggestive of the simultaneous binding to both C1 domains. A complication in all of the above
studies is that the linker simultaneously fulfils two functions; it provides a potential limitation
on the separation of the two binding moieties and it provides the hydrophobicity which is well
ACS Paragon Plus Environment
4

Page 5 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
characterized as being required for ligand activity. In the present study, a separate hydrophobic
moiety is included in each of the binding moieties, although the linker also provides additional
hydrophobicity. Additionally, the PKC translocation induced by bivalent ligands has not been
investigated previously. In view of the functional differences between the C1a and C1b domains,
monovalent and bivalent ligands might have different effects on the translocation of PKC. For a
binding moiety in the present study DAGꢀlactones were chosen. DAGꢀlactones are
conformationally constrained diacylglycerol analogs which are synthetically tractable and have
been optimized to provide potent PKC binding affinity as well as selectivity between classes of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
9
C1 domain containing targets. We describe here the design and synthesis of dimeric
derivatives of DAGꢀlactones and characterize their PKC binding affinities and effects on PKC
translocation.
ACS Paragon Plus Environment
5

Bioconjugate Chemistry
Page 6 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
R
ESULTS AND DISCUSSION
Design and Synthesis of Dimeric Derivatives of DAGꢀLactones. To obtain a monomeric
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
moiety derived from DAG with sufficient affinity for PKC, an αꢀalkylidine group was
connected to the snꢀ2 carbonyl group on the DAGꢀlactone template and an alkyl chain
containing 9 carbons was attached to the snꢀ1 carbonyl group according to the previous reports
2
9ꢀ32
(
Figure 1).
The alkyl chain moiety is considered to be necessary to interact with cell
membranes or the C1 domain and was chosen to provide appropriate lipophilicity as determined
2
0
from previous studies. A unique feature of this series of compounds, compared to other series
previously examined, is that the 9 carbon alkyl chain is present both in our monomeric ligand as
well as in the dimeric derivatives. In previous studies, only the linker provided substantial
hydrophobicity for the dimeric derivatives. For choice of a linker moiety, a previous report
concerning dimeric benzolactams found that dimeric ligands with alkyl chains containing linear
14 carbons as linkers increased by 200ꢀfold their binding affinities for PKCα and for PKCδ
compared to the monomeric ligands. In addition, replacement of the alkyl chain with an
2
5
oligoethylene glycol chain caused a 4000ꢀ to 7000ꢀfold decrease in affinity for PKCα. Thus, in
the present study alkyl chains were adopted as linkers for the dimeric DAGꢀlactones. Olefin
metathesis was utilized for concise synthesis of the dimeric form of DAGꢀlactones, and the
motion of the resulting linker moiety with olefin was restricted.
ACS Paragon Plus Environment
6

Page 7 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
Figure 1. The design of the dimeric derivatives of DAGꢀlactones (6aꢀg) and the structure of the
monomer (5h). DAGꢀlactones, which are C1(a and b) domain binding moieties, are shown
inside light brown ellipses.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The synthesis of dimeric derivatives of DAGꢀlactones is described in Scheme 1. Several
alkyl halides containing the terminal olefin were introduced at the αꢀposition of tertꢀbutyl
decanoate 2 to yield αꢀalkylated compounds 3aꢀg. After removal of the tertꢀbutyl group of 3aꢀg
by TFA treatment, the resulting acids were conjugated with the hydroxy group of a
33,34
DAGꢀlactone derivative 4
to yield the esters 5aꢀg by the conversion of carboxylic acids to
acid chlorides using oxalyl chloride and DMF. Intermolecular olefin metathesis using the
second generation Grubbs catalyst gave dimeric derivatives of DAGꢀlactones 6aꢀg. ESIꢀTOF
mass spectra of 6aꢀg were measured after flash column chromatography, indicating that the
mixture of byproducts with the loss of methylene units occurred by the isomerization in the
35
olefin metathesis process. The products with impurities were employed for biological
36,37
experiments.
ACS Paragon Plus Environment
7

Bioconjugate Chemistry
Page 8 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. The synthesis of dimeric derivatives of DAGꢀlactones. Reagents and conditions: (a)
tertꢀbutyl 2,2,2ꢀtrichloroacetimidate, BF
3
·OEt
ꢀbromoꢀ1ꢀbutene for 3a: 47%, 5ꢀbromoꢀ1ꢀpentene for 3b: 93%, 6ꢀbromoꢀ1ꢀhexene for 3c: 66%,
ꢀbromoꢀ1ꢀheptene for 3d: 72%, 8ꢀbromoꢀ1ꢀoctene for 3e: 66%, 9ꢀbromoꢀ1ꢀnonene for 3f: 38%,
0ꢀbromoꢀ1ꢀdecene for 3g: 59%. (c) TFA, CH Cl . (d) (COCl) , cat. DMF, CH Cl . (e) 4, Et N,
2 2 2
, CH Cl , 45%. (b) LDA, TBAI, THF/HMPA,
4
7
1
2
2
2
2
2
3
THF, 5a: 35%, 5b: 52%, 5c: 41%, 5d: 58%, 5e: 54%, 5f: 27%, 5g: 35% (3 steps). (f) the second
generation Grubbs catalyst, CH Cl . Optical purities of products were not determined. Yields
2
2
are described in Materials and Methods.
Evaluation of the Binding Affinities of the Dimeric Derivatives of DAGꢀLactones for
PKCδ or for Its C1b Domain. The binding affinities of the synthetic compounds were assessed
ACS Paragon Plus Environment
8

Page 9 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
3
by their ability to competitively inhibit [ H]PDBu binding to PKC and to its C1 domain. The K
i
values were determined both for the PKCδ C1b (δC1b) domain as well as for the full length
recombinant PKCδ and PKCα (Table 1, Figure 2). While the standard practice in the field is
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
that K
i
values are reported in nM units, it is recognized that the C1 domains recognize the
38
membrane partitioned ligand rather than that in aqueous solution. Nonetheless, under both the
usual in vitro and cellular assay conditions, the membrane concentrations of ligand are
proportional to the nominal aqueous concentration, thus this measure is appropriate for
prediction of cellular response.
Table 1. Ki and CLogP values of the monomeric DAGꢀlactone and its dimeric derivatives.
number of
carbon atoms
in the linker
Purity PKCδ C1b
PKCδ
PKCα
ClogP
a
b
b
b
compound
(%)
K
i
(nM)
K
i
(nM)
i
K (nM)
5h
monomer
5.47 ± 0.81 2.43 ± 0.10 10.4 ± 0.31
2.07 ± 0.08 2.67 ± 0.59 8.4 ± 1.4
4.54
6a
6
73
43
46
47
40
32
39
10.65
11.71
12.77
13.82
14.88
15.94
16.70
6b
8
1.17 ± 0.05 3.11 ± 0.22 15.9 ± 1.4
0.80 ± 0.05 6.31 ± 1.30 17.4 ± 5.2
6
c
10
12
14
16
18
6d
1.17 ± 0.34 13.2 ± 1.9
1.86 ± 0.41 16.6 ± 2.1
2.52 ± 0.46 27.2 ± 2.0
3.00 ± 0.76 12.3 ± 2.1
50 ± 13
48.5 ± 6.3
173± 17
6e
6f
6g
30.9 ± 3.3
a
b
Purities were analyzed by ESIꢀmass spectra.
values were measured by a competitive binding assay against [ H]PDBu binding and are the
3
K
i
mean ± SEM of three independent experiments.
ACS Paragon Plus Environment
9

Bioconjugate Chemistry
Page 10 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Relation between the binding affinity of the dimeric derivatives of DAGꢀlactones and
3
the carbon chain length of linkers. K
i
values were determined against [ H]PDBu binding to full
length PKC or the C1b domain. Bars indicate the value from triplicate independent experiments
of the K ± SEM for the C1b domain of PKCδ (black), full length PKCδ (dark gray) or full
i
length PKCα (light gray). [Change 0 to “monomer” in the figure]
The full length PKCδ and PKCα each contain two individual C1 domains capable of
binding to DAGꢀlactones but the PKCs differ in the separation between the two domains. The
3
K values for binding of [ H]PDBu to the PKCδ C1b and to the full length PKCδ were 0.19 nM
d
and 0.33 nM, respectively. All of the dimeric derivatives of DAGꢀlactones (6aꢀg) showed
ACS Paragon Plus Environment
10

Page 11 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
stronger binding affinity for the δC1b domain than did the monomer 5h. Three factors should
contribute to this enhanced affinity. First, the dimeric constructs contain two DAGꢀlactone
moieties per molecule, whereas the monomeric DAGꢀlactone contains only one. If everything
else remained equal, this twoꢀfold increase in the relative concentration of the DAGꢀlactone
moiety would predict a 2ꢀfold increase in apparent potency. Indeed, at least a twoꢀfold
enhancement was observed for linkers between 6 and 16 carbons, and even the 18 carbon linker
afforded an enhancement of 1.8ꢀfold. Second, lipophilicity has been clearly shown to influence
binding potency of DAGꢀlactones, with reduced potency if the lipophilicity is either too low or
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20
too high. While CLogP values below 4 were clearly disadvantageous, the effect of CLogP
20
values above 5ꢀ6 were reported to be modest. For comparison, the CLogP of the monomer, 5h
was 4.54 and all of the dimeric compounds were above this range (e.g. 6a, CLogP, 10.65, values
calculated by ChemDraw 15.1). The dimeric DAGꢀlactone containing a 10 carbon atoms (10C)
linker (6c: K
compounds. Significantly lower affinity for δC1b was observed as the linker length was longer
or shorter than 10C (K [nM]: 6a (6C); 2.07, 6b (8C); 1.17, 6d (12C); 1.17, 6e (14C); 1.86, 6f
16C); 2.52, 6g (18C); 3.00). The results suggest that the 10 carbon linker, which has a distance
i
= 0.80 nM) showed the highest binding affinity for δC1b among all of the tested
i
(
of approximately 14 Å (by Spartan’14; Wavefunction, Irvine, CA, USA), might be capturing
simultaneously two molecules of the C1b domain. This is consistent with a hypothetical
distance of 15 Å between two centers of the adjacently coherent C1a and C1b domains, which
was obtained from the model of the C1a and C1b domains that was calculated by Newton and
coworkers using two C1b crystal structures without the sequence located between the C1a and
26
C1b domains. The decrease of the interaction with δC1b of the dimeric DAGꢀlactones (6a and
6
b) containing shorter linkers than 10 carbons might be caused by the repulsion of two
molecules of the C1b domain, thus 6 carbons and 8 carbons are exceedingly short to capture two
ACS Paragon Plus Environment
11

Bioconjugate Chemistry
Page 12 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
δC1b domains simultaneously. On the other hand, the decrease of the interaction with δC1b of
the dimeric DAGꢀlactones (6dꢀg) containing longer linkers than 10 carbons might be caused by
interference of the long linker with the proper spread of the binding moieties, since the long,
hydrophobic linker might be expected to insert into the lipid bilayer rather than to extend at the
membrane surface. It is important to note that the DAGꢀlactones are racemic and that binding to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
9
PKC is stereospecific. Consequently, only one fourth of the dimeric molecules would actually
have two active binding moieties capable of binding to two C1 domains. Half would have a
single active moiety, and a fourth would be entirely inactive. Of course, an alternative
interpretation for the optimum binding by the dimeric derivative with a linker length of 10
carbons is that the hydrophobicity is simply optimal.
In the case of full length recombinant PKCδ, the dimeric derivatives of DAGꢀlactones
containing 6 carbon (6a: K
values similar to that for the monomer 5h (K
16 carbons, the affinity for PKCδ decreased (K
i
= 2.67 nM) and 8 carbon (6b: K
= 2.43 nM). As the linker length increased up to
[nM]: 6c (10C); 6.31, 6d (12C); 13.2, 6e (14C);
6.6, 6f (16C); 27.1), with some increase in binding affinity then occurring for the dimeric
i i
= 3.11 nM) linkers showed K
i
i
1
DAGꢀlactone with the 18 carbon linker (6g, K
i
= 12.3 nM). For PKCα, the dependence of the K
i
values on the number of carbons of the linker showed a similar pattern to that for PKCδ (K
i
[
nM]: 6a (6C); 8.40, 6b (8C); 15.9, 6c (10C); 17.4, 6d (12C); 49.6, 6e (14C); 48.5, 6f (16C);
173, 6g (18C), 30.9), indicating no evidence for isozyme selectivity (Figure 2). The clear
interpretation is that there is no strong evidence for both DAGꢀlactone moieties in the dimeric
constructs binding simultaneously to the C1a and C1b domains of either PKC isoform. One
possible explanation is that, because the binding represents a ternary complex of DAGꢀlactone,
PKC, and lipid bilayer, the constraints imposed by the lipid bilayer interfere with positioning of
the second binding moiety of the dimeric DAGꢀlactone with the second C1 domain in the intact
ACS Paragon Plus Environment
12

Page 13 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
PKC. One possible source of this interference is that the acyl linker might be expected to insert
into the lipid bilayer so as to interact with the acyl chains of the phospholipids. This interaction
would draw the two DAGꢀlactone moieties together, preventing the desired separation predicted
for an extended chain. As a further consequence, the second, now incorrectly positioned
DAGꢀlactone moiety might then generate steric interference with the remainder of the PKC. A
further factor could be that the DAGꢀlactone moieties showed a sufficient difference in affinity
for the C1a and C1b domains so that they only bound to the preferred domain although their
positioning would permit binding to both. The rationale is that the C1a and C1b domains
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
d
display different affinities for phorbol esters and for DAG. Thus, we reported K values of
PDBu for the isolated PKCδ C1a and C1b domains of 2.0 nM and 0.33 nM for a 6ꢀfold
difference and 130 nM versus 2.8 nM, for a 46ꢀfold difference, in full length PKCδ constructs
4
0
possessing only a C1a domain or a C1b domain, respectively (with the other domain deleted).
The difference was even greater for DAG, with values of 2,400 nM versus 13 nM (difference of
185 fold) for the full length PKCδ constructs possessing only a C1a domain or a C1b domain,
respectively.
Observation of PKC Translocation Caused by the Dimeric Derivatives of
DAGꢀLactones. PKC translocation induced by ligand binding was observed using a confocal
laser scanning microscope. CHO Flpꢀin cells that stably express EGFPꢀtagged PKCδ were
utilized. The effect on the translocation of EGFPꢀPKCδ by the addition of 10 ꢀM of the dimeric
derivative of the DAGꢀlactone containing the 6 carbon linker (6a), the monomer DAGꢀlactone
(
5h) and PDBu are shown in Figure 3(a). The addition of the dimer (6a) caused the
translocation of EGFPꢀPKCδ from cytosol to plasma membranes. The monomer (5h) also
3
2
caused the translocation of PKCδ, however, the translocation caused by 6a was more clearly
observed than that by 5h at 5 min after the addition of ligands. The translocation to plasma
ACS Paragon Plus Environment
13

Bioconjugate Chemistry
Page 14 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
membranes induced by PDBu was not observed more clearly than those by monomeric and
dimeric DAGꢀlactones. To examine the effect of the longer linker length on the PKCδ
translocation, experiments employing dimeric DAGꢀlactones, which contain the 10 carbon (6c),
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
12 carbon (6d) and 18 carbon (6g) linkers, were performed (Figure 3b). The membrane
translocation induced by 6c was observed similarly or slightly more clearly compared to that
induced by 6a. The best translocation was seen for 6c (10 carbons) with some translocation still
evident for 6d (12 carbons) and 6g (18 carbons). Interestingly, the potent dimeric DAGꢀlactones
only induced plasma membrane translocation. In contrast, both the monomeric DAGꢀlactone
and the PDBu also showed translocation adjacent to the nucleus, consistent with Golgi
localization.
(
a)
(
b)
ACS Paragon Plus Environment
14

Page 15 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Translocation of EGFPꢀtagged PKCδ by the addition of compounds in transiently
transfected CHO cells. (a) EGFPꢀPKCδꢀexpressing CHO cells were treated with 10 ꢀM
compound (6a: the dimeric derivative of DAGꢀlactone containing the 6 carbon linker, 5h: the
monomer of DAGꢀlactone, PDBu: phorbol 12,13ꢀdibutyrate). Fluorescent images were
observed with confocal fluorescence microscopy at 0, 5 and 10 min after the addition of
compounds. (b) EGFPꢀPKCδꢀexpressing CHO cells were treated with dimeric derivatives of
DAGꢀlactones containing longer linkers (10 ꢀM) (6c: 10 carbon linker, 6d: 12 carbon linker,
6g: 18 carbon linker). DIC: differential interference contrast.
To check whether another PKC isoform might behave substantially differently, limited
experiments were also conducted with another PKC isoform, PKCε, using the human pancreatic
cancer cell line LNCaP. PKCε provides a contrast to PKCδ in several respects. The C1a and
4
1
C1b domains of PKCε have affinities within 3ꢀfold of one another for PMA and for DAG.
ACS Paragon Plus Environment
15

Bioconjugate Chemistry
Page 16 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
PKCε has a less complicated translocation profile than does PKCδ, in that it only translocates to
the plasma membrane in response to PMA (phorbol 12ꢀmyristate 13ꢀacetate), whereas PKCδ
translocates to both plasma membrane and internal membranes in response to PMA, depending
on time of exposure and the extent to which PMA has equilibrated with internal membranes.
Finally, the LNCaP cells possess a different phospholipid composition in their plasma
membrane than do CHO cells, with greater negative charge, since they are mutant in PTEN and
thus have elevated levels of phosphatidylinositolꢀ3,4,5ꢀtriphosphate. Despite these differences,
the pattern of response was similar to that seen for PKCδ in the CHO cells (Figure 4). PMA (1
µM) gave dramatic plasma membrane translocation of PKCε. Modest translocation was
observed for the monomer and somewhat less for 6a (6 carbons). The other compounds failed to
show translocation except for faint translocation for 6g (18 carbons).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
16

Page 17 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Translocation of EGFPꢀtagged PKCε by the addition of compounds in transiently
transfected LNCaP cells. Cells were treated with PMA (phorbol 12ꢀmyristate 13ꢀacetate, 1 µM),
with the monomeric DAGꢀlactone, or with the dimeric DAGꢀlactones 6a and 6g (3 µM) and
imaged by confocal microscopy at the indicated times. Images are from a single experiment.
C
ONCLUSIONS
The dimeric derivatives of the DAGꢀlactones bearing the same DAGꢀlactone moieties linked
ACS Paragon Plus Environment
17

Bioconjugate Chemistry
Page 18 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with variable lengths of an alkyl linker were synthesized and evaluated for their PKC binding
and translocation activity. A very impressive result was that it was possible to achieve
subnanomolar potencies using a DAGꢀlactone moiety. This is significant because the
DAGꢀlactone scaffold is far more synthetically accessible than that of the phorbol esters or other
high affinity PKC ligands.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As already discussed, one of the elements for the design of selective dimeric ligands would
be optimization of the two binding moieties for the C1a and C1b domains, and the
DAGꢀlactones are well suited for such optimization. Like the other dimeric constructs reported
using other scaffolds, the current compounds failed to achieve a dramatic gain in binding
affinity over the monomeric construct. A plausible explanation is that, after binding of the first
DAGꢀlactone moiety, the linker did not favorably position the second DAGꢀlactone moiety to
interact with the second C1 domain. Precise interpretation is of course not possible because the
linker changes the lipophilicity of the compound and it has been clearly established that
lipophilicity is a critical parameter for binding affinity, with reduced affinity if a compound is
either not sufficiently lipophilic or is too lipophilic. This could well explain why the dimeric
compound with the 10 carbon linker was modestly more effective for the isolated PKCδ C1b
domain than was the monomeric compound or the dimeric compounds with either shorter or
longer linkers. Alternatively, it is possible that a 10 carbon linker or longer was sufficient to
allow access of a second isolated C1b domain to the remaining free DAGꢀlactone moiety,
accounting for the improved affinity over dimers with a shorter linker, and that the decrease
with the longer linkers was related to excess lipophilicity.
For the intact PKCα and PKCδ the behavior was distinctly different from that of the
isolated PKCδ C1b domain. Here, the shortest DAGꢀlactone dimer had similar affinity to the
monomer and affinity decreased progressively up to the 16 carbon linker. The affinity then
ACS Paragon Plus Environment
18

Page 19 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
increased slightly for the 18 carbon linker, yielding an affinity similar to that of the compound
with the 14 carbon linker. This pattern of behavior suggests that the dimeric compound is only
binding to the intact PKC through a single binding moiety and the absolute difference in affinity
compared to that of the isolated C1b domain is not surprising, since in the intact PKC ligand
binding is coupled to conformational changes in the protein.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The translocation results are again quite promising for evaluation of the synthetic strategy.
Translocation of PKCδ and of PKCε by the most active derivatives was observed and behavior
largely mirrored that found in the binding assays. A noteworthy finding was that the dimeric
derivatives did not cause Golgi accumulation of PKCδ, unlike the monomeric derivative and
PDBu. This difference probably reflects the slower penetration of the dimeric derivatives into
the interior of the cells over the time course examined.
Our current strategy yielded potent compounds suitable for evaluation in vitro and in intact
living cells. Despite the use of linkers predicted to permit a separation that could span the
distance between the tandem C1 domains in intact PKC isoforms, we saw little evidence for
actual bivalent binding. An underlying concern is that aliphatic linkers, being structurally
flexible, in the presence of membranes will themselves insert into the membranes, collapsing
the actual separation between the two binding domains. A further problem may be that, by using
a hydrophobic linker region, it complicates the optimization of lipophilicity.
Several strategies may help to address these concerns. For linkers, a polyproline structure
might provide rigidity. An oligoethylene glycol linker could provide a hydrophilic linker that
would sit on top of the membrane surface, with the required hydrophobicity provided separately
by acyl chains such as those in the current structures. While bivalent benzolactams with an
oligoethylene glycol linker showed very little activity, these compounds did not have additional
elements to render the binding domains hydrophobic. Finally, heteroꢀbivalent ligands containing
ACS Paragon Plus Environment
19

Bioconjugate Chemistry
Page 20 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
DAGꢀlactones and other C1aꢀselective ligands could be designed to efficiently bind to the C1a
as well as to C1b domains. Our present study provides important insights into development of
new chemical tools for biological studies on PKC and useful information for discovery of
therapeutic agents.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
20

Page 21 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
MATERIALS AND METHODS
General. All reactions utilizing airꢀ or moistureꢀsensitive reagents were performed in dried
glassware under an atmosphere of nitrogen, using commercially supplied solvents and reagents
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2 2 2
unless otherwise noted. CH Cl was distilled from CaH and stored over molecular sieves. For
thinꢀlayer chromatography (TLC), precoated silica gel plates (Merck 60F254) were employed.
TLC plates were visualized by fluorescence quenching under UV light and by staining with
phosphomolybdic acid, pꢀanisaldehyde, or ninhydrin. Flash chromatography was performed
with Wakogel Cꢀ200 (Wako Pure Chemical Industries, Ltd., Japan) and silica gel 60 N (Kanto
1
13
Chemical Co., Inc., Japan). HꢀNMR (400 or 500 MHz) and CꢀNMR (125 MHz) spectra were
recorded on a Bruker Avance III 400 spectrometer or a Bruker AVANCE 500 spectrometer.
4 3
Chemical shifts are reported in δ (ppm) relative to Me Si (in CDCl ) as the internal standard.
Infrared (IR) spectra were recorded on a JASCO FT/IR 4100, and are reported as wavenumber
–
1
(
cm ). Lowꢀ and highꢀresolution mass spectra were recorded on a Bruker Daltonics
micrOTOFꢀ2focus (ESIꢀMS) spectrometer in the positive or negative detection mode. For
confocal laser scanning fluorescence images, cells were observed with a FluoView FV10i laser
scanning confocal microscope (OLYMPUS, Japan) or a Zeiss LSM 510 confocal microscopy
system (Carl Zeiss, Inc.).
Synthesis of DAGꢀlactone Derivatives.
tertꢀButyl 2ꢀ(butꢀ3ꢀenꢀ1ꢀyl)decanoate (3a) (Representative compound of 3aꢀ3g): LDA was
prepared by addition of nꢀBuLi in hexane (2.56 mL of 1.64 M solution, 4.2 mmol, 1.4 eq) to a
solution of diisopropylamine (0.63 mL, 4.5 mmol, 1.5 eq.) in THF (1.26 mL) at 0 °C, and the
mixture allowed to sit for 10 min at room temperature. To the reaction mixture, 2 (0.685 g, 3
mmol, 1 eq.) dissolved with THF (0.34 mL) was added at room temperature, and the reaction
o
mixture was stirred for 30 min at ꢀ78 C. To the mixture were then added TBAI (50 mol%),
ACS Paragon Plus Environment
21

Bioconjugate Chemistry
Page 22 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HMPA (0.4 mL) and 4ꢀbromoꢀ1ꢀbutene (0.48 mL, 4.8 mmol, 1.6 eq.) at this temperature. After
stirring, the reaction mixture was warmed to ambient temperature and stirred for 40 h. The
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
reaction was diluted with hexane and quenched with saturated aqueous NH Cl. The aqueous
layer was extracted three times with EtOAc. The combined organic layer was dried over MgSO4.
Concentration under reduced pressure followed by flash chromatography (hexane/EtOAc =
1
3
0:1) gave compound 3a (0.4 g, 47%). R
CDCl
2H), 2.18ꢀ2.26 (m, 1H), 4.94ꢀ5.04 (m, 2H), 5.74ꢀ5.84 (m, 1H); CꢀNMR (125 MHz, CDCl
4.0, 22.6, 27.2, 28.1, 29.2, 29.5, 31.6, 31.8, 32.5, 35.6, 45.8, 79.8, 114.6, 138.1, 175.6; HRMS
f
0.5 (hexane/EtOAc = 30:1); HꢀNMR (400 MHz,
3
) δ 0.86ꢀ0.89 (m, 3H), 1.24ꢀ1.28 (m, 12H), 1.52 (s, 9H), 1.63ꢀ1.72 (m, 2H), 2.00ꢀ2.08 (m,
13
3
) δ
1
+
(
2
ESI), m/z calcd for C18H34NaO [M+Na] 305.2451, found 305.2446.
tertꢀButyl 2ꢀ(pentꢀ4ꢀenꢀ1ꢀyl)decanoate (3b): By use of a procedure similar to that described
for the synthesis of compound 3a, alkylation of 2 (0.685 g, 3 mmol) using 5ꢀbromoꢀ1ꢀpenten
f
(0.57 mL, 4.8 mmol) gave compound 3b (0.824 g, 93%). R 0.5 (hexane/EtOAc = 30:1);
1
HꢀNMR (400 MHz, CDCl ) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.30 (m, 12H), 1.34ꢀ1.41 (m, 4H), 1.44
3
(
(
s, 9H), 1.51ꢀ1.60 (m, 2H), 2.01ꢀ2.07 (m, 2H), 2.17ꢀ2.23 (m, 1H), 4.93ꢀ5.03 (m, 2H), 5.74ꢀ5.85
13
m, 1H); CꢀNMR (125 MHz, CDCl
3
) δ 14.0, 26.6, 27.3, 28.1 (2C), 29.2, 29.4, 29.5, 31.8, 32.0,
+
3
3
2.6, 33.6, 46.3, 79.7, 114.4, 138.6, 175.8; HRMS (ESI), m/z calcd for C19
19.2608, found 319.2605.
2
H36NaO [M+Na]
tertꢀButyl 2ꢀ(hexꢀ5ꢀenꢀ1ꢀyl)decanoate (3c): By use of a procedure similar to that described for
the synthesis of compound 3a, alkylation of 2 (0.685 g, 3 mmol) using 6ꢀbromoꢀ1ꢀhexene (0.64
1
mL, 4.8 mmol) gave compound 3c (0.618 g, 66%). R
400 MHz, CDCl
.51ꢀ1.59 (m, 2H), 2.01ꢀ2.07 (m, 2H), 2.16ꢀ2.22 (m, 1H), 4.92ꢀ5.01 (m, 2H), 5.74ꢀ5.84 (m, 1H);
f
0.5 (hexane/EtOAc = 30:1); HꢀNMR
(
3
) δ 0.86ꢀ0.89 (m, 3H), 1.24ꢀ1.29 (m, 12H), 1.35ꢀ1.41 (m, 6H), 1.44 (s, 9H),
1
1
3
CꢀNMR (125 MHz, CDCl
3
) δ 14.0, 22.6, 26.8, 27.3, 28.1, 28.8, 29.2, 29.4, 29.5, 31.8, 32.4,
ACS Paragon Plus Environment
22

Page 23 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
+
3
2
2.6, 33.6, 46.6, 79.7, 114.3, 138.8, 175.8; HRMS (ESI), m/z calcd for C20H38NaO [M+Na]
333.2764, found 333.2759.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
tertꢀButyl 2ꢀ(heptꢀ6ꢀenꢀ1ꢀyl)decanoate (3d): By use of a procedure similar to that described
for the synthesis of compound 3a, alkylation of 2 (0.685 g, 3 mmol) using 7ꢀbromoꢀ1ꢀheptene
(
0.73 mL, 4.8 mmol) gave compound 3d (0.699 g, 72%). R
f
0.5 (hexane/EtOAc = 30:1);
) δ 0.89ꢀ0.91 (m, 3H), 1.25ꢀ1.32 (m, 16H), 1.37ꢀ1.42 (m, 4H), 1.46
s, 9H), 1.54ꢀ1.59 (m, 2H), 2.03ꢀ2.08 (m, 2H), 2.18ꢀ2.24 (m, 1H), 4.94ꢀ5.03 (m, 2H), 5.79ꢀ5.88
1
HꢀNMR (400 MHz, CDCl
3
(
13
(m, 1H); CꢀNMR (125 MHz, CDCl
3
) δ 14.0, 22.6, 27.2, 27.3, 28.1, 29.0, 29.2, 29.4, 29.5, 31.8,
3
2.5, 32.6, 33.6, 46.5, 79.7, 114.1, 139.0, 176.0; HRMS (ESI), m/z calcd for C21H40NaO
2
+
[
M+Na] 347.2921, found 347.2916.
tertꢀButyl 2ꢀoctyldecꢀ9ꢀenoate (3e): By use of a procedure similar to that described for the
synthesis of compound 3a, alkylation of 2 (0.685 g, 3 mmol) using 8ꢀbromoꢀ1ꢀoctene (0.80 mL,
1
4.8 mmol) gave compound 3e (0.667 g, 66%). R
MHz, CDCl
.54ꢀ1.59 (m, 2H), 2.03ꢀ2.08 (m, 2H), 2.18ꢀ2.24 (m, 1H), 4.94ꢀ5.03 (m, 2H), 5.79ꢀ5.88 (m, 1H);
f
0.5 (hexane/EtOAc = 30:1); HꢀNMR (400
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.31 (m, 18H), 1.34ꢀ1.40 (m, 4H), 1.44 (s, 9H)
1
13
3
CꢀNMR (125 MHz, CDCl ) δ 14.0, 22.6, 27.3, 28.1 (2C), 28.8, 28.9, 29.2, 29.3, 29.4, 29.5,
3
1.8, 32.5, 32.6, 33.7, 46.5, 79.6, 114.1, 139.1, 175.9; HRMS (ESI), m/z calcd for C22H42NaO
2
+
[
M+Na] 361.3077, found 361.3077.
tertꢀButyl 2ꢀoctylundecꢀ10ꢀenoate (3f): By use of a procedure similar to that described for the
synthesis of compound 3a, alkylation of 2 (0.685 g, 3 mmol) using 9ꢀbromoꢀ1ꢀnonene (0.89 mL,
1
4
.8 mmol) gave compound 3f (0.401 g, 38%). R
MHz, CDCl
.51ꢀ1.59 (m, 2H), 2.01ꢀ2.06 (m, 2H), 2.16ꢀ2.20 (m, 1H), 4.91ꢀ5.00 (m, 2H), 5.77ꢀ5.85 (m, 1H);
f
0.5 (hexane/EtOAc = 30:1); HꢀNMR (400
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.30 (m, 20H), 1.35ꢀ1.39 (m, 4H), 1.44 (s, 9H)
1
1
3
CꢀNMR (125 MHz, CDCl
3
) δ 14.0, 14.1, 21.0, 22.6, 27.3, 28.8, 28.9, 29.1, 29.2, 29.3, 29.4,
ACS Paragon Plus Environment
23

Bioconjugate Chemistry
Page 24 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
9.5, 31.8, 32.6, 33.7, 46.5, 60.3, 79.6, 114.0, 139.1, 176.0; HRMS (ESI), m/z calcd for
+
C
23
H
45
O
2
[M+H] 353.3414, found 353.3414.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
tertꢀButyl 2ꢀoctyldodecꢀ11ꢀenoate (3g): By use of a procedure similar to that described for the
synthesis of compound 3g, alkylation of 2 (0.685 g, 3 mmol) using 10ꢀbromoꢀ1ꢀdecene (0.96
1
mL, 4.8 mmol) gave compound 3g (0.649 g, 59%). R
400 MHz, CDCl
.51ꢀ1.60 (m, 2H), 2.00ꢀ2.07 (m, 2H), 2.16ꢀ2.22 (m, 1H), 4.01ꢀ5.02 (m, 2H), 5.76ꢀ5.86 (m, 1H);
f
0.5 (hexane/EtOAc = 30:1); HꢀNMR
(
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.29 (m, 22H), 1.33ꢀ1.42 (m, 4H), 1.45 (s, 9H),
1
1
3
3
CꢀNMR (125 MHz, CDCl ) δ 14.0, 22.6, 27.3, 28.1 (2C), 29.0, 29.2, 29.3 (2C), 29.4, 29.5
(
2C), 31.8, 32.6 (2C), 33.7, 46.5, 79.6, 114.0, 139.1, 180.0; HRMS (ESI), m/z calcd for
+
C
24
H
47
O
2
[M+H] 367.3571, found 367.3567.
(
2ꢀ(Hydroxymethyl)ꢀ5ꢀoxoꢀ4ꢀ(propanꢀ2ꢀylidene)tetrahydrofuranꢀ2ꢀyl)methyl
ꢀoctyldecꢀ9ꢀenoate (5e) (Representative compound of 5aꢀ5g): To a solution of a tertꢀbutyl
ester derivative 3e (0.108 g, 0.32 mmol, 1 eq.) in CH Cl (2 mL) was added TFA (2 mL) at 0 °C,
2
2
2
and the mixture was stirred at room temperature for 1 h followed by concentration under
reduced pressure. The obtained carboxylic acid was used without further purification. To the
carboxylic acid in CH
2
Cl
2
(2.7 mL) were added oxalyl chloride (0.27 mL, 3.2 mmol, 10 eq.) and
o
cat. DMF (3 drops) at 0 C, and the mixture was then stirred at room temperature for 2 h
followed by concentration under reduced pressure. The obtained acid chloride was used without
3
further purification. To the acid chloride in THF (6.4 mL) Et N (0.13 mL, 0.96 mmol, 3 eq.) and
33,34
compound 4, which was synthesized as in the previously reported procedure
(0.12 g, 0.64
o
mmol, 2 eq.), were added at 0 C, and the mixture was then stirred at room temperature for 40 h.
After stirring, the solvent was removed in vacuo followed by flash chromatography
(
CHCl
3 f 3
/MeOH = 20:1) to give the title compound 5e (77 mg, 54% yield). R 0.5 (CHCl /MeOH
1
=
20:1), HꢀNMR (400 MHz, CDCl
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.33 (m, 12H), 1.39ꢀ1.65 (m,
ACS Paragon Plus Environment
24

Page 25 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
1
2H), 1.87 (s, 3H), 2.01ꢀ2.06 (m, 2H), 2.27 (s, 3H), 2.34ꢀ2.39 (m, 1H), 2.63ꢀ2.84 (m, 2H),
13
3.60ꢀ3.69 (m, 2H), 4.13ꢀ4.29 (m, 2H), 4.93ꢀ5.01 (m, 2H), 5.76ꢀ5.85 (m, 1H); CꢀNMR (125
MHz, CDCl ) δ 14.0, 19.9, 22.6, 24.5, 27.3, 27.4, 28.8, 28.9, 29.2, 29.3, 29.5, 31.8, 32.1, 32.2
2C), 33.4, 45.6 (2C), 64.9, 65.1, 81.0, 114.2, 118.8, 139.0, 151.7; HRMS (ESI), m/z calcd for
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
(
+
C
27 5
H46KO [M+K] 489.2977, found 489.2968.
(
2ꢀ(Hydroxymethyl)ꢀ5ꢀoxoꢀ4ꢀ(propanꢀ2ꢀylidene)tetrahydrofuranꢀ2ꢀyl)methyl
ꢀ(butꢀ3ꢀenꢀ1ꢀyl)decanoate (5a): By use of a procedure similar to that described for the
2
synthesis of compound 5e, through the removal of the tertꢀbutyl group from 3a (0.155 g, 0.68
mmol) and the subsequent formation of the acid chloride using oxalyl chloride (0.58 mL, 6.8
mmol) and a catalytic amount of DMF in DCM (5.7 mL), coupling with 4 (0.253 g, 1.36 mmol)
using Et
3
f
N (0.28 mL, 2.04 mmol) in THF (13.6 mL) gave compound 5a (0.094 g, 35%). R 0.5
1
(
CHCl
3
/MeOH = 20:1), HꢀNMR (400 MHz, CDCl
3
) δ 0.86ꢀ0.89 (m, 3H), 1.25ꢀ1.29 (m, 12H),
1
.44ꢀ1.77 (m, 4H), 1.87 (s, 3H), 2.00ꢀ2.14 (m, 2H), 2.26 (s, 3H), 2.43ꢀ2.63, (m, 1H), 2.73ꢀ2.94
1
3
(m, 2H), 3.60ꢀ3.74 (m, 2H), 4.15ꢀ4.29 (m, 2H), 4.97ꢀ5.06 (m, 2H), 5.70ꢀ5.83 (m, 1H); CꢀNMR
125 MHz, CDCl ) δ 14.0, 19.9, 22.6, 24.5, 27.3, 29.2, 29.3, 29.5, 31.1, 31.2, 31.5, 31.8, 32.0,
2.1, 44.8, 44.9, 64.9, 65.2, 80.9, 115.2, 118.8, 137.6; HRMS (ESI), m/z calcd for C23
(
3
3
39 5
H O
+
[
M+H] 395.2792, found 395.2794.
(
2ꢀ(Hydroxymethyl)ꢀ5ꢀoxoꢀ4ꢀ(propanꢀ2ꢀylidene)tetrahydrofuranꢀ2ꢀyl)methyl
2ꢀ(pentꢀ4ꢀenꢀ1ꢀyl)decanoate (5b): By use of a procedure similar to that described for the
synthesis of compound 5e, through the removal of the tertꢀbutyl group from 3b (0.119 g, 0.4
mmol) and the subsequent formation of the acid chloride using oxalyl chloride (0.34 mL, 4
mmol) and a catalytic amount of DMF in DCM (4 mL), coupling with 4 (0.150 g, 0.8 mmol)
using Et
CHCl
3
f
N (0.09 mL, 1.2 mmol) in THF (8 mL) gave compound 5b (0.085 g, 52%). R 0.5
1
(
3
/MeOH = 20:1), HꢀNMR (400 MHz, CDCl
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.31 (m, 12H),
ACS Paragon Plus Environment
25

Bioconjugate Chemistry
Page 26 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
.39ꢀ1.50 (m, 6H), 1.88 (s, 3H), 2.02ꢀ2.12 (m, 2H), 2.27 (s, 3H), 2.34ꢀ2.41 (m, 1H), 2.63ꢀ2.84
1
3
(m, 2H), 3.60ꢀ3.72 (m, 2H), 4.14ꢀ4.29 (m, 2H), 4.93ꢀ5.03 (m, 2H), 5.73ꢀ5.82 (m, 1H); CꢀNMR
(125 MHz, CDCl ) δ 14.0, 20.0, 22.6, 24.6, 25.1, 26.6, 27.4, 28.1, 29.2, 29.4, 29.5, 31.8, 32.1,
3.5, 35.6, 45.4, 64.9, 65.1, 80.9, 114.8, 138.2, 151.7, 168.8; HRMS (ESI), m/z calcd for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
+
C
24 5
H40KO [M+K] 447.2507, found 447.2507.
(
2ꢀ(Hydroxymethyl)ꢀ5ꢀoxoꢀ4ꢀ(propanꢀ2ꢀylidene)tetrahydrofuranꢀ2ꢀyl)methyl
ꢀ(hexꢀ5ꢀenꢀ1ꢀyl)decanoate (5c): By use of a procedure similar to that described for the
2
synthesis of compound 5e, through the removal of the tertꢀbutyl group from 3c (0.124 g, 0.4
mmol) and the subsequent formation of the acid chloride using oxalyl chloride (0.34 mL, 4
mmol) and a catalytic amount of DMF in DCM (4 mL), coupling with 4 (0.150 g, 0.8 mmol)
using Et
3
f
N (0.09 mL, 1.2 mmol) in THF (8 mL) gave compound 5c (0.069 g, 41%). R 0.5
1
(
CHCl
3
/MeOH = 20:1), HꢀNMR (400 MHz, CDCl
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.31 (m, 12H),
1
.35ꢀ1.61 (m, 8H), 1.88 (s, 3H), 2.01ꢀ2.12 (m, 2H), 2.27 (s, 3H), 2.32ꢀ2.39 (m, 1H), 2.63ꢀ2.84
1
3
(m, 2H), 3.60ꢀ3.71 (m, 2H), 4.13ꢀ4.30 (m, 2H), 4.92ꢀ5.02 (m, 2H), 5.74ꢀ5.84 (m, 1H); CꢀNMR
125 MHz, CDCl ) δ 14.0, 19.9, 22.6, 24.5, 25.1, 26.8, 27.3, 28.7, 29.2, 29.4, 29.5, 31.8, 32.1,
3.5, 45.5, 45.6, 64.8, 65.1, 81.0, 114.4, 118.8, 138.7, 151.6, 168.8; HRMS (ESI), m/z calcd for
(
3
3
+
C
25 5
H42KO [M+K] 461.2664, found 461.2659.
(
2ꢀ(Hydroxymethyl)ꢀ5ꢀoxoꢀ4ꢀ(propanꢀ2ꢀylidene)tetrahydrofuranꢀ2ꢀyl)methyl
2ꢀ(heptꢀ6ꢀenꢀ1ꢀyl)decanoate (5d): By use of a procedure similar to that described for the
synthesis of compound 5e, through the removal of the tertꢀbutyl group from 3d (0.103 g, 0.32
mmol) and the subsequent formation of the acid chloride using oxalyl chloride (0.27 mL, 3.2
mmol) and a catalytic amount of DMF in DCM (2.7 mL), coupling with 4 (0.120 g, 0.64 mmol)
using Et
CHCl
3
f
N (0.13 mL, 0.96 mmol) in THF (6.4 mL) gave compound 5d (0.082 g, 58%). R 0.5
1
(
3
/MeOH = 20:1), HꢀNMR (400 MHz, CDCl
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.31 (m, 12H),
ACS Paragon Plus Environment
26

Page 27 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
1
.35ꢀ1.64 (m, 10H), 1.87 (s, 3H), 2.00ꢀ2.05 (m, 2H), 2.27 (s, 3H), 2.32ꢀ2.39 (m, 1H), 2.63ꢀ2.84
1
3
(m, 2H), 3.60ꢀ3.70 (m, 2H), 4.13ꢀ4.29 (m, 2H), 4.92ꢀ5.01 (m, 2H), 5.75ꢀ5.84 (m, 1H); CꢀNMR
(125 MHz, CDCl ) δ 14.0, 19.9, 22.6, 24.5, 27.1, 27.2, 27.3, 27.4, 28.6, 28.9, 29.1, 29.4, 29.5,
1.8, 32.0, 32.1, 32.2, 33.6, 45.0, 45.6, 64.9, 65.1, 81.0, 114.3, 138.9; HRMS (ESI), m/z calcd
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
3
+
5
for C26H44NaO [M+Na] 459.3081, found 459.3071.
(
2ꢀ(Hydroxymethyl)ꢀ5ꢀoxoꢀ4ꢀ(propanꢀ2ꢀylidene)tetrahydrofuranꢀ2ꢀyl)methyl
2
ꢀoctylundecꢀ10ꢀenoate (5f): By use of a procedure similar to that described for the synthesis
of compound 5e, through the removal of the tertꢀbutyl group from 3f (0.110 g, 0.32 mmol) and
the subsequent formation of the acid chloride using oxalyl chloride (0.27 mL, 3.2 mmol) and a
3
catalytic amount of DMF in DCM (2.7 mL), coupling with 4 (0.12 g, 0.64 mmol) using Et N
(
f 3
0.13 mL, 0.96 mmol) in THF (6.4 mL) gave compound 5f (0.04 g, 27%). R 0.5 (CHCl /MeOH
1
=
20:1), HꢀNMR (400 MHz, CDCl
3
) δ 0.86ꢀ0.89 (m, 3H), 1.23ꢀ1.33 (m, 12H), 1.39ꢀ1.65 (m,
1
4H), 1.87 (s, 3H), 2.01ꢀ2.06 (m, 2H), 2.27 (s, 3H), 2.34ꢀ2.39 (m, 1H), 2.63ꢀ2.84 (m, 2H),
13
3.60ꢀ3.69 (m, 2H), 4.13ꢀ4.29 (m, 2H), 4.92ꢀ5.01 (m, 2H), 5.76ꢀ5.85 (m, 1H); CꢀNMR (125
MHz, CDCl ) δ 14.0, 14.1, 19.9, 21.0, 22.6, 24.5, 27.3, 28.8, 28.9, 29.2 (2C), 29.3, 29.4, 29.5,
1.8, 32.1, 32.2, 32.2, 33.7, 45.0, 45.6, 60.3 (2C), 64.9, 65.1, 81.0, 114.1, 139.1, 151.7; HRMS
3
3
+
(
ESI), m/z calcd for C28
5
H48KO [M+K] 503.3133, found 503.3126.
(
2ꢀ(Hydroxymethyl)ꢀ5ꢀoxoꢀ4ꢀ(propanꢀ2ꢀylidene)tetrahydrofuranꢀ2ꢀyl)methyl
2ꢀoctyldodecꢀ11ꢀenoate (5g): By use of a procedure similar to that described for the synthesis
of compound 5g, through the removal of the tertꢀbutyl group from 3g (0.147 g, 0.4 mmol) and
the subsequent formation of the acid chloride using oxalyl chloride (0.505 mL, 4 mmol) and a
catalytic amount of DMF in DCM (4 mL), coupling with 4 (0.15 g, 0.8 mmol) using Et
3
N (0.12
mL, 0.86 mmol) in THF (8 mL) gave compound 5f (0.067 g, 35%). R 0.5 (CHCl /MeOH =
f 3
1
2
0:1), HꢀNMR (400 MHz, CDCl
3
) δ 0.88ꢀ0.91 (m, 3H), 1.24ꢀ1.33 (m, 12H), 1.36ꢀ1.58 (m,
ACS Paragon Plus Environment
27

Bioconjugate Chemistry
Page 28 of 38
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
6H), 1.90 (s, 3H), 2.03ꢀ2.13 (m, 2H), 2.29 (s, 3H), 2.35ꢀ2.41 (m, 1H), 2.66ꢀ2.82 (m, 2H),
13
3.63ꢀ3.72 (m, 2H), 4.16ꢀ4.31 (m, 2H), 4.94ꢀ5.03 (m, 2H), 5.80ꢀ5.89 (m, 1H); CꢀNMR (125
MHz, CDCl ) δ 14.0, 19.9, 22.6, 24.5, 27.4 (2C), 28.8, 29.0, 29.2, 29.3, 29.4 (2C), 29.5 (2C),
1.8, 32.1 (2C), 33.7 45.6, 64.9, 65.1, 81.0, 114.1, 118.8, 139.2, 151.7, 168.8, 176.4; HRMS
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
+
(
ESI), m/z calcd for C29
H50NaO
5
[M+Na] 501.3550, found 501.3550.
6
c (Representative compound of 6aꢀ6g): To a solution of 5c (56 mg, 0.13 mmol) in CH Cl (3.3
2 2
mL) was added the second generation Grubbs catalyst (0.011 g, 0.013 mmol) at room
temperature, and the mixture was then stirred at that temperature for 24 h. After stirring, the
solvent was removed under reduced pressure followed by flash chromatography to give the title
compound 6c (37 mg, 70%).
+
LowꢀMS (ESI), 6a m/z calcd for C44
H
72NaO10 [M+Na] 783.5, found 783.5. 6b m/z calcd for
+
+
C
46
H
76NaO10 [M+Na] 811.5, found 811.6. 6c m/z calcd for C48
H
80NaO10 [M+Na] 839.6, found
+
839.6. 6d m/z calcd for C50
H
84NaO10 [M+Na] 867.6, found 867.6. 6e m/z calcd for
+
+
C
52
H
88NaO10 [M+Na] 895.6, found 895.6. 6f m/z calcd for C54
H
92NaO10 [M+Na] 923.7, found
+
9
23.7. 6g m/z calcd for C56H96NaO10 [M+Na] 951.7, found 951.7.
Yields of compounds 6aꢀ6g are described in the following table.
ACS Paragon Plus Environment
28

Page 29 of 38
Bioconjugate Chemistry
1
2
3
4
5
6
7
8
9
starting materials
products (yields)
6a (4.9 mg, 65%)
6b (30 mg, 64%)
6c (37 mg, 70%)
6d (9.8 mg, 14%)
6e (23 mg, 31%)
6f (15 mg, 45%)
6g (33 mg, 65%)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5a (8 mg, 0.02 mmol)
5b (51 mg, 0.12 mmol)
5c (56 mg, 0.13 mmol)
5d (76 mg, 0.17 mmol)
5e (76 mg, 0.17 mmol)
5
f (35 mg, 0.076 mmol)
g (54 mg, 0.11 mmol)
5
Assessment of Ki Values of the Synthesized Compounds. Binding affinities of the
synthesized compounds for the C1b domain of PKCδ, full length PKCδ, and full length PKCα were
3
3
assayed in vitro by competition for binding of [20ꢀ H]phorbol 12,13ꢀdibutyrate ([ H]PDBu)(13.5
Ci/mmol, from Perkin Elmer, Boston, MA, USA) in the presence of 100 µg/ml of
42
phosphatidylserine (PS) as described previously. Incubation was for 5 min at 37°C. Values
represent the mean ± SEM of triplicate binding experiments. In each experiment, competition was
determined for 7 concentrations of ligand, with the increasing concentrations of ligand spaced at
halfꢀlog intervals, and compared to binding in the absence of competing ligand. In each experiment,
triplicate measurements at each concentration of ligand were performed.
Translocation Analysis of EGFPꢀtagged PKCδ in CHO Cells. CHO cells were cultured in
Ham’s F12 medium supplemented with 10% fetal bovine serum and antibiotics (penicillin 50
units/mL and streptomycin 0.05 mg/mL). CHO cells plated on glass bottom dishes (Matsunami Glass
Ind., Ltd., Osaka, Japan) were transfected with EGFPꢀtagged PKCδ derived from the insertion of
PKCδ into pEGFPꢀN1 (ClontechꢀTakara Bio, Kusatsu, Japan) using Lipofectamine LTX (Invitrogen,
29
ACS Paragon Plus Environment