Amides as bioisosteres of triazole-based geranylgeranyl diphosphate synthase inhibitors

Article abstract of DOI:10.1016/j.bmc.2020.115604

Geranylgeranyl diphosphate synthase (GGDPS) inhibitors are of potential therapeutic interest as a consequence of their activity against the bone marrow cancer multiple myeloma. A series of bisphosphonates linked to an isoprenoid tail through an amide linkage has been prepared and tested for the ability to inhibit GGDPS in enzyme and cell-based assays. The amides were designed as analogues to triazole-based GGDPS inhibitors. Several of the new compounds show GGDPS inhibitory activity in both enzyme and cell assays, with potency dependent on chain length and olefin stereochemistry.

Full text of DOI:10.1016/j.bmc.2020.115604

Journal Pre-proofs
Amides as Bioisosteres of Triazole-Based Geranylgeranyl Diphosphate Syn‐
thase Inhibitors
Daniel B. Goetz, Michelle L. Varney, David F. Wiemer, Sarah A. Holstein
PII:
DOI:
Reference:
S0968-0896(20)30434-X
https://doi.org/10.1016/j.bmc.2020.115604
BMC 115604
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
23 January 2020
21 June 2020
22 June 2020
Please cite this article as: Goetz, D.B., Varney, M.L., Wiemer, D.F., Holstein, S.A., Amides as Bioisosteres of
Triazole-Based Geranylgeranyl Diphosphate Synthase Inhibitors, Bioorganic & Medicinal Chemistry (2020), doi:
https://doi.org/10.1016/j.bmc.2020.115604
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
©
2020 Elsevier Ltd. All rights reserved.

Amides as Bioisosteres of Triazole-Based Geranylgeranyl
Diphosphate Synthase Inhibitors
a
b
a,c
b*
Daniel B. Goetz, Michelle L. Varney, David F. Wiemer and Sarah A. Holstein
aDepartment of Chemistry, University of Iowa, Iowa City, IA 52242-1294, US
Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE 68198,
US
b
c
Department of Pharmacology, University of Iowa, Iowa City, IA 52242-1109, US
Corresponding author:
Sarah A. Holstein, MD, PhD
Division of Oncology and Hematology
Department of Internal Medicine
University of Nebraska Medical Center
Omaha, NE 68142
Academic Office Phone: 402-559-8500
Academic Office Fax: 402-559-6520
Email: sarah.holstein@unmc.edu
KEYWORDS. bisphosphonate, geranylgeranyl diphosphate synthase, inhibition, isoprenoid
biosynthesis, amide, triazole, bioisostere, myeloma
1

Abstract
Geranylgeranyl diphosphate synthase (GGDPS) inhibitors are of potential therapeutic interest as
a consequence of their activity against the bone marrow cancer multiple myeloma. A series of
bisphosphonates linked to an isoprenoid tail through an amide linkage has been prepared and
tested for the ability to inhibit GGDPS in enzyme and cell-based assays. The amides were
designed as analogues to triazole-based GGDPS inhibitors. Several of the new compounds show
GGDPS inhibitory activity in both enzyme and cell assays, with potency dependent on chain
length and olefin stereochemistry.
2

1
. Introduction
Multiple myeloma is a malignancy characterized by proliferation of clonal plasma cells
that produce monoclonal protein. As a result of this activity, these cells are inherently
susceptible to agents that disrupt protein homeostasis and trafficking pathways.1-2 Our efforts to
therapeutically exploit this potential vulnerability have focused on developing inhibitors of the
3
enzyme geranylgeranyl diphosphate synthase (GGDPS), a key component of the isoprenoid
biosynthesis pathway (IBP, Figure 1). Inhibition of this enzyme leads to a diminished pool of
geranylgeranyl diphosphate (GGDP), which is the isoprenoid donor used in protein
geranylgeranylation reactions. The Rab family of small GTPases require geranylgeranylation in
4
order to function as regulators of intracellular trafficking events. Mutated Rab proteins that are
not geranylgeranylated do not localize properly in the cell and are essentially non-functional.5
Our work has demonstrated that impairment of Rab geranylgeranylation, either as a result of
depletion of GGDP or via direct inhibition of the Rab geranylgeranyl transferase, leads to
6
disruption of intracellular monoclonal protein trafficking in myeloma cells. In turn, this
disruption in protein trafficking leads to induction of the unfolded protein response and
apoptosis.6-7 While the two classes of clinically used IBP inhibitors (statins (HMG-CoA
reductase inhibitors) and nitrogenous bisphosphonates (farnesyl diphosphate synthase (FDPS)
6
inhibitors)) disrupt Rab geranylgeranylation in vitro, these agents are not ideal from an anti-
8
-9
myeloma therapeutic perspective given their off-target effects at high doses (statins ) and
10
limited systemic biodistribution (nitrogenous bisphosphonates ). We therefore have focused on
the development of GGDPS inhibitors as a more clinically relevant means by which to
selectively disrupt protein geranylgeranylation in vivo.7, 11
3

HMG CoA
reductase
O
O
O
HO
HO
HO
SCoA
HO
OH
HMG-CoA
Mevalonic Acid
Statins
OPP
OPP
IPP isomerase
IPP
DMAPP
Nitrogenous
Bisphosphonates
Farnesyl Diphosphate
Synthase (FDPS)
Steroids
Farnesyl
Transferase
OPP
Prenylated Ras
proteins
Farnesyl Diphosphate
Terpenoid
Bisphosphonates
inc. these studies)
Geranylgeranyl Diphosphate
Synthase (GGDPS)
(
Geranylgeranyl
Transferase I
Prenylated Rho
proteins
(
GGTase I)
OPP
Geranylgeranyl
Transferase II
Geranylgeranyl Diphosphate
Prenylated Rab
proteins
(GGTase II)
Figure 1. The mammalian isoprenoid biosynthesis pathway (IBP) with key enzymes and
inhibitors.
During the course of our efforts to develop GGDPS inhibitors, we reported a series of di-
and monoalkyl isoprenoid bisphosphonates that selectively inhibit GGDPS over other enzymes
in the IBP.12-17 Subsequently, our efforts have focused on the development of triazole-based
GGDPS inhibitors,18-24 prepared using “click” chemistry to introduce the isoprenoid unit. One
benefit of the click strategy is that it allows reasonably efficient preparation of groups of
compounds from a common intermediate. Structure-activity studies with triazole
25
bisphosphonates have revealed that both chain length and olefin stereochemistry have significant
impact on inhibitor potency.3, 18-20, 22 For example, those studies established that while triazoles
4

bearing C monoterpenoid chains were more active than either the C or C analogues, the
1
0
5
15
homologated C terpenoids were more active than either normal-length terpenoids or
1
1
bishomologated terpenoids.18-20, 22 Given the predominance of trans isoprenoids in both protein
26
27
prenylation and plant metabolism, it was surprising that the cis-olefin isomers were
consistently more active than trans-olefin isomers.18, 20, 22 In fact, one of the most potent single
isomer triazole-based GGDPS inhibitors is compound (1) that contains a homoneryl substituent
2
0
i.e. a C substituent with a cis-olefin (IC of 75 nM in an enzyme assay with GGDPS).
1
1
50
Finally, the triazole moiety might interact favorably with the metal ions known to be held in the
enzyme active site15, 20 but this has not been established rigorously.
(
NaO) (O)P
2
N
(
NaO) (O)P
N N
2
1
One disadvantage of the click strategy is that when allylic azides are employed, they
undergo a facile [3,3] sigmatropic rearrangement that isomerizes the alkene stereochemistry and
2
8
affords a mixture of olefinic products. We have developed a synthetic sequence that
2
9
circumvents this issue through use of an epoxide as an olefin protecting group, although this
necessarily extends the synthetic sequence. The most direct synthesis of the
homogeranyl/homoneryl compounds is based on addition of cyclopropylmagnesium bromide to a
methyl ketone, and also generates a mixture of olefin isomers.19 Strategies to connect an
isoprenoid chain to a bisphosphonate head group that do not afford mixed olefin isomers would
be welcome, particularly if they could afford products that also might allow complexation with
metal ions. The triazole ring system shares many features with an amide functional group and
can be viewed as a bioisosteric replacement.30-31 From that perspective, it is reasonable to
question whether an amide could be employed in place of a triazole moiety and still preserve the
5

biological activity as a GGDPS inhibitor. To explore this concept, we have prepared a set of new
bisphosphonates linked to an isoprenoid chain through an amide bond. The synthesis of these
new compounds and their biological activity are the subjects of this report.
2
. Chemical Synthesis.
Amides of the general structure 2 can be seen to arise from the amine 3 and a variety of
carboxylic acids (4) via amide formation followed by hydrolysis of the resulting phosphonate
esters (Scheme 1), as long as the amide bond proves stable to the conditions of the standard
McKenna hydrolysis.32 The amine 3 already has been synthesized through a 3-step process33-35
from tetraethyl methylenebisphosphonate (5). The necessary carboxylic acids 4 can be obtained
from readily available terpenoid alcohols (6) through various means depending on the degree of
homologation. The normal-length carboxylic acids could be derived by oxidation of the alcohol
3
6
37
to the corresponding aldehydes followed by Pinnick oxidation to the carboxylic acids.
Homologated aldehydes are known intermediates in a synthetic route which was employed in our
synthesis of some homologated triazoles.21 Alternatively, these carboxylic acids can be seen as a
product of oxidation of the known homologated terpenoid alcohols, achievable through a
sequence including oxidation to the aldehyde, Wittig condensation with methylene
3
8
triphenylphosphorane, and hydroboration-oxidation, or by hydrolysis of geranyl and neryl
nitriles.39 Finally, the bishomologated acids 4h, 4i, and 4j are known to be available through a
2
2
malonic ester synthesis, again starting with an isoprenoid alcohol 6.
6

O
H
N
O
(
EtO)2P
(EtO)2P
H
O
H
N
(
EtO)2P
(EtO)2P
(
NaO)2P
R
O
3
O
O
5
(NaO)2P
O
O
2
a-j
HO
R
4
a-j
O
O
O
HO
HO
HO
HO
O
4
4
a
b
4
h
HO
HO
R1
Malonic ester
synthesis
4d
O
O
[O]
R2
HO
4
i
O
6
O
4
e
HO
HO
O
4j
4
c
HO
HO
O
4
f
4g
Scheme 1. Retrosynthesis of amide bisphosphonates.
Our initial synthetic efforts were focused on optimizing amide formation with the
commercially available carboxylic acid 4a, a mixed geranyl/neryl (E/Z) olefin isomer that
affords the amide 10a (Scheme 2). While there is some precedent for amide formation in similar
3
3
systems, in our hands the reaction of amine 3 with the commercial acid 4a in the presence of
N,N’-dicyclohexylcarbodiimide (DCC) afforded a complex mixture. Attention then turned to
amide formation from the acyl halide 9a. Both traditional Schotten-Baumann conditions and the
direct reaction of the acyl halide and amine 3 afforded the amide 10a in only moderate yield.
4
0
Significant improvements were found with a protocol from Zhang et al. in which the acyl
halide is added slowly to a cooled solution of the amine in the presence of an excess of
anhydrous potassium carbonate. Subsequent hydrolysis of the phosphonate esters in the presence
3
2
of trimethylsilyl bromide and collidine, followed by brief treatment with NaOH, afforded the
terpenoid amide bisphosphonate salt 2a.
7

1
2
) NaOH
) (COCl)2
pyr (cat)
O
O
HO
Cl
9
7%
4
a
9a
O
(
EtO) P
NH2
2
K CO
THF
60%
(
EtO) P
2
3
2
3
O
O
H
(
EtO) P
N
2
(
EtO) P
O
2
1
0a
O
1
2
) TMSBr, collidine
) NaOH
9
4%
O
H
N
(
NaO) P
2
(
NaO) P
O
2
O
2
a
Scheme 2. Synthesis of the amide bisphosphonate 2a.
Once acceptable conditions for formation of the terpenoid amides were in hand, a series
of analogues was prepared by parallel reactions (Scheme 3). Single isomers of both geranic acid
(4b) and nerolic acid (4c) were obtained in good to excellent yield by oxidation of the alcohols
3
6
6
b and 6c, first with TEMPO to afford the aldehydes 11b and 11c and then by Pinnick
oxidation to afford the respective carboxylic acids 4b and 4c.37 Commercial (S)-citronellol (6d)
was oxidized to (S)-citronellal (11d) under Swern conditions.41 Subsequent Pinnick oxidation of
both this (S)-citronellal (11d) and commercial (R)-citronellal (11e) afforded the citronellic acids
4
d and 4e respectively. Homologation of geranial and neral (aldehydes 11b and 11c) could be
accomplished by Wittig reaction with methylene triphenylphosphorane to afford the respective
3
8
olefins, which were then subjected to a known hydroboration-oxidation scheme with 9-BBN in
tetrahydrofuran.24, 42 The resulting alcohols (6f and 6g) were oxidized first with PCC to the
aldehydes, which were then immediately subjected to Pinnick conditions to yield the carboxylic
8

acids 4f and 4g. Finally, the bishomologated carboxylic acids 4h, 4i, and 4j were obtained
through a known malonic ester synthesis starting from alcohols geraniol (6b), nerol (6c), and
prenol (6j).22
X
HO
TEMPO
TEMPO
O
6
b
1
4
1b X = H
b
NaClO2
X = OH
HO
X
O
1
4
1c X = H
X = OH
6
6
c
d
NaClO2
NaClO2
c
(
COCl)2
HO
X
DMSO
O
1
4
1d X = H
X = OH
d
NaClO2
H
HO
O
O
1
1e
4e
1
2
) (C H ) PCH
6 5 3 2
O
) 9BBN
then H O2
1) PCC
2) NaClO2
2
HO
HO
1
1b
6
f
4f
1
2
) (C H ) PCH
2
O
6
5 3
) 9BBN
then H O2 HO
1) PCC
2) NaClO2 HO
2
1
1c
6
g
4g
HO
HO
HO
O
O
O
4
h
4i
4j
Scheme 3. Synthesis of the carboxylic acids 4b-j.
Subsequent conversion of the carboxylic acids 4 to their respective amides 10 through the
acid chlorides 9 proceeded in good to fair yield (Table 1). Hydrolysis of the bisphosphonate
esters upon treatment with trimethylsilyl bromide and collidine in dichloromethane afforded the
9

terpenoid amide bisphosphonate salts 2 in moderate to good yield, with the exception of the
homogeranyl amide 10f. In all attempts to hydrolyze this specific amide, either incomplete
4
3
hydrolysis or decomposition was observed via NMR or ESIMS analysis. The other amide
bisphosphonates were then subjected to a variety of bioassays.
1
2
) NaOH
) (COCl)2
pyr (cat)
O
O
HO
R
Cl
9a-j
R
4
a-j
3
, K CO (xs)
2 3
THF (0.025 M)
O
H
N
(
RO) P
R
2
(
RO) P
O
2
O
1
2
) TMSBr
collidine
) NaOH
1
2
0a-j R = Et
a-j R = Na
RCO H
9
10
2
2
4
4
4
4
4
a
b
c
d
e
97%
90%
66%
62%
60%
85%
62%
71%
86%
82%
60%
48%
35%
61%
58%
37%
50%
39%
70%
70%
94%
87%
79%
18%
53%
0%
87%
20%
54%
99%
4
f
4
4
g
h
4
4
i
j
Table 1. Yields for conversion of carboxylic acids 4 to amide bisphosphonates 2.
3
. Biological Results
All nine novel amide bisphosphonates were subjected to enzymatic (GGDPS and FDPS)
as well as cellular assays (immunoblot analysis and ELISA) in order to determine the potency
1
0

and selectivity of these agents as GGDPS inhibitors. Immunoblot analysis of H-Ras (a
representative farnesylated protein) and Rap1a (a representative substrate of GGTase I) enables
detection of impaired farnesylation and geranylgeranylation while accumulation of intracellular
light chain (as measured by ELISA) is a marker for disrupted Rab geranylgeranylation.6
Lovastatin, an HMG-CoA reductase inhibitor, was included as a positive control, as it globally
6
disrupts prenylation. Table 2 summarizes the results of these assays.
Table 2. Summary of the bioassay results for the novel bisphosphonate amides.
Compound
GGDPS IC50 FDPS IC (M)
Fold-selectivity for
GGDPS compared to
LEC
(M)
50
1
(M)
FDPS
445
11.7
4.5
0.8
5.1
14.3
2.1
17.9
28.8
3.1
120
0.075 + 0.013
2.4 + 1.2
4.7 + 1.2
37.7 + 3.2
5.4 + 1.5
3.4 + 0.8
3.7 + 0.6
0.8 + 0.3
1.1 + 0.5
14.4 + 4.7
33.3 + 5.6
28.1 + 5.9
21.1 + 3.1
29.9 + 6.6
27.8 + 3.7
48.5 + 14.8
7.8 + 1.9
14.3 + 3.7
31.7 + 14.6
44.4 + 16.8
0.03
10
5
50
50
5
5
1
0.5
50
2
2
2
2
2
2
2
2
2
a
b
c
d
e
g
h
i
j
1
Cellular LEC (lowest effective concentration) is defined as the lowest concentration for which an
unmodified Rap1a band is visible in the immunoblot and a statistically significant increase in intracellular
lambda light chain is observed in the ELISA.
These studies demonstrated the impact of both chain length and stereochemistry on
GGDPS inhibitor potency. As shown in Figure 2, the mixture of the geranyl and neryl amides
(2a) was approximately 5-fold more potent than the bishomoprenyl compound (2j) with respect
to disruption of Rap1a geranylgeranylation and accumulation of intracellular light chains. Those
1
1

Figure 2. Cellular activity of compounds 2j and 2a are consistent with GGDPS inhibition.
RPMI-8226 myeloma cells were incubated for 48 hours in the presence or absence of lovastatin
(“Lov”) or varying concentrations of compounds 2j or 2a. A) Immunoblot analysis of H-Ras,
Rap1a (antibody detects only unmodified protein) and -tubulin (as a loading control) was
performed. B) Intracellular lambda light chain concentrations were determined via ELISA. Data
are expressed as percentage of control (mean + SD, n=3). The * denotes p<0.05 per t-test and
compares treated cells to untreated control cells.
results, in conjunction with the lack of effect of the compounds on H-Ras prenylation, and the
results from the enzymatic studies, confirmed the specificity of these compounds as GGDPS
inhibitors. The preparation of the pure geranyl (2b) and neryl (2c) isomers allowed for evaluation
1
2

Figure 3. Cellular activity of novel bisphosphonate amides. RPMI-8226 cells were incubated for
8 hours in the presence or absence of lovastatin (“Lov”) or varying concentrations of
4
compounds 2b and 2c (A), 2g (B), 2h and 2i (C) or 2e and 2d (D). Immunoblot analysis of H-
Ras, Rap1a (antibody detects only unmodified protein) and -tubulin (as a loading control) was
performed.
of the effects of olefin stereochemistry on potency. As shown in Figure 3A, the geranyl isomer
was approximately 10-fold more potent than the neryl isomer with respect to inhibition of Rap1a
geranylgeranylation. ELISA results confirmed potency as determined by the Rap1a immunoblots
(data not shown). Interestingly, however, the geranyl isomer also disrupted H-Ras farnesylation.
These results conflicted with the enzyme assay results, in which the geranyl isomer was found to
1
3

be a more selective inhibitor against GGDPS compared to FDPS, while the neryl isomer
inhibited both enzymes with similar potency. To explore these findings further, FTase and
GGTase I enzyme assays were performed. The geranyl isomer 2b was found to be a weak FTase
inhibitor (IC 47.5 + 8.6 M) and to have no inhibitory activity against GGTase I. In contrast,
5
0
the neryl isomer 2c, displayed less than 30% inhibitory activity at 100 M against FTase.
Extension of the chain by one carbon to yield the homoneryl amide (2g) resulted in a 10-
fold increase in GGDPS inhibitor activity with respect to the neryl compound (2c), as shown in
Figure 3B. Extension to the bishomologated length further enhanced cellular potency (Figure
3
C). The most potent inhibitor of the series was revealed to be bishomoneryl compound 2i which
was slightly more potent than the bishomogeranyl 2h in the cellular assays, although equipotent
in the enzyme assays. Finally, evaluation of the citronellyl isomers demonstrated that the R-
isomer 2e was 10-fold more potent than the corresponding S-isomer 2d in cellular assays (Figure
3
D). None of these compounds (2g, 2h, 2i, 2e, 2d) disrupted H-Ras farnesylation and no
inhibitory activity was observed in the FTase and GGTase I assays at concentrations as high as
00 M (data not shown).
1
4
. Discussion
Here we present a series of novel isoprenoid amide bisphosphonate-based GGDPS
inhibitors. These amide-based inhibitors were designed as bioisosteres of some of our previously
reported triazole-based inhibitors.18-19, 21 As shown in Figure 4, we hypothesized that the activity
of the bishomoneryl amide 2i would be similar to that of the homoneryl triazole 1, as the two
compounds have a high degree of overlap using three-dimensional rendering in which the
4
4
molecules were minimized with a universal force field. As shown in Figure 5, when drawn
1
4

Figure 4. Three dimensional rendering of the amide 2i and the triazole 1, as well as an overlay
of the two compounds. Both structures were generated in an energy-minimized conformation
4
4
45
using Avogadro and then displayed using Chimera.
in an extended conformation, compounds 2h and 2i place the first olefin at nearly the same
distance from the bisphosphonate head group. In contrast, the two amides with shorter chains (2b
and 2c) place this olefin significantly closer to the polar head group than found in the triazole 1.
1
5

Figure 5. Impact of olefin isomer and chain length on potency of amide bisphosphonate
inhibitors of GGDPS compared to the homoneryl triazole bisphosphonate. The IC values from
5
0
the in vitro enzymatic assay are shown.
Both of these amides have higher IC values, with the Z-olefin 2c ~4-fold less potent than the Z-
5
0
olefin 2i. Placement of the terminal olefin in compound 2i also approximates the position of this
functionality in the triazole 1 more closely than any other amide. Thus, while it was perhaps not
1
6

surprising that the most potent amide in this series did turn out to be compound 2i (LEC of 0.5
M), we note that it is approximately 17-times less potent than the corresponding triazole 1.19

Although there is a marked similarity in the orientation of the majority of the atoms within
compounds 1 and 2i (Figure 5), it is apparent that the supposedly subtle differences induced by
the amide moiety in comparison to the triazole moiety, must be sufficient to lead to more
significant differences in interaction with key components of the enzyme’s binding sites.
Our previous structure-function work with the triazole-based GGDPS inhibitors
consistently demonstrated that the cis-olefin isomers are more potent than the trans-olefin
isomers, regardless of chain length.18, 20, 22 In the present series, while differences in potency
between the geranyl/neryl (2b/2c) and bishomogeranyl/bishomoneryl (2h/2i) analogues were
observed, there was not a consistent pattern, suggesting that the amides interact with the
isoprenoid binding sites in a manner that is disparate from the triazoles. Computational modeling
studies with 1 and its corresponding trans-olefin isomer indicated that while both isomers can
occupy either the farnesyl pyrophosphate (FDP) substrate site or the GGDP product site, the
homoneryl isomer preferentially occupies the FDP site while the homogeranyl isomer
2
0
preferentially occupies the GGDP site. Whether the cis- and trans-olefin isomers of the
bisphosphonate amides have differences in affinity to the FDP and GGDP sites is not known and
future modeling studies and co-crystallization studies would be necessary to further understand
the interaction/s between these novel inhibitors and the substrate/product binding sites of
GGDPS. Furthermore, whether the potency of the bisphosphonate amides can be enhanced by
methylation at the alpha-carbon position, as was demonstrated with the corresponding triazole
2
3
bisphosphonates, or by other modifications at the alpha-carbon position or distal to the amide
moiety remains to be determined.
1
7

In conclusion, these studies represent the first description of amide-based isoprenoid
bisphosphonate GGDPS inhibitors. While thus far, these inhibitors have been found to be less
potent than their triazole counterparts, these compounds may still be useful if they are found to
have more favorable biodistribution properties in vivo. Our prior preclinical studies with key
triazole bisphosphonate GGDPS inhibitors have revealed a complex structure-function
relationship with respect to olefin stereochemistry, presence of an alpha-carbon substituent,
pharmacokinetic parameters and biodistribution profiles,7, 11 and it is unknown whether
replacement of the triazole moiety with the amide will further impact these parameters.
Experimental Procedures and methods
5
.1. General Experimental Procedures. Acetonitrile was distilled from calcium hydride prior
to use. All other reagents and solvents were purchased from commercial sources and used
without further purification. All reactions in non–aqueous solvents were conducted in oven–
dried glassware under a positive pressure of argon or nitrogen and with magnetic stirring. For
TLC analyses, pre-coated silica polyester- (200 μm thickness, UV254 indicator) or glass-backed
plates were used. Flash column chromatography was carried out using silica gel (60 Å, 40–63
μm). The purity of the final compounds was corroborated by HPLC analysis using an Agilent
1
120 infinity LC solvent delivery system with a variable wavelength UV detector, and
compounds for bioassay were >95% pure at 254 nm. All NMR spectra were obtained at either
1
13
31
3
00 or 400 MHz for H, 75 or 100 MHz for C, and 121 or 162 MHz for P with internal
1
1
13
1
standards of (CH ) Si ( H, 0.00 ppm) or CDCl ( H, 7.27; C, 77.2 ppm) or CD OD ( H, 3.31;
3
4
3
3
1
3
1
13
1
13
C, 49.0 ppm) or CD C(O)CD ( H, 2.05; C, 206.3 ppm) or CD CN ( H, 1.94; C, 118.3 ppm)
3
3
3
1
31
for non–aqueous samples or D O ( H, 4.80 ppm) for aqueous samples. The P chemical shifts
2
1
8

are reported in ppm relative to 85% H PO (external standard). High-resolution mass spectra
3
4
were obtained by TOF MS ES+ at the University of Iowa Mass Spectrometry Facility.
3
7
5
.2. (3S)‐3,7‐Dimethyloct‐6‐enoic acid (4d). In a procedure modified from Kim, et al. , (S)-
4
1
citronellal 11d (0.93 g, 6 mmol) was dissolved in a mixture of tert-butyl alcohol (30 mL) and
-methyl-2-butene (6.4 mL, 60 mmol, 10 eq). The resulting solution was stirred and cooled to 0
C. A mixture of NaClO (80% mixture with NaCl, 0.95 g, 8.4 mmol, 1.4 eq) and NaH PO
2
°
2
2
4
(0.72 g, 6 mmol, 1 eq) was dissolved in a minimum volume of water and then added to the
reaction. The mixture was stirred and monitored via TLC until complete consumption of the
aldehyde was observed. The reaction was diluted with ethyl acetate and brine and the layers were
separated. The aqueous layer was acidified with NaHSO (aq) to pH 3 and extracted three times
4
with ethyl acetate. Additional acid was added between extractions as necessary to maintain pH 3.
The combined organic layers were dried (Na SO ) and concentrated under reduced pressure.
2
4
Flash chromatography of the residual oil afforded carboxylic acid 4d (5.6 g, 6 mmol, >99%
1
46
yield) with H NMR data consistent with that in the literature.
.3. (3R)‐3,7‐dimethyloct‐6‐enoic acid (4e). Following the general procedure for carboxylic
5
acid 4d, Pinnick oxidation of commercial (R)-citronellal (11e, 1.12 mL, 5.6 mmol) afforded
1
carboxylic acid 4e (0.95 g, 5.6 mmol, >99% yield) with H NMR data consistent with that in the
literature.46
2
1
5
.4. (3Z)‐4,8‐Dimethylnona‐3,7‐dienoic acid (4g). To a solution of alcohol 6g (1.18 g, 7
mmol) in dichloromethane (70 mL) was added flame-dried celite (11 g) and pyridinium
dichromate (2.26 g, 10.5 mmol, 1.5 eq). The reaction was stirred at room temperature until
complete consumption of the alcohol was observed by TLC. The reaction was filtered through a
bed of Florisil and the solids were washed 3 times with ether. The combined filtrate was
1
9

concentrated under reduced pressure to afford crude homoneral 11g which was used without
further purification. Following the general procedure for (S)-citronellic acid 4d, homoneral 11g
1
13
was oxidized to the title compound (1.12 g, 6.1 mmol, 88%). The H and C NMR data matched
that in the literature.47
5
.5 Diethyl
[1‐(diethoxyphosphoryl)‐3‐[3,7‐dimethylocta‐2,6‐dienamido]propyl]phosphonate (10a). In
4
0
a procedure modified from Zhang, et al. commercial carboxylic acid 4a (1 mL, 5 mmol) was
stirred with a 3.8 M aqueous solution of sodium hydroxide (1.4 mL, 5.25 mmol, 1.05
equivalents) for 30 minutes, after which water was removed on a lyophilizer. The resulting
sodium carboxylate was added to a round-bottom flask. Toluene was added and then removed
under reduced pressure three times, and the solids were stirred under vacuum for hours.
Anhydrous toluene was added (25 mL) followed by pyridine (0.06 mL, 0.75 mmol, 0.15 eq). The
resulting mixture was cooled to 0 °C and a 1 M solution of oxalyl chloride in anhydrous toluene
(1.00 mL, 11.5 mmol, 2.3 eq) was added dropwise. The reaction was allowed to stir for 45
minutes, the solids were removed via filtration, and the filtrate was concentrated under reduced
pressure to afford the acyl chloride 9a (866 mg, 4.6 mmol, 93%) which was used immediately
without further purification.
To a round-bottomed flask fitted with a magnetic stirbar was added amine 333-35 (930 mg,
2
.8 mmol), THF (280 mL), and anhydrous potassium carbonate (3.1 g, 22.4 mmol, 8 eq). After
the reaction mixture was stirred and cooled to -10 °C, the acyl halide 9a (836 mg, 4.5 mmol, 1.6
eq) was added dropwise over 2 hours as a 0.25 M solution in THF. The reaction was allowed to
stir with cooling and monitored via TLC until the amine was consumed. The mixture was
filtered, the solids were washed with ethyl acetate, and the filtrate was concentrated under
2
0

reduced pressure. The resulting residue was dissolved in ethyl acetate and extracted with water.
The aqueous layer was extracted three times with ethyl acetate, after which the combined organic
layers were dried (MgSO ) and filtered. The filtrate was concentrated under reduced pressure and
4
the residue was purified by flash chromatography to afford the desired amide 10a as a yellow oil
1
(
785 mg, 1.68 mmol, 60% yield): H NMR (400 MHz, CDCl ) δ 6.52 (t, J = 5.8 Hz, 1H minor),
3
6
.45 (t, J = 5.3 Hz, 1H major), 5.58 (s, 1H minor), 5.56 (s, 1H major), 5.16 (t, J = 7.1 Hz, 1H
minor), 5.07 (t, J = 6.7 Hz, 1H major), 4.18 (quint d, J = 7.3, 3.4 Hz, 8H), 3.48 (m, 2H), 2.39 (m,
H), 2.13 (m, 9H), 1.83 (s, 3H minor), 1.68 (s, 3H major), 1.62 (s, 3H minor), 1.61 (s, 3H
1
1
3
major), 1.57 (s, 3H minor), 1.35 (t, J = 7.1 Hz, 12H); C (100.6 MHz, CDCl ) δ 167.8, 166.8,
3
1
53.9, 153.8, 132.0, 131.6, 124.0, 123.2, 118.8, 118.1, 62.7 (t, J = 6.4 Hz, 4C), 40.7, 38.1 (t,
PC
J = 7.0 Hz), 34.3 (t, J = 133.7 Hz), 32.9, 32.3 (?), 26.7, 26.1, 25.6, 25.1 (t, J = 4.7 Hz),
PC
PC
PC
3
1
2
4.7, 18.0, 17.5, 16.2 (d, J = 6.2 Hz, 4C); P (162 MHz, CDCl ) δ 23.5. HRMS (ESI) m/z
PC 3
+
calculated for C H NO P [M + H] 482.2437, found 482.2440
2
1
42
7 2
5
.6. Diethyl
[1‐(diethoxyphosphoryl)‐3‐[(2E)‐3,7‐dimethylocta‐2,6‐dienamido]propyl]phosphonate
(10b). Following the general procedure for formation of acyl chloride 9a, the acid chloride 9b
was prepared from geranic acid 4b36-37 (673 mg, 4 mmol) as a yellow oil (670 mg, 3.6 mmol,
0% yield). Subsequently, following the general procedure for formation of amide 10a, the title
9
compound was obtained from the acid chloride 9b (390 mg, 2.1 mmol) and bisphosphonate
1
amine 3 (500 mg, 1.5 mmol) as a yellow oil (340 mg, 0.70 mmol, 48%): H NMR (400 MHz,
CDCl ) δ 6.46 (t, J = 4.8 Hz, 1H), 5.55 (s, 1H), 5.06 (t, J = 6.0 Hz, 1H), 4.17 (m, 8H), 3.49 (q, J
3
=
5.8 Hz, 2H), 2.39 (tt, JPH = 24.3 Hz, J = 5.9 Hz, 1H), 2.2-2.0 (m, 9H), 1.67 (s, 3H), 1.59 (s,
1
3
3
H), 1.34 (t, J = 7.0 Hz, 12H); C (100.6 MHz, CDCl ) δ 167.4, 154.2, 132.4, 123.4, 118.2,
3
2
1

6
2.97 (d, J = 6.9 Hz, 2C), 62.89 (d, J = 6.8 Hz, 2C), 40.9, 38.2 (t, J = 6.9 Hz), 34.6 (t, J
PC PC PC PC
3
1
=
133.9 Hz), 26.2, 25.7, 25.1 (t, J = 4.8 Hz), 18.2, 17.7, 16.4 (d, J = 6.2 Hz, 4C); P (162
PC PC
+
MHz, CDCl ) δ 23.6. HRMS (ESI) m/z calculated for C H NO P [M + H] 482.2437, found
3
21 42
7 2
4
82.2439
5
.7. Diethyl
[1‐(diethoxyphosphoryl)‐3‐[(2Z)‐3,7‐dimethylocta‐2,6‐dienamido]propyl]phosphonate
(10c). Following the general procedure for formation of acyl chloride 9a, the acid chloride 9c
was prepared from nerolic acid 4c36-37 (673 mg, 4 mmol) as a yellow oil (495 mg, 2.65 mmol,
6%). Following the general procedure for formation of amide 10a, the title compound was
6
obtained from the acid chloride 9c (420 mg, 2.25 mmol) and bisphosphonate amine 3 (500 mg,
1
1
1
7
.5 mmol) as a yellow oil (251 mg, 0.52 mmol, 35%): H NMR (400 MHz, CDCl ) δ 6.45 (br,
3
H), 5.54 (s, 1H), 5.14 (t, J = 6.7 Hz, 1H), 4.15 (bm, 8H), 3.46 (q, J = 4.8 Hz, 2H), 2.61 (t, J =
.7 Hz, 2H), 2.36 (tt, JPH = 24.2, J = 5.6 Hz, 1H), 2.14 (m, 4H), 1.80 (s, 3H), 1.65 (s, 3H), 1.60
1
3
(
s, 3H), 1.32 (t, J = 6.8 Hz, 12H); C (100.6 MHz, CDCl ) δ 166.9, 154.3, 132.0, 124.2, 118.9,
3
6
2.9 (t, J = 5.7 Hz, 4C), 38.3 (t, J = 6.6 Hz), 34.6 (t, J = 143.4 Hz), 33.2, 27.0, 25.8, 25.2
PC PC PC
3
1
(
t, J = 4.4 Hz), 25.0, 17.8, 16.5 (d, J = 5.9 Hz, 2C); P (121.5 MHz, CDCl ) δ 23.6. HRMS
P
C
3
+
(
ESI) m/z calculated for C H NO P [M + H] 482.2437, found 482.2439.
21 42 7 2
5
.8. Diethyl
[1‐(diethoxyphosphoryl)‐3‐[(3S)‐3,7‐dimethyloct‐6‐enamido]propyl]phosphonate (10d).
Following the general procedure for formation of acyl chloride 9a, the acid chloride 9d was
prepared from (S)-citronellic acid 4d (400 mg, 2.35 mmol) as a yellow oil (275 mg, 1.5 mmol,
6
2%). Following the general procedure for formation of amide 10a, the title compound was
obtained from the acid chloride 9d (226 mg, 1.2 mmol) and bisphosphonate amine 3 (248 mg,
2
2

1
0
.75 mmol) as a yellow oil (221 mg, 0.46 mmol, 61%): H NMR (300 MHz, CDCl ) δ 6.55 (t, J
3
=
4.6 Hz, 1H), 5.08 (tm, J = 7.3, 1H), 4.17 (d quint, J = 7.2, 2.9 Hz, 8H), 3.46 (q, J = 6.0 Hz,
2
H), 2.9-1.8 (m, 9H), 1.66 (s, 3H), 1.58 (s, 3H), 1.34 (t, J = 7.1 Hz, 12H), 1.16 (m, 1H) 0.91 (d, J
1
3
=
6.3 Hz, 3H); C (100.6 MHz, CDCl ) δ 172.8, 131.3, 124.4, 62.8 (t, J = 6.2 Hz, 4C), 44.5,
3 PC
3
1
8.6 (t, J = 7.2 Hz), 37.0, 34.7 (t, J = 133.9 Hz), 30.4, 25.7, 25.5, 25.1 (t, J = 4.6 Hz), 19.5,
PC PC PC
3
1
7.7, 16.4 (d, J = 5.3 Hz, 4C); P (121.5 MHz, CDCl ) δ 23.5. HRMS (ESI) m/z calculated
PC
3
+
for C H NO P [M + H] 484.2593, found 484.2594
2
1
44
7 2
5
.9. Diethyl
[1‐(diethoxyphosphoryl)‐3‐[(3R)‐3,7‐dimethyloct‐6‐enamido]propyl]phosphonate (10e).
Following the general procedure for formation of acyl chloride 9a, the acid chloride 9e was
prepared from (R)-citronellic acid 4e (460 mg, 2.7 mmol) as a yellow oil (304 mg, 1.6 mmol,
6
0%). Following the general procedure for formation of amide 10a, the title compound was
obtained from the acid chloride 9e (304 mg, 1.6 mmol) and bisphosphonate amine 3 (381 mg,
1
31
1
.15 mmol) as a yellow oil (320 mg, 0.66 mmol, 58%). Both H and P NMR data matched that
given above for the (S)-enantiomer 10d.
.10. Diethyl
5
[1‐(diethoxyphosphoryl)‐3‐[(3E)‐4,8‐dimethylnona‐3,7‐dienamido]propyl]phosphonate
(10f). Following the general procedure for formation of acyl chloride 9a, the acid chloride 9f was
prepared from homogeranic acid 4f (54 mg, 0.3 mmol) as a yellow oil (51 mg, 0.25 mmol, 85%).
Following the general procedure for formation of amide 10a, the title compound was obtained
from the acid chloride 9f (51 mg, 0.25 mmol) and bisphosphonate amine 3 (53 mg, 0.16 mmol)
1
as a yellow oil (29 mg, 0.06 mmol, 37%): H NMR (400 MHz, CDCl ) δ 6.42 (t, J = 5.2 Hz, 1H),
3
5
.30 (t, J = 7.3 Hz, 1H), 5.07 (t, J = 6.3, 1H), 4.16 (quint d, J = 7.2, 4.1 Hz, 8H), 3.44 (q, J = 6.3
2
3

Hz, 2H), 2.95 (d, J = 7.4 Hz, 2H), 2.32 (tt, JPH = 24.0 Hz, J = 6.2 Hz, 1H), 2.2-2.0 (m, 6H), 1.68
1
3
(
s, 3H), 1.63 (s, 3H), 1.60 (s, 3H), 1.34 (t, J = 7.0 Hz, 12H); C (100.6 MHz, CDCl ) δ 171.9,
3
1
40.9, 132.0, 124.0, 116.8, 62.97 (d, J = 6.7 Hz, 2C), 62.92 (d, J = 7.1 Hz, 2C), 39.8, 38.7 (t,
PC PC
J = 7.4 Hz), 36.2, 34.6 (t, J = 133.8), 26.6, 25.9, 25.4 (t, J = 4.9), 17.9, 16.6 (d, J = 6.6
PC
PC
PC
PC
3
1
Hz, 4C), 16.4; P (162.0 MHz, CDCl ) δ 23.3. HRMS (ESI) m/z calculated for C H NO P Na
3
22 43
7 2
+
[
M + Na] 518.2412, found 518.2409.
.11. Diethyl
1‐(diethoxyphosphoryl)‐3‐[(3Z)‐4,8‐dimethylnona‐3,7‐dienamido]propyl]phosphonate
10g). Following the general procedure for formation of acyl chloride 9a, the acid chloride 9g
5
[
(
was prepared from homonerolic acid 4g (365 mg, 2 mmol) as a yellow oil (250 mg, 1.25 mmol,
2%). Following the general procedure for formation of amide 10a, the title compound was
6
obtained from the acid chloride 9g (250 mg, 1.3 mmol) and bisphosphonate amine 3 (265 mg,
1
0
5
.8 mmol) as a yellow oil (200 mg, 0.4 mmol, 50%): H NMR (400 MHz, CDCl ) δ 6.40 (t, J =
3
.0 Hz, 1H), 5.30 (t, J = 7.0 Hz, 1H), 5.08 (m, 1H), 4.17 (m, 8H), 3.45 (q, J = 6.4 Hz, 2H), 2.95
(d, J = 6.2 Hz, 2H), 2.33 (tt, JPH = 24.0 Hz, J = 6.2 Hz, 1H), 2.2-2.0 (m, 6H), 1.76 (s, 3H), 1.67
1
3
(
s, 3H), 1.60 (s, 3H), 1.35 (t, J = 7.0 Hz, 12H); C (75.5 MHz, CDCl ) δ 172.0, 140.8, 132.2,
3
1
3
23.8, 117.4, 62.9 (t, J = 5.8 Hz, 4C), 38.6 (t, J = 7.3 Hz), 35.9, 34.6 (t, J = 133.7 Hz),
PC PC PC
3
1
2.1, 26.5, 25.8, 25.4 (m), 23.6, 17.8, 16.5 (d, J = 6.1 Hz, 4C); P (162 MHz, CDCl ) δ 23.3.
PC
3
+
HRMS (ESI) m/z calculated for C H NO P [M + H] 496.2593, found 496.2599.
2
2
44
7 2
5
.12. Diethyl
[1‐(diethoxyphosphoryl)‐3‐[(4E)‐5,9‐dimethyldeca‐4,8‐dienamido]propyl]phosphonate
(10h). Following the general procedure for formation of acyl chloride 9a, the acid chloride 9h
2
2
was prepared from bishomogeranic acid 4h (589 mg, 3 mmol) as a yellow oil (457 mg, 2.1
2
4

mmol, 71%). Following the general procedure for formation of amide 10a, the title compound
was obtained from the acid chloride 9h (270 mg, 1.26 mmol) and bisphosphonate amine 3 (300
1
mg, 0.9 mmol) as a yellow oil (180 mg, 0.312 mmol, 39%): H NMR (300 MHz, CDCl ) δ 6.52
3
(t, J = 5.0 Hz, 1H), 5.09 (q, J = 5.3 Hz, 2H), 4.18 (d quint, J = 7.3, 3.1 Hz, 8H), 3.46 (q, J = 6.1
1
3
Hz, 2H), 2.5-1.9 (m, 11 H), 1.67 (s, 3H), 1.61 (s, 3H), 1.60 (s, 3H), 1.34 (t, J = 7.1 Hz, 12H); C
100.6 MHz, CDCl ) δ 173.1, 136.5, 131.4, 124.2, 122.7, 62.9 (t, J = 6.3 Hz, 4C), 39.7, 38.6 (t,
(
3
PC
J = 7.0), 36.8, 34.6 (t, J = 133.8), 26.7, 25.7, 25.1 (t, J = 4.7 Hz), 24.3, 17.7, 16.4 (d, J =
PC
PC
PC
PC
3
1
6
.2 Hz, 4C), 16.0; P (121.5 MHz, CDCl ) δ 23.5. HRMS (ESI) m/z calculated for C H NO P
3 23 46 7 2
+
[
M + H] 510.2750, found 510.2749.
.13. Diethyl
1‐(diethoxyphosphoryl)‐3‐[(4Z)‐5,9‐dimethyldeca‐4,8‐dienamido]propyl]phosphonate
10i). Following the general procedure for formation of acyl chloride 9a, the acid chloride 9i was
5
[
(
2
2
prepared from bishomonerolic acid 4i (589 mg, 3 mmol) as a yellow oil (553 mg, 2.6 mmol,
8
6%). Following the general procedure for formation of amide 10a, the title compound was
obtained from the acid chloride 9i (346 mg, 1.61 mmol) and bisphosphonate amine 3 (381 mg,
1
1
.15 mmol) as a yellow oil (400 mg, 0.81 mmol, 70%): H NMR (300 MHz, CDCl ) δ 6.54 (t, J
3
=
5.1 Hz, 1H), 5.06 (t, J = 6.9 Hz, 1H), 5.06 (t, J = 6.9 Hz, 1H), 4.14 (d quint, J = 7.3, 3.1 Hz,
8
H), 3.41 (q, J = 6.2 Hz, 2H), 2.4-2.0 (m, 11H), 1.64 (s, 6H), 1.57 (s, 3H), 1.31 (t, J = 7.1 Hz, 12
1
3
H); C (75.5 MHz, CDCl ) δ 173.0, 136.7, 131.7, 124.3, 123.5, 62.9 (br, 4C), 38.6 (t, J = 6.7
3
PC
Hz), 37.0, 34.7 (t, J = 134 Hz), 32.0, 26.7, 25.8, 25.2, 24.1, 23.5, 17.7, 16.5 (d, J = 5.9 Hz,
PC
PC
3
1
+
4
5
C); P (121.5 MHz, CDCl ) δ 23.4. HRMS (ESI) m/z calculated for C H NO P [M + H]
3 23 46 7 2
10.2750, found 510.2756.
2
5

5
.14. Diethyl [1‐(diethoxyphosphoryl)‐3‐(5‐methylhex‐4‐enamido)propyl]phosphonate (10j).
Following the general procedure for formation of acyl chloride 9a, bishomoprenoyl chloride 9j
2
2
was prepared from bishomoprenoic acid 4j (360 mg, 2.8 mmol) as a yellow oil (335 mg, 2.30
mmol, 82%). Following the general procedure for formation of amide 10a, the title compound
was obtained from bishomoprenoyl chloride 9j (308 mg, 2.1 mmol) and bisphosphonate amine 3
1
(
497 mg, 1.5 mmol) as a yellow oil (466 mg, 1.05 mmol, 70%): H NMR (400 MHz, CDCl ) δ
3
6
6
1
=
.57 (t, J = 4.8 Hz, 1H), 5.08 (t, J = 7.0 Hz, 1H), 4.18 (quint d, J = 7.3, 4.0 Hz, 8H), 3.45 (q, J =
.1 Hz, 2H), 2.4-2.2 (m, 3H), 2.2-2.0 (m, 4H), 1.68 (s, 3H), 1.61 (s, 3H), 1.35 (t, J = 7.2 Hz,
1
3
2H); C (100.6 MHz, CDCl ) δ 173.2, 133.0, 122.9, 62.99 (d, J = 7.1 Hz, 2C), 62.94 (d, J
PC
3
PC
6.8 Hz, 2C), 38.6 (t, J = 7.1 Hz), 38.8, 34.5 (t, J = 134.1 Hz), 25.8, 25.1 (t, J = 4.3 Hz),
PC
PC
PC
3
1
2
4.4, 17.8, 16.5 (d, J = 6.7 Hz, 4C); P (162 MHz, CDCl ) δ 23.5. HRMS (ESI) m/z calculated
PC 3
+
for C H NO P [M + H] 442.2124, found 442.2124.
1
8
38
7 2
5
.15. Disodium
1‐[bis(sodiooxy)phosphoryl]‐3‐[3,7‐dimethylocta‐2,6‐dienamido]propyl)phosphonate (2a).
To a round-bottomed flask fitted with a magnetic stirbar was added bisphosphonate 10a (93 mg,
.2 mmol), collidine (0.19 mL, 1.4 mmol, 7 equivalents), and dichloromethane (7.5 mL, 10 mL
per mmol bisphosphonate). The flask was cooled to 0 °C and stirred and trimethylsilyl bromide
0.22 mL, 1.68 mmol, 8 equivalents) was added dropwise. The reaction mixture was allowed to
(
0
(
react for 24 hours while warming gradually to room temperature. The volatile materials were
removed under reduced pressure. Residual volatiles were removed by diluting the residue with
anhydrous toluene and removing the toluene under reduced pressure. This process was repeated
three times to ensure complete removal of any residual bromotrimethylsilane. The remaining
material was allowed to stand for 5 hours under vacuum. After aqueous sodium hydroxide
2
6

solution (4 equivalents) was added, the mixture was stirred for at least 15 minutes and then the
water was removed via lyophilization. The residual material was dissolved in a minimum volume
of warm water, and then diluted carefully with anhydrous acetone. The mixture was cooled to 0
°
C for 30 minutes, and the precipitate was collected by decantation and filtration. The precipitate
1
was subjected to repetitive dissolution and precipitation until no collidine was detected in the H
NMR spectrum. Final dissolution in water followed by lyophilization afforded the title
1
compound 2a as a white solid (86 mg, 0.19 mmol, 94%): H NMR (300 MHz, D O) δ 6.18 (b,
2
1
2
1
=
H major), 5.94 (br, 1H minor), 5.23 (br, 1H), 3.88 (t, J = 6.9 Hz, 2H), 2.47 (br, 1H minor), 2.4-
1
3
.1 (m, 6H), 2.1-1.8 (m, 4H), 1.73 (s, 3H), 1.67 (s, 3H); C (100.6 MHz, D O) δ 170.0, 169.7,
2
54.0, 153.8, 133.97, 133.92, 123.6, 123.4, 118.4, 117.7, 40.1 (t, J = 8.1 Hz), 39.6, 37.9 (t, J
PC
PC
3
1
114.1 Hz), 32.4, 27.6, 26.1 (br), 25.4, 24.9, 24.8, 23.6, 17.7, 17.5, 16.92, 16.90; P (121.5
-
MHz, D O) δ 19.6 (major), 19.5 (minor). HRMS (ESI) m/z calculated for C H NO P [M – H]
2
13 24
7 2
3
68.1028, found 368.1019.
5
.16. Disodium
(1‐[bis(sodiooxy)phosphoryl]‐3‐[(2E)‐3,7‐dimethylocta‐2,6‐dienamido]propyl)phosphonate
(2b). Following the general procedure for preparation of salt 2a, the title compound was obtained
1
from amide 10b (26 mg mg, 0.055 mmol) as a white solid (22 mg, 87%): H NMR (400 MHz,
D O) δ 5.62 (s, 1H), 5.09 (tt, J = 6.2, 1.4 Hz, 1H), 3.29 (t, J = 7.6 Hz, 2H), 2.1-2.0 (m, 4H), 2.0-
2
1
3
1
.8 (m, 5H), 1.62 (tt, J = 21.6 Hz, J = 6.3 Hz, 1H), 1.59 (s, 3H), 1.52 (s, 3H); C (100.6 MHz,
PH
D O) δ 170.2, 167.9 (Na CO ), 153.7, 134.0, 123.5, 117.8, 40.2 (br), 39.6, 38.2 (t, J = 114.2
2
2
3
CP
3
1
Hz), 26.2 (Br), 25.4, 24.8 (Br), 17.7, 16.9; P (162 MHz, D O) δ 19.6. HRMS (ESI) m/z
2
-
calculated for C H NO P [M – H] 368.1028, found 368.1019.
1
3
24
7 2
2
7

5
.17. Disodium
(1‐[bis(sodiooxy)phosphoryl]‐3‐[(2Z)‐3,7‐dimethylocta‐2,6‐dienamido]propyl)phosphonate
(2c). Following the general procedure for preparation of salt 2a, the title compound was obtained
from amide 10c (45 mg, 0.09 mmol) as a white-yellow solid (32.5 mg, 0.071 mmol, 79% yield):
1
H NMR (400 MHz, D O) δ 5.71 (s, 1H), 5.17 (t, J = 6.8 Hz, 1H), 3.39 (t, J = 6.3 Hz, 2H), 2.49
2
13
(
t, J = 7.2 Hz, 2H), 2.15 (m, 2H), 2.1-1.9 (bm, 3H), 1.83 (s, 3H), 1.67 (s, 3H), 1.60 (s, 3H); C
(
100.6 MHz, D O) δ 170.0, 154.0, 134.1, 123.6, 118.3, 39.2 (t, J = 8.1 Hz), 36.8 (t, J = 116.6
2
PC
PC
3
1
Hz), 32.4, 26.0, 25.1 (br), 24.9, 23.5, 16.9; P (162.0 MHz, D O) δ 19.5. HRMS (ESI) m/z
2
calculated for C H NO P 368.1028, found 368.1030.
1
3
24
7 2
5
.18. Disodium
(1‐[bis(sodiooxy)phosphoryl]‐3‐[(3S)‐3,7‐dimethyloct‐6‐enamido]propyl)phosphonate (2d).
Following the general procedure for preparation of salt 2a, the title compound was obtained from
amide 10d (111 mg, 0.23 mmol) as a white solid. Purification by flash chromatography on C18-
functionalized silica gel (3:7 H CCN:H O to 100% H CCN) recovered 13.9 mg of pure sodium
3
2
3
1
31
salt (0.041 mmol, 18%): H and P NMR data matched that for R-enantiomer 2e.
.19. Disodium
1‐[bis(sodiooxy)phosphoryl]‐3‐[(3R)‐3,7‐dimethyloct‐6‐enamido]propyl)phosphonate (2e).
5
(
Following the general procedure for preparation of salt 2a, the title compound was obtained from
amide 10e (179 mg, 0.37 mmol) as a white solid. Purification by flash chromatography on C18-
functionalized silica gel (3:7 H CCN:H O to 100% H CCN) afforded 65 mg of pure sodium salt
3
2
3
1
2
e (0.20 mmol, 53%): H NMR (300 MHz, CDCl ) δ 5.25 (t, J = 6.6 Hz, 1H), 3.42 (t, J = 6.2 Hz,
3
2
H), 2.30 (dd, J = 13.4 Hz, J = 6.4 Hz, 1H), 2.2-1.8 (m, 7H), 1.73 (s, 3H), 1.66 (s, 3H), 1.4-1.2
1
3
(
m, 3H), 0.95 (d, J = 6.5 Hz, 3H); C (75.5 MHz, D O) δ 176.5, 133.5, 124.9, 43.6, 39.7 (t, J =
2 PC
2
8

3
1
8
.2 Hz), 37.1 (t, J = 115.5 Hz), 36.2, 30.4, 25.3 (br), 25.0, 24.9, 18.8, 17.0; P (121.5 MHz,
PC
-
D O) δ 19.5. HRMS (ESI) m/z calculated for C H NO P [M – H] 370.1185, found 370.1179.
2
13 26
7 2
5
.20. Disodium
(1‐[bis(sodiooxy)phosphoryl]‐3‐[(3Z)‐4,8‐dimethylnona‐3,7‐dienamido]propyl)phosphonate
(2g). Following the general procedure for preparation of salt 2a, the title compound was obtained
1
from amide 10g (42 mg, 0.085 mmol) as a yellow solid (35 mg, 0.074 mmol, 87% yield): H
NMR (300 MHz, D O) δ 5.34 (t, J = 7.0 Hz, 1H), 5.21 (m, 1H), 3.40 (t, J = 7.5 Hz, 2H), 3.02 (d,
2
1
3
J = 7.4 Hz, 2H), 2.2-1.8 (m, 7H), 1.78 (s, 3H), 1.71 (s, 3H), 1.65 (s, 3H); C (75.5 MHz, D O)
2
1
2
75.5, 141.7, 134.0, 124.1, 116.7, 40.0 (br), 37.2 (t, J = 114.2 Hz), 35.1, 31.3, 25.8, 25.4 (br),
PC
3
1
5.1, 22.7, 17.1; P (121.5 MHz, D O) δ 19.43. HRMS (ESI) m/z calculated for C H NO P
7 2
2
14 26
-
[
M – H] 382.1185, found 382.1190.
5
.21. Disodium
(1‐[bis(sodiooxy)phosphoryl]‐3‐[(4E)‐5,9‐dimethyldeca‐4,8‐dienamido]propyl)phosphonate
(2h). Following the general procedure for preparation of salt 2a, the title compound was obtained
1
from amide 10h (102 mg, 0.2 mmol) as a white solid (19.6 mg, 0.040 mmol, 20% yield): H
NMR (300 MHz, D O) δ 5.11 (m, 2H), 3.28 (t, J = 7.4 Hz, 2H), 2.3-2.1 (m, 4H), 2.1-1.7 (m,
2
7
H), 1.60 (s, 3H), 1.54 (s, 6H); 13C (75.5 MHz, D O) δ 176.3, 138.0, 133.6, 124.5, 122.5, 40.0
2
3
1
(
br), 38.9, 37.3 (br), 36.1, 26.0, 25.3 (br), 24.9, 24.1, 17.1, 15.3; P (121.5 MHz, D O) δ 19.3.
2
-
HRMS (ESI) m/z calculated for C H NO P [M – H] 396.1341, found 396.1338.
1
5
28
7 2
5
.22. Disodium
(1‐[bis(sodiooxy)phosphoryl]‐3‐[(4Z)‐5,9‐dimethyldeca‐4,8‐dienamido]propyl)phosphonate
(2i). Following the general procedure for preparation of salt 2a, the title compound was obtained
1
from amide 10i (168 mg, 0.34 mmol) as a white solid (89.8 mg, 0.19 mmol, 54% yield): H
2
9