Synthesis of an azide-bearing N-mustard analogue of S-adenosyl-L-methionine

Article abstract of DOI:10.1021/jo2019637

The synthesis of an azide-bearing N-mustard S-adenosyl-L-methionine (SAM) analogue, 8-azido-5′-(diaminobutyric acid)-N-iodoethyl-5′- deoxyadenosine, has been accomplished in 10 steps from commercially available 2′,3′-isopropylidene adenosine. Critical to this success was executing C8 azidation prior to derivatizing the 5′-position of the ribose sugar and the late stage alkylation of the 5′ amino group with bromoethanol, which was necessitated by the reactivity of the aryl azide moiety. The azide-bearing N-mustard is envisioned as a useful biochemical tool by which to probe DNA and protein methylation patterns (Figure presented).

Full text of DOI:10.1021/jo2019637

Note
pubs.acs.org/joc
Synthesis of an Azide-Bearing N-Mustard Analogue of S-Adenosyl-L-
methionine
Van Mai and Lindsay R. Comstock*
Department of Chemistry, Wake Forest University, Winston-Salem, North Carolina 27109 United States
*
S Supporting Information
ABSTRACT: The synthesis of an azide-bearing N-mustard S-adenosyl-
L-methionine (SAM) analogue, 8-azido-5′-(diaminobutyric acid)-N-
iodoethyl-5′-deoxyadenosine, has been accomplished in 10 steps from
commercially available 2′,3′-isopropylidene adenosine. Critical to this
success was executing C8 azidation prior to derivatizing the 5′-position
of the ribose sugar and the late stage alkylation of the 5′ amino group
with bromoethanol, which was necessitated by the reactivity of the aryl
azide moiety. The azide-bearing N-mustard is envisioned as a useful biochemical tool by which to probe DNA and protein
methylation patterns.
S-adenosyl-L-methionine (SAM)-dependent methylation of
nucleic acids and proteins is now recognized as playing an
absolutely vital role in the careful regulation of gene
base, Rajski simplified and increased the versatility of 1 by
11
synthesizing analogues containing azides. It was envisioned
that such analogues would be capable of undergoing the
Staudinger ligation or copper-catalyzed azide−alkyne cyclo-
addition (CuAAC or Click) chemistry, to covalently label DNA
and protein in a methyltransferase-dependent fashion. As a
result, first-generation azide-bearing SAM analogues 2 and 3
(Figure 1) were generated and shown to be transferred to both
oligonucleotides and plasmid DNA using a variety of DNA
1
−3
transcription.
Changes in the activity or expression levels
of the eukaryotic DNA methyltransferase DNMT1 have been
integrally linked to various diseases, including carcinogenesis
4
−6
and obesity.
Although agents capable of exploiting such
flawed traits of DNMT1 may serve as a launching point for the
development of new therapeutics, substances capable of
undergoing transfer to nucleic acids and proteins in an
enzyme-dependent fashion also hold tremendous promise as
biochemical tools by which to dissect and understand biological
methylation. Our interests in developing these tools are
reflected in the synthesis and biochemical study of SAM-
based analogues containing reactive functionalities capable of
undergoing chemoselective ligations.
12,13
methyltransferases.
It was also demonstrated that these
azide-modified DNAs undergo efficient Staudinger ligation with
13,14
functionalized triarylphosphines
fluorophore-alkyne conjugate.
and Click chemistry with a
15
Although it was demonstrated that 2 and 3 can be transferred
to DNA substrates in an enzyme-dependent fashion and
undergo subsequent ligation, reaction conditions required a
13
The design of SAM analogues containing these function-
alities was inspired by Weinhold et al. Originally exploiting an
aziridine-containing SAM mimic (1, Figure 1) to alkylate DNA
1
00-fold excess of analogue. This was easily attributed to the
reduced binding of 2 and 3 to the methyltransferase due to
their core structures lacking the amino acid recognition moiety
present in SAM for optimal binding and enzymatic activity.
Thus, an N-mustard analogue bearing this recognition moiety
(
4, Figure 1) was synthesized and has been shown to generate a
highly reactive aziridinium at physiological pH. Incorporating
this amino acid functionality at the 5′-position of the ribose
sugar generated a more effective SAM analogue compared to
1
−3, as demonstrated by its enzymatic incorporation onto both
16,17
DNA and peptide substrates.
Recent reports indicate that
alternate SAM analogues bearing reactive functionalities, such
18
19−22
as ketones and alkynes,
at the sulfonium center are also
Figure 1. First-generation SAM analogues.
capable of labeling small molecule, DNA, and protein
substrates. The utility of the alkyne-containing SAM analogues
was further demonstrated through successful protein labeling
studies via Click chemistry. Although these alkyne-bearing
7
in a sequence-specific fashion, his more recent work has
demonstrated that derivatives of 1 bearing large functional
6
probes at the C8 and N positions of the adenine base are
8−10
capable of labeling DNA.
As an alternative to Weinhold’s
Received: September 22, 2011
Published: November 3, 2011
approach incorporating a large, bulky probe on the adenine
©
2011 American Chemical Society
10319
dx.doi.org/10.1021/jo2019637 | J. Org. Chem. 2011, 76, 10319−10324

The Journal of Organic Chemistry
Note
analogues show promise for future biochemical studies, their
utility is currently limited to a small number of protein lysine
methyltransferases.
In this work, we describe the further advancement of
compounds 2 and 4 by synthesizing a second-generation azide-
bearing analogue of SAM, 8-azido-5′-(diaminobutyric acid)-N-
iodoethyl-5′-deoxyadenosine, 5. As shown in Figure 2, this new
underwent facile S 2 reaction with methylbromoacetate,
N
followed by its reduction with DIBALH, to generate amino
alcohol 8 in 80% yield over two steps. Reductive amination
with tert-butyl (S)-2-[N-(tert-butoxycarbonyl)amino]-4-oxobu-
20−22
23
tanoate, in the presence of NaCNBH3 and acetic acid,
provided the fully N-protected amino alcohol 9. The azide
functionality was easily introduced using NaN to generate 10
3
in moderate yield.
Completion of the synthetic pathway required iodination of
the primary alcohol, followed by a global deprotection. On the
16
basis of precedence to generate 4, both of these steps were
expected to proceed smoothly in providing the desired azide-
containing N-mustard 5. Although iodination of 10 proceeded
smoothly to obtain 11, its subsequent deprotection using 4 N
HCl in dioxane failed to completely remove the TBS protecting
groups. In an attempt to determine whether complete
deprotection of the TBS protecting groups could be achieved,
several methodologies were explored. The use of other acids
Figure 2. 8-Azido-5′-(diaminobutyric acid)-N-iodoethyl-5′-deoxyade-
nosine, 5.
10% HCl in acetone, 50% TFA in CH Cl , and Dowex-H+
2
4
25
(
2 2
SAM-based analogue was engineered to bring together the
advantage of containing an azide on the adenine base for
subsequent chemoselective ligation reactions in addition to
incorporating the amino acid moiety for improved enzyme
interactions. Ultimately, 5 is hypothesized to serve as a
biochemical tool for future studies in identifying both sites
and substrates of biological methylation.
26
in ACN ) failed to carry out the global deprotection. Attempts
to employ TBAF to remove the TBS groups from 11 appeared
promising following TLC, but NMR and MS analysis failed to
confirm the presence of the diol and only indicated residual
TBAF salt along with several side products that could not be
identified. Although additional deprotection strategies could
have been explored, switching to the more labile triethylsilyl
On the basis of the synthetic steps that were used to generate
(
TES) protecting group was deemed to be most productive at
1
2,16
2
and 4,
a pathway to synthesize 5 was developed and
this stage in generating 5.
relied on literature precedence that derivatized the C8 position
of the adenine base prior to incorporating the amino acid
functionality at the 5′ position of the ribose sugar (Scheme 1).
It was determined that installation of the azide functionality
would be best carried out following the requisite reductive
amination to incorporate the amino acid functionality, as the
Literature precedence and previous preparation of 4 in our
laboratory have demonstrated that TES ethers on the ribose
sugar can be completely removed using HCl in dioxane.
Thus, an alternate synthetic route incorporating the TES
protecting group was developed, following the synthetic route
shown in Scheme 1. Preparation of the TES-protected
intermediates of 6−9 was facile and proceeded in similar yields
as compared to their TBS-protected analogues. Following the
successful reductive amination to yield the TES-protected
version of 9, the next step required the installation of the azide
16
use of NaBH CN was foreseen to reduce the azide functional
3
group to a primary amine. Thus, our synthesis began with
previously reported 6, and phthalimide removal easily afforded
12
7
using previously described methods. The resulting amine
Scheme 1. Proposed Synthetic Pathway to 5
1
0320
dx.doi.org/10.1021/jo2019637 | J. Org. Chem. 2011, 76, 10319−10324

The Journal of Organic Chemistry
Note
Scheme 2. Alternate Synthetic Pathway Incorporating Azide Earlier in Pathway
at the C8 position via bromide displacement. Interestingly,
reaction analysis indicated the formation of a small percentage
of the desired azide-containing product, in addition to several
side products identified as either the mono-TES-protected
product or the corresponding diol. This finding was confirmed
upon repeating the azidation, and all attempts to determine if
the resulting bromide or heat was the cause failed to provide
any conclusive results.
Scheme 3. Improved Synthesis of 13
On the basis of these difficulties incorporating the azide, an
alternative approach was envisioned to incorporate the azide
earlier in the synthetic pathway. Although this reaction was a
feasible option, our major concern was whether or not the use
of NaBH CN would reduce the azide to a primary amine
3
during the reductive amination. Model studies investigating this
conversion using 1.5 equiv of hydride indicated a 1:1 ratio of
azide to amine. Reaction optimization found that reducing the
acid moiety via reductive amination was then conducted via
treatment with tert-butyl (S)-2-[N-(tert-butoxycarbonyl)-
number of equivalents of NaBH CN to 1.1 allowed for a
3
amino]-4-oxobutanoate and NaBH CN to yield 15 in 51%
3
marked increase in product yield. Having achieved control over
the formation of the amine byproduct, we developed an
alternative synthetic pathway (Scheme 2) that incorporated the
azide on the adenine base prior to the reductive amination. In
re-engineering our pathway to 5, we relied on the previously
yield after the elimination of the contaminating primary amine
byproduct. Subsequent 5′-alkylation was carried out most
efficiently using bromoethanol in the presence of N,N-
diisopropylethylamine. It is interesting to note that the use of
iodoethanol produced a low yield of 16 (approximately 20%),
as the formation of a secondary alkylated product (addition of a
1
2
reported procedures in generating 13. We began with
27
previously reported 12, and bromination of the adenine
base occurred readily. Subsequent incorporation of the azide
afforded 13, but yields fluctuated between 39 and 68% and
were attributed to removal of the TES groups, as indicated by
TLC analysis and mass spectrometry. Reducing the reaction
time did not improve reaction yield, as unreacted starting
material remained and comigrated with 13 by TLC. Under
these conditions, it was hypothesized that retention of the
isopropylidene on the ribose sugar would be a more robust
protecting group.
12
second ethyl alcohol to the 5′-nitrogen) was prevalent.
Having successfully synthesized 16, iodination and subse-
quent global deprotection was carried out. Although the
literature indicated that iodination should be carried out at 0
16
°
C, we consistently observed unreacted starting material that
was independent of reaction time. While raising the temper-
ature to ambient conditions failed to yield complete conversion,
it was found that a gentle reflux in CH Cl facilitated iodination
2
2
in a near quantitative yield. Having the penultimate product in
hand, initial attempts to globally deprotect 16 using 4 N HCl in
dioxane failed to go to completion. Although 5 was the major
product obtained, a derivative of 5 retaining the tert-butyl ester
was also present. Further reaction optimization to increase the
overall yield of 5 indicated that complete global deprotection
To test the viability of the isopropylidene group during the
bromination and azidation steps, a pathway that involved the
formation of 8-azido-5′-phthalimide-2′,3′-O-isopropylidene ad-
enosine (19) was developed (Scheme 3). We began with
27
previously described 17, and bromination and azidation were
easily carried out using the chemistry described above to
generate 18 and 19, respectively, in high yields. Subsequent
isopropylidene cleavage and reprotection as the TES ethers
produced 13 in a yield of 86%, which was significantly higher
compared to the conversion of 12 to 13.
was achieved under a gentle reflux in CH Cl for 3 h.
2 2
In summary, 8-azido-5′-(diaminobutyric acid)-N-iodoethyl-
5′-deoxyadenosine, 5, was successfully synthesized on the basis
of the two previously reported SAM analogues, 2 and 4. The
use of suitable protecting groups was found to be a key factor in
several transformations to successfully obtain the final azide-
bearing N-mustard SAM analogue. Preliminary work in our
laboratory has indicated that 5 can be enzymatically transferred
After obtaining 13 in sufficient quantity, the synthetic
pathway (see Scheme 2) was continued upon removal of the
phthalimide to primary amine 14. Introduction of the amino
1
0321
dx.doi.org/10.1021/jo2019637 | J. Org. Chem. 2011, 76, 10319−10324

The Journal of Organic Chemistry
Note
by methyltransferase TaqI to plasmid DNA using a restriction/
protection assay in comparable yields to N-mustard analogue
4.35 (dd, J = 4.6, 1.9 Hz, 1H), 4.25−4.21 (m, 1H), 3.79−3.69 (m,
2
H), 3.02−2.91 (m, 3H), 2.84−2.77 (m, 2H), 0.96 (s, 9H), 0.78 (s,
13
2
8
9H), 0.14 (s, 3H), 0.13 (s, 3H), −0.11 (s, 3H), −0.44 (s, 3H);
C
4.
On the basis of these results, we anticipate that 5 will be
NMR (CDCl ) δ 154.8, 152.9, 150.6, 128.4, 120.6, 90.6, 85.3, 73.9,
3
useful to a variety of DNA and protein methyltransferases and
serve as a unique biochemical tool to detect sites of biological
7
2.7, 60.5, 51.2, 50.8, 26.0, 25.8, 18.2, 17.9, −4.4, −4.4, −4.5, −5.3;
+
HRMS-ESI calcd for C H BrN O Si (M + Na ) 639.2116, obsd
2
9
24 45
6
4
2
methylation. Further biochemical analysis of 5 is ongoing and
6
39.2128.
will be reported shortly.
8
-Bromo-5′-(N-Boc-diaminobutyric acid-O-tert-butyl ester)-N-
ethanolamine-5′-deoxy-2′,3′-bis-(O-tert-butyldimethylsilyl) Adeno-
sine (9). To a solution of 8-bromo-5′-ethanolamine-5′-deoxy-2′,3′-bis-
EXPERIMENTAL SECTION
■
(
O-tert-butyldimethylsilyl) adenosine (8) (0.400 g, 0.647 mmol) and
tert-butyl (S)-2-[N-(tert-butoxycarbonyl)amino]-4-oxobutanoate
0.161 g, 0.588 mmol) in 2.82 mL of dry MeOH were added
General Experimental Methods. All reagents and solvents were
purchased from commercial sources and used without additional
purification. Anhydrous solvents were obtained from a Meyer solvent
system except for DMSO and toluene, which were purchased.
Reactions were performed at room temperature under argon, unless
(
NaBH CN (0.055 g, 0.882 mmol) and AcOH (33.7 μL, 0.588 mmol).
3
The reaction mixture was stirred overnight. After the reaction was
diluted with EtOAc and saturated NaHCO , the organic layer was
1
13
3
otherwise indicated. H and C NMR spectra were recorded on a 300
or 500 MHz spectrometer using solvent as the internal reference.
Chemical shifts are reported in δ (ppm). Analytical and semi-
preparative HPLC was performed with the UV−Vis detector set at 254
nm. An Alltima C18 analytical column (100 Å, 5 μm, 250 × 4.6 mm)
and Alltima C18 preparative column (100 Å, 5 μm, 250 × 10.0 mm)
were used for analysis and isolation of 5. All mobile phases were
filtered through a 0.22 μm membrane filter prior to use.
Experimental Procedures. 8-Bromo-5′-amine-5′-deoxy-2′,3′-
bis-(O-tert-butyldimethylsilyl) Adenosine (7). Ethylenediamine (668
μL, 9.984 mmol) was added to 8-bromo-5′-phthalimide-5′-deoxy-2′,3′-
bis-(O-tert-butyldimethylsilyl) adenosine (6) (1.405 g, 1.997 mmol) in
washed with saturated NaHCO , dried over Na SO , and evaporated
3
2
4
in vacuo. Column chromatography (10:4:2:1 petroleum ether/EtOAc/
1
CH Cl /MeOH) yielded 9 (0.353 g, 72%) as a solid: H NMR
2
2
(
CDCl ) δ 8.27 (s, 1H), 5.94 (d, J = 5.2 Hz, 1H), 5.64 (bs, 2H), 5.43
3
(bd, J = 7.8 Hz, 1H), 5.31 (t, J = 4.9 Hz, 1H), 4.39 (t, J = 4.1 Hz, 1H),
4
2
1
−
1
5
.20−4.15 (m, 2H), 3.55−3.51 (m, 2H), 2.96−2.90 (m, 2H), 2.69−
.56 (m, 4H), 1.99−1.92 (m, 1H), 1.73−1.64 (m, 1H), 1.42 (s, 9H),
.42 (s, 9H), 0.96 (s, 9H), 0.81 (s, 9H), 0.16 (s, 3H), 0.14 (s, 3H),
13
0.06 (s, 3H), −0.32 (s, 3H); C NMR (CDCl ) δ 172.0, 155.6,
3
54.6, 152.8, 150.8, 128.7, 120.6, 91.0, 85.8, 82.0, 79.8, 74.4, 72.4, 59.6,
6.6, 56.2, 52.9, 50.5, 29.7, 28.5, 28.1, 26.0, 25.8, 18.2, 18.0, −4.2, −4.3,
+
−
8
4.5, −5.0; HRMS-ESI calcd for C H BrN O Si (M + Na )
3
7
68
7
8
2
6
5.5 mL of EtOH. The reaction was heated at 70 °C and stirred for 5
96.3743, obsd 896.3715.
h. The solvent was evaporated in vacuo and chromatographed on silica
8
-Azido-5′-(N-Boc-diaminobutyric acid-O-tert-butyl ester)-N-
pretreated with 2% TEA (4:2:1 EtOAc/CH Cl /MeOH) to yield 7
2
2
1
ethanolamine-5′-deoxy-2′,3′-bis-(O-tert-butyldimethylsilyl) Adeno-
(
=
0.807 g, 70%) as a solid: H NMR (CDCl ) δ 8.29 (s, 1H), 5.99 (d, J
3
sine (10). NaN (0.056 g, 0.858 mmol) was added to a solution of 8-
3
6.4 Hz, 1H), 5.57 (bs, 2H), 5.26 (dd, J = 6.6, 4.7 Hz, 1H), 4.41 (dd,
bromo-5′-(N-Boc-diaminobutyric acid-O-tert-butyl ester)-N-ethanol-
amine-5′-deoxy-2′,3′-bis-(O-tert-butyldimethylsilyl) adenosine (9)
J = 4.6, 2.4 Hz, 1H), 4.11−4.07 (m, 1H), 3.06−3.04 (m, 2H), 1.65 (bs,
2
3
1
H), 0.96 (s, 9H), 0.80 (s, 9H), 0.14 (s, 3H), 0.14 (s, 3H), −0.08 (s,
(
0.188 g, 0.215 mmol) in 1.82 mL of DMSO. The reaction mixture
was heated to 85 °C and stirred overnight. Upon cooling to rt, an
aqueous workup was performed (saturated NaHCO , CH Cl , brine).
The resulting organic was dried over Na SO , evaporated in vacuo, and
13
H), −0.43 (s, 3H); C NMR (CDCl ) δ 154.7, 152.6, 150.6, 128.1,
3
20.4, 90.4, 87.3, 73.1, 72.6, 43.7, 25.8, 25.6, 18.0, 17.7, −4.5, −4.6,
+
3
2
2
−
4.7, −5.4; HRMS-ESI calcd for C H BrN O Si (M + Na )
22
41
6
3
2
2
4
5
95.1853, obsd 595.1858.
-Bromo-5′-amino-acetic acid methyl ester-5′-deoxy-2′,3′-bis-
O-tert-butyldimethylsilyl) Adenosine. To 8-bromo-5′-amine-5′-
deoxy-2′,3′-bis-(O-tert-butyldimethylsilyl) adenosine (7) (0.825 g,
.439 mmol) in 7.9 mL of dry THF was added TEA (240 μL, 1.727
column chromatographed (8:4:2:1 petroleum ether/EtOAc/CH Cl /
2
2
8
1
MeOH) to yield 10 (0.087 g, 48%) as a light yellow solid: H NMR
(
(
CDCl ) δ 8.21 (s, 1H), 5.73 (d, J = 5.0 Hz, 1H), 5.45 (bd, J = 7.9 Hz,
3
1
4
2
0
−
H), 5.35 (bs, 2H), 5.08 (t, J = 4.4 Hz, 1H), 4.40 (t, J = 3.6 Hz, 1H),
.16−4.11 (m, 2H), 3.55−3.48 (m, 2H), 2.93−2.87 (m, 2H), 2.66−
.59 (m, 4H), 2.00−1.89 (m, 1H), 1.73−1.60 (m, 1H), 1.41 (s, 18H),
.94 (s, 9H), 0.81 (s, 9H), 0.13 (s, 3H), 0.12 (s, 3H), −0.06 (s, 3H),
1
mmol). Methylbromoacetate (164 μL, 1.727 mmol) in 4 mL of dry
THF was then added dropwise to the solution. The reaction mixture
was stirred overnight. The resulting precipitate was filtered off and
1
3
0.28 (s, 3H); C NMR (CDCl ) δ 172.0, 155.7, 153.7, 151.9, 150.3,
washed with anhydrous Et O; the resulting organic was evaporated in
3
2
1
5
46.2, 128.9, 118.4, 88.4, 83.1, 82.0, 79.8, 74.3, 72.8, 59.3, 56.7, 56.3,
vacuo. Column chromatography (3:1 EtOAc/CH Cl to 15:5:1
2
2
3.0, 50.6, 34.3, 29.7, 28.5, 28.1, 26.1, 25.8, 18.2, 18.0, −4.1, −4.3,
EtOAc/CH Cl /MeOH) yielded 8-bromo-5′-amino-acetic acid methyl
2
2
+
−
4.4, −4.9; HRMS-ESI calcd for C H N O Si (M + Na )
ester-5′-deoxy-2′,3′-bis-(O-tert-butyldimethylsilyl) adenosine (0.770 g,
37 68 10
8
2
1
859.4652, obsd 859.4667.
8
3%) as a white solid: H NMR (CDCl ) δ 8.28 (s, 1H), 5.98 (d, J =
3
5
′-Phthalimide-5′-deoxy-2′,3′-bis-(O-triethylsilyl) Adenosine
6
=
.7 Hz, 1H), 5.48 (bs, 2H), 5.30 (dd, J = 6.7, 4.7 Hz, 1H), 4.38 (dd, J
4.7, 2.2 Hz, 1H), 4.20−4.16 (m, 1H), 3.70 (s, 3H), 3.46 (d, J = 2.2
(
12). Note: Although this product has been previously described in
27
1
the literature, incorrect H NMR data was reported and is corrected
here. Additionally, our procedure deviated from that reported and was
based on the previously described tert-butyldimethylsilyl-protected
Hz, 2H), 2.99 (dd, J = 12.3, 3.7 Hz, 1H), 2.92 (dd, J = 12.3, 6.3 Hz,
1
3
1
2
H), 2.64 (bs, 1H), 0.95 (s, 9H), 0.79 (s, 9H), 0.13 (s, 3H), 0.13 (s,
13
H), −0.10 (s, 3H), −0.45 (s, 3H); C NMR (CDCl ) δ 172.5, 154.6,
3
12
version of 12. To commercially available 2′,3′-isopropylidene
52.6, 150.6, 128.1, 120.4, 90.4, 85.8, 73.6, 72.4, 51.5, 51.0, 50.8, 25.7,
adenosine (1.005 g, 3.257 mmol) in 11 mL of dry THF were added
5.6, 17.9, 17.6, −4.6, −4.6, −4.7, −5.5; HRMS-ESI calcd for
+
phthalimide (0.495 g, 3.366 mmol) and PPh (0.852 g, 3.250 mmol).
C H BrN O Si (M + Na ) 667.2066, obsd 667.2071.
3
25
45
6
5
2
After the mixture was stirred for 10 min, DIAD (640 μL, 3.250 mmol)
was added to the mixture, which was stirred an additional 2.5 h. The
8
-Bromo-5′-ethanolamine-5′-deoxy-2′,3′-bis-(O-tert-butyldime-
thylsilyl) Adenosine (8). To 8-bromo-5′-amino-acetic acid methyl
ester-5′-deoxy-2′,3′-bis-(O-tert-butyldimethylsilyl) adenosine (0.770 g,
white precipitate was filtered off and washed with cold Et O to yield
2
1
.192 mmol) in 25.6 mL of dry THF at 0 °C was slowly added 1 M
the product that was taken forward. 5′-Phthalimide-5′-deoxy-2′,3′-
isopropylidene adenosine (1.175 g, 2.693 mmol) was dissolved in 27
DIBALH in hexanes (5.96 mmol). The reaction was kept cold for a 10
min, warmed to rt, and stirred for an additional 5 h. Saturated
potassium sodium tartrate tetrahydrate (Rochelle’s salt, 25.6 mL) was
added to the reaction, and the mixture was stirred vigorously
mL of 3:1:1 TFA/H O/THF, and the mixture was stirred for 2 h. The
2
solvent was evaporated in vacuo and coevaporated with EtOH (× 3).
The resulting material was dissolved in 5.93 mL of dry DMF, followed
by the addition of imidazole (0.917 g, 13.47 mmol) and TESCl (994
μL, 5.925 mmol). The reaction was stirred overnight, followed by an
overnight. The organic was washed (saturated Rochelle’s salt, H O,
2
EtOAc, brine), dried over Na SO , and evaporated in vacuo to obtain
2
4
1
8
(0.714 g, 97%) as a white solid: H NMR (CDCl ) δ 8.27 (s, 1H),
aqueous workup (saturated NH Cl (× 2), EtOAc, brine), dried over
3
4
6.00 (d, J = 7.0 Hz, 1H), 5.72 (bs, 2H), 5.29 (dd, J = 6.9, 4.7 Hz, 1H),
Na SO , and evaporated in vacuo. Column chromatography (4:4:2:1
2
4
1
0322
dx.doi.org/10.1021/jo2019637 | J. Org. Chem. 2011, 76, 10319−10324

The Journal of Organic Chemistry
Note
hexanes/EtOAc/CH Cl /MeOH) yielded 12 (1.148 g, 68%) as a
(t, J = 7.9 Hz, 9H), 0.69 (q, J = 7.7 Hz, 6H), 0.38 (qq, J = 15, 7.6 Hz,
2
2
1
13
white solid: H NMR (CDCl ) δ 8.12 (s, 1H), 8.00 (s, 1H), 7.88−7.85
6H); C NMR (CDCl ) δ 152.9, 150.8, 149.3, 145.1, 117.6, 87.5,
3
3
(
m, 2H), 7.75−7.72 (m, 2H), 5.86 (d, J = 6.4 Hz, 1H), 5.67 (bs, 2H),
86.9, 73.3, 72.7, 44.2, 7.6, 7.3, 5.8, 5.3; HRMS-ESI calcd for
+
5
7
0
.28 (dd, J = 6.3, 4.1 Hz, 1H), 4.35−4.30 (m, 2H), 4.24 (dd, J = 13.7,
.2 Hz, 1H), 3.93 (dd, J = 13.7, 5.4 Hz, 1H), 0.93 (t, J = 7.8 Hz, 9H),
.80 (t, J = 7.9 Hz, 9H), 0.61 (q, J = 8.0 Hz, 6H), 0.38 (qq, J = 15, 7.6
C H N O Si (M + Na ) 558.2763, obsd 558.2763.
2
2
41
9
3
2
8-Azido-5′-N-Boc-diaminobutyric acid-O-tert-butyl ester-5′-
deoxy-2′,3′-bis-(O-triethylsilyl) Adenosine (15). In an adaptation of
13
27
16
Hz, 6H); C NMR data matched that previously described.
a reported procedure, NaBH CN (0.066 g, 1.056 mmol) and AcOH
3
8
-Bromo-5′-phthalimide-5′-deoxy-2′,3′-bis-(O-isopropylidene)
(55 μL, 0.960 mmol) were added to a solution of 8-azido-5′-amine-5′-
deoxy-2′,3′-bis-(O-triethylsilyl) adenosine (14) (0.566 g, 1.056 mmol)
and tert-butyl (S)-2-[N-(tert-butoxycarbonyl)amino]-4-oxobutanoate
(0.262 g, 0.960 mmol) in 5.6 mL of dry MeOH. The reaction mixture
was stirred for 2 h. After diluting the reaction with EtOAc and
Adenosine (18). To 5′-phthalimide-5′-deoxy-2′,3′-bis-(O-isopropyli-
dene) adenosine (17) (5.123 g, 11.74 mmol) in 176 mL of 7:4
dioxane/0.5 M NaOAc (pH 5.2) was added Br (1.21 mL, 23.48
mmol). The reaction was stirred for 3.5 h and followed by an aqueous
workup (saturated Na S O , CH Cl , brine), dried over Na SO , and
2
saturated NaHCO , the organic layer was washed with saturated
2
2
3
2
2
2
4
3
evaporated in vacuo. Column chromatography (32:4:1 EtOAc/
NaHCO , dried over Na SO , and evaporated in vacuo. Column
3
2
4
1
CH Cl / MeOH) yielded 18 (5.879 g, 97%) as a yellow solid: H
chromatography (12:4:2:0.6 petroleum ether/EtOAc/CH Cl /
2
2
2 2
1
NMR (CDCl ) δ 8.04 (s, 1H), 7.78−7.75 (m, 2H), 7.69−7.65 (m,
MeOH) yielded 15 (0.431 g, 51%) as a light yellow solid: H NMR
3
2
1
1
H), 6.16 (d, J = 1.3 Hz, 1H), 5.97 (bs, 2H), 5.74 (dd, J = 6.2, 1.3 Hz,
H), 5.34 (dd, J = 6.2, 3.2 Hz, 1H), 4.57−4.51 (m, 1H), 3.99 (dd, J =
4.1, 5.7 Hz, 1H), 3.87 (dd, J = 14.1, 6.8 Hz, 1H), 1.57 (s, 3H), 1.38
(CDCl ) δ 8.21 (s, 1H), 5.76 (d, J = 6.7 Hz, 1H), 5.56 (bd, J = 6.5 Hz,
3
2H), 5.43 (bd, J = 8.0 Hz, 1H), 5.25 (dd, J = 6.5, 4.8 Hz, 1H), 4.35
(dd, J = 4.6, 2.2 Hz, 1H), 4.27−4.22 (m, 1H), 4.12−4.08 (m, 1H),
2.87 (d, J = 4.9 Hz, 2H), 2.74−2.58 (m, 2H), 2.16 (bs, 1H), 2.02−1.89
(m, 1H), 1.80−1.69 (m, 1H), 1.45 (s, 9H), 1.44 (s, 9H), 1.01 (t, J =
7.9 Hz, 9H), 0.80 (t, J = 7.9 Hz, 9H), 0.68 (q, J = 7.8 Hz, 6H), 0.35
(
s, 3H); 13C NMR (CDCl ) δ 167.9, 154.4, 152.7, 150.2, 133.8, 131.8,
3
1
27.5, 123.2, 120.0, 114.0, 91.1, 85.7, 83.5, 82.6, 39.3, 27.0, 25.3;
+
HRMS-ESI calcd for C H BrN O (M + Na ) 537.0493, obsd
21
19
6
5
1
3
537.0485.
(qq, J = 15, 7.6 Hz, 6H); C NMR (CDCl ) δ 170.8, 154.4, 152.8,
3
8
-Azido-5′-phthalimide-5′-deoxy-2′,3′-bis-(O-isopropylidene)
150.7, 149.3, 145.2, 117.7, 87.4, 85.2, 81.5, 79.3, 74.0, 72.5, 53.0, 51.8,
46.3, 33.4, 28.8, 28.5, 7.6, 7.2, 5.8, 5.3; HRMS-ESI calcd for
Adenosine (19). NaN (1.602 g, 24.62 mmol) was added to a solution
3
+
of 8-bromo-5′-phthalimide-5′-deoxy-2′,3′-bis-(O-isopropylidene) ad-
enosine (18) (3.172 g, 6.155 mmol) in 51.9 mL of DMSO. The
reaction mixture was heated to 85 °C and stirred for 10.5 h. The
organic was washed (saturated NaHCO , CH Cl , brine), dried over
C H N O Si (M + Na ) 815.4390, obsd 815.4390.
3
5
64 10
7
2
8-Azido-5′-(N-Boc-diaminobutyric acid-O-tert-butyl ester)-N-
ethanolamine-5′-deoxy-2′,3′-bis-(O-triethylsilyl) Adenosine (16).
N,N-diisopropylethylamine (1.09 mL, 6.235 mmol) and 2-bromoe-
thanol (443 μL, 6.235 mmol) were added to a solution of 8-azido-5′-
N-Boc-diaminobutyric acid-O-tert-butyl ester-5′-deoxy-2′,3′-bis-(O-trie-
thylsilyl) adenosine (15) (0.380 g, 0.480 mmol) in 2.51 mL of dry
toluene. The reaction was heated to 70 °C and stirred for 27 h. Upon
3
2
2
Na SO , and evaporated in vacuo. Column chromatography (16:2:1
2
4
EtOAc/CH Cl /MeOH) yielded 19 (2.632 g, 90%) as a light yellow
2
2
1
solid: H NMR (CDCl ) δ 8.01 (s, 1H), 7.78−7.74 (m, 2H), 7.70−
3
7
6
3
.66 (m, 2H), 5.98 (d, J = 1.5 Hz, 1H), 5.64 (bs, 2H), 5.60 (dd, J =
.3, 1.5 Hz, 1H), 5.25 (dd, J = 6.3, 3.3 Hz, 1H), 4.52−4.47 (m, 1H),
.99 (dd, J = 14.1, 5.7 Hz, 1H), 3.89 (dd, J = 14.1, 6.7 Hz, 1H), 1.55
cooling, an aqueous workup (saturated NaHCO , EtOAc, brine) was
3
performed. The resulting organic was dried over Na SO and
2
4
(
s, 3H), 1.36 (s, 3H); 13C NMR (CDCl ) δ 167.9, 153.6, 151.7, 149.6,
evaporated in vacuo. Column chromatography on silica (1:0.5
3
1
45.1, 133.8, 131.8, 123.1, 117.7, 114.1, 88.4, 85.2, 83.4, 82.5, 39.4,
petroleum ether/EtOAc to 1:2 petroleum ether/EtOAc) yielded 16
+
1
27.0, 25.4; HRMS-ESI calcd for C H N O (M + Na ) 500.1401,
(0.229 g, 57%) as a light yellow solid: H NMR (CDCl ) δ 8.20 (s,
21
19
9
5
3
obsd 500.1395.
1H), 5.73 (d, J = 5.3 Hz, 1H), 5.69 (bs, 2H), 5.43 (bd, J = 8.2 Hz, 1H),
5.19 (t, J = 4.9 Hz, 1H), 4.34 (t, J = 4.1 Hz, 1H), 4.19−4.11 (m, 2H),
3.55−3.47 (m, 2H), 2.96−2.82 (m, 2H), 2.70−2.56 (m, 4H), 1.97−
1.90 (m, 1H), 1.76−1.66 (m, 1H), 1.42 (s, 9H), 1.42 (s, 9H), 1.00 (t, J
= 7.9 Hz, 9H), 0.81 (t, J = 7.9 Hz, 9H), 0.66 (q, J = 7.8 Hz, 6H), 0.39
8
-Azido-5′-phthalimide-5′-deoxy-2′,3′-bis-(O-triethylsilyl) Adeno-
sine (13). 8-Azido-5′-phthalimide-5′-deoxy-2′,3′-bis-(O-isopropyli-
dene) adenosine (19) (2.632 g, 5.512 mmol) was dissolved in 55.3
mL of 3:1:1 TFA/H O/THF and stirred for 2.5 h. The solvent was
2
1
3
evaporated in vacuo and coevaporated with EtOH (× 3). The resulting
material was dissolved in 12.1 mL of dry DMF, followed by the
addition of imidazole (1.876 g, 27.56 mmol) and chlorotriethylsilane
(qq, J = 15, 7.5 Hz, 6H); C NMR (CDCl ) δ 171.9, 155.4, 153.6,
3
151.6, 150.0, 145.9, 118.1, 87.9, 83.0, 81.7, 79.5, 74.4, 72.4, 59.1, 56.5,
56.1, 52.8, 50.5, 29.6, 28.2, 27.9, 6.8, 6.5, 4.9, 4.5; HRMS-ESI calcd for
+
(
TESCl) (2.04 mL, 12.13 mmol). The reaction was stirred overnight,
C H N O Si (M + H ) 837.4833, obsd 837.4810.
3
7
68 10
8
2
followed by an aqueous workup (saturated NH Cl (× 2), EtOAc,
brine). The organic was dried over Na SO and evaporated in vacuo.
8-Azido-5′-(diaminobutyric acid)-N-iodoethyl-5′-deoxyadeno-
4
sine Ammonium Hydrochloride (5). I (0.0623 g, 0.2453 mmol)
2
4
2
Column chromatography (8:4:2:0.5 hexanes/EtOAc/CH Cl /MeOH
was added to triphenylphosphine (0.0623 g, 0.2375 mmol) and
2
2
to 8:4:2:1 hexanes/EtOAc/CH Cl /MeOH) yielded 13 (3.139 g,
imidazole (0.0162 g, 0.2375 mmol) in 623 μL of CH Cl at 0 °C. The
2
2
2
2
1
8
7
=
6%) as a light yellow solid: H NMR (CDCl ) δ 7.99 (s, 1H), 7.85−
components were stirred until TLC indicated complete consumption
3
.82 (m, 2H), 7.73−7.70 (m, 2H), 5.79 (d, J = 6.9 Hz, 1H), 5.54 (dd, J
of PPh . 8-Azido-5′-(N-Boc-diaminobutyric acid-O-tert-butyl ester)-N-
3
6.9, 4.5 Hz, 1H), 5.47 (bs, 2H), 4.47 (dd, J = 4.5, 1.3 Hz, 1H), 4.29−
ethanolamine-5′-deoxy-2′,3′-bis-(O-triethylsilyl) adenosine (16)
(0.1308 g, 0.1562 mmol) in CH Cl (623 μL) was then added and
4
.20 (m, 2H), 3.94−3.86 (m, 1H), 0.92 (t, J = 7.9 Hz, 9H), 0.81 (t, J =
2
2
7.9 Hz, 9H), 0.60 (q, J = 7.7 Hz, 6H), 0.37 (qq, J = 15, 7.6 Hz, 6H);
warmed to 40 °C slowly. After stirring for 30 min, the reaction was
13
C NMR (CDCl ) δ 167.0, 152.6, 150.7, 149.4, 145.3, 133.3, 131.2,
cooled and diluted with ice-chilled CH Cl and H O. The organic
3
2
2
2
1
22.6, 117.5, 87.0, 83.5, 73.9, 71.3, 39.9, 7.5, 7.4, 5.7, 5.3; HRMS-ESI
layer was washed with H
dioxane (4N, 898 μL) was added to the iodinated product in CH
2
O (× 3) and evaporated in vacuo. HCl/
Cl
2
+
calcd for C H N O Si (M + Na ) 688.2818, obsd 688.2805.
2
30
43
9
5
2
8
-Azido-5′-amine-5′-deoxy-2′,3′-bis-(O-triethylsilyl) Adenosine
(1.80 mL), and the mixture was heated to 40 °C for 3 h. Upon cooling,
(14). To 8-azido-5′-phthalimide-5′-deoxy-2′,3′-bis-(O-triethylsilyl) ad-
ice-chilled H O was added, and the aqueous layer was extracted with
2
enosine (13) (0.917 g, 1.377 mmol) in 45 mL of EtOH was added
ethylenediamine (460 μL, 6.887 mmol). The reaction was heated to 70
CH Cl (× 3) prior to lyophilization to afford a light yellow solid. This
2
2
was dissolved in minimal MeOH, and EtOAc was then added dropwise
to precipitate a light yellow solid, which was then further purified by
reverse-phase HPLC. Compound separation utilized a gradient system
comprising of 0.1% TFA in water (solvent A) and HPLC grade ACN
(solvent B) using a flow rate of 1.0 mL/min (analytical scale) or 4.0
mL/min (preparative scale). The gradient was run isocratically with
5% B for 2 min followed by a linear gradient of 5−10% B over a 16
min period. The gradient was then increased to 100% B over the next
°
C and stirred for 5 h. The solvent was evaporated in vacuo, and the
product was chromatographed on silica pretreated with 2% TEA
4:2:1 EtOAc/CH Cl /MeOH) to yield 14 (0.566 g, 77%) as a light
(
2
2
1
yellow solid: H NMR (CDCl ) δ 8.23 (s, 1H), 5.77 (d, J = 6.3 Hz,
3
1
2
H), 5.55 (bs, 2H), 5.24 (dd, J = 6.3, 4.8 Hz, 1H), 4.37 (dd, J = 4.7,
.8 Hz, 1H), 4.03−3.98 (m, 1H), 3.02 (dd, J = 13, 4.0 Hz, 1H), 2.97
(
dd, J = 13, 5.8 Hz, 1H), 1.72 (bs, 2H), 1.02 (t, J = 7.9 Hz, 9H), 0.81
1
0323
dx.doi.org/10.1021/jo2019637 | J. Org. Chem. 2011, 76, 10319−10324

The Journal of Organic Chemistry
Note
1
min and run isocratically for an additional 6 min. Under these
(21) Wang, R.; Zheng, W.; Yu, H.; Deng, H.; Luo, M. J. Am. Chem.
Soc. 2011, 133, 7648−7651.
conditions, the azide-bearing N-mustard adenosine (5) eluted at 16
min. 5 was obtained following product lyophilization (0.025 g, 28%):
(22) Islam, K.; Zheng, W.; Yu, H.; Deng, H.; Luo, M. ACS Chem.
Biol. 2011, 6, 679−684.
1
H NMR (1% TFA in CD OD) δ 8.39 (s, 1H), 5.96 (d, J = 3.8 Hz,
3
1
H), 4.80 (dd, J = 5.4, 4.0 Hz, 1H), 4.49 (t, J = 5.7 Hz, 1H), 4.42−4.38
(23) Fukuyama, T.; Lin, S.-C.; Li, L. J. Am. Chem. Soc. 1990, 112,
7050−7051.
(
m, 1H), 4.08 (dd, J = 7.4, 5.6 Hz, 1H), 3.73 (dd, J = 14, 10 Hz, 1H),
3
2
1
5
5
.64−3.59 (m, 3H), 3.54−3.49 (m, 1H), 3.47−3.41 (m, 3H), 2.43−
(24) Eggert, J. P. W.; Luning, U.; Nather, C. Eur. J. Org. Chem. 2005,
1107−1112.
13
.35 (m, 1H), 2.29−2.22 (m, 1H); C NMR (1% TFA in CD OD) δ
3
71.1, 150.6, 150.2, 149.4, 145.3, 119.1, 90.9, 80.3, 73.6, 73.5, 57.4,
5.9, 52.0, 51.7, 26.2, 7.7; HRMS-ESI calcd for C H IN O (M )
(25) Yamazaki, S.; Iwata, Y.; Fukushima, Y. Org. Biomol. Chem. 2009,
7, 655−659.
+
5
+
16
24
10
63.0970, obsd 563.0969. Note: The 13C NMR spectrum of 5 contains
(26) Comstock, L. R.; Denu, J. M. Org. Biomol. Chem. 2007, 5,
3087−3091.
minor TFA signals in the range of δ 116−117 and δ 161−163.
(
(
(
27) Petersen, S. G.; Rajski, S. R. J. Org. Chem. 2005, 70, 5833−5839.
28) Mai, V. M.S. Thesis, Wake Forest University, 2011.
ASSOCIATED CONTENT
■
29) Although incorporation of an azide on the C8 position of the
*
S
Supporting Information
adenine base has been tolerated by a small panel of DNA
methyltransferases, it is anticipated that its addition may perturb
binding with other DNA and protein methyltransferases. Ref 9
evaluates various methyltransferase crystal structures in complex with
SAM (or its analogue) and indicates that some enzymes are likely to
accept this substitution, whereas others will not. For those without
crystal structures, the ability of 5 to bind within the active site must be
determined experimentally.
1
13
H and C NMR spectra for new compounds 5, 7−10, 13−16,
1
8, and 19 and HPLC chromatogram of 5. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*
E-mail: comstolr@wfu.edu.
ACKNOWLEDGMENTS
■
We thank our colleagues Dr. Rajsekhar Guddneppanavar and
Dr. Yuhao Du for very helpful discussions. This material is
based upon work supported in part by the North Carolina
Biotechnology Center. Additionally, we thank Wake Forest
University for generous startup funding.
REFERENCES
■
(
(
(
(
1) Cheng, X.; Blumenthal, R. M. Biochemistry 2010, 49, 2999−3008.
2) Razin, A. EMBO J. 1998, 17, 4905−4908.
3) Zhang, Y.; Reinberg, D. Genes Dev. 2001, 15, 2343−2360.
4) Momparler, R. L.; Bovenzi, V. J. Cell. Physiol. 2000, 183, 145−
154.
(
(
(
5) Robertson, K. D. Oncogene 2001, 20, 3139−3155.
6) Franks, P. W.; Ling, C. BMC Med. 2010, 8, 88−92.
7) Pignot, M.; Siethoff, C.; Linscheid, M.; Weinhold, E. Angew.
Chem., Int. Ed. 1998, 37, 2888−2891.
8) Pljevaljcic, G.; Pignot, M.; Weinhold, E. J. Am. Chem. Soc. 2003,
25, 3486−3492.
9) Pljevaljcic, G.; Schmidt, F.; Weinhold, E. ChemBioChem 2004, 5,
65−269.
10) Schmidt, F. H.; Huben, M.; Gider, B.; Renault, F.; Teulade-
Fichou, M. P.; Weinhold, E. Bioorg. Med. Chem. 2008, 16, 40−48.
11) The addition of an azide at the C8 position of SAM has been
(
1
(
2
(
(
successful in generating photoaffinity probes to investigate SAM−
methyltransferase interactions. See articles cited within ref 12.
(
12) Comstock, L. R.; Rajski, S. R. J. Org. Chem. 2004, 69, 1425−
428.
13) Comstock, L. R.; Rajski, S. R. Nucleic Acids Res. 2005, 33, 1644−
652.
14) Comstock, L. R.; Rajski, S. R. J. Am. Chem. Soc. 2005, 127,
4136−14137.
15) Comstock, L. R. Ph.D. Dissertation, University of Wisconsin −
Madison, 2005.
1
(
1
(
1
(
(
16) Weller, R. L.; Rajski, S. R. ChemBioChem 2006, 7, 243−245.
(
17) Osborne, T.; Roska, R. L.; Rajski, S. R.; Thompson, P. R. J. Am.
Chem. Soc. 2008, 130, 4574−4575.
18) Lee, B. W. K.; Sun, H. G.; Zang, T.; Kim, B. J.; Alfaro, J. F.;
Zhou, Z. S. J. Am. Chem. Soc. 2010, 132, 3642−3643.
19) Dalhoff, C.; Lukinavicius, G.; Klimasauskas, S.; Weinhold, E.
Nat. Chem. Biol. 2006, 2, 31−32.
20) Binda, O.; Boyce, M.; Rush, J. S.; Palaniappan, K. K.; Bertozzi,
C. R.; Gozani, O. ChemBioChem 2011, 12, 330−334.
(
(
(
1
0324
dx.doi.org/10.1021/jo2019637 | J. Org. Chem. 2011, 76, 10319−10324