Oxidative substitution of boranephosphonate diesters as a route to post-synthetically modified DNA

Article abstract of DOI:10.1021/ja511145h

The introduction of modifications into oligonucleotides is important for a large number of applications in the nucleic acids field. However, the method of solid-phase DNA synthesis presents significant challenges for incorporating many useful modifications that are unstable to the conditions for preparing synthetic DNA. Here we report that boranephosphonate diesters undergo facile nucleophilic substitution in a stereospecific manner upon activation by iodine. We have subsequently used this reactivity to post-synthetically introduce modifications including azides and fluorophores into DNA by first synthesizing boranephosphonate-linked 2′-deoxyoligonucleotides and then treating these oligomers with iodine and various nucleophiles. In addition, we show that this reaction is an attractive method for preparing stereodefined phosphorus-modified oligonucleotides. We have also examined the mechanism of this reaction and show that it proceeds via an iodophosphate intermediate. Beyond nucleic acids synthesis, due to the ubiquity of phosphate derivatives in natural compounds and therapeutics, this stereospecific reaction has many potential applications in organophosphorus chemistry.

Full text of DOI:10.1021/ja511145h

Subscriber access provided by GEORGIAN COURT UNIVERSITY
Article
Oxidative Substitutions of Boranephosphonate
Diesters as a Route to Post Synthetically Modified DNA
Sibasish Paul, Subhadeep Roy, Luca Monfregola, Shiying
Shang, Richard Shoemaker, and Marvin H. Caruthers
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja511145h • Publication Date (Web): 13 Feb 2015
Downloaded from http://pubs.acs.org on February 18, 2015
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.
Journal of the American Chemical Society 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 39
Journal of the American Chemical Society
1
2
3
4
5
Oxidative Substitutions of Boranephosphonate
6
7
8
Diesters as a Route to Post Synthetically Modified DNA
9
†
†
Sibasish Paul Subhadeep Roy , Luca Monfregola, Shiying Shang,
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
,
Richard Shoemaker and Marvin H. Caruthers*
Department of Chemistry and Biochemistry, University of Colorado, Boulder,
CO 80303 USA
ABSTRACT: The introduction of modifications into oligonucleotides is important for a
large number of applications in the nucleic acids field. However the method of solid
phase DNA synthesis presents significant challenges for incorporating many useful
modifications that are unstable to the conditions for preparing synthetic DNA. Here we
report that boranephosphonate diesters undergo facile nucleophilic substitution in a
stereospecific manner upon activation by iodine. We have subsequently used this
reactivity to introduce modifications including azides and fluorophores, postꢀ
synthetically into DNA by first synthesizing boranephosphonate linked 2’ꢀ
deoxyoligonucleotides followed by treatment of these oligomers with iodine and various
nucleophiles. In addition we show that this reaction is an attractive method for preparing
stereodefined phosphorous modified oligonucleotides. Further we have also examined
the mechanism of this reaction and show that it proceeds via an iodophosphate
intermediate. Beyond nucleic acid synthesis, due to the ubiquity of phosphate derivatives
in natural compounds and therapeutics, this stereospecific reaction has many potential
applications in organophosphorous chemistry.
INTRODUCTION
Chemically modified oligodeoxynucleotides1,2,3,4,5 (ODNs) are used for
applications in diagnostics6, biomedical research, therapeutics7, forensics8 and material
sciences.9,10,11 For therapeutic applications, the modifications are introduced to confer
properties such as enhanced nuclease stability, improved pharmacokinetics, stronger and
more selective binding to complementary DNA/RNA, and improved cellular uptake. On
the other hand the use of DNA based sensors, nanomaterials, diagnostic tools and
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 39
1
2
3
4
5
6
catalysts require the attachment of labels such as fluorophores12, quenchers, metal
complexes13 and EPR labels.14
7
8
9
The phosphoramidite approach15 is the most commonly used procedure for
chemical synthesis of DNA, RNA and their derivatives. This method requires the stepꢀ
wise treatment of the solid support linked, growing oligonucleotide with the following
reagents: (i) a strong acid, (ii) coupling with a phosphoramidite synthon to generate a
phosphite triester, and (iii) oxidation of the phosphite to phosphate. Treatment with a
strong base is required to remove the protecting groups from phosphate and the
nucleoside bases. Thus chemical modifications that are unstable towards acids and
strong bases or react with P (III) compounds lead to side reactions with resultant
diminishing yields of the desired product. Therefore a general method for introducing
modifications after completion of the synthesis would greatly expand the repertoire 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
Figure 1. (A) The reaction of diastereomerically
pure 9 or 10 with 30 equivalents of iodine and a 0.4
M solution of methylamine in tetrahydrofuran leads
to formation of methylphosphoramidates 11 or 12 in
a stereospecific manner. (B) and (C) show the 31P
NMR spectrum of the crude reaction mixtures
obtained from 9 and 10 as starting materials
respectively.
ACS Paragon Plus Environment

Page 3 of 39
Journal of the American Chemical Society
1
2
3
4
nucleic acids analogs that can be synthesized.
5
6
7
8
9
Here we report the use of a boranephosphonate diester when incorporated into the
DNA backbone (bpDNA)16,17 as a highly versatile analog that can be used to postꢀ
synthetically introduce modifications into oligodeoxynucleotides. The method is based
upon iodine activation of boranephosphonates towards substitution by nucleophiles
(including alcohols, thiols, primary and secondary amines as well as carbon
nucleophiles). Furthermore as this reaction occurs in a stereospecific manner, it is ideally
suited for the purpose of synthesizing chiral phosphate derivatives. As a proofꢀofꢀ
principle we show that this reaction allows the incorporation of stereospecific backbone
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
O
OR1
O
OR1
P
+ NHR'R'' + I2
P
H3B OR2
R''R'N OR2
31P NMR ! (ppm)
ESI+ (m/z)
Entry
NHR'R''
Theoretical
Observed
Fast isomer
Slow isomer
10.37 (13)
774.31a
774.33
10.43 (14)
NH3
1.
CH3NH2
Et2NH
9.97 (11)
9.49 (15)
8.03 (17)
819.41b
903.64b
867.40b
819.39
903.48
867.38
10.28 (12)
10.15 (16)
2.
3.
NH2
8.56 (18)
4.
5.
6.
931.50b
850.40a
931.44
850.36
O
NH
6.92 (19)
2.14 (21)
7.39 (20)
2.62 (22)
NH2
957.62b
868.54a
957.45
868.39
8.89 (23)
8.76 (25)
9.32 (24)
8.59 (26)
7.
8.
N3
NH2
NH2
N
H
NH
NH2.HCl
9.
ClH.H2N
N
945.60a
945.43
6.75 (28)
8.49 (30)
6.57 (27)
8.06 (29)
H
CO2Me
NH2
1062.61 b
1062.60
10.
Table 1. Iodine activated substitution of boranephosphonate dimers (9 and 10) by
amines. Reactions were performed in acetonitrile using 0.01 mmol of the
boranephosphonate dimer, 3.0 equivalents of iodine and 10.0 equivalents of the
amine. Diisopropylethylamine (20 equivalents) was also added for entry 9. The mass
of the positively charged adduct of the phosphoramidates and the corresponding
a
b
ammonium species ( ) or the mass of the protonated phosphoramidate ( ). R1: 5’ꢀ
(tertꢀbutyldimethylsilyl)ꢀ2’ꢀdeoxythymidyl;
R2:
3’ꢀ(tertꢀbutyldimethylsilyl)ꢀ2’ꢀ
deoxythymidyl.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 39
1
2
3
4
modifications into 2’ꢀdeoxyoligonucleotides.
5
6
7
8
9
We also note that because this chemistry is orthogonal to other methods of DNA
modification such as copper activated click reactions18 or palladium catalyzed Stille
couplings19, it opens avenues for labelling DNA with multiple functional groups and
fluorophores. Finally, although the focus of the present report is towards modification of
oligodeoxynucleotides, this reaction is of broader interest in fields involving the synthesis
of stereodefined phosphate containing compounds such as glycophosphate
analogs20,21,22,23 and nucleoside triphosphates.24
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
RESULTS
Substitution Reactions of Boranephosphonate Dimers Activated by Iodine
We have previously reported that boranephosphonate diesters reduce strongly
oxidizing metal ions such as Au (III), Pt (II) and Ag (I) and produce metallic
nanoparticles.25 Concomitantly the boranephosphonate diester undergoes nucleophilic
substitution by the solvent (water or methanol) to generate the phosphate diester or
triester respectively. These observations hinted at a useful pathway to oxidatively
activate the highly stable boranephosphonate internucleotide linkage towards
nucleophilic substitution reactions. However the use of metal ions for carrying out this
process with a broad range of nucleophiles proved to be problematic due to the crossꢀ
reactivity of Au (III) with many organic functional groups such as amines or thiols and
also because the formation of nanoparticles in the reaction rendered workꢀup and
purification difficult. These challenges therefore led us to explore alternate oxidizing
agents that would similarly activate a boranephosphonate towards substitution. In this
regard, the instability of boranephosphonates towards aqueous iodine solutions has been
observed in the past26 but there have not been any investigations into the products formed
from this reaction.
boranephosphonates, we chose to study the reaction with diastereomerically pure 5’ꢀ(tertꢀ
butyldimethylsilyl)ꢀ2’ꢀdeoxythymidyl, 3’ꢀ(tertꢀbutyldimethylsilyl)ꢀ2’ꢀdeoxythymidyl
In order to investigate whether iodine could activate
boranephosphonates (9 and 10) as model substrates. (Scheme S1; details of synthesis and
ACS Paragon Plus Environment

Page 5 of 39
Journal of the American Chemical Society
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
Table 2. Iodine activated substitution of boranephosphonate dimers 9 and 10 by alcohols, thiols and
CH3MgBr. a Denotes [M+Na]+ and b denotes [M+(iꢀPr)2EtNH]+ ions. All the reactions were performed
in acetonitrile with 0.01 mmol of 9 or 10 using acetonitrile as solvent; 3.0 equiv of iodine, 5.0 equiv of
DIPEA and 10.0 equiv of the nucleophile. For entries 3 and 4 the corresponding alcohol was the
solvent. 20.0 equiv. of nucleophiles were used for entries 6 and 7. R1: 5’ꢀ(tertꢀbutyldimethylsilyl)ꢀ2’ꢀ
deoxythymidyl; R2: 3’ꢀ (tertꢀbutyldimethylsilyl)ꢀ2’ꢀdeoxythymidyl
purification are provided in the supplemental section). We refer to 9 and 10 as the fast
and slow isomers respectively based upon their mobility on a silica column.
Initially we reacted each diastereomer with a thirtyꢀfold excess of iodine
dissolved in an anhydrous solution of 0.4 M methylamine and THF (Figure 1A). The 31P
NMR of the crude reaction mixtures (Figures 1B and 1C) showed formation of single
peaks at 9.97 and 10.28 ppm when using 9 and 10 as starting materials respectively. The
resonances observed in each case were distinct which demonstrated the formation of only
one diastereomer via a stereospecific reaction. The products were subsequently isolated
and analyzed by proton NMR and mass spectrometry (Table 1, Entry 2, also see
supplementary section). These results confirmed the formation of the phosphoramidates
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 39
1
2
3
4
5
6
7
8
9
(11 and 12) in accordance with our hypothesis of iodine activated, nucleophilic
substitution of the BH3 group. We also varied the amount of iodine added to these
reaction mixtures, while keeping an excess of the methylamine (Figure S1). The results
from these experiments indicated that the complete conversion of boranephosphonate to
the phosphoramidate required one equivalent of iodine. This was an interesting
observation as it points to the fact that it is not necessary to oxidize all three hydrides on
borane in order to convert it into a good leaving group.
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
Subsequently we carried out the reaction of 9 and 10 with several different types
of amines. The amines were added at a tenꢀfold excess over 9 and 10. Although only an
equimolar amount of iodine was needed with respect to the boranephosphonate, three
equivalents were used for these NMR scale experiments in order to avoid variations in
results due to random experimental error. Reactions with ammonia (Table 1, Entry 1), a
primary amine (Table 1, Entry 2, methyl amine), secondary amines (Table 1, Entries 3
and 5, diethyl amine, morpholine) and an aromatic amine (Table 1, Entry 6, aniline)
occurred quantitatively under these conditions (as estimated by 31P NMR). Propargyl
amine and azido propanamine were also tested because they are known as “click”
counterparts (Entries 4 and 7). Reactions with azides are particularly noteworthy as they
are prone to undergo Staudinger reactions with P (III) compounds and cannot be
incorporated using current methods of DNA synthesis. We were also able to successfully
introduce arginine (Table 1, Entry 9). This observation suggests that amino acids and
peptides can be successfully linked to DNA as well. While pyrrole did not react under
these conditions, 2ꢀ(2ꢀaminoethyl)pyrrole readily formed the phosphoramidates (Table 1,
Entry 8). The efficacy of the reaction with amines was further demonstrated by
quantitative conversion of the hindered camphenylamine (Table 1, Entry 10, also see
Figure S2 and S3 for representative 31P NMR spectra) to the corresponding amidates.
Similarly, when these dimers were treated with a solution of iodine in methanol,
the expected methyl phosphotriesters were formed quantitatively based on 31P NMR
(Figure S4). During the reaction with methanol we found that the 5´ꢀsilyl protecting
group (tertꢀbutyldimethylsilyl, TBDMS) was removed. Presumably the production of
hydroiodic acid from iodine oxidation of the BH3 hydrides causes silyl ethers to be
ACS Paragon Plus Environment

Page 7 of 39
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
activated towards nucleophilic substitution (by methanol in this case). In contrast to the
efficient reactivity with amines and methanol, we found that the reaction of 9 and 10 with
nꢀpropanol (either as the solvent or in 10 fold excess with THF as the solvent) did not
occur. However, addition of a tertiary amine (diisopropylethylamine; DIPEA) to
scavenge hydroiodic acid, led to the desired nꢀpropylphosphate triesters as determined by
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
31
observing a sharp singlet in the P NMR at ꢀ1.96 ppm and ꢀ1.88 ppm when using 9 and
10 respectively. The formation of the correct product was further ascertained by ESIꢀMS
(Table 2, Entry 1). In these experiments we found that the order of addition of the tertiary
amine was important. The presence of DIPEA prior to addition of iodine prevented any
reaction. This was likely due to coordination of iodine with the tertiary amine. Under
optimized conditions, with DIPEA added following iodine oxidation, the formation of
phosphotriesters could be accomplished with several alcohols including 2,2,2ꢀ
trifluoroethanol and 2ꢀpropanol (Table 2, Entries 2 and 3). As with amines, the reaction
with alcohols was also useful for introducing groups such as alkynes and azides (Table 2,
Entries 4 and 5).
Finally, propanethiol (Table 2, Entry 6) and methyl magnesium bromide (Table 2;
Entry 7) were also found to be effective nucleophiles for testing the synthesis of
phophorothioate esters and alkylphosphonates respectively. As these nucleophiles are
reactive towards iodine, the reactions in these cases were carried out in two steps.
Scheme 1. Postꢀsynthetic modification of 2’ꢀdeoxyoligonucleotides by iodine activated substitution of
bpDNA oligomers.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 39
1
2
3
4
5
6
7
8
9
Initially three equivalents of iodine were added to the boranephosphonate solution. After
stirring this mixture for 5 minutes, an excess (20 equivalents) of the nucleophile and 5.0
equivalents of the base (DIPEA) were added. The excess thiol or Grignard reagent
neutralized any unreacted iodine present in the reaction mixture. The precipitates
generated from these side reactions were removed by filtration or an aqueous work up.
Subsequent analysis revealed formation of the desired products.
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
Phosphoramidate DNA Synthesis from Boranephosphonate Oligomers.
The highly efficient reaction of boranephosphonate dinucleotides with amines
encouraged us to explore the use of bpDNA as a substrate for generating modified
oligodeoxynucleotides having phosphoramidates at predefined positions within DNA.
When coupled with the synthesis of bpDNA, we expected that these studies would lead to
a methodology for introducing postꢀsynthetically a large number of modifications. The
approach (Scheme 1) was as follows: 1.) Synthesize 2’ꢀdeoxyoligonucleotides having
boranephosphonates at specific sites. 2.) Remove all protecting groups from
oligodeoxynucleotides but retain attachment to the support. 3.) Introduce
phosphoramidates at sites that contain boranephosphonate linkages. 4.) Use a mild, brief
treatment to hydrolyze the succinate linkage that attaches the ODN to the solid support.
Our current method for synthesizing boranephosphonate DNA27 proved to be particularly
attractive for developing this approach as it relies upon using the diꢀtertꢀbutylisobutylsilyl
group (BIBS) for protection of the nucleobase exocyclic amines. Following synthesis,
the methyl protection on phosphorous is removed using disodiumꢀ2ꢀcarbamoylꢀ2ꢀ
cyanoethyleneꢀ1, 1ꢀdithiolate and BIBS protection on the nucleobases is eliminated with
a fluoride ion solution buffered to neutral pH. At this point, ODNs free of protecting
groups remain joined to the support.
Following siteꢀspecific modification of
boranephosphonate linkages using iodine and a nucleophile, the ODN can be released
from the support by a 30 min treatment with aqueous ammonia (or other similarly mild
conditions such as a 1 hour treatment with ethylenediamine). In order to use even milder
conditions for removal of ODNs from the supports, both the Q and oxalyl linkers will be
tested. Thus modifications are introduced by an orthogonal approach and are not exposed
to reagents of DNA synthesis or prolonged ammonia treatment.
ACS Paragon Plus Environment

Page 9 of 39
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Mol. Wt.
(Obs)
Mol. Wt.
(theor)
ODN No.
Sequence
Amine
NH3
ODN 1 T*TTTTT
ODN 2 T*TTTTT
ODN 3 T*TTTTT
ODN 4 T*TTTTT
1761.326
1817.381
1799.336
1844.372
1761.336
1817.399
1799.352
1844.385
(CH3CH2)2NH
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
ODN 5 T*TTTTT
1897.451
1897.462
ODN 6 T*TTTTTTTTTTTTTTTTTTT
ODN 7 T*TTTTTTTTTTTTTTTTTTT
ODN 8 TTTTTTTTTTTTTTTTTTT*T
6060.019
6058.902
6058.89
6061.02
6058.003
6058.003
ODN 9 T*TTTTTTTT*TTTTTTTTTT*T
ODN 10 T*TTTTTTTTTTTTTTTTTTT
7018.48
7018.64
6354.073
6353.178
ODN 11 TTTTTTTTTTTTTTTTTTT*T
ODN 12 T*TTTTTTTTT*TTTTTTTTT*T
ODN 13 T*T*T*TTTTTTTTTTTTTT*T*T*T
6103.928
6018.929
6016.001
6319.038
6103.036
6018.019
6015.068
6318.13
NH3
NH3
NH3
ODN 14 T*TTTT*TTT*TTT*TTT*TTT*TTT*T
ODN 15 TGTAAACCATG*ATGTGCTGCTAT
ODN 16 TGTAAACCATG*ATGTGCTGCTAT
7090.266
7135.16
7091.7
7136.733
7091.701
7136.733
7050.709
7045.789
ODN 17 T*GTAAACCATGATGTGCTGCTAT
ODN 18 T*GTAAACCATGATGTGCTGCTAT
ODN 19 T*GTAAACCATG*ATGTGCTGCTA*T
7091.175
7136.169
7051.108
7046.239
NH3
NH3
ODN 20
T* GTA* AAC* CAT* GAT* GTG* CTG* CTA* T
T*T*T*T*T*T*T*T*T*T*T*T*T*T*T*T*T*
T*T*T
ODN 21
NH3
6002.56
6003.18
Table 3. 2’ꢀdeoxyoligonucleotides containing boranephosphonate diester linkages (denoted by *) were
synthesized and reacted with various amines to produce the corresponding phosphoramidate
oligodeoxynucleotides. For ODN 21 the mass spectra was recorded by MALDIꢀTOF mass
spectrometry, whereas LCꢀMS was used for the other ODNs.
In order to carry out a systematic evaluation of this methodology, bpDNAs
containing boranephosphonate diester internucleotide linkages at defined positions were
synthesized (Table 3) and protecting groups removed. The support linked 2’ꢀ
deoxyoligonucleotides were then reacted with a solution of 0.05 M iodine and various
amines at a concentration of 0.2 M in anhydrous acetonitrile for 48 hrs. The resins were
repeatedly washed with acetonitrile and the ODNs cleaved using 30% NH4OH (30
minutes at room temperature). After removal of ammonia under reduced pressure, the
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 39
1
2
3
4
5
6
31
oligomers were dissolved in water and analyzed by LCꢀMS and P NMR (Table 3 and
Figure S5ꢀS25).
7
8
9
Initially we treated a resin bound 2’ꢀdeoxyhexathymidine containing a single
boranephosphonate diester linkage adjacent to the 5’ end with a solution of 0.05 M I2 and
0.4 M ammonia in dioxane (ODN 1, Table 3). Upon cleavage and LCꢀMS analysis, we
found that the desired oligomer having one phosphoramidate linkage was produced in
quantitative yield. Similarly, the reaction proceeded smoothly with diethylamine and the
highly hindered isopinocamphenyl amine (Table 3, ODNs 2 and 5 respectively). In order
to demonstrate the introduction of other functional groups (alkyne and azide) into DNA,
the reaction was carried out with propargyl amine (Table 3, ODN 3) and 3ꢀazidoꢀ1ꢀ
propanamine (Table 3, ODN 4). These experiments also demonstrated that none of the
amines led to measurable cleavage of the succinate linkage and loss of product. This
conclusion was based upon the observation that the yields of the crude phosphoramidate
containing oligodeoxynucleotides were close to the amount expected from synthesis on a
0.2 µmol scale.
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
In order to evaluate the effect of the position of the boranephosphonate linkage
within the oligomer towards substitution by amines, we carried out similar reactions on
20ꢀmers containing a single boranephosphonate diester linkage adjacent to either the 5’ or
3’ ends (Table 3, ODN 6, 7, 8, 10 and 11). Substitution with propargyl amine, 3ꢀazidoꢀ1ꢀ
propanamine, azidoꢀPEGꢀamine or 2ꢀmethylaziridine showed that conversion to the
phosphoramidate occurred quantitatively at the 5’ or the 3’ end. Thus the solid support
matrix does not appear to interfere with this substitution reaction even when the
boranephosphonate is located at the first internucleotide linkage adjacent to the support.
We next synthesized bpDNA containing three, six or eight, boranephosphonate
diester linkages as well as a fully boronated 20ꢀmer bpDNA sequence (Table 3, ODN 12,
13, 14 and 21 respectively) in order to determine whether multiple modifications could be
introduced within the same ODN. Substitution of borane with ammonia upon activation
by iodine proceeded efficiently and analysis of the crude reaction mixtures by both LCꢀ
31
MS (Figures S16ꢀS18) and P NMR (Figures S25 (A), (B), (C), and (D)) revealed that
the desired phosphoramidates were formed in near quantitative yields. We also reacted
ACS Paragon Plus Environment

Page 11 of 39
Journal of the American Chemical Society
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 2. LCꢀMS and denaturing gel electrophoresis analysis of mixed sequence oligonucleotides
(ODN 15ꢀ20). Each set of data shows the LC trace along with a zoomed in view of the base peak in the
mass spectra. The base peak in each case was a ꢀ4 charged species. The ODN numbers are indicated at
the top of each lane in the gel image. Full mass spectra are provided in the supplementary section.
azidoꢀPEGꢀamine and iodine with
a
bpDNA 20mer that contained three
boranephosphonate diester linkages (one each at the 5’ and 3’ ends and one internally)
(Table 3, ODN 9). Three products were obtained. This was in contrast to the reaction of
ODN 10 (one boranephosphonate at the 5’ꢀ linkage) where conversion to the
phosphoramidate with azidoꢀPEGꢀamine was quantitative. Based on LCꢀMS data, the
product that was present in approximately 50% yield contained three azidoꢀPEGꢀamine
groups. In addition, two products having one or two azidoꢀPEGꢀamines were also
detected (Figure S12). Use of higher equivalents of the amine (up to 2.0 M) or iodine
(1.0 M), longer reaction times (96 hrs), carrying out the reaction at elevated temperatures
o
(up to 55 C) and alternate solvents (dichloromethane, dimethylformamide) did not lead
to a significant improvement in yield of ODN 9 having three azidoꢀPEGꢀamine
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 39
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. A DNA oligomer (5’ꢀ T*GTAAACCATGATGTGCTGCTAT ꢀ3’; where * denotes a
boranephosphonate diester linkage) was reacted with iodine (0.05 M and 0.2M aminoꢀdansyl, 48h,
room temperature) as described above, released from the solid support (30% NH4OH, 30 min, room
temperature) and analyzed by LCꢀMS. The peak at 24.5 minutes corresponded to the dansyl
conjugated product, whereas the peak at 23.73 min was due to conversion of the boranephosphonate to
the phosphate. The inset shows expansion of the relevant regions of the mass spectra extracted from
these peaks. The theoretical values for the ꢀ4 charged peaks for these oligomers are 1841.8 and 1762.4
respectively.
phosphoramidates. It is possible that conjugation of one hydrophobic group causes the
polystyrene bound ODN to adhere to the solid matrix and thereby slowing down further
reactions.
We also synthesized 2’ꢀdeoxyoligonucleotides containing all four nucleobases
with variable locations and numbers of boranephosphonate diester linkages. Consistent
with previous results, propargyl amine and 3ꢀazidopropanamine reacted with
boranephosphonate diesters located internally as well as at the 5’ end of the
oligodeoxynucleotide (ODNs 15ꢀ18). LCꢀMS analysis of the unpurified mixtures
demonstrated formation of the desired phosphoramidates in high yields (Figure 2 and
Figures S19ꢀ22). Mixed base ODNs having ammonia substituted for borane at up to
eight sites in bpDNA could be synthesized without any detectable “failure” products
(Table 3 entries 19 and 20, Figure 2 and Figures S23 and S24). This is evident from the
31P NMR of the unpurified ODN 20 (Figure S25 (C)). The unpurified reaction mixtures
for the mixed base modified oligomers (ODNs 15ꢀ20) were also analyzed by denaturing
gel electrophoresis. The only significant band in each case was the expected product
(Figure 2).
ACS Paragon Plus Environment

Page 13 of 39
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Finally we tested whether this method could be used to attach common
functionalities such as a fluorophores to DNA oligomers. We carried out the reaction of
an amino derivative of the dansyl dye with a 23ꢀmer mixed based oligodeoxynucleotide
(see Figure 3) in which the last internucleotide linkage adjacent to the 5’ end was
boranephosphonate. The desired product was obtained in 40% yield. Internucleotide
phosphates are typically inaccessible and do not efficiently undergo reactions. Thus
fluorophores and other groups are usually conjugated to the 5’ hydroxyl or through a
linker attached to the C5 position of pyrimidines. In this light, the reactivity observed
here is significant as probes can be potentially introduced throughout the oligomer at any
internucleotide linkage. Many useful fluorophores such as the cyanine family of dyes are
sensitive to strong base and therefore must be incorporated via solution phase conjugation
reactions after cleavage of the oligomers from the solid support. This adds extra
synthetic and purifications steps. Using the present chemistry, fluorophores would be
introduced after removal of all protecting groups and need to survive only a very mild
base for cleavage of modified oligomers from the support.
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
Mechanism of Iodine Activated Substitution of Boranephosphonate Diesters
Iodine
activated
substitution of the borane
group, when carried out using
dieastereomerically pure 9 or
10, yielded a single peak in the
31P NMR (Figure 1) that
indicates
a
stereospecific
reaction pathway. These results
are significant as synthesis of
stereoꢀdefined
phosphorous
containing compounds such as
oligodeoxynucleotide
Figure 4. Characterization of the iodophosphate intermediate
45. 31P NMR spectrum in 2,2,2ꢀtrifluoroethanol shows a peak
at ꢀ43.24 ppm with 9 as the starting material. Inset: ESIꢀMS
(+ mode) shows that the recorded isotope distribution (top)
matches the theoretical isotope distribution (bottom) of 45.
Theoretical mass: 656.0380, Observed: 656.0371
derivatives, phosphotones and
other phosphate containing
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 39
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
Scheme 2. Mechanism of iodine activated substitution of boranephosphonate diesters.
therapeutics are challenging. These factors therefore motivated us to further investigate
the potential mechanism of this reaction. Similar to analogous reactions of amino
boranes28 we expected the present reaction to begin via oxidation of the borane hydrides
by iodine. Moreover based on our observation that one equivalent of iodine is sufficient
to activate the phosphorous center towards nucleophilic attack, an active intermediate of
the type (OR1)(OR2)(PO)(BH2I) (44; Scheme 2) was hypothesized.
When 9 reacted with three equivalents of iodine in acetonitrile, we observed the
formation of a precipitate after approximately ten minutes of stirring. Concomitantly a
31
loss in P NMR signal in the supernatant was observed, while addition of methylamine
to the precipitate produced the methyamino phosphoramidate. This result indicated that
the precipitate was likely the sought after reactive intermediate. Initial attempts to
characterize the precipitate were unsuccessful as it was found to be insoluble in nonꢀpolar
solvents such as dichloromethane, toluene or ethylacetate while in polar, non nucleophilic
solvents such as DMSO, DMF and pyridine, various adducts (based on 31P NMR spectra)
that could not be characterized were produced (data not shown). During the course of
investigating the reaction of boranephosphonate dimers with iodine and alcohols as
nucleophiles (Table 2), we had found that the reaction with 2,2,2ꢀtrifluoroethanol, which
is a polar solvent but a poor nucleophile, was quite slow. Letting this reaction mixture
stand overnight did lead to the 2,2,2ꢀtrifluoroethylphosphate triester. This result
indicated that the reactive intermediate was stable in this solvent for a long period of
time. Consistent with these observations we dissolved the precipitate in 2,2,2ꢀ
trifluoroethanol and added a few drops (approximately 50 µL) of CD2Cl2 as a lock
31
solvent for NMR. The P NMR of this solution revealed a single resonance at ꢀ43 ppm
(Figure 4). Although produced via a different pathway, a previous report from our
ACS Paragon Plus Environment

Page 15 of 39
Journal of the American Chemical Society
1
2
laboratory,29 has described the formation of a reactive species having a similar P NMR
resonance that was postulated to be the iodophosphate intermediate. We also analyzed a
solution of the precipitate dissolved in 2,2,2ꢀtrifluoroethanol by diluting it with
anhydrous acetonitrile and directly infusing this solution into an ESIꢀMS (Figure 4). The
resulting mass spectrum showed a high degree of correspondence between the theoretical
and observed isotopic distribution for an iodophosphate (45) (Figure 4 inset). We noted
that as per the mass spectrum, the precipitated product no longer contained the TBDMS
protecting groups. Loss of TBDMS is likely caused by acid activation by the hydroiodic
acid produced during oxidation of the hydrides of the borane group by iodine. The loss
of the hydrophobic TBDMS groups is consistent with this species precipitating from the
solution.
3
31
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
Thus we conjecture that the initial reaction proceeds by oxidation of a hydride on
the borane group to produce an iodoboranephosphonate intermediate (44, Scheme 2).
Subsequently nucleophilic attack by iodide generates the iodophosphate (45). A second
substitution reaction occurs at phosphorous through displacement of iodide with a
nucleophile. Nucleophilic substitutions at phosphorous proceed via inversion of
configuration.30 Therefore this reaction pathway, with two successive inversions of
configuration, implies that formation of the final product occurs with an overall retention
of configuration at phosphorous. These observations are supported by comparing the
NOESY spectra (Figure S26) of the starting boranephosphonate diastereomers (9 and 10)
with the methyphosphotriesters that are produced from each isomer (47 and 48
respectively). 9 showed much stronger NOE peaks between the BH3 hydrogens and the
hydrogens on the ribose moiety when compared to 10. Similarly 46 (the product from 9)
also showed stronger NOE signals between the methyl hydrogens and the sugar
hydrogens than 47. These results support the conclusion that the BH3 group of 9 and the
OCH3 group of 46 are in the same relative orientation at phosphorous.
Stereospecific Synthesis of 2’-Deoxyoligonucleotide Analogs.
Finally, the stereospecific nature of this reaction encouraged us to investigate
whether the reactivity of boranephosphonates could be used to introduce stereospecific
internucleotide phosphate linkages within 2’ꢀdeoxyoligonucleotides. We carried out the
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 39
1
2
3
4
31P NMR (ppm)
5
6
7
8
fast
slow
TT
CT
AT
GT
94.97 (63)
93.38 (65)
93.75 (67)
93.90 (69)
93.74 (64)
94.54 (66)
94.71 (68)
95.11 (70)
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
Table 4. 31P NMR chemical shifts corresponding
to the boranephosphonate moiety in the dimer
phosphoramidites. The compound numbers are
indicated in parenthesis.
synthesis of diastereomerically pure 5’ꢀOꢀ
dimethoxytrityl
dideoxynucleotide
boranephosphonate
3’ꢀphosphoramidites
d(TT), d(CT), d(AT) and d(GT) (63-70,
Scheme S2). Detailed synthetic procedure
and methods for separation are provided in the
supplementary information. Briefly, after the
tetrazole mediate coupling of 5’ꢀOꢀ
dimethoxytritylꢀNꢀBIBS
protected
2’ꢀ
deoxynucleoside
3’ꢀOꢀcyanoethylꢀN,Nꢀ
diisopropylphosphoramidites with 3’ꢀOꢀacetyl
2’ꢀdeoxythymidine, the resulting triesters
were boronated, the cyanoethyl groups
Figure 5. 31P NMR of purified ODN 22 and ODN
23.
removed and the two diasteromers were separated via silica gel column chromatography.
Subsequently, the 3’ꢀacetyl was removed and the resulting free hydroxyl phosphitylated
to yield the desired dimer phosphoramidites. Comparison of the 31P NMR of the
diastereomeric pairs revealed distinct chemical shifts in each case (Table 4). These
studies demonstrate the feasibility of synthesizing and more importantly separating the
diastereomeric pairs containing any of the four nucleobases. In fact in our experience we
have found that the bulky BIBS group allows more facile separation of the diastereomers
by silica gel chromatography when compared to nucleobases containing standard amide
protecting groups such as benzoyl and acetyl.
ACS Paragon Plus Environment

Page 17 of 39
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Subsequently we carried out solid phase synthesis of 21ꢀmer 2’ꢀ
deoxyoligonucleotides where each ODN had one of the diastereomers (Table 4)
incorporated at four positions. These ODNs were then reacted with 0.05 M iodine and 0.4
M ammonia in dioxane to yield oligomers containing stereospecific phosphoramidate
internucleotide linkages (Table 5). The incorporation of the d(TT) and d(CT) dimers
occurred in high yields, although longer coupling time of 15 minutes were required.
Unfortunately, the d(AT) and d(GT) dimers did not couple efficiently under these
conditions and in amounts sufficient for further studies. Future research will be carried
out to optimize conditions that will allow incorporation of these compounds as well.
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 oligomer pairs incorporating the d(TT) or d(CT) dimers (ODN 22 from 63
and ODN 23 from 64 or ODN 24 from 65, ODN 25 from 66 respectively) were purified
by reverse phase HPLC and analyzed by LCꢀMS (Figures S27ꢀ S30). The 31P NMR
(Figure 5) of ODN 22 and ODN 23 shows efficient conversion of the
boranephosphonates to amidates. Interestingly we found that multiple batches of ODN
22 showed the formation of a broader phosphoramidate peak (at 12 ppm) when compared
to the corresponding peak for ODN 23. Moreover this oligomer also proved to be
challenging to purify by HPLC as evidenced by a higher than expected integration value
of the phosphate peak (Figure 5). In contrast ODN 22 yielded the correct 1:4 ratio of the
areas under the phosphoramidate:phosphate peaks. These observed differences may be
ODN No.
Sequence
Dimer Used
Tm
TTTTTTTT*TT*TT*TT*TTTTTTT
TTTTTTTT*TT*TT*TT*TTTTTTT
ODN 22
ODN 23
63
64
65
66
44.1 ± 0.6
45.8 ± 0.5
57.4 ± 0.01
55.4 ± 1.0
ODN 24 TTTTTTTC*TC*TC*TC*TTTTTTT
ODN 25 TTTTTTTC*TC*TC*TC*TTTTTTT
ODN 26 TTTTTTTTTTTTTTTTTTTTT
ꢀ
ꢀ
52.87 ± 0.8
64.5 ± 0.4
TTTTTTTCTCTCTCTTTTTTT
ODN 27
Table 5. 2’ꢀdeoxyoligonucleotides containing stereospecific phosphoramidate linkages (denoted by
*
) synthesized starting from stereopure boranephosphonate dimer phosphoramidites. Melting
temperatures were measured for binding to a complementary RNA sequence. ODN 26 and ODN 27
correspond to the unmodified DNA sequences.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 39
1
2
3
4
5
6
explained by different conformations adopted by oligomers containing distinct
diastereomeric phosphoramidates.
7
8
9
Further confirmation that the reaction for converting the boranephosphonate to
phosphoramidate occurs stereospecifically was obtained by measuring the melting
temperature of these 2’ꢀdeoxyoligonucleotides with complementary RNA sequences. As
expected in each case one of the oligomers was found to bind more strongly to the
complementary sequence than its diastereomeric pair (Table 5).
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
DISCUSSION
Our discovery that the stable boranephosphonate internucleotide linkage is
activated by iodine oxidation toward substitution reactions with various nucleophiles has
created several possible avenues for using this new synthetic pathway not only in the
nucleic acids field but other areas as well. This manuscript outlines a plausible
mechanism for this reaction and also presents initial results illustrating how we can
utilize the boranephosphonate analog for various applications.
The high stability of boranephosphonates is based upon the fact that the borane
moiety is a poor leaving group due to its low apicophilicity. It is well established that
during substitutions at phosphorous centers, the incoming and leaving groups need to be
in the apical positions of the trigonal bipyramidal intermediate31. Apicophilicity is
positively correlated with electronegativity.32,33 Thus theoretical calculations have shown
that reaction intermediates with borane at the apical position are high energy and
unfavorable. Additionally, comparisons of boranephosphonates with analogous aminoꢀ
boranes, which are relatively unstable and readily hydrolyze in protic solvents such as
methanol or water, suggest that the PꢀBH3 bond is significantly stronger than NꢀBH3.
The added stability of this interaction is probably due to back donation of electrons from
the σ orbitals of the borane to empty (perhaps σ *)34 orbitals on phosphorous. In the
present reaction the initial oxidation of one or more of the hydrides on borane leads to the
formation of an iodoborane derivative (such as 44) that likely causes disruption of such
bonding interactions. Concomitantly the electronegative iodine atom makes the borane
ACS Paragon Plus Environment

Page 19 of 39
Journal of the American Chemical Society
1
2
3
4
5
6
moiety more apicophilic and thus a better leaving group. The overall effect is to cause
substitution at phosphorous to proceed very efficiently.
7
8
9
Wada and coworkers have reported a separate reaction pathway35,36 whereby
boranephosphonates react with trityl cations to produce an Hꢀphosphonate – a P (III)
compound. The present reaction however clearly proceeds through a P (V) intermediate.
Taken together these complementary reactivities illustrate that in spite of their “P (V)
like” stability, boranephosphonates represents an intermediate oxidation state allowing
access to both P (III) as well as P (V) compounds depending on the reaction conditions
used. We hope that such unparalleled versatility of these synthons will find broad usage
in organophosphorous reaction schemes as well as further studies into gaining better
understanding of the structure and bonding in these compounds.
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 synthetic methodology for preparing bpDNA oligomers we have developed
previously27 has a twoꢀfold benefit in enabling the versatile method of DNA modification
described here. Firstly the ability to prepare 25ꢀ30mer bpDNA, having all four nucleoside
bases, in essentially quantitative yield far exceeds any previously reported
procedure16,37,38,39,40,41. Thus we are able to produce modified oligomers derived from
boranephosphonates of lengths that are necessary for applications such as DNA based
materials, antisense agents and sensors. Secondly, the diꢀtertꢀbutylisobutylsilyl group
used to protect the exocyclic amines is removed by treatment with mild, buffered fluoride
solutions. Thus by introducing the modifications through substitution of
boranephosphonates at the end of the synthetic cycle and after removal of nucleobase
protecting groups, there is no need to expose the final ODN product to either strong acids
or bases as is the case in all previously outlined chemistries42,43,44
.
Here we illustrate the strength of this approach through the synthesis of ODNs
having phosphoramidate internucleotide linkages. However based on experiments with
boranephosphonate dimers we anticipate that this approach will also apply to the
synthesis of ODNs having alkyl/aryl phosphonate and phosphate triester linkages as well.
The approach would be the sameꢀcomplete synthesis and removal of protecting groups
prior to conversion of the boranephosphonate linkages to phosphonates or triesters. Thus
the synthesis of ODNs having these linkages, which are highly susceptible to base and
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 39
1
2
3
4
5
6
7
8
9
fluoride, should now be possible in very high yields and with excellent purity. For
example previous research has shown that alkynyl phosphonates 45 (as clickable reagents
46,47
when reacted with azide to yield triazoylphosphonates), phosphonoacetates
and
48
phosphonoformates
all show biological and biochemical activity. However with
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
current chemistries such as the phosphoramidite approach, stepwise yields of 95%
severely limits their value. The approach as outlined here will generate very high yields
of boranephosphonate DNA that can then be converted via Grignard, Reformatski, or
alkyl/aryl lithium reagents to the corresponding phosphonates.
This reaction pathway also opens opportunities to generate new analogs that are
limited by current chemistries. Here we show that azide functionalities (which previously
could not be introduced according to the phosphoramidite approach) are accessible in
high yield simply by iodine oxidation of boranephosphonates following ODN synthesis.
Thus this pathway will be useful, via azides, for introducing fluorophores, EPR probes,
and other diagnostic reagents throughout an ODN at any predetermined internucleotide
linkage.
Oligonucleotides containing phosphorous modifications provide stability against
nucleases and therefore have been extensively used in oligonucleotide based therapeutics.
In these DNA analogs, the presence of a chiral phosphorous has significant implications
for their properties such as binding to complementary sequences. Although the field of
stereo defined oligonucleotide synthesis has advanced significantly, perhaps the most
relevant advances are the synthesis of stereo defined phosphorothioate DNAs and
boranephosphonate DNAs using monomer synthons 49,50,51, 52. For all these methods there
are several limitations such as cycle yields during synthesis of oligomers, difficult
purifications of isomers, and stereo excess. Here we use iodine activated substitution of
boranephosphonates with ammonia to produce stereodefined phosphoramidate
internucleotide linkages in 2’ꢀdeoxythymidine and 2’ꢀdeoxycytidine ODNs. The
approach involves first the synthesis in high yield of 5’ꢀOꢀdimethoxytritylꢀ2’ꢀ
nucleosidylꢀ3’ 3’ꢀOꢀacetylthymidylꢀ5’ boranephosphonates with sequences TT and TC.
These diastereomers are separated on silica gel, the acetyl group removed, the resulting
dimers converted to the 3’ꢀ phosphoramidites, and these synthons used in the preparation
ACS Paragon Plus Environment

Page 21 of 39
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
of 2’ꢀdeoxyoligonucleotides having four stereocontrolled internucleotide linkages
(21mers). As is the case with other dimer synthesis strategies, yields are relatively low
due to the slow reaction rates of these dimers. This is currently a limitation of our
research with stereo defined dimers of AT and GT. Although the yields of AT and GT
dimers as boranephosphonates are very good (see the supplementary experimental
section) and separation of the isomers is straightforward via silica gel chromatography,
the incorporation of these synthons into ODNs proceeds slowly at best and further
research with activators, more reactive 3’ꢀphosphoramidites, longer reaction times and
double coupling must be carried out. However given the efficient reaction of these
boranephosphonates dimers with iodine and a variety of nucleophiles, current work is
underway in our laboratory to use this reaction pathway to prepare various stereodefined,
Pꢀmodified oligomers such as phosphotriesters, phosphorthioates and alkylphosphonates.
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
Finally, the discovery of this new reaction pathway has stimulated us to
investigate several novel applications. For example we are currently attempting to use
this reaction sequence to prepare imidoamidate DNA (both nonꢀlinking valencies have
nitrogen substituted for oxygen) and as a method for templated chemical ligation of two
ODNs to produce longer ODNs linked via a natural phosphodiester bond. Another
research direction we are exploring is to use this chemistry to insert organometallic
reagents (most are impossible at this time because of the chemistries used to synthesize
DNA) at phosphorus and specifically complementary duplexes. We have also recently
demonstrated that natural internucleotide linkages and morpholino DNAs can be prepared
via iodine oxidation of boranephosphonate synthons.
CONCLUSIONS
Boranephosphonate diesters were shown to undergo efficient substitution
reactions upon oxidation of the BH3 hydrides by iodine. Reactions with amines, alcohols,
propanethiol and methylmagnesium bromide led to the formation in near quantitative
yields
of
phosphoramidates,
phosphotriesters,
phosphorothioates
and
methylphosphonates respectively. This reaction proceeds through an iodophosphate
intermediate in a stereospecific manner with an overall retention of configuration. We
have used this reaction to develop a method for the incorporation of modifications into
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 39
1
2
3
4
5
6
7
8
DNA in a postꢀsynthetic manner. Given the presence of phosphate groups in a large
number of biologically relevant molecules, the present reaction pathway has the potential
to open up many new avenues in synthetic organophosphorous chemistry.
9
EXPERIMENTAL SECTION
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 of Phosphoramidate, Phosphotriester, Phosphorothioate and
Alkylphosphonate Dinucleotides from Boranephosphonate Dinucleotides. The
boranephosphonate dinucleotide (0.01 mmol) (9 or 10) was added to a 10 mL pear
shaped flask stoppered with a rubber septum under an atmosphere of argon. Anhydrous
acetonitrile (1.0 mL) was added to the flask via a syringe. In a separate flask, iodine (0.03
mmol) was dissolved in acetonitrile (0.5 mL). For reactions listed in Table 1, the
nucleophile (0.1 mmol) was added to the flask containing the boranephosphonate diester
followed by addition of the iodine solution. For reactions in Table 2, the iodine solution
was first added and allowed to stir for approximately 1 minute followed by addition of
diisopropylethylamine. Immediately afterwards, the nucleophile was added. Reaction
mixtures were then allowed to stir overnight. Except for entries 6 and 7 in Table 2, an
aliquot of each reaction mixture was transferred to an NMR tube and 50 µL of CD2Cl2
was added as the lock solvent for recording the 31P NMR spectrum. For reactions with
propanethiol or methylmagnesium bromide (Table 2, Entries 6 and 7), a white precipitate
formed upon addition of the nucleophile. After stirring overnight, these reaction mixtures
were diluted with dichloromethane (5 mL) and extracted with water (twice, 5mL). The
organic layers were evaporated to dryness. The residues were dissolved in CD2Cl2 for
recording the 31P NMR spectra or in methanol for mass spectrometric analyses.
For mass spectrometric analysis, each reaction mixture was stirred overnight,
diluted with dichloromethane (5 mL), washed with aqueous solutions of sodium
thiosulfate (twice, 5 mL each), and once with water (5 mL). The organic layers were
diluted with methanol and infused into the ESIꢀMS.
Post-synthetic Modification of ODNs. ODNs containing boranephosphonate linkages
were synthesized as described previously27 at a 0.2 µmol scale on highly crossꢀlinked
polystyrene resin (Glen Research). Subsequently each resin linked oligodeoxynucleotide
ACS Paragon Plus Environment

Page 23 of 39
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
was treated with disodiumꢀ2ꢀcarbamoylꢀ2ꢀcyanoethyleneꢀ1,1ꢀdithiolate in N,Nꢀ
dimethylformamide (1.0 M) for 45 minutes in order to remove the methyl protecting
group on phosphorous, washed repeatedly with anhydrous DMF and acetonitrile and
dried under a stream of argon. The resins were then treated with a fluoride deprotection
mixture (940 µL DMF + 470 µL Et3N + 630 µL Et3N.(HF)3) overnight in order to
remove the silyl protecting groups on the 2’ꢀdeoxynucleoside bases. Subsequently the
resins were washed repeatedly first with DMF and then with anhydrous acetonitrile, dried
under a flow of argon and transferred to glass vials. A solution of iodine (0.05M) and
amine (0.2 M) in anhydrous acetonitrile was added to each glass vial. In the case of
reactions with ammonia, appropriate amounts of iodine were dissolved in solutions of
ammonia (0.4 M) in dioxane and these solutions were added to the resins. The glass vials
were then placed on a mechanical shaker for 48 hrs.
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
Vials containing the resins were centrifuged at 4000 rpm and the supernatants
removed with a pipette. Subsequently the resins were washed 4ꢀ5 times with 2 mL
aliquots of acetonitrile, shaken vigorously, placed in a centrifuge at 4000 rpm for 5
minutes and supernatants removed. The resins were then treated with 1 mL of 30%
ammonium hydroxide in water for 30 minutes and the ammonia was removed by
evaporation in a SpeedVac. The oligodeoxynucleotides were dissolved by adding 800 µL
of water. The polystyrene resin beads were then removed using a 0.2 µm centrifugal
filters. The filtrates were used for subsequent analyses.
ASSOCIATED CONTENT
Supporting Information
Detailed synthetic procedures, characterization information, LCꢀMS spectra of
oligomers synthesized as well as NOESY spectra are provided. This information is
available free of charge via the Internet at http://pubs.acs.org/
AUTHOR INFORMATION
Author Contributions
†
These authors contributed equally to this work.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 39
1
2
3
4
Corresponding Author
5
6
marvin.caruthers@colorado.edu
7
8
9
ACKNOWLEDGEMENTS
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
We thank Dr. Douglas J. Dellinger (Agilent Technologies) for providing access to
an LCꢀMS instrument. This research was supported by the University of Colorado.
REFERENCES
1. Singh, R.; Murat, P.; Defrancq, E. Chem. Soc Rev., 2010, 39, 2054ꢀ2070.
2. Juliano, R. L.; Ming, X.; Nakagawa, O. Acc. Chem. Res., 2012, 45, 1067ꢀ1076.
3. Marx, A.; Weisbrod, S. H. Chem. Commun., 2008, 5675ꢀ5685.
4. Cook, P. D.; Current Protocols in Nucleic Acid Chemistry, 2000, 4.1.1ꢀ4.1.17.
5. Singh, Y.; Murat, P.; Spinelli, N.; Defrancq, E., in Erdmann V. A.; Barciszewski,
J., editors. Nucleic Acids Sequences to Molecular Medicine: Springer Berlin
Heidelberg, 2012, 85ꢀ120.
6. Lee, J. H.; Yigit, M. V.; Mazumdar, D.; Lu, Y. Adv. Drug. Deliv. Rev. 2010, 62,
592ꢀ605.
7. Lönnberg, H. Bioconj. Chem. 2009, 20, 1065ꢀ1094.
8. Fascione, N.; Thorogate, R.; Daniel, B.; Jickells, S.; Analyst., 2011, 137, 508ꢀ512.
9. Jones, M. R.; Osberg, K. D.; Macfarlane, R. J.; Langille, M. R.; Mirkin, C. A.
Chem. Rev. 2011, 111, 3736ꢀ3827.
10. Wilner, O. I.; Willner, I. Chem. Rev., 2012, 112, 2528ꢀ2556.
11. Paul, S.; Jana, S.; Bhadra, J.; Sinha, S. Chem. Comm.2013, 49, 11278ꢀ11280.
12. (a) Stadler, A. L.; Delos Santos, J. O.; Stensrud, E. S.; Dembska, A.; Silva, G. L.;
Liu, S.; Shank, N.; KunttasꢀTatli, E.; Sobers, C. J.; Gramlich, P. M. E.; Carell, T.;
Peteanu, L. A.; McCartney, B. M.; Armitage, B. A. Bioconj. Chem., 2011, 22,
1491ꢀ1502. (b) Jager, A.; Levy, M. J.; Hecht S. M. Biochemistry, 1988, 27, 7237ꢀ
7246.
13. (a) Stewart, K. M.; McLaughlin, L. W. J. Am. Chem. Soc., 2004, 126, 2050ꢀ2057.
(b) Stewart, K. M.; Rojo, J.; McLaughlin, L. W. Angew. Chem. Int. Ed., 2004, 43,
5808ꢀ5811. (c) VargasꢀBaca, I.; Mitra, D., Zulyniak, H. J.; Banerjee, J.; Sleiman,
H. F. Angew. Chem. Int. Ed., 2001, 40, 4629ꢀ4632. (d) Yang, H.; Sleiman, H. F.
Angew. Chem. Int. Ed., 2008, 47, 2443ꢀ2446. (e) Ghosh, S.; Pignotꢀpaintrand, I.;
Dumy, P.; Defrancq, E. Org. Biomol. Chem., 2009, 2729ꢀ2737.
14. Qin, P. Z.; Dieckmann, T. Curr. Opin. Struct. Biol., 2004, 14, 350ꢀ359.
15. Beaucage, S. L.; Caruthers, M. H. Tet. Lett., 1981, 22, 1859ꢀ1862.
16. Li, P.; Sergueeva, Z. A.; Dobrikov, M.; Shaw, B. R. Chem. Rev. 2007, 107, 4746ꢀ
4796.
ACS Paragon Plus Environment

Page 25 of 39
Journal of the American Chemical Society
1
2
3
4
5
17. Sood, A.; Shaw, B. R.; Spielvogel, B. F. J. Am. Chem. Soc. 1990, 112, 9000ꢀ
9001.
6
7
8
9
18. (a) Gierlich, J.; Burley, G. A.; Gramlich, P. M. E.; Hammond, D. M.; Carell, T.
Org. Lett., 2006, 8, 3639ꢀ3642. (b) Krishna, H.; Caruthers M. H. J. Am. Chem.
Soc., 2012, 134, 11618ꢀ11631.
19. Wicke, L.; Engels, J. W.; Bioconj. Chem., 2012, 23, 627ꢀ642.
20. Darrow, J. W.; Drueckhammer, D. G. Bioorg. Med. Chem., 1996, 1341ꢀ1348.
21. Hanessian, S.; Galéotti, N. Bioorg. Med. Chem. Lett., 1994, 2763ꢀ2768.
22. Ferry, A.; Guinchard, X.; Retailleau, P.; Crich, D. J. Am. Chem. Soc., 2012, 134,
12289ꢀ12301.
23. Clarion, L.; Jacquard, C.; SainteꢀCatherine, O.; Loiseau, S.; Filippini, D.;
Hirlemann, MꢀH.; Volle, JꢀV.; Virieux, D.; Lecouvey, M.; Pirat, JꢀL; Bakalara, N.
J. Med. Chem., 2012, 55, 2196ꢀ2211.
24. Burgess, K.; Cook, D. Chem. Rev., 2000, 100, 2047ꢀ2059.
25. Roy, S.; Olesiak, M.; Padar, P.; McCuen, H.; Caruthers, M. H. Org. Biomol.
Chem., 2012, 10, 9130ꢀ9133.
26. Brummel, H. A. “Chemical synthesis and biochemical analysis of
boranephosphonate DNA” PhD Dissertation: University of Colorado, Boulder,
CO, 2002
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
27. Roy, S.; Olesiak, M.; Shang, S.; Caruthers, M. H. J. Am. Chem. Soc., 2013, 135,
6234ꢀ6241.
28. Ryschkewitsch, G. E. J. Am. Chem. Soc., 1967, 89, 3145ꢀ3148.
29. Nielsen, J.; Caruthers, M. H. J. Am. Chem. Soc., 1988, 110, 6285ꢀ6276.
30. Oivanen, M.; Kuusela, S.; Lönnberg, H. Chem. Rev. 1998, 98, 961ꢀ990.
31. Thatcher, G. R. J.; Kluger, R. Adv. Phys. Org. Chem. 1989, 25, 99.
32. Thatcher, G. R. J; Campbell, A. S. J. Org. Chem., 1993, 58, 2272ꢀ2281.
33. McDowell, R. S.; Streitweiser Jr., A. J. Am. Chem. Soc., 1985, 107, 5849ꢀ5855.
34. Gilheany,D.G. Chem. Rev., 1994, 94, 1339ꢀ1374.
35. Kawanaka, T., Shimizu, M., and Wada, T. Tet. Lett., 2007, 48, 1973ꢀ1976
36. Kawanaka, T., Shimizu, M., Shintani, N., and Wada, T. Bioorg. Med. Chem. Lett.
2008, 18, 3783ꢀ3786.
37. Higashida, R., Oka, N., Kawanaka, T., and Wada, T. Chemical Communication,
2009, 2466ꢀ2468
38. Uehara, S., Hiura, S., Higashida, R., Oka, N., and Wada, T. J. Org. Chem.
2014, 79, 3465ꢀ3472.
39. Higson, A., Sierzchala, A., Brummel, H., Zhao, Z., and Caruthers, M. Tet. Lett.
1998, 39, 3899ꢀ3902.
40. Brummel, H., and Caruthers, M. Tet. Lett. 2002, 43, 749ꢀ751.
41. McCuen, H., Noe, M., Sierzchala, A., Higson, A., and Caruthers, M. J. Am.
Chem.Soc. 2006, 128, 8138ꢀ8139.
42. Kraszewski, A., and Stawinski, J. Pure and Applied Chemistry, 2007, 79, 2217ꢀ
2227.
43. Peyrottes, S., Vasseur, JꢀJ., Imbach, JꢀL., and Rayner, B. Nucleic Acids Res.
1996, 24, 1841ꢀ1848.
44. Asseline, U., Chassignol, M., Draus, J., Durand, M., and Maurizot, JꢀC.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 39
1
2
3
4
5
6
7
8
9
Bioorganic and Medicinal Chemistry, 2003, 11, 3499ꢀ3511.
45. Krishna, H., and Caruthers, M. J. Am. Chem. Soc. 2012, 134, 11618ꢀ11631.
46. Dellinger, D., Sheehan, D., Christensen, N., Lindberg, J., and Caruthers, M. J.
Am. Chem. Soc. 2003, 125, 940ꢀ950.
47. Yamada, C., Dellinger, d., and Caruthers, M. Nucleosides Nucleotides Nucleic
Acids, 2007, 26, 539ꢀ546.
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
48. Yamada, C., Dellinger, D., and Caruthers, M. J. Am. Chem. Soc. 2006, 128, 5251ꢀ
5261.
49. Lu, Y., Mini-Reviews in Medicinal Chemistry, 2006, 6, 319ꢀ330.
50. Guga, P., Okruszek, A., Stec, W. Topics Current Chemistry, 2002, 220, 169ꢀ200.
51. Wilk, A., Grajkowski, A., Phillips, L., and Beaucage, S., J. Am. Chem. Soc. 2000,
122, 2149ꢀ2156.
52. Iwamoto, N., Natsuhisa, O., Wada, T. Tet. Lett. 2012, 53, 4361ꢀ4364
ACS Paragon Plus Environment

Page 27 of 39
Journal of the American Chemical Society
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
Table Of Contents Figure
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 39
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 1(A) The reaction of diastereomerically pure 9 or 10 with 30 equivalents of iodine and a 0.4 M
solution of methylamine in tetrahydrofuran leads to formation of methylphosphoramidates 11 or 12 in a
stereospecific manner. (B) and (C) show the 31P NMR spectrum of the crude reaction mixtures obtained from
9 or 10 as starting materials respectively.
141x146mm (72 x 72 DPI)
ACS Paragon Plus Environment

Page 29 of 39
Journal of the American Chemical Society
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 2LC-MS and denaturing gel electrophoresis analysis of mixed sequence oligonucleotides (ODN 15-
20). Each set of data shows the LC trace along with a zoomed in view of the base peak in the mass
spectra. The base peak in each case was a -4 charged species. The ODN numbers are indicated at the top
of each lane in the gel image. Full mass spectra are provided in the supplementary section.
539x420mm (72 x 72 DPI)
ACS Paragon Plus Environment