Versatile phosphoramidation reactions for nucleic acid conjugations with peptides, proteins, chromophores, and biotin derivatives

Article abstract of DOI:10.1021/bc1001505

Chemical conjugations of nucleic acids with macromolecules or small molecules are common approaches to study nucleic acids in chemistry and biology and to exploit nucleic acids for medical applications. The conjugation of nucleic acids such as oligonucleotides with peptides is especially useful to circumvent cell delivery and specificity problems of oligonucleotides as therapeutic agents. However, current approaches are limited and inefficient in their ability to afford peptide - oligonucleotide conjugates (POCs). Here, we report an effective and reproducible approach to prepare POCs and other nucleic acid conjugates based on a newly developed nucleic acid phosphoramidation method. The development of a new nucleic acid phosphoramidation reaction was achieved by our successful synthesis of a novel amine-containing biotin derivative used to systematically optimize the reactions. The improved phosphoramidation reactions dramatically increased yields of nucleic acid - biotin conjugates up to 80% after 3 h reaction. Any nucleic acids with a terminal phosphate group are suitable reactants in phosphoramidation reactions to conjugate with amine-containing molecules such as biotin and fluorescein derivatives, proteins, and, most importantly, peptides to enable the synthesis of POCs for therapeutic applications. Polymerase chain reactions (PCRs) to study incorporation of biotin or fluorescein-tagged DNA primers into the reaction products demonstrated that appropriate controls of nucleic acid phosphoramidation reactions incur minimum adverse effects on inherited base-pairing characteristics of nucleotides in nucleic acids. The phosphoramidation approach preserves the integrity of hybridization specificity in nucleic acids when preparing POCs. By retaining integrity of the nucleic acids, their effectiveness as therapeutic reagents for gene silencing, gene therapy, and RNA interference is ensured. The potential for POC use was demonstrated by two-step phosphoramidation reactions to successfully synthesize nucleic acid - tetraglycine conjugates. In addition, phosphoramidation reactions provided a facile approach to prepare nucleic acid - BSA conjugates with good yields. In summary, the new approach to phosphoramidation reactions offers a universal method to prepare POCs and other nucleic acid conjugates with high yields in aqueous solutions. The methods can be easily adapted to typical chemistry or biology laboratory setups which will expedite the applications of POCs for basic research and medicine.

Full text of DOI:10.1021/bc1001505

1642
Bioconjugate Chem. 2010, 21, 1642–1655
Versatile Phosphoramidation Reactions for Nucleic Acid Conjugations with
Peptides, Proteins, Chromophores, and Biotin Derivatives
Tzu-Pin Wang,*,† Yi-Jang Chiou,† Yi Chen,† Eng-Chi Wang,† Long-Chih Hwang,† Bing-Hung Chen,‡
Yen-Hsu Chen,§ and Chun-Han Ko| Department of Medicinal and Applied Chemistry, Department of
Biotechnology, Graduate Institute of Medicine, and Department of Internal Medicine, Kaohsiung Medical
University, Kaohsiung, 80708, Taiwan, and School of Forestry and Resource Conservation, National Taiwan
University, Taipei, 10617, Taiwan Received March 25, 2010; Revised Manuscript Received July 9, 2010
Chemical conjugations of nucleic acids with macromolecules or small molecules are common approaches to
study nucleic acids in chemistry and biology and to exploit nucleic acids for medical applications. The conjugation
of nucleic acids such as oligonucleotides with peptides is especially useful to circumvent cell delivery and specificity
problems of oligonucleotides as therapeutic agents. However, current approaches are limited and inefficient in
their ability to afford peptide-oligonucleotide conjugates (POCs). Here, we report an effective and reproducible
approach to prepare POCs and other nucleic acid conjugates based on a newly developed nucleic acid
phosphoramidation method. The development of a new nucleic acid phosphoramidation reaction was achieved
by our successful synthesis of a novel amine-containing biotin derivative used to systematically optimize the
reactions. The improved phosphoramidation reactions dramatically increased yields of nucleic acid-biotin
conjugates up to 80% after 3 h reaction. Any nucleic acids with a terminal phosphate group are suitable reactants
in phosphoramidation reactions to conjugate with amine-containing molecules such as biotin and fluorescein
derivatives, proteins, and, most importantly, peptides to enable the synthesis of POCs for therapeutic applications.
Polymerase chain reactions (PCRs) to study incorporation of biotin or fluorescein-tagged DNA primers into the
reaction products demonstrated that appropriate controls of nucleic acid phosphoramidation reactions incur minimum
adverse effects on inherited base-pairing characteristics of nucleotides in nucleic acids. The phosphoramidation
approach preserves the integrity of hybridization specificity in nucleic acids when preparing POCs. By retaining
integrity of the nucleic acids, their effectiveness as therapeutic reagents for gene silencing, gene therapy, and
RNA interference is ensured. The potential for POC use was demonstrated by two-step phosphoramidation reactions
to successfully synthesize nucleic acid-tetraglycine conjugates. In addition, phosphoramidation reactions provided
a facile approach to prepare nucleic acid-BSA conjugates with good yields. In summary, the new approach to
phosphoramidation reactions offers a universal method to prepare POCs and other nucleic acid conjugates with
high yields in aqueous solutions. The methods can be easily adapted to typical chemistry or biology laboratory
setups which will expedite the applications of POCs for basic research and medicine.
Conjugation of fluorophores with nucleic acids has provided
sensitive detection methods to study tertiary structure transitions
relevant to RNA folding (6-9) and ribozyme catalysis (7, 10, 11),
genomics (12, 13), and DNA microarray preparation (14).
Finally, oligonucleotides conjugated with peptides to afford
peptide-oligonucleotide conjugates (POCs) hold promise as an
effective therapeutic reagent to treat viral infections and genetic
diseases (15-17). Applications of POCs in basic and clinical
research have produced nucleic acids with improved biological
stability (18), cellular uptake efficiency (18, 19), and in vivo
cell-specific targeting (20).
Preparation of POCs is essential for therapeutic applications
of nucleic acids such as oligonucleotides. The direct use of
oligonucleotides in medicine generally falls short of disease
treatment expectations due to poor cell specificity and uptake
of nucleic acids, and inaccessibility of nucleic acids to cell
nuclei (15-17). The conjugation of oligonucleotides with
peptides is thus the most common approach to circumvent cell
delivery and specificity problems of oligonucleotides (15-17).
However, methods currently available to prepare POCs are
inefficient and inconvenient for typical research laboratories.
The difficulties experienced when preparing desired effective
POCs restrict the broad applications of POCs in achieving their
much-needed medical applications.
INTRODUCTION
Covalent modifications of nucleic acids have been used to
explore the diverse functions of nucleic acids and to characterize
their intrinsic biochemical properties in biological systems.
Continuous development of covalent conjugation methods for
nucleic acids further facilitates applications of nucleic acids as
research tools for chemical and biomedical studies, improves
potency, and promotes specificity of nucleic acids as therapeutic
reagents in medicine. For example, the combination of co-
valently modified nucleic acids and in vitro selection methods
has contributed to the discovery of novel catalytic nucleic acids
such as ribozymes (catalytic RNA) able to perform the
Diels-Alder cycloaddition (1), the Michael addition (2), the
aldol condensation (3), NAD+-dependent alcohol dehydroge-
nation (4), and NADH-dependent aldehyde hydrogenation (5).
* Corresponding author. E-mail: tzupinw@cc.kmu.edu.tw. Contact
address: Department of Medicinal and Applied Chemistry, Kaohsiung
Medical University, Kaohsiung, 80708, Taiwan, Tel: +886-07-312-
1101, ext. 2756, Fax: +886-07-312-5339.
† Department of Medicinal and Applied Chemistry, Kaohsiung
Medical University.
‡ Department of Biotechnology, Kaohsiung Medical University.
§ Graduate Institute of Medicine and Department of Internal Medi-
cine, Kaohsiung Medical University.
POCs are generally prepared by post solid-phase synthesis
to couple peptides with oligonucleotides (fragment coupling
| National Taiwan University.
10.1021/bc1001505  2010 American Chemical Society
Published on Web 08/06/2010

Nucleic Acid Phosphoramidation and Applications
Bioconjugate Chem., Vol. 21, No. 9, 2010 1643
strategy) or stepwise solid-phase synthesis (online solid-phase
synthesis) (15-17). POC preparation is generally more feasible
using fragment coupling strategy to avoid chemical intolerance
between peptides and oligonucleotides during standard solid-
phase synthesis (15, 16). The generic fragment coupling strategy
separately synthesizes peptides and oligonucleotides by standard
solid-phase synthesis protocols, and then, both are covalently
linked together post solid-phase synthesis. There are limited
reactions available for POC formation because peptides prima-
rily use unstable phosphodiester linkages when conjugating to
5′- or 3′-termini or ester linkages when linking to 2′-positions
of riboses in oligonucleotides (17). Diverse and stable chemical
linkages between peptides and oligonucleotides can be attained
only if nonconventional nucleotides are used during solid-phase
synthesis of oligonucleotides in order to introduce additional
functionalities. The need to incorporate nonconventional nucle-
otides into oligonucleotides increases the cost of oligonucleotide
solid-phase synthesis and deters current use of POCs. The
development of a facile approach to use hydroxyl or phosphate
groups in standard oligonucleotides to afford stable POCs with
high purity and yields is highly desirable.
An ideal method to prepare POCs is nucleic acid phospho-
ramidation which offers the advantage of providing a stable
phosphoramidate bond for conjugations between peptides and
oligonucleotides. Stability of the phosphoramidate linkage has
long been recognized and extensively used to acquire POCs by
solid-phase synthesis (17). The methodology and applications
of nucleic acid phosphoramidation in aqueous solutions were
previously studied by Orgel’s lab (21). They exploited 1-ethyl-
3-[3-dimethylaminopropyl]carbodiimide (EDC) activation to
transiently generate reactive phosphorimidazolide intermediates
of nucleotides, followed by the coupling reaction with a
nucleophile at rt for 24 h to result in oligonucleotide conjugates
with maximum yields of 50% (21). The concept of phospho-
ramidation reactions for nucleic acid conjugations is analogous
to the effective amidation reaction achieved by the intermediate
step to form isolatable reactive N-hydroxysuccinimide (NHS)
esters of carboxylic acids (22). The aqueous-phase phosphora-
midation reaction, however, has received little attention and
failed to gain popularity as a standard method for nucleic acid
conjugation. Failure to gain support was likely due to cumber-
some and time-consuming procedures and low yields. A later
report of an aqueous-phase phosphoramidation method claimed
to have quantitative yields of reactions between RNA and
ethylenediamine after 4-8 h of incubation (23). However,
further exploration of more extensive applications of phospho-
ramidation for nucleic acid conjugations was not conducted.
An effective and efficient aqueous-phase nucleic acid phos-
phoramidation method became important during our studies of
RNA conjugations and ribozyme catalysis. We thus synthesized
a new biotin derivative (11) to systematically optimize the
nucleic acid phosphoramidation methods of Chu et al. (21) with
the goal to improve effectiveness and efficiency. The synthesis
of 11 was successfully accomplished with an excellent overall
yield of 38.7%. The afforded 11 has a biotin tag and a primary
amine which can be used to optimize phosphoramidation
reactions with nucleic acids. The new aqueous-phase phospho-
ramidation method provided up to 80% yield for 11-nucleic
acid conjugates after 3 h of reaction. The method was able to
complete conversions for 1,6-hexanediamine-nucleic acid con-
jugates after only 10-20 min of reaction. The phosphorami-
dation reaction between a model peptide tetraglycine (12) and
nucleic acids was also successful in producing the desired POCs.
The integrity of hybridization characteristics of nucleic acids
after phosphoramidation reactions was demonstrated by poly-
merase chain reactions (PCRs) in which almost all the phos-
phoramidation-prepared DNA-11 conjugate primers were
integrated into PCR products. The phosphoramidation reactions
were also applied to effective preparation of nucleic acid-BSA
conjugates. The hallmark of the nucleic acid phosphoramidation
reactions is that they offer a facile and universal approach to
prepare POCs with no restrictions on compositions of either
nucleic acids or peptides and without compromising base-pairing
preference in nucleotides. The new aqueous-phase phosphora-
midation reaction will provide an advantageous method for
nucleic acid conjugate preparation and expedite medical ap-
plications of POCs as therapeutic reagents for gene silencing,
gene therapy, and RNA interference (15-17, 24).
MATERIALS AND METHODS
All reagents were purchased from commercial sources
(Sigma-Aldrich, Acros, Alfa Aesar, Mallinckrodt Baker, and
Merck KGaA) except where noted and were further purified if
necessary. Both disuccinimidyl suberate (DSS) and disuccin-
imidyl glutarate (DSG) for oligonucleotide-peptide conjugates
were synthesized according to published methods (25). The
procedures modified from those of Boger et al. (26) were used
to synthesize H2N-Gly-Gly-Gly-Gly-OH (tetraglycine; 12). 1H
and 13C NMR spectra were recorded using either a Varian 200
or 400 MHz spectrometer (Varian, Inc., Palo Alto, CA, USA)
at Kaohsiung Medical University, Taiwan. NMR samples were
prepared in CD3OD, DMSO-d6, or CDCl3, and the chemical
1
shifts of H signals were given in parts per million downfield
from TMS. 13C signals were given in parts per million based
on the internal standard of each deuteriated solvent. ESI high-
resolution mass spectra were acquired from Department of
Chemistry, National Sun Yat-Sen University, Taiwan, on a
Bruker APEX II Fourier-transfer mass spectrometry (FT-MS;
Bruker Daltonics Inc., Taiwan). Nucleic acid conjugates were
analyzed by urea polyacrylamide gel electrophoresis (PAGE)
or streptavidin (SAv) gel shift assay in urea PAGE, visualized,
and quantified by an Amersham Typhoon phosphoimager (GE
Healthcare Bio-Sciences AB, Uppsala, Sweden).
Synthesis of the Biotin Derivative (11). 5-(2-Oxo-hexahydro-
thieno[3,4-d]imidazol-4-yl)-pentanoic Acid 2,5-Dioxo-pyrrolidin-
1-yl Ester (2). The method for synthesis of 2 was modified from
that of Bayer and Wilchek (27). In brief, biotin (1; 0.31 g, 1.26
mmol) was completely dissolved in DMF (9 mL) with gentle
warming. After cooling the 1 solution to room temperature
without reprecipitation, it was added to N,N′-dicyclohexylcar-
bodiimide (DCC; 0.26 g, 1.28 mmol) and pyridine (0.10 mL,
1.26 mmol) while stirring for 5 min at room temperature,
followed by the addition of N-hydroxysuccinimide (NHS;
0.19 g, 1.65 mmol) with continuous stirring for 24 h. The
reaction product was filtered and concentrated under reduced
pressure. The obtained solid was redissolved in 2-propanol (50
mL) with gentle heating, cooled down to room temperature,
and reprecipitated at 4 °C overnight. The product was scraped
out of the glassware, washed with cold 2-propanol, and air-
1
dried to give 2 as white solid (0.36 g; 82.4%). H NMR (400
MHz) (DMSO) δ: 6.46 (s, 1H, CONH), 6.39 (s, 1H, CONH),
4.31 (t, 1H, CHN), 4.16 (t, 1H, CHN), 3.12 (dd, 1H, CHS),
2.83 (s, 4H, CH2 of NHS), 2.71 (d, 1H, CHHS), 2.69 (t, 2H,
CH2CO), 2.59 (d, 1H, CHHS), 1.43-1.68 (m, 6H). 13C NMR
(100.67 MHz) (DMSO) δ: 170.42, 162.85, 61.01, 59.28, 55.33,
30.08, 27.91, 27.68, 25.51. HRMS (ESI) calculated for
C14H19N3O5S, [M+Na]+ 364.0943 (calcd.), 364.0940 (found).
6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoy-
lamino]-hexanoic Acid (3). 5-(2-Oxo-hexahydro-thieno[3,4-d]imi-
dazol-4-yl)-pentanoic acid 2,5-dioxo-pyrrolidin-1-yl ester (2; 0.35 g,
1.03 mmol) was completely dissolved in DMF (3 mL) with
gentle warming (28). After cooling the 2 solution to room
temperature without reprecipitation, it was mixed with a solution
of ε-aminocaproic acid (0.18 g, 1.37 mmol) in sodium bicarbon-

1644 Bioconjugate Chem., Vol. 21, No. 9, 2010
Wang et al.
ate buffer (0.1 M, pH 8.0; 4 mL) while stirring. After 4 h at
room temperature, the reaction mixture was acidified by HCl
solution (1 N) to assist in the formation of white precipitate. The
white solid was washed with cold HCl solution, dried over ether,
5 (0.25 g, 1.14 mmol) in DMF (3 mL) dropwise while stirring.
After overnight reaction at room temperature, the mixture was
filtered, evaporated in vacuo, and loaded to a silica column for
further purification. The products were separated by eluting with
mobile phases of 5-20% MeOH in DCM to afford 6 (0.45 g;
93.8%) as a yellowish solid. 1H NMR (400 MHz) (CD3OD) δ:
4.53 (t, 1H, CHN), 4.34 (t, 1H, CHN), 3.25 (dd, 1H, CHS),
3.20 (t, 2H, CH2NH), 3.08 (t, 2H, CH2NH), 2.96 (d, 1H, CHHS),
2.73 (d, 1H, CHHS), 2.22 (t, 4H, CH2CO), 1.38-1.80 (m, 29H).
13C NMR (100.67 MHz) (DMSO) δ: 176.02, 175.97, 166.05,
158.50, 79.50, 63.39, 61.63, 57.01, 41.24, 41.04, 40.24, 36.99,
36.82, 30.90, 30.35, 30.13, 29.78, 29.49, 28.80, 27.56, 26.92,
26.74. HRMS (ESI) calculated for C27H49N5O5S, [M+Na]+
578.3352 (calcd.), 578.3349 (found).
1
and air-dried to afford 3 (0.30 g; 80.5%). H NMR (400 MHz)
(DMSO) δ: 7.78 (s, 1H, CONH), 6.46 (s, 1H, CONH), 6.38 (s,
1H, CONH), 4.32 (t, 1H, CHN), 4.14 (t, 1H, CHN), 3.11 (dd,
1H, CHS), 3.02 (t, 2H, CH2NH), 2.84 (d, 1H, CHHS), 2.59 (d,
1H, CHHS), 2.18 (t, 2H, CH2COOH), 2.05 (t, 2H, CH2CO),
1.23-1.68 (m, 12H). 13C NMR (100.67 MHz) (DMSO) δ:
174.45, 171.79, 162.70, 61.03, 59.18, 55.43, 35.21, 33.60, 28.90,
28.02, 25.97, 25.33, 24.22. HRMS (ESI) calculated for
C16H27N3O4S, [M+Na]+ 380.1620 (calcd.), 380.1617 (found).
6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoy-
lamino]-hexanoic Acid 2,5-Dioxo-pyrrolidin-1-yl Ester (4). Syn-
thesis of 4 with high purity was achieved following the method
of Wilchek and Bayer with critical changes in workup proce-
dures (28). 6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-
pentanoylamino]-hexanoic acid (3; 0.31 g, 0.87 mmol) was
completely dissolved in DMF (20 mL) with gentle warming.
After cooling the 3 solution to room temperature without
reprecipitation, DCC (0.34 g, 1.65 mmol) and pyridine (0.07
mL, 0.83 mmol) were added to the solution while stirring for 5
min, followed by the addition of NHS (0.16 g, 1.46 mmol) with
continuous stirring for 18 h. The reaction product was filtered
and concentrated under reduced pressure. The remaining solid
was redissolved in 2-propanol (40 mL) with gentle heating,
cooled down to room temperature, and reprecipitated at 4 °C
overnight. The product was scraped out of the glassware, washed
with cold 2-propanol, and air-dried to obtain 4 (0.22 g; 56.8%)
as a white solid. 1H NMR (400 MHz) (DMSO) δ: 7.77 (s, 1H,
CONH), 6.45 (s, 1H, CONH), 6.38 (s, 1H, CONH), 4.31 (t,
1H, CHN), 4.15 (t, 1H, CHN), 3.11 (dd, 1H, CHS), 3.03 (t,
2H, CH2NH), 2.83 (s, 4H, CH2 of NHS), 2.79 (d, 1H, CHHS),
2.67 (t, 2H, CH2COOH), 2.60 (d, 1H, CHHS), 2.06 (t, 2H,
CH2CO), 1.04-1.75 (m, 12H). 13C NMR (100.67 MHz)
(DMSO) δ: 171.90, 170.36, 169.03, 162.79, 61.13, 59.27, 55.53,
38.14, 35.32, 30.24, 28.74, 28.32, 28.14, 25.54, 25.41, 24.05.
HRMS (ESI) calculated for C20H30N4O6S, [M+Na]+ 477.1784
(calcd.), 477.1784 (found).
6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoy-
lamino]-hexanoic Acid (6-amino-hexyl)-amide, TFA Salt (7). (6-
{6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoy-
lamino]-hexanoylamino}-hexyl)-carbamic acid tert-butyl ester
(6; 0.23 g, 0.39 mmol) was dissolved in trifluoroacetic acid
(TFA) and stirred at room temperature for 45 min. The product
was concentrated under reduced pressure, washed by ether, and
again concentrated under reduced pressure to obtain oil-like 7
(0.22 g; 100%) as TFA salts. 1H NMR (400 MHz) (CD3OD) δ:
4.58 (t, 1H, CHN), 4.35 (t, 1H, CHN), 3.25 (dd, 1H, CHS),
3.21 (t, 4H, CH2NH), 2.96 (d, 1H, CHHS), 2.95 (t, 2H,
CH2NH3+), 2.75 (d, 1H, CHHS), 1.22-1.80 (m, 20H). 13C NMR
(100.67 MHz) (DMSO) δ: 176.67, 166.14, 63.38, 61.63, 57.03,
41.05, 40.64, 40.19, 40.01, 36.98, 36.81, 30.17, 29.78, 29.50,
28.50, 27.53, 27.33, 26.96, 26.72, 26.27. HRMS (ESI) calculated
for C22H41N5O3S, [M+H]+ 456.3008 (calcd.), 456.3009 (found).
Carbamoylmethyl-carbamic Acid tert-Butyl Ester (8a). Gly-
cine (0.45 g, 6.05 mmol) was dissolved in 6 mL of dioxane/
water mixture (dioxane:water ) 4:2) with stirring, followed by
the addition of 1 N NaOH (2 mL). The acquired glycine solution
was immersed in an ice-water bath and mixed with di-tert-
butyl dicarbonate (1.38 g, 6.31 mmol) dropwise while stirring
to initiate the reaction. The reaction was allowed to proceed at
room temperature overnight. The reaction was diluted with water
(2 mL) and EA (10 mL), extracted with 1 N HCl twice, and
extracted with water twice. The final organic phase was dried
over MgSO4 and evaporated under reduced pressure to give oil-
like 8a (0.94 g; 90%). 1H NMR (400 MHz) (CDCl3) δ: 5.42 (s,
1H, CONH), 3.91 (d, 2H, CH2CO), 1.44 (s, 9H, CCH3). 13C
NMR (100.67 MHz) (CDCl3) δ: 177.0, 156.0, 79.6, 35.8, 34.3,
28.2.
(6-Amino-hexyl)-carbamic Acid tert-Butyl Ester (5). 1,6-
Hexanediamine (1.0 g, 8.61 mmol) was completely dissolved
in dichloromethane (DCM; 7 mL) (29). The solution was chilled
to 0 °C, followed by the addition of di-tert-butyl dicarbonate
(0.63 g, 2.87 mmol) with stirring. The reaction was allowed to
proceed at room temperature overnight. The acquired reaction
product was diluted with chloroform (7.5 mL), extracted with
5% Na2CO3 twice, and its organic phase concentrated under
reduced pressure. The obtained solid was dissolved in 1 N HCl
and extracted with ether twice. The pH of the acquired aqueous
phase was adjusted to 10 with NaOH and extracted with ethyl
acetate (EA) five times. The final organic phase was dried over
MgSO4 and concentrated under reduced pressure to give oil-
tert-Butoxycarbonylamino-acetic Acid 2,5-Dioxo-pyrrolidin-1-
yl Ester (8b). Carbamoylmethyl-carbamic acid tert-butyl ester
(8a; 0.61 g, 3.48 mmol) and NHS (0.61 g, 5.34 mmol) were
dissolved in THF (7 mL), followed by dropwise addition of
DCC (1.42 g, 6.9 mmol) in THF (7 mL) with stirring (30). After
overnight reaction at room temperature, the mixture was
quenched by the addition of three drops of glacial acetic acid,
stirred for 1 h, and filtered to remove suspension. The acquired
filtrate was concentrated under reduced pressure, resuspended
in 2-propanol (25 mL) while stirring for 1 h, and filtered to
separate the suspension from its filtrate. The remaining solid
was washed by 2-propanol and dried to afford white-colored
1
like 5 (0.25 g; 39.8%). H NMR (400 MHz) (CDCl3) δ: 4.54
(s, 1H, CONH), 3.11 (t, 2H, CH2NH), 2.68 (t, 2H, CH2NH2),
1.40-1.50 (s, 9H, CH3; m, 4H, CH2), 1.25-1.37 (m, 4H, CH2).
13C NMR (100.67 MHz) (DMSO) δ: 77.00, 41.98, 33.55, 29.95,
28.33, 26.48. HRMS (ESI) calculated for C11H24N2O2, [M+H]+
217.1916 (calcd.), 217.1917 (found).
1
8b (0.62 g; 68.8%). H NMR (400 MHz) (CDCl3) δ: 5.00 (s,
1H, CONH), 4.27 (d, 2H, CH2CO), 2.85 (s, 4H, CNOCH2CH2),
1.43 (s, 9H, CCH3). 13C NMR (100.67 MHz) (CDCl3) δ: 168.6
(CONH), 77.0 (CH2N), 40.2 (CH2CO), 37.4 (CH2CO), 25.5
(CH3CO).
(6-{6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pen-
tanoylamino]-hexanoylamino}-hexyl)-carbamic Acid tert-Butyl Es-
ter (6). 6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-
pentanoylamino]-hexanoic acid 2,5-dioxo-pyrrolidin-1-yl ester
(4; 0.39 g, 0.86 mmol) was completely dissolved in DMF (15
mL) with gentle warming. After cooling the solution to room
temperature without reprecipitation, it was mixed with triethy-
lamine (Et3N; 0.52 g, 5.15 mmol) followed by the addition of
[(6-{6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pen-
tanoylamino]-hexanoylamino}-hexylcarbamoyl)-methyl]-carbam-
ic Acid tert-Butyl Ester (9). tert-Butoxycarbonylamino-acetic acid
2,5-dioxo-pyrrolidin-1-yl ester (8b; 0.18 g, 0.69 mmol) dissolved
in DMF (5 mL) was added slowly to a DMF solution (3 mL)

Nucleic Acid Phosphoramidation and Applications
Bioconjugate Chem., Vol. 21, No. 9, 2010 1645
containing 6-[5-(2-oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-
pentanoylamino]-hexanoic acid (6-amino-hexyl)-amide (7; 0.16 g,
0.27 mmol) and Et3N (0.1 mL, 0.72 mmol) while stirring at
room temperature. The reaction was stopped after 3 h by
evaporating DMF in vacuo, and the remaining residue was
redissolved in a limited volume of a DCM/MeOH (19:1)
solution and loaded to a preequilibrated silica column. The
products were separated by eluting with mobile phases of
1-6.67% MeOH in DCM to afford 9 (0.10 g; 62.2%) as a
presence of ethanol, and redissolved in 50 mM KCl. The GMP-
primed TW17 RNA (87-mer; 5′-GGGAUCGUCAGUGCA-
UUGAGAAGUGCAGUGUCUUGCGCUGGGUUCGAGCG-
GUCCGUGGUGCUGGCCCGGUGGUAUCCCCAAGGGGUA-
3′) was transcribed from the PCR-amplified double-stranded
TW17 DNA carrying T7 promoter sequences under the T7 RNA
polymerase runoff reaction in the presence of 6.25 mM each of
ATP, CTP, GTP, and UTP, 10 mM GMP (guanosine mono-
phosphate) and with or without 5 µCi [R-32P]UTP (Izotop,
Hungary). The GMP-primed TW17 RNA transcript was sepa-
rated by 8% urea PAGE (220 V, 120 min) followed by ethanol
precipitation. Acquired nucleic acids were quantified by their
1
yellow solid. H NMR (400 MHz) (CD3OD) δ: 4.30 dd, 1H,
CHN), 4.15 (dd, 1H, CHN), 3.94 (d, 2H, CH2CO), 3.32 (quin,
4H, CH2NH), 2.99 (q, 2H, CH2N), 2.92 (d, 1H, CHHS), 2.75
(d, 1H, CHHS), 2.20 (t, 4H, CH2CO), 1.44-1.88 (m, 29H). 13
C
absorption at A260
.
NMR (100.67 MHz) (CD3OD) δ: 175.9 (CO), 63.7 (CHN), 61.6
(CHN), 56.9 (CHS), 49.0 (CH2N), 41.0 (CH2CO), 40.2 (CH2N),
36.9 (CH2S), 30.1 (CH2CO), 28.6 (CH3CO), 26.7 (CH2CH2).
HRMS (ESI) calculated for C29H52N6O6S, [M+Na]+ 635.3567
(calcd.), 635.3564 (found).
Nucleic Acid Radiolabeling. The TW17 RNA was body-
labeled with 32P during T7 RNA polymerase runoff transcription
stated previously. The double-stranded and single-stranded DNA
molecules were 32P-labeled at their 5′-ends with the procedures
following the manufacture’s instruction (Promega, Madison, WI,
USA) and briefly described below. The double-stranded or
single-stranded DNA (200 pmol) was dissolved in 20 µL of
DEPC (diethyl dicarbonate) water, followed by the addition of
10× reaction buffer (3 µL), alkaline phosphatase (3 U), and
DEPC water (4 µL). The reaction was stopped after 2 h at 37
°C, and the reaction product was purified by phenol/chloroform
and chloroform extractions, followed by ethanol precipitation.
The 5′-OH nucleic acids were redissolved in 7 µL of DEPC
water, mixed with 10× reaction buffer (1 µL), [γ-32P]ATP
(Izotop, Hungary) and T4 polynucleotide kinase (10 U; Prome-
ga, Madison, WI, USA), and reacted 2 h at 37 °C. 32P-labeled
double-stranded DNA was purified by phenol/chloroform and
chloroform extractions, followed by ethanol precipitation.
Single-stranded DNA with 32P-labeled 5′-end was purified by
20% urea PAGE.
Nucleic Acid Phosphoramidation Reactions. One-Step
Phosphoramidation Reactions. The standard one-step phospho-
ramidation reaction for RNA was carried out by dissolving the
GMP-primed TW17 RNA (92 pmol) and EDC (6.52 µmol) in
20 µL urea-imidazole buffer (8 M urea, 0.1 M imidazole, pH
6.0) and activating at rt for 90 min, followed by the addition of
7.5 µL urea-EPPS buffer (8 M urea, 100 mM EPPS, 2 mM
EDTA, pH 8.0) and 5 µL of 11 (187.2 mM in DMF) to react at
41 °C for 3 h. The double-stranded TW17 DNA was also
conjugated with 11 according to the same phosphoramidation
reaction stated above. For single-stranded DNA (the 3′-primer
DNA) the phosphoramidation reaction was modified by dis-
solving single-stranded DNA (92 pmol) and EDC (6.52 µmol)
in 20 µL imidazole buffer (0.1 M imidazole, pH 6.0) and
activating at rt for 90 min, followed by the addition of 7.5 µL
EPPS buffer (100 mM EPPS, 2 mM EDTA, pH 8.0) and 5 µL
of 11 (187.2 mM in DMF) with subsequent reactions at 41 °C
for 3 h. No urea was used here to ensure higher reaction yields.
All resulting nucleic acid-substrate conjugates were purified
twice by ethanol precipitation, and their reaction yields were
analyzed by SAv gel shift assay (8% urea PAGE for the TW17
DNA and RNA, and 20%/8% biphasic urea PAGE for single-
stranded DNA), visualized, and quantified by an Amersham
Typhoon phosphoimager.
6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoy-
lamino]-hexanoic Acid [6-(2-Amino-acetylamino)-hexyl]-amide,
TFA Salt (10). [(6-{6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-
4-yl)-pentanoylamino]-hexanoylamino}-hexylcarbamoyl)-methyl]-
carbamic acid tert-butyl ester (9; 0.38 mg, 0.62 mmol) was
dissolved in TFA (3 mL) and stirred for 45 min. The product
was evaporated under reduced pressure, washed by chloroform,
and again concentrated under reduced pressure to remove chloro-
1
form and to obtain brown oil-like 10 (0.32 g, 100%). H NMR
(400 MHz) (CD3OD) δ: 4.50 (dd, 1H, CHN), 4.30 (dd, 1H,
CHN), 3.64 (s, 2H, CH2CO), 3.13-3.23 (m, 4H, CH2NH), 2.99
(q, 2H, CH2N), 2.93 (dd, 1H, CHHS), 2.72 (d, 1H, CHHS), 2.20
(t, 4H, CH2CO), 1.29-1.88 (m, 20H). 13C NMR (100.67 MHz)
(CD3OD) δ: 175.9 (CO), 63.7 (CHN), 61.6 (CHN), 56.9 (CHS),
49.0 (CH2N), 41.0 (CH2CO), 40.2 (CH2N), 36.9 (CH2S), 30.1
(CH2CO), 26.7 (CH2CH2). HRMS (ESI) calculated for
C24H44N6O4S, [M+H]+ 513.3223 (calcd.), 513.3226 (found).
6-[5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoy-
lamino]-hexanoic Acid [6-(2-Amino-acetylamino)-hexyl]-amide
(11). The TFA salt of 6-[5-(2-oxo-hexahydro-thieno[3,4-d]imi-
dazol-4-yl)-pentanoylamino]-hexanoic acid [6-(2-amino-acety-
lamino)-hexyl]-amide, TFA salt (10; 0.52 g, 0.88 mmol) was
dissolved in 5 mL of DCM/MeOH mixture (DCM:MeOH )
1:1) and mixed with 10 equiv of Amberlyst A-21 (1.1 mg, 8.8
mmol) while shaking for 30 min (31). After filtering to remove
Amberlyst A-21, the filtrate of the reaction products was
concentrated under reduced pressure to afford yellow oil-like
1
11 (0.42 mg, 93.8%). H NMR (400 MHz) (CD3OD) δ: 4.49
(dd, 1H, CHN), 4.30 (dd, 1H, CHN), 3.32 (s, 2H, CH2CO),
3.14-3.29 (m, 4H, CH2N), 2.99 (q, 2H, CH2N), 2.86 (s, 1H,
CHS), 2.72 (d, 1H, CHS), 2.20 (q, 4H, CH2CO), 1.29-1.88
(m, 20H). 13C NMR (100.67 MHz) (CD3OD) δ: 175.9 (CO),
63.7 (CHN), 61.6 (CHN), 56.9 (CHS), 49.0 (CH2N), 41.0
(CH2CO), 40.2 (CH2N), 36.9 (CH2S), 30.1 (CH2CO), 26.7
(CH2CH2). HRMS (ESI) calculated for C24H44N6O4S, [M+H]+
513.3223 (calcd.), 513.3226 (found).
Nucleic Acid Preparation. Single-stranded DNA (the 5′ primer,
5 ′ - A A C A C G C A T A T G T A A T A C G A C T C A C T A T -
AGGGATCGTCAGTGCATTGAG-3′; the 3′ primer, 5′-TAC-
CCCTTGGGGATACCACC-3′) was purchased from Purigo
Biotech, Inc., Taiwan, and purified by PAGE. Both the 5′ and
3′ primers were used in a standard PCR reaction to amplify the
double-stranded TW17 DNA (119 bp; 5′-GGTAACACG-
CATATGTAATACGACTCACTATAGGGATCGTCAGTGCATT-
GAGAATGTCAGTGTCTTGCGCTGGGTTCGAGCGGTCCG-
TGGTGCTGGCCCGGTGGTATCCCCAAGGGGTA-3′) from
a plasmid derived from the pGEM-T vector (Promega, Madison,
WI, USA; Wang et al., unpublished results). PCR products were
extracted by phenol-chloroform solutions, precipitated in the
Two-Step Phosphoramidation Reactions. The standard two-
step phosphoramidation reaction for RNA was carried out by
dissolving the GMP-primed TW17 RNA (92 pmol) and EDC
(6.52 µmol) in 20 µL urea-imidazole buffer (8 M urea, 0.1 M
imidazole, pH 6.0) and activating at rt for 90 min. The resulting
5′-phosphorimidazolide was purified by ethanol precipitation,
redissolved in 27.5 µL of urea-EPPS buffer (8 M urea, 100
mM EPPS, 2 mM EDTA, pH 7.5), and added to 5 µL of 11
(187.2 mM in DMF) under reaction temperature at 41 °C for
3 h. For the single-stranded 3′-primer DNA, the two-step

1646 Bioconjugate Chem., Vol. 21, No. 9, 2010
Wang et al.
phosphoramidation reaction was modified by dissolving single-
stranded DNA (92 pmol) and EDC (6.52 µmol) in 20 µL
imidazole buffer (0.1 M imidazole, pH 6.0) and activating at rt
for 90 min. Again, the resulting 5′-phosphorimidazolides were
also purified by ethanol precipitation, redissolved in 27.5 µL
of EPPS buffer (100 mM EPPS, 2 mM EDTA, pH 7.5), and
added to 5 µL of 11 (187.2 mM in DMF) to react at 41 °C for
3 h. The double-stranded TW17 DNA was also conjugated with
11 according to the same two-step phosphoramidation reaction
for the single-stranded DNA. It is noted that urea was not used
in two-step phosphoramidation reactions for the doubled-
stranded and single-stranded DNA molecules to attain higher
reaction yields. All acquired nucleic acid-11 conjugates were
purified twice by ethanol precipitation, and their reaction yields
were analyzed by SAv gel shift assay in the same urea PAGE
described for one-step phosphoramidation reaction analysis,
visualized,andquantifiedbyanAmershamTyphoonphosphoimager.
was also purified by a standard phenol-chloroform extraction
procedure, analyzed by urea PAGE, and visualized under an
Amersham Typhoon phosphoimager using an excitation wave-
length of 488 nm and an emission filter set at 525 nm.
Preparation of Nucleic Acid-Tetraglycine (12) Conju-
gates by Phosphoramidation Reactions. The phosphorami-
dation reaction between nucleic acids and tetraglycine 12
followed the standard two-step phosphoramidation method stated
above with 11 substituted with 12. The afforded nucleic
acid-tetraglycine conjugates were purified twice by ethanol
precipitation, and the reaction yields were analyzed by high-
resolution 8% urea PAGE for the TW17 RNA-12 and the
TW17 DNA-12 conjugates or high-resolution 20% urea PAGE
for the 3′-primer DNA-12 conjugate. All electrophorized
samples were visualized and quantified by an Amersham
Typhoon phosphoimager.
Nucleic Acid-BSA (Bovine Serum Albumin) Conjugate
Preparation. The standard procedures for preparation of nucleic
acid-BSA conjugates were modified from those of Jablonski
et al. (33). Oligonucleotide-1,6-hexanediamine conjugates were
dissolved in 10 µL of EDTA-bicarbonate buffer (0.1 M sodium
bicarbonate, 2 mM EDTA, pH 8.25) to give a concentration of
0.02 mM. Nucleic acid-1,6-hexanediamine conjugate solutions
were then reacted with 20 µL of DSS or DSG solutions (182
mM in DMSO) for 5 min at rt in the dark. Reaction products
were immediately purified twice by ethanol precipitation to
avoid hydrolysis of the remaining succinimidyl esters. The
afforded nucleic acid conjugates were again redissolved in 30
µL of BSA (10 mg/mL) in sodium phosphate buffer (0.1 M,
pH 7.5) to give a molar ratio of BSA to nucleic acid conjugates
of 2. The reaction was allowed to proceed for 2-16 h at rt, and
the reaction product was purified twice by ethanol precipitation.
Yields of nucleic acid-BSA conjugates were also determined
by urea PAGE analysis and visualized through an Amersham
Typhoon phosphoimager.
Nucleic Acid-1,6-Hexanediamine Conjugate Preparation.
The phosphoramidation reaction between the TW17 RNA and
1,6-hexanediamine followed the standard one-step phosphora-
midation method stated above but with 11 replaced by 1,6-
hexanediamine. The phosphoramidation reaction between DNA
and 1,6-hexanediamine was carried out by the standard one-
step phosphoramidation method with the following modifica-
tions: (1) 11 replaced by 1,6-hexanediamine, (2) EDC activation
time shortened to 15 min, (3) nucleophilic coupling reaction
time shortened to 10 min for both the double-stranded TW17
DNA and the single-stranded 3′ primer DNA. The afforded
nucleic acid-1,6-hexanediamine conjugates were purified twice
by ethanol precipitation, and the reaction yields were analyzed
by 8% (the TW17 DNA and RNA) or 20% (the 3′-primer DNA)
urea PAGE, visualized, and quantified by an Amersham
Typhoon phosphoimager.
Single-Stranded 3′-Primer DNA-Fluorescein Conjugate
Preparation. The single-stranded 3′-primer DNA-fluorescein
conjugate was prepared from its corresponding single-stranded
3′-primer DNA-1,6-hexanediamine conjugate described previ-
ously. In a typical reaction for the preparation of single-stranded
DNA-fluorescein conjugates, the single-stranded 3′-primer
DNA-1,6-hexanediamine conjugate was redissolved in 20 µL
sodium carbonate buffer (0.1 M, pH 9.0) to give a final
concentration of 2 µM (32). Under a darkened laboratory setup,
fluorescein isothiocyanate (FITC, Acros) was weighted and
dissolved in dry DMSO for a final concentration of 1 mg/mL
(2.57 µM). In the same darkened laboratory, 20 µL of the FITC
solution was slowly added to the singled-stranded 3′-primer
DNA-1,6-hexanediamine conjugate solution while gentle mix-
ing. The reaction was allowed to proceed for 8 h at 4 °C in the
dark. The afforded single-stranded DNA-fluorescein conjugate
was purified twice by ethanol precipitation.
RESULTS
Synthesis of the Biotin Derivative (11) and Coupling to
the TW17 RNA for Optimization of Phosphoramidation
Reactions. The synthesis of 11 started from biotin (1) which
was activated by carbodiimide to give its NHS ester (2),
followed by coupling 2 with 4-aminocaproic acid to produce 3
which was again activated by carbodiimide to afford the
corresponding NHS ester 4 (Scheme 1). We initially observed
that the yield for 4 was always higher than 100%, an indication
of ineffective product purification despite following the method
of Wilchek and Bayer (28). We thus changed the workup
procedures by reprecipitating the product in 2-propanol to
1
acquire a fairly clean 4 (56.8%) as confirmed by H and 13C
NMR and HRMS (ESI). Compound 4 was then reacted with
previously synthesized 5 to afford 6 (93.8%), followed by TFA
Boc-deprotection (7) and amidation with previously synthesized
8b to give 9 (86.4%). The final product 11 (93.8%) was almost
quantitatively obtained after TFA deprotection of 9 and TFA
PCR Reactions to Determine DNA Amplification Efficiency
for the Biotin-Labeled 3′-Primer DNA or to Demonstrate
Formation of the 3′-Primer DNA-Fluorescein Conjugate.
Standard PCR procedures were applied to determine DNA
amplification efficiency using the 11-modified 3′-primer DNA
under the condition that the normal 3′-primer (100 µM, 0.75
µL) was substituted with a mixture of the normal 3′-primer (100
µM, 0.15 µL) and 32P-labeled and 11-modified 3′-primer (1.9
µM, 10 µL). After 15 cycles of the PCR reaction, the product
was purified by a standard phenol-chloroform extraction
procedure, analyzed by biphasic urea PAGE (top gels of 6%
acrylamide and bottom gels of 20% acrylamide), and visualized
through an Amersham Typhoon phosphoimager. For the fluo-
rescein-labeled 3′-primer DNA, its formation was studied by
the same PCR reaction except that 32P-labeled and 11-modified
3′-primer was substituted with nonradiolabeled fluorescein-
modified 3′-primer (1.9 µM, 10 µL). The PCR reaction product
1
removal of 10, as confirmed by H and 13C NMR and HRMS
(ESI).
Successful synthesis of 11 paved the way for the optimization
of aqueous-phase nucleic acid phosphoramidation reactions. As
noted previously, aqueous-phase nucleic acid phosphoramidation
reactions exploit EDC to activate 5′-terminal phosphate groups
in nucleic acids and to transiently form reactive phosphorimi-
dazolide intermediates of nucleic acids (Supporting Information
Scheme S1). Phosphorimidazolide intermediates of nucleic acids
can be purified or used directly to couple with amine-containing
nucleophiles to result in desired nucleic acid conjugates. We
believed that a key factor affecting final yields for phosphora-
midation-prepared nucleic acid conjugates would rely on

Nucleic Acid Phosphoramidation and Applications
Bioconjugate Chem., Vol. 21, No. 9, 2010 1647
Scheme 1. Synthesis of the Biotin Derivative 11 for Optimization
of Nucleic Acid Phosphoramidation Reactions
providing a reasonable high concentration of phosphorimida-
zolide intermediates for nucleophilic attack by 11. We thus
refined phosphoramidation reactions to provide a steady supply
of phosphorimidazolide intermediates to 11 during the course
of the reactions. Refinements included systematic modifications
of reaction conditions such as reactant and EDC concentrations,
activation time, reaction time with nucleophile 11, reaction
temperature, pH, and surfactant effects. Both one-step and two-
step phosphoramidation reactions were studied with the ultimate
goal of improving reaction efficiency and effectiveness.
The optimized one-step phosphoramidation reaction provided
an excellent yield of 79% for conjugation of the TW17 RNA
with 11 after 3 h reactions (Figure 1A). Further improvement
in the phosphoramidation reaction yield was not achieved,
probably reflecting the lability of phosphorimidazolide inter-
mediates and inaccessibility of phosphorimidazolide intermedi-
ates to 11. A constant reaction temperature is essential for
attaining high phosphoramidation reaction yields. The choice
of 41 °C as a reaction temperature balances an enhanced reaction
rate and prevention of phosphorimidazolide hydrolysis and RNA
hydrolysis at higher temperatures. Maintenance of denaturing
conditions by 6.77 M urea during the phosphoramidation
reaction plays a pivotal role in achieving higher yields by
preventing formation of RNA secondary structures. A very small
portion of the GMP-primed TW17 RNA suffered from degrada-
tion during the reaction, an amount too insignificant to decrease
overall reaction efficiency and yield.
Figure 1. Phosphoramidation reactions of the 32P-labeled GMP-
primed TW17 RNA with 11. (A) Results from the one-step
phosphoramidation reaction were analyzed by streptavidin (SAv)
gel shift assay in 8% urea PAGE and visualized by autoradiogram.
(B) Results from the two-step phosphoramidation reaction were again
analyzed by SAv gel shift assay in 8% urea PAGE and autoradio-
gram. The samples in each figure, starting from the left, are the
TW17 RNA without the phosphoramidation reaction, the TW17
RNA conjugated with 11 by the phosphoramidation reaction without
the addition of 6.77 M urea, and the TW17 RNA conjugated with
11 by the phosphoramidation reaction in the presence of 6.77 M
urea, respectively. In both figures, the top arrow indicates the
location of the TW17 RNA-11 conjugate; the bottom arrow
represents migration of the GMP-primed TW17 RNA.
yields than the one-step phosphoramidation reactions (52% in
the presence of 6.77 M urea and 40% without 6.77 M urea;
Figure 1B). Lower yields for two-step phosphoramidation
reactions are attributed to phosphorimidazolide intermediate
hydrolysis during its workup step. Both two-step phosphora-
midation reactions, with or without 6.77 M urea, still provided
reasonably good yields for the TW17 RNA-11 conjugates,
comparable to those prepared by the one-step phosphoramidation
reactions. Two-step phosphoramidation reactions can be an
advantageous approach to conjugate carbodiimide-sensitive
nucleophiles with nucleic acids (vide infra). However, similar
to the one-step reactions, RNA degradation was also observed.
RNA degradation was thus not caused by prolonged carbodi-
imide activation but rather by the nucleophile attack step or by
a higher reaction temperature at 41 °C.
Chu et al. also performed a two-step nucleic acid phospho-
ramidation method attained by purification of phosphorimida-
zolide intermediates of DNA which were subsequently reacted
with nucleophiles to afford DNA phosphoramidation conjugates
(21). We therefore also exploited the two-step phosphorami-
dation method for the TW17 RNA conjugation with 11 to study
its effectiveness in preventing RNA degradation observed during
the one-step phosphoramidation reaction. Yields of the two-
step phosphoramidation reactions for the TW17 RNA-11
conjugate preparation were consistent with expectations of lower

1648 Bioconjugate Chem., Vol. 21, No. 9, 2010
Wang et al.
Figure 2. Phosphoramidation reactions of the 32P-labeled DNA with 11. Results from one-step phosphoramidation reactions were analyzed by SAv
gel shift assays in (A) 8% urea PAGE for the double-stranded DNA (the TW17 DNA) conjugations or (B) 20% urea PAGE for the single-stranded
DNA (the 3′-primer) conjugations, and visualized by autoradiogram. Results from two-step phosphoramidation reactions were again analyzed by
SAv gel shift assays in (C) 8% urea PAGE for the double-stranded DNA (the TW17 DNA) conjugations or (D) 20% urea PAGE for the single-
stranded DNA (the 3′-primer) conjugations, and also visualized by autoradiogram. The samples in each figure, starting from the left, are the DNA
without the phosphoramidation reaction, the DNA conjugated with 11 by a phosphoramidation reaction without the addition of 6.77 M urea, and
the DNA conjugated with 11 by a phosphoramidation reaction in the presence of 6.77 M urea, respectively. Also in each figure, the top arrow
indicates the location of the DNA-11 conjugate; the bottom arrow represents migration of the DNA without phosphoramidation reactions.
Preparation of Phosphoramidated DNA-11 Conjugates.
One-step and two-step phosphoramidation reactions were carried
out to prepare DNA-11 conjugates and study universal ap-
plications of the phosphoramidation reaction protocol developed
for the TW17 RNA-11 conjugate. The one-step phosphora-
midation reactions for the conjugations of either the single-
stranded 3′-primer or the double-stranded TW17 DNA with 11
in the presence of 6.77 M urea offered a yield of 80% for the
TW17 DNA-11 conjugate and 60% for the 3′-primer DNA-11
conjugate (Figure 2A,B). These yields were comparable to the
79% yield for the TW17 RNA-11 conjugate. The addition of
6.77 M urea significantly improved the yields for the TW17
DNA-11 conjugates from 23%, without urea, to 80% in the
presence of urea. These yields parallel the positive effect of
urea during the phosphoramidation reactions to prepare the
TW17 RNA-11 conjugates and are consistent with the proposed
function of urea to disassemble secondary structures in nucleic
acids and increase the availability of terminal phosphate groups
for phosphoramidation reactions. The positive effect of urea on
phosphoramidation reaction yields was, however, not observed
upon preparation of the 3′-primer DNA-11 conjugates. Slight
deterioration of the phosphoramidation reaction yield for the
3′-primer DNA-11 conjugates was observed in the presence
of urea. This finding indicates very limited secondary structure
formation in a relatively small 3′-primer DNA and adverse
effects of urea on single-stranded DNA phosphoramidation
reactions. In contrast to results of the TW17 RNA phosphora-
midation reactions, DNA degradation was not observed during
the one-step DNA phosphoramidation reactions.
The two-step phosphoramidation reactions for conjugation
of either the 3′-primer or the TW17 DNA with 11 again provided
yields lower than those for the one-step reactions (Figure 2C,D).
Interestingly, both phosphoramidation reactions performed better
in the absence of 6.77 M urea in which the yield for the TW17
DNA-11 conjugate was 38% and that for the 3′-primer
DNA-11 conjugate was 30%. The inhibitory effect of urea on
phosphoramidation reactions was clearly more prominent than
its beneficial effect on denaturation of nucleic acids to disrupt

Nucleic Acid Phosphoramidation and Applications
Bioconjugate Chem., Vol. 21, No. 9, 2010 1649
their secondary structures in the two-step DNA phosphorami-
dation reaction. DNA degradation during the two-step DNA
phosphoramidation reaction was not observed. Overall, we
demonstrated the feasibility of using two phosphoramidation
reaction formats to prepare the DNA-11 conjugates and the
stability of DNA during phosphoramidation reactions.
Preparation of Nucleic Acid-1,6-Hexanediamine Conju-
gates. The success in using 11 to develop effective nucleic acid
phosphoramidation methods encouraged us the further applica-
tions of the phosphoramidation reaction to the preparation of
nucleic acid-1,6-hexanediamine conjugates. These conjugates
can serve as key intermediates for covalently linking chro-
mophores or oligopeptides to nucleic acids. One-step phospho-
ramidation reactions were chosen to attain higher phosphora-
midation yields for nucleic acid-1,6-hexanediamine conjugates.
We were encouraged by observing quantitative production of
the TW17 RNA-1,6-hexanediamine conjugates according to
the same phosphoramidation reaction for the TW17 RNA-11
conjugate (Figure 3A). Quantitative formation of the TW17
DNA-1,6-hexanediamine or the 3′-primer DNA-1,6-hexanedi-
amine conjugates was also attained even under the conditions
of reduced times for EDC activation (15 min) and for 1,6-
hexanediamine coupling (10 min for both DNA molecules) in
the phosphoramidation reactions (Figure 3B,C). Unexpectedly,
multiple conjugations of 1,6-hexanediamine to either the TW17
DNA or the 3′-primer DNA were very significant, but the
conjugates were never observed to produce TW17 RNA-1,6-
hexanediamine conjugates. Substitutions of 1,6-hexanediamine
with ethylenediamine significantly decreased phosphoramidation
reaction yields for all nucleic acid conjugates (Wang et al.,
unpublished results). These results imply that the pKa, lipo-
philicity and diamine characteristics of 1,6-hexadiamine play
an essential role in accomplishing higher yields and multiple
conjugations in nucleic acid phosphoramidation reactions.
Preparation of the 3′-Primer DNA-Fluorescein Conju-
gate. The quantitative 1,6-hexanediamine-3′ primer DNA
conjugates were applied to conjugations with FITC through
covalent thiourea linkages between newly acquired primary
amines in the 5′-termini of nucleic acids and FITC. Conventional
high-resolution urea PAGE methods to determine yields for the
3′-primer DNA conjugate preparation could not be used due to
multiple conjugations of the 3′-primer DNA with 1,6-hexanedi-
amine (Figure 3C). Instead of obtaining quantitative measure-
ments, we pursued qualitative data to demonstrate formation
of the 3′-primer DNA-fluorescein conjugates by applying the
fluorescein-tagged 3′-primer DNA to PCR amplification. For-
mation of the 3′-primer DNA-fluorescein conjugates was
clearly shown in the results of PCR amplification. The fluo-
rescence signals from either the PCR products or the fluorescein-
tagged 3′-primer DNA were vividly observed (Supporting
Information Figure S1). The signals corresponding to the PCR
products were less than 30% of the total fluorescein signal, a
sharp contrast to quantitative yields for the 3′-primer DNA-1,6-
hexanediamine conjugate preparation. Nevertheless, the PCR
results supported formation of the covalently linked 3′-primer
DNA-fluorescein conjugates.
Characteristics of Hybridization Specificity in Nucleic
Acids after Phosphoramidation Reactions. The effectiveness
of POCs as therapeutic reagents strongly depends on intact
hybridization specificity in nucleic acids after preparation. Thus,
assurance of the integrity of hybridization specificity in nucleic
acids after phosphoramidation reactions is essential to POC
preparation reactions. Retention of hybridization specificity in
the 3′-primer DNA after phosphoramidation and followup
coupling reactions was ambiguous according to our qualitative
studies of the 3′-primer DNA-fluorescein conjugate preparation
(Supporting Information Figure S1). The low PCR product yield
Figure 3. Phosphoramidation reactions of the 32P-labeled DNA or RNA
with 1,6-hexanediamine. Products from each one-step phosphorami-
dation reaction were analyzed by (A) 8% urea PAGE for the GMP-
primed TW17 RNA conjugations, (B) 8% urea PAGE for the doubled-
stranded TW17 DNA conjugations, or (C) 20% high-resolution urea
PAGE for the single-stranded 3′-primer DNA conjugations, and
visualized by autoradiogram. The samples in Figure 3A, starting from
the left, are the RNA without phosphoramidation reactions, and
formation of the RNA-1,6-hexanediamine conjugate by the phospho-
ramidation reaction with 60 min EDC activation and 180 min coupling
reactions of with 1,6-hexanediamine in the presence of 6.77 M urea,
respectively. The samples in Figure 3B,C, starting from the left, are
DNA without phosphoramidation reactions, and formation of DNA-1,6-
hexanediamine conjugates by the phosphoramidation reactions with 15
min EDC activation and 5, 10, 20, and 30 min coupling reactions of
1,6-hexanediamine in the presence of 6.77 M urea, respectively. Also
in each figure, the top arrow indicates location(s) of nucleic acid-1,6-
hexanediamine conjugate(s); the bottom arrow represents migration of
the nucleic acid without phosphoramidation reactions.

1650 Bioconjugate Chem., Vol. 21, No. 9, 2010
Wang et al.
DNA-fluorescein conjugate. Since the discrepancy in PCR
product yields was not likely caused by differences between
biotin and fluorescence moieties in nucleic acid conjugates, we
questioned whether changes of PCR amplification efficiency
could be attributed to the extent of multiple conjugations from
phosphoramidation reactions. Consistent with predictions, the
phosphoramidation reaction between the 3′-primer DNA and
11 entailed fewer extents of multiple conjugations, likely a key
factor in sustaining hybridization specificity in nucleic acids
(Figure 4B). Nucleic acid phosphoramidation reactions under
appropriate controls provide an outstanding method to efficiently
and effectively prepare nucleic acid conjugates, including POCs,
without loss of hybridization specificity in nucleic acids
moieties.
Preparation of Nucleic Acid-Tetraglycine (12) Conju-
gates. Understanding the basic chemistry for nucleic acid
phosphoramidation provided essential information to develop
appropriate and simple protocols for POC preparation. Our
results initially implied that one-step phosphoramidation reac-
tions were superior in preparing POCs because of their better
reaction yields (Figures 1 and 2). Phosphoramidation reactions
of DNA with 1,6-hexanediamine, however, revealed that the
one-step approach was too reactive to produce multiple conjuga-
tions in DNA molecules and had adverse effects on hybridization
specificity (Supporting Information Figure S1). In addition,
generic one-step nucleic acid phosphoramidation reactions have
restrictions on nucleophile carbodiimide-sensitive functionalities,
which demands the use of carboxyl-protected peptides to
participate in one-step nucleic acid phosphoramidation reactions
for POC preparation. The need to modify peptides for one-step
nucleic acid phosphoramidation reactions can impair the intrinsic
binding affinity to their natural counterpart receptors on cell
membranes (increase of Kd) to prohibit effective POC uptake
by cells. Furthermore, peptide modifications can transform
peptides into ligands of undesirable receptors on nontargeted
cell membranes to incur side effects or lethal complications in
clinical applications. It is a delicate balance between producing
POCs with high yields and developing POCs for effective
medical treatments with minimum adverse effects.
We anticipated that the combination of two-step nucleic acid
phosphoramidation reactions and natural peptides free from any
chemical modifications could be harnessed to afford POCs
devoid of the drawbacks of using modified peptides. To test
the concept, we synthesized tetraglycine (12) as a model peptide
(26) and allowed it to react with nucleic acids by two-step
phosphoramidation reactions. Consistent with the previous
studies of nucleic acid conjugations with 11 by two-step
phosphoramidation reactions, the acquired nucleic acid-12
conjugates after 3 h of coupling reactions with nucleophile 12
had the following yields: the TW17 RNA-12 conjugate of 44%,
the TW17 DNA-12 conjugate of 39%, and the 3′-primer
DNA-12 conjugate of 27% (Figure 5). The molecular mass
difference between two DNA strands in the double-stranded
TW17 DNA, which is over 500 Da, contributes to the observa-
tion of a dual DNA signal before and after the phosphorami-
dation reaction (Figure 5B; calculated molecular mass: the
positive strand of the TW17 DNA, 36 963.8 g/mol; the negative
strand of the TW17 DNA, 36 448.6 g/mol). The product yields
are comparable to the two-step phosphoramidation reactions for
nucleic acid-11 conjugate preparation. Extending the reaction
time for 12 to overnight (16 h) improved DNA-12 conjugate
yields, but not for the TW17 RNA-12 conjugate. A lower yield
for the TW17 RNA-12 conjugate after the overnight coupling
reaction could be attributed to instability of RNA during a
prolonged coupling period. However, 3 h of reactions with
peptide nucleophiles in two-step nucleic acid phosphoramidation
Figure 4. Demonstration of intact hybridization characteristics and
minor multiple conjugations for the 3′-primer DNA after the phospho-
ramidation reaction with 11. (A) The one-step phosphoramidation-
prepared 3′-primer DNA-11 conjugate (19 pmol) was mixed into the
nonlabeled 3′-primer DNA (15 pmol) and used for PCR reactions. The
samples of the 3′-primer DNA-11 conjugates before (the left sample)
and after (the right sample) a PCR reaction were analyzed by a biphasic
urea PAGE with the top gel of 6% acrylamide and the bottom gel of
20% acrylamide to attain good resolutions. The top arrow indicates
the location of the 119-mer DNA PCR products; the bottom arrow
represents migration of the 3′-primer DNA-11 conjugate. (B) The same
samples in Figure 2B were analyzed by a 20% urea PAGE without the
addition of SAv and visualized by autoradiogram.
could not be attributed to the poor coupling efficiency of FITC
with the 3′-primer DNA-1,6-hexanediamine conjugates or to
changes in hybridization specificity of the 3′-primer DNA-
fluorescein conjugate. We thus applied the 3′-primer DNA-11
conjugate, with a known product yield (Figure 2B), to PCR
reactions to clearly demonstrate intact hybridization specificity.
Results from the PCR reaction indicated that 68% of the
radiolabeled 3′-primer DNA was incorporated into the PCR
DNA product (Figure 4A). As the yield for the 3′-primer
DNA-11 conjugate preparation, in the absence of 6.77 M urea,
was 69% (Figure 2B), we concluded that all of the phospho-
ramidation-prepared 3′-primer DNA-11 conjugates had ef-
fectively hybridized with the provided DNA template to produce
the desired DNA product. The impressive PCR amplification
efficiency of the 3′-primer DNA-11 conjugate is in sharp
contrast to the poor PCR amplification efficiency of the 3′-primer

Nucleic Acid Phosphoramidation and Applications
Bioconjugate Chem., Vol. 21, No. 9, 2010 1651
Phosphoramidation Reactions for Preparation of Nucleic
Acid-BSA Conjugates. We further explored the use of
phosphoramidation reactions to develop a generic approach for
the preparation of nucleic acid-protein conjugates. Starting with
the previously prepared nucleic acid-1,6-hexanediamine con-
jugates (Figure 3), they were first reacted with either DSS or
DSG to effectively convert -NH2 to the NHS ester of a
carboxylic acid in 5′-termini of the nucleic acids (33). Activated
acyl group characters of NHS esters allowed easy formation of
new amide linkages between nucleic acids and proteins such
as BSA to acquire nucleic acid-protein conjugates. Incorpora-
tion of 3 M NaCl in coupling reactions was proposed to be
indispensible to diminish electric repulsions between nucleic
acids and proteins to attain higher yields for nucleic acid-protein
conjugates (33). Therefore, we performed the coupling reactions
in the presence or absence of 3 M NaCl to ascertain any benefit.
The yields for the TW17 RNA-BSA conjugates were only
8% using DSS linkers and 22% using DSG linkers after 2 h
reactions in the absence of 3 M NaCl (Figure 6A). Extension
of the reaction time from 2 to 16 h, however, improved the
yields for the TW17 RNA-BSA conjugates to 44% using DSS
linkers and 34% using DSG linkers. Despite neutral pH reaction
conditions (pH ) 7.4), RNA suffered severe degradation for
both DSS and DSG linker conjugates after prolonged coupling
reactions. Preparation of the TW17 RNA-BSA conjugates after
2 h reactions in the presence of 3 M NaCl offered reasonable
good yields by both DSS and DSG linkers and was comparable
to the overnight (16 h) coupling reactions in the absence of 3
M NaCl (Figure 6B). Unexpectedly, extending the reaction time
to 16 h was not necessary to improve reaction yields. The loss
of the TW17 RNA-BSA conjugates during overnight reactions
was observable; RNA degradation was again prevalent and more
intensive during the longer reaction time. Overall, additions of
3 M NaCl when preparing the TW17 RNA-BSA conjugates,
although not significantly increasing the conjugate yields,
reduced the coupling time from 16 to 2 h, which is critical in
preventing RNA degradation during POCs preparation.
Similar coupling yields were never obtained for the prepara-
tion of the TW17 DNA-BSA conjugates. Yields were less than
20% when using DSS or DSG linkers in the absence of 3 M
NaCl (Supporting Information Figure S2A). No improvements
in yields for the coupling reactions between BSA and the TW17
DNA derivatives were observed when the reaction times were
increased from 2 to 16 h or for reactions carried out in the
presence of 3 M NaCl (Supporting Information Figure S2B).
Such results defy the beneficial effects of prolonged reaction
times and the use of 3 M NaCl to enhance the formation of
DNA-alkaline phosphatase conjugates (33). The studies of
nucleic acid-BSA conjugate preparations underscore the fun-
damental difference between DNA and RNA toward phospho-
ramidation reactions which significantly effect yields of nucleic
acid-BSA conjugates. Yields for all phosphoramidation reac-
tions are provided in Table 1.
Figure 5. Phosphoramidation reactions of the 32P-labeled DNA or RNA
with tetraglycine (12) to prepare nucleic acid-tetraglycine conjugates.
Products from each two-step phosphoramidation reaction were analyzed
by (A) 8% high-resolution urea PAGE for the GMP-primed TW17 RNA
conjugations, (B) 8% high-resolution urea PAGE for the doubled-
stranded TW17 DNA conjugations, or (C) 20% high-resolution urea
PAGE for the single-stranded 3′-primer DNA conjugations, and
visualized by autoradiogram. The samples in each figure, starting from
the left, are the nucleic acid without phosphoramidation reactions, the
nucleic acid conjugated with 12 by phosphoramidation reactions without
the addition of 6.77 M urea, and the nucleic acid conjugated with 12
by phosphoramidation reactions in the presence of 6.77 M urea,
respectively. Also in each figure, the top arrow indicates the location
of the nucleic acid-12 conjugate; the bottom arrow represents migration
of the original nucleic acid.
DISCUSSION
We have successfully synthesized a new amine-containing
biotin derivative 11 suitable for conjugation with biomolecules
such as proteins or nucleic acids for versatile applications. We
harnessed the amine functionality and the biotin moiety in 11
to develop the phosphoramidation reactions for preparation of
nucleic acid-11 conjugates with excellent yields and efficiency,
but without compromising hybridization specificity of nucleic
acids. Both two-step and one-step phosphoramidation reactions
were studied, and as expected, one-step phosphoramidation
reactions always performed better than two-step formats, which
can be attributed to phosphorimidazolide intermediate lability
in aqueous solutions. A one-step phosphoramidation reaction
is enough to prepare suitable amounts of POCs. No multiple
conjugations were observed in any nucleic acid-12 conjugates,
implying the integrity of hybridization specificity in nucleic acid
moieties. These results clearly validate the two-step phospho-
ramidation reaction to prepare appropriate POCs for medical
applications.

1652 Bioconjugate Chem., Vol. 21, No. 9, 2010
Wang et al.
research laboratories and will enable POCs to be used for
important therapeutic and scientific applications such as gene
therapy, gene silencing, or RNA interference (24) (Scheme 2).
Besides the facile approach and suitable yields, the hallmark
of two-step phosphoramidation reactions for POC preparation
is the general requirements of peptides and nucleic acids
involved in reactions. Essentially, any nucleic acids with 5′-
phosphate groups and any peptides with no previous modifica-
tion are suitable reactants in phosphoramidation reactions to
prepare desired POCs. A prerequisite of 5′-phosphate groups
in nucleic acids ensures that any nucleic acid, ranging from DNA
to RNA and natural to artificial nucleotides, will participate in
phosphoramidation reactions. Natural or synthetic peptides with
no protection groups can follow two-step phosphoramidation
reactions to conjugate with nucleic acids to afford POCs. There
is no need to protect carbodiimide-sensitivity functionality in
peptides prior to coupling reactions. Peptides with any composi-
tions of amino acid residues and no previous modifications can
participate in phosphoramidation reactions to obtain POCs.
Furthermore, two-step phosphoramidation reactions can be the
only method to acquire POCs when the introduction of
protection group(s) to peptides severely deteriorates binding
affinities between modified peptides and their receptors to inhibit
effective uptake of POCs into cells (vide infra). However, two-
step phosphoramidation reactions have disadvantages in a lower
coupling efficiency than the one-step approach. The disadvan-
tage was demonstrated by the two-step phosphoramidation
reactions between 11 and nucleic acids (Figures 1 and 2). We
are exploring additional approaches to enhance the half-life of
phosphorimidazolide intermediates during two-step phospho-
ramidation reactions to improve POC yields.
One-step nucleic acid phosphoramidation reactions with
higher product yields remain an effective approach to acquire
POCs under selected circumstances. As previously noted, the
validity of one-step phosphoramidation reactions for POC
preparation is warranted only if we can demonstrate that
structure-modified peptides preserve the same levels of binding
affinity and specificity to their intrinsic receptors. It is thus
essential to determining the difference in dissociation constants
Kd between protected and unprotected peptides when adsorbed
to specific protein receptors before one-step phosphoramidation-
prepared POCs can be exploited for in vivo applications and
biomedical studies. The assurance of the integrity of hybridiza-
tion specificity for nucleic acid moieties in one-step phospho-
ramidation-prepared POCs is especially important. The findings
presented on one-step phosphoramidation reactions indicate that
these reactions are very reactive to attain higher product yields
and multiple conjugations, especially in cases of DNA (Figure
3). Significant multiple conjugations by one-step phosphora-
midation reactions were particularly problematic to DNA in
which extents of multiple conjugations paralleled a loss of
hybridization specificity in DNA (Figure 4 and Supporting
Information Figure S1). In essence, one-step nucleic acid
phosphoramidation reactions require appropriate controls of the
reaction conditions to obtain high product yields and minimize
multiple conjugations to retain integrity of hybridization speci-
ficity in nucleic acids.
Figure 6. Preparation of the TW17 RNA-BSA conjugates. The 32P-
labeled GMP-primed TW17 RNA was conjugated with 1,6-hexanedi-
amine using the one-step phosphoramidation reaction stated previously,
followed by reactions with DSS or DSG and finally coupled with BSA
(A) in the absence of 3 M NaCl or (B) in the presence of 3 M NaCl.
The samples in both (A) and (B) were analyzed by 8% urea PAGE
and visualized by autoradiogram. The samples in each figure, starting
from the left, are the TW17 RNA without any modification reactions
and formation of the TW17 RNA-BSA conjugates by using the DSS
linker with 2 h coupling reactions, the DSG linker with 2 h coupling
reactions, the DSS linker with 16 h coupling reactions, and the DSG
linker with 16 h coupling reactions, respectively. The top arrow indicates
the location of the TW17 RNA-BSA conjugate; the middle arrow
stands for the migration of the TW17 RNA conjugates before coupling
with BSA; the bottom arrow represents migration of the original GMP-
primed TW17 RNA before phosphoramidation reactions.
is recommended when higher conjugation yields are a priority.
Two-step phosphoramidation reactions, in spite of lower reaction
yields, can be an advantageous approach. Two-step formats
circumvent the problems of nucleophiles when containing a
carbodiimide-sensitive functionality and the intricate phenomena
of multiple conjugations, which deteriorate hybridization speci-
ficity in nucleic acids.
The outstanding properties of phosphoramidation reactions
demonstrate their potential to afford diverse POCs not limited
by sequences of peptides and nucleic acids or by properties of
nucleic acids as DNA or RNA. We demonstrated the universal
applications of this approach by effectively conjugating the
model peptide tetraglycine (12) to RNA, single-stranded DNA,
or doubled-stranded DNA with reasonably good yields. The
facile procedures of two-step phosphoramidation reactions will
alleviate difficulties in preparing POCs in typical biomedical
The cause of multiple conjugations in nucleic acids during
phosphoramidation reactions can be traced back to the structures
of nucleic acids and nucleophiles. The findings presented
indicate that DNA molecules have a far greater tendency to
experience multiple conjugations during phosphoramidation
reactions than RNA. The difference between DNA and RNA is
the presence of 2′-hydroxyl group in RNA and an extra methyl
group in thymine of DNA when compared to uracil in RNA.
2′-Hydroxyl groups in RNA have no reason to be activated by
EDC and react with nucleophiles to contribute multiple conjuga-

Nucleic Acid Phosphoramidation and Applications
Bioconjugate Chem., Vol. 21, No. 9, 2010 1653
Table 1. Yields for Nucleic Acid Conjugates Directly or Indirectly Prepared from Phosphoramidation Reactions in Current Studies
nucleophiles
TW17 RNA
TW17 DNA
3′ primer
11 (one-step phosphoramidation)
11 (two-step phosphoramidation)
12 (two-step phosphoramidation)
1,6-hexanediamine (one-step phosphoramidation)
BSA (one-step phosphoramidation)f
BSA (one-step phosphoramidation)l
70%a, 79%b
23%a, 80%b
69%a, 60%b
30%a, 21%b
27%c, 34%d
100%e
40%a, 52%b
38%a, 32%b
44%c, 40%d
39%c, 42%d
100%
100%
8%g, 44%h, 45%i, 26%j
22%g, 34%h, 32%i, 44%j
5%g, 5%h, 2%i, 4%j
10%g, 12%h, 5%i, 7%j
N/Ak
N/Ak
a In the absence of 6.77 M urea. b In the presence of 6.77 M urea. c For a 3 h phosphoramidation reaction. d For a 16 h phosphoramidation reaction.
e Under the conditions of 15 min EDC activation and 20 min 1,6-hexanediamine coupling reactions. f Using DSS as the linker. g For a 2 h coupling
reaction with BSA in the absence of 3 M NaCl. h For a 16 h coupling reaction with BSA in the absence of 3 M NaCl. i For a 2 h coupling reaction with
BSA in the presence of 3 M NaCl. j For a 16 h coupling reaction with BSA in the presence of 3 M NaCl. k Not analyzed. l Using DSG as the linker.
Scheme 2. Versatile Applications of Nucleic Acid Phosphoramidation Reactions in Science and Medicinea
a dsDNA, double-stranded DNA; ssDNA, single-stranded DNA.
tions. It is likely that the presence of the thymine base is one
reason to confer multiple conjugations in DNA during phos-
phoramidation reactions. The ability of the extra methyl group
in uracil to transform thymine in DNA into a relatively good
electrophile in phosphoramidation reactions remains unan-
swered. The other factor to bestow multiple conjugations in
nucleic acids during phosphoramidation reactions is the pKa of
the amine moieties in nucleophiles and the structures of
nucleophiles, particularly their lipophilicity and NH2 functional-
ity contents. According to the results from nucleic acid
phosphoramidation reactions with 11, 12, ethylenediamine, and
1,6-hexanediamine, the increase of amine pKa, lipophilicity or
NH2 contents in nucleophiles promotes multiple conjugations
in nucleic acids. We thus observed the vigorous activity of 1,6-
hexanediamine in phosphoramidation reactions to produce
multiple conjugations in nucleic acids (Figure 3). Appropriate
designs of nucleophile lipophilicity and NH2 contents can afford
high yields of nucleic acid conjugates while avoiding multiple
conjugations in nucleic acids during one-step phosphoramidation
reactions.
The problem of multiple conjugations in nucleic acids during
phosphoramidation reactions is the negative effect on nucleic
acid hybridization specificity manifested by the single-stranded
DNA-fluorescein conjugate studies. The decline of nucleic acid
hybridization specificity is intimately related to the extent of
multiple conjugations in nucleic acid, especially modifications
which damage the base-pairing property of thymine. The extents
of multiple conjugations may also be a determinant to contribute
to lower yields of nucleic acid-BSA conjugates (Figure 6 and
Supporting Information Figure S2). These results contradict our
earlier assumption that more conjugations in nucleic acids would
provide more reactive groups to aid in covalently linking BSA
to nucleic acids and thereby increasing yields of nucleic
acid-BSA conjugates. In contrast, the extent of multiple
conjugations in phosphoramidation-prepared nucleic acid con-
jugates had the opposite effect on yields of nucleic acid-BSA
conjugates (Figure 6 and Supporting Information Figure S2).
The TW17 RNA-1,6-hexanediamine conjugates with a minor
multiple conjugations offered better coupling efficiency with
BSA compared to the same coupling reactions using DNA-1,6-
hexanediamine conjugates with more extensive multiple con-
jugations. The actual cause of higher extents of multiple
conjugations in DNA to render low yields of DNA-BSA
conjugates is under investigation currently. However, results
from these nucleic acid-BSA conjugate preparations suggest
that two-step phosphoramidation reaction schemes are a desir-
able approach to acquire DNA-protein conjugates with better
yields.
The controversial role of high concentration of urea in
nucleic acid phosphoramidation reaction was intriguing. Urea
of 6.77 M was introduced to nucleic acid phosphoramidation
reactions to disrupt secondary structures in nucleic acids and
to expose 5′-phosphate groups for phosphoramidation reac-
tions. Urea in high concentrations is a common approach in
biochemical research to denature nucleic acids because of
its typical inertness toward chemical reactions. We thus
expected to have higher yields for nucleic acid phosphora-
midation reactions in the presence of 6.67 M urea, yet only
RNA molecules had better phosphoramidation reaction yields
with additions of 6.77 M urea. Furthermore, high concentra-
tions of urea had definite negative effects on two-step DNA
phosphoramidation reactions (Figures 1B and 2C,D). The
unexpected results can be explained by the ambient nucleo-
phile characteristics in urea (34-36). The current results
suggested that urea plays a dual role in nucleic acid
phosphoramidation reactions: one as a surfactant to denature
nucleic acids, and the other as a nucleophile to divert
phosphoramidation reactions and to reduce product yields.
The observed product yield from nucleic acid phosphorami-
dation reactions is the competition between two counteracting
activities. The greater adverse effect of urea in DNA
phosphoramidation reactions can be attributed to the presence
of a more reactive thymine base in DNA which reacts with
urea to decrease product yields. The negative effect of urea
in two-step DNA phosphoramidation reactions is further
ascribed to the hydrolysis of phosphorimidazolide intermedi-
ates of nucleic acids during the workup procedures. We
conclude that high concentrations of urea will have beneficial

1654 Bioconjugate Chem., Vol. 21, No. 9, 2010
Wang et al.
(10) Ikawa, Y., Moriyama, S., and Furuta, H. (2008) Facile
syntheses of BODIPY derivatives for fluorescent labeling of the
3′ and 5′ ends of RNAs. Anal. Biochem. 378, 166–170.
(11) Tinsley, R. A., Furchak, J. R. W., and Walter, N. G. (2007)
Trans-acting glmS catalytic riboswitch: locked and loaded. RNA
13, 468–477.
(12) Smith, L. M., Sanders, J. Z., Kaiser, R. J., Hughes, P., Dodd,
C., Connell, C. R., Heiner, C., Kent, S. B. H., and Hood, L. E.
(1986) Fluorescence detection in automated DNA sequence
analysis. Nature 321, 674–679.
(13) Prober, J., Trainor, G., Dam, R., Hobbs, F., Robertson, C.,
Zagursky, R., Cocuzza, A., Jensen, M., and Baumeister, K.
(1987) A system for rapid DNA sequencing with fluorescent
chain-terminating dideoxynucleotides. Science 238, 336–341.
(14) Niemeyer, C. M., and Blohm, D. (1999) DNA Microarrays.
Angew. Chem., Int. Ed. 38, 2865–2869.
(15) Lu, K., Duan, Q.-P., Ma, L., and Zhao, D.-X. (2010) Chemical
strategies for the synthesis of peptide-oligonucleotide conjugates.
Bioconjugate Chem. 21, 187–202.
(16) Eritja, R. (2007) Solid-phase synthesis of modified oligo-
nucleotides. Int. J. Pept. Res. Ther. 13, 53–68.
(17) Zatsepin, T. S., Turner, J. J., Oretskaya, T. S., and Gait, M. J.
(2005) Conjugates of oligonucleotides and analogues with cell
penetrating peptides as gene silencing agents. Curr. Pharm. Des.
11, 3639–3654.
(18) Fraley, A. W., Pons, B., Dalkara, D., Nullans, G., Behr, J.-
P., and Zuber, G. (2006) Cationic oligonucleotide-peptide
conjugates with aggregating properties enter efficiently into cells
while maintaining hybridization properties and enzymatic rec-
ognition. J. Am. Chem. Soc. 128, 10763–10771.
(19) Rogers, F. A., Manoharan, M., Rabinovitch, P., Ward, D. C.,
and Glazer, P. M. (2004) Peptide conjugates for chromosomal
gene targeting by triplex-forming oligonucleotides. Nucleic Acids
Res. 32, 6595–6604.
effects only when applied to RNA phosphoramidation reac-
tions. The substitution of urea with more inert surfactants to
increase one-step and two-step nucleic acid phosphoramida-
tion reaction product yields is being undertaken.
In summary, we successfully synthesized a new biotin
derivative 11 which was exploited to develop facile one-
step and two-step nucleic acid phosphoramidation reactions
with excellent yields. One-step phosphoramidation reactions
offered better yields of nucleic acid conjugates than two-
step phosphoramidation approaches. One-step phosphorami-
dation reactions were, however, more reactive than two-step
phosphoramidation reactions and resulted in multiple con-
jugations accompanied by deterioration of hybridization
specificity in nucleic acids. Multiple conjugations in phos-
phoramidation-prepared nucleic acid conjugates were most
apparent in DNA and nucleophiles contained multiple NH2
functional groups. Understanding the basic chemistry under-
lying phosphoramidation reactions allowed an effective
application of the reactions to nucleic acid-biotin and nucleic
acid-protein conjugate preparations for biomolecule detec-
tion and biomedical research (Scheme 2). Most importantly,
the phosphoramidation approach provides versatile POCs
without scarifying hybridization specificity in nucleic acids.
Facile phosphoramidation reactions to achieve effective
POC preparation can facilitate therapeutic applications of
POCs and fulfill the goals of using POCs for RNA interfer-
ence, gene silencing, and gene therapy in science and
medicine.
ACKNOWLEDGMENT
We thank Dr. Susan Fetzer for critical reading of the
manuscript. This work was supported with grants from the
National Science Council of Taiwan (96-2113-M-037-004-MY2
and 98-2113-M-037-002-) awarded to T.-P. W.
(20) Kwok, K. Y., Park, Y., Yang, Y., Mckenzie, D. L., Liu, Y.,
and Rice, K. G. (2003) In vivo gene transfer using sulfhydryl
cross-linked PEG-peptide/glycopeptide DNA co-condensates.
J. Pharm. Sci. 92, 1174–1185.
(21) Chu, B. C., Wahl, G. M., and Orgel, L. E. (1983) Derivati-
zation of unprotected polynucleotides. Nucleic Acids Res. 11,
6513–6529.
Supporting Information Available: Full details of a reaction
scheme, NMR and MS spectra for compounds 2-11, and
supporting figures referenced in the text. This material is
available free of charge via the Internet at http://pubs.acs.org.
(22) Anderson, G. W., Zimmerman, J. E., and Callahan, F. M.
(1964) The use of esters of N-hydroxysuccinimide in peptide
synthesis. J. Am. Chem. Soc. 86 1839–1842.
(23) Qin, P. Z., and Pyle, A. M. (1999) Site-specific labeling of
RNA with fluorophores and other structural probes. Methods 18,
60–70.
(24) Achenbach, T. V., Brunner, B., and Heermeier, K. (2003)
Oligonucleotide-based knockdown technologies: antisense versus
RNA interference. ChemBioChem 4, 928–935.
(25) Berg, T. A., v, d., Feringa, B. L., and Roelfes, G. (2007)
Double strand DNA cleavage with a binuclear iron complex.
Chem. Commun. 180–182.
LITERATURE CITED
(1) Seelig, B., and Jaschke, A. (1999) A small catalytic RNA motif
with Diels-Alderase activity. Chem. Biol. 6, 167–176.
(2) Sengle, G., Eisenfuhr, A., Arora, P. S., Nowick, J. S., and
Famulok, M. (2001) Novel RNA catalysts for the Michael
reaction. Chem. Biol. 8, 459–473.
(3) Fusz, S., Eisenfuhr, A., Srivatsan, S. G., Heckel, A., and
Famulok, M. (2005) A ribozyme for the aldol reaction. Chem.
Biol. 12, 941–950.
(4) Tsukiji, S., Pattnaik, S. B., and Suga, H. (2003) An alcohol
dehydrogenase ribozyme. Nat. Struct. Biol. 10, 713–717.
(5) Tsukiji, S., Pattnaik, S. B., and Suga, H. (2004) Reduction of
an aldehyde by a NADH/Zn2+-dependent redox active ribozyme.
J. Am. Chem. Soc. 126, 5044–5045.
(26) Boger, D. L., Zhou, J., Winter, B., and Kitos, P. A. (1995)
Key analogs of the tetrapeptide subunit of RA-VII and deoxy-
bouvardin. Bioorg. Med. Chem. 3, 1579–1593.
(27) Bayer, E., and Wilchek, M. (1974) Insolubilized biotin for
the purification of avidin. Methods Enzymol. 34, 265–267.
(28) Wilchek, M., and Bayer, E. A. (1990) Biotin-containing
reagents. Methods Enzymol. 184, 123–138.
(6) Pereira, M. J. B., Nikolova, E. N., Hiley, S. L., Jaikaran, D.,
Collins, R. A., and Walter, N. G. (2008) Single VS ribozyme
molecules reveal dynamic and hierarchical folding toward
catalysis. J. Mol. Biol. 382, 496–509.
(29) Callahan, J. F., Ashton-Shue, D., Bryan, H. G., Bryan, W. M.,
Heckman, G. D., Kinter, L. B., McDonald, J. E., Moore, M. L.,
Schmidt, D. B., Silvestri, J. S., et al. (1989) Structure-activity
relationships of novel vasopressin antagonists containing C-
terminal diaminoalkanes and (aminoalkyl)guanidines. J. Med.
Chem. 32, 391–396.
(7) Boots, J. L., Canny, M. D., Azimi, E., and Pardi, A. (2008)
Metal ion specificities for folding and cleavage activity in the
Schistosoma hammerhead ribozyme. RNA 14, 2212–2222.
(8) Mortimer, S. A., and Weeks, K. M. (2009) C2′-endo nucleotides
as molecular timers suggested by the folding of an RNA domain.
Proc. Natl. Acad. Sci. U.S.A. 106, 15622–15627.
(9) Shi, X., Mollova, E. T., Pljevaljci, G., Millar, D. P., and
Herschlag, D. (2009) Probing the dynamics of the P1 helix within
the Tetrahymena group I intron. J. Am. Chem. Soc. 131, 9571–
9578.
(30) Laurent, S., Botteman, F., Vander Elst, L., and Muller, R. N.
(2004) Relaxivity and transmetallation stability of new benzyl-
substituted derivatives of gadoliniumDTPA complexes. HelV.
Chim. Acta 87, 1077–1089.

Nucleic Acid Phosphoramidation and Applications
Bioconjugate Chem., Vol. 21, No. 9, 2010 1655
(31) Srinivasan, N., Yurek-George, A., and Ganesan, A. (2005)
Rapid deprotection of N-Boc amines by TFA combined with
freebase generation using basic ion-exchange resins. Mol. DiVers.
9, 291–293.
(32) Hermanson, G. T. (1996) Bioconjugate techniques, Academic
Press, San Diego, pp 670-671.
(33) Jablonski, E., Moomaw, E. W., Tullis, R. H., and Ruth, J. L.
(1986) Preparation of oligodeoxynucleotide-alkaline phosphatase
conjugates and their use hybridization probes. Nucleic Acids Res.
14, 6115–6128.
(35) Kluger, R., and Adawadkar, P. D. (1976) A reaction proceed-
ing through intramolecular phosphorylation of a urea. A chemical
mechanism for enzymic carboxylation of biotin involving cleav-
age of adenosine 5′-triphosphate. J. Am. Chem. Soc. 98, 3741–
3742.
(36) Tamaru, Y., Hojo, M., Higashimura, H., and Yoshida, Z.
(1988) Urea as the most reactive and versatile nitrogen
nucleophile for the palladium(2+)-catalyzed cyclization of
unsaturated amines. J. Am. Chem. Soc. 110, 3994–4002.
(34) Bar, G. L. J., Lloyd-Jones, G. C., and Booker-Milburn, K. I.
(2005) Pd(II)-catalyzed intermolecular 1,2-diamination of con-
jugated dienes. J. Am. Chem. Soc. 127, 7308–7309.
BC1001505