Advanced aqueous-phase phosphoramidation reactions for effectively synthesizing peptide-oligonucleotide conjugates trafficked into a human cell line

Article

pubs.acs.org/bc

Advanced Aqueous-Phase Phosphoramidation Reactions for

Eﬀectively Synthesizing Peptide−Oligonucleotide Conjugates

Traﬃcked into a Human Cell Line

Tzu-Pin Wang,*^,†Ni Chien Ko,^†Yu-Chih Su,^†Eng-Chi Wang,^†Scott Severance,^‡Chi-Ching Hwang,^§

Ying Ting Shih,^⊥Min Hui Wu,^†and Yen-Hsu Chen*^,⊥,∥

§

⊥

^†Department of Medicinal and Applied Chemistry, Department of Biochemistry, Faculty of Medicine, School of Medicine,

∥

Graduate Institute of Medicine, College of Medicine, and Division of Infectious Diseases, Department of Internal Medicine,

Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, 80708, Taiwan

^‡Department of Science and Mathematics, Baptist Bible College, S. Abington Township, Pennsylvania 18411, United States

S

* Supporting Information

ABSTRACT: Peptide−oligonucleotide conjugates (POCs) have

held promise as eﬀective therapeutic agents in treating microbial

infections and human genetic diseases including cancers. In

clinical applications, POCs are especially useful to circumvent

cellular delivery and speciﬁcity problems of oligonucleotides. We

previously reported that nucleic acid phosphoramidation

reactions performed in aqueous solutions have the potential for

facile POC synthesis. Here, we carried out further studies to

signiﬁcantly improve aqueous-phase two-step phosphoramida-

tion reaction yield. Optimized reactions were employed to

eﬀectively synthesize POCs for delivery into human A549 cells.

We achieved optimization of aqueous-phase two-step phosphor-

amidation reaction and improved reaction yield by (1)

determining appropriate co-solutes and co-solute concentrations to acquire higher reaction yields, (2) exploring a diﬀerent

nucleophilicity of imidazole and its derivatives to stabilize essential nucleic acid phosphorimidazolide intermediates prior to POC

formation, and (3) enhancing POC synthesis by increasing reactant nucleophilicity. The advanced two-step phosphoramidation

reaction was exploited to eﬀectively conjugate a well-studied cell penetrating peptide, the Tat₄₈₋₅₇peptide, with oligonucleotides,

bridged by either no linkers or a disulﬁde-containing linker, to have the corresponding POC yields of 47−75%.

Phosphoramidation-synthesized POCs showed no cytotoxicity to human A549 cells at studied POC concentrations after 24

h inoculation and were successfully traﬃcked into the human A549 cell line as demonstrated by ﬂow cytometry, ﬂuorescent

microscopy, and confocal laser scanning microscopy study. The current report provides insight into aqueous-phase

phosphoramidation reactions, the knowledge of which was used to develop eﬀective strategies for synthesizing POCs with crucial

applications including therapeutic agents for medicine.

The direct use of nucleic acids for treating diseases, however,

faces serious hurdles. Diﬃculties include cell speciﬁcity,

ineﬃcient cellular uptake of nucleic acids, and inaccessibility

of nucleic acids to cell nuclei, due primarily to ineﬀective

translocation of nucleic acids across biological barriers after

administration.^6,7Consequently, successful use of nucleic acids

in clinical practice will not be achieved until there are better

strategies for targeted and eﬃcient delivery of nucleic acids to

cells and tissues.⁴⁻⁷The critical issue of eﬃcient target delivery

for nucleic acids has been studied by many laboratories through

chemical modiﬁcation of nucleic acids to improve stability and

cellular delivery properties of nucleic acids in vivo. Developed

INTRODUCTION

■

Extensive research endeavors have been devoted to exploring

the therapeutic potential of nucleic acids, including siRNA,

miRNA, catalytic RNA (ribozymes), aptamer oligonucleotides

(oligonucleotides with exquisite roles similar to protein

receptors), and antisense oligonucleotides.¹⁻⁷Theoretically,

when designed appropriately, nucleic acids delivered into

biological systems will participate in cellular activities, such as

RNA interference or gene silencing, to abolish speciﬁc gene

expression in cells and to attain more precise therapeutic

targeting than typical small molecule drugs. Nucleic acid-based

therapeutics have shown promise for treating a variety of

human genetic diseases and microbial infections. Recent

progress has resulted in some antisense oligonucleotides and

aptamer RNA reaching clinical applications, while a signiﬁcant

number of clinical trials for siRNA are underway.^1,3,5,6

Received: August 14, 2012

Revised: November 3, 2012

A

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

methods include conjugating cellular surface receptor-speciﬁc

ligands with various nucleic acid-containing nanocarriers or

covalently linking cellular surface receptor-speciﬁc ligands with

nucleic acids directly.⁷⁻¹¹

synthesize bioactive POCs which can stimulate broader

applications of POCs in research and clinics.

MATERIALS AND METHODS

■

The use of peptides as ligands to traﬃc nucleic acids across

the plasma membrane has been extensively investigated in the

development of eﬀective nucleic acid-based therapeutic

agents.^7,8,11,12Conjugating nucleic acids such as oligonucleo-

tides with cell-penetrating peptides (CPPs) or cell-targeting

peptides (CTPs) to acquire peptide−oligonucleotide conju-

gates (POCs) has created appropriate designs to circumvent

cellular delivery or cell speciﬁcity problems inherited from

administrating only oligonucleotides in clinics.^7,11,12CPPs,

including the Tat, the Antennapedia, the CyLoP-1, and the

(KFF)₃K peptides, are either protein-derived or artiﬁcially

developed short sequences (10−16 amino acids), and they are

able to spontaneously cross cellular barriers when provided in

extracellular media.^8,12−18The unique cell permeability proper-

ties of CPPs signiﬁcantly improve the uptake eﬃciency of

oligonucleotides in POCs by cells and facilitate broader uses of

POCs in science and medicine.

POCs are primarily prepared by coupling peptides with

oligonucleotides after solid-phase synthesis (fragment coupling

strategy) or directly synthesized through stepwise solid-phase

reactions (online solid-phase synthesis).¹⁹To achieve peptide−

oligonucleotide conjugations, current POC synthesis methods

typically require previous incorporations of additional function-

alities in peptides, oligonucleotides, or both. The requirement

to have additional functional groups renders these methods

ineﬃcient, inconvenient, and not cost-eﬀective for academia or

industry. The development of a facile approach to exploit

readily available functionalities, such as hydroxyl or phosphate

groups, in standard oligonucleotides in order to synthesize

POCs with high purity and yields is crucial to the advancement

of POC applications.

We previously demonstrated that aqueous-phase two-step

phosphoramidation reactions are an ideal strategy to synthesize

POCs without compromising oligonucleotide base-pairing

speciﬁcity, an essential criterion when administrating oligonu-

cleotides as therapeutic agents.²⁰The original two-step

phosphoramidation reaction easily accommodates multiple

active functionalities in peptides for POC synthesis; however,

it suﬀers from a drawback of relatively low POC yield, typically

ranging from 27% to 44%. Low POC yield in the original two-

step phosphoramidation reaction was ascribed to the interfering

presence of the co-solute urea, the instability of nucleic acid

phosphorimidazolide intermediates in aqueous solutions during

reactions, and peptide nucleophilicity.²⁰

Here, we report further study of the two-step phosphor-

amidation reaction to glean essential information for improving

product yield and synthesizing POCs more eﬀectively. We

successfully employed the optimized two-step phosphoramida-

tion reaction to eﬀectively conjugate a well-studied CPP, the

TAT₄₈₋₅₇peptide,^12,18with DNA or RNA to produce

corresponding POC yields of 47−75%. Phosphoramidation-

synthesized POCs showed no cytotoxicity to human A549 cells

at studied POC concentrations after 24 h inoculation and were

successfully traﬃcked into the human A549 cell line as

demonstrated by ﬂow cytometry, ﬂuorescent microscopy, and

confocal laser scanning microscopy study. The following report

conﬁrms that the optimized two-step aqueous-phase phosphor-

amidation reaction is a facile and eﬀective approach to

All reagent-grade chemicals were purchased from commercial

sources (Sigma-Aldrich, Acros, Alfa Aesar, Mallinckrodt Baker,

and Merck KGaA) except where noted, and were further

puriﬁed if necessary. Synthesis of 1 was reported previously.²⁰

Procedures modiﬁed from a published method were used to

synthesize H₂N-Gly-Gly-Gly-Gly-OH (tetraglycine; 3).²¹Com-

pound 3 was also obtained from commercial sources (Acros

and TCI) to avoid the time-consuming organic synthesis. Five

peptides (Tat₄₈₋₅₇, H₂N-GRKKRRQRRR-CONH₂; Tat₄₇₋₅₇

,

H₂N-YGRKKRRQRRR-CONH₂; (KFF)₃K, H₂N-

KFFKFFKFFK-CONH₂; G(KFF)₃K, H₂N-GKFFKFFKFFK-

CONH₂; C(KFF)₃K, H₂N-CKFFKFFKFFK-CONH₂) were

purchased from Peptide 2.0 (Chantilly, VA, USA) and used

1

without further puriﬁcation. H 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 shifts of ¹H 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). FAB

high-resolution mass spectra were obtained from Department

of Chemistry, National Cheng Kung University, Taiwan, on a

JEOL JMS-700 instrument (JEOL Ltd., Tokyo, Japan).

Radiolabeled nucleic acid conjugates were analyzed by urea

polyacrylamide gel electrophoresis (urea-PAGE) or streptavidin

(SAv) gel shift assay in urea-PAGE, visualized and quantiﬁed by

an Amersham Typhoon PhosphorImager (GE Healthcare Bio-

Sciences AB, Uppsala, Sweden) at Kaohsiung Medical

University, Taiwan. Molecular mass of POCs was measured

by using an Autoﬂex III TOF/TOF analyzer (Bruker Daltonics,

Taiwan) available at Kaohsiung Medical University, Taiwan.

POC uptake by human A549 cells was analyzed by a BD

FACSCalibur cytometer (BD, Franklin Lakes, NJ, USA), a

Nikon ECLIPSE Ti−U Inverted microscope (Nikon, Shinjuku,

Tokyo, Japan) and a FluoView 1000 confocal laser scanning

microscope (Olympus, Tokyo, Japan). These instruments were

located at Kaohsiung Medical University, Taiwan.

Synthesis of the Diamino Derivatives of Tetraglycine

(7). {2-[2-(2-tert-Butoxycarbonylamino-acetylamino)-acety-

lamino]-acetylamino}-acetic Acid (4). Tetraglycine (3; 50 mg,

0.2 mmol) was dissolved in 16 mL of a dioxane/NaOH/water

(4:2:2) solution to which was slowly added di-tert-butyldicar-

bonate (355 mg, 1.6 mmol) within 3 min and reacted at rt for 2

h. The ﬁnal reaction mixture was concentrated under reduced

1

pressure to give the white-colored solid 4 (70 mg, 95%). H

NMR (400 MHz) (DMSO-d₆) δ: 8.32 (t, 1H, GIy⁴-NH), 8.26

(t, 1H, Gly³-NH), 7.33 (t, 1H, GIy²-NH), 7.05 (t, 1H, Gly¹-

NH), 3.74 (d, 2H, Gly^4α-H), 3.68 (d, 2H, Gly^3α-H), 3.59 (d,

2H, Gly^2α-H), 3.32 (d, 2H, Gly^1α-H), 1.37 [s, 9H, CO₂C-

(CH₃)₃]. ¹³C NMR (DMSO-d₆) δ: 171.2, 169.9, 169.3, 167.7,

155.9, 79.2, 78.2, 43.8, 43.4, 42.3, 42.1, 28.2. HRMS (ESI)

calculated for C₁₃H₂₂N₄O₇Na, [M+Na]⁺369.1386 (calcd.),

369.1388 (found).

[({[({[(2-tert-Butoxycarbonylamino-ethylcarbamoyl)-

methyl]-carbamoyl}-methyl)-carbamoyl]-methyl}-carbamo-

B

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

yl)-methyl]-carbamic Acid tert-Butyl Ester (6a). {2-[2-(2-tert-

Butoxycarbonylamino-acetylamino)-acetylamino]-acetylami-

no}-acetic acid (4; 70 mg, 0.20 mmol) was ﬁrst sonicated to

completely dissolve in 20 mL DMF, followed by submerging

the reaction mix in an ice−water bath. EDC (40 mg, 0.40

mmol) and HOBt (65 mg, 0.4 mmol) were then added to the

reaction mixture to activate 4 for 15 min, followed by addition

of a solution of mono-t-Boc-ethylenediamine (5a; 80 mg, 0.50

mmol) and DIPEA (0.035 mL, 0.40 mmol) in 5 mL of DMF to

initiate the coupling reaction at rt for 6 h. Compound 5a was

previously synthesized according to similar reported procedures

for 5b synthesis.²⁰The ﬁnal reaction solution was ﬁltrated,

evaporated in vacuo, resuspended in 25% MeOH in DCM, and

loaded to a silica column for further puriﬁcation. The products

were separated by eluting with mobile phases of 25% MeOH in

colored oil-like TFA salt of 7a. The obtained TFA salt of 7a was

further dissolved in 5 mL of a DCM/MeOH mixture (DCM/

MeOH = 1:1) and mixed with 10 equiv of Amberlyst A-21

(0.14 mg, 1.1 mmol) while rocking for 30 min.²⁰After ﬁltering

to remove Amberlyst A-21, the ﬁltrate of the reaction products

was concentrated under reduced pressure to give white solid 7a

1

(30 mg, 95%). H NMR (400 MHz) (DMSO-d₆) δ: 8.65 (t,

1H, NHCH₂CH₂), 8.32 (t, 1H, Gly⁴-NH), 8.17 (t, 1H, GIy³-

NH), 8.06 (t, 1H, Gly²-NH), 7.80−8.02 (br, 4H, Gly¹-NH₂,

CH₂CH₂NH₂), 3.85 (d, 2H, Gly^4α-H), 3.77 (d, 2H, Gly^3α-H),

3.71 (d, 2H, Gly^2α-H), 3.62 (s, 2H, Gly^1α-H), 3.30 (q, 2H,

CH₂NH), 2.85 (q, 2H, NHCH₂). ¹³C NMR (DMSO-d₆) δ:

169.6, 169.1, 168.8, 166.4, 42.0, 42.0., 42.0, 39.0, 38.5, 36.3.

HRMS (ESI) calculated for C₁₀H₂₁N₆O₄, [M+H]⁺289.1624

(calcd.), 289.1626 (found).

1

2-Amino-N-{[({[(6-amino-hexylcarbamoyl)-methyl]-carba-

moyl}-methyl)-carbamoyl]-methyl}-acetamide (7b). [6-(2-

{2-[2-(2-tert-Butoxycarbonylamino-acetylamino)-acetylamino]-

acetylamino}-acetylamino)-hexyl]-carbamic acid tert-butyl ester

(6b; 60 mg, 0.11 mmol) was dissolved in a TFA/DCM solution

(1:1; 4 mL), stirred at 0 °C for 1 h, and then reacted at rt for 1

h. The product was evaporated under reduced pressure, washed

by ether and chloroform, sequentially, and again concentrated

under reduced pressure to remove ether and chloroform to

obtain the brown oil-like TFA salt of 7b. The acquired TFA salt

of 7b (0.52 g, 0.88 mmol) was further dissolved in 5 mL of

DCM/MeOH mixture (DCM/MeOH = 1:1) and mixed with

10 equiv of Amberlyst A-21 (0.14 mg, 1.1 mmol) while rocking

for 30 min.²⁰After ﬁltering to remove Amberlyst A-21, the

ﬁltrate of the reaction products was concentrated under

DCM to aﬀord 6a (50 mg; 50%) as light yellowish solid. H

NMR (400 MHz) (DMSO-d₆) δ: 8.17 (t, 1H,

CONHCH₂CH₂), 8.10−8.04 (m, 2H, Gly¹-NH, CH₂CH₂NH

CO₂C(CH₃)₃), 7.80 (t, 1H, Gly⁴-NH), 6.99 (t, 1H, GIy³-NH),

6.80 (t, 1H, Gly²-NH), 3.74 (t, 4H, Gly^4α-H, Gly^1α-H), 3.66 (d,

2H, Gly^3α-H), 3.58 (d, 2H, Gly^2α-H), 3.08 (q, 2H,NHCH₂),

2.98 (q, 2H, CH₂NH), 1.37 [d, 18H, CO₂C(CH₃)₃]. ¹³C

NMR (DMSO-d₆) δ: 169.9, 169.4, 169.0, 168.8, 155.8, 155.6,

78.2, 77.7, 48.6, 43.3., 42.1, 42.1, 42.0, 40.1, 28.2, 28.2. HRMS

(ESI) calculated for C₂₀H₃₆N₆O₈Na, [M+Na]⁺511.2492

(calcd.), 511.2490 (found).

[6-(2-{2-[2-(2-tert-Butoxycarbonylamino-acetylamino)-

acetylamino]-acetylamino}-acetylamino)-hexyl]-carbamic

Acid tert-Butyl Ester (6b). {2-[2-(2-tert-Butoxycarbonylamino-

acetylamino)-acetylamino]-acetylamino}-acetic acid (4; 100

mg, 0.29 mmol) was ﬁrst sonicated to completely dissolve in

20 mL DMF, followed by immersing the reaction mix in an

ice−water bath. EDC (110 mg, 0.58 mmol) and HOBt (78 mg)

were initially added to the reaction mix to activate 4 for 15 min,

followed by addition of a solution of mono-t-Boc-hexanedi-

amine (5b; 63 mg, 0.29 mmol)²⁰and DIPEA (0.1 mL, 0.29

mmol) in 5 mL of DMF to commence the coupling reaction at

rt for 6 h. The ﬁnal reaction mixture was ﬁltrated, evaporated in

vacuo, redissolved in 10% MeOH in DCM, and loaded to a

silica column for further puriﬁcation. The products were

separated by eluting with mobile phases of 10% MeOH in

DCM to aﬀord 6b (85 mg; 54%) as a yellowish solid. ¹H NMR

(400 MHz) (DMSO-d₆) δ: 8.16 (t, 1H, CONHCH₂CH₂),

8.04−8.08 [m, 2H, Gly⁴-NH, NH CO₂C(CH₃)₃], 7.70 (t, 1H,

Gly³-NH), 7.00 (t, 1H, Gly¹-NH), 6.75 (t, 1H, GIy²-NH),, 3.75

(d, 2H, Gly^1α-H), 3.72 (d, 2H, Gly^4α-H), 3.65 (d, 2H, Gly^3α-H),

3.58 (d, 2H, Gly^2α-H), 3.03 (q, 2H, NHCH₂), 2.88 (q, 2H,

CH₂NH), 1.37 [d, 22H, CO₂C(CH₃)₃, NHCH₂CH₂], 1.21−

1.23 (m, 4H, CH₂). ¹³C NMR (DMSO-d₆) δ: 169.9, 169.4,

169.0, 168.4, 155.9, 155.6, 78.2, 77.3, 43.3, 42.2., 42.1, 42.0,

38.9, 38.5, 29.5, 29.1, 28.3, 28.2, 26.1, 26.0. HRMS (ESI)

calculated for C₂₄H₄₄N₆O₈Na, [M+Na]⁺567.3118 (calcd.),

567.3115 (found).

1

reduced pressure to aﬀord white solid 7b (50 mg, 95%). H

NMR (400 MHz) (DMSO-d₆) δ: 8.68 (t, 1H, NHCH₂CH₂),

8.33 (t, 1H, Gly⁴-NH), 8.13 (t, 1H, Gly³-NH), 7.84−8.08 (br,

4H, GIy¹-NH₂, CH₂CH₂NH₂), 7.79 (t, 1H, Gly²-NH), 3.84 (d,

2H, Gly^4α-H), 3.75 (d, 2H, Gly^3α-H), 3.66 (d, 2H, Gly^2α-H),

3.62 (s, 2H, Gly^1α-H), 3.04 (q, 2H, CH₂NH), 2.76 (t, 2H,

CH₂NH₂), 1.18−1.60 (m, 8H, CH₂). ¹³C NMR (DMSO-d₆) δ:

169.0, 168.9, 168.5, 166.5, 42.1, 42.5, 42.0, 40.2, 38.8, 38.4,

28.9, 27.0,25.8,25.5. HRMS (ESI) calculated for C₁₄H₂₉N₆O₄,

[M+H]⁺345.2250 (calcd.), 345.2251 (found).

Synthesis of a N-Maleoyl Amino Acid Succinimidyl

Ester, AMAS (9). (2,5-Dioxo-2,5-dihydro-pyrrol-1-yl)-acetic

acid 2,5-dioxo-pyrrolidin-1-yl ester (9) was synthesized from

(2,5-dioxo-2,5-dihydro-pyrrol-1-yl)-acetic acid (8) which was

prepared by a previously reported method.²²The acquired 8

(69 mg, 0.45 mmol) and NHS (103 mg, 0.89 mmol) were ﬁrst

dissolved in THF (10 mL), followed by dropwise addition of

DCC (184 mg, 0.89 mmol) in THF (10 mL) while stirring.

After overnight reaction at rt, the mixture was quenched by

addition of three drops of glacial acetic acid and stirring for 1 h

and ﬁltered to remove suspension. The acquired ﬁltrate was

concentrated under reduced pressure, washed by EtOH twice,

resuspended in 2-propanol (30 mL) while stirring for 1 h, and

ﬁltered to separate the suspension from its ﬁltrate. The ﬁnal

solid was washed by 2-propanol and dried to aﬀord a white-

colored 9 (96 mg; 86%). ¹H NMR (400 MHz) (DMSO-d₆) δ:

7.20 (s, 2H, CHCH), 4.73 (s, 2H, CH₂), 2.81 (s 4H, CH₂CH₂)

¹³C NMR (DMSO-d₆) δ: 169.7, 169.7, 164.4, 135.1, 36.3, 25.4.

HRMS (ESI) calculated for C₁₀H₈N₂O₆Na, [M+Na]⁺275.0280

(calcd.), 275.0279 (found).

2-Amino-N-{[({[(2-amino-ethylcarbamoyl)-methyl]-carba-

moyl}-methyl)-carbamoyl]-methyl}-acetamide (7a).

[({[({[(2-tert-Butoxycarbonylamino-ethylcarbamoyl)-methyl]-

carbamoyl}-methyl)-carbamoyl]-methyl}-carbamoyl)-methyl]-

carbamic acid tert-butyl ester (6a; 50 mg, 0.12 mmol) was

dissolved in a TFA/DCM solution (1:1; 4 mL), stirred at 0 °C

for 1 h, and then reacted at rt for 1 h. The product was

evaporated under reduced pressure, washed by ether and

chloroform, sequentially, and again concentrated under reduced

pressure to remove ether and chloroform to give the yellow-

Synthesis of the Thiol-Containing Biotin Derivative

(13). 5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-penta-

noic acid (2-{2-[5-(2-oxo-hexahydro-thieno[3,4-d]imidazol-

C

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

4-yl)-pentanoylamino]-ethyldisulfanyl}-ethyl)-amide (12). 5-

(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoic acid

2,5-dioxo-pyrrolidin-1-yl ester (11; 0.19 g, 0.55 mmol) was

synthesized from 10 according to our previously published

method²⁰and was completely dissolved in DMF (7 mL) with

gentle warming. After slowly cooling the 11 solution to rt

without recurring precipitation, it was mixed with an aqueous

solution (1 mL) containing cystamine dihydrochloride (0.068

g, 0.30 mmol) and Et₃N (0.14 g, 1.38 mmol) with stirring. The

reaction proceeded at rt overnight with stirring and the reaction

products were concentrated in vacuo. The remaining solid was

redissolved in 2-propanol (20 mL) with gentle heating, cooled

down to room temperature, and reprecipitated at 4 °C

overnight. The obtained white solid was recovered from the

glassware, washed with cold 2-propanol and air-dried to give 12

(0.14 g; 82.1%). ¹H NMR (400 MHz) (DMSO) δ: 8.01 (s, 2H,

CONH), 6.54 (s, 2H, CONH), 6.39 (s, 2H, CONH), 4.31 (t,

2H, CHN), 4.15 (t, 2H, CHN), 3.11 (dd, 2H, CHS), 2.85 (d,

2H, CHHS), 2.78 (t, 4H, CH₂S), 2.58 (d, 2H, CHHS), 2.09 (t,

4H, CH₂CO), 1.31−1.61 (m, 12H). ¹³C NMR (100.67 MHz)

(DMSO) δ: 172.26, 162.74, 61.04, 59.21, 55.43, 41.10, 37.88,

37.31, 35.15, 28.11, 25.23. HRMS (ESI) calculated for

C₂₄H₄₀N₆O₄S₄Na, [M+Na]⁺627.1891 (calcd.), 627.1893

(found).

5-(2-Oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-penta-

noic acid (2-mercapto-ethyl)-amide (13). 5-(2-Oxo-hexahy-

dro-thieno[3,4-d]imidazol-4-yl)-pentanoic acid (2-{2-[5-(2-

oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)-pentanoylamino]-

ethyldisulfanyl}-ethyl)-amide (12; 0.14 g, 0.23 mmol) was

completely dissolved in DMF (8 mL) with gentle warming.

After cooling, the 12 solution to rt without recurring

precipitation, ^DL-dithiothreitol (DTT; 0.11 g, 0.68 mmol) and

Et₃N (0.002 g, 0.02 mmol) were added to the solution with

stirring. Additional DTT (0.11 g, 0.68 mmol) was replenished

to the reaction mixture 1 h later. The reaction proceeded at rt

for 2 more hours, followed by evaporation in vacuo. The

remaining solid was washed by DCM (10 mL) three times and

concentrated under reduced pressure to give white-colored 13

(0.07 g; 52.7%). ¹H NMR (400 MHz) (DMSO) δ: 7.98 (s, 1H,

CONH), 6.45 (s, 1H, CONH), 6.38 (s, 1H, CONH), 4.32 (t,

1H, CHN), 4.14 (t, 1H, CHN), 3.19 (t, 2H, CH₂SH), 3.11 (dd,

1H, CHS), 2.83 (d, 1H, CHHS), 2.58 (d, 1H, CHHS), 2.37 (t,

1H, SH), 1.32−1.61 (m, 6H). ¹³C NMR (100.67 MHz)

(DMSO) δ: 172.38, 162.81, 61.07, 59.23, 55.44, 42.04, 35.14,

28.21, 28.03, 25.25, 23.52. HRMS (FAB) calculated for

C₁₇H₃₀O₅N₅SNa, [M+Na]⁺439.5112 (calcd.), 439.1865

(found).

GMP-primed TW17 RNA (0.32 nmol) and EDC (4.17 μmol)

in 4 μL 4(5)-methylimidazole-Triton X-100 buﬀer [0.1 M 4(5)-

methylimidazole, 15% Triton X-100, pH 6.0] and activating at

rt for 90 min. The resulting 5′-phosphorimidazolide RNA was

puriﬁed by ethanol precipitation and resuspended in 5.5 μL of

EPPS-Triton X-100 buﬀer (100 mM EPPS, 15% Triton X-100,

2 mM EDTA, pH 7.5). One microliter of 1 (187.2 mM in

DMF) was then added to the 5′-phosphorimidazolide RNA

solution to allow a phosphoramidation reaction at 41 °C for 3

h.

For the single-stranded 3′-primer DNA, the optimized two-

step phosphoramidation reaction was performed by dissolving

the single-stranded DNA (0.32 nmol) and EDC (4.17 μmol) in

4 μL of 4(5)-methylimidazole buﬀer [0.1 M 4(5)-methyl-

imidazole, pH 6.0] and activating at rt for 90 min. Similarly, the

resulting 5′-phosphorimidazolide DNA was puriﬁed by ethanol

precipitation and redissolved in 5.5 μL of EPPS buﬀer (100

mM EPPS, 2 mM EDTA, pH 7.5). A solution of 1 (1 μL; 187.2

mM in DMF) was later added to the 5′-phosphorimidazolide

DNA solution to allow a phosphoramidation reaction at 55 °C

for 3 h. No co-solute was used in the two-step phosphor-

amidation reaction of the single-stranded DNA to attain higher

reaction yield.

All resulting nucleic acid−substrate conjugates were puriﬁed

twice by ethanol precipitation, analyzed by urea-PAGE and SAv

gel shift assay (8% urea-PAGE for the TW17 RNA, and 20%

urea-PAGE for single-stranded DNA), visualized, and quanti-

ﬁed by an Amersham Typhoon PhosphorImager to determine

reaction yield.

Nucleic Acid-Tetraglycine Derivative Conjugate Prep-

aration. The phosphoramidation reactions between the TW17

RNA and 6a/6b adhered to the optimized two-step

phosphoramidation reaction stated above but with 1 replaced

by the same concentration of 6a/6b. The phosphoramidation

reactions between the 3′ primer DNA and 6a/6b was also

carried out by following the previously optimized two-step

phosphoramidation reaction for single-stranded DNA but with

the same concentration of 6a/6b substituting for 1. The

synthesized nucleic acid-tetraglycine derivative conjugates were

puriﬁed twice by ethanol precipitation, analyzed by 8% (the

TW17 RNA) or 20% (the 3′ primer DNA) urea-PAGE to

separate reaction products, visualized and quantiﬁed by an

Amersham Typhoon PhosphorImager.

Two-Step Phosphoramidation Reactions for Syn-

thesis of Nucleic Acid−Tat₄₈₋₅₇Peptide Conjugates.

The single-stranded 3′-primer DNA was conjugated with the

Tat₄₈₋₅₇peptide according to the optimized two-step

phosphoramidation reaction described previously but with the

following modiﬁcations. First, the scale of the reaction was

increased ﬁve times to acquire the suﬃcient conjugates for

subsequent ex vivo studies. In addition, the pH of the

conjugation reaction between the 5′-phosphorimidazolide

DNA and the Tat₄₈₋₅₇peptide was increased to 8.0 achieved

by addition of concentrated EPPS buﬀer (600 mM EPPS, 5

mM EDTA, pH 8.0) to attain higher yield. Finally, only 20 mM

of the Tat₄₈₋₅₇peptide was required in the coupling reaction to

have the best yield.

For RNA-Tat₄₈₋₅₇conjugates, preparation also followed the

optimized two-step RNA phosphoramidation reaction but with

following modiﬁcations: (1) Only 20 mM of the Tat₄₈₋₅₇

peptide was required in conjugation reactions, and (2)

concentrated EPPS-Triton X-100 buﬀer (600 mM EPPS, 15%

Triton X-100, 5 mM EDTA, pH 7.5) was added to the

Nucleic Acid Preparation and Radiolabeling. The 3′

primer (5′-TACCCCTTGGGGATACCACC-3′), a single-

stranded DNA, was purchased from Purigo Biotech, Inc.,

Taiwan, and used without puriﬁcation. The TW17₁₋₁₇RNA

(5′-GGGAUCGUCAGUGCAUU-3′) was purchased from

Bioneer (Daejeon, South Korea) and also used without

puriﬁcation. The 5′ GMP-primed TW17 RNA (87-mer; 5′-

GGGAUCGUCAGUGCAUUGAGAAGUGCAGUGUCUUG-

CGCUGGGUUCGAGCGGUCCGUGGUGCUGGC-

CCGGUGGUAUCCCCAAGGGGUA-3′) was prepared as

previously described.²³The TW17 RNA body-labeled with³²P

and the 3′ primer DNA ³²P-labeled at the 5′-end were prepared

according to the previously reported procedures.^20,23

Optimization of Two-Step Nucleic Acid Phosphor-

amidation Reactions. The optimized two-step phosphor-

amidation reaction for RNA was carried out by dissolving the

D

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

conjugation reaction between the 5′-phosphorimidazolide RNA

and the Tat₄₈₋₅₇peptide to retain the buﬀering capacity.

The synthesized nucleic acid-Tat₄₈₋₅₇conjugates were also

puriﬁed twice by ethanol precipitation, analyzed by 8% (the

TW17 RNA) or 20% (the 3′ primer DNA and the TW17₁₋₁₇

RNA) urea-PAGE, visualized and quantiﬁed by an Amersham

Typhoon PhosphorImager.

Two-Step Phosphoramidation Reactions for Syn-

thesis of Nucleic Acid−Cystamine Conjugates. Similar

optimized two-step phosphoramidation reactions were applied

when preparing nucleic acid−cystamine conjugates and are

described below. The RNA−cystamine conjugate synthesis was

carried out by dissolving the GMP-primed TW17₁₋₁₇RNA

(0.32 nmol) and EDC (4.17 μmol) in 4 μL of 4(5)-

methylimidazole-Triton X-100 buﬀer and activating at rt for

90 min. The resulting 5′-phosphorimidazolide RNA was

puriﬁed by ethanol precipitation and then resuspended in 5.5

μL of concentrated EPPS-Triton X-100 buﬀer with the addition

of 1 μL of cystamine (187.2 mM in water) to allow a reaction at

41 °C for 3 h.

For the single-stranded 3′-primer DNA, the cystamine

conjugate was prepared by dissolving the DNA (1.59 nmol)

and EDC (26 μmol) in 20 μL 4(5)-methylimidazole buﬀer and

activating at rt for 90 min. The resulting 5′-phosphorimidazo-

lide DNA was puriﬁed by ethanol precipitation, redissolved in

27.5 μL of concentrated EPPS buﬀer. Five microliters of

cystamine (187.2 mM in DEPC water) was then added to the

5′-phosphorimidazolide DNA solution to allow reaction at 55

°C for 3 h. Again, no co-solute was used in the two-step

phosphoramidation reaction of the single-stranded DNA to

attain higher reaction yield.

The products of the conjugation reactions between cyst-

amine and nucleic acids were separated by 20% urea-PAGE,

visualized and quantiﬁed by an Amersham Typhoon

PhosphorImager to determine reaction yield.

Optimization of AMAS (9) Reactions to Eﬀectively

Conjugate Nucleic Acids with the C(KFF)₃K Peptide. The

heterobifunctional 9 served as the linker to conjugate

cystamine-modiﬁed nucleic acids with peptides (Scheme S1).

Before having 9 react with cystamine-modiﬁed nucleic acids

and peptides, each reaction was optimized by using one of the

model compounds 11 and 13 to attain the highest yield. The

optimization reactions are brieﬂy described below.

For the reaction between a cystamine-modiﬁed nucleic acid

and 9, 11 with an activated carboxyl group was substituted for

9. The substitution simpliﬁed the acylation reaction allowing

the determination of the appropriate condition to provide the

highest yield as stated below. The cystamine-modiﬁed 3′ primer

DNA (900 pmol) was dissolved in 22.5 μL of concentrated

EPPS-7.3 buﬀer (600 mM EPPS, pH 7.3) to produce 40 μM of

the DNA solution in a microfuge tube. After addition of 11

solution (9 μL; 0.01−1 M in DMF), the microfuge tube was

wrapped with aluminum foil, and reacted at rt with rocking for

1.5 h. The resulting 3′ primer DNA-11 conjugates were puriﬁed

twice by ethanol precipitation, analyzed by 20% urea-PAGE,

visualized and quantiﬁed by an Amersham Typhoon

PhosphorImager to determine reaction yield.

The conjugation of the 9-modiﬁed 3′ primer DNA with 13

was studied to determine optimal conditions for the Michael

addition reaction between the 9-modiﬁed nucleic acids and

cysteine-containing peptides and to attain high yield (Scheme

S1; Supporting Information). The reaction optimization is

brieﬂy described below. Assuming no DNA loss during ethanol

precipitation, the acquired 9-modiﬁed 3′ primer DNA reaction

mixture (0.9 nmol) was dissolved in 50 μL of concentrated

EPPS-7.3 buﬀer, followed by the slow addition of 50 μL of 13

solution (36−3600 μM in DMF), and reacted at rt with rocking

for 16 h. The acquired 3′ primer DNA-13 conjugates were

puriﬁed twice by ethanol precipitation, analyzed by 20% urea-

PAGE, visualized and quantiﬁed by an Amersham Typhoon

PhosphorImager to determine reaction yield.

Synthesis of Nucleic Acid−C(KFF)₃K Peptide Con-

jugates Using the Cystamine-9 Linker. All nucleic acid−

C(KFF)₃K peptide conjugates were prepared by following the

optimized AMAS (9) reactions described previously and brieﬂy

stated below. The TW17₁₋₁₇RNA- C(KFF)₃K peptide

conjugate synthesis was initiated by dissolving the previously

prepared cystamine-modiﬁed TW17₁₋₁₇RNA (0.22 nmol) in

5.55 μL of concentrated EPPS-7.3 buﬀer to yield a 40 μM RNA

solution. After addition of 22 μL of 9 (10 mM in DMSO) to

the RNA solution, the microfuge tube containing the RNA

reaction mixture was wrapped with aluminum foil and reacted

at rt with rocking for 1.5 h. The 9-modiﬁed RNA was puriﬁed

by standard ethanol precipitation to remove excess 9, and was

redissolved in 50 μL of concentrated EPPS-7.3 buﬀer. A

solution of C(KFF)₃K solution (50 μL; 4.4 nmol) was then

slowly added to the 9-modiﬁed RNA solution and reacted at rt

with rocking for 16 h (Scheme S1).

For DNA conjugates, the cystamine-modiﬁed 3′ primer DNA

(0.9 nmol) was dissolved in 22.5 μL of concentrated EPPS-7.3

buﬀer to produce a 40 μM DNA solution in a microfuge tube.

A solution of 9 (90 μL; 10 mM in DMSO) was added to the

DNA solution and followed by wrapping the reaction mixture-

containing microfuge tube with aluminum foil and rocking the

reaction mixture at rt for 1.5 h. The reaction products were

puriﬁed twice by ethanol precipitation to remove excess 9. The

obtained 9-modiﬁed 3′ primer DNA (∼0.9 nmol) was

redissolved in 50 μL of concentrated EPPS-7.3 buﬀer, followed

by slow addition of 50 μL of the C(KFF)₃K peptide solution

(18 nmol in DEPC water), and reacted at rt with rocking for 16

h (Scheme S1). The obtained nucleic acid−C(KFF)₃K peptide

conjugate was puriﬁed twice by ethanol precipitation, analyzed

by 20% urea-PAGE, visualized and quantiﬁed by an Amersham

Typhoon PhosphorImager to obtain reaction yield.

MALDI-TOF MS for POC Characterizations. Prior to

MALDI-TOF analysis, each POC sample was puriﬁed by urea-

PAGE and desalted by passing through a Sephadex G-10 (GE

Healthcare Life Sciences, Taipei, Taiwan). MALDI-TOF mass

spectra were recorded on an Autoﬂex III TOF/TOF analyzer

(Bruker Daltonics, Taiwan). Each MALDI-TOF sample was

prepared by mixing a puriﬁed POC (100 pmol in 0.5 μL water)

with the MALDI-TOF matrix solution (0.5 μL), deposited on a

target plate and allowed to dry before the POC mass

measurement. A fresh MALDI-TOF matrix solution was

prepared daily by adopting a volume ratio of 1:2:1 for 2,3,4-

trihydroxyacetophenone (0.2 M in 50% acetonitrile), 2,4,6-

trihydroxyacetophenone (0.2 M in 50% acetonitrile), and

ammonium citrate (0.3 M), respectively.

Cell Culture. Cells of A549, a human lung carcinoma cell

line, were maintained in Ham’s F12K (Biowest, France) and

supplemented with 10% fetal bovine serum (FBS; Hyclone,

USA), 1% penicillin−streptomycin−amphotercin (PSA; Bio-

logical Industries, Israel), and 1% ^L-glutamine solution

(Biological Industries, Israel).

Cytotoxicity Assays. The viability of mammalian cells in

the presence of POCs was determined by the methylthiazolete-

E

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

trazolium (MTT) conversion assay described previously.²⁴

A549 cells were cultured in 96 well plates (2 × 10⁴cells/well;

0.1 mL medium/well), washed, and maintained in medium

containing serum overnight prior to assay. After inoculation

with 0−10 μM of the 3′ primer, DNA- Tat₄₈₋₅₇peptide

conjugate, cells were maintained at 37 °C for 24 h and further

incubated with MTT (0.5 mg/mL) at 37 °C for 4 h. By the end

of MTT incubation, the medium was removed and the

insoluble formazan product was dissolved by adding DMSO

(100 μL/well) with thoroughly mixing. MTT conversion was

determined by colorimetric analysis at 570 nm. Each experi-

ment was performed in triplicate; the reported result was mean

of triplicate experiments SD.

cold PBS (1 mL), transferred to Falcon tubes (BD Bioscience,

San Jose, CA, USA) and placed on ice in the dark before each

analysis. Each experiment was performed in triplicate; in each

case, the ﬂuorescence of at least 10 000 viable cells was

measured by using a BD FACSCalibur cytometer (BD, Franklin

Lakes, NJ, USA).

Confocal Laser Scanning Microscopy and Fluorescent

Microscopy. A549 cells were seeded in an 8-well Lab-Tek

coverglass chamber (0.3 mL medium per well; Nalge Nunc,

Rochester, NY, USA) with an appropriate density of 1 × 10⁴

cells per well and incubated for 24 h before the experiment.

The medium in the overnight cultures was removed,

replenished with a fresh medium containing 0 or 5 μM of

the FITC-labeled Tat₄₈₋₅₇peptide or 3′ primer DNA−Tat₄₈₋₅₇

peptide conjugate, and further maintained at 37 °C for 24 h.

The peptide/POC-containing medium was discarded after

incubation, followed by washing the cells with PBS three times

before confocal imaging analysis. The pictures were obtained

using a FluoView 1000 confocal laser scanning microscope

(Olympus, Tokyo, Japan). All images were acquired using living

and nonﬁxed cells. Each experiment was performed in triplicate.

Experimental procedures similar to those described above were

used in ﬂuorescent microscopy analysis.

Fluorescent Labeling of DNA and Its Conjugation

with the Tat₄₈₋₅₇Peptide. In order to detect the uptake of

DNA in POCs by mammalian cells using ﬂow cytometry,

ﬂuorescent microscopy, and confocal laser scanning micros-

copy, the DNA was required to be labeled with ﬂuorophores

before conjugating with the Tat₄₈₋₅₇peptide for POC synthesis.

Fluorescent labeling of DNA was achieved by a DNA

depurination method²⁵and brieﬂy described below. DNA

depurination was accomplished by resuspension of the 3′

primer DNA (1−10 mg) in 0.2 N HCl (10 mL) and allowed to

react at 37 °C for 90 min. The depurinated DNA solution was

diluted with 40 mL of ethylenediamine buﬀer (650 mM, pH

7.6) and reacted at 37 °C for 3 h. The solution of the imine-

containing DNA was further reacted with freshly prepared

NaBH₄(0.1 M in water, 4 mL) at rt for 30 min. The acquired

diamino-derivatized DNA sample was puriﬁed by adding 1 mL

of lithium perchlorate (2% in acetone) to precipitate the DNA,

washing the DNA with acetone twice and air−drying. After

redissolving the diamino-derivatized DNA in carbonate buﬀer

(0.1 M, pH 9.0; 20 μL), a solution of ﬂuorescein isothiocyanate

(FITC; 2.57 mM in DMSO, 20 μL) was slowly added to the

DNA solution and reacted in the dark for 12 h. The reaction

products were precipitated by ethanol and further puriﬁed by

20% urea-PAGE to obtain the FITC-labeled 3′ primer DNA.

The FITC-labeled DNA was conjugated with the Tat₄₈₋₅₇

peptide according to the optimized zero-linker two-step

phosphoramidation reaction described in this study. The

POC, the FITC-labeled 3′ primer DNA- Tat₄₈₋₅₇peptide

conjugate, was again puriﬁed by 20% urea-PAGE in which the

migration of the FITC-labeled POC was visualized by an

Amersham Typhoon PhosphorImager with the settings of the

excitation wavelength at 488 nm and the emission wavelength

at 525 nm for ﬂuorescein detection. It is noted that the FITC-

labeled Tat₄₈₋₅₇peptide was also prepared according to a

previously described method.²⁰

Flow Cytometry. A549 cells were seeded into a 12-well

plate (2 × 10⁵cells in 1 mL medium per well) and cultured

overnight. The medium in the overnight cultures was removed,

replenished with a fresh medium containing 0 or 5 μM of the

FITC-labeled Tat₄₈₋₅₇peptide or 3′ primer DNA−Tat₄₈₋₅₇

peptide conjugate, and further maintained at 37 °C for 24 h.

The peptide/POC-containing medium was discarded after

incubation in which the cells were washed by PBS once and

detached from the plate surface by treating with trypsin

(trypsin−EDTA 1×, 0.5 mL per well; Biowest, France) for 3−5

min. The trypsinized cells were resuspended in the medium

(0.5 mL) and centrifuged at 700 rcf for 4 min. After removal of

the supernatants, the cells were resuspended in ice-cold PBS (1

mL) and subjected to another round of centrifugation (700 rcf

for 4 min). The ﬁnal cell pellets were again resuspended in ice-

RESULTS

■

General Considerations in Optimizing the Aqueous-

Phase Two-Step Phosphoramidation Reactions for POC

Synthesis. We envisioned that three aspects of two-step

nucleic acid phosphoramidation reactions critical to POC

synthesis could be modiﬁed to increase product yields (Scheme

1). First, inert co-solutes such as Tween 20 and Triton X-100

were substituted for urea (8 M in the original reactions)²⁰in

phosphoramidation reactions to ensure co-solutes would carry

out their proposed roles of denaturing nucleic acids without

participating in nucleophilic substitution reactions. Our

previous study has revealed that using urea as a co-solute

could lead to less than optimal reaction yield. Second, we

sought to attain higher POC yields by modestly increasing the

nucleophilicity of the imidazole moiety. Appropriately stabiliz-

ing phosphorimidazolide intermediates would extend their half-

lives in aqueous solutions while not deterring subsequent

nucleophilic substitution reaction with peptides. Third, POC

yields could be further improved by chemically modifying

peptides to increase peptide nucleophilicity. We adopted these

three strategies to improve the aqueous-phase two-step

phosphoramidation reactions and present the results in the

following sections.

Identiﬁcation of an Appropriate Co-Solute and

Concentration to Facilitate POC Synthesis. We exploited

previously synthesized 1 (Scheme 1) as the nucleophile in

aqueous-phase two-step phosphoramidation reactions to

optimize conjugation with nucleic acids. First, we substituted

one of ﬁve diﬀerent co-solutes (Triton X-100, Tween 20, PEG-

6000, PEG-8000, and glycerol) for urea in the phosphor-

amidation reactions while varying co-solute concentrations

from 0% to 33% (w/v) to identify appropriate co-solutes and

concentrations for DNA and RNA conjugation with the

nucleophile 1. These co-solutes were chosen because each of

their structures only contains the hydroxyl functionality, which

is unlikely to deprotonate and participate in phosphoramidation

reactions as nucleophiles at neutral pH. The TW17 RNA (87-

mer nt) clearly favored 15% Triton X-100 in the phosphor-

amidation reactions to have the best yield (Figure 1A). The 3′

F

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

Scheme 1. Strategies to Optimize Aqueous-Phase Two-Step

Phosphoramidation Reactions for Conjugating Nucleic Acid

with Nucleophiles, Such as Peptides, to Synthesize Peptide−

a

Oligonucleotide Conjugates (POCs)

Figure 1. Eﬀects of co-solutes on yields of aqueous-phase two-step

phosphoramidation reactions when conjugating 1 with (A) the ³²P-

labeled TW17 RNA or (B) the ³²P-labeled single-stranded 3′ primer

DNA. (A) The studied co-solutes included the following: 1, urea; 2,

Tween 20; 3, Triton X-100; 4, PEG 6000; 5, PEG 8000; 6, glycerol.

Each co-solute had a concentration of 15% (w/v) in both buﬀers for

the phosphoramidation reactions. The reaction products were puriﬁed

by ethanol precipitation, analyzed by 8% urea-PAGE, visualized and

quantiﬁed by an Amersham Typhoon PhosphorImager. a, unreacted

TW17 RNA; b, TW17 RNA-1 conjugate. (B) The relationship of co-

solute concentrations and DNA-1 conjugate yields for each studied co-

solute. The data were quantiﬁed by results from several urea-PAGE

analyses similar to that in (A). Each data point is determined by

calculating the average of the yields from triplicate experiments

accompanied with the standard deviation represented as a vertical bar.

The studied co-solutes in the phosphoramidation reactions were urea

a

Dashed boxes indicate the approaches to improve the reaction yields.

Each bold, superscripted, lower-case letter represents the speciﬁc

method to explore reaction conditions to achieve best reaction yields.

These methods include the following: a, identifying an appropriate co-

solute and concentration; b, investigating imidazole derivatives to

stabilize phosphorimidazolide intermediates and facilitate subsequent

phosphoramidation reactions; c, increasing numbers of nucleophilic

functionalities in nucleophiles which were peptides in this study.

Compound 1 was one of the model nucleophiles in this study.

Compound 2d was identiﬁed as the best imidazole derivative for POC

synthesis revealed by the results noted in the text.

■

▲

(closed square, ), Tween 20 (closed triangle, ), Triton X-100

⧫

○

(open circle, ), PEG 6000 (closed diamond, ), PEG 8000 (closed

primer DNA (20-mer nt), however, had the best yield when no

co-solutes were included in the phosphoramidation reactions

(Figure 1B). Acquiring better yield with no co-solutes in DNA

phosphoramidation reactions is consistent with previous

ﬁndings²⁰and likely reﬂects short single-stranded DNA, such

as the 3′ primer DNA, having less of a tendency to fold into

secondary structures which would interfere in phosphoramida-

tion reactions. Having co-solute concentrations higher than

32.5% in two-step phosphoramidation reactions might promote

product formation as implicated in Figure 1B, but this is an

impractical approach as reaction mixtures became too thick to

handle.

Survey of Imidazole Derivatives with Potential to

Enhancing POC Formation. After determining optimal co-

solute conditions in the two-step phosphoramidation reactions

for nucleic acids, we substituted one of six imidazole derivatives

(2a−f; Figure S1) for imidazole in reactions. All the imidazole

derivatives have electron-donating substitutions in the aromatic

ring which presumably strengthens phosphoramidate bonds

and improves stability of the key phosphorimidazolide

intermediates in water (boxed in b in Scheme 1). Results of

the studied two-step phosphoramidation reactions in the

□

●

circle, ), and glycerol (open square, ).

presence of these imidazole derivatives, however, indicated

that only 2d [4(5)-methylimidazole); Scheme 1] could

substantially increase product yields for both DNA and RNA

conjugation (Figure 2). The ability of compound 2d as the only

imidazole derivative to enhance product formation suggests

that not only the nucleophilicity in the imidazole ring but also

the steric hindrance imposed on the nucleophilic N in the

imidazole moiety determine product yield in phosphoramida-

tion reactions. Increasing nucleophilicity in the imidazole ring is

a double-edged sword approach; it could stabilize the

phosphoramidate bond in a phosphorimidazolide intermediate

so eﬀectively as to deter subsequent nucleophilic substitution

required to attain desired products. We are currently

synthesizing imidazole derivatives based on 2d to further

improve two-step phosphoramidation reaction yield.

Increase of Peptide Nucleophilicity to Promote POC

Synthesis. We explored the relationship between peptide

nucleophilicity and phosphoramidation yield. We previously

demonstrated that diamino nucleophiles provided higher

G

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

Figure 2. Inﬂuence of imidazole nucleophilicity on yields of two-step phosphoramidation reactions when conjugating 1 with (A) the ³²P-labeled

TW17 RNA or (B) the ³²P-labeled single-stranded 3′ primer DNA. Either imidazole or one of its derivatives (2a−f, Figure S1) was individually

incorporated into a phosphoramidation reaction. The reaction product was puriﬁed by ethanol precipitation, analyzed by 8% urea-PAGE for the

RNA-1 conjugate and 20% urea-PAGE for the DNA-1 conjugate, visualized and quantiﬁed by an Amersham Typhoon PhosphorImager. a, unreacted

TW17 RNA; b, TW17 RNA-1 conjugate; c, unreacted 3′ primer DNA; d, 3′ primer DNA-1 conjugate.

a

product yield in signiﬁcantly shorter reaction time than

monoamino nucleophile 1 in phosphoramidation reaction.²⁰

We thus reasoned that using modiﬁed peptides carrying

additional amino groups would improve peptide nucleophilicity

and further increase phosphoramidation reaction yields. Since

we previously synthesized tetraglycine (3)²⁰and it is also

commercially available, we decided to use 3 as the starting

material to yield two of its amino derivatives, the novel peptides

7a and 7b, each carrying an additional amino group appended

on two or six carbon linkers (Scheme 2).

Scheme 2. Synthesis of the Tetraglycine Derivatives 7

Synthesis of 7a and 7b started with Boc protection of 3 to

give 4. The Boc-protected 4 was initially acylated with 5a/5b

following our previous method for similar reactions in synthesis

of 3 (EDC and Et₃N in DMF). Unexpectedly, the desired 6a/

6b was not produced after the reactions and their formations

were not improved when DMAP was included in the reactions.

We eventually adopted standard peptide coupling conditions²⁶

using EDC, HOBt, and DIPEA as reagents for the reactions and

successfully synthesized 6a and 6b. Boc deprotection of 6a and

6b (TFA in DCM) and subsequent TFA removal by Amberlyst

A-21²⁰ﬁnally led to eﬀective synthesis of the required 7a and

7b.

The successful synthesis of 7a and 7b allowed study of the

two-step phosphoramidation reaction using 7a and 7b as the

nucleophiles in the optimized reaction conditions. As expected,

7a and 7b were better nucleophiles than 1 and 3 when coupled

with either the 3′ primer DNA or the TW17 RNA to give

higher reaction yield (Figure 3). Increasing the extent of amino

functionality in the nucleophiles could contribute to higher

yields in two-step phosphoramidation reactions. The results

a

Reaction conditions: a, Di-t-Boc, NaOH/H₂O/dioxane; b, (1) EDC,

HOBt, (2) 5a/5b, DIPEA, DMF; c, (1) TFA, DCM, (2) Amberlyst A-

21, MeOH/DCM.

encouraged us to exploit the optimized two-step phosphor-

amidation reaction to synthesize POCs other than nucleic acid-

3 conjugates because typical peptides usually have more amino

groups than 7a and 7b. However, before employing the

optimized two-step phosphoramidation reactions for synthesis

H

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

Figure 3. Phosphoramidation yield inﬂuenced by the nucleophilicity of nucleophiles when reacting with (A) the ³²P-labeled 17-mer TW17₁₋₁₇RNA

and (B) the ³²P-labeled single-stranded 3′ primer DNA. Each nucleophile (1, 3, 7a, or 7b) was conjugated with the nucleic acids according to the

two-step phosphoramidation reactions optimized in this study. The reaction products were puriﬁed by ethanol precipitation, analyzed by 20% urea-

PAGE, visualized and quantiﬁed by an Amersham Typhoon PhosphorImager. a, unreacted TW17₁₋₁₇RNA; b, TW17₁₋₁₇RNA-3 conjugate; c,

TW17₁₋₁₇RNA-7a (or 7b) conjugate; d, TW17₁₋₁₇RNA-1 conjugate; e, unreacted 3′ primer DNA; f, 3′ primer DNA-3 conjugate; g, 3′ primer

DNA-7a (or 7b) conjugate; h, 3′ primer DNA-1 conjugate.

Figure 4. Autoradiogram of nucleic acids 1 (3′ primer DNA), 2 (two-step phosphoramidation reaction between the 3′ primer DNA and the Tat₄₈₋₅₇

peptide), 3 (TW17 RNA), 4 (two-step phosphoramidation reaction between the TW17 RNA and 200 mM of the Tat₄₈₋₅₇peptide), and 5 (two-step

phosphoramidation reaction between the TW17 RNA and 20 mM of the Tat₄₈₋₅₇peptide) analyzed by 8% urea−PAGE for the TW17 RNA samples

and 20% urea−PAGE for the 3′ primer DNA samples, visualized and quantiﬁed by an Amersham Typhoon PhosphorImager. The arrows indicate

migrations of a, 3′ primer DNA; b, 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate; c, TW17 RNA; d, TW17 RNA−Tat₄₈₋₅₇peptide conjugate.

of bioactive POCs, we desired to further increase POC yields

by carrying out the reactions for DNA at higher temperatures.

Eﬀects of Higher Temperatures on DNA Phosphor-

amidation Reactions. We previously concluded that RNA

phosphoramidation reactions at 41 °C is the optimal reaction

temperature because it balances reaction yield and RNA

degradation at higher temperatures.²⁰DNA, on the other hand,

is far more chemically stable than RNA. Therefore, we

investigated whether we could substantially improve yields of

two-step DNA phosphoramidation reactions when performed

above 41 °C (Figure S2). Indeed, we observed a steady increase

of product yields when having two-step phosphoramidation

reactions at temperatures higher than 45 °C. DNA in the two-

step phosphoramidation reactions, however, suﬀered from

signiﬁcant degradation at reaction temperatures above 55 °C.

We, therefore, determined that 55 °C was the optimal

temperature for two-step DNA phosphoramidation reactions

and was used subsequently.

POC Synthesis by the Optimized Two-Step Phosphor-

amidation Reactions Using the Zero Linker Approach.

We ﬁrst employed the developed two-step phosphoramidation

reactions to directly conjugate nucleic acids with a well-studied

I

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

Table 1. Survey of Aqueous-Phase Two-Step Nucleic Acid Phosphoramidation Reactions for Identifying the Optimal Reaction

a

Condition to Give the Best Reaction Yields and to Be Exploited for POC Synthesis

b

c,d

DNA /RNA

co-solute

imidazole and derivatives

nucleophile

temperature (°C)

yield (%)

c

RNA

N/A

imidazole

1

41

45

50

55

60

55

41

55

41

55

41

25%

18%

30%

33%

32%

31%

30%

27%

21%

26%

18%

28%

26%

22%

36%

19%

15%

42%

30%

33%

54%

64%

61%

50%

61%

54%

31%

33%

37%

47%

75%

72%

60%

59%

75%

c

RNA

urea, 15% (w/v)

imidazole

1

c

RNA

Tween 20, 15% (w/v)

imidazole

1

c

RNA

Triton X-100, 15% (w/v)

imidazole

1

c

RNA

PEG 6000, 15% (w/v)

imidazole

1

c

RNA

PEG 8000, 15% (w/v)

imidazole

2a

1

c

RNA

glycerol, 15% (w/v)

1

b

DNA

N/A

1

b

DNA

N/A

1

b

DNA

N/A

2b

1

b

DNA

N/A

2c

1

b

DNA

N/A

2d

1

b

DNA

N/A

2e

1

b

DNA

N/A

2f

1

c

RNA

Triton X-100, 15% (w/v)

2a

1

c

RNA

Triton X-100, 15% (w/v)

2b

1

c

RNA

Triton X-100, 15% (w/v)

2c

1

c

RNA

Triton X-100, 15% (w/v)

2d

1

c

RNA

Triton X-100, 15% (w/v)

2e

1

c

RNA

Triton X-100, 15% (w/v)

2f

1

c

RNA

Triton X-100, 15% (w/v)

2d

3

c

RNA

Triton X-100, 15% (w/v)

2d

7a

7b

3

c

RNA

Triton X-100, 15% (w/v)

2d

b

DNA

N/A

2d

b

DNA

N/A

2d

7a

7b

1

b

DNA

N/A

2d

b

DNA

N/A

2d

b

DNA

N/A

2d

1

b

DNA

N/A

2d

1

b

DNA

N/A

2d

1

b

DNA

N/A

2d

Tat₄₈₋₅₇peptide

cystamine

c

RNA

Triton X-100, 15% (w/v)

N/A

2d

b

DNA

2d

c

RNA

Triton X-100, 15% (w/v)

N/A

2d

cystamine

b

e

DNA

2d

cystamine, C(KFF)₃K peptide

d

e

RNA

Triton X-100, 15% (w/v)

2d

cystamine, the C(KFF)₃K peptide

a

It is noted that only the most important results from the co-solute eﬀect study have been provided. The yields for the optimized two-step

b

c

d

e

phosphoramidation reactions and for POC synthesis are bolded. 3′ primer DNA. TW17 RNA. TW17₁₋₁₇RNA. Conjugated to the nucleic acids

by Michael addition.

CPP, the Tat₄₈₋₅₇peptide,^12,18in the presence of no linking

molecules between them, the zero linker approach. The

Tat₄₈₋₅₇peptide was used to serve the purpose of a vector to

traﬃc DNA into mammalian cells validated by subsequent

biological analysis. The Tat₄₈₋₅₇peptide has 9 nucleophilic

amino groups (3 primary amino and 6 secondary amino

groups) in each molecule. Since increasing numbers of

nucleophilic functionalities in nucleophiles was likely to

enhance POC yields (Figure 3), we reasoned that using the

Tat₄₈₋₅₇peptide as the nucleophile could decrease the molar

ratios of nucleophiles to nucleic acids well below 585, the

required ratio for 1 and 3 as the nucleophiles, while retaining

similar reaction yields in the optimized two-step phosphor-

amidation reactions.

Table 1). Figure 4 also clearly indicates that only a nucleic acid

molecule was covalently linked to the Tat₄₈₋₅₇peptide even

though the Tat₄₈₋₅₇peptide has 9 nucleophilic functionalities

for phosphoramidation reactions. In addition, multiple

nucleophilic functionalities in the Tat₄₈₋₅₇peptide suggested

that each amino group in the Tat₄₈₋₅₇peptide has the potential

to engage in a phosphoramidation reaction to contribute to the

observed POC in Figure 4. Interestingly, the product

complexity did not seem to be an issue, because we observed

signiﬁcantly lower POC yields in the same phosphoramidation

reactions when replacing the Tat₄₈₋₅₇peptide, which has

glycine in the N-terminus, with the Tat₄₇₋₅₇peptide, which has

tyrosine in the N-terminus (Figure S3). Similar preference for

glycine in the N terminus was also observed when an E. coli

CPP, the (KFF)₃K peptide, and its derivatives were

phosphoramidated with the 3′ primer DNA (Figure S3).

These results implicated the N-terminus amino group in a

peptide as the primary nucleophile in phosphoramidation

reactions and its reactivity strongly aﬀected by steric hindrance

As expected, the optimized two-step phosphoramidation

reactions by the zero linker approach only required 62.5 as ratio

of Tat₄₈₋₅₇peptide to nucleic acids to achieve a 47% yield for

the 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate and a 75% yield

for the TW 17 RNA−Tat₄₈₋₅₇peptide conjugate (Figure 4 and

J

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

imposed by the side chain of the N-terminus amino acid

residue. The N-terminus glycine in a peptide emerges as the

best nucleophile in phosphoramidation reactions because very

limited steric hindrance would be enacted by the hydrogen

atom, the side chain of glycine. The preference of glycine over

other amino acids in the N-terminus of peptides for

phosphoramidation-prepared POC formation is also consistent

with the properties of 2d [4(5)-methylimidazole)] to give the

best phosphoramidation yield (Figure 2). Consequently,

despite the presence of primary and secondary amino groups

in typical peptides, the N-terminus amino group is clearly the

best nucleophile to participate in nucleophilic substitution

reactions with nucleic acid phosphorimidazolides. The N-

terminus amino group of a peptide as the favorite nucleophile

in nucleic acid phosphoramidation attains more speciﬁc

reactions and produces relatively structurally uniform POC

conjugates. Moreover, glycine in the N-terminus of peptides

should be the choice for synthesis of phosphoramidation-

prepared POCs because it has the least steric hindrance among

amino acids.

We characterized the acquired POCs by MALDI-TOF MS to

determine the molecular mass. We initially wondered whether

the molecular mass of a phosphoramidation-prepared POC

could be measured by more sensitive positive-mode MALDI-

TOF MS. Adopting positive-mode MALDI-TOF MS required

the use of acidic MALDI-TOF matrices, reported to be

detrimental to phosphoramidate bonds in DNA during analysis

and lead to DNA degradation.²⁷Therefore, we performed a

control experiment prior to positive-mode MALDI-TOF MS

analysis by incubating the gel-puriﬁed 3′ primer DNA−Tat₄₈₋₅₇

peptide conjugate in an acidic MALDI-TOF matrix solution

composed of 2,3,4-trihydroxyacetophenone, 2,4,6-trihydroxya-

cetophenone, and ammonium citrate for 3 h. We were pleased

to observe that the phosphoramidate bond linking the 3′

primer DNA and the Tat₄₈₋₅₇peptide was signiﬁcantly inert in

the MALDI-TOF matrix solution because only 5% of the

phosphoramidation-prepared POC were hydrolyzed at the end

of incubation (Figure S4).

will be wasted during POC synthesis. We were thus motivated

to pursue the cystamine-9 linker strategy for more eﬀective

POC synthesis.

To achieve synthesis of the cystamine-9 linker-included

POCs, we decided to synthesize 9, a N-maleoyl amino acid

succinimidyl ester. In addition, we desired to develop a method

for general synthesis of N-maleoyl amino acid succinimidyl

esters for POC synthesis in the future. We succeeded in

synthesizing 9 (Scheme 3) by adopting a reported procedure to

Scheme 3. Synthesis of the N-Maleoyl Amino Acid

a

Succinimidyl Ester 9

a

Reaction conditions: a, (1) glycine, glacial AcOH, 16 h (2) reﬂux, 8

h; b, NHS, DCC, THF, 16 h.

acquire 8²²and following up a NHS ester formation reaction to

give the desired 9 with good yield. We are improving the

method for synthesis of 9 and developing a more universal

approach to synthesize other useful N-maleoyl amino acid

succinimidyl esters.

When applying 9 for POC synthesis, we soon realized the

importance of optimizing the conjugation reactions involving

both functionalities in 9, i.e., acylation and Michael addition

reactions to relieve the cost burden when purchasing peptides

from commercial vendors. Therefore, we synthesized 11 and 13

(Scheme 4) as model compounds to optimize the reactions of 9

a

Scheme 4. Synthesis of 13, a Thiol Derivative of Biotin

The molecular mass of the 3′ DNA primer−Tat₄₈₋₅₇peptide

conjugate was ultimately conﬁrmed by MALDI-TOF MS ([M

+H]⁺= 7416.096; Table 1 and Supporting Information). We,

however, were not able to determine the molecular mass of the

TW17 RNA−Tat₄₈₋₅₇peptide conjugate because the TW17

RNA was synthesized by in vitro transcription of T7 RNA

polymerase to render a heterogeneous molecular mass of the

RNA.²⁸In addition, the size of the TW17 RNA−Tat₄₈₋₅₇

peptide conjugate (∼30 kD) prevents an accurate measurement

of the molecular mass by MALDI-TOF MS.

a

Optimizing the Two-Step Phosphoramidation Reac-

tions for Synthesis of POCs Containing the Cystamine-9

Linker. We synthesized POCs by using cystamine in two-step

phosphoramidation reactions in order to introduce a cystamine-

9 linker in nucleic acids for subsequent peptide coupling

reactions. The cystamine-9 linker was harnessed for two

purposes: (1) the disulﬁde linkage in cystamine potentially

reduced in vivo to release oligonucleotides from POCs which

could be essential for oligonucleotides to more speciﬁcally

silence target genes in biological systems, and (2) the

maleimide group in 9 to perform the more eﬀective Michael

addition reaction which would lower ratios of peptides to

nucleic acids in POC synthesis. Reducing ratios of peptide to

nucleic acids in synthesis of phosphoramidation-prepared

POCs has important economic signiﬁcance, as fewer peptides

Reaction conditions: a, NHS, DCC, pyridine, DMF, 24 h; b,

cystamine, Et₃N, DMF/H₂O, 16 h; c, DTT, Et₃N, DMF, 3 h.

for more eﬀective conjugation of nucleic acids with peptides.

The reasons for using 11 and 13 to determine optimal reaction

conditions of 9 for POC synthesis were that these compounds

have appropriate functionalities to engage in one of the

conjugation reactions. In addition, both compounds have biotin

tags to separate reaction products from reactants by SAv gel

shift assay²³in case of an inability to separate nucleic acid

samples by urea-PAGE. Synthesis of 11 was reported by us

previously.²⁰Starting from 11, a facile acylation reaction

successfully produced 12 (82%), which was subsequently

reduced in the presence of DTT to give the desired 13 (53%).

K

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

Figure 5. Optimization of the reactions involving AMAS (9) and synthesis of POCs in the presence of the cystamine-9 linker. (A) The reactions of 9

were optimized by using 11 and 13. The ³²P-labeled single-stranded 3′ primer DNA (a) was ﬁrst coupled with cystamine to acquire the DNA−

cystamine conjugate (Chart S1; Supporting Information) (b). The DNA−cystamine conjugate was subsequently reacted with various molar

equivalents of 11 to determine the appropriate molar equivalent able to give the best yield of the DNA−cystamine−11 conjugate (Chart S1;

Supporting Information) (c). A molar equivalent of 1000 was adopted to react the DNA−cystamine conjugate with 9 to yield the DNA−cystamine−

9 conjugate (Chart S1; Supporting Information) (c), which was later coupled with various molar equivalents of 13. It was determined that the molar

equivalent of 20 was suﬃcient to have the best yield of the DNA−cystamine−9−13 conjugate (Chart S1; Supporting Information) (d). Each

reaction product was puriﬁed by ethanol precipitation, analyzed by 20% urea−PAGE, visualized, and quantiﬁed by an Amersham Typhoon

PhosphorImager. a, the unreacted 3′ primer DNA; b, the 3′ primer DNA−cystamine conjugate; c, the 3′ primer DNA-cystamine-11 (or -9)

conjugate; d, the 3′ primer DNA-9-13 conjugate. (B) The cystamine-9 linker-prepared POCs were synthesized by the optimized reaction conditions

developed in Figure 5A. The nucleic acids in the autoradiogram contains 1 (the 3′ primer DNA), 2 (the cystamine-modiﬁed 3′ primer DNA

preparation reaction), 3 (the 3′ primer DNA-C(KFF)₃K peptide conjugate preparation reaction), 4 (the 17-mer TW17₁₋₁₇RNA), 5 (the cystamine-

modiﬁed TW17₁₋₁₇RNA preparation reaction), and 6 (the TW17₁₋₁₇RNA−C(KFF)₃K peptide conjugate preparation reaction). The arrows

indicate migrations of e, the 3′ primer DNA or TW17₁₋₁₇RNA; f, the cystamine-modiﬁed oligonucleotides; g, the 3′ primer DNA−C(KFF)₃K

peptide conjugate; h, the TW17₁₋₁₇RNA−C(KFF)₃K peptide conjugate.

We ﬁrst used 11 to optimize the acylation reaction between

the cystamine-modiﬁed 3′ primer DNA and 9. The 3′ primer

DNA was initially conjugated with cystamine according to the

optimized two-step phosphoramidation reaction reported in

this study with a yield of 72% (Figure 5A and Table 1). When

reacting the cystamine-modiﬁed 3′ primer DNA with 11, we

were delighted to observe that the ³²P-labeled 3′ primer DNA

and its structure-modiﬁed derivatives were well separated in

urea-PAGE so that the potentially ambiguous SAv gel shift

assay was not required. Moreover, the results from the reactions

of cystamine-modiﬁed 3′ primer DNA with various concen-

trations of 11 clearly indicated that a value of 1000 for the

molar ratio of 11 to the 3′ primer DNA gave the best yield for

the acylation reaction and was adopted for the similar reaction

between the cystamine-modiﬁed 3′ primer DNA and 9. The

acquired 9-cystamine-modiﬁed 3′ primer DNA was employed

to react with a series of 13 concentrations and to identify the

appropriate molar ratio of 13 to nucleic acids for attaining the

best yield in the Michael addition reaction (Figure 5A). A value

of 20 for the molar ratio of 13 to the 9-cystamine-modiﬁed 3′

primer DNA was suﬃcient to provide the best yield for the

studied Michael addition reaction and was determined to be the

optimal condition for POC synthesis.

The optimized reactions for 9 were successfully applied to

conjugating nucleic acids with the C(KFF)₃K peptide, the

derivative of a previously reported synthetic CPP (KFF)₃K

peptide,¹⁷to obtain the corresponding POCs with excellent

yields (59−75%; Figure 5B and Table 1). It is noted that we

used a shorter version of TW17 RNA, TW17₁₋₁₇RNA, in order

to measure the molecular mass of the desired RNA POC,

TW17₁₋₁₇RNA−C(KFF)₃K peptide conjugate, subsequently

by MALDI-TOF MS. As expected, the identity of each

synthesized POCs was successfully characterized by MALDI-

TOF MS to determine the molecular mass (Table 1 and

L

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

Supporting Information). The two-step phosphoramidation

reactions for POC synthesis in the presence of the cystamine-9

linker demonstrated advantages over POC synthesis by the

phosphoramidation reactions performed with the zero linker

approach. The incorporation of the cystamine-9 linker in POC

synthesis oﬀers both higher reaction yields (Figures 4 and 5B)

and a lower demand for peptides in POC synthesis (20 as the

molar ratio of peptide to oligonucleotide in this reaction) than

the zero linker strategy for POC synthesis. We expect that the

use of linkers similar to the cystamine-9 bridge will have more

leverage for future POC synthesis.

Cytotoxicity of Phosphoramidation-Prepared POCs to

a Human Cell Line. We understood that the phosphoramidate

linkage has to exhibit low to no toxicity to human beings before

phosphoramidation-prepared POCs would be administrated in

clinical applications. Therefore, we exploited the two-step

phosphoramidation reactions using the zero linker approach to

synthesize more of the POC, the 3′ primer DNA−Tat₄₈₋₅₇

peptide conjugate (Figure 4), and to study toxicity of the POC

to human A549 cells. We chose the zero linker approach rather

than using the cystamine-9 linker strategy to synthesize POC

for the toxicity study because we want to focus only on the

toxic eﬀects of the introduced phosphoramidate bond in the 3′

primer DNA−Tat₄₈₋₅₇peptide conjugate. We were satisﬁed to

ﬁnd that the inoculated 3′ primer DNA−Tat₄₈₋₅₇peptide

conjugate with concentrations up to 10 μM showed no toxicity

to human A549 cells after 24 h inoculation (Figure 6).

DNA by human A549 cells, we labeled the 3′ primer DNA with

the ﬂuorophore FITC by a depurination method²⁵before the

DNA was conjugated with the Tat₄₈₋₅₇peptide by the two-step

phosphoramidation reaction developed in this study. Acquis-

ition of the FITC-labeled 3′ primer DNA−Tat₄₈₋₅₇peptide

conjugate allowed us to study permeability of the POC into

human A549 cells by ﬂow cytometry, ﬂuorescent microscopy,

and confocal laser scanning microscopy.

The results from ﬂow cytometry analysis clearly showed that

the FITC-labeled 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate

was signiﬁcantly taken up by human A549 cells after 24 h

inoculation of the POC (Figure 7). In addition, the association

of the FITC-labeled 3′ primer DNA−Tat₄₈₋₅₇peptide

conjugate with A549 cells was revealed by ﬂuorescent

microscopy studies which indicated colocalization of ﬂuo-

rescence signals with intact A549 cells (Figure S5). Finally, we

performed confocal laser scanning microscopy to determine

whether the observed ﬂuorescence signals were contributed by

the FITC-labeled 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate

traﬃcked inside A549 cells or by the FITC-labeled POC

adsorbed on the cell membrane. Consistent with our

expectation, the results from the confocal laser scanning

microscopy study unequivocally demonstrated that the FITC-

labeled 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate was taken

up by A549 cells rather than merely adsorbed on the cell

membrane after 24 h inoculation (Figure 7). The results of

these biological uptake studies support the premise that a

phosphoramidate bond does not alter cellular permeability

activity of CPPs in phosphoramidation-prepared POCs and

would not have a negative impact on cellular delivery and

targeting of phosphoramidation-prepared POCs when admin-

istrated as therapeutic agents in the clinic.

DISCUSSION

■

We have signiﬁcantly improved reaction yields of two-step

nucleic acid phosphoramidation by studying the contribution of

three crucial factors, co-solute eﬀect, imidazole nucleophilicity

and structure, and the advantage of multiple nucleophilic

functionalities in nucleophiles (Figures 1−3, Scheme 1, and

Table 1). Moreover, we successfully employed the zero linker

strategy of the optimized two-step phosphoramidation

reactions to eﬀectively conjugate a well-studied CPP

TAT₄₈₋₅₇peptide with DNA or RNA with corresponding

POC yields of 47−75% (Figure 4, Table 1, and Scheme 5). We

also developed a more eﬀective phosphoramidation method for

DNA/RNA POC synthesis by exploiting a cystamine-9 linker

to acquire corresponding POC yields of 59−75% (Figure 5,

Table 1, and Scheme 5). The phosphoramidation-synthesized

DNA POC, 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate,

showed no cytotoxicity to human A549 cells at POC

concentrations up to 10 μM after 24 h inoculation and

established the potentially safe use of phosphoramidation-

synthesized POCs in clinical applications (Figure 6). In

addition, the FITC-labeled 3′ primer DNA−Tat₄₈₋₅₇peptide

conjugate was successfully traﬃcked into the human A549 cell

line as demonstrated by studies of ﬂow cytometry, ﬂuorescent

microscopy, and confocal laser scanning microscopy (Figures 7

and S5). These results provide evidence that the presence of

phosphoramidate bonds in CPP-containing POCs will not

impair the cell permeability of CPPs in POCs and suggests the

potential development of target-delivered therapeutic reagents

based on nucleic acid phosphoramidation reactions. The

current study corroborates the assertion that the advanced

Figure 6. Undetectable cytotoxicity of phosphoramidation-prepared

POCs to human A549 cells after 24 h inoculation. Human A549 cells

were incubated with increased concentrations of the phosphoramida-

tion-prepared POC, the 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate

(0−10 μM). After 24 h inoculation, cell viability for each sample

containing a speciﬁed POC concentration was determined by the

MTT assay.

Therefore, there is no detectable toxicity of phosphoramidate

linkages in POCs to human cells in these studied concen-

trations after 24 h inoculation, and we validated the potentially

safe use of phosphoramidation-prepared POC as therapeutic

agents for treating human diseases.

Uptake of Phosphoramidation-Prepared POCs by a

Human Cell Line. We had already shown that phosphor-

amidate bonds in POCs had no toxicity to human A549 cells

after 24 h inoculation (Figure 6). The potential evidence of

human safety of phosphoramidate linkages in POCs

encouraged us to carry out additional studies to explicate

eﬀects of phosphoramidate bonds on the cellular permeability

ability of CPPs in phosphoramidation-prepared POCs. We

again used the 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate as

the model POC to study the uptake of POCs by human A549

cells. To unambiguously demonstrate uptake of the 3′ primer

M

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

Figure 7. Uptake of phosphoramidation-prepared POCs by human A549 cells. Human A549 cells were inoculated with 5 μM of the FITC-labeled

Tat₄₈₋₅₇peptide or the phosphoramidation-prepared POC, the FITC-labeled 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate, for 24 h and followed by

analysis with ﬂow cytometry and confocal laser scanning microscopy. For ﬂow cytometry, cells were washed and harvested with trypsin/EDTA

before analysis. A549 cells for the confocal laser scanning microscopy study were washed three times with PBS and analyzed immediately. The

studied A549 cell samples included: A, cells not treated with either the Tat₄₈₋₅₇peptide or the 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate; B, cells

incubated with the FITC-labeled Tat₄₈₋₅₇peptide for 24 h; C, cells incubated with the FITC-labeled 3′ primer DNA−Tat₄₈₋₅₇peptide conjugate for

24 h. PC, phase contrast.

Scheme 5. Optimized Aqueous-Phase Two-Step Phosphoramidation Reactions for POC Synthesis

aqueous-phase two-step phosphoramidation reactions are an

eﬃcient and eﬀective approach to synthesize bioactive POCs.

The use of urea and other co-solutes in phosphoramidation

reactions deserves additional discussion to explicate their roles.

The current study clearly demonstrates that urea could not

carry out its expected function of denaturing nucleic acids to

confer higher product yields for phosphoramidation reactions.

Instead, urea caused reduction of reaction yield because of

direct participation of urea in nucleophilic substitution

preventing the desired product formation (Figure 1).²⁰We

had reached a similar conclusion on DNA phosphoramidation

reactions but had observed the beneﬁcial presence of urea in

RNA phosphoramidation reactions previously.²⁰The discrep-

ancy of the eﬀect of urea in RNA phosphoramidation reactions

is attributed to the shortcomings of the SAv gel shift assay used

in the previous study. A standard SAv gel shift assay relies on

strong interaction between biotin and SAv to separate biotin-

tagged from non-biotin-tagged biomacromolecules. Unfortu-

nately, in previously studied nucleic acid phosphoramidation

reactions, SAv-retarded nucleic acids co-migrated with poly-

merized nucleic acids, byproducts of the phosphoramidation

reactions,²⁹and were indistinguishable in the standard SAv gel

shift assay. The co-migration resulted in an overestimation of

the desired RNA reaction products (Ko and Wang, unpublished

results). We had carefully optimized the concentrations and

ratios of EDC to nucleic acids to signiﬁcantly suppress the

unwanted polymerization of nucleic acids (less than 5%) in not

only the two-step phosphoramidation reactions reported in this

study, but also the one-step phosphoramidation versions

(Wang et al., manuscript in preparation). The more accurate

determination of the phosphoramidation reaction yield was

attained by analysis of urea-PAGE, which depends solely on the

size and the charge of nucleic acids to achieve separation and

unambiguously resolved the molecular diﬀerence of reactants

and phosphoramidation products, resulting in more faithful

calculations of product formation. Therefore, urea-PAGE was a

major analytical method employed in the current study.

Consequently, inclusion of urea as co-solute in any nucleic

N

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

acid phosphoramidation reactions should be avoided, as urea

has a negative impact on the reactions by reducing the yield of

the desired product. On the contrary, more inert and hydroxyl

group-based co-solutes such as Triton X-100 will perform the

anticipated function of RNA denaturation and deliver higher

product yield in phosphoramidation reactions (Figure 1A).

In addition to the eﬀects of co-solutes in phosphoramidation

reactions, the chemoselectivity of nucleic acids toward speciﬁc

primary amino groups of peptides in the zero linker approach

of the phosphoramidation reactions is signiﬁcant (Figure S3).

Speciﬁcally, we observed that nucleic acids in zero-linker two-

step phosphoramidation reactions primarily reacted with the α-

amino group in N-terminus glycine residue of a peptide and

only one nucleic acid was conjugated to each peptide molecule.

This phenomenon came as a pleasant surprise, as we were

initially concerned about the multiple nucleophilic function-

alities in a typical peptide and their potential to engage in

phosphoramidation reactions. We anticipated that it would be

unlikely to have speciﬁc phosphoramidation reactions for

controllable POC synthesis and expected undesirable compli-

cated structures of POC products which might impede the use

of phosphoramidation reactions for synthesis of POCs as

therapeutic agents or scientiﬁc tools. Therefore, it was very

exciting to ﬁnd that the phosphorimidazolide intermediates of

nucleic acids had chemoselectivity and preferred to react with

the N-terminal α-amino group and with glycine in the N-

terminus of a peptide.

The chemoselective property of the phosphoramidation

reactions between nucleic acids and peptides is intriguing

because the basicity of the primary amino group in the side

chain of lysine (pK_a= 10.54),³⁰which is one of the major

components in the peptides in this study, is higher than that of

the α-amino group in glycine (pK_a= 9.78).³⁰When considering

the basicity and nucleophilicity of the primary amino groups in

glycine and lysine, we would predict that the phosphorimida-

zolide intermediates of nucleic acids are more likely to react

with the ε-amino group of lysine than with the α-amino group

of glycine in these peptides. In addition, multiple primary

amino groups in the peptides would also implicate that

conjugations of nucleic acids with a peptide by zero-linker two-

step phosphoramidation reactions should produce complicated

structural variations of POC products because of the likelihood

of multiple nucleic acid conjugations with each peptide.

Fortunately, the ε-amino groups of lysine in the peptides did

not demonstrate appreciably reactivity with the phosphor-

imidazolide intermediates of nucleic acids, and each peptide

was only conjugated with one molecule of nucleic acid (Figure

4). We speculate that the ε-amino groups of lysine in the

peptides might have been buried in the interior of three-

dimensional peptide structures and were unavailable for

phosphoramidation reactions. NMR analysis and computational

studies of these peptide structures in water are required to

provide essential information on the steric environments of

these primary amino groups in the peptides and their

relationship to the chemoselectivity of nucleic acid phosphor-

amidation reactions.

the CPPs studied here all have ample lysine or lysine and

arginine residues in their structures to contribute high positive

charge density in the peptides at neutral pH. Consequently,

when the α-amino group in the N-terminal glycine of a CPP

underwent a phosphoramidation reaction with a nucleic acid,

the phosphoramidate bond brought these oppositely charged

components in close proximity to enact strong electrostatic

forces between the CPP and the nucleic acid. Extensive

electrostatic forces from CPP−nucleic acid interactions would

immediately shield other primary amino groups in the CPP

from having further phosphoramidation reactions and attain a

single nucleic acid conjugation in each CPP. In fact, the

electrostatic attractive forces between nucleic acids and CPPs

were so strong that even boiling the nucleic acid−CPP

complexes for minutes in the presence of 8 M urea did not

completely remove CPPs from nucleic acids as revealed by

standard urea-PAGE (Su and Wang, unpublished results).

Moreover, the sizes of nucleic acids (6037 amu for the 3′

primer DNA, 5440 amu for the TW17₁₋₁₇RNA, and ∼28 000

amu for the TW17 RNA) are far larger than those of CPPs

(1396 amu for the Tat₄₈₋₅₇peptide and 1517 amu for the

C(KFF)₃K peptide) so that a nucleic acid molecule could

adsorb and exclude each CPP from contacting a second nucleic

acid. Similarly, strong electrostatic attraction between the

Tat₄₈₋₅₇peptide and nucleic acids could play a prominent role

to enhance POC formation observed in Figure 4.

Using a nucleic acid-CPP bridge similar to the cystamine-9

linker further alleviates the potential complexity problem of

POC products synthesized by phosphoramidation reactions.

The use of molecular bridges such as the cystamine-9 linker for

POCs synthesized by phosphoramidation reactions took

advantage of a more deﬁned conjugation reaction between

nucleic acids and CPPs, which was achieved by the more

speciﬁc Michael addition reaction between the thiol group in

CPPs and the maleimide moiety in phosphoramidation-

modiﬁed nucleic acids. Maleimides have speciﬁc reactivity for

thiol groups in the pH range 6.5−7.5.³¹The other beneﬁt of

introducing molecular bridges analogous to the cystamine-9

linker is a signiﬁcantly reduced requirement for the amount of

CPP for eﬀective POC synthesis (Figure 5B), which again

could be attributed to eﬀectiveness of the Michael addition

reaction. The outstanding properties of Michael addition to

convey more eﬀective and chemospeciﬁc reactions and to

deliver more structure-deﬁned POC products justify the

preferred choice of adopting the maleimide-containing bridges

for POCs synthesized by aqueous-phase two-step phosphor-

amidation reactions.

In summary, this study signiﬁcantly improved the aqueous-

phase two-step phosphoramidation reactions and demonstrated

the potential of either zero linker or linkers containing

maleimide moieties of phosphoramidation reactions for

eﬀective and eﬃcient POC synthesis. In addition, chemo-

selectivity of the advanced two-step nucleic acid phosphor-

amidation reactions supports the synthesis of POCs with

expected structures, which is essential to evaluating POC

functions in clinical applications. Furthermore, the versatility of

phosphoramidation reactions enables us to append diverse

functional groups to nucleic acids for subsequent reactions and

further facilitate POC synthesis. Finally, the advanced nucleic

acid phosphoramidation reaction is not restricted to POC

synthesis for acquiring nucleic acid-based therapeutic agents

useful in clinic. It can be exploited to conjugate nucleic acids

with various tag molecules invaluable in research. For example,

The achievement of only one nucleic acid unit conjugated

with each peptide molecule in zero-linker two-step phosphor-

amidation reactions provided relatively uniform and predictable

POC structures and could be explained by a strong electrostatic

attraction between nucleic acids and peptides. Nucleic acids are

well-known polyelectrolytes laden with negatively charged

phosphate groups at a physiological pH. On the other hand,

O

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

(2) Douglas, S. M., Bachelet, I., and Church, G. M. (2012) A logic-

gated nanorobot for targeted transport of molecular payloads. Science

335, 831−834.

(3) Juliano, R., Bauman, J., Kang, H., and Ming, X. (2009) Biological

barriers to therapy with antisense and siRNA oligonucleotides. Mol.

Pharmaceutics 6, 686−695.

(4) Whitehead, K. A., Langer, R., and Anderson, D. G. (2009)

Knocking down barriers: advances in siRNA delivery. Nat. Rev. Drug

Discovery 8, 129−138.

(5) Watts, J. K., and Corey, D. R. (2012) Silencing disease genes in

the laboratory and the clinic. J. Pathol. 226, 365−379.

(6) Burnett, J. C., and Rossi, J. J. (2012) RNA-based therapeutics:

current progress and future prospects. Chem. Biol. 19, 60−71.

(7) Juliano, R. L., Ming, X., and Nakagawa, O. (2012) The chemistry

and biology of oligonucleotide conjugates. Acc. Chem. Res. 45, 1067−

1076.

we had developed a facile and low-cost universal ﬂuorescent

labeling method for DNA and RNA based on the advanced

two-step phosphoramidation reaction (Wang et al., manuscript

in preparation). We expect broader applications of nucleic acid

phosphoramidation reactions in research and in clinic in the

future.

ASSOCIATED CONTENT

■

S

* Supporting Information

Full details of supporting ﬁgures, scheme and chart referenced

in the text, NMR and MS spectra for new compounds (4, 6a,

6b, 7a, 7b, 9, 12, and 13) and MALDI-TOF spectra for three

POCs. This material is available free of charge via the Internet

at http://pubs.acs.org.

(8) Ivanova, G. D., Arzumanov, A., Abes, R., Yin, H., Wood, M. J. A.,

Lebleu, B., and Gait, M. J. (2008) Improved cell-penetrating peptide-

PNA conjugates for splicing redirection in HeLa cells and exon

skipping in mdx mouse muscle. Nucleic Acids Res. 36, 6418−6428.

(9) Ming, X., Alam, M. R., Fisher, M., Yan, Y., Chen, X., and Juliano,

R. L. (2010) Intracellular delivery of an antisense oligonucleotide via

endocytosis of a G protein-coupled receptor. Nucleic Acids Res. 38,

6567−6576.

(10) Nakagawa, O., Ming, X., Huang, L., and Juliano, R. L. (2010)

Targeted intracellular delivery of antisense oligonucleotides via

conjugation with small-molecule ligands. J. Am. Chem. Soc. 132,

8848−8849.

(11) Said Hassane, F., Saleh, A., Abes, R., Gait, M., and Lebleu, B.

(2010) Cell penetrating peptides: overview and applications to the

delivery of oligonucleotides. Cell. Mol. Life Sci. 67, 715−726.

(12) Vives, E., Schmidt, J., and Pelegrin, A. (2008) Cell-penetrating

and cell-targeting peptides in drug delivery. Biochim. Biophys. Acta

1786, 126−138.

(13) Derossi, D., Calvet, S., Trembleau, A., Brunissen, A., Chassaing,

G., and Prochiantz, A. (1996) Cell internalization of the third helix of

the antennapedia homeodomain is receptor-independent. J. Biol. Chem.

271, 18188−18193.

AUTHOR INFORMATION

■

Corresponding Author

*Department of Medicinal and Applied Chemistry, Kaohsiung

Medical University, Kaohsiung, 80708, Taiwan; E-mail:

tzupinw@cc.kmu.edu.tw; Tel: +886-07-312-1101, ext. 2756;

Fax: +886-07-312-5339 (T.-P. Wang). School of Medicine,

Graduate Institute of Medicine, College of Medicine,

Kaohsiung Medical University, Kaohsiung, 80708, Taiwan;

Division of Infectious Diseases, Department of Internal

Medicine, Kaohsiung Medical University Hospital, Kaohsiung

Medical University, Kaohsiung, 80708, Taiwan; E-mail:

infchen@gmail.com (Y.-H. Chen).

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

We thank Dr. Susan Fetzer for critical reading of the

manuscript. We also thank Drs. Shih-Shin Liang and Chi-Yu

Lu for providing expertise in MALDI-TOF MS analysis. This

work was supported with grants from the National Science

Council of Taiwan awarded to T.-P. W. (99-2113-M-037-008-

and 100-2113-M-037-008-). M. H. W. is supported by the

undergraduate research program funded by National Science

Council of Taiwan (101CFD2100044).

(14) Good, L., Awasthi, S. K., Dryselius, R., Larsson, O., and Nielsen,

P. E. (2001) Bactericidal antisense effects of peptide-PNA conjugates.

Nat. Biotechnol. 19, 360−364.

(15) Jha, D., Mishra, R., Gottschalk, S., Wiesmuller, K.-H., Ugurbil,

̈

K., Maier, M. E., and Engelmann, J. (2011) CyLoP-1: A novel

cysteine-rich cell-penetrating peptide for cytosolic delivery of cargoes.

Bioconjugate Chem. 22, 319−328.

ABBREVIATIONS

■

(16) Saleh, A. F., Arzumanov, A., Abes, R., Owen, D., Lebleu, B., and

Gait, M. J. (2010) Synthesis and splice-redirecting activity of branched,

arginine-rich peptide dendrimer conjugates of peptide nucleic acid

oligonucleotides. Bioconjugate Chem. 21, 1902−1911.

POCs, peptide−oligonucleotide conjugates; Tat₄₈₋₅₇peptide,

amino acid residues 48 to 57 of the HIV Tat protein; siRNA,

small interference RNA; miRNA, microRNA; CPPs, cell-

penetrating peptides; CTPs, cell-targeting peptides; urea-

PAGE, urea polyacrylamide gel electrophoresis; SAv, strepta-

vidin; HOBt, 1-hydroxybenzotriazole; DIPEA, N,N-diisopropy-

lethylamine; EDC, 1-ethyl-3-(3-dimethylaminopropyl)-

carbodiimide; TFA, triﬂuoroacetic acid; AMAS, N-(α-maleimi-

doacetoxy) succinimide ester; NHS, N-hydroxysuccinimide;

DCC, N,N′-dicyclohexylcarbodiimide; Et₃N, triethylamine;

DTT, ^DL-dithiothreitol; EPPS, 3-[4-(2-hydroxyethyl)-1-

piperazinyl]propanesulfonic acid; DEPC, diethylpyrocarbonate;

MTT, methylthiazoletetrazolium; FITC, ﬂuorescein isothiocya-

nate; PBS, phosphate buﬀer saline; DMAP, 4-dimethylamino-

pyridine.

(17) Vaara, M., and Porro, M. (1996) Group of peptides that act

synergistically with hydrophobic antibiotics against gram-negative

enteric bacteria. Antimicrob. Agents Chemother. 40, 1801−1805.

̀

(18) Vives, E., Brodin, P., and Lebleu, B. (1997) A truncated HIV-1

Tat protein basic domain rapidly translocates through the plasma

membrane and accumulates in the cell nucleus. J. Biol. Chem. 272,

16010−16017.

(19) 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.

(20) Wang, T.-P., Chiou, Y.-J., Chen, Y., Wang, E.-C., Hwang, L.-C.,

Chen, B.-H., Chen, Y.-H., and Ko, C.-H. (2010) Versatile

phosphoramidation reactions for nucleic acid conjugations with

peptides, proteins, chromophores, and biotin derivatives. Bioconjugate

Chem. 21, 1642−1655.

REFERENCES

■

(21) Boger, D. L., Zhou, J., Winter, B., and Kitos, P. A. (1995) Key

analogs of the tetrapeptide subunit of RA-VII and deoxybouvardin.

Bioorg. Med. Chem. 3, 1579−1593.

(1) Bennett, C. F., and Swayze, E. E. (2010) RNA targeting

therapeutics: molecular mechanisms of antisense oligonucleotides as a

therapeutic platform. Annu. Rev. Pharmacol. Toxicol. 50, 259−293.

P

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Bioconjugate Chemistry

Article

(22) Allen, V. C., Robertson, C. C., Turega, S. M., and Philp, D.

(2010) A simple network of synthetic replicators can perform the

logical OR operation. Org. Lett. 12, 1920−1923.

(23) Wang, T.-P., Su, Y.-C., Chen, Y., Liou, Y.-M., Lin, K.-L., Wang,

E.-C., Hwang, L.-C., Wang, Y.-M., and Chen, Y.-H. (2012) In vitro

selection and characterization of a novel Zn(II)-dependent phosphor-

othiolate thiolesterase ribozyme. Biochemistry 51, 496−510.

(24) Lin, C.-M., Chiu, J.-H., Wu, I. H., Wang, B.-W., Pan, C.-M., and

Chen, Y.-H. (2010) Ferulic acid augments angiogenesis via VEGF,

PDGF and HIF-1α. J. Nutr. Biochem. 21, 627−633.

(25) Proudnikov, D., and Mirzabekov, A. (1996) Chemical methods

of DNA and RNA fluorescent labeling. Nucleic Acids Res. 24, 4535−

4542.

(26) Grauer, A., Riechers, A., Ritter, S., and Konig, B. (2008)

̈

Synthetic receptors for the differentiation of phosphorylated peptides

with nanomolar affinities. Chem.Eur. J. 14, 8922−8927.

(27) Shchepinov, M. S., Denissenko, M. F., Smylie, K. J., Worl, R. J.,

̈

Leppin, A. L., Cantor, C. R., and Rodi, C. P. (2001) Matrix-induced

fragmentation of P3′-N5′ phosphoramidate-containing DNA: high-

throughput MALDI-TOF analysis of genomic sequence polymor-

phisms. Nucleic Acids Res. 29, 3864−3872.

(28) Milligan, J. F., Groebe, D. R., Witherell, G. W., and Uhlenbeck,

O. C. (1987) Oligoribonucleotide synthesis using T7 RNA polymerase

and synthetic DNA templates. Nucleic Acids Res. 15, 8783−8798.

(29) Ho, N. W. Y., Duncan, R. E., and Gilham, P. T. (1981)

Esterification of terminal phosphate groups in nucleic acids with

sorbitol and its application to the isolation of terminal polynucleotide

fragments. Biochemistry 20, 64−67.

(30) Dawson, R. M. C., Elliott, D. C., Elliott, W. H., and Jones, K. M.

(1986) Data for Biochemical Research, 3rd ed.; Oxford Science

Publication.

(31) Hermanson, G. T. (1996) Bioconjugate techniques, Academic

Press, San Diego, USA.

Q

dx.doi.org/10.1021/bc300444y | Bioconjugate Chem. XXXX, XXX, XXX−XXX

Article Doi

Advanced aqueous-phase phosphoramidation reactions for effectively synthesizing peptide-oligonucleotide conjugates trafficked into a human cell line

DOI: 10.1021/bc300444y

Source and publish data:

Authors:

Article abstract of DOI:10.1021/bc300444y

Full text of DOI:10.1021/bc300444y

Products guided by the article

R&D Labs maybe for 25021-08-3

Relevant to this article

Hot Product