6
A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivativesZhongyu Li and Ying Chau * Received 14th October 2010, Accepted 15th December 2010 DOI: 10.1039/c0py00339e Allyl-(PEG-OH) 2 , a dual-branched poly(ethylene glycol) (PEG) with a latent group amenable for modification at the junction, was successfully synthesized using trimethylolpropane allyl ether (TMPAE), a commercially available compound, as the initiator for anionic polymerization. To demonstrate the versatility of this approach, derivatives of the branched PEG were formed using simple modification. The chain ends of PEG were modified to inert methoxy groups and active functional groups. Using an orthogonal reaction procedure, the allyl junction was modified to carboxyl and amino group. The synthesis route was short, quantitative, and easily controlled. No cumbersome purification was needed. The branched PEG and its derivatives were characterized by SEC, 1 H and 13 C NMR, and MALDI-TOF mass spectroscopy. Introduction Poly(ethylene glycol) (PEG), a polymer consisting of ethylene oxide as repeating units, is widely used in biotechnology and medicine due to its excellent safety record in clinical use and its many peculiar properties, including good solubility in a wide range of organic and aqueous media, polymer backbone flexi- bility, stability in physiological conditions, non-adhesiveness to protein, biocompatibility and the ease of excretion from living organisms. 1 The most common applications are found in drug delivery and biosensor preparation, in which biological macro- molecules, colloidal carriers, low molecular weight compounds and device surfaces are modified with PEG. 1,2 In these applications, some advantages have been observed for dual-branched PEG (PEG 2 ) over linear PEG. Nanomaterials, including carbon nanotubes, gold nanoparticles, and gold nanorods, when grafted by branched polyethylene glycol, were found to possess high aqueous solubility and ultra-long circula- tion half-life upon intravenous injection into mice. 3 Protein conjugated with a branched PEG has longer in vivo circulation half-life, improved stability against proteolysis, and reduced immunogenicity compared to linear PEG-proteins. This is explained by the greater hydrodynamic volume of branched PEG-proteins, which slows the rate of renal excretion, and the ‘‘umbrella-like’’ structure of the branched PEG around the protein molecule, which provides better shielding. 4 The conju- gation of interferon-a2b (IFN-a2b) to a branched PEG (PEG 2 , 40 K) was studied extensively from experiments to clinical pha- ses 4d,5 and is now a product for the treatment of hepatitis C under the trade name PEGASYS. Until now, almost all branched PEGs have been synthesized by cumbersome methods involving tri-functional linkers. A common method uses lysine to couple two methoxy PEG (mPEG) chains to its alpha and epsilon amino groups and leaves the carboxylic group to be activated for subsequent protein conjugation. 4d,5a,5b Similarly, 2-(2-aminoethoxy)ethanol 6 and tert-butyl protected N,N-bis(2-hydroxyethyl)glycine (bicine) 7 have been used as tri-functional linkers for the synthesis of branched PEG. The disadvantages are: 1) multiple and lengthy reactions steps; 2) difficult and expensive purification (using chromatography) for the separation of PEG 2 and linear PEG chains; 3) the limitation of the active functional group to be carboxyl; 4) the water susceptibility of the bond between PEG chains and the tri-functional linker (e.g., urethane 4d,6 and ester 7 linkage). Branched PEG synthesized by the polymerization of ethylene oxide (EO) from an initiator containing two hydroxyl groups and one protected or latent group seems to circumvent these problems. Protected glycerol is an example of this kind of initi- ator. From this initiator, a series of branched PEG derivatives have been synthesized and commercialized by NOF Corporation Ltd. (Tokyo, Japan). 8 However, this particular initiator has different reactivity at the alpha and beta hydroxyl positions, making it difficult to control the evenness of PEG chain lengths at the branches. We report herein a one-pot, inexpensive and high-yielding method to synthesize branched PEG carrying two chains of the statistically same length. Our strategy is to initiate anionic polymerization of ethylene oxide (EO) from trimethylolpropane allyl ether (TMPAE), a commercially available chemical with Department of Chemical and Biomolecular Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong, China. E-mail: [email protected]; [email protected]; Fax: +852-2358 0054; Tel: +852-2358 8935 † Electronic supplementary information (ESI) available: The SEC trace of allyl-(PEG-OH) 2 . See DOI: 10.1039/c0py00339e This journal is ª The Royal Society of Chemistry 2011 Polym. Chem., 2011, 2, 873–878 | 873 Dynamic Article Links C < Polymer Chemistry Cite this: Polym. Chem., 2011, 2, 873 www.rsc.org/polymers PAPER Downloaded on 19 March 2013 Published on 04 January 2011 on http://pubs.rsc.org | doi:10.1039/C0PY00339E View Article Online / Journal Homepage / Table of Contents for this issue

A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivatives

  • Upload
    ying

  • View
    218

  • Download
    1

Embed Size (px)

Citation preview

Page 1: A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivatives

Dynamic Article LinksC<PolymerChemistry

Cite this: Polym. Chem., 2011, 2, 873

www.rsc.org/polymers PAPER

Dow

nloa

ded

on 1

9 M

arch

201

3Pu

blis

hed

on 0

4 Ja

nuar

y 20

11 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0P

Y00

339E

View Article Online / Journal Homepage / Table of Contents for this issue

A facile synthesis of branched poly(ethylene glycol) and its heterobifunctionalderivatives†

Zhongyu Li and Ying Chau*

Received 14th October 2010, Accepted 15th December 2010

DOI: 10.1039/c0py00339e

Allyl-(PEG-OH)2, a dual-branched poly(ethylene glycol) (PEG) with a latent group amenable for

modification at the junction, was successfully synthesized using trimethylolpropane allyl ether

(TMPAE), a commercially available compound, as the initiator for anionic polymerization. To

demonstrate the versatility of this approach, derivatives of the branched PEG were formed using simple

modification. The chain ends of PEG were modified to inert methoxy groups and active functional

groups. Using an orthogonal reaction procedure, the allyl junction was modified to carboxyl and amino

group. The synthesis route was short, quantitative, and easily controlled. No cumbersome purification

was needed. The branched PEG and its derivatives were characterized by SEC, 1H and 13C NMR, and

MALDI-TOF mass spectroscopy.

Introduction

Poly(ethylene glycol) (PEG), a polymer consisting of ethylene

oxide as repeating units, is widely used in biotechnology and

medicine due to its excellent safety record in clinical use and its

many peculiar properties, including good solubility in a wide

range of organic and aqueous media, polymer backbone flexi-

bility, stability in physiological conditions, non-adhesiveness to

protein, biocompatibility and the ease of excretion from living

organisms.1 The most common applications are found in drug

delivery and biosensor preparation, in which biological macro-

molecules, colloidal carriers, low molecular weight compounds

and device surfaces are modified with PEG.1,2

In these applications, some advantages have been observed for

dual-branched PEG (PEG2) over linear PEG. Nanomaterials,

including carbon nanotubes, gold nanoparticles, and gold

nanorods, when grafted by branched polyethylene glycol, were

found to possess high aqueous solubility and ultra-long circula-

tion half-life upon intravenous injection into mice.3 Protein

conjugated with a branched PEG has longer in vivo circulation

half-life, improved stability against proteolysis, and reduced

immunogenicity compared to linear PEG-proteins. This is

explained by the greater hydrodynamic volume of branched

PEG-proteins, which slows the rate of renal excretion, and the

‘‘umbrella-like’’ structure of the branched PEG around the

protein molecule, which provides better shielding.4 The conju-

gation of interferon-a2b (IFN-a2b) to a branched PEG (PEG2,

Department of Chemical and Biomolecular Engineering, The Hong KongUniversity of Science and Technology, Clear Water Bay, Hong Kong,China. E-mail: [email protected]; [email protected]; Fax: +852-2358 0054;Tel: +852-2358 8935

† Electronic supplementary information (ESI) available: The SEC traceof allyl-(PEG-OH)2. See DOI: 10.1039/c0py00339e

This journal is ª The Royal Society of Chemistry 2011

40 K) was studied extensively from experiments to clinical pha-

ses4d,5 and is now a product for the treatment of hepatitis C under

the trade name PEGASYS.

Until now, almost all branched PEGs have been synthesized

by cumbersome methods involving tri-functional linkers. A

common method uses lysine to couple two methoxy PEG

(mPEG) chains to its alpha and epsilon amino groups and leaves

the carboxylic group to be activated for subsequent protein

conjugation.4d,5a,5b Similarly, 2-(2-aminoethoxy)ethanol6 and

tert-butyl protected N,N-bis(2-hydroxyethyl)glycine (bicine)7

have been used as tri-functional linkers for the synthesis of

branched PEG. The disadvantages are: 1) multiple and lengthy

reactions steps; 2) difficult and expensive purification (using

chromatography) for the separation of PEG2 and linear PEG

chains; 3) the limitation of the active functional group to be

carboxyl; 4) the water susceptibility of the bond between PEG

chains and the tri-functional linker (e.g., urethane4d,6 and ester7

linkage).

Branched PEG synthesized by the polymerization of ethylene

oxide (EO) from an initiator containing two hydroxyl groups and

one protected or latent group seems to circumvent these

problems. Protected glycerol is an example of this kind of initi-

ator. From this initiator, a series of branched PEG derivatives

have been synthesized and commercialized by NOF Corporation

Ltd. (Tokyo, Japan).8 However, this particular initiator has

different reactivity at the alpha and beta hydroxyl positions,

making it difficult to control the evenness of PEG chain lengths

at the branches.

We report herein a one-pot, inexpensive and high-yielding

method to synthesize branched PEG carrying two chains of the

statistically same length. Our strategy is to initiate anionic

polymerization of ethylene oxide (EO) from trimethylolpropane

allyl ether (TMPAE), a commercially available chemical with

Polym. Chem., 2011, 2, 873–878 | 873

Page 2: A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivatives

Dow

nloa

ded

on 1

9 M

arch

201

3Pu

blis

hed

on 0

4 Ja

nuar

y 20

11 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0P

Y00

339E

View Article Online

two uniform alpha hydroxyl groups and one allyl group. After

polymerization, the allyl group can be modified to carboxyl or

amino group with simple procedures, while the PEG chain ends

can be readily functionalized. These dual-branched, hetero-

biofunctional PEGs will be useful for drug and protein delivery,

surface coating, and for preparing block copolymers with func-

tionalized terminals.

Experimental part

Materials

All starting compounds were used as received without additional

purification except for those specified. Chemicals were purchased

from Aldrich unless otherwise indicated. Tetrahydrofuran

(THF) (Merck, 99%) was refluxed over sodium wire and distilled

from sodium naphthalenide solution. Trimethylolpropane allyl

ether (TMPAE) (Fluka, 98%) and dimethyl sulfoxide (DMSO)

(Merck, 98%) were distilled over CaH2 under reduced pressure

just before use. Methyl iodide (Riedel deHaer, 99%) was used

directly. Ethylene oxide (EO, 99.7%) was purchased from Hong

Kong Special Gas Company and used directly. Diphenylmethyl

potassium (DPMK) was prepared as described elsewhere.9

Methods

1H NMR and 13C NMR spectra were obtained on a DMX 400

MHz spectrometer with tetramethylsilane (TMS) as the internal

standard and CDCl3 as the solvent. Size exclusion chromato-

graphy (SEC) was performed in 0.1 M NaNO3 at 40 �C with an

elution rate of 0.5 mL min�1 on a Waters HPLC system equipped

with a G1310A pump and a G1362A refractive index (RI)

detector. Ultrahydrogel 250 (Waters) and Ultrahydrogel 1000

(Waters) columns were used in series and calibrated by poly-

ethylene glycol standards (Polymer Source, Inc., Canada).

MALDI-TOF MS spectra were recorded using Bruker REFLEX

III. a-cyano-4-hydroxycinnamic acid (CHCA) was used as the

matrix for the ionization operated in the reflection mode.

Scheme 1 Synthesis of branched PEG (a

874 | Polym. Chem., 2011, 2, 873–878

The synthesis steps of the branched PEG and its derivatives are

illustrated in Scheme 1. Details are provided in the following

sections.

Synthesis of allyl-(PEG-OH)2 (1)

A 150 mL stainless steel kettle was vacuumed at 80 �C for 24 h

and cooled to room temperature and then to 0 �C under icy water

bath. Anhydrous trimethylolpropane allyl ether (TMPAE)

(1.74 g, 0.01mol) was dissolved in 50 mL of mixed solvents of

DMSO and THF (v/v: 3/2). A solution of DPMK in THF

(6.7 mL, 0.6 M solution) was slowly added. The orange-red color

of DPMK was changed to yellow when alkoxide was formed.

The homogeneous initiator solution was introduced into the

cooled kettle by a syringe, followed with the addition of ethylene

oxide (EO). After the solution was stirred at 50 �C for 24 h,

polymerization was terminated by adding of a few drops of

acidified methanol (0.1 M HCl in methanol). All the solvents

were removed by reduced distillation. The crude product was

dissolved in CH2Cl2, filtered, and dried over anhydrous MgSO4,

then precipitated in diethyl ether. Allyl-(PEG-OH)2 (1) was

obtained as a white powder at a reaction yield of 99%. 1H NMR

(ppm) (400 MHz, CDCl3): 0.84 (t, J ¼ 7.57 Hz, CH3CH2–), 1.38

(q, J ¼ 7.56 Hz, CH3CH2–), 3.25 and 3.28 (s, –C(CH2O–)3, 3.45–

3.80 (m, –CH2CH2O– of PEG main chain), 4.02 (d, J ¼ 5.38 Hz,

–O–CH2–CH]CH2), 5.01–5.22 (dd, J ¼ 17.33 Hz and J ¼ 10.25

Hz, –CH]CH2), 5.82–5.95 (m, J ¼ 5.13 Hz, –CH]CH2); 13C

NMR (ppm) (400 MHz, CDCl3): 8.78 (CH3CH2–), 24.5

(CH3CH2–), 43.2 (–C(CH2O–)3, 70.4–71.5 (–CH2CH2O– of PEG

main chain), 73.6 (–C(CH2O–)3, 74.2 (–O–CH2–CH]CH2),

121.2 (–CH]CH2), 132.4 (–CH]CH2); SEC: Mn¼ 5.27� 103 g

mol�1, Mw/Mn ¼ 1.08.

Synthesis of allyl-(mPEG)2 (2)

Synthesis of allyl-(mPEG)2 was performed according to

a method adapted from a previous report.10 One gram (0.38

mmol hydroxyl groups) of allyl-(PEG-OH)2 (Mn ¼ 5.27 � 103 g

mol�1) was removed moisture by azeotropic distillation with

llyl-(PEG-OH)2) and its derivatives.

This journal is ª The Royal Society of Chemistry 2011

Page 3: A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivatives

Dow

nloa

ded

on 1

9 M

arch

201

3Pu

blis

hed

on 0

4 Ja

nuar

y 20

11 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0P

Y00

339E

View Article Online

toluene just before use, then reacted with 0.01 g of NaH (0.42

mmol) in 10 mL of anhydrous tetrahydrofuran (THF) at room

temperature under a nitrogen stream for 20 min. This was fol-

lowed by the addition of 0.10 g of methyl iodide (0.7 mmol)

under vigorous stirring at room temperature for 24 h. After

neutralization by 0.1 M HCl solution, the solvent was removed

by rotary evaporation. The product was dissolved in water and

was extracted by dichloromethane (DCM). The organic phase

was dried over anhydrous MgSO4. After filtration and concen-

tration, the polymer was precipitated in cold diethyl ether twice

to afford a white powder (yield ¼ 95%). 1H NMR (ppm) (400

MHz, CDCl3): 0.84 (t, J ¼ 7.57 Hz, CH3CH2–), 1.38 (q, J ¼ 7.56

Hz, CH3CH2–), 3.25 and 3.28 (s, –C(CH2O–)3, 3.38 (s, CH3O–),

3.45–3.80 (m, –CH2CH2O– of PEG main chain), 4.02 (d, J¼ 5.38

Hz, –O–CH2–CH]CH2), 5.01–5.22 (dd, J ¼ 17.33 Hz and J ¼10.25 Hz, –CH]CH2), 5.82–5.95 (m, J ¼ 5.13 Hz, –CH]CH2);

SEC: Mn ¼ 5.28 � 103 g mol�1, Mw/Mn ¼ 1.08.

Synthesis of COOH-(mPEG)2 (3)

The carboxylation reaction of the allyl terminus of polymer allyl-

(mPEG)2 was conducted by the radical addition reaction using 3-

mercaptopropionic acid.10,11 Allyl-(mPEG)2 (1 g, 0.2 mmol) with

moisture removed by azeotropic distillation with toluene just

before use, was mixed with a solution containing 424 mg of 3-

mercaptopropionic acid (4 mmol, 20 equivalent) and 36.0 mg of

azobisisobutyronitrile (AIBN) (0.1 mmol, 1 equivalent) in 5 mL

anhydrous dimethylformamide (DMF). The reaction mixture

was stirred at 65 �C for 24 h under nitrogen atmosphere. The

polymer was precipitated twice in a large excess of diethyl ether.

The polymer COOH-(mPEG)2 was obtained as a white power

(924 mg, yield ¼ 84%). 1H NMR (ppm) (400 MHz, CDCl3): 0.84

(t, J¼ 7.57 Hz, CH3CH2–), 1.38 (q, J¼ 7.56 Hz, CH3CH2–), 1.76

(p, J ¼ 6.71 Hz, –OCH2CH2CH2S–), 2.56 (t, J ¼ 7.08 Hz,

–CH2CH2–S–CH2– and t, J ¼ 7.32 Hz, –S–CH2CH2COOH),

2.96 (t, J¼ 7.32 Hz and J¼ 7.08 Hz, –CH2COOH), 3.25 and 3.28

(s, –C(CH2O–)3, 3.38 (s, CH3O–), 3.45–3.80 (m, –CH2CH2O– of

PEG main chain and –OCH2CH2CH2S–); SEC: Mn¼ 5.28� 103

g mol�1, Mw/Mn ¼ 1.08.

Synthesis of NH2-(mPEG)2 (4)

In a typical reaction, 400 mg of allyl-(mPEG)2 (0.077 mmol) in 10

mL of anhydrous DMF was reacted with 175 mg of 2-amino-

ethanethiol hydrochloride (1.54 mmol, 20 equiv.) in the presence

of 12.6 mg of AIBN (0.077 mmol, 1 equiv). The reaction mixture

was stirred at 65 �C for 24 h under nitrogen atmosphere. After

the reaction, the solution was precipitated in diethyl ether twice.

The resulting white product was dissolved in methanol, and

4.3 mg (0.077 mmol, 1 equiv) of potassium hydroxide dissolved

in water was added. The mixture was stirred for approximately

4 h. Then, methanol was partially evaporated and diluted with

water (30 mL), and extracted by dichloromethane (3 � 50 mL).

The combined organic layer was dried over MgSO4, filtered, and

concentrated. The polymer was reprecipitated from an excess

volume of ether twice to afford a white powder (412 mg, yield ¼79.5%). 1H NMR (ppm) (400 MHz, CDCl3): 0.84 (t, J¼ 7.57 Hz,

CH3CH2–), 1.38 (q, J ¼ 7.56 Hz, CH3CH2–), 1.76 (p, J ¼ 6.71

Hz, –OCH2CH2CH2S–), 2.41 (t, J ¼ 7.57 Hz, –CH2–S–CH2–),

This journal is ª The Royal Society of Chemistry 2011

2.66 (t, J ¼ 7.57 Hz, –CH2–S–CH2–), 2.94 (t, J ¼ 7.08 Hz,

–CH2CH2NH2), 3.25 and 3.28 (s, –C(CH2O–)3, 3.38 (s, CH3O–),

3.45–3.80 (m, –CH2CH2O– of PEG main chain and

–OCH2CH2CH2S–); SEC: Mn¼ 5.28� 103 g mol�1, Mw/Mn¼ 1.08.

Synthesis of allyl-(PEG-alkyne)2 (5)

Synthesis of allyl-(PEG-alkyne)2 was performed using a method

adapted from a previous report.12 One gram (0.38 mmol

hydroxyl groups)of allyl-(PEG-OH)2 (Mn ¼ 5.27 � 103 g mol�1),

with moisture removed by azeotropic distillation with toluene

just before use was mixed with 0.12 g NaH (0.5 mmol) in 10 mL

anhydrous THF under nitrogen atmosphere at room tempera-

ture for 1 h, then 0.060 g propargyl bromide (0.5 mmol) was

added at room temperature for 24 h. After neutralization by

hydrogen chloride solution, the solvent was removed by rotary

evaporation. The crude product was dissolved in water and

extracted by DCM (3� 50 mL). The combined organic layer was

dried over MgSO4, filtered, and concentrated. The polymer was

reprecipitated from an excess volume of diethyl ether twice to

afford a white powder (835 mg, yield ¼ 83%). 1H NMR (ppm)

(400 MHz, CDCl3): 0.84 (t, J ¼ 7.57 Hz, CH3CH2–), 1.38 (q, J ¼7.56 Hz, CH3CH2–), 2.40 (s, J ¼ 2.19 Hz, –CCH), 3.25 and 3.28

(s, –C(CH2O–)3, 3.45–3.80 (m, –CH2CH2O– of PEG main

chain), 4.02 (d, J ¼ 5.38 Hz, –O–CH2–CH]CH2), 4.17 (s, J ¼2.44 Hz, –CH2–CCH), 5.01–5.22 (dd, J¼ 17.33 Hz and J¼ 10.25

Hz, –CH]CH2), 5.82–5.95 (m, J ¼ 5.13 Hz, –CH]CH2); SEC:

Mn ¼ 5.28 � 103 g mol�1, Mw/Mn ¼ 1.08.

Synthesis of allyl-(PEG-N3)2 (7)

The azide end group was introduced by the tosylation of the

hydroxyl terminus and the subsequent substitution with sodium

azide, in accordance with a previously reported method.12 One

gram (0.38 mmol hydroxyl groups) of allyl-(PEG-OH)2 (Mn ¼5.27 � 103 g mol�1) with moisture removed by azeotropic

distillation with toluene just before use was dissolved in anhy-

drous THF (10 mL), followed by the addition of triethylamine

(40 mg, 0.4 mmol). The mixture was then added to a solution of

p-toluenesulfonyl chloride (57 mg, 0.5 mmol) in THF (8 mL)

under nitrogen atmosphere, and stirred overnight at room

temperature. After the reaction, THF was partially evaporated

under reduced pressure. The residue was dissolved in water

(20 mL) and extracted with dichloromethane (3 � 50 mL). The

organic layers were combined and dried over anhydrous MgSO4.

After filtration and concentration, the polymer was recovered by

precipitation into diethyl ether and dried in vacuo, yielding

a white powder (6) (yield ¼ 87%). Allyl-(PEG-OTs)2: 1H NMR

(ppm) (400 MHz, CDCl3): 0.84 (t, J ¼ 7.57 Hz, CH3CH2–), 1.38

(q, J ¼ 7.56 Hz, CH3CH2–), 2.45 (s, CH3–C6H4–), 3.25 and 3.28

(s, –C(CH2O–)3, 3.45–3.80 (m, –CH2CH2O– of PEG main

chain), 4.02 (d, J¼ 5.38 Hz, –O–CH2–CH]CH2), 5.01–5.22 (dd,

J ¼ 17.33 Hz and J ¼ 10.25 Hz, –CH]CH2), 5.82–5.95 (m, J ¼5.13 Hz, –CH]CH2); 7.35 (d, J ¼ 7.82 Hz, two CH in phenyl

ring close to –CH3), 7.79 (d, J¼ 7.81 Hz, other two CH in phenyl

ring).

This tosylated polymer (6) (500 mg, 0.1 mmol) was dissolved in

anhydrous DMF (12 mL), followed by sodium azide (1.3 mg,

2 mmol) addition, and was stirred for 2 days at 30 �C. DCM

Polym. Chem., 2011, 2, 873–878 | 875

Page 4: A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivatives

Dow

nloa

ded

on 1

9 M

arch

201

3Pu

blis

hed

on 0

4 Ja

nuar

y 20

11 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0P

Y00

339E

View Article Online

(100 mL) was then added, and the reaction mixture was washed

five times with water and brine. The organic layer was dried over

anhydrous MgSO4, filtered, concentrated, and then precipitated

in diethyl ether. Allyl-(PEG-N3)2 (7) was obtained as a white

powder (447 mg, yield ¼ 92%). 1H NMR (ppm) (400 MHz,

CDCl3): 0.84 (t, J ¼ 7.57 Hz, CH3CH2–), 1.38 (q, J ¼ 7.56 Hz,

CH3CH2–), 1.78 (m, J¼ 7.28 Hz, –CH2CH2N3); 3.25 and 3.28 (s,

–C(CH2O–)3, 3.45–3.80 (m, –CH2CH2O– of PEG main chain

and –CH2CH2N3), 4.02 (d, J ¼ 5.38 Hz, –O–CH2–CH]CH2),

5.01–5.22 (dd, J ¼ 17.33 Hz and J ¼ 10.25 Hz, –CH]CH2),

5.82–5.95 (m, J ¼ 5.13 Hz, –CH]CH2); 13C NMR (ppm) (400

MHz, CDCl3): 8.78 (CH3CH2–), 24.5 (CH3CH2–), 43.2

(–C(CH2O–)3, 50.8 (–CH2CH2N3), 69.5 (–CH2CH2N3), 70.4–

71.5 (–CH2CH2O– of PEG main chain), 73.6 (–C(CH2O–)3, 74.2

(–O–CH2–CH]CH2), 117.5 (–CH]CH2), 132.4 (–CH]CH2);

SEC: Mn ¼ 5.28 � 103 g mol�1, Mw/Mn ¼ 1.08.

Synthesis of COOH-(PEG-N3)2 (8) and NH2-(PEG-N3)2 (9)

The synthesis of COOH-(PEG-N3)2 and NH2-(PEG-N3)2 are

similar to the synthesis of COOH-(mPEG)2 and NH2-(mPEG)2,

respectively. Instead of using allyl-(mPEG)2 (2) as the polymer

for modification, (allyl-(PEG-N3) 2 (7) was used.

COOH-(PEG-N3)2 (8) was obtained as a white powder with

a reaction yield of 94%. 1H NMR (ppm) (400 MHz, CDCl3): 0.84

(t, J¼ 7.57 Hz, CH3CH2–), 1.38 (q, J¼ 7.56 Hz, CH3CH2–), 1.76

(p, J ¼ 6.71 Hz, –OCH2CH2CH2S–), 1.82 (t, J ¼ 7.28 Hz,

–CH2CH2N3), 2.56 (t, J ¼ 7.08 Hz, –CH2CH2–S–CH2– and t,

J ¼ 7.32 Hz, –S–CH2CH2COOH), 2.96 (t, J ¼ 7.32 Hz and J ¼7.08 Hz, –CH2COOH), 3.25 and 3.28 (s, –C(CH2O–)3, 3.38 (s,

CH3O–), 3.45–3.80 (m, –CH2CH2O– of PEG main chain,

–OCH2CH2CH2S– and –CH2CH2N3); 13C NMR (ppm) (400

MHz, CDCl3): 8.78 (CH3CH2–), 24.5 (CH3CH2–), 27.1

(–SCH2CH2COOH), 34.8 (–SCH2CH2COOH), 40.2 (–C(CH2O–)3,

50.8 (–CH2CH2N3), 69.5 (–CH2CH2N3), 70.4–71.5 (–CH2CH2O–

of PEG main chain), 73.6 (–C(CH2O–)3, 173.4 (–SCH2CH2-

COOH); SEC: Mn ¼ 5.29 � 103 g mol�1, Mw/Mn ¼ 1.08.

NH2–(PEG-N3)2 (9) was obtained as a white powder with

a reaction yield of 91%. 1H NMR (ppm) (400 MHz, CDCl3): 0.84

(t, J¼ 7.57 Hz, CH3CH2-), 1.38 (q, J¼ 7.56 Hz, CH3CH2–), 1.76

(p, J ¼ 6.71 Hz, –OCH2CH2CH2S–), 1.82 (t, J ¼ 7.28 Hz,

–CH2CH2N3), 2.41 (t, J ¼ 7.57 Hz, –CH2–S–CH2–), 2.66 (t, J ¼7.57 Hz, –CH2–S–CH2–), 2.94 (t, J ¼ 7.08 Hz, –CH2CH2NH2),

3.25 and 3.28 (s, –C(CH2O–)3, 3.38 (s, CH3O–), 3.45–3.80 (m,

–CH2CH2O– of PEG main chain, –OCH2CH2CH2S– and

–CH2CH2N3); 13C NMR (ppm) (400 MHz, CDCl3): 8.78

(CH3CH2–), 24.5 (CH3CH2–), 28.5 (–SCH2CH2NH2), 40.3

(–SCH2CH2NH2), 43.2 (–C(CH2O–)3, 50.8 (–CH2CH2N3), 69.5

(–CH2CH2N3), 70.4–71.5 (–CH2CH2O– of PEG main chain), 73.6

(–C(CH2O–)3; SEC: Mn ¼ 5.30 � 103 g mol�1, Mw/Mn ¼ 1.08.

Results and discussion

Much attention has been given to improve the synthesis of

branched PEG.4d,6–8 We are motivated to develop a novel

procedure to overcome the drawbacks of conventional coupling

methods, which require multiple synthesis steps and difficult

purification. The choice of trimethylolpropane allyl ether

(TMPAE) as the initiator is inspired by the facile synthesis of

876 | Polym. Chem., 2011, 2, 873–878

heterobifunctional PEG initiated from allyl alcoholate.10,11

TMPAE allows the synthesis of branched PEG with equal chain

lengths via anionic polymerization of EO because the allyl group

is inert for the polymerization and does not affect the reactivity

of the two pendent alpha hydroxyl groups. In addition, the allyl

group is versatile and enables easy chemical modifications such

as addition reactions. The polymerization was carried out in

mixed solvent (THF/DMSO ¼ 3/2) as reported by us before9 and

the reaction time was dependent on the pre-designed number

average molecular weight (Mn). It was found that the polymer-

ization of branched PEG of Mn ¼ 5k and 20k required 24 h and

96 h respectively. After precipitation into ether, the product was

obtained as a white powder. Results of polymerization using

TMPAE as an initiator under different conditions are summa-

rized in Table 1. The molecular weight of the polymers (allyl-

(PEG-OH)2) (1) determined from SEC was close to that calcu-

lated by the initial monomer/initiator ratio, supporting that the

polymerization was complete and without detrimental side

reactions. The resulting polymers are unimodal with narrow

polydispersity (PDI) (Mw/Mn). The value of the Mn could be

controlled by the initial monomer/initiator ratio while retaining

a narrow polydispersity due to the nature of anionic polymeri-

zation.

The structure of allyl-(PEG-OH)2 was confirmed by 1H NMR

and 13C NMR. In the 1H NMR spectrum (Fig. 1A), the signals of

the protons of the allyl group are detected at d 4.02 ppm (d, –O–

CH2–CH]CH2), 5.01–5.22 ppm (dd, –CH]CH2), 5.82–5.95

ppm (m, –CH]CH2), respectively. The number average molec-

ular weight (Mn) of the polymers was determined by 1H NMR

spectrum on the basis of end group analysis using the following

equation: Mn ¼ 44:035�3AEO

4Amþ 174:23 Where AEO and Am are

the peak area of sum of protons in the PEG main chain at d ¼3.45 � 3.80 ppm and methyl protons at d ¼ 0.84 ppm respec-

tively; 174.13 and 44.05 are the molecular weight of TMPAE and

EO respectively. The Mn calculated by 1H NMR is very close to

that measured by SEC (Table 1).

Allyl-(PEG-OH)2 of pre-designed Mn, at 5k was further

characterized by MALDI-TOF MS spectroscopy (Fig. 2). The

polymer was confirmed to be unimodal and narrowly distributed

(Mw/Mn ¼ 1.03). Presence of side reaction product was not

indicated. The molecular weight found using MALDI-TOF was

5.36k, was in good agreement with the SEC and NMR results.

The major series of the molecular masses of the product is

expressed in the following equation:

Mw(MALDI-TOF) ¼ 44.035n(EO) + 174.23(TMPAE) +

22.99(sodium) (2)

where n is an integer, that confirms the initiator is TMPAE.

The polydispersities (PDIs) of allyl-(PEG-OH)2 characterized

by MOLDI-Tof MS are lower than those characterized by SEC

(Table 1). This phenomenon has been discussed in previous

report13 and the authors considered that the reason may be

a small amount of SEC axial dispersion.

From (1), we capped the terminal hydroxyl groups with methyl

iodine to prepare allyl-(mPEG)2 (2). Fig. 1A and Fig. 1B show

the 1H NMR spectrum of PEG before and after methylation. The

new peak at d ¼ 3.38 ppm was assigned to methoxy end group in

This journal is ª The Royal Society of Chemistry 2011

Page 5: A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivatives

Table 1 Results of anionic polymerizations of ethylene oxide (EO) with TMPAE as initiator

[EO]0/[TMPAE]0

10�3 � Mnb (g mol�1) Polydispersity Mw/Mn

g

Calcdc NMRd SECe MSf SECg MSh

1a 110 5.01 5.24 5.27 5.36 1.08 1.032a 450 20.02 20.64 20.85 ND 1.09 ND

a Reaction time of 1 and 2 are 24 h and 96 h respectively, and the yields of product 1 and 2 are 99% and 99% respectively. b Mn denotes number averagemolecular weight. c Determined from the following equation: Mn(calcd)¼Mw(EO)[EO]0/[TMPAE]0 + Mw(TMPAE)¼ 44.05 [EO]0/[TMPAE]0 + 174.23(3). d The number average molecular weight (Mn) of the polymers was determined by 1H NMR spectrum on the basis of end group analysis using the eqn(1). e Determined by SEC. f Determined by MALDI-TOF mass spectroscopy. g Polydispersity Index (PDI) ¼ Mw/Mn. Mw denotes weight averagemolecular weight. h The PDI according to MALDI-TOF MS.

Fig. 1 1H NMR spectra of allyl-(PEG-OH)2 (A), allyl-(mPEG)2 (B) and

COOH-(mPEG)2 (C) respectively, (CDCl3 at 20 �C).

Fig. 2 MALDI-TOF Mass spectrum of allyl-(PEG-OH)2.

Fig. 3 1H NMR spectrum of allyl-(PEG-alkyne)2 (CDCl3 at 20 �C).

Dow

nloa

ded

on 1

9 M

arch

201

3Pu

blis

hed

on 0

4 Ja

nuar

y 20

11 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0P

Y00

339E

View Article Online

Fig. 1B. The peak area ratio of terminal methoxy group (d¼ 3.38

ppm) to methyl group (d ¼ 0.84 ppm) was 2 : 1, showing that the

hydroxyl groups were completely methylated. The SEC result

indicated the modification maintained the unimodality and

narrow polydispersity of the polymer.

The allyl group is not an active group for bioconjugation, but

it can be modified to carboxyl, amino or hydroxyl group easily by

thiol-ene ‘‘click’’ reaction,14 a hydrothiolation of a double bond

with chemicals containing thiol group, such as 3-mercaptopro-

pionic acid, 2-aminoethanethiol hydrochloride and 2-mercap-

toethanol. To activate the branched allyl-(mPEG)2, the radical

addition of 3-mercaptopropionic acid and 2-aminoethanethiol

hydrochloride to the allyl middle group of the polymer was

performed. Taking COOH-(mPEG)2 (3) as an example, the 1H

NMR spectrum (Fig. 1C) of the polymer revealed the complete

This journal is ª The Royal Society of Chemistry 2011

disappearance of the signals assigned to the allyl protons at 4.02

ppm (d, –O–CH2–CH]CH2), 5.01–5.22 ppm (dd, –CH]CH2),

and 5.82–5.95 ppm (m, –CH]CH2). Concomitantly, new signals

were clearly observed at 1.76 ppm (m, –OCH2CH2CH2S–), 2.56

ppm (m, –CH2–S–CH2–), and 2.96 ppm (m, –CH2COOH), cor-

responding to the resulting structure via the addition of 3-mer-

capopropionic acid to allyl group. Synthesis of NH2-(mPEG)2

was also carried out (Scheme 1). The 1H NMR spectrum of this

polymer also revealed that the allyl groups were completely

transformed into amino groups based on the complete dis-

appearance of the signals assigned to the allyl protons and the

appearance of the signals at 1.76 ppm (m, –OCH2CH2CH2S–),

2.41 ppm (m, –CH2–S–CH2–), 2.66 ppm (m, –CH2–S–CH2–) and

2.94 ppm (m, –CH2CH2NH2) (Spectrum not shown). Thus two

functionalized dual-branched mPEGs (COOH–(mPEG)2 and

NH2–(mPEG)2) have been successfully synthesized. They are

particularly suitable for drug, peptide, or protein pegylation.

Allyl-(PEG-OH)2 is a starting material for preparing hetero-

bifunctional branched PEG derivatives containing two uniform

end groups and another active group at the mid-junction. Five

heterobifunctional dual-branched PEGs have been successfully

synthesized: allyl-(PEG-alkyne)2 (5), allyl-(PEG-OTs)2 (6), allyl-

(PEG-N3)2 (7), COOH-(PEG-N3)2 (8), NH2-(PEG-N3)2 (9).

From the 1H NMR spectrum (Fig. 3) of allyl-(PEG-alkyne)2, new

signals at 2.40 ppm (s, –CCH) and 4.17 ppm (s, –CH2–CCH),

along with the 2 : 1 molar ratio of alkyne group (d¼ 2.40 and 4.17

ppm) to methoxy group (d¼ 0.84 ppm), showed that the hydroxyl

groups were completely changed to alkyne group. Alkyyne group

is amenable for ‘‘click chemistry’’, a highly specific cycloaddition

that can take place in mild aqueous conditions.15

Polym. Chem., 2011, 2, 873–878 | 877

Page 6: A facile synthesis of branched poly(ethylene glycol) and its heterobifunctional derivatives

Fig. 4 13C NMR spectra of allyl-(PEG-N3)2 (A) and COOH-(PEG-N3)2

(B), (CDCl3 at 20 �C).

Dow

nloa

ded

on 1

9 M

arch

201

3Pu

blis

hed

on 0

4 Ja

nuar

y 20

11 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0P

Y00

339E

View Article Online

Two other clickable heterobifunctional branched PEG,

COOH-(PEG-N3)2 (8) and NH2-(PEG-N3)2 (9), were success-

fully synthesized. Aside from 1H NMR (with identical spectra of

(3) and (8), and of (4) and (9)), 13C NMR gave strong evidence

about the synthesis of these derivatives. As Fig. 4 shows, the

corresponding carbon signals at 121.2 ppm (–CH]CH2) and

132.4 ppm (–CH]CH2) completely disappeared and new peaks

at 27.1 ppm (–SCH2CH2COOH), 34.8 ppm (–SCH2CH2COOH),

and 173.4 ppm (–SCH2CH2COOH) were observed, indicating

the conversion of allyl group to carboxyl group. Meanwhile, no

side reaction of the azido group such as radical scavenging was

detected. These results indicated the successful preparation of

heterobifunctional, dual-branched PEG possessing carboxyl

group on the chain middle and azido groups on the two chain

ends. These heterobifunctional dual-branched PEGs may find

applications in combination drug delivery16 or targeting drug

delivery,17 and biosensors18 and surface modification.19 It should

be pointed out that in addition to those derivatives demonstrated

in this report, a large number of heterobifunctional dual-

branched PEGs can be designed and synthesized due to the

versatility of the hydroxyl groups present on allyl-(PEG-OH)22,12

Conclusion

In conclusion, a new and well-defined dual-branched PEG, allyl-

(PEG-OH)2, was successfully synthesized with a commercial

chemical TMPAE as the initiator for the anionic polymerization

of ethylene oxide. The polymer is modifiable for a number of

derivatives. The preparation of branched PEGs with methoxy

terminals, having carboxyl or amino group in the mid-junction,

and heterobifunctional branched PEGs was demonstrated. The

synthesis and purification procedures were short and simple.

These dual-branched mPEG and heterofunctional dual-

branched PEG can find applications in drug and protein delivery

and surface modification of biosensors.

Acknowledgements

The authors gratefully acknowledge financial support from the

Hong Kong Research Grant Council (General Research Fund

600207).

878 | Polym. Chem., 2011, 2, 873–878

Notes and references

1 (a) J. M. Harris, Poly(ethylene glycol) Chemistry: Biotechnical andBiomedical Applications, Plenum Publishing corporation, NewYork, 1992; (b) J. M. Harris and S. Zalipsky, Poly(ethylene glycol)Chemistry and Biological Applications, American Chemical Society,Washington. 1997.

2 (a) J. M. Harris and R. B. Chess, Nat. Rev. Drug Discovery, 2003, 2,214–221; (b) M. J. Roberts, M. D. Bentley and J. M. Harris, Adv.Drug Delivery Rev., 2002, 54, 459–476; (c) F. M. Veronese,Biomaterials, 2001, 22, 405–417; (d) F. M. Veronese and G. Pasut,Drug Discovery Today, 2005, 10, 1451–1458; (e) F. M. Veronese andG. Pasut, Adv. Drug Delivery Rev., 2009, 61, 1177–1188.

3 G. Prencipe, S. M. Tabakman, K. Welsher, Z. Liu, A. P. Goodwin,L. Zhang, J. Henry and H. J. Dai, J. Am. Chem. Soc., 2008, 131,4783–4787.

4 (a) C. J. Fee, Biotechnol. Bioeng., 2007, 98, 725–731; (b)F. M. Veronese, P. Caliceti and O. Schiavon, J. Bioact. Compat.Polym., 1997, 12, 196–207; (c) F. M. Veronese, C. Monfardini,P. Caliceti, O. Schiavon, M. D. Scrawen and D. Beer, J. ControlledRelease, 1996, 40, 199–209; (d) C. Monfardini, O. Schiavon,P. Caliceti, M. Morpurgo, J. M. Harris and F. M. Veronese,Bioconjugate Chem., 1995, 6, 62–69.

5 (a) P. Bailon, A. Palleroni, C. A. Schaffer, C. L. Spence, W. J. Fung,J. E. Porter, G. K. Ehrlich, W. Pan, Z. X. Xu, M. W. Modi, A. Faridand W. Berthold, Bioconjugate Chem., 2001, 12, 195–202; (b)K. R. Reddy, M. W. Modi and S. Pedder, Adv. Drug Delivery Rev.,2002, 54, 571–586; (c) A. Kozlowski, S. A. Charles andJ. M. Harris, BioDrugs, 2001, 15, 419–429; (d) K. R. Reddy,T. L. Wright, P. J. Pockros, M. Shiffman, G. Everson,R. Reindollar, M. W. Fried, P. P. Purdum, D. Jensen, C. Smith,W. M. Lee, T. D. Boyer, A. Lin, S. Pedder and J. DePamphilis,Hepatology, 2001, 33, 433–438; (e) R. J. Motzer, A. Rakhit,J. Thompson, H. Gurney, P. Selby, R. Figlin, S. Negrier, S. Ernst,M. Siebels, M. Ginsberg, K. Rittweger and L. Hooftman, Ann.Oncol., 2002, 13, 1799–1805; (f) S. Zeuzem, S. V. Feinman,J. Rasenack, E. J. Heathcote, M. Y. Lai, E. Gane, J. O’Grady,J. Reichen, M. Diago, A. Lin, J. Hoffman and M. J. Brunda, N.Engl. J. Med., 2000, 343, 1666–1672; (g) E. J. Heathcote,M. L. Shiffman, W. G. E. Cooksley, G. M. Dusheiko, S. S. Lee,L. Balart, R. Reindollar, R. K. Reddy, T. L. Wright, A. Lin,J. Hoffman and J. De Pamphilis, N. Engl. J. Med., 2000, 343, 1673–1680; (h) S. J. Keam and R. S. Cvetkovic, Drugs, 2008, 68, 1273–1319; (i) S. J. Keam and R. S. Cvetkovic, BioDrugs, 2009, 23, 63–68.

6 A. Martinez, A. Pendri, J. Xia and R. B. Greenwald, Macromol.Chem. Phys., 1997, 198, 2489–2498.

7 H. Zhao, K. Yang, A. Martinez, A. Basu, R. Chintala, H. C. Liu,A. Janjua, M. L. Wang and D. Filpula, Bioconjugate Chem., 2006,17, 341–351.

8 (a) NOF Corp., M. Kouzou and K. Yoshiyuki, 1999-08-24; (b) http://www.nof.co.jp/business/dds/apeg/english/con02/con02-2.html-2010-10-14.

9 Z. Y. Li, P. P. Li and J. L. Huang, J. Polym. Sci., Part A: Polym.Chem., 2006, 44, 4361–4371.

10 S. Cammas, Y. Nagasaki and K. Kataoka, Bioconjugate Chem., 1995,6, 226–230.

11 S. Hiki and K. Kataoka, Bioconjugate Chem., 2007, 18, 2191–2196.

12 Z. Y. Li and Y. Chau, Bioconjugate Chem., 2009, 20, 780–789.13 T. H. Mourey, A. J. Hoteling, S. T. Balke and K. G. Owens, J. Appl.

Polym. Sci., 2005, 97, 627–639.14 A. B. Lowe, Polym. Chem., 2010, 1, 17–36.15 J. F. Lutz and Z. Zarafshani, Adv. Drug Delivery Rev., 2008, 60, 958–

970.16 A. Warnecke and F. Kratz, Bioconjugate Chem., 2003, 14, 377–

387.17 G. Pasut, F. Canal, L. D. Via, S. Arpicco, F. A. Veronese and

O. Schiavon, J. Controlled Release, 2008, 127, 239–248.18 R. Schlapak, P. Pammer, D. Armitage, R. Zhu, P. Hinterdorfer,

M. Vaupel, T. Fruhwirth and S. Howorka, Langmuir, 2006, 22,277–285.

19 A. K. Pannier, J. A. Wieland and L. D. Shea, Acta Biomater., 2008, 4,26–39.

This journal is ª The Royal Society of Chemistry 2011