"Click chemistry" in the preparation of porous polymer-based particulate stationary phases for μ-HPLC separation of peptides and proteins

Article abstract of DOI:10.1021/ac060006s

With the use of the copper(I)-catalyzed (3 + 2) azide-alkyne cycloaddition, an element of "click chemistry," stationary phases carrying long alkyl chains or soybean trypsin inhibitor have been prepared for use in HPLC separations in the reversed-phase and affinity modes, respectively. The ligands were attached via a triazole ring to size monodisperse porous beads containing either alkyne or azide pendant functionalities. Alkyne-containing beads prepared by direct copolymerization of propargyl acrylate with ethylene dimethacrylate were allowed to react with azidooctadecane to give a reversed-phase sorbent. Azide-functionalized beads were prepared by chemical modification of glycidyl methacrylate particles. Subsequent reaction with a terminal aliphatic alkyne produced a reversed-phase sorbent similar to that obtained from the alkyne beads. Soybean trypsin inhibitor was functionalized with N-(4-pentynoyloxy) succinimide to carry alkyne groups and then allowed to react with the azide-containing beads to produce an affinity sorbent for trypsin. The performance of these stationary phases was demonstrated with the HPLC separations of a variety of peptides and proteins.

Full text of DOI:10.1021/ac060006s

Anal. Chem. 2006, 78, 4969-4975
“
Click Chemistry” in the Preparation of Porous
Polymer-Based Particulate Stationary Phases for
µ-HPLC Separation of Peptides and Proteins
Michael Slater, Marian Snauko,†,§ Frantisek Svec,†,‡ and Jean M. J Fr e´ chet*,†,‡
†
Department of Chemistry, University of California, Berkeley, California 94720-1460, and The Molecular Foundry,
E. O. Lawrence Berkeley National Laboratory, Berkeley, California 94720-8197
With the use of the copper(I)-catalyzed (3 + 2) azide-
alkyne cycloaddition, an element of “click chemistry,”
stationary phases carrying long alkyl chains or soybean
trypsin inhibitor have been prepared for use in HPLC
separations in the reversed-phase and affinity modes,
respectively. The ligands were attached via a triazole ring
to size monodisperse porous beads containing either
alkyne or azide pendant functionalities. Alkyne-containing
beads prepared by direct copolymerization of propargyl
acrylate with ethylene dimethacrylate were allowed to react
with azidooctadecane to give a reversed-phase sorbent.
Azide-functionalized beads were prepared by chemical
modification of glycidyl methacrylate particles. Subse-
quent reaction with a terminal aliphatic alkyne produced
a reversed-phase sorbent similar to that obtained from the
alkyne beads. Soybean trypsin inhibitor was functionalized
with N-(4-pentynoyloxy)succinimide to carry alkyne groups
and then allowed to react with the azide-containing beads
to produce an affinity sorbent for trypsin. The performance
of these stationary phases was demonstrated with the
HPLC separations of a variety of peptides and proteins.
The two-step approach is particularly attractive as it enables
the formation of numerous stationary phases from a single type
of material with optimized size and porous properties. For
example, glycidyl methacrylate beads have proven to be rather
versatile and have been used to prepare a wide range of stationary
phases for HPLC in a variety of separation modes.1 The beads
are typically modified using standard electrophile-nucleophile
displacement chemistry, or via ring-opening of a pendant reactive
moiety such as an epoxide or an azlactone. Although these
reactions work well, they limit the type of functionality that can
be presented on the surface of the beads as some sensitive
functional groups may not survive the modification reaction
intact. For example, immobilization of a protein ligand carried
out in aqueous solution using beads with azlactone or epoxide
groups to form stationary phases for affinity chromatography is
accompanied with a spontaneous hydrolysis of the reactive
functionalities of the support.
With emerging targets in fields as diverse as environmental
studies, proteomics, metabolomics, and biotechnology, high-
performance liquid chromatography faces new challenges. Most
of these new areas require highly efficient separations of very
complex mixtures for which currently available columns may not
perform adequately. Therefore, the development of new, versatile
separation media is desirable. One approach to achieve this goal
is the development of stationary phases with the desired chemistry
attached in a controlled manner via the selective reaction of
complementary functionalities orthogonal to those that may be
found on the ligand to be attached to the bead. The copper(I)-
-6
There are many types of columns available for HPLC, both
commercial products and those developed in academic laborato-
ries. Most of these use packings based on only two types of porous
beads made either of silica or of synthetic polymers. Polymeric
stationary phases may be prepared with a broad range of
functionalities in a single step by the copolymerization of functional
monomers. Although simple, this approach requires reoptimiza-
tion of the entire preparation process to control the ultimate
properties of the beads each time a new monomer is used. Silica-
based stationary phases are typically obtained in a two-step
process: first, the silica beads themselves are prepared, and then
their surface is modified in a separate functionalization step. A
similar two-step approach is also used with polymers; a classical
example of this process is the preparation of ion-exchange resins
by chemical modification of poly(styrene-co-divinylbenzene) beads.
catalyzed (3 + 2) azide-alkyne cycloaddition, an element of the
7
“
click chemistry” popularized by Sharpless’ group, is a very
efficient coupling reaction that provides an ideal reactivity profile
(
(
(
(
(
1) Svec, F.; Hradil, J.; Coupek, J.; Kalal, J. Angew. Makromol. Chem. 1975,
48, 135-143.
2) Turkova, J.; Blaha, K.; Malanikova, M.; Vancurova, J.; Svec, F.; Kalal, J.
Biochim. Biophys. Acta 1978, 524, 162.
3) Ageev, A. N.; Aratskova, A. A.; Belyakova, L. D.; Gvozdich, T. N.; Kiselev,
A. V.; Yashin, Y. I.; Kalal, J.; Svec, F. Chromatographia 1983, 17, 545-548.
4) Tennikova, T. B.; Horak, D.; Svec, F.; Kolar, J.; Coupek, J.; Trushin, S. V.;
Maltsev, A. G.; Belenkii, B. G. J. Chromatogr. 1988, 435, 357-362.
5) Tennikova, T. B.; Horak, D.; Svec, F.; Tennikov, M. B.; Kever, E. E.; Belenkii,
B. G. J. Chromatogr. 1989, 457, 187-194.
*
Corresponding author. Phone: 510 643 3077. E-mail: frechet@
berkeley.edu.
(6) Mal’tsev, V. G.; Nasledov, D. G.; Trushin, S. A.; Tennikova, T. B.;
Vinogradova, L. V.; Volokitina, I. N.; Zgonnik, V. N.; Belenkii, B. G. J. High
Resolut. Chromatogr. 1990, 13, 185-192.
(7) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed. 2001, 40,
2004-2021.
†
University of California.
Lawrence Berkeley National Laboratory.
On leave from Polymer Institute, Slovak Academy of Sciences, Dubravska
‡
§
cesta 9, 842 36 Bratislava, Slovakia.
1
0.1021/ac060006s CCC: $33.50 © 2006 American Chemical Society
Analytical Chemistry, Vol. 78, No. 14, July 15, 2006 4969
Published on Web 06/02/2006

for this purpose. Equation 1 shows a general scheme of the
Cu(I)-catalyzed reaction of a terminal alkyne with a terminal azide
affording a 1,4-disubstituted triazole ring:
containing ligand was then used for the selective separation of a
specific antibody.
The objective of this communication is to demonstrate for the
first time the suitability of the versatile copper-catalyzed azide-
alkyne (3 + 2) cycloaddition for the preparation of efficient
stationary phases for the separation of peptides and proteins in
HPLC mode.
EXPERIMENTAL SECTION
This reaction has already been widely used for the synthesis
Materials and Methods. HPLC grade water, acetonitrile
(ACN), dichloromethane (DCM), dimethyl sulfoxide (DMSO), and
tetrahydrofuran (THF) were purchased from Fisher Scientific
(Pittsburgh, PA). All proteins and peptides including ribonuclease
A, lysozyme, cytochrome c, myoglobin, albumin, trypsin, Phe-Gly-
Phe-Gly, bradykinin, angiotensin II, angiotensin I, insulin, and
Kunitz-type soybean trypsin inhibitor (SBTI) were obtained from
Sigma (St. Louis, MO). Octadecyne was purchased from GFS
Chemicals. All other reagents were purchased from Aldrich.
Monomers, glycidyl methacrylate, propargyl acrylate, and ethylene
dimethacrylate, were distilled under vacuum prior to use; other
reagents were used as provided. Polystyrene seed particles were
obtained from Bangs Laboratories (Fishers, IN). Polymer beads
were physically characterized by isothermal nitrogen adsorption/
desorption using an ASAP 2010 sorptometer from Micromeritics
(Norcross, GA); surface area was calculated using the BET model,
while mesopore information was derived using the BJH model.
Reactions were monitored for completion by TLC on fluorescent
silica plates, visualizing with UV and/or phosphomolybdic acid
8
-12
of substituted triazoles in solution.
Because azide and alkyne
functionalities do not interfere with common biological processes,
the copper-catalyzed azide-alkyne cycloaddition has been used
1
3-17
for profiling enzyme activity
and to incorporate new function-
alities in proteins, cells, and viruses.20 This “click” reaction was
18
19
also applied for the preparation and functionalization of den-
2
1,22
23
24
drimers
as well as for the synthesis and modification of
linear polymers. In addition, the utility of this reaction has also
been demonstrated in chemical modifications performed on solid
polymer supports. For example, it has been integrated with solid-
2
5,26
phase peptide synthesis to produce peptidotriazoles
the installation, via a “click linker”, of sensitive functionalities on
Merrifield resin for solid-phase organic synthesis.
and for
27,28
Finally, Puna
et al. have recently prepared an azide derivative of diaminodipro-
pylamine agarose beads and used this click reaction to immobilize
29
alkyne-containing versions of biotin and a hexapeptide. They also
used a complementary alkyne derivative of agarose to attach
4
-azido-N-(4-(4-oxobutoxy)phenyl)butyramide; this aldehyde-
4
stain. Organic extracts were dried over MgSO and solvents were
(
8) Demko, Z. P.; Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 41, 2113-
116.
9) Demko, Z. P.; Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 41, 2110-
113.
10) Lewis, W. G.; Green, L. G.; Grynszpan, F.; Radic, Z.; Carlier, P. R.; Taylor,
P.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 41, 1053-
057.
11) Rodionov, V. O.; Fokin, V. V.; Finn, M. G. Angew. Chem 2005, 44, 2210-
215.
12) Wang, Q.; Chittaboina, S.; Barnhill, H. N. Lett. Org. Chem. 2005, 2, 293-
01.
13) Speers, A. E.; Adam, G. C.; Cravatt, B. F. J. Am. Chem. Soc. 2003, 125,
686-4687.
removed in vacuo using a rotary evaporator. Elemental analyses
were performed at the UC Berkeley Microanalysis Facility. IR
spectra were recorded on a Mattson Genesis II FT-IR as thin films
between NaCl disks or as dispersions in KBr. NMR spectra were
recorded on Bruker AV-400 or DRX-500 instruments as solutions
in chloroform-d. NMR chemical shifts are reported as δ in ppm
relative to TMS (δ 0.00), and coupling constants are given as J
2
(
2
(
1
(
(
(
2
13
3
values in Hz; DEPT experiments were used to assign the C NMR
resonances as CH , CH , CH, or C. Teflon-coated 251 µm i.d.
3
2
4
capillaries were purchased from Polymicro Technologies (Phoe-
nix, AZ).
(
(
(
(
14) Speers, A. E.; Cravatt, B. F. Chem. Biol. 2004, 11, 535-546.
15) Speers, A. E.; Cravatt, B. F. ChemBioChem 2004, 5, 41-47.
16) Speers, A. E.; Cravatt, B. F. J. Am. Chem. Soc. 2005, 127, 10018-10019.
17) Evans, M. J.; Saghatelian, A.; Sorensen, E. J.; Cravatt, B. F. Nat. Biotechnol.
005, 23, 1303-1307.
18) Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. J. Am. Chem. Soc. 2004, 126,
5046-15047.
19) Link, A. J.; Tirrell, D. A. J. Am. Chem. Soc. 2003, 125, 11164-11165.
20) Wang, Q.; Chan, T. R.; Hilgraf, R.; Fokin, V. V.; Sharpless, K. B.; Finn, M.
G. J. Am. Chem. Soc. 2003, 125, 3192-3193.
21) Wu, P.; Feldman, A. K.; Nugent, A. K.; Hawker, C. J.; Scheel, A.; Voit, B.;
Pyun, J.; Fr e´ chet, J. M. J.; Sharpless, K. B.; Fokin, V. V. Angew. Chem., Int.
Ed. 2004, 43, 3928-3932.
Care in Handling Azides. Azides must be handled with care,
and all work must be performed behind a safety shield. Avoid
high temperatures and reactions that would lead to products
containing a high weight percent of the azide moiety in the final
isolated product.
2
(
1
(
(
1-Azidooctadecane (1). This was prepared by a modification
(
30
of a procedure described by Alvarez and Alvarez. Sodium azide
(
7.15 g, 110 mmol) and tetrabutylammonium iodide (20 mg, 54
µmol) were dissolved in DMSO (220 mL); 1-bromooctadecane
33.3 g, 100 mmol) was added, and the biphasic mixture was
(
(
(
(
(
22) Helms, B.; Mynar, J. L.; Hawker, C. J.; Fr e´ chet, J. M. J. J. Am. Chem. Soc.
2004, 126, 15020-15021.
(
23) Gao, H.; Louche, G.; Sumerlin, B. S.; Jahed, N.; Golas, P.; Matyjaszewski,
heated at 80 °C for 16 h. Upon cooling, the mixture was poured
into water (200 mL) and then extracted with hexanes (3 × 200
mL); the combined organic extracts were washed with water (3
K. Macromolecules 2005, 38, 8979-8982.
24) Malkoch, M.; Thibault, R. J.; Drockenmuller, E.; Messerschmidt, M.; Voit,
B.; Russell, T. P.; Hawker, C. J. J. Am. Chem. Soc. 2005, 127, 14942-14949.
25) Tornoe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67, 3057-
064.
26) Punna, S.; Kuzelka, J.; Wang, Q.; Finn, M. G. Angew. Chem., Int. Ed. 2005,
4, 2215-2220.
27) L o¨ ber, S.; Gmeiner, P. Tetrahedron 2004, 60, 8699-8702.
28) L o¨ ber, S.; Rodriguez-Loaiza, P.; Gmeiner, P. Org. Lett. 2003, 5, 1753-1755.
29) Punna, S.; Kaltgrad, E.; Finn, M. G. Bioconjugate Chem. 2005, 16, 1536-
×
200 mL) and brine (300 mL) and worked up as usual to afford
3
a straw-colored oil. This was purified by flash chromatography
on 150 g of silica gel, eluting with petroleum ether; the product 1
was obtained as a clear, colorless oil (28.66 g, 97.0 mmol, 97%
4
(
(
(
1
541.
(30) Alvarez, S. G.; Alvarez, M. T. Synthesis 1997, 413-414.
4970 Analytical Chemistry, Vol. 78, No. 14, July 15, 2006

1
yield). H NMR (400 MHz): δ 3.24 (t, 2H), 1.59 (quin, 2H), 1.26
and MeCN (5 mL) were combined and sonicated briefly to
produce an emulsion. PA-EDMA beads were added, and the
mixture was rotated using a rotary evaporator motor at ambient
temperature for 18 h; phase separation occurred during this time.
The beads were isolated by filtration on a fine glass frit and washed
with DCM (20 mL), MeCN (40 mL), 10 mmol/L HCl in 75%
MeOH (20 mL), 75% MeOH (20 mL), 5 mmol/L NaOH in 75%
MeOH (20 mL), 75% MeOH (20 mL), MeOH (40 mL), THF (40
1
3
(
br m, 30 H), 0.88 (t, 3H). C NMR (125 MHz): δ 14.0 (CH
2.7 (CH ), 26.7 (CH ), 28.8 (CH ), 29.2 (CH ), 29.4 (CH ), 29.50
), 29.56 (CH ), 29.64 (CH ), 29.67 (CH ), 29.69 (CH ), 29.706
), 29.714 (CH ), 29.72 (CH ), 31.9 (CH ), 51.4 (CH ); the
peaks were not sufficiently dispersed to identify each C, even when
resolution-enhancing processing was used. Elemental Anal. Calcd
3
),
2
2
2
2
2
2
(
(
CH
CH
2
2
2
2
2
2
2
2
2
2
for C18
H
37
N
3
: C, 73.2%; H, 12.6%; N, 14.2%. Found: C, 73.6%; H,
2.8%; N, 14.4%.
N-(4-Pentynoyloxy)succinimide (2). 4-Pentynoic acid (1.00
1
mL), and Et O (40 mL). The beads were air-dried and then placed
2
in a vacuum oven at 40 °C for 14 h. An amount of 237 mg of beads
g, 10.2 mmol), N-hydroxysuccinimide (2.35 g, 20.4 mmol), N-ethyl-
N′-(3-dimethylaminopropyl) carbodiimide hydrochloride (2.15 g,
5
were obtained as a white powder. Elemental analysis of nitro-
gen indicated that the beads contained 0.54 mmol/g triazole
units: C, 63.0%; H, 7.75%; N, 2.29%. νmax/cm : 3276, 2977, 2924,
1
1.2 mmol), and 4-(dimethylamino)pyridine (415 mg, 3.40 mmol)
-1
were combined under an Ar atmosphere, and dry DCM (20 mL)
was introduced. After stirring for 3 days, the mixture was poured
into water containing 5% v/v acetic acid. The layers were
separated, and the aqueous phase was extracted with DCM (2 ×
1
714(br), 1521, 1475, 1429, 1389; the intensity of the band due to
the alkyne C-H stretch (3276) is smaller relative to that of other
bands as compared to the spectrum of the starting material.
Azide-Modified GMA-EDMA Beads (6). GMA-EDMA
beads 3 (2.00 g, ca. 5 mmol epoxide groups based on monomer
ratios) and sodium azide (1.95 g, 30 mmol) were added to a flask,
and water (7 mL) and DMF (14 mL) were added. The flask was
rotated in an oil bath at 50 °C for 20 h using a rotary evaporator
motor. The beads 6 with 3-azido-2-hydroxypropyl functionalities
were isolated by filtration on a fine glass frit and washed with
DMF (50 mL), water (100 mL), methanol (60 mL), THF (60 mL),
2
5 mL). The combined organic extracts were washed with brine
(60 mL) and worked up as usual giving a clear, colorless oil. This
was purified using chromatography on 21 g of silica gel, eluting
with hexanes with an ethyl acetate gradient (0-100%); the product
was obtained as a white solid (0.99 g, 5.1 mmol, 50% yield). Mp:
1
7
2
0-72 °C. H NMR (400 MHz): δ 2.08 (t, 1H, J ) 2.5), 2.61 (td,
H, J ) 2.5, J ) 6), 2.84 (s, 4H), 2.89 (t, 2H, J ) 6). 1 C NMR (100
3
MHz): δ 13.4 (CH
66.8 (C), 169.2 (C). νmax/cm : 2255, 1816, 1790, 1743, 1431, 1373.
HRMS (EI): m/z found [MH+] 196.0603, C requires
96.0610; found [M+] 195.0532, C NO requires 195.0532.
Elemental Anal. Calcd for C NO : C, 55.4%; H, 4.65%; N, 7.18%.
2 2 2
), 25.1 (CH ), 29.6 (CH ), 69.8 (C), 80.8 (CH),
2
and Et O (30 mL) before drying in a vacuum oven at 40 °C for 12
h. Incorporation of azide groups was determined by elemental
analysis for nitrogen: C, 52.4%; H, 7.73%; N, 6.65%, corresponding
-1
1
9
H10NO
4
1
9
H
9
4
-1
to 1.6 mmol/g azide groups. νmax (cm ): 3324 (v br), 2947, 2892,
103, 1718 (br), 1475, 1456, 1388.
18 RP Sorbent from Azide-Containing Beads (7). CuI (7.6
mg, 40 µmol) and 1-octadecyne (250 mg) were mixed in MeCN
5 mL) and sonicated briefly to aid in mixing; the emulsion took
H
9 9
4
2
Found: C, 55.1%; H, 4.69%; N, 7.12%.
C
Monodisperse Porous Polymer Beads. Poly(glycidyl meth-
acrylate-co-ethylene dimethacrylate) (GMA-EDMA) beads 3 were
prepared via a templated suspension polymerization described
(
on a pale yellow hue. Azide-modified beads 6 (250 mg, ∼0.4 mmol
azide) were added, and the mixture was rotated at ambient
temperature for 2 days. The beads 7 were isolated by filtration
on a fine glass frit and washed with CHCl
mL), 3 mmol/L disodium EDTA in 10% MeOH (30 mL), H
mL), MeOH (30 mL), THF (45 mL), and Et O (30 mL), then dried
3
1
previously. Poly(propargyl acrylate-co-ethylene dimethacrylate)
PA-EDMA) beads 4 were prepared in an analogous manner as
(
follows: polystyrene seed particles with a diameter of 560 nm (0.05
mL total bead volume) were first activated by swelling in an
emulsion prepared by sonicating dibutyl phthalate (0.33 mL) in
3
(30 mL), MeOH (30
2
O (30
2
1
5 mL of an aqueous solution of SDS (37.5 mg, 0.25% w/v). After
in a vacuum oven at 40 °C for 13 h. Elemental analysis: C, 53.8%;
H, 7.62%; N, 6.46%.
primary swelling leading to particles with a diameter of 1 µm was
complete, they were added to an emulsion consisting of the
porogenic solvents cyclohexanol (18.8 mL) and 1-dodecanol (2.1
mL), monomers EDMA (8.4 mL) and propargyl acrylate (5.6 mL),
and initiator AIBN (290 mg) in 80 mL of an aqueous solution of
SDS (200 mg, 0.25% w/v). The secondary swelling was allowed
to proceed until the calculated bead diameter of 5 µm was
Alkyne-Modified Soybean Trypsin Inhibitor (8). SBTI (478
mg) was added to a 50 mL polypropylene vial followed by 25 mL
of 10 mmol/L sodium phosphate buffer (pH 7.0); the pH of the
resulting solution was 5.6, and some material remained undis-
solved. NaOH (0.5 mol/L) was added until all of the protein
dissolved, and then the pH was adjusted to 7 by dropwise addition
of 0.5 mol/L HCl. N-(4-Pentynoyloxy)succinimide 9 was dissolved
with mild heating in 5.0 mL ethylene glycol, and the solution was
added at once to the protein solution. The resulting suspension
was rotated for 15 h at ambient temperature; all of the solids
dissolved after about 10 h. The pH of the solution fell to 5.3 during
the course of the reaction, and some protein precipitated near
the end of the reaction time; prior to workup, the pH was raised
to 7 using 0.5 mol/L NaOH. The crude protein solution was
obtained. Sufficient 4% w/v poly(vinyl alcohol) (PVA, M
10 000, 100% hydrolyzed) solution was added to the mixture
to afford a final concentration of 1% w/v PVA in the polymeriza-
tion mixture, and NaNO (16 mg) was added to suppress
polymerization in the aqueous phase. The mixture was sparged
with N for 20 min and then polymerized in an orbiting shaker
n
106 000-
1
2
2
bath at 70 °C for 16 h. The beads were washed by repeated
decantation with water, methanol, and THF and then isolated by
filtration.
purified by gel filtration on 50 g of Sephadex G-25 in H
resulting solution was analyzed by MALDI-TOF MS using R-cyano-
-hydroxycinnamic acid (CHCA) as matrix: [M+] m/z 20 500.
2
O. The
C18 RP Stationary Phase from PA-EDMA Beads (5).
1
-Azidooctadecane (1.06 g, 3.6 mmol), CuI (23 mg, 120 µmol),
4
(
31) Smigol, V.; Svec, F. J. Appl. Polym. Sci. 1992, 45, 1439-1448.
The observed mass increase relative to that of unmodified SBTI
Analytical Chemistry, Vol. 78, No. 14, July 15, 2006 4971

(
[M+] m/z 20 000) indicates acylation of ca. 6 of 10 possible amine
Table 1. Properties of Azide- and Alkyne-Containing
Porous Polymer Beads
sites on the protein.
Attachment of Alkyne-Modified SBTI to Azide-Containing
Beads (10). CuI (190 µg, 1.0 µmol) was added as 17.2 µL of a
specific surface
pore volume
mL/g
pore size
nm
2
beads
area, m /g
1
2
1.1 mg/mL suspension in H O to tris-(1-benzyl-[1,2,3]-triazol-4-
4
6
243
76
0.83
0.35
10
16
32
ylmethyl)amine (1.1 mg, 2 µmol) in 0.5 mL MeOH. The mixture
was sonicated until homogeneous, and then the methanol was
evaporated under a stream of argon. The resulting white solid
was dissolved in 10 mL of 70 µmol/L alkyne-modified SBTI (as
obtained from the gel filtration step above), and azide beads (50
mg, 80 µmol azide) were added. The mixture was slowly agitated
for 16 h at ambient temperature. The beads 10 were isolated by
centrifugation (60 rpm, 5 min) and then washed by resuspending
the solids in water (15 mL), agitating for a few minutes, and
centrifuging; this wash step was repeated seven times. The final
supernatant was decanted, and the wet beads were packed in a
capillary as described below. The absorbance of the supernatant
at 280 nm was measured and compared to the value before the
reaction; on the basis of this analysis the protein loading of the
beads was estimated to be ca. 2 µmol/g. Elemental analysis: C,
volumes of 1% trifluoroacetic acid (TFA) at a temperature of 80
°C and flow rate 1 µL/min. At selected times, the temperature in
the thermostat was decreased to 35 °C and the column cleaned
with a blank gradient of 0-100% ACN in water in 15 min before
the test substances were injected. Separations of acetone and
benzene were carried out using 50% aqueous ACN. The separa-
tions of proteins were achieved using a linear gradient of from
5% to 100% ACN in 0.1% aqueous TFA in 15 min. All these
separations were performed at a flow rate of 6 µL/min.
RESULTS AND DISCUSSION
5
2.2%; H, 7.49%; N, 7.61%; S, 0.48%.
Chromatography. Stationary phases for reversed-phase chro-
Preparation of Stationary Phases. The triazole forming
“click” reaction we used to modify the stationary phases is derived
matography were packed into the capillary from 6 mg/mL slurry
in an isodensity solvent mixture consisting of a 25:75 (v/v) mixture
of THF and DCM at a constant pressure of 13.8 MPa. The packed
capillary was attached to an HPLC system consisting of an Ultra
Plus II pump and pump controller (Micro-Tech Scientific, Vista,
CA) and pressurized to 27.6 MPa using 50:50 (v/v) ACN-water
mixture until no further compression of sorbent within the
capillary was observed. Typically, a compression amounting to
from the original 1,3-dipolar cycloaddition of organic azides to
carbon-carbon multiple bonds discovered by Huisgen et al.33 and
7
later enhanced and popularized by Sharpless. Since both azide
and alkyne functional groups are relatively stable to many
synthetic procedures, either group can be incorporated on the
stationary phase as desired. Thus, the solid support can include
alkyne functionality to later react with a ligand bearing an azide
group or vice versa. To demonstrate this versatility, we prepared
both types of porous polymer beads: azide beads and alkyne
3
-5 mm was observed for a 400 mm long packed segment. The
packed capillary was cut into 150 mm columns that were then
fitted with inlet and outlet fittings. A similar procedure was used
for packing columns with sorbent containing immobilized soybean
trypsin inhibitor except that water was used instead of organic
solvents. The column length was 100 mm.
beads. First, using a procedure developed earlier in our labora-
tory,34 we prepared uniformly sized particles 3 containing reactive
epoxide functionalities by the staged, templated polymerization
of glycidyl methacrylate. A simple, single-step reaction of these
beads with sodium azide in water-DMF afforded beads 6 with
the desired azide groups present in 2-hydroxy-3-azidopropyl
moieties. Using an analogous approach, we also prepared beads
4 featuring pendant alkyne groups by copolymerization of prop-
argyl acrylate and ethylene dimethacrylate. The properties of these
beads are summarized in Table 1.
Both types of beads were then reacted with the complementary
reagent, 1-octadecyne and 1-azidooctadecane, respectively, to
afford the two analogous hydrophobic stationary phases suitable
for reversed-phase HPLC. In both cases the alkyl chains are
attached to the polymer support via a triazole ring (see Schemes
1 and 2).
Hydrophobicity and Methylene Selectivity. The effect of
surface chemistry on separation properties is easily followed by
determination of methylene selectivity. This value is a measure
of the ability of the stationary phase to separate alkyl-substituted
homologues of benzene. Plots of retention factors for columns
packed with the original beads and their alkylated counterparts
as recorded for a homologous series of alkylbenzenes, including
benzene and its methyl, ethyl, propyl, butyl, and pentyl derivatives,
are shown in Figure 1. Clearly, the original propargyl acrylate-
The HPLC system consisted of the Ultra Plus II pump and
pump controller and a UVIS 205 detector (Linear, Thermo
Electron, Waltham, MA). The column temperature was held
constant at 35 °C using a Polaratherm series 9000 oven (Selerity
Technologies Inc., Salt Lake City, UT), unless stated otherwise.
The data were processed by Chrom Perfect Spirit (Justice
Laboratory, Denville, NJ). Baseline was subtracted in chromato-
grams of proteins and peptides to eliminate the effects of refractive
index variations resulting from the use of the gradient. Typically,
the injected sample volume was 0.6 µL. Mixtures of proteins and
peptides were prepared by mixing 1 mg/mL stock solutions of
individual components. Retention factors k for alkyl benzenes were
calculated using equation
k ) (V - V )/V
o
a
o
where V
o a
and V are the retention volumes of acetone and analyte,
respectively.
Column Stability. The capillary column packed with sorbent
5
was flushed for almost 400 h representing about 3300 column
(
32) Chan, T. R.; Hilgraf, R.; Sharpless, K. B.; Fokin, V. V. Org. Lett. 2004, 6,
853-2855.
(33) Huisgen, R.; Szeimies, G.; Moebius, L. Chem. Ber. 1967, 100, 2494-2507.
(34) Smigol, V.; Svec, F. J. Appl. Polym. Sci. 1992, 46, 1439-1448.
2
4972 Analytical Chemistry, Vol. 78, No. 14, July 15, 2006

Scheme 1
Scheme 2
Figure 1. Effect of the number of carbon atoms in the alkyl group
of substituted benzene homologues on their retention in columns
packed with 2-hydroxy-3-azidopropyl methacrylate (a), alkylated
2
-hydroxy-3-azidopropyl methacrylate (b), propargyl acrylate (c), and
alkylated propargyl acrylate beads (d). Conditions: column 150 mm
0.251 mm i.d.; mobile phase 50:50 (v/v) acetonitrile-water; flow
×
rate 6 µL/min; column temperature 35 °C.
based stationary phase itself exhibits significant retentivity, and
its alkylation leads to only a small increase in methylene selectivity
as demonstrated by the increase in slope of the solid straight lines
from 0.12 to 0.18. In contrast, the stationary phase with 2-hydroxy-
3
-azidopropyl functionalities is rather hydrophilic. Therefore, it
shows very little retention toward alkylbenzenes and exhibits a
marginal selectivity as indicated by the slope value of only 0.03.
This value increases considerably after modification by click
chemistry to a slope of 0.17, which is almost identical to that of
the alkylated propargyl acrylate-based stationary phase. This
increase clearly confirms that the chemical modification using
click chemistry affords a retentive hydrophobic surface. On the
basis of the results of the elemental analysis of nitrogen, as shown
in the Experimental Section, the surface coverage with C16 alkyl
Figure 2. Gradient elution of peptides using columns packed with
the original propargyl acrylate beads 4 (a) or the alkylated propargyl
acrylate beads 5 (b). Conditions: column 150 mm × 0.251 mm i.d.;
mobile phase A, 0.1% aqueous TFA; mobile phase B, acetonitrile;
gradient 0-65% B in A in 15 min; flow rate 6 µL/min; column
temperature 35 °C. Peaks: Phe-Gly-Phe-Gly (1), bradykinin (2),
angiotensin II (3), angiotensin I (4), insulin (5).
2
chains is 2.2 µmol/m . This value is on par with typical monomeric
stationary phases for reversed-phase HPLC that contain 2-3.5
2
35
µmol/m . It is worth noting that in contrast to silica-based phases,
the alkyl chains of the polymer backbone also contribute to the
overall hydrophobicity of the packing which may then appear
higher than expected if only the C16 chains are taken into account.
Typical C18 silica columns exhibit methylene selectivity in the
range from 1.3 to 1.6 calculated as the ratio of retention factors
of 5 leads to somewhat longer retention times for both peptides
and proteins. In addition, the performance of this column is better
in terms of peak shape and resolution.
36-38
for toluene and benzene.
The values for our stationary phases
are 1.44 for 4, 1.58 for 5, 0.96 for 6, and 1.51 for 7. These high
selectivities show that the high hydrophobicities of alkylated
packings 5 and 7 exceed those of many ODS columns.
The situation is different for the other pair of stationary phases
derived from glycidyl methacrylate beads. As shown in Figure 4,
sorbent 6 based on 2-hydroxy-3-azidopropyl methacrylate beads
is hydrophilic due to the presence of hydroxyl groups, and thus
it does not retain small peptides. The only retained peak is insulin.
Similarly, as shown in Figure 5, the separation of proteins using
this stationary phase is also poor. As expected, after alkylation of
the beads the retention of peptides is significantly enhanced, and
very good separation of angiotensin I and insulin can easily be
achieved (Figure 4b). The unusual elution pattern with broad
tailing peaks of bradykinin and angiotensin II results from the
large volume of the gradient mixer and capillary connecting it
with the injector. As a result, the gradient of the mobile phase is
delayed by about 14 min. Therefore, these two slightly retained
peptides elute in the isocratic mode as poor peaks. Since the Phe-
Gly-Phe-Gly peptide is not retained and elutes close to the column
dead volume, its peak is sharp. However, this stationary phase
Separation of Proteins and Peptides. Both alkylated station-
ary phases 5 and 7 were used for the reversed-phase separation
of peptides and proteins in the gradient elution mode, and their
separation abilities were compared to those of the nonalkylated
beads 4 and 6. As expected from the methylene selectivity data
indicating similar properties for 4 and 5, separations using
columns packed with these stationary phases are similar, as shown
in Figures 2 and 3. However, the slightly higher hydrophobicity
(
35) Poole, C. F. The Essence of Chromatography; Elsevier: Amsterdam, 2003; p
81.
36) Rimmer, C. A.; Sander, L. C.; Wise, S. A.; Dorsey, J. G. J. Chromatogr., A
003, 1007, 11-20.
2
(
2
(
37) Sander, L. C.; Wise, S. A. J. Sep. Sci. 2003, 26, 283-294.
(
38) Rimmer, C. A.; Sander, L. C.; Wise, S. A. Anal. Bioanal. Chem. 2005, 382,
6
98-707.
Analytical Chemistry, Vol. 78, No. 14, July 15, 2006 4973

Figure 3. Gradient elution of proteins using columns packed with
the original propargyl acrylate beads 4 (a) or with alkylated propargyl
acrylate beads 5 (b). Conditions: column 150 mm × 0.251 mm i.d.;
mobile phase A, 0.1% aqueous TFA; mobile phase B, acetonitrile;
gradient 5-100% B in A in 15 min; flow rate 6 µL/min; column
temperature 35 °C. Peaks: ribonuclease A (1), cytochrome c (2),
lysozyme (3), myoglobin (4).
Figure 5. Gradient elution of proteins using columns packed with
2-hydroxy-3-azidopropyl methacrylate beads 6 (a) or with alkylated
2-hydroxy-3-azidopropyl methacrylate beads 7 (b). Conditions: col-
umn 150 mm × 0.251 mm i.d.; mobile phase A, 0.1% aqueous TFA;
mobile phase B, acetonitrile; gradient 5-100% B in A in 15 min; flow
rate 6 µL/min; column temperature 35 °C. Peaks: ribonuclease A
(1), cytochrome c (2), lysozyme (3), myoglobin (4), albumin (5).
Scheme 3
Figure 4. Gradient elution of peptides using columns packed with
2-hydroxy-3-azidopropyl methacrylate beads 6 (a) or with alkylated
2-hydroxy-3-azidopropyl methacrylate beads 7 (b). Conditions: col-
umn 150 mm × 0.251 mm i.d.; mobile phase A, 0.1% aqueous TFA;
mobile phase B, acetonitrile; gradient 0-65% B in A in 15 min; flow
rate 6 µL/min; column temperature 35 °C. Peaks: Phe-Gly-Phe-Gly
This behavior confirms the expected formation of a reversible
complex between trypsin and SBTI.
(1), bradykinin (2), angiotensin II (3), angiotensin I (4), insulin (5).
Column Stability. We also assessed the stability of the
stationary phases containing long alkyl chains attached via the
triazole ring under extremely harsh conditions. The capillary
column packed with sorbent 5 was flushed with 1% trifluoroacetic
acid (TFA) for over 380 h, representing about 3300 column
volumes, at a high temperature of 80 °C. Despite these treatment
conditions, repeated measurements of the retention of both
enables an excellent baseline separation of all five proteins as
shown in Figure 5b since all of them are sufficiently retained.
Protein Immobilization and Affinity Chromatography. As
discussed above, the chromatographic experiments with 2-hy-
droxy-3-azidopropyl methacrylate beads clearly demonstrated their
hydrophilicity. This feature is particularly advantageous in the
design of separation systems for affinity chromatography where
nonspecific hydrophobic interactions should be minimized. Ac-
cordingly, we prepared a stationary phase for the affinity-based
separation of trypsin through the immobilization of soybean
trypsin inhibitor (SBTI) onto the hydrophilic beads. As shown in
Scheme 3, the SBTI was modified using a mild acylating agents
the N-hydroxysuccinimide ester of 4-pentynoic acid, 2sto pro-
vide the necessary complementary alkyne functionality. This
modified protein was then attached to the hydrophilic azide beads
using a ligand-accelerated version of the copper mediated coupling
reaction.
Figure 6. Affinity chromatography of trypsin using a column packed
with 2-hydroxy-3-azidopropyl methacrylate beads containing im-
mobilized soybean trypsin inhibitor 10. Conditions: injection of 1.0
µg of trypsin in an aqueous solution of pH 7; desorption at pH 2 (0.1%
aqueous TFA); flow rate 6 µL/min; column temperature 23 °C.
Figure 6 shows that a column packed with beads 10 containing
this affinity ligand binds a significant amount of trypsin at neutral
pH. Trypsin release is achieved only after the pH is lowered to 2.
4974 Analytical Chemistry, Vol. 78, No. 14, July 15, 2006

CONCLUSION
In this short communication we have presented the application
of “click chemistry” in HPLC and demonstrated that this approach
can afford stationary phases with well-defined chromatographic
properties. Two approaches were used in this work; the reaction
of alkyne-carrying beads with azide-containing small molecules
gave similar end results as the inverse arrangement of functional
groups, as good surface coverage is obtained in each case.
However, the azide-functionalized support proved to be more
versatile as its hydrophilic nature can be exploited in systems
where the undesirable nonspecific interactions of the supporting
matrix with biological polymers such as proteins should be
minimized. A closer look at the attachment chemistry shown in
Schemes 1 and 2 reveals its similarity with columns featuring polar
groups such as amide, carbamate, urea, sulfonamide, and others
Figure 7. Effect of treatment of the column using 1.0% aqueous
trifluoroacetic acid at 80 °C on the retention of acetone and ben-
zene. Conditions: column 150 mm × 0.251 mm i.d.; mobile phase
embedded between the hydrophobic alkyl chain and the support
5
0:50 (v/v) acetonitrile-water; flow rate 6 µL/min; column temperature
5 °C.
surface.3
9-41
These stationary phases are gaining popularity since
3
they enable excellent separations of basic compounds, change
selectivity, and increase wettability, a particularly valuable property
in separations requiring mobile phases with a large proportion of
water. Although this aspect is not targeted in this report, it will
certainly be interesting to use this functionalization approach to
prepare stationary phases for small molecules using silica-based
supports and study this effect.
While this report has only shown the power of “click chem-
istry” to prepare stationary phases for the separation of peptides
and proteins in both reversed-phase and affinity chromatography
modes, many other types of functionalization are possible. For
example, click chemistry using functional azides or alkynes could
be used to prepare packings for other chromatographic modes
such as normal phase, ion-exchange, or hydrophobic interaction.
Because of its gentle nature, click chemistry seems especially well
suited for the immobilization of proteins, a feature that can be
exploited for the immobilization of enzymes and the creation of
enzymatic reactors. We see a major opportunity for using these
reactions in conjunction with the photografting techniques that
Figure 8. Separation of standard proteins using a column packed
with beads 7 immediately after its preparation and after 381 h of
treatment with 1.0% aqueous trifluoroacetic acid at 80 °C. Condi-
tions: column 150 mm × 0.251 mm i.d.; mobile phase A, 0.1%
aqueous TFA; mobile phase B, acetonitrile; gradient 5-100% B in A
in 15 min; flow rate 6 µL/min; column temperature 35 °C. Peaks:
ribonuclease A (1), cytochrome c (2), lysozyme (3), myoglobin (4),
albumin (5).
42
we have recently developed in the field of microfluidics. The
orthogonal character of the click reactions with respect to many
other functionalization techniques enables the attachment of a
variety of ligands without compromising the native functionalities
of the biopolymer. Experiments in this direction are currently in
progress.
acetone and benzene shown in Figure 7 do not exhibit any
appreciable variation. Similarly, separations of proteins pre-
sented in Figure 8 vary very little. Interestingly, the retention of
the less retained proteins slightly increases suggesting an increase
in hydrophobicity of the stationary phase. These observations
clearly confirm the excellent stability of the support-ligand
linkage.
ACKNOWLEDGMENT
Support of this work by a Grant of the National Institute of
General Medical Sciences, National Institutes of Health (GM44885)
and by the Materials Sciences and Engineering Division of the
U.S. Department of Energy under Contract No. DE-AC02-
0
5CH11231 is gratefully acknowledged.
(39) Wilson, N. S.; Gilroy, J.; Dolan, J. W.; Snyder, L. R. J. Chromatogr., A 2004,
1
026, 91-100.
(
40) Euerby, M. R.; Petersson, P. J. Chromatogr., A 2005, 1088, 1-15.
(
41) Silva, C. R.; Jardim, I. C. S. F.; Airoldi, C. J. Chromatogr., A 2003, 987,
Received for review January 2, 2006. Accepted April 4,
006.
1
27-138.
2
(
42) Rohr, T.; Hilder, E. F.; Donovan, J. J.; Svec, F.; Fr e´ chet, J. M. J.
Macromolecules 2003, 36, 1677-1684.
AC060006S
Analytical Chemistry, Vol. 78, No. 14, July 15, 2006 4975