Purification of synthetic peptides using a catching full-length sequence by polymerization approach

Article abstract of DOI:10.1021/ol403426u

During automated solid-phase peptide synthesis, failure sequences were capped with acetic anhydride. After synthesis, a polymerizable methacrylamide tag was attached to the full-length sequences. Peptide purification was then achieved by polymerizing the

Full text of DOI:10.1021/ol403426u

Letter
pubs.acs.org/OrgLett
Purification of Synthetic Peptides Using a Catching Full-Length
Sequence by Polymerization Approach
Mingcui Zhang, Durga Pokharel, and Shiyue Fang*
Department of Chemistry, Michigan Technological University, 1400 Townsend Drive, Houghton, Michigan 49931 United States
S
* Supporting Information
ABSTRACT: During automated solid-phase peptide syn-
thesis, failure sequences were capped with acetic anhydride.
After synthesis, a polymerizable methacrylamide tag was
attached to the full-length sequences. Peptide purification
was then achieved by polymerizing the full-length sequences,
washing away impurities, and cleaving the peptide product
from the polymer.
ynthetic peptides have wide applications. For example,
of harmful solvents, which results in high waste to product
Sthere are 65 therapeutic peptides on the market and around ratios, usually more than 1,000. Other methods for peptide
purification include fluorous affinity purification,3 antigen−
antibody affinity purification,4 covalent capture with a solid
matrix,5 and lipophilic tag assisted chromatography.6 All these
methods are extensively studied but are still not ideal for large
scale peptide drug purification.
Recently, we reported a catching by polymerization strategy
for purification of oligodeoxynucleotides (ODN).7 In one
method, at the end of automated synthesis, a polymerizable
methacrylamide group is attached to the full-length sequence.
ODN purification is then achieved by polymerizing the full-
length sequences, washing away failure sequences and other
impurities, and cleaving product from the polymer.7a The
method avoided many drawbacks associated with chromatog-
raphy and other known methods and could be used for large
scale and high throughput ODN purification. Here we report
our results on using the strategy for synthetic peptide
purification.
To demonstrate feasibility, compound 1, which contains an
acid-labile linker and a reactive 4-nitrophenyl carbonate
function, was designed. Because we planned to use the
commercially available 2-Cl-trityl polystyrene resin as the
solid support for peptide synthesis, the linker must be stable in
1% TFA, which are the conditions for cleaving peptides from
the resin.2,8 However, the linker must be readily cleavable under
more acidic conditions to obtain good recovery yield in the
purification process. These goals were achieved by careful
tuning the electron density of the benzene ring in the linker.
The molecule was also designed to be easy to synthesize. As
shown in Scheme 1, 4-hydroxybenzyl alcohol was alkylated with
methyl chloroacetate to give 2,9 which was reacted at room
temperature with 2,2′-(ethylenedioxy)bis(ethylamine) to give
270 peptides in clinical trials.1 Peptides are mostly synthesized
on a solid support by stepwise addition of amino-protected
amino acid (aa) monomers. Each synthetic cycle accomplishes
the addition of one aa. It usually consists of three steps, which
are coupling, capping, and deprotection. Excess reagents and
side products in each step are removed by washing. After
synthesis, the product is cleaved from the support and fully
deprotected. Besides the desired full-length peptide, the major
impurities in the crude product include failure sequences and
small molecules from side chain protecting groups. The latter
can usually be removed by precipitation with cold diethyl ether
from a trifluoroacetic acid (TFA) solution. However, the failure
sequences have physical properties similar to those of the full-
length peptide and cannot be removed in the process.
Currently, the first industrial choice for peptide purification is
reversed-phase (RP) HPLC. However, it has limitations. For
example, in the manufacture of enfuvirtide, which is a 36-aa
peptide used to treat HIV infection, RP HPLC is the primary
tool for purification. When a linear solid-phase synthesis
method was used, the purity of the crude product was 30−40%,
which is acceptable considering the length of the synthesis and
the convenience of the procedure. Unfortunately, the product
had to be purified by a complicated procedure involving
difficult, low-concentration, low-throughput, and two-pass
HPLC.2 To ease purification, later a combined solid-phase
and solution-phase synthesis method was developed, and the
purity of product was improved to ∼75%. However, the
purification was not much easier. The product was loaded in
multiple portions onto a column for preparative HPLC.2 In
general, the disadvantages of HPLC for large scale purification
include high capital costs on instrument and column, labor
intensiveness, high energy demand for solvent evaporation, and
inability to resolve peptides with multiple higher order
structures. Most importantly, HPLC consumes large volumes
Received: November 26, 2013
Published: February 14, 2014
© 2014 American Chemical Society
1290
dx.doi.org/10.1021/ol403426u | Org. Lett. 2014, 16, 1290−1293

Organic Letters
Letter
capped with excess acetic anhydride. The steps for removing
the Fmoc protecting group were normal. After assembling the
sequence, the full-length peptide was reacted with compound 1
in the presence of a base at room temperature overnight to give
6 (see Scheme 2). The failure sequences (7) could not react
because they were capped with acetic anhydride. The crude
product, which mainly contained 6 and 7, were cleaved from
the resin using 1% TFA to give full-length peptide 8 and failure
sequences 9. Under these conditions, the side chain protecting
groups and the 4-hydroxybenzyl alcohol linker were stable.2
The mixture was analyzed with RP HPLC (trace a, Figure 1).
The full-length peptide appeared at 61 min. To generate a
control HPLC profile, a portion of the crude was treated with a
TFA cocktail containing 81.5% TFA, 1.0% triisopropylsilane
(TIPS), 5.0% water, 2.5% ethane dithiol (EDT), 5.0%
thioanisole, and 5.0% phenol to globally remove protecting
groups and to cleave the methacrylamide tag. After Et2O
precipitation, the crude product was analyzed with RP HPLC
(trace b). The fully deprotected full-length peptide (5)
appeared at 23 min.
Another portion of the crude containing 8 and 9 (with side
chain protecting groups and methacrylamide tag not removed)
was subjected to polymerization under typical polyacrylamide
gel formation conditions [N,N-dimethylacrylamide, N,N′-
methylenebis(acrylamide), (NH4)2S2O8, TMEDA]. The full-
length peptide 8 was incorporated into the polymer 10. The
failure sequences 9 and other impurities remained in solution
(Scheme 2). The gel was washed extensively with solvents such
as water, DMF, and MeCN to remove impurities. Full-length
peptide was then cleaved from the gel and fully deprotected by
treating with the TFA cocktail described above. At this stage,
the peptide was still contaminated with small molecules from
side chain protecting groups and the cocktail. Removing them
was achieved by precipitation of peptide with cold Et2O. The
pure peptide 5 was analyzed with RP HPLC (trace c, Figure 1).
Scheme 1. Synthesis of Compound 1
3.10 Methacrylation of 3 gave compound 4, which was
converted to 1 by reacting with 4-nitrophenylchloroformate.
The synthesis did not use any expensive and highly reactive
reagents and is expected to be readily scalable.
Since peptide synthesis and purification are much more
complicated than ODN, at the beginning of the project, we
used the short 6-aa peptide 5 (LWTRFA) to demonstrate the
feasibility of the concept. The 2-Cl-trityl polystyrene resin
preloaded with Ala was selected as the solid support. The
Fmoc-protected Phe, Arg(Pbf), Thr(tBu), Trp(Boc), and Leu
were used as aa monomers. Synthesis was carried out on an
automated synthesizer. In the coupling steps, only 2.1 equiv of
monomers was used, which was lower than the amounts usually
suggested by manufacturers of peptide synthesizers. The
lowered amount was expected to generate more failure
sequences and to better show the effectiveness of the new
purification method. After coupling, failure sequences were
Scheme 2. Peptide Cleavage from Resin and Purification by Catching Full-Length Sequence by Polymerization
1291
dx.doi.org/10.1021/ol403426u | Org. Lett. 2014, 16, 1290−1293

Organic Letters
Letter
purification. Some drugs need to be purified for more than one
time, and ideally, each time uses a different purification
technique. The catching by polymerization method may be an
excellent choice for initial removal of large quantities of failure
sequences. This is predicted to be particularly useful for the
purification of peptide sequences that have inherently low
coupling yields in certain cycles. Furthermore, due to the ease
of removal of failure sequences using the method, it is possible
for drug manufacturers to reduce the equivalents of aa
monomers and their activators during synthesis. The rebalance
of reagent costs and purification costs will provide new
opportunities to lower the overall expenses of drug production.
In conclusion, we have successfully developed a new method
for peptide purification. The method used a concept that is
completely different from any previously reported peptide
purification techniques. It is easy to use and readily scalable and
gives peptides with good purity and recovery yields. Adapting
the method for purification of several peptide drugs such as
Bivalirudin (20 aa),12 Ziconotide (25 aa),13Thymalfasin (28
aa),14 Enfuvirtide (36 aa),15 and Pramlintide (39 aa)16 is
underway.
ASSOCIATED CONTENT
Figure 1. RP HPLC profiles of peptides. (a) Crude 6-aa peptide
containing full-length sequences 8 and failure sequences 9; the side
chain protecting groups and the methacrylamide tag on 8 were on. (b)
Crude peptide with side chain protecting groups and methacrylamide
tag removed. (c) Peptide 5 purified using the catching full-length
sequences by polymerization approach.
■
S
* Supporting Information
Experimental procedures, compound characterization, and
images of NMR spectra, ESI-MS, and HPLC profiles. This
material is available free of charge via the Internet at http://
pubs.acs.org.
The purity of the peptide was more than 98%. The recovery
yield of the purification process was estimated to be 66% by
comparing the areas of the peaks at 23 min in traces b and c.
The identity of the peptide was confirmed with ESI-MS and
NMR (see Supporting Information).
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: shifang@mtu.edu.
Notes
To further demonstrate the effectiveness of the purification
technique, the 16-aa peptide 11 (KNRWEDPGKQLYNVEA),
which is a peptide derived from human C3d and carries the
LYNVEA CR2 binding sequence,11 was synthesized under
similar conditions as described for 5. Fmoc aa Glu(OtBu), Val,
Asn(Trt), Tyr(tBu), Leu, Gln(Trt), Lys(Boc), Gly, Pro,
Asp(OtBu), Trp(Boc), and Arg(Pbf) were used as the
monomers. To further increase the difficulty of the purification
task, the amount of aa monomers was lowered to 1.5 equiv.
The purification process was exactly the same. The RP HPLC
profiles for the crude peptide with the methacrylamide tag and
side chain protecting groups on, that with tag and protecting
groups removed, and purified full-length peptide are included in
Supporting Information. The purity of the peptide (11) is more
than 96%. The recovery yield was estimated to be 73%. The
identity of 11 was confirmed with ESI MS and NMR.
The catching by polymerization peptide purification method
has significant advantages over other techniques. Compared
with chromatographic methods such as HPLC and lipophilic
tag assisted purification,6 the method does not need any
expensive instrument, column, or large volumes of harmful
solvents. Compared with RP cartridge, fluorous, and antigen−
antibody affinity purification methods,3,4 the method does not
require a disposable column. Compared with the methods
using covalent capture,5 the method does not need a
functionalized solid matrix. Importantly, using the method,
purification is achieved by simple manipulations such as
shaking, washing, extraction, and precipitation. As a result, it
is expected to be particularly useful for large scale peptide drug
The authors declare the following competing financial
interest(s): The authors are inventors of a provisional US
patent filed by Michigan Technological University.
ACKNOWLEDGMENTS
■
Financial support from US NSF (CHE-0647129 and CHE-
1111192), MTU Biotech Research Center, MUCI and MTU
REF Technology Commercialization; the assistance from Mr.
Dean W. Seppala (electronics) and Mr. Jerry L. Lutz (NMR);
and an NSF equipment grant (CHE-9512445 and CH-
1048655) are gratefully acknowledged.
REFERENCES
■
(1) (a) Bellmann-Sickert, K.; Beck-Sickinger, A. G. Trends Pharmacol.
Sci. 2010, 31, 434. (b) Vlieghe, P.; Lisowski, V.; Martinez, J.;
Khrestchatisky, M. Drug Discovery Today 2010, 15, 40−56.
(2) Bray, B. L. Nat. Rev. Drug Discovery 2003, 2, 587−593.
(3) (a) Akcay, G.; Kumar, K. J. Fluorine Chem. 2009, 130, 1178−
1182. (b) Montanari, V.; Kumar, K. J. Am. Chem. Soc. 2004, 126,
9528−9529. (c) Filippov, D. V.; van Zoelen, D. J.; Oldfield, S. P.; van
der Marel, G. A.; Overkleeft, H. S.; Drijfhout, J. W.; van Boom, J. H.
Tetrahedron Lett. 2002, 43, 7809−7812. (d) Fustero, S.; Sancho, A. G.;
Chiva, G.; Sanz-Cervera, J. F.; del Pozo, C.; Acena, J. L. J. Org. Chem.
2006, 71, 3299−3302. (e) de Visser, P. C.; van Helden, M.; Filippov,
D. V.; van der Marel, G. A.; Drijfhout, J. W.; van Boom, J. H.; Noort,
D.; Overkleeft, H. S. Tetrahedron Lett. 2003, 44, 9013−9016.
(f) Montanari, V.; Kumar, K. Eur. J. Org. Chem. 2006, 874−877.
(g) Montanari, V.; Kumar, K. J. Fluorine Chem. 2006, 127, 565−570.
1292
dx.doi.org/10.1021/ol403426u | Org. Lett. 2014, 16, 1290−1293

Organic Letters
Letter
(4) (a) Shogren-Knaak, M. A.; Imperiali, B. Tetrahedron Lett. 1998,
39, 8241−8244. (b) Bang, D.; Kent, S. B. H. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 5014−5019.
(5) (a) Vizzavona, J.; Villain, M.; Rose, K. Tetrahedron Lett. 2002, 43,
8693−8696. (b) Canne, L. E.; Winston, R. L.; Kent, S. B. H.
Tetrahedron Lett. 1997, 38, 3361−3364. (c) Funakoshi, S.; Fukuda, H.;
Fujii, N. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 6981−6985.
(6) Ball, H. L.; Mascagni, P. Int. J. Pept. Protein Res. 1996, 48, 31−47.
(7) (a) Fang, S.; Fueangfung, S. Org. Lett. 2010, 12, 3720−3723.
(b) Fang, S.; Fueangfung, S.; Lin, X.; Zhang, X.; Mai, W.; Bi, L.; Green,
S. A. Chem. Commun. 2011, 47, 1345−1347. (c) Yuan, Y.; Fueangfung,
S.; Lin, X.; Pokharel, D.; Fang, S. RSC Adv. 2012, 2, 2803−2808.
(8) (a) Barlos, K.; Gatos, D.; Kapolos, S.; Poulos, C.; Schafer, W.;
Yao, W. Q. Int. J. Pept. Protein Res. 1991, 38, 555−561. (b) Barlos, K.;
Gatos, D.; Kapolos, S.; Papaphotiu, G.; Schafer, W.; Yao, W. Q.
Tetrahedron Lett. 1989, 30, 3947−3950. (c) Athanassopoulos, P.;
Barlos, K.; Gatos, D.; Hatzi, O.; Tzavara, C. Tetrahedron Lett. 1995, 36,
5645−5648. (d) Harre, M.; Nickisch, K.; Tilstam, U. React. Funct.
Polym. 1999, 41, 111−114.
(9) Naganawa, A.; Matsui, T.; Ima, M.; Saito, T.; Murota, M.;
Aratani, Y.; Kijima, H.; Yamamoto, H.; Maruyama, T.; Ohuchida, S.;
Nakai, H.; Toda, M. Bioorg. Med. Chem. 2006, 14, 7121−7137.
(10) Tang, W.; Fang, S. Y. Tetrahedron Lett. 2008, 49, 6003−6006.
(11) Frade, R.; Hermann, J.; Barel, M. Biochem. Biophys. Res.
Commun. 1992, 188, 833−842.
(12) Garner, W. L.; Linden, J. A.; Chrysant, G. S. Cardiovasc.
Hematol. Agents Med. Chem. 2013, 11, 44−48.
(13) Pope, J. E.; Deer, T. R. Expert Opin. Pharmacother. 2013, 14,
957−966.
(14) Ciancio, A.; Rizzetto, M. Ann. NY Acad. Sci. 2010, 1194, 141−
146.
(15) Joly, V.; Jidar, K.; Tatay, M.; Yeni, P. Expert Opin. Pharmacother.
2010, 11, 2701−2713.
(16) Younk, L. M.; Mikeladze, M.; Davis, S. N. Expert Opin.
Pharmacother. 2011, 12, 1439−1451.
1293
dx.doi.org/10.1021/ol403426u | Org. Lett. 2014, 16, 1290−1293