J. Sep. Sci. 2006, 29, 1245 – 1250 S. Dube et al.

1245

Simiso Dube1 Original Paper

Mkhululi T. Khumalo1
Nelson Torto2
Jacob A. Nyati3
1
Department of Applied
Chemistry, National University
of Science and Technology,
Ascot, Bulawayo, Zimbabwe
2 An MEKC method was developed for the separation and characterization of phenyl-
Department of Chemistry,
University of Botswana,
Gaborone, Botswana
3
Department of Textile
Technology, National University
of Science and Technology,
Ascot, Bulawayo, Zimbabwe
Characterization of amino acids in silk sericin
protein from Gonometa rufobrunnae by MEKC
with phenyl isothiocyanate derivatization
isothiocyanate (PITC)-labeled amino acids derived from Gonometa rufobrunnae silk-
worm after microdialysis sample cleanup. The influence of the buffer and SDS con-
centration on the resolution of the amino acids was investigated. A buffer system
consisting of 25 mM phosphate, 10 mM borate buffer at pH 9.00, and 70 mM SDS
showed the best results, with 13 PITC-amino acid derivatives being resolved out of
15 possible amino acids that were under study. Microdialysis sampling demonstra-
ted its efficiency as a sample cleanup technique. Sericin protein from G. rufobrunnae
was found to be characterized by at least 11 positively identified amino acids. These
included His, Tyr, Ser, Ala, Phe, Lys, Gly, Arg, Cys, Glu, and Asp. Leu/Met and Val/Thr
were coeluting pairs and hence could not be positively confirmed.
Keywords: Amino Acids / Gonometa rufobrunnae / MEKC / Phenyl isothiocyanate / Sericin protein /
Received: January 20, 2006; revised: February 21, 2006; accepted: February 23, 2006
DOI 10.1002/jssc.200600045

1 Introduction from fibroin by a process known as degumming and

Silks are fibrous proteins of interest in the textile indus-
try, which have also demonstrated superior properties
such as good mechanical and environmental stability,
biocompatibility, and biodegradability [1–6] in the bio-
material industry. It is because of these attractive proper-
ties that silk-like proteins have been produced by genetic
engineering for various applications, especially as medi-
cal and degradable biomaterials and drug delivery car-
riers [4, 5, 7–11]. The largest contributor toward commer-
cial production of silk is Bombyx mori, a domesticated silk-
worm cultivated in Asia. Most of the research work on
silks has been focused on this type of silkworm in an
attempt to comprehend its chemical properties, improve
its physical properties, and hence broaden its application
in biotechnology, biomaterial, and biomedical fields.
then discarded as waste. It has been suggested that seri-
cin could be recovered from waste not only to reduce the
environmental impact but also to utilize it for economic
and social purposes. Zhang [12] published a detailed
review of various applications of sericin protein in the
biomaterial industry. Some of these applications could
be correlated to the chemical composition, particularly
the amino acid profile of the protein. For example, the
moisture absorbing and desorption properties of sericin
that renders it suitable for production of cosmetics and
as a biodegradable material depend on the amount of
serine present in the protein [13]. To our knowledge, very
little has been reported on both chemical and physical
characteristics of sericin derived from Gonometa rufobrun-
nae [14], a silkworm found in abundance in Southern
Africa.

B. mori cocoon consists of two major proteins, fibroin,
which is utilized for silk fibers, and sericin, a gum-like
protein, which holds fibroin together. The latter consti-
tutes about 25–30% [12] of silk protein and is separated
Correspondence: Dr. Simiso Dube, Department of Applied
Chemistry, National University of Science and Technology, P. O.
Box AC 939, Ascot, Bulawayo, Zimbabwe.
E-mail: simi@nust.ac.zw.
Fax: +263-9-286-390.
Abbreviations: PITC, phenylisothiocyanate; PTC, phenyl thiocar-
bamyl; UHP, ultrahigh purity
Protein profiling remains a fundamental study in biolo-
gical, biochemical, medical, clinical, and natural prod-
uct fields. Characterization of proteins involves the
hydrolysis of the protein into free amino acids followed
by separation using highly efficient or selective tech-
niques typified by HPLC [14–17], CZE [18–20], CEC [21],
and MEKC [22–25].
In general, most amino acids are UV/Vis or fluorescence
inactive because they lack chromophores or fluoro-
phores. Hence, derivatization is a prerequisite when
employing electrically driven capillary column tech-

i 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

1246 S. Dube et al.
niques due to their inherent poor sensitivity as a result
of the short path length. Precolumn derivatization using
phenyl isothiocyanate has been successfully applied in
the determination of amino acids using CE with UV
detection [21–24]. The reagent reacts with the alpha
amine of the amino acids under basic conditions form-
ing phenyl thiocarbamyl (PTC) derivatives. Phenyl-
isothiocyanate (PITC) derivatization is usually favored
because it is rapid and applicable to both primary and
secondary amino acids and, furthermore, it yields rela-
tively stable derivatives.
This work evaluates the capability of MEKC with PITC pre-
column derivatization as a method for profiling amino
acids in sericin protein. The characterization of G. rufo-
brunnae sericin protein is essential in developing an
understanding and appreciation of its capabilities and
potential in the fields of biotechnology and biomaterials.
Effects of SDS concentration and of microdialysis sample
cleanup on the resolution of the amino acids are investi-
gated.
2 Experimental
2.1 Reagents and chemicals
G. rufobrunnae silkworm cocoons were originally collected
from Gwanda (Zimbabwe). Phenyl isothiocyanate (98.0%
purity) used for derivatization of amino acids and SDS
were obtained from Sigma (Steinheim, Germany). Fifteen
amino acids standards were used in this study and were
all of analytical grade. Tyrosine, leucine, valine, glycine,
arginine were from Saarchem (Muldersdrift, RSA), gluta-
mic acid, lysine, alanine, and serine were purchased
from Merck (Darmstadt, Germany). Histidine and aspar-
tic acid were obtained from Aldrich (Dorset, UK), cystine
was from Fluka (Switzerland), methionine and threonine
were from Sigma (St. Louis, USA), and phenylalanine was
purchased from BDH Chemicals (Poole, UK).
Sodium dihydrogen orthophosphate, di-sodium hydro-
gen phosphate, and triethylamine were purchased from
Saarchem (Muldersdrift, RSA). Sodium tetraborate was
from NT Laboratory Supplies (Johannesburg, RSA) and
ethanol (99%) was from Ultrafine (Finchley, London, UK).
All buffer solutions and other aqueous reagents were pre-
pared in ultrahigh purity (UHP) water and processed
through a Millipore Quantum Ultrapure Ionex Gradient
A10 purification system (Millipore, Molsheim, France).
2.2 Instrumentation
Separation conditions for amino acids were developed
using a Crystal CE (Prince Technologies, The Nether-
lands) equipped with a Crystal 100 variable UV/Vis detec-
tor. WinPrince software (Prince Technologies) was used
for controlling the CE system whereas DAx software ver-
sion 6.1 (Prince Technologies) was used for data acquisi-
i 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2006, 29, 1245 – 1250
tion and analysis. CE separations were achieved using
uncoated fused-silica capillary tubing of 50 lm id,
375 lm od, 95 cm total length, and 35 cm effective
length, and were purchased from Composite Metal Ser-
vices (Worcester, UK).
2.3 Analytical procedures
2.3.1 Extraction of sericin protein from silkworm
cocoons
Sericin was extracted from the silkworm cocoons using
the “degumming” procedure. In this process cocoons
were boiled in water in a steam vessel using an oil bath at
a temperature of 1208C for a period of 60 min. Insoluble
traces of fibroin and other solid impurities were sepa-
rated from the aqueous sericin solution by filtration. The
sericin gum was further heated at 1008C to evaporate the
sample to dryness. The remaining solid mass (brown in
color) was allowed to cool at ambient temperature.
2.3.2 Acid hydrolysis of sericin
The dried sericin gum was crushed in a mortar and about
100 mg of the resulting powder was weighed into pre-
viously acid-washed Eppendorff vials. One milliliter of
6 M HCl was then added into the vials and heated in an
oven at 1108C for 24 h. The acid hydrolysis was carried
out in tightly sealed vials. The resulting acid hydrolysates
were allowed to cool to room temperature and filtered
through a 0.45 lm Millex-HV filter membrane (Millipore,
Bedford, MA, USA) prior to further sample treatment.
2.3.3 Microdialysis sample cleanup
Microdialysis technique was used for sample cleanup.
Two-hundred microliters of sericin acid hydrolysates was
neutralized using 1200 lL of 1.0 M NaOH. Microdialysis
sampling was accomplished using a tunable microdialy-
sis probe equipped with a polysulfone hollow fiber mem-
brane with a 15 mm effective dialysis length and 3 kDa
MW cutoff, A/G Technology (Needham, MA). UHP water
was used as the perfusion liquid at a flow rate of 2 lL/
min. A 5 ml gas tight syringe and a CMA/400 microinjec-
tion pump, both supplied by CMA/Microdialysis (Solna,
Sweden), were used to continuously pump the perfusion
liquid into the probe during sampling and for the sam-
ple cleanup. The dialysate fractions from the sericin
hydrolysates were collected with a CMA/142 fraction col-
lector also from CMA/Microdialysis, and subsequently
injected into the CE system.
2.3.4 Precolumn derivatization of amino acids with
PITC
The derivatization procedure was adapted from the
method described by Siebert et al. [15]. Amino acid stan-
dard stock solutions were prepared in 0.1 M HCl at a con-
www.jss-journal.com
J. Sep. Sci. 2006, 29, 1245 – 1250 Characterization of silk sericin protein from G. rufobrunnae by MEKC
centration of 2 mM. A mixture of 15 amino acids was pre-
pared by using equivolumes of the stock solutions such
that each amino acid had a final concentration of
1.3610 – 4 M. This mixture (100 lL) was dried under
vacuum using a dessicator attached to an Alcatel vacuum
pump (Vacutec, Roosevelt Park, RSA). Individual amino
acid stock solutions and the reagent blank were dried
under similar conditions. Prior to derivatization the
dried samples were resuspended in 20 lL of the drying
reagent (ethanol-water-triethylamine, 2:2:1 v/v/v) and
redried for 1 h under vacuum. PITC derivatization was
accomplished by adding 20 lL of the derivatization
reagent (ethanol-water-triethylamine-PITC, 7:1:1:1, v/v/v/
v) and incubating for 20 min at room temperature. The
final step in the derivatization process involved drying
the samples under vacuum for at least 6 h.
Derivatized samples were redissolved into 1000 lL of
22 mM phosphate buffer (pH 6.4), ultrasonicated for a
few seconds, and vortexed for 60 s. The solutions were
finally filtered through 0.45 lm Millex-HV filter mem-
branes and transferred to CE vials ready for analysis. In
cases where storage was required, the vacuum dried sam-
ples were kept in the freezer.
Both the microdialysis cleaned and the uncleaned sericin
acid hydrolysates (100 lL) were dried under vacuum and
also derivatized as described earlier. Sample blanks were
prepared for both samples. To confirm peak identifica-
tion it was necessary to employ a spiking technique in
which 100 lL of the amino acid mixture was spiked into
the sericin sample prior to microdialysis cleaning and
derivatization.
2.3.5 CE procedures
Various buffer systems were prepared and evaluated for
the MEKC studies. These included 25 mM phosphate buf-
fer consisting of various amounts of SDS or a combina-
tion of 25 mM phosphate, 10 mM borate, and varying
concentrations of SDS concentrations.
A new capillary column was conditioned by rinsing with
1.0 M and 0.2 M NaOH for 30 min each followed by
10 min rinses with UHP water and the running buffer.
Two-minute rinses were carried out using the running
buffer after each run in order to improve reproducibility.
Daily capillary column conditioning steps involved rin-
sing with 0.2 M NaOH followed by UHP water and the
buffer for 30, 10, and 5 min, respectively.
3 Results and discussions
3.1 Separation of PTC-amino acid derivatives by
CZE
In our preliminary studies the separation of PITC-labeled
amino acids was attempted using the conventional CZE
i 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1247
Figure 1. Effect of borate buffer concentration on the sep-
aration of PTC-amino acid derivatives by CZE. CE condi-
tions: 10 mM borate buffer; applied voltage of 20 kV, electro-
kinetic injection at 5.0 kV for 0.5 min; UV detection at
254 nm. See Fig. 2 for peak identification.
mode. The effect of borate buffer concentration on the
resolution of amino acids was evaluated. Previously,
other researchers successfully employed borate buffer
for the separation of different types of labeled amino
acids [24–26]. Figure 1 demonstrates a separation that
was achieved when 10 mM borate was used. Only 7 out of
the possible 15 peaks were resolved under these condi-
tions. In an attempt to improve the resolution, the con-
centration was increased to 12.5 and 25 mM. Unfortu-
nately, this resulted in loss of sensitivity particularly at
25 mM borate buffer concentration (results not shown).
This anomaly was in contrast with earlier observations
by previous workers when they used higher concentra-
tions of borate buffer for the separation of derivatized
amino acids [24, 26]. A similar phenomenon was however
reported by Takizawa and Nakamura [25] when they
used a running buffer consisting of 80 mM sodium
borate at pH 9.2 and 45 mM a-CD for separating 18 FITC-
amino acid derivatives. They observed a decrease in
detection sensitivity of amino acids when using both the
UV and LIF detection. The authors speculated that the
phenomenon could have resulted from the possible light
scattering effects that could occur when the a-CD concen-
tration is increased. In this present work, the source for
such a phenomenon was not clearly understood or eluci-
dated.
Because of the limitations encountered with borate buf-
fer, the BGE was changed to 25 mM sodium phosphate
buffer (pH 7.12). The sensitivity improved under these
conditions (results not shown) although resolution was
still poor. Unfortunately increasing the phosphate buffer
concentration to 50 mM also resulted in a decrease in
sensitivity. Consequently, subsequent separations were
accomplished using 25 mM phosphate since the peaks
observed under these conditions showed reasonable sen-
sitivity.
www.jss-journal.com
1248 S. Dube et al. J. Sep. Sci. 2006, 29, 1245 – 1250

3.2 Separation of amino acids by MEKC

3.2.1 Effect of SDS concentration on resolution of
PTC-amino acid derivatives
Selectivity is determined by the mechanism that is
involved in the separation. The capability to control
selectivity can improve resolution and hence yield infor-
mation that is useful for the validation of a method. In
CE selectivity is achieved by varying the pH of the run-
ning buffer or by using buffer additives. Thus, introduc-
ing a surfactant into a running buffer at concentrations
greater than its CMC offers selectivity since the analytes
would electromigrate on the basis of the differences in
their distribution between a pseudostationary phase and
a running buffer.
Figure 2 shows a great improvement in the resolution of
15 PTC-amino acid derivatives when 35 mM SDS was
incorporated into 25 mM phosphate buffer (pH 7.12).
Ten peaks were well resolved; however, four of the amino
acids (Leu, Met, Thr, and Val) coeluted as a single peak at
7.33 min. Similarly, Phe and Ser coeluted as a single peak
at 7.64 min. Increasing the SDS concentration to 50 mM
resulted in loss of resolution for the early eluting amino
acids. Interestingly the peak areas were greater under
these conditions indicating that there was an increase in
sensitivity.
In an attempt to improve the separation of different buf-
fer systems, comprising a mixture of 25 mM phosphate
buffer and 10 mM borate buffer at pH 9.00, they were
investigated at varyious concentrations of SDS. Generally
the combination of the two buffers had the effect of
improving the resolution of the amino acids. Eleven
peaks were well resolved using 35 mM SDS (Fig. 3A)
whereas 13 peaks were separated using both 50 mM and
70 mM SDS (Fig. 3B and C, respectively). Baseline resolu-
tion was however achieved with 70 mM SDS concentra-

Figure 3. Electropherograms of the separation of PTC-

amino acid derivatives using 25 mM phosphate buffer,
10 mM borate buffer at pH 9.00 for (A) 35 mM SDS, (B)
50 mM SDS, and (C) 70 mM SDS; other conditions as in Fig.
1 and peak identifications as in Fig. 2.

tion. The Phe/Ser pair was completely resolved to indivi-

Figure 2. MEKC of PTC-amino acid derivatives using 25 mM
phosphate buffer at pH 7.12 and 35 mM SDS; other condi-
tions as in Fig. 1. Peak ID: 1, histidine; 2, tyrosine; 3, leucine;
4, methionine; 5, threonine; 6, valine; 7, lysine; 8, serine; 9,
phenylalanine; 10, alanine; 11, glycine; 12, arginine; 13,
cystine; 14, glutamic acid; 15, aspartic acid; U, unknown; R,
reagent.
dual peaks. The peak originally due to Leu/Met/Thr/Val
was separated into two distinct peaks of Leu/Met and
Thr/Val pairs. The buffer system consisting of 70 mM SDS
was used for subsequent separation of amino acids in the
sericin sample as it offered the best resolution.
3.2.2 Microdialysis sample cleanup
Microdialysis sampling is an in situ sampling and sample
cleanup technique known for its efficient clean up of

i 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2006, 29, 1245 – 1250
Figure 4. Electropherograms of the separation of PTC-
amino acid derivatives in G. rufobrunnae sericin protein
using 25 mM phosphate buffer, 10 mM borate buffer at
pH 9.00 and 70 mM SDS for (A) microdialysis cleaned sam-
ple, (B) uncleaned sample, and (C) microdialysis cleaned
spiked sample; other conditions as in Fig. 1 and peak identifi-
cation as in Fig. 2.
complex biological matrices [27]. The process of sample
cleanup in a microdialysis experiment is driven by diffu-
sion as analytes are selectively allowed to pass the micro-
dialysis membrane as governed by their size and the
membrane cutoff. If coupled on-line, microdialysis sam-
pling provides a chromatographically clean sample (see
Fig. 4A).
i 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Characterization of silk sericin protein from G. rufobrunnae by MEKC 1249
The electropherogram shown in Fig. 4A was obtained
from the sample that had been previously processed
through microdialysis using a 3 kDa polysulfone dialysis
membrane at a perfusion flow rate of 2 lL/min. The sam-
ple analyzed after microdialysis sampling gave a cleaner
electropherogram with narrower peaks. In comparison,
the filtered sample showed higher background and
broader peaks (see Fig. 4B), which were attributed to the
relatively higher concentrations of the analytes in the
sample and higher matrix effects. Although microdialy-
sis is an excellent cleanup method, it should be noted
that by virtue of its process it results in the dilution of
the sample. Consequently, sensitivity was reduced in the
microdialysis cleaned up sample. Also the migration
times of the amino acids were longer in the microdialysis
cleaned up sample probably due to the reduction of the
salt effects.
Peak identification was confirmed by the spiking
method approach, in which 100 lL of the amino acids
standard mixture was added to the sericin acid hydroly-
sate prior to microdialysis and derivatization. The elec-
tropherograms for the spiked (Fig. 4C) and the unspiked
samples (Fig. 4A) have a comparable number of peaks
illustrating that the amino acids in the sample corre-
sponded well with those of the standard mixture.
3.2.3 Characterization of sericin silk protein
At least 13 amino acid peaks were detected in the sericin
silk protein. However, only 11 amino acids were posi-
tively identified as His, Tyr, Ser, Ala, Phe, Lys, Gly, Arg,
Cys, Glu, and Asp. The other two peaks could have been
due to Leu, Met, Val, and Thr or a combination of these
amino acids (i. e., Leu/Met and Val/Thr) since they could
not be resolved under the separation conditions used.
The amino acids identified in this work are consistent
with those found by Freddi et al. [14] when they studied
sericin protein from G. rufobrunnae using an HPLC
method. However, in their work, they identified 18
amino acids in sericin protein.
4 Concluding remarks
MEKC proved to be a superior separation method for the
separation of PTC-amino acid derivatives than conven-
tional CZE. As much as borate buffer appeared to yield
better resolution than phosphate buffer it resulted in the
reduction of sensitivity, a phenomenon, which was not
clearly understood. The best buffer system consisted of
25 mM phosphate, 10 mM borate buffer at pH 9.00, and
70 mM SDS. Addition of SDS improved the resolution of
amino acids. Using the separation method developed, 11
out of the 13 peaks observed were positively identified.
This method can be used in further work on quantitative
determination of amino acids in sericin protein.
www.jss-journal.com
1250 S. Dube et al.
The authors would like to thank the Southern and Eastern African
Network of Analytical Chemists (SEANAC) for supporting S. D. The
National University of Science and Technology is acknowledged
for granting leave for S. D. We are grateful to the University of
Botswana for providing the facilities.
5 References
[1] Demura, M., Asakura, T., Biotechnol. Bioeng. 1989, 33,
598 – 603.
[2] Demura, M., Asakura, T., Nakamura, E., J. Biotechnol.
1989, 10, 113 – 120.
[3] Demura, M., Asakura, T., Kuroo, T., Biosensors 1989, 4,
361 – 372.
[4] Minoura, N., Aiba, S.-I., Gotoh, Y., Tsukada, M., Imai, Y.,
J. Biomed. Mater. Res. 1995, 29, 1215 – 1221.
[5] Mori, H., Tsukada, M., J. Biotechnol. 2000, 74, 95 – 103.
[6] Yoshimizu, H., Asakura, T., J. Appl. Polym. Sci. 1990, 40,
127 – 134.
[7] Asakura, T., Nitta, K., Yang, J., Nakazawa, Y., Kaplan, D.,
Biomacromolecules 2003, 4, 815 – 820.
[8] Peng, X., Zhang, L., Pan, X., Acta Polym. Sin. 2000, 4, 502 –
504.
[9] Katayama, H., Issiki, M., Yoshitomi, H., Biol. Pharmaceut.
Bull. 2000, 23, 1229 – 1253.
[10] Lee, K. Y., Kong, S. J., Park, W. H., Ha, W. S., Kwon, I. C., J.
Biomater. Sci. Polym. Ed. 1998, 9, 905 – 914.
[11] Asakura, T., Tanaka, C., Yang, M., Yao, J., Kurokawa, M.,
Biomaterials 2004, 25, 617 – 624.
[12] Zhang, Y. Q., Biotechnol. Adv. 2002, 20, 91 – 100.
i 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2006, 29, 1245 – 1250
[13] Kurioka, A., Kurioka, F., Yamazaki, M., Biosci. Biotechnol.
Biochem. 2004, 68, 774 – 780.
[14] Freddi, G., Svilokos, A. B., Ishikawa, H., Tsukada, M., J.
Appl. Polym. Sci. 1993, 48, 99 – 106.
[15] Siebert, J., Palmer, R. J., Hirsch, P., Appl. Environ. Microbiol.
1991, 57, 879 – 881.
[16] Battaglia, A., Bertoluzza, A., Calbucci, F., Eusebi, V. et al.,
J. Chromatogr. B 1999, 730, 81 – 93.
[17] G
mez-Ariza, J. L., Villegas-Portero, M. J., Bernal-Daza,
V., Anal. Chim. Acta 2005, 540, 221 – 230.
[18] Veledo, M. T., de Frutos, M., Diez-Masa, J. C., J. Chroma-
togr. A 2005, 1079, 335 – 343.
[19] Kaml, I., Vcelakova, K., Kenndler, E., J. Sep. Sci. 2004, 27,
161 – 166.
[20] Chen, F., Wang, S., Guo, W., Hu, M., Talanta 2005, 66,
755 – 761.
[21] Qi, M., Li, X.-F., Stathakis, C., Dovichi, N. J., J. Chromatogr.
A 1999, 853, 131 – 140.
[22] Liu, J., Hsieh, Y.-Z., Wiesler, D., Novotny, M., Anal. Chem.
1991, 63, 408 – 412.
[23] Ueda, T., Mitchell, R., Kitamura, F., J. Chromatogr. 1992,
593, 265 – 274.
[24] Okutsa, K., Terabe, S., Ando, T., J. Chromatogr. 1985, 332,
219 – 226.
[25] Takizawa, K., Nakamura, H., Anal. Sci. 1998, 14, 925 –
928.
[26] Trapp, G., Sydow, K., Dulay, M. T., Chou, T. et al., J. Sep.
Sci. 2004, 27, 1483 – 1490.
[27] Torto, N., Laurell, T., Gorton, L., Marko-Varga, G., Anal
Chim Acta 1998, 374, 111 – 135.
www.jss-journal.com