Journal of Biotechnology 106 (2003) 101–112

Degumming of silk fabric with several proteases

Giuliano Freddi a,∗ , Raffaella Mossotti b , Riccardo Innocenti b
a Stazione Sperimentale per la Seta, via Giuseppe Colombo 83, 20133 Milano, Italy
b C.N.R-ISMAC-Sezione di Biella, Corso Giuseppe Pella 16, 13900 Biella, Italy

Received 31 March 2003; received in revised form 11 August 2003; accepted 2 September 2003

Abstract

A crêpe silk fabric was treated with different alkaline (3374-L, GC 897-H), neutral (3273-C), and acid (EC 3.4 23.18)
proteases with the aim to study their effectiveness as degumming agents. Proteases were used under optimum conditions of pH
and temperature, while enzyme dosage (0.05–2 U/g fabric) and treatment time (5–240 min) were changed in order to study the
kinetics of sericin removal. Degumming loss with soap and alkali was 27 wt.%. The maximum amount of sericin removed in 1 h
was 17.6, 24, and 19 wt.% for 3374-L (2 U/g fabric), GC 897-H (1 U/g fabric), and 3273-C (0.1 U/g fabric), respectively. Under
the experimental conditions adopted, EC 3.4 23.18 was almost ineffective as a degumming agent. Degumming loss increased as
a function of the treatment time, reaching a value of 25 wt.% with 1 U/g fabric of 3374-L. The morphological analysis showed
that sericin was completely removed from the warp yarns of the crêpe fabric, while the highly twisted weft yarns still exhibited
the presence of sericin deposits within the most internal parts of the close fibre texture. The chromatographic pattern of soluble
sericin peptides changed as a function of the kind of enzyme used, enzyme dosage, and treatment time. A mixture of peptides
from 5 to 20 kDa in weight, with a weight-average molecular weight of about 12 kDa was obtained.
© 2003 Elsevier B.V. All rights reserved.

Keywords: Silk fabric; Sericin; Degumming; Proteases

1. Introduction ical -(ala–gly)n - repeating motifs (Lotz and Colonna

The silk filament spun by the silkworm Bombyx
mori is a composite material formed by two fibroin
filaments surrounded by a cementing layer of sericin.
Both fibroin and sericin, which account for about 75
and 25 wt.%, respectively, are proteins. Fibroin, the
real fibrous component, is a high molecular weight
polypeptide (∼ =350 kDa), whose primary structure
is rich in glycine, alanine, and serine amino acids
(∼
=85 mol%) in the molar ratio 3:2:1, which form typ-
∗ Corresponding author. Tel.: +39-02-2665990.
E-mail address: freddi@ssiseta.it (G. Freddi).
Cesari, 1979; Zhou et al., 2000). In the fibre, fibroin
chains are aligned along the fibre axis, held together
by a close network of interchain hydrogen bonds,
with adjacent -(ala–gly)n - sequences forming the
well known ␤-sheet crystals (Takahashi et al., 1991).
Sericin, the silk gum glueing the fibroin filaments, is a
complex mixture of 5–6 polypeptides widely differing
in size (40–400 kDa), chemical composition, structure
and properties, such as: solubility, hydrophylicity, and
stickiness (Gamo et al., 1977; Couble et al., 1987).
Fine details of the primary structure of four sericin
proteins have been reported (Garel et al., 1997).
The amino acid composition is charaterized by an

0168-1656/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.jbiotec.2003.09.006
102 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112
extremely high concentration of serine, which ranges
from 16 to 38 mol% in the different sericins. These
proteins, which are synthesized, secreted, and stored
in the middle silk gland, form a sheath around the fi-
broin core during spinning of the silk filament. Their
physiological function is to lower the shear stress
and to absorb the water squeezed from the stretched
fibroin mass during the process of fibre formation.
Silk processing from cocoons to the finished cloth-
ing articles consists of a series of steps which include:
reeling, weaving, degumming, dyeing or printing, and
finishing (Zahn, 1993). Degumming is a key process
during which sericin is totally removed and silk fibres
gain the typical shiny aspect, soft handle, and elegant
drape highly appreciated by the consumers. The indus-
trial process takes advantage of the different chemical
and physical properties of the two silk components, fi-
broin and sericin. While the former is water-insoluble
owing to its highly oriented and crystalline fibrous
structure, the latter is readily solubilized by boiling
aqueous solutions containing soap, alkali, synthetic
detergents, or organic acids (Svilokos Bianchi and
Colonna, 1992; Freddi et al., 1996). Nowadays, batch
degumming of silk is mostly carried out in alkaline
baths containing soap and alkali. Soap is replaced
by synthetic detergents in continuous degumming
systems, because it cannot compensate the acidity
of sericin hydrolysis products accumulating in the
bath, thus limiting the use of the degumming bath for
weekly degumming cycles. The mechanism of sericin
removal in chemical degumming is a combination
of various effects such as: dispersion/solubilization
and hydrolysis of the different sericin polypeptides
(Freddi et al., 1996). Hydrolysis prevails when strong
alkaline compounds are added to the degumming bath.
Therefore, suitable procedures for controlling process
parameters, such as temperature, time, pH, and alka-
linity must be implemented on an industrial scale in
order to attain effective sericin removal without trig-
gering the hydrolytic degradation of fibres, which can
be easily induced by the presence of harsh chemicals
in the treatment bath. Fibre degradation often appears
as loss of aesthetic and physical properties, such
as dull appearance, surface fibrillation, poor handle,
drop of tensile strength, as well as uneven dyestuff
absorption during subsequent dyeing and printing.
In recent years, various studies have dealt with the
removal of sericin by using proteolytic enzymes. Sev-
eral acidic, neutral, and alkaline proteases have been
used on silk yarn as degumming agents. Alkaline pro-
teases performed better than acidic and neutral ones in
terms of complete and uniform sericin removal, reten-
tion of tensile properties, and improvement of surface
smoothness, handle, and lustre of silk (Gulrajani et al.,
1996, 1998, 2000b). The combination of a lipase and
a protease resulted in effective de-waxing and degum-
ming, with positive effects on wettability of silk fibres
(Gulrajani et al., 2000a). Enzyme degummed silk fab-
ric displayed a higher degree of surface whiteness, but
higher shear and bending rigidity, lower fullness, and
softness of handle than soap and alkali degummed
fabric, owing to residual sericin remaining at the cross
over points between warp and weft yarns (Chopra
et al., 1996). To overcome these drawbacks and to
enhance the effectiveness of the enzymatic process,
the application of an ultrasonic field to an enzymatic
degumming bath has been proposed (Krasowski et al.,
1999). However, the lower performance of enzyme
degummed silk fabrics in terms of handle, as well as
the higher cost of enzymes compared to chemicals
and some concerns about the use of enzymes in con-
tinuous degumming plants, have so far limited the
development of industrial processes alternative to the
traditional degumming with chemical agents.
The increasing awareness of legislators and citizens
for the ecological sustainability of industrial pro-
cesses has recently stimulated the interest of scientists
and technologists for the application of biotechnol-
ogy to textile processing (Duran and Duran, 2000;
Cavaco-Paulo, 1998; Gübitz and Cavaco-Paulo, 2001).
Silk degumming is a high resource consuming process
as far as water and energy are concerned. Moreover, it
is ecologically questionable for the high environmen-
tal impact of effluents. The development of an effective
degumming process based on enzymes as active agents
would entail savings in terms of water, energy, chemi-
cals, and effluent treatment. This could be made possi-
ble by the milder treatment conditions, the recycling of
processing water, the recovery of valuable by-products
such as sericin peptides, and the lower environmental
impact of effluents. The present study focuses on the
application of proteolytic enzymes characterized by
high stability and efficiency to silk degumming. The
aim is to study the kinetics and mechanism of sericin
removal and to set up an experimental model that may
assist in developing enzyme-based degumming pro-
 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112 103

cess competitive with the traditional one in terms of
effectiveness, costs, and quality of the final product.
2. Materials and methods
2.1. Silk fabric
The treatments were carried out on a 100% raw silk
fabric (crêpe). Construction parameters are listed in
Table 1. Before degumming, the fabric was extracted
with a mixture of methanol and toluene (75/25 vol.%)
to remove sizing agents.
2.2. Proteolytic enzymes
AAApNA), resulting in an increase in absorbance at
405 nm. The activity was expressed in MPU and GSU
for 3374-L and GC 897-H, respectively.
The assay for protease 3273-C is based on the ability
of the enzyme to hydrolyze casein as a substrate at pH
6 and 40 ◦ C. FCCPU (FCC Papain Unit) is defined as
the quantity of enzyme which liberates the equivalent
of 1 ␮g of tyrosine per hour under the conditions of
the assay.
Enzymatic assay for protease EC 3.4.23.6 was made
by using hemoglobin as a substrate at pH 2.8 and
37 ◦ C. One unit (U/mg) is defined as the quantity of
enzyme which hydrolyzes hemoglobin to produce a
color equivalent to 1 ␮mol of tyrosine per minute.

The commercial proteases used for enzymatic 2.4. Silk degumming

degumming of the silk fabric are listed in Table 2.
Standard degumming was carried out in an alka-
They represent different protease types generally used
line solution containing 10 g/l Marseille soap and
in industrial laundry and food applications. 3374-L,
1 g/l sodium carbonate, at 98 ◦ C, for 1 h. Degummed
GC 897-H, and 3273-C were kindly provided by
silk was thoroughly rinsed with warm distilled water,
Genencor Inc., USA. EC 3.4 23.18 was purchased
dried at room temperature, and then extracted with
from Sigma–Aldrich.
petroleum ether to remove residual fatty acids.
The experimental conditions used for enzymatic
2.3. Determination of enzymatic activity
degumming are listed in Table 3. Silk fabric samples
of about 0.1 g were immersed in 20 ml of buffer solu-
The assay for proteases 3374-L and GC 897-H is
tion (material-to-liquor ratio 1:200) containing differ-
based on the ability of the enzyme to cleave a synthetic
ent amounts of enzyme. Blank samples were obtained
peptide, N-succinyl-ala-ala-ala-p-nitroanilide (succ-
by treating silk with buffer alone, without enzyme. Op-
timum pH and temperature for each enzyme were used
Table 1 throughout the tests. Enzyme dosage (0.05–2 Units/g
Construction properties of the silk fabric (crêpe) fabric) and treatment time (5–240 min) were changed.
Warp yarn Weft yarn Degumming tests were carried out in a thermostatic
Number of ends (cm−1 ) 120 40
bath under gentle shaking. Inactivation of proteases
Counts (den) 29.2 23.3 × 3 was made at 85 ◦ C for 10 min. At the end of the treat-
Torsions (m−1 ) 2S (2400) 2Z (2500) ment, silk fabrics were rinsed with distilled water and
Fabric weight (gm−2 ) 79.6 dried at room temperature. All degumming tests were
performed in duplicate.
Table 2
Proteolytic enzymes used for silk degumming
Enzyme code Origin Characteristics pHa T (◦ C)a Activitya

3374-L (A) Bacillus subtilis Oxidative-stable endopeptidase 7.5–12 (10) 20–60 (60) 55.9 MPU/g
(genetically modified)
GC 897-H (B) Bacillus lentus Bacterial high alkaline strain 7–12 (10) 40–65 (65) 44.7 GSU/ml
(genetically modified)
3273-C (C) Carica papaya Papain, thiol protease 3.5–9 (5–7) 65–78 (65) 58.349 FCCPU/g
EC 3.4 23.18 (D) Aspergillus saitoi Aspergillus pepsin I 2.5–6.5 (2.5–3) 30–60 (50) 1 U/mg
a Optimum pH and temperature values in parentheses.
104 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112

Table 3
Experimental conditions used for enzymatic degumming of silk
Enzyme code Buffer pH T (◦ C) Enzyme concentration (Units/g fabric) Time (min)

3374-L Tris–HCl 0.1 M 10 60 0.05–2 5–240

GC 897-H Tris–HCl 0.1 M 10 65 0.05–2 5–240
3273-C Citric acid–Na phosphate, 0.1 M 6 65 0.05–2 5–240
EC 3.4 23.18 Citric acid–Na phosphate 0.2 M 3 50 0.05–60 5–240

2.5. Degumming loss Ltd.). Samples were observed at 10 kV acceleration

Degumming loss, which represents a quantitative
evaluation of the degumming efficiency, indicates the
weight loss of the fabric (expressed as a percentage
of the initial weight) after standard or enzymatic
degumming. Before weight measurement, samples
were conditioned at 20 ◦ C and 65% relative humidity
for 24 h.
2.6. Scanning electronic microscopy (SEM)
Morphological characterization of silk fabrics was
performed by means of scanning electron microscopy
(SEM) (Stereoscan 440, LEO Electron Microscopy
voltage, after gold sputtering.
2.7. High performance-size exclusion
chromatography (HP-SEC) of sericin peptides
HP-SEC runs were performed with a Waters Cor-
porate (USA) chromatographic system controlled by
a Maxima 820 Workstation, which included a GPC
data handling module (Freddi et al., 2000). At the end
of degumming, the silk fabric was removed and the
solution containing sericin peptides was pooled with
washing waters and immediately freeze-dried. For HP-
SEC analysis, freeze-dried samples were dissolved in
a fixed volume of 50 mM sodium phosphate buffer, pH

Fig. 1. Degumming kinetics of the crêpe silk fabric at different enzyme dosage (0.05–2 U/g fabric). 3374-L and GC 897-H: alkaline
proteases. 3273-C: neutral protease. Time: 1 h. Other parameters as in Table 3. The dotted line shows the level of degumming loss obtained
with soap and alkali (27 wt.%). Results are the average of duplicate tests.
 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112 105

7.2, containing 0.15 M KCl, filtered, and analyzed with
a protein Pak-60 column (Waters Corporate, USA).
Injection volumes ranged from 50 to 100 ␮l, flow rate
was 0.5 ml/min. Eluate was detected at 254 nm. Molec-
ular weight calibration was performed by using the
LMW gel filtration calibration kit (Amersham Bio-
sciences, Sweden).
3. Results and discussion
3.1. Study of the degumming kinetics: effect of
enzyme dosage and time
The effect of enzyme dosage on the extent of sericin
removal was studied by treating silk fabric samples for

30

Inactivation
3374-L
Degumming loss (%)

20

10

0
5 10 30 60 120 180 240
(a) Time (min)

30

Inactivation
GC897-H
Degumming loss (%)

20

10

0
5 10 30 60 120 180 240
(b) Time (min)

Fig. 2. Time dependence of the enzymatic degumming of the crêpe silk fabric (5–240 min). Enzyme dosage: 3374-L 2 U/g fabric; GC 897-H
1 U/g fabric; 3273-C 0.1 U/g fabric. Other parameters as in Table 3. Dark + light bars: total degumming loss. Light bars: contribution of
the inactivation step to the degumming loss.
106 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112
30
3273-C
Degumming loss (%)
20
10
0
5 10
(C)
1 h with different amounts of proteases. Fig. 1 shows
the results obtained with the alkaline proteases A and
B, and with the neutral protease C, (Table 2 for en-
zyme codes), whose amount ranged from 0.05 to 2 U/g
fabric. It is worth noting that the acid protease D dis-
played a very low degumming efficiency (7.6 wt.%
degumming loss after 3 h treatment with 60 U/g fab-
ric). For these reasons, this protease was not included
in Fig. 1 and was not used for further tests. The dotted
line drawn in the graph at about 27 wt.% degumming
loss indicates the target value for complete degum-
ming, as obtained under standard conditions with soap
and alkali. Without enzymes, the degumming loss was
negligible (2–3 wt.%), owing to the low treatment tem-
perature. In fact, it is well known that sericin can be
removed by using water alone, but high temperature
is needed to attain complete degumming (110–120 ◦ C,
under pressure).
Degumming loss increased linearly as the amount
of protease B increased until 2 U/g fabric, attaining
a value of 24 wt.%. On the other hand, proteases
A and C reached a plateau at 1 and 0.1 U/g fabric,
corresponding to 17.6 and 19 wt.% degumming loss,
respectively. The values of maximum turnover at sat-
uration, expressed as degumming loss, were 0.29 and
30 60 120 180 240
Time (min)
Fig. 2. (Continued ).
3.22 wt.% for protease A and C, respectively. Enzyme
concentration at half maximum turnover was 0.15 U/g
fabric for the alkaline protease A and 0.04 U/g fabric
for the neutral protease B. These preliminary values
are of chief interest in view of characterizing the ac-
tivity of different proteases. However, a deeper kinetic
investigation of the proteolytic degradation of sericin
is needed because the enzyme-substrate system under
consideration is quite complicated. In fact, as it will
be discussed later in more detail, several factors may
significantly affect the protease activity, such as fabric
texture, accessibility of the cleavage sites, and treat-
ment conditions (agitation). Moreover, competition
of already solubilized sericin peptides for the active
sites of the enzyme cannot be excluded.
To study the effect of the treatment time on the
extent of sericin removal, the amount of enzyme was
kept constant, while the time was changed in the
range 5–240 min. The enzyme dosage was 2, 1, and
0.1 U/g fabric for proteases A, B, and C, respectively.
These values were chosen on the basis of the results
reported in Fig. 1, because they resulted in almost
similar levels of degumming loss (18–19 wt.%). Fig. 2
shows the time dependence of sericin removal for the
three proteases. The total bar height corresponds to
 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112 107

the degumming loss obtained in a complete degum-
ming cycle, in which the enzymatic treatment was
followed by the inactivation step at 85 ◦ C for 10 min.
The amount of sericin removed reached 15–20 wt.%
after 5–10 min, and then tended to increase further as
the treatment time increased. Alkaline proteases per-
formed better than the neutral one. In particular, pro-
tease A attained a degumming loss of about 25 wt.%,
very close to the target value. It is worth noting the
significant contribution of the inactivation step to the
final value of the degumming loss, especially at short
treatment times. This behavior can be attributed to
various factors. The temperature rise from 60–65 ◦ C
to 85 ◦ C probably contributed to enhance the solu-
bility of partially hydrolyzed sericin fractions still
adhering to the fibrous core of the silk. Moreover,
the extension of the treatment time probably exposed
sericin to the action of a residual enzymatic activity.
None of the proteases used in this study allowed
the attainment of the target degumming loss (27 wt.%)

Fig. 3. Crêpe silk fabric degummed under standard (a) and enzymatic (b) conditions (GC 897-H, 2 U/g fabric, 60 min).
108 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112
under the experimental conditions adopted. However,
this result is only apparently negative, because the
crêpe fabric is one of the most difficult substrates to
treat, owing to the presence of highly twisted weft
yarns. In fact, higher amounts of alkalis are needed to
achieve complete sericin removal also during standard
chemical degumming. It is likely that other factors be-
side the lack of hydrolytic power of proteases have
played a role in lowering the extent of degumming.
The morphological features of the samples which got
very close to the target degumming loss seem to con-
firm this hypothesis.
3.2. Morphological characterization
The surface morphology of the silk fabric after stan-
dard and enzymatic degumming is shown in Fig. 3.
Removal of sericin resulted in the separation of the
individual silk filaments, which were glued together
by sericin and sizing agents in the raw fabric. The
Fig. 4. Details of Fig. 3b. (a) Warp yarn. (b–d) Weft yarn.
volume of the yarns increased, especially for the un-
twisted warp yarn, while the weft yarn shrank due to
the high number of twists. This conferred on the fab-
ric the denser texture and the rough surface typical of
light weight crêpe silk fabrics. The dull appearance
and stiff handle of the raw fabric disappeared, and the
degummed fabric became shiny, soft, and scroopy.
By comparing the fabrics degummed with the stan-
dard and enzymatic methods, it is possible to observe
that the latter exhibited a flatter surface, mostly at-
tributable to incomplete shrinkage of the weft yarns,
which appeared straighter and thinner than in the ref-
erence fabric. The closer SEM examination of the
warp yarn showed that the individual silk filaments
split off and their surface was clean and free of sericin
(Fig. 4a). On the other hand, weft yarns still showed
the presence of sericin residues (Fig. 4b–d). It is in-
teresting to note that the deposits were mainly located
at the cross over points between warp and weft yarns
and/or in the more internal parts of the weft yarns.
 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112 109

These sericin residues still sticking the silk filaments
together hampered the weft yarns from reaching the
expected degree of shrinkage.
The morphological observations confirm that pro-
teases failed to completely remove sericin from the
highly twisted weft yarns. Since the ability of these en-
zymes to hydrolyze sericin is unquestionable, the ob-
served behavior can be attributed to other factors, such
as the lack of an effective mechanical agitation in the
lab-scale degumming system used for this study. This
probably limited penetration of enzymes into the close
texture of weft yarns. The influence of the mechanical
agitation on the enzymatic treatment of cellulose-
based textiles has been stressed by various authors
(Cavaco-Paulo et al., 1996; Morgado et al., 2000;
Radhakrishnaiah et al., 1999; Traore and Buschle-
Diller, 1999). It has been shown that not only the con-
tact of the enzyme with the substrate was drastically
improved by applying suitable mechanical agitation,
but also the kinetics and mechanism of the enzyme
action was influenced. Moreover, the mechanical agi-
tation during enzyme treatment has proved to change
the aesthetic, tactile, thermal, and comfort properties
of the fabric. The surface of silk fabrics is very del-
icate and sensitive to mechanical stresses during wet
treatments. These may result in severe faults known
as chafe marks, that is, surface fibrillation at creases,
which impart a typical dull appearance on the fabric
and irremediably impair its final quality. Therefore,
the degumming system must be optimized with the
implementation of a suitable bath and/or material
movement that assists the hydrolytic action of the en-
zyme and allows to obtain completely degummed silk
fabrics without causing significant damages to the

15.8

12.9
18.0

11.3
A 254 nm (a.u.)

7.7 5.9 4.9

3273-C

3374-L
9.5

GC897-H

Buffer

15 20 25
Time (min)

Fig. 5. HP-SEC profiles of soluble sericin peptides obtained by enzymatic degumming with 3374-L, GC 897-H, and 3273-C, for 1 h.
Enzyme dosage: 2 U/g fabric. Other parameters as in Table 3.
110 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112

20

3374-L
GC897-H
18 3273-C
Average MW (kDa)

16

14

12

10
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2

(a) Enzyme (units/g fabric)

20

3374-L
GC897-H
18 3273-C
Average MW (kDa)

16

14

12

10
0 50 100 150 200 250

(b) Time (min)

Fig. 6. Enzyme dosage (a) and time (b) dependence of the weight-average molecular weight of soluble sericin peptides obtained by
enzymatic degumming.
 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112
textile goods. Lab-scale degumming tests focusing on
these aspects are in progress.
3.3. HP-SEC screening of sericin peptides
Fig. 5 shows the HP-SEC curves of sericin peptides
obtained by degumming the silk fabric with alkaline
and neutral proteases. A common feature of the three
chromatographic profiles is that the whole peaks fall
in the low molecular weight range, from about 5 to
20 kDa. It is worth noting that native sericin proteins
cover a much wider molecular weight range, from 40
to 400 kDa (Garel et al., 1997). The narrower molecu-
lar weight distribution of sericin by-products is mostly
dependent on the specific mechanism of action of pro-
teases, that is, the selective cleavage of target peptide
bonds along the protein chains. This resulted in main
chain fission and formation of a range of small size
soluble peptides.
The two alkaline proteases A and B displayed a
rather similar chromatographic pattern of the hy-
drolyzed sericin peptides, with a major peak centered
at about 12.9 kDa. The neutral protease C resulted in
two main peaks at 18.0 and 15.8 kDa. This feature
can be attributed to the different substrate specificity
of the proteases, that is, the chemical structure of
the target cleavage site. Bacterial proteases are usu-
ally characterized by a low degree of specificity, and
this may explain the similarity in the pattern of the
resulting sericin peptides. The neutral protease (pa-
pain) hydrolyzes the “X–Y” peptide bonds where X
is arginine, lysine, or phenylalanine (Mathews and
van Holde, 1994), and therefore resulted in a different
pattern of soluble sericin peptides.
To further characterize the properties of the sericin
peptides produced by the proteases used for silk
degumming, the following quantitative parameters
of the molecular weight distribution were calculated
from the raw chromatographic data: weight-average
(M̄w ) and number-average (M̄n ) molecular weights,
which provide a measure of the average chain weight
and length; peak molecular weight (Pmw ), which gives
the weight of the most abundant polypeptide frac-
tion in the sample; polydispersity index (M̄w /M̄n ),
which indicates the breadth of the molecular weight
distribution. As shown in Table 4, the average size of
the sericin peptides (M̄w ) was in the following order:
A > B > C. However, the difference among the sam-
111
Table 4
Molecular weight distribution parameters of sericin by-products
Samplea M̄w (kDa) M̄n (kDa) Pmw (kDa) M̄w /M̄n
3374-L 12.9 11.8 12.9 1.092
GC 897-H 11.9 11.1 12.9 1.078
3273-C 11.4 10.5 15.8 1.087
a Reaction conditions: 2 U/g fabric, 1 h.
ples was very low, despite the above noticed changes
in peak distribution and intensity. Accordingly, the
values of the polydispersity indexes were similar.
Fig. 6 shows the behavior of weight-average molec-
ular weight as a function of the enzyme dosage and
the treatment time. The plot of M̄w versus enzyme
dosage (Fig. 6a) shows that the size of sericin peptides
decreased sharply with increasing the amount of the
proteases A, B, and C until 0.5 U/g fabric, and then
remained constant. The plot M̄w versus treatment time
(Fig. 6b) indicates that the size of the soluble sericin
peptides remained essentially unchanged regardless of
the degumming time. These results suggest that the
soluble sericin peptides tended to reach a minimum
size, which was probably dependent on the number of
sites available for enzymatic cleavage, and that the ex-
tent of the proteolytic attack towards the peptide frac-
tions in solution was negligible.
4. Conclusions
Alkaline and neutral proteases effectively degum-
med silk fabrics. Almost complete sericin removal was
obtained by using a crêpe silk fabric as the substrate.
Owing to the presence of the highly twisted weft yarns,
this fabric is one of the most difficult substrates to
degum, even by using the standard chemical degum-
ming method.
Hydrolytic degradation of sericin took place by se-
lective peptide bond cleavage. Degumming kinetics
were dependent on enzyme dosage and treatment time.
Moreover, chemical properties of soluble sericin pep-
tides changed as a function of the kind of enzyme used.
A mixture of peptides ranging from 5 to 20 kDa, with
a weight-average molecular weight of about 12 kDa,
was obtained with the alkaline and neutral proteases
used in this study. Recovery from waste water and
reuse of these peptides as additives for cosmetic prod-
112 G. Freddi et al. / Journal of Biotechnology 106 (2003) 101–112
ucts appears feasible, due to the lack of contaminants
such as alkalis and detergents.
Further optimization of the degumming conditions
is needed, with particular reference to the study of
the mechanical agitation effect on the degumming
efficiency of proteases. This parameter is likely to
enhance enzyme penetration within the closest parts
of the fabric texture and to make sericin removal
more effective. As a consequence of complete sericin
removal, the quality of silk goods in terms of han-
dle, appearance and tensile properties is expected to
increase, due to the lower extent of chemical and
physical stresses to which silk is subjected to during
enzymatic processing, as compared to the traditional
chemical process. To this purpose, tests on pilot plants
are in progress, with the aim to scale-up the process
for its implementation on industrial equipments.
Acknowledgements
Authors wish to express their thanks to Dr. Mee
Young Yoon of Genencor Inc. (USA) for providing
alkaline (3374-L, GC 897-H), and neutral (3273-C)
proteases, for the determination of enzymatic activi-
ties, as well as for scientific and technical information
and support.
References
Cavaco-Paulo, A., 1998. Processing textile fibres with enzymes:
an overview. In: Eriksson, K.E.L., Cavaco-Paulo, A. (Eds.),
Enzyme Applications in Fibre Processing, ACS Symposium
Series 687, American Chemical Society, Washington, DC,
pp. 180–189.
Cavaco-Paulo, A., Almeida, L., Bishop, D., 1996. Effects of
agitation and endoglucanase pre-treatment on the hydrolysis
of cotton fabrics by a total cellulose. Textile Res. J. 66, 287–
294.
Chopra, S., Chattopadhyay, R., Gulrajani, M.L., 1996. Low stress
mechanical properties of silk fabric degummed by different
methods. J. Textile Inst. 87, 542–553.
Couble, P., Michaille, J.-J., Garel, A., Couble, M.-L., Prudhomme,
J.-C., 1987. Developmental switches of sericin mRNA splicing
in individual cells of Bombyx mori silkgland. Dev. Biol. 124,
431–440.
Duran, N., Duran, M., 2000. Enzyme applications in the textile
industry. Rev. Prog. Coloration 30, 41–44.
Freddi, G., Allara, G., Candiani, G., 1996. Degumming of silk
with tartaric acid. JSDC 112, 191–195.
Freddi, G., Berlin, A., Tsukada, M., Dubini Paglia, E., 2000. Use
of HP-size exclusion chromatography to study the degree of
polymerization of silk (Bombyx mori) fibroin fibres. Sericologia
40, 363–373.
Gamo, T., Inokuchi, T., Laufer, H., 1977. Polypeptides of
fibroin and sericin secreted from the different sections of
the silk gland in Bombyx mori. Insect Biochem. 7, 285–
295.
Garel, A., Deleage, G., Prudhomme, J.-C., 1997. Structure and
organization of the Bombyx mori sericin 1 gene and of the
sericins 1 deduced from the sequence of the Ser 1B cDNA.
Insect Biochem. Mol. Biol. 27, 469–477.
Gübitz, G.M., Cavaco-Paulo, A., 2001. Editorial: Biotechnology
in the textile industry – perspectives for the new millennium.
J. Biotechnol. 89, 89–90.
Gulrajani, M.L., Gupta, S.V., Gupta, A., Suri, M., 1996.
Degumming of silk with different protease enzymes. Indian J.
Fibre Textile Res. 21, 270–275.
Gulrajani, M.L., Sen, S., Soria, A., Suri, M., 1998. Efficacy of
proteases on degumming of dupion silk. Indian J. Fibre Textile
Res. 23, 52–58.
Gulrajani, M.L., Agarwal, R., Grover, A., Suri, M., 2000a.
Degumming of silk with lipase and protease. Indian J. Fibre
Textile Res. 25, 69–74.
Gulrajani, M.L., Agarwal, R., Chand, S., 2000b. Degumming of
silk with fungal protease. Indian J. Fibre Textile Res. 25, 138–
142.
Krasowski, A., Müller, B., Föhles, J., Höcker, H., 1999.
Degumming of silk in an ultrasonic field. Melliand
Textilberichte 80, 543–545.
Lotz, B., Colonna Cesari, F., 1979. The chemical structure and the
crystalline structures of Bombyx mori silk fibroin. Biochimie
61, 205–214.
Mathews, C.K., van Holde, K.E., 1994. Biochemistry, C.E.A.,
Milano, p. 146.
Morgado, J., Cavaco-Paulo, A., Rousselle, M.-L., 2000. Enzymatic
treatment of lyocell: clarification of depilling mechanisms.
Textile Res. J. 70, 696–699.
Radhakrishnaiah, P., Meng, X., Huang, G., Buschle-Diller, G.,
Walsh, W.K., 1999. Mechanical agitation of cotton fabrics
during enzyme treatment and its effect on tactile properties.
Textile Res. J. 69, 708–713.
Svilokos Bianchi, A., Colonna, G.M., 1992. Developments in the
degumming of silk. Melliand Textilberichte 73, 68–75.
Takahashi, Y., Gehoh, M., Yuzuriha, K., 1991. Crystal structure of
silk (Bombyx mori). J. Polym. Sci., Part B:Polym. Phys. 29,
889–891.
Traore, M.K., Buschle-Diller, G., 1999. Influence of wetting
agents and agitation on enzymatic hydrolysis of cotton. Textile
Chemist and Colourist and American Dyestuff Reporter 1, 51–
56.
Zahn, H., 1993. Silk. Ullmann’s Encyclopedia of Industrial
Chemistry, VCH Publisher Inc., vol. A24, pp. 95–106.
Zhou, C.Z., Confalonieri, F., Medina, N., Zivanovic, Y., Esnault,
C., Yang, T., Jacquet, M., Janin, J., Duguet, M., Perasso, R., Li,
Z.G., 2000. Fine organization of Bombyx mori fibroin heavy
chain gene. Nucleic Acids Res. 28, 2413–2419.