Bioresource Technology 98 (2007) 53–57

Extraction and characterisation of gelatine

from Atlantic salmon (Salmo salar) skin
Jan Arne Arnesen, Asbjørn Gildberg *

Norwegian Institute of Fisheries and Aquaculture Ltd., P.O. Box 6122, 9291 Tromsø, Norway

Received 24 May 2005; received in revised form 23 September 2005; accepted 23 November 2005
Available online 10 January 2006

Abstract

Gelatine was extracted from Atlantic salmon and Atlantic cod skin by the acid extraction process. After filtration and ion exchange
treatment the extracts were colourless and free from fishy odour. In three separate experiments the average yields of gelatine from salmon
and cod skins were 39.7% (±2.2%) and 44.8% (±0.2%) respectively, on a dry matter basis. Gelatine from salmon contained slightly more
hydroxyproline and proline (16.6%) than cod gelatine (15.4%), whereas the content of serine was lower (4.6% versus 6.3%). Salmon gel-
atine expressed slightly higher gelling temperature (12 C) than cod gelatine (10 C), and higher initial gel strength. During storage at
10 C, gel strengths were increased and more so with gels made from cod than from salmon gelatine. Hence, gels made from cod and
salmon gelatines extracted at 56 C achieved the same gel strength (195 g) after 7 days of storage. Gelatines extracted at a higher tem-
perature (65 C) gave lower gel strengths.
 2005 Elsevier Ltd. All rights reserved.

Keywords: Fish gelatine; Atlantic salmon; Skin; Gelatine; Gel strength; Viscosity

1. Introduction
Fifty years ago there was still a considerable production
of fish glue from fish skin and bones. Fish glue is a crude
gelatine product only suitable for technical application.
At present, the fish gelatine production is minor, yielding
about 1% of the annual world gelatine production of
250,000 tons. Factors such as the outbreak of BSE (‘‘mad
cow disease’’) and increasing demand for non-mammalian
gelatine for halal and kosher food markets have revived
interest in gelatine produced from fish raw materials.
Recently, several studies of gelatines from the skin of var-
ious fish species have been published (Gudmundsson, 2002;
Haug et al., 2004; Kolodziejska et al., 2004; Muyonga
et al., 2004; Sadowska et al., 2003; Simon et al., 2002;
Terao et al., 2003).
It is well established that many physical properties and,
particularly, the temperature characteristics of collagen
*
Corresponding author. Tel.: +47 776 29000; fax: +47 776 29100.
E-mail address: asbjorn.gildberg@fiskeriforskning.no (A. Gildberg).
and gelatine from fish skin reflect the fish habitat tempera-
ture (Andreeva, 1971; Gustavson, 1953). The content of
imino acids (proline and hydroxyproline) is of particular
importance regarding both gelatine gel strength and melt-
ing point. Due to the rigidity of their R groups the imino
acids provide rigidity to triple helix structures both in
intact collagen and gelatine gels (Babel, 1996; Haug
et al., 2004). Apparently, high contents of hydrophobic
amino acids have a similar effect, although less prominent
(Badii and Howell, 2005). A high content of hydroxylated
amino acids, such as serine and threonine, may promote
regeneration of triple helix structures from random coil,
by creating regular water structures both close to the mac-
romolecules and by local formation of hydrogen bonds
within and between individual helices (Babel, 1996; Priva-
lov et al., 1971).
Gelatine with low molecular weight distribution has
inferior gelling properties compared to high molecular
weight gelatine (Badii and Howell, 2005). Like every pro-
tein, gelatine is more susceptible to chemical hydrolysis at
increasing acidity and temperature. Hence, the extraction

0960-8524/$ - see front matter  2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2005.11.021
54 J.A. Arnesen, A. Gildberg / Bioresource Technology 98 (2007) 53–57
conditions applied greatly influence the gelling properties
of the gelatine.
Although collagen from cold water fish species normally
contains less imino acids than collagen from warm water
fish species and mammalians, the contents of both hydro-
phobic and hydroxylated amino acids, as well as other
properties such as molecular weight distribution and gela-
tine viscosity, seem to be species specific (Gómez-Guillén
et al., 2002; Gudmundsson, 2002; Johnston-Banks, 1990).
To reveal such properties, gelatine from each actual species
must be studied.
In Northern Europe farmed Atlantic salmon has
become a significant fish resource. The annual production
in Norway alone amounts to about 400,000 tons. Until
now salmon skin has not been available in large quantities,
since most of the fish have been exported as skin-on
products. At present, a market for more processed salmon
products without skin is being developed, and the skin,
comprising about 5% of the whole fish, has become an
interesting raw material for gelatine production. Although
salmon collagen has been studied briefly (Eckhoff et al.,
1998; Simpson et al., 1999; Yunoki et al., 2003), little is
known about salmon gelatine.
The aim of the present work was to study extraction
yield and some chemical and physical characteristics of gel-
atine from the skin of Atlantic salmon. Gelatines from pig
and Atlantic cod skin, which have been thoroughly studied
previously, were used as references.
2. Methods
2.1. Chemicals and raw materials
All chemicals used in this study were of analytical grade.
DGF Stoess Eberbach Germany provided porcine gelatine
type A (extracted by the acid process). Salmon skin was
provided by Fjord Aqua Group AS, Kristiansund,
Norway. The skins used for gelatine extraction were
obtained by deep skinning of farmed Atlantic salmon
(Salmo salar) (3–5 kg). The skin was frozen ( 20 C)
before and during transport to Tromsø. The frozen skin
was cut into pieces (approximately 10 · 10 cm) with a meat
saw. After thawing overnight at 4 C, residual meat and fat
was removed manually with a filleting knife. Atlantic cod
skin was provided by a local fish processing plant and
stored frozen ( 20 C) until further use.
2.2. Gelatine preparation
Gelatine was prepared by the acid extraction method,
essentially as given by Gudmundsson and Hafsteinsson
(1997). One kilogram of skin was stirred gently for
30 min in 3.3 l volumes of various solutions. The skin
was first washed in cold water (about 8 C ) and then incu-
bated twice in cold sodium hydroxide solutions (0.04 N) for
30 min each time. After washing out NaOH with cold
water, two successive acid incubations were performed,
each for 30 min, first in a sulphuric (0.12 M) and then in
a citric acid solution (0.005 M). The acid solutions were
drained and then the samples were washed with cold water.
After the final washing in cold water, a two-step gelatine
extraction (each step 2 h) was performed by gentle stirring,
first in 1 l water at 56 C and then in 1 l at 65 C. Solubi-
lized gelatine was separated from residual skin fragments
by filtration through a 250 lm mesh nylon filter. Gelatine
was further filtered through a Whatman No. 4 and GF/C
filter. The gelatine solution was demineralized by adding
Purolite cation- and anion exchange resins. After 16 h the
resins were removed by filtration through a 250 mesh nylon
filter. The gelatine extract was then concentrated in a rota-
vapor (40 C, 30 min) before drying as a thin film in plastic
trays at 50 C in a heated cabinet until the water content
was below 6% (w/w).
2.3. Chemical and physical analyses
Molecular weight distribution in gelatine extracts was
estimated by gel filtration chromatography on a HPLC
instrument as described by Arnesen and Gildberg (2002).
A TSKgel 4000 SWXL column (Tosohaas) equilibrated
with a buffer (pH 5.3) containing 0.01 M NaH2PO4,
0.1 M Na2SO4 and 1% SDS was used, and the flow rate
was 0.5 ml/min. Compounds leaving the column were
detected by measuring optical density at 227 nm and data
were analyzed by EZChrom (Scientific Software). Viscosity
and gel strength were also determined as described pre-
viously (Arnesen and Gildberg, 2002) except that in the
present work a Stable Micro Systems TA.HDi texture
analyser was used for gel strength measurements. Gelatine
solutions (6.67% w/v) were made by dissolving dry powder
in distilled water and heating to 60 C. Viscosity was deter-
mined with a Brookfield RVT viscometer equipped with a
No. 1 spindle at 100 rpm by starting at 60 C and then
reducing temperature (0.25 C/min) until the solutions
were gelling. Gel strength was determined by forcing a
plunger (U = 12.7 mm) 4 mm (1 mm/s) into a 6.67% gel
in a Bloom jar. Amino acid analysis was performed on
hydrolyzed gelatine samples essentially as described by
Pedersen et al. (2003). Gelatine samples were hydrolyzed
for 24 h at 110 C in 6 M HCl with 0.05% phenol under
an atmosphere of nitrogen. Phenylisothiocyanate derivates
of the free amino acids were analyzed by RP-HPLC with a
gradient of acetonitrile in Pico Tag Eluent on a C18 col-
umn (Waters, USA).
3. Results and discussion
3.1. Extraction recovery
Deep skinning leaves a layer of subcutaneous fat on the
skin. This is done to get a better quality of the fillet. The
‘‘deep skinned skin’’ contained approximately 64.3%
(±0.6%) (w/w) of muscle tissues and 35.7% (±0.6%) skin.
This muscle fraction is suitable for production of salmon
 J.A. Arnesen, A. Gildberg / Bioresource Technology 98 (2007) 53–57
oil and fish protein hydrolysate. The remaining salmon
skin contained 36% dry matter. The first extraction of sal-
mon gelatine at 56 C gave 11.3% (±1.4%) and the second
at 65 C gave 4.0% (±0.8%) yield (dry gelatine/wet
skin · 100%). The corresponding yields of cod skin gelatine
were 10.1% (±0.2%) and 1.8% (±0.1%) respectively. How-
ever, since the dry matter content of cod skin was only
25%, the total gelatine yield on a dry matter basis was
slightly lower from salmon skin (39.7% ± 2.2%) than from
cod skin (44.8% ± 0.2%). The higher content of dry matter
measured in salmon- than in cod skin may have been
caused by some water evaporation from the skin during
the manual removal of fat muscle tissues. Reporting gela-
tine yield as dry gelatine weight compared to the weight
of wet skin is common, but not very reliable. Water content
may vary because of different treatment of the skin (freez-
ing, salting, scraping, draining, etc.). Therefore gelatine
yield should rather be reported as the amount of dry gela-
tine compared to the amount of dry matter in skin.
Gelatine recoveries from salmon and cod at 56 C were
74.3% and 84.8% respectively. The difference in gelatine
recovery from the two species can be attributed to differ-
ences in the extent of intermolecular crosslinking in the
collagens. Most probably this is due to a slightly higher
thermal adaptation temperature of the salmon connective
tissues (Andreeva, 1971). Hence, a moderate increase in
extraction temperatures may improve the gelatine yield
from salmon skin to the same level as for cod skin.
During the treatment with ion exchange resins, the pH
in the gelatine solutions increased from 4.4 to 6.4 due to
net removal of anions. After filtration and ion exchange
treatment the gelatine extracts were colourless and free
from fishy odour.
3.2. Molecular weight distribution
HPLC-gel filtration (Fig. 1) showed that gelatines from
both salmon and cod extracted at 65 C were more hydro-
lyzed than the gelatines extracted at 56 C. The major part
Porcine
Relative OD (220 nm)
Salmon 56 °C
Salmon 65 °C
Cod 56 °C
Cod 65 °C
10000 1000 100 10 1 0.1
MW (kDa)
Fig. 1. HPLC gel chromatography of gelatines extracted from the skins of
pig, Atlantic salmon and Atlantic cod. Relative OD at 220 nm is given as a
function of molecular weight.
55
of the peptides in both salmon gelatine extracts apparently
occurred between the corresponding fractions of cod
gelatine. Gelatine from cod extracted at 56 C contained
peptides with molecular weight above 800 kDa whereas sal-
mon gelatine extracted at the same temperature did not.
Whereas the major peak of cod gelatine above 300 kDa
completely disappeared when the extraction temperature
was increased from 56 to 65 C, only minor changes
occurred in the molecular weight pattern of the salmon gel-
atines. This indicates that not only intermolecular, but also
intramolecular crosslinks are more thermostable in salmon
than in cod collagen. Hydrolysis in gelatine is the combined
effect of splitting peptide bonds and intramolecular cross-
links between peptide chains. Since the primary structure
is fairly similar in collagens from various species, the nature
and frequency of intermolecular crosslinks must be of pre-
mium importance for molecular weight composition of gel-
atine preparations (Babel, 1996).
3.3. Gel strength
Fig. 2 shows the Bloom value (gel strength after storage
for 18 h at 10 C) and gel strength as a function of storage
time of the different gelatines. The porcine gelatine had the
highest Bloom value (216 g). Salmon gelatine extracted at
56 C had a higher Bloom value (108 g) than cod gelatine
(71 g) extracted at the same temperature. After 6 days of
further storage, however, the gel strengths of the two sam-
ples were identical (195 g). It is well known that gelatine
gels strengthen during storage. It should be noted, how-
ever, that the relative strengthening of fish gelatines was
found to be much higher than the strengthening of porcine
gelatine. The gel strength of cod gelatine extracted at 65 C
increased by 250% with further storage for 6 days after the
Bloom value was measured, whereas the corresponding
strength increase of porcine gelatine gel was only 23%.
Hence, the measurement of standard Bloom value may give
an incorrect impression of the potential gel strength of fish
gelatines. Gel strengthening during storage is mainly attrib-
uted to regeneration of helical structures between collagen
peptide chains, but is also due to formation of hydrogen
300
250
Gel strength (g)
Porcine
200
Salmon 56 °C
150 Salmon 65 °C
Cod 56 °C
100
Cod 65 °C
50
0
0 2 4 6 8
Time (days)
Fig. 2. Gel strengths of gels made from pig, Atlantic salmon and Atlantic
cod skin gelatines as functions of storage time at 10 C.
56 J.A. Arnesen, A. Gildberg / Bioresource Technology 98 (2007) 53–57

bonds between hydroxylated amino acids and incorporated 100

water molecules (Babel, 1996; Haug et al., 2004). Porcine
90
Salmon 56 °C
3.4. Amino acid composition 80
Salmon 65 °C
70 Cod 56 °C

Viscosity (MPas)
Table 1 shows that gelatine from salmon skin had a 60
1.2% higher molar content of imino acids than gelatine
from cod skin. Apart from this, the only significant differ- 50
ence was that cod gelatine had a higher content of serine. 40
Probably the hydroxyl side chains of serine and other
30
hydroxylated amino acids play an important role in the
generation of hydrogen bonds and helical structures during 20
the storage gel strengthening. The amino acid compositions 10
of gelatines from each species were identical regardless of
0
extraction temperature. Accordingly, the extent and ther- 0 10 20 30 40 50 60 70
mostability of intermolecular crosslinks must be an essen- Temperature (˚C)
tial factor influencing the extractability of gelatine from
various skin tissues at different temperatures. Fig. 3. Viscosities of solutions of gelatine extracted from pig, Atlantic
salmon and Atlantic cod skin as functions of temperature.

3.5. Viscosity

Fig. 3 shows the viscosity of gelatines from salmon, cod 4. Conclusions

and pig. At high temperature, gelatines from cod and sal-
mon extracted at 56 C had higher viscosity than porcine The physical properties of salmon skin gelatine were
gelatine. As the temperature approached the gelling point, quite similar to those of cod skin gelatine extracted under
the viscosity increased dramatically. It is well known that the same conditions. The salmon gelatine had a slightly
porcine gelatine has a much higher gelling point than gela- higher content of imino acids, and this is probably the main
tines from cold water fishes. The gelling point of gelatine reason for the slightly higher gel strength and melting point.
from salmon extracted at 56 C was slightly higher The major chemical difference was the lower content of ser-
(12 C) than that for cod gelatine (10 C) extracted at the ine in salmon gelatine. It is suggested that the high content
same temperature. This corresponds with the higher con- of this hydroxylated amino acid in Atlantic cod gelatine,
tent of imino acids in salmon gelatine (Table 1). compared to all other gelatines investigated (Badii and
Howell, 2005; Arnesen and Gildberg, 2002; Gómez-Guillén
et al., 2002), may promote a slow regeneration of helical
structures resulting in the considerable gel strengthening
Table 1 observed during long term storage at low temperature.
Amino acid composition of gelatine from skin of salmon and cod
extracted at 56 and 65 C
Acknowledgements
Amino acid Salmon Cod
56 C 65 C 56 C 65 C Financial support from The Research Council of Nor-
Aspartic acid 5.4 5.2 5.2 5.2 way is acknowledged.
Glutamic acid 7.4 7.4 7.1 7.3
Hydroxyproline 6.0 6.0 5.6 5.6 References
Serine 4.6 4.6 6.3 6.3
Glycine 36.6 36.4 35.8 36.2
Andreeva, A.P., 1971. Thermostability of cutaneous collagen of some
Histidine 1.3 1.5 1.2 1.2
species and subspecies of the cod family. Tsitologiya 13, 1004–1008.
Arginine 5.3 5.4 5.3 5.3
Arnesen, J.A., Gildberg, A., 2002. Preparation and characterisation of
Threonine 2.3 2.3 2.3 2.3
gelatine from the skin of harp seal (Phoca groendlandica). Bioresource
Alanine 10.4 10.5 10.3 10.2
Tech. 82, 191–194.
Proline 10.6 10.6 9.8 9.7
Babel, W., 1996. Gelatine—a versatile biopolymer. Chemie in unserer
Tyrosine 0.3 0.3 0.5 0.4
Zeit, 06/98, Information from Deutsche Gelatine-Fabriken Stoess AG,
Valine 1.5 1.5 1.7 1.7
10p.
Methionine 1.8 1.8 1.7 1.6
Badii, F., Howell, N.K., 2005. Fish gelatin: Structure, gelling properties
Isoleucine 0.9 1.0 1.1 1.1
and interaction with egg albumen proteins. Food Hydrocoll., in press,
Leucine 1.9 1.9 2.0 2.0
doi:10.1016/j.foodhyd.2005.06.006.
Phenylalanine 1.3 1.3 1.2 1.2
Eckhoff, K.M., Aidos, I., Hemre, G.-I., Lie, Ø., 1998. Collagen content in
Lysine 2.4 2.3 2.7 2.6
farmed Atlantic salmon (Salmo salar) and subsequent changes in
Hyp + Pro 16.6 16.6 15.4 15.3
solubility during storage on ice. Food Chem. 62 (2), 197–200.
Figures show moles per 100 mol of amino acids in hydrolyzed gelatine Gómez-Guillén, M.C., Turnay, J., Fernández-Dı́az, M.D., Ulmo, N.,
samples. Lizarbe, M.A., Montero, P., 2002. Structural and physical properties
 J.A. Arnesen, A. Gildberg / Bioresource Technology 98 (2007) 53–57
of gelatine extracted from different marine species: a comparative
study. Food Hydrocoll. 16, 24–34.
Gudmundsson, M., 2002. Rheological properties of fish gelatins. J. Food
Sci. 67, 2172–2176.
Gudmundsson, M., Hafsteinsson, H., 1997. Gelatin from cod skins as
affected by chemical treatments. J. Food Sci. 52, 37–39.
Gustavson, K.H., 1953. Hydrothermal stability and intermolecular
organization of collagen from mammalian and teleost skins. Svensk
Kem. Tidskr. 65, 70–77.
Haug, I.J., Draget, K.I., Smidsrød, O., 2004. Physical and rheological
properties of fish gelatin compared to mammalian gelatin. Food
Hydrocoll. 18, 203–213.
Johnston-Banks, F.A., 1990. Gelatin. In: Harris, P. (Ed.), Food Gels.
Elsevier Appl. Food Sci. Publ., London, pp. 233–289.
Kolodziejska, I., Kaczorowski, K., Piotrowska, B., Sadowska, M., 2004.
Modification of the properties of gelatine from skins of Baltic cod
(Gadus morhua) with transglutaminase. Food Chem. 86, 203–209.
Muyonga, J.H., Cole, C.G.B., Duodu, K.G., 2004. Extraction and
physico-chemical characterisation of Nile perch (Lates niloticus) skin
and bone gelatine. Food Hydrocoll. 18, 581–592.
Pedersen, G.M., Gildberg, A., Steiro, K., Olsen, R.O., 2003. Histone-like
proteins from Atlantic cod milt: stimulatory effect on Atlantic salmon
leucocytes in vivo and in vitro. Comp. Biochem. Phys. 134B, 407–416.
57
Privalov, P.L., Serdyuk, I.N., Tiktopulo, E.I., 1971. Thermal conforma-
tional transformation of tropocollagen 2. Viscometric and light-
scattering studies. Biopolymers 10, 1777.
Sadowska, M., Kolodziejska, I., Niecikowska, C., 2003. Isolation of
collagen from the skins of Baltic cod (Gadus morhua). Food Chem.
81, 257–262.
Simpson, B.K., Awafo, V., Ramaswamy, H., 1999. Biotechnological
approaches for the production of value added substances from by-
products of seafood harvesting. In: Trends in Food Science and
Technology. Proceedings of IV International Food Convention
(IFCON ’98). Jwalamukhi Job Press, Bangalore, pp. 151–157.
Simon, A., Vandanjon, L., Levesque, G., Bourseau, P., 2002. Concentra-
tion and desalination of fish gelatine by ultrafiltration and continuous
diafiltration processes. Desalination 144, 313–318.
Terao, K., Nagasawa, N., Nishida, H., Furusawa, K., Mori, Y., Yoshii,
F., Dobashi, T., 2003. Reagent-free crosslinking of aqueous gelatine:
manufacture and characteristics of gelatine irradiated with gamma-
ray and electron beam. J. Biomater. Sci. Polym. Edn. 14, 1197–
1208.
Yunoki, S., Suzuki, T., Takai, M., 2003. Stabilization of low denaturation
temperature collagen from fish by physical cross-linking methods.
J. Biosci. Bioeng. 96, 575–577.