ORIGINAL PAPER
Jose´ Roberto Machado Cunha da Silva
Ide`rcio Luis Sinhorini Æ Bernard Ernesto Jensch-Junior
Lae´rcio Ribeiro Porto-Neto
Francisco Javier Hernadez-Blazquez
Bruno Cossermelli Vellutini
Leandro Nogueira Pressinotti
Frederico Azevedo Costa Pinto Æ Edwin Lowell Cooper
Joa˜o Carlos Shimada Borges
Kinetics of induced wound repair at 0
C in the Antarctic fish
(Cabec¸uda) Notothenia coriiceps
Received: 7 October 2003 / Revised: 22 February 2004 / Accepted: 23 February 2004 / Published online: 7 May 2004
 Springer-Verlag 2004
Abstract Notothenia coriiceps were kept at 0±2
C.
Following anaesthesia, a square excision (4 cm2) was
made on the dorsal lateral anterior regions on both sides.
The process of wound healing was monitored after 0, 1, 2,
7, 15, 23, 60 and 90 days. The wounds were processed for
scanning electron microscopy. Following surgery, haem-
orrhage was abundant; after 24–48 h, the wound was
covered by mucous exudes. At 7–15 days, there was an
epidermal migration towards the centre of the wound,
while the borders became oedematous. During
23–30 days, all the wounds were closed like a sphincter
and there was retraction of the borders. After 60–90 days,
the wounds contracted and became black. No scales were
seen in the wounded area. The regenerating epidermis
migration speed was 33.3–43.4 lm/day. This work shows
for the first time the kinetics of wound repair at 0
C.
Introduction
The intensity and velocity of wound healing in ecto-
thermic vertebrates varies according to seasonality and
local temperature (Reddan and Rothstein 1965; Finn
and Nielsen 1971; Grout and Morris 1987; Hardie et al.
1994). Tissue repair always seeks to close the wounded
region by means of tissue regeneration, resulting usually
in a scar, in order to recover the original function
(Majno and Joris 1996).
Despite some studies on the influence of fish accli-
matisation from tropical and temperate waters on the
regenerating process (Mittal and Munshi 1974; Ander-
son and Roberts 1975; Phromsuthirak 1977; Bullock
et al. 1978; Majno and Joris 1996; Quilhac and Sire
1998, 1999), there are no studies on wound repair under
polar temperatures. However, inflammatory processes
(Silva et al. 1998, 1999) and phagocytosis have already
been described at 0
C (Silva and Peck 2000; Silva et al.
2001, 2002; Borges et al. 2002).
Notothenia coriiceps (Richardson 1844), also called
‘‘Cabec¸uda’’, is a shallow-water benthic, primarily
predator species, and has widespread distribution,
probably circum-Antarctic on the continental shelf. This
Antarctic species lives at depth range of 0–550 m, at
temperatures of )1.8 to near 0
C (Gon and Heemstra
1990; Prisco et al. 1991).
During fieldwork, we collected one N. coriiceps with a
conic-shaped scar that suggests a possible seal bite.
Therefore, it appears that the skin is able to regenerate
properly at low temperatures, close to 0
C. But, do these
processes of wound healing fit well with those at higher
temperatures? The present study analyses the kinetics
after experimentally inducing wounds in N. coriiceps at
0
C, using scanning electron microscopy. The data from
tissues and cells involved in the wound repair are being
J. R. M. C. da Silva (&) Æ B. E. Jensch-Junior
L. R. Porto-Neto Æ B. C. Vellutini Æ L. N. Pressinotti
J. C. S. Borges
Department of Histology and Embryology,
Institute of Biomedical Sciences, University of Sa˜o Paulo,
Av. Prof. Lineu Prestes, 1524, sala 409, 04601-000 Sa˜o Paulo,
Brazil
E-mail: jrmcs@usp.br
I. L. Sinhorini Æ F. A. C. Pinto
Department of Pathology, Faculty of Veterinary Medicine,
University of Sa˜o Paulo, Sa˜o Paulo, Brazil
E. L. Cooper
Laboratory of Comparative Neuroimmunology,
Department of Neurobiology,
David Geffen School of Medicine at UCLA,
University of California, Los Angeles, CA
190095-4763, USA
F. J. Hernadez-Blazquez
Department of Surgery, Faculty of Veterinary Medicine,
University of Sa˜o Paulo, Sa˜o Paulo, Brazil
Polar Biol (2004) 27: 458–465
DOI 10.1007/s00300-004-0611-7

processed, and will be published in another work. The
process of wound healing is compared in the available
literature on temperate and tropical teleost fishes.
Materials and methods
Fish
Naturally occurring
One N. coriiceps was found with a wounded scar on the
posterior ventral lateral region. This fish was photo-
graphed and processed for scanning electron microscopy
as described below.
Experimental
Twenty-five N. coriiceps (Richardson 1844) (according
to Gon and Heemstra 1990), weight 523–1.163 g
(843±320 g) and total size 33.2–43.4 cm (38.3±5.1 cm)
were used in this experiment. They were collected in
January-February 2000 (n=14) and from December to
April 2001 (n=11) at Admiralty Bay, King George Is-
land, in the South Shetland Islands (62
06¢S and
58
23¢W). The fish were kept in fibreglass tanks (2000 l)
filled with seawater in an acclimatised room at
0.0±1.0
C and fed twice weekly. The experiments were
carried out in the Biology Laboratories of the Brazilian
Antarctic Station ‘‘Comandante Ferraz’’.
Surgical lesions and anaesthesia
Surgical lesions were made, after anaesthesia with ben-
zocaine 50.0 ppm (Silva et al. 2002), in the dorsal-lateral
anterior region on both sides of experimental fish
(n=22). The scales, epidermis, dermis, and most of the
perimysium were removed on 4.0 cm2 (total of 44 le-
sions). Three more fishes were used as controls on nor-
mal skin in this region.
Gross analysis of wound healing
After time spans of 0 (immediately after the wound), 1,
2, 7, 15, 23, 30, 45, 60 and 90 days, the fishes were killed
by an overdose of benzocaine. For each period, photo-
graphs of the wound were taken (Canon eos-300 with
macro lens-100 mm) on both sides (along with a scale in
millimetres) until their sacrifice, in order to study mac-
roscopically the kinetics of wound healing.
Scanning electron microscopy
The tissue samples from the whole wound area after
15 days (n=2), 30 days (n=3), 60 days (n=4) and
90 days (n=4) were fixed in cold marine McDowell
solution (McDowell and Trump 1976) for 24 h at 0
C,
dehydrated in a graded series of ethanol, dried through
the critical point method with CO2, covered with a gold
layer and examined in an SEM-JEOL (6100) and LEO
(435VP).
Quantification of regenerated tissue
The photographs of the left and right wounded regions
of each fish were scanned using the software SigmaScan.
The wounded area and the regenerated area were mea-
sured (in mm2) and the percentage of the regenerated
area was calculated.
Results
Normal skin
In N. coriiceps, the skin is scaled and the outer epidermal
surface is ornamented with fingerprints. There are also
mucous cells whose pores are identified with scanning
electron micrographs (SEM) (Fig. 1).
Naturally found wound
We noticed the presence of a scar in one of the collected
N. coriiceps. The scar was complete and the centre was
smooth, bright and dark-brown to black. The wound
presented smaller scales on the periphery, and irregular
borders (Fig. 2a). With SEM, the majority of the scar
border cells did not show the typical fingerprint surface
while none of the wound centre epidermal cells had the
fingerprint surface (Fig. 2b, c).
Induced wounds at different time spans
After the lesion
Immediately after surgery, haemorrhage was visible
mainly on the wound borders (but also in the central
Fig. 1 Scanning electron micrograph of N. coriiceps skin surface,
showing the flat epidermal cells with their typical fingerprints
surface pattern
459

area). Small black spots can be observed all over the
surface of the remaining perimysium that covers the
muscular tissues. Figure 3 shows the sequence of wound
healing on the same fish region after the different time
spans.
Aspect of the wounded region at days 1–15
At day 1, the wounded region presented blood clots,
mainly at the wound borders; large amount of mucus
and exude can also be seen. After 2 days, observations
were the same as on day 1, with the beginning of a
discrete oedema on the wound borders (Fig. 3). On
day 7, from the wound borders, a thin layer of
migrating epidermis can be seen over the coagulated
blood. This projection is about 1–3 mm in length all
over the border with some pigmented areas (unpub-
lished data). There is an increase of the wound border
oedema, under the remaining scales (the original
square wound became rounded). The centre of the
wound is similar to day 1 (Fig. 3). On day 15, there is
an increase of the wound border oedema. The regen-
erating epidermis covers almost half of the wound, and
has a length of 4–8 mm; it is possible to see some
pigmented radial black lines, appearing from the
periphery to the central region (Fig. 3), composed of
melanocytes (unpublished data). The epidermis border
has an irregular surface (Fig. 4), with some different
patterns of fingerprints (Fig. 5). The central wound
region not covered by the epidermis (i.e. exposed to
the seawater) presents a disorganised pattern, com-
pared to the normal skin surface, with numerous fila-
mentous structures under SEM, suggesting presence of
bacteria (Fig. 6).
Aspects of the wound region after 23 and 30 days
The wound face is completely covered by the regener-
ating epidermis in the majority of the experimental
fishes. The oedema is still present under the scales at the
wound border. The central region is more pigmented. In
some wounds, the centre of the regenerated epidermis
cover is circular, but in others, it is a short straight line
(Fig. 3). At day 30, the oedema under the scales near the
border is still present and the wound has been com-
pletely closed. Thicker black radial lines, being almost
homogeneously pigmented in the central region, form
the pigmented area. Observed with SEM, the wound
central area resembles a sphincter (Figs. 3,7). There are
no longer epidermal borders, suggesting fusion.
Aspects of the wound region after 45–60 days
After 45 days, the oedema was still present and the
wound area became more homogeneously pigmented in
general, with many small black spots (Fig. 3). At day 60,
the wound region is almost entirely black, it being no
longer possible to identify the radial lines and wound
centre. The surface became more homogeneously pig-
mented with many black spots, larger than in previous
time spans and, consequently, closer to each other. The
wound border oedema remains prominent under the
scales. The wound borders have changed from angled to
round (Fig. 3).
Fig. 2 a Naturally found scar at the ventral lateral region of a N.
coriiceps. The scales are absent and the central region is black. b
Central region SEM of the same material, showing that epidermal
cells do not have the fingerprint surface. c Wound borders SEM.
Only a few epidermal cells present the typical fingerprint surface
(compare to Fig. 1)
460

The condition at 90 days
The wound border oedema under the scales has
diminished but is still present, and the wound general
aspect is darker and more rounded. A dark tissue
(darker than the rest of the animal) covers the whole
wounded region homogeneously. The original squared
shape of the wound is no longer discernible. Only a
rounded depression is visible (Fig. 8a). At the SEM
level, in the wound centre, the cells are flat and most of
them present the typical fingerprint surface (Fig. 8b).
No regenerating scales were seen in the wounded
region (Fig. 8).
Fig. 3 Photographs of the induced anterior dorsum lateral
wounds of two different N. coriicepsafter different time spans—0,
1, 2, 7, 14, 23, 30, 60 days. Haemorrhage occurred immediately
after the lesion was inflicted, turning into blood clots between 0
and 2 days. The oedema that had begun to accumulate in the
early days is more prominent at the borders after 7 days,
changing the angulated corners of the wound to a rounded
shape, and a thin epidermal layer from the borders to the centre is
already present. After 15 days, there is an increase of the
regenerating epidermis-covered area; at day 23 the wound was
completely closed in the left column, forming a line, and after
30 days in the right column, both with pigmented radial black
lines. After 30 days, thicker pigmented lines can be seen in the left
column and after 60 days in the right column. After 60 days, the
wound became homogeneously black. All photographs are at the
same scale and the bar is in millimetres
461

Kinetic studies
The number of wounds and associated data and regen-
erating epidermis surface are presented in Table 1 and
Fig. 9. The kinetics of the wound healing process can be
seen in Fig. 9. The percentages of the regenerated epi-
dermis compared to the induced wound surface after
different time spans were calculated.
Discussion
This study is the first reporting on the wound repair
kinetics in fish integument at 0
C. Only a few studies are
concerned with the influence of the acclimatisation of
fishes from tropical and temperate fresh waters on skin
regeneration. This includes wound repair in a tropical
cyprinid Tanichthys albonubes at 10–30
C and the
Atlantic salmonid Salmo salar (temperate fish) at 5–
23
C. Results suggest that the rate of wound healing in
both species was directly correlated with environmental
temperature, while temperature stress had little effect on
healing rates (Anderson and Roberts 1975). The wound
repair also depends on other factors, such as size of the
lesion, blood and nerve supply, and contamination
among others (Majno and Joris 1996).
Bullock et al. (1978), using Pleuronectes platessa
reared at different temperatures (5, 10 and 15
C), ob-
served that a lesion of 5·1 mm was completely closed
after 9 h at 10
C, and after 12 h at 5
C, by Langerhans-
type cell migration of the epidermis. The authors poin-
ted out that under low temperatures, such as 5
C, the
main mechanism of epidermal wound repair in teleost
fishes is cellular migration, an epidermal layer sliding
through the injury. In N. coriiceps, there was no quick
epidermal sliding after the first hours wound repair. The
regeneration process was affected by the epidermis slid-
ing slowly from the wound borders, moving towards the
Fig. 5 SEM. Wound surface after 15 days. Higher magnification of
the regenerating epidermis surface at the wound border. The
epidermal cells are rounded with different and incipient patterns of
fingerprints
Fig. 6 SEM. Wound central region, after 15 days, is covered by
cell debris and filaments
Fig. 4 SEM. The wound surface after 15 days. Migration front of
the regenerating epidermis. The tongue tip can be seen (bottom of
the figure)
Fig. 7 SEM. Wound surface after 30 days. The wound surface is
completely closed and the regenerated epidermis presents many
radial lines converging towards the wound centre in a sphincter-like
fashion
462

centre (4 cm2). This process required 23–30 days in or-
der to cover the whole area.
The healing speed studied by Quilhac and Sire (1999)
on the tropical teleost Hemichromis bimaculatus was
500 lm/h, after excision of the scales and part of the
epidermis on a 1-cm2 area. The re-epithelisation speed of
the tropical fish Prochilodus scrofa (20±2
C), using the
same methodology as in the present work except for
smaller lesion (1.0 cm2), was on average 6.9 lm/h (per-
sonal unpublished data). In N. coriiceps, 23–30 days
were necessary to cover 1 cm2 with a speed of regener-
ation ranging from 1.4 to 1.8 lm/h, i.e. about 300 times
slower than the data found by Quilhac and Sire (1999)
and about 4 times slower than observed for Prochilodus
scrofa (personal data).
The differences in wound healing speed found
between H. bimaculatus and Prochilodus scrofa when
compared to N. coriiceps are relative to the tempera-
ture (25, 20, and 0
C) and to the wound size (1.0, 1.0
and 2.0 cm2). The difference observed between the
wound healing speed of H. bimaculatus, described in
the studies of Quilhac and Sire (1998), and Prochilodus
scrofa (unpublished data), both with 1.0-cm2 wound,
is probably the permanence of connective tissue in
the wound. Those authors removed only the scales
and part of the epidermis, whilst we removed scales,
dermis, hypoderm and most of the perimysium,
probably complicating the sliding of the regenerating
epidermis.
There is a faster epidermal closing mechanism of the
wound by fishes living at higher temperatures, in order
to reduce the contamination risk (Bullock et al. 1978).
The supporting basis for cell migration is composed of
matrix, membrane debris and fibrils. The literature
indicates a leap-frog type re-epithelisation, due to the
facility of cell adhesion to the substrate (Quilhac and
Sire 1999). The migrating epidermis central layer slides
between other layers and after adhesion and differenti-
ation of the basal layer, a new substrate is formed to give
continuity to the sliding process until fusion, making a
stronger barrier to the external environment. N. corii-
ceps were able to deal with this problem due to the cel-
lular debris and tissue (Fig. 7) which possibly isolates
muscle cells from the hyperosmotic environment. After
2 days, N. coriiceps seems to adopt a similar strategy, i.e.
of sliding the epidermis over this necrotic area, but much
more slowly, with the tip region of the tongue detached
from the underlying substrate as described by Quilhac
and Sire (1999).
Quilhac and Sire (1999), analysing wound healing
in H. bimaculatus, observed that the migrating epi-
dermis borders made a fold over the wound after one
side reached the opposite border epidermis, suggesting
that the migration inhibitory mechanism does not
occur immediately. This does not seem to be the
strategy adopted by N. coriiceps, where there is
immediate fusion of the migrating epidermis borders
after contact.
No regenerating scales were seen on N. coriiceps
wound sections after 90 days (unpublished data),
pointing to the importance of retraction to cover or
minimise old wounds, like the naturally found wound
observed with scales around the scar. Despite the long
period needed to close the wound (23–30 days) in all
studied fishes, there were no symptoms of contamination
by bacteria or fungus retarding the process. This finding
suggests either a very effective skin-mucus antiseptic
action or the low pathogenic capacity of the Antarctic
seawater germs, or both. This study indicates wound
healing adaptation to the Antarctic temperatures (0
C)
despite the differences with the other teleost species
studied. N. coriiceps is able to deal with large extension
lesions (4 cm2) without death or contamination, dem-
onstrating the high efficiency of the wound healing
process in these animals from the Antarctic environ-
ment.
Fig. 8 a Anterior dorsum lateral induced wound of N. corii-
cepsafter 90 days (top) with a visible retraction of the borders. No
scales can be seen in wounded area. b Wound surface SEM, after
90 days, where the majority of the epidermal cells present
fingerprint surface
463

Acknowledgements Financial support was given by CNPq (68.0047/
00-0 and 48.0262/00-4 grants), PROANTAR, SECIRM, and lo-
gistic support was given by the Brazilian Navy. We also thank
Edson R. de Oliveira for the photographs, Dr. M.A. Miglino from
FMVZ-USP for the use of the SEM, and Yara Shimada Brotto for
the English revision of the paper.
References
Anderson CD, Roberts RJ (1975) A comparison of the effects of
temperature on wound healing in a tropical and a temperate
teleost. J Fish Biol 7:173–182
Borges JCS, Porto-Neto LR, Mangiaterra MBCD, Jensch-Junior
BE, Silva JRMC (2002) Phagocytosis in vivo and in vitro in the
Antarctic Sea Urchin Sterechinus neumayeri(Meissner) at 0
C.
Polar Biol 25:891–897
Bullock AM, Marks R, Roberts RJ (1978) The cell kinetics of
teleost fish epidermis: epidermal mitotic activity in relation to
wound healing at varying temperatures in plaice (Pleuronectes
platessa). J Zool 185:197–204
Finn JP, Nielsen NO (1971) The effect of temperature variation on
the inflammatory response of rainbow trout. J Pathol 105:257–
268
Gon O, Heemstra PC (1990) Fish of the Southern Ocean. J.L.B.
Smith Institute of Ichthyology, Grahmstown
Grout BWW, Morris GJ (1987) Low temperatures on biological
systems. Richard Clay ltda., Singapore, p 501
Hardie LJ, Fletcher TC, Secombes CJ (1994) Effect of temperature
on macrophage activation and the production of macrophage
activating factor by rainbow trout (Oncorhynchus mykiss) leu-
cocytes. Dev Comp Immunol 18(1):57–66
Majno G, Joris I (1996) Cells, tissues, and disease—principles of
general pathology. Blackwell, Oxford, p 974
McDowell EM, Trump BF (1976) Histologic fixatives suitable for
diagnostic light and electron microscopy. Arch Pathol Lab Med
100:405–414
Mittal AK, Munshi JSD (1974) On the regeneration and repair of
superficial wounds in the skin of Rita rita (Ham.) (Bagridae,
Pisces). Acta Anat 88:424–442
Phromsuthirak P (1977) Electron microscopy of wound healing in
the skin of Gasterosteus aculeatus. J Fish Biol 11:193–206
Prisco G, Maresca B, Tota B (1991) Biology of Antarctic fish.
Springer, Berlin Heidelberg New York, p 285
Fig. 9 Area of the regenerated
epidermis and the wound area
of N. coriiceps after different
time spans. The wound borders
start to retract after 7 days and
this process persists until the
longest period studied
Table 1 Number of wounded regions examined in the kinetics studies after different time spans, showing the wound surface, the
regenerated surface and the percentage of regenerated surface
Days after injury 0 1 2 7 15 23 30 60 90
No. of wounds measured (n) 43a 43 43 43 41 40 38 20 4
Wound surface (mm2)
Average 417.62 451.62 456.32 506.98 503.91 487.88 469.58 456.70 341.24
Standard deviation 65.45 78.47 69.42 80.93 79.71 87.10 83.69 73.12 32.07
Maximum value 544.32 604.42 645.50 719.81 674.52 767.14 677.76 672.58 375.30
Minimum value 288.40 319.73 330.68 357.08 372.15 340.37 336.37 360.02 299.69
Regenerated surface (mm2)
Average 0.00 0.00 13.70 164.95 340.82 456.41 465.98 456.70 341.24
Standard deviation 0.00 0.00 22.18 50.89 72.61 79.68 81.93 73.12 32.07
Maximum value 0.00 0.00 75.21 261.71 476.96 731.06 677.76 672.58 375.30
Minimum value 0.00 0.00 0.00 79.60 223.70 318.79 336.67 360.02 299.69
Percentage of regenerated
surfaceb
Average 0.00 0.00 3.07 32.65 68.15 94.05 99.30 100.00 100.00
Standard deviation 0.00 0.00 4.86 9.10 12.51 8.17 2.40 0.00 0.00
Maximum value 0.00 0.00 15.26 49.38 96.65 100.00 100.09 100.00 100.00
Minimum value 0.00 0.00 0.00 15.22 41.40 67.81 86.39 100.00 100.00
aOne wound was damage due to mechanical shock in the tank and was not included in this study
bAll the percent values of regenerated epidermis were significantly different (Turkey-Krammer-multiple comparison test)—GRAPHPAD
INSTAT
464

Quilhac A, Sire JY (1998) Restoration of the subepidermal tissues
and scale regeneration after wounding a cichlid fish, Hemichr-
omis bimaculatus. J Exp Zool 281:305–327
Quilhac A, Sire JY (1999) Spreading, proliferation, and differen-
tiation of the epidermis after wounding a cichlid fish, He-
michromis bimaculatus. Anat Rec 254:435–451
Reddan JR, Rothstein H (1965) Influence of temperature on
wound healing in a poikilotherm. Exp Cell Res 40:442–445
Silva JRMC, Hernadez-Blazquez FJ, Barbieri RL (1998) Induced
inflammatory process in the Antarctic fish Notothenia neglecta.
Polar Biol 20:206–212
Silva JRMC, Stainess N, Parra OM, Hernadez-Blazquez FJ (1999)
Experimental studies on the response of the fish (Notothenia
coriiceps Richardson, 1844) to parasite (Pseudoterranova deci-
piens Krabbe, 1878) and other irritant stimuli at Antarctic
temperatures. Polar Biol 22:417–424
Silva JRMC, Peck L (2000) Induced in vitro phagocytosis of the
Antarctic starfish Odontaster validus (Koehler, 1906) at 0
C.
Polar Biol 23:225–230
Silva JRMC, Hernandez-Blazquez FJ, Porto-Neto LR, Borges JCS
(2001) Comparative study of in vivo and in vitro phagocytosis
including germicide capacity in Odontaster validus (Koehler
1906) at 0
C. J Invertebr Pathol 77:180–185
Silva JRMC, Staines N, Hernandez-Blazquez FJ, Porto-Neto LR,
Borges JCS (2002) Phagocytosis and giant cell formation at 0
C
by macrophage (M/) of Notothenia coriiceps. J Fish Biol
60:466–487
465