sequential changes in trace metal, metallothionein and ...€¦ · 2.11.98). other metals found in...
TRANSCRIPT
J. Anat. (1999) 195, pp. 375–386, with 12 figures Printed in the United Kingdom 375
Sequential changes in trace metal, metallothionein and
calmodulin concentrations in healing skin wounds
A. B. G. LANSDOWN,1 B. SAMPSON1 AND A. ROWE2
"Departments of Clinical Chemistry, Imperial College School of Medicine and Charing Cross Hospital, and #Clinical
Pharmacology, Imperial College School of Medicine, Chelsea and Westminster Hospital, London, UK
(Accepted 4 May 1999)
Metalloenzymes have an important role in repair and regenerative processes in skin wounds. Demands for
different enzymes vary according to the phase in the healing cascade and constituent events. Sequential
changes in the concentrations of calcium, copper, magnesium and zinc were studied in the incisional wound
model in the rat over a 10 d period. Copper levels remained low (! 10 µg}g dry weight) throughout, but
calcium, magnesium and zinc increased from wounding and peaked at about 5 d at a time of high
inflammation, granulation tissue formation and epidermal cell proliferation. Metal concentrations declined
to normal by 7 d when inflammation had regressed, re-epithelialisation of the wound site was complete and
the ‘normalisation’ phase had commenced. Although the wound was overtly healed by 10 d, the epidermis
was still moderately hyperplastic. In view of competitive binding of trace metals at membrane receptors and
carrier proteins, the ratios or balance between these trace metals was examined and the significance is
discussed. Using immunocytochemistry, we demonstrated increases in metallothionein immunoreactivity as
an indication of zinc and copper activity in the papillary dermis and in basal epidermal cells near the wound
margin 1–5 d after wounding. This is consistent with metalloenzyme requirements in inflammation and
fibrogenesis. Calmodulin, a major cytosolic calcium binding protein was highest in maturing keratinocytes
and in sebaceous gland cells of normal skin; it was notably more abundant in the epidermis near the wound
margin and in re-epithelialising areas at a time when local calcium levels were highest.
Key words : Skin wounds; repair ; trace metalloenzymes; metal ion balance; zinc ; calcium; copper ; magnesium.
Wound healing in the skin and in other tissues of the
human body depends upon the availability of com-
petent cells to carry out repair processes, appropriate
activation by hormones, growth factors, cytokines
and chemotactic factors, and a microenvironment
favouring cell movement, proliferation and functional
maturation (Grotendorst, 1992; Hunt & Hussain,
1992). Metal ions and metalloenzymes feature promi-
nently in this wound healing environment and
experimental and clinical studies are available to show
that deficiencies in zinc, copper, calcium, iron or
magnesium are potential causes of abnormal homeo-
static mechanisms and impaired wound healing
(Moynahan, 1974; Lansdown, 1995). Although at-
Correspondence to Dr A. B. G. Lansdown, Department of Clinical Chemistry, Division of Investigative Sciences, Imperial College School of
Medicine, Charing Cross Campus, London, W6 8RP, UK.
tention recently has focused upon the role of metallo-
proteinases in the degradation of extracellular matrix
in the early phases of wound healing (Nwomeh et al.
1998), it is to be expected that the demands for these
and other metalloenzymes will vary greatly as the
profile of wound healing proceeds. Surprisingly, the
action and interaction of trace metals in the skin
under normal conditions is not well documented, and
sequential changes in the healing wound are not
published for any species (Lansdown, 1995).
Much of the trace metal requirement for the body is
derived from the diet, but many topical dressings,
healing creams and other commercial preparations
containing zinc and calcium are now available to
improve wound healing. There is a general lack of
published data to show how much ‘active ’ metal ion
is absorbed percutaneously from these preparations,
and we can only speculate on how much additional
ion is necessary to advance the wound healing process.
Much attention has centred upon the value of
supplementary zinc in wound healing; zinc is a
constituent of more than 70 enzyme systems, many of
which are involved in wound healing (Lansdown,
1993, 1996). However, whereas zincated creams can
be beneficial under some circumstances, excessive zinc
may retard healing (Lansdown & Sampson, unpub-
lished studies). This is attributed to the fact that zinc,
copper and calcium interact at binding sites on carrier
proteins. Excess of one ion is liable to inhibit essential
processes modulated by others (Klevay, 1975; Heng et
al. 1993). It follows that any xenobiotic ion absorbed
into the skin that impairs the availability of zinc or
other trace metals, is a potential cause of delayed or
nonhealing of wounds (Lansdown, 1996).
Recent studies have demonstrated that concen-
trations of trace metals in normal skin are specific for
the region of the body, its keratinisation patterns and
contact with the environment (Lansdown, 1985, 1995;
Lansdown & Sampson, 1997). Zinc, calcium and
magnesium concentrations are highest in areas of
pressure keratinisation and parakeratosis, which are
normally characterised by high epidermal cell turn-
over. In the healing wound, the demand for trace
metals, mainly in the form of metalloenzyme com-
plexes, can be expected to vary according to the
sequential events of haemostasis, inflammation}granulation tissue formation, cell proliferation and
normalisation of the wound site (Lansdown, 1995).
Whereas demands for calcium will be high during
haemostasis and periods of keratinocyte proliferation
and maturation (Menon et al. 1985), zinc metallo-
enzymes will feature in matrix metalloproteases, RNA
and DNA polymerases and enzymes involved in
protein synthesis, mitosis, collagenesis and tissue
remodelling (Halstead et al.1974). Copper exhibits a
ubiquitous role in cellular metabolism in the skin, but
is notable as a component in lysyl oxidase necessary in
elastic and collagen fibre cross-linking (Chou et al.
1968).
Present studies were conducted with a view to
Fig. 1. Metallothionein distribution in the basal layer of the epidermis of the back skin of a young adult Sprague–Dawley rat. Immuno-
cytochemistry using antimetallothionein antibody (DAKO, Copenhagen). ¬150.
Fig. 2. Calmodulin distribution in the outer layers of the epidermis of the back skin of a young adult Sprague–Dawley rat. Immuno-
cytochemistry using anticalmodulin polyclonal antibody (Santa Cruz Biotechnology Inc, CA.) ¬150.
Fig. 3. Incisional wound site in the rat after 24 h showing the inflammatory cell infiltrate at the wound margin and debris in the wound core.
Note lack of inflammation in adjacent tissue. H & E. ¬75.
Fig. 4. Narrow tongue of epidermal cells migrating between the acute inflammatory cells and wound debris at deeper aspects of the 48 h
incisional wound. H & E. ¬150.
identifying sequential changes in the mineral and trace
metal content of incisional wounds in the rat model,
and correlating these with cellular events charac-
terising the major events of the wound healing cascade
(Mast, 1992). The rat has been used for many years to
study wound healing profiles in the skin and has
proved a valuable model in which to evaluate the
influence of intrinsic and environmental influences on
repair processes (Lansdown & Pate, 1993). In this
work, we sought to define our experimental model for
use in the evaluation of dietary constituents, topical
preparations or medicated dressings which are po-
tentially able to alter trace metal ion uptake and
metabolism in the wound environment. From earlier
studies, we know that silver can interact with zinc to
advance experimental wound healing in the rat
(Lansdown et al. 1997). We are unclear at this stage,
whether the effect is achieved through applying silver
nitrate or silver sulphadiazine throughout the wound
healing period, or only during phases of high zinc
metalloenzyme activity. In another situation, the
xenobiotic metal cadmium was shown to induce a
marked increase in ‘bound’ zinc in intact skin
(Lansdown & Sampson, 1996). In wounded skin, this
would be expected to seriously impair wound healing
through inhibiting zinc metalloenzymes involved in
matrix degradation, cell proliferation and connective
tissue formation.
Animals
Adult male Sprague-Dawley rats of the CFY strain
(180–200 g body weight) were used in all sections of
this study. They were bred under barrier-maintained
specified pathogen free conditions at the Charing
Cross Campus of Imperial College School of Medi-
cine. They were housed under conditions of 22³2 °C,
45–55% relative humidity, and 12 h d}night cycles as
specified in the Animals [Scientific Procedures] Act,
1986. The animals were housed in groups of 5 in
plastic solid bottomed cages and provided with sterile
dust free sawdust as bedding. Pelletted small rodent
diet (CRM, Special Diet Services, Witham, Essex)
376 A. B. G. Lansdown, B. Sampson and A. Rowe
1 2
3 4Fig. 1. For legend see opposite.
Trace metals and skin wound healing 377
and tap water was provided ad libitum. The mineral
content of this diet includes copper (19 mg}kg),
zinc (65 mg}kg), magnesium (0.22%), and calcium
(0.80%) (Special Diet Services, Information Service,
2.11.98). Other metals found in this diet either as
trace metals or as contaminants are not present in
sufficiently high levels to significantly influence the
metabolism of zinc, calcium, copper or magnesium in
wound repair (Lansdown, 1995).
Experimental details
Incisional full thickness skin wounds (15 mm long)
were made surgically with a scalpel in the closely
shaved middorsal skin of fully anaesthetised rats. A
single middorsal wound was made in each animal.
Skin sites were cleansed with 70% ethyl alcohol and
animals anaesthetised with a single intraperitoneal
injection of Hypnorm (fentanyl}fluanisone) (Janssen
Animal Health) and midazolam combination an-
aesthetic (0.27 ml}100 g body weight). This provided
anaesthesia for " 30 min with good muscle relaxation
(Flecknall, 1987). Wounds were sutured at 3 equi-
distant sites using polyamide monofibre (Ethicon,
Edinburgh, Scotland). To standardise wounding, all
procedures were conducted during the period 09.00–
10.30 h.
Groups of 10 rats received wounds as above and
were euthanised by carbon dioxide asphyxiation after
1, 2, 3, 5, 7 and 10 d. Earlier studies have demonstrated
that rat wounds heal well within 10 d (Lansdown &
Pate, 1993). At autopsy, wound sites were excised,
with normal tissue to within 1.5 mm of the incision
line and samples taken for trace metal analysis,
histology and immunocytochemistry. Wounds from
5 animals for each time period, were cryopreserved at
–20 °C for metal analysis, 2 were preserved in ice-
cold paraformaldehyde (4% w}v in phosphate
buffered saline, PBS) for immunocytochemistry, and 3
fixed in cold (®4 °C) 10% PBS for paraffin wax
embedding and routine histology. Samples (5 mm¬5 mm) of intact skin from an uninjured area of back
skin were cryopreserved for metal analysis as no
useful data exist on trace metal content in normal rat
skin. Histological sections were stained with haema-
toxylin and eosin, or by Masson’s trichrome for
connective tissue.
To examine early changes in the wound, 2 wound
bearing rats were euthanised 12 h after surgery and
the wound site preserved in 10% phosphate buffered
formalin for histology and haematoxylin and eosin
stained sections.
Immunocytochemistry
Dissected skin wound samples were processed through
several changes of sucrose (15% w}v in PBS) over
36 h and then frozen in ‘Cryo-M-Bed’ (Bright Ltd, St
Ives, Cambridge, UK) at ®60 °C. Tissues were
sectioned at 6 µm and immunocytochemistry per-
formed (Rowe et al. 1997) using Vectastain ABC
Elite reagents (Vector Laboratories, Burlingham,
CA, USA) according to the supplier’s protocol. Rat-
adsorbed biotinylated antimouse IgG (Vector Labora-
tories) was used for the detection of IgG primary
antibodies. The chromagen employed was 3,3«-di-
aminobenzene (Vector Laboratories) with Harris’s
haematoxylin counterstain. Monoclonal mouse anti-
metallothionein antibody was kindly donated by
DAKO Laboratories (Copenhagen, Denmark). It was
supplied in liquid form in 0.05 Tris HCl, 15 m
NaN3, pH 7.5, 1% bovine serum albumin (BSA).
DAKO-metallothionein antibody is inhibited specifi-
cally and equally well by glutaraldehyde polymerised
human, horse, sheep or rat metallothioneins (MT-I
and MT-II) suggesting that its is directed against a
single and hightly conserved epitope ((DAKO Infor-
mation Sheet). It was diluted 1:150. Anticalmodulin
goat polyclonal antibody (Santa Cruz Biotechnology,
Santa Cruz, CA, USA) was diluted 1:100. Negative
control sections were incubated with no primary
antibody.
Metal analysis
Tissue samples were dried to a constant weight at
120 °C and then dissolved in concentrated nitric acid
(Aristar Grade, BDH, Poole, Dorset, UK) by heating
overnight at 90 °C. The resulting solutions were
diluted to an appropriate volume using ultrapure
water. Appropriate reference standards were analysed
at the same time (see Lansdown & Sampson, 1997).
Calcium, magnesium and zinc concentrations were
measured by flame atomic spectrophotometry (Uni-
cam 939, ATI Unicam) after dilution with lanthanum
chloride (for calcium and magnesium) or 6% butan-
1-ol (for zinc). Copper was measured by electro-
thermal atomisation flame atomic spectrophotometry
(Unicam 939 with GF90 electrothermal atomiser).
The skin on the back of an adult rat is densely haired
and is devoid of sweat glands, although sebaceous
glands are prominent in association with the nu-
merous hair follicles. The epidermis is 4–6 cells thick
and comprises characteristic layers of basal, spinous,
378 A. B. G. Lansdown, B. Sampson and A. Rowe
granular and keratinised cells. Under normal con-
ditions, basal cells express cytoplasmic metallo-
thionein and low levels of calmodulin (Figs 1, 2).
Minimal amounts of metallothionein are present in
more superficial layers of the epidermis, but cal-
modulin is more noticeable in maturing keratinocytes.
The dermis comprises collagen-rich connective
tissue interspersed with a modest vascular supply and
nervous tissue. Under normal conditions this tissue
displays low levels calmodulin and metallothionein.
Macrophages, fibroblasts and the paniculus carnosus
muscle (PCM) in the deeper aspects of the zona
reticularis exhibit immunoreactivity for both proteins.
Healing profiles in the skin wound
Haemostatic and early post wound period (0–12 h)
All animals recovered well from anaesthesia and
exhibited no signs of ill health at any time. No
fatalities occurred. Following surgery, there was a
minimal haemostatic phase (! 30 min). Wounds
dried quickly and showed no appreciable adverse
response to the suture material used. Histological
examination of 12 h wounds showed local oedema in
the region of the incision line but minimal evidence of
inflammatory cell infiltration in marginal areas. The
epidermis at the wound margin was necrotic. Aggre-
gations of red blood cells were evident at all levels
from the surface of the skin to the hypodermis.
Inflammatory and proliferative phase (1–7 d)
Crusting and scab formation occurred along the
wound suture line within 24 h and this persisted for
approximately 5 d. After this time, we noted a
progressive loss of sutures and prominent hair growth.
Where sutures had been lost, insertion marks re-
mained.
By 24 h the wounds exhibited a prominent central
core of cell debris with haemolysed erythrocytes
extending through the epidermis and dermis into the
hypodermis. In the dermis, this mass was separated
from mildly oedematous but otherwise histologically
normal tissue by a zone of acute inflammatory cells
(mainly neutrophils). Initially, this infiltrate was more
pronounced in the reticular zone than in the papillary
area (Fig. 3). The epidermis at the wound margin
consisted of a cap of fragmented and necrotic tissue
adjacent to histologically normal keratinocytes. Mi-
totic activity was not noted at this stage in histo-
logically normal-looking keratinocytes adjacent to the
wound margin. In the deeper reticular dermis, the
severed PCM was associated with infiltrates of
neutrophils, monocytes and a few eosinophils. At this
stage, epidermal concentrations of calmodulin near
the wound margin had increased appreciably and
were present in basal cells, which in intact skin are
usually low in this protein.
Metallothionein levels in the epidermis and dermal
fibroblasts near the wound margin after 24 h exhibited
a marginal increase.
After 48 h, there was a pronounced increase in
the dermal inflammatory cell infiltrate in the wound
margin. Importantly, by this stage mitotic activity was
present in the epidermis at the wound margin and
tongues of epithelial cells had begun to migrate
downward with the effect of separating the central
core of tissue debris and inflammatory cells from the
‘peripheral ’ zone of histologically normal tissue (Fig.
4). There was minimal evidence of an epidermal
basement membrane in the early stages of re-
epithelialisation. A diffuse inflammatory cell infiltrate
and some oedema persisted in the dermis near the
wound margin, this being more obvious in the deeper
reticular region, where vascular dilation was noted.
By 48 h postwounding, dermal inflammatory cell
infiltrates included monocytes, some fibroblasts and
macrophages. Minimal evidence of collagenesis was
seen until after 72 h when the fibroblast population of
the wound site had markedly increased both in the
reticular and papillary regions. We noted a follicular
cell hyperplasia in the region of the wound, such that
follicles cut during surgery were contributing to the
re-epithelialisation process.
The re-epithelialisation process continued until
approximately 5 d after surgery, when the wound
site resembled a crater lined by epithelial cells and
containing a central core of wound debris and
inflammatory cells. In deeper aspects of the crater the
epidermis was not well defined and lacked a clear
basement membrane (Fig. 5). A few chronic inflam-
matory cells were present. Except in these deeper
layers, keratohyalin granules were prominent in the
reformed epidermis and in places extended over up to
60% of the epidermal thickness. In the 5 d wound, the
reformed epidermis was up to 3 times the normal
thickness and mitoses were numerous in the basal
layer, both in the new epidermis and in adjacent areas.
The stratum corneum was identifiable as a thin strand
in the early re-epithelialisation process but by 5 d
extended across the wound site. Some parakeratosis
was present in the deeper regions of the wound crater.
At this time, the re-epithelialised epidermis showed
minimal evidence of the columnar epidermal cell
organisation although more superficial cells were
flattened parallel with the skin surface.
Trace metals and skin wound healing 379
5 6
7 8Fig. 5. Reepithelialised incisional skin wound in the rat after 5 d. Base of the crater to demonstrate the residual inflammatory cell infiltrate
(mainly lymphocytes), fibroblasts and early fibrogenesis. Note the lack of basement membrane at the epidermal-dermal interface. Masson’s
trichrome method. x150.
Fig. 6. Metallothionein distribution in the epidermis and papillary dermis of the incisional skin wound after 24 h. ¬150.
380 A. B. G. Lansdown, B. Sampson and A. Rowe
9 10Fig. 9. Rat incisional wound site after 10 d demonstrating the fibrous scar tissue, minimal inflammatory cell infiltration and modestly
hyperplastic epidermis. H & E. ¬75.
Fig. 10. Increased metallothionein concentrations in macrophages in the hypodermis under an incisional wound site in the rat 10 d following
surgery. Immunocytochemistry using DAKO antibody. ¬75.
The dermal inflammatory cell population after 5 d,
was predominantly of fibroblasts and lymphocytes
with macrophages, eosinophils and a few mast cells
present at deeper layers and in the region of the PCM.
Collagenesis was active beneath the wound and in the
papillary epidermis adjacent to the wound site.
Granulation tissue with prominent vascular dilatation
was notable, particularly in the deep dermis and
hypodermis.
The inflammatory and proliferative phases of
wound healing were marked by an increase in
metallothionein in epidermal cells at the wound
margin and in regenerating tissue (Fig. 6). Interest-
ingly, we noted a pronounced increase in metallo-
thionein in fibroblasts and macrophages of the
papillary epidermis in wounds examined up to 5 d.
Fig. 7. Increases in calmodulin activity in the wound margin of the 24 h skin wound. Note the presence of calmodulin reaction product in
basal cells and in all postmitotic cells of the epidermis. Calmodulin antibody is seen in sebaceous gland cells. ¬150.
Fig. 8. Rat incisional wound site after 7 d showing the hyperplastic epidermis, with a prominent granular layer, early evidence of realignment
of epidermal cells and hyperkeratinisation. Scar tissue formation is shown. Masson’s trichrome. ¬75.
Calmodulin in contrast was essentially epidermal in
distribution. It increased markedly over the first 4 d of
re-epithelialisation in the tongues of migrating keratin-
ocytes (Fig. 7). Increased epidermal calmodulin
persisted in the re-epithelialised epidermis well into
the normalisation phase.
Normalisation phase (7–10 d)
Healing in the rat wound from 5 d was associated
with a pronounced decline in the inflammatory cell
infiltrate in and near the wound margin, and a
progressive reduction in the crater formation with loss
of the wound debris and a realignment of epidermal
cells. Thus after 7 d, the wound site was characterised
Trace metals and skin wound healing 381
Fig. 11. Sequential changes in zinc, calcium, copper and magnesium
in the healing wound site.
by a pit-like configuration and lined by a hyperplastic
epidermis with a prominent keratohyalin granulation
and hyperkeratinisation (Fig. 8). We noted a pro-
nounced formation of rete pegs at the epidermal–
dermal interface of the new epidermis, which was up
to 3 times the normal thickness. There was no evidence
of repair in the PCM but collagen deposition}scar
tissue formation was evident at this site. After 10 d,
the normalisation process in the rat wound was well
advanced (Fig. 9), but the epidermis was still
marginally thicker than normal and cellular realign-
ment incomplete. Epidermal mitoses were appreciably
less obvious than during the 3–5 d period but some
follicular cell hyperplasia persisted. Dermal cellularity
was greatly reduced, and blood vessels appreciably
less obvious than at earlier stages. Scar tissue was
prominent.
From 5 d after wounding, metallothionein levels
declined in the epidermis, papillary dermis and
follicular epithelium to normal levels. A slightly
increased level of immunoreactivity was seen in the
region of the PCM and in macrophages in the
hypodermis (Fig. 10). Calmodulin concentrations
were appreciably higher than normal in the hyper-
Fig. 12. Sequential changes in the ratio of zinc :calcium, calcium:
magnesium, and zinc:copper in the healing wound site.
plastic epidermis in maturing keratinocytes but not in
most basal cells.
Trace metal analyses
The mean concentrations of zinc, copper, calcium and
magnesium in uninjured shaved back skin of the
young adult Sprague Dawley rat are illustrated in
Figure 11. In the skin wound site, we noted a
progressive increase in zinc, calcium and magnesium
up 5 d postwounding and then a decline to normal
values by 7 d. Zinc concentrations were marginally
higher than normal 10 d after wounding. Copper
concentrations were very low at all stages in the
wound healing profile, but did show a marginal
increase in the first 2 d.
Because certain metals compete for binding sites on
carrier proteins, it was of interest to examine changes
in the ratio of metal concentrations in the period of
study (Fig. 12). Thus, there was a progressive decline
in the zinc:calcium ratio over first 5 d after wounding.
The zinc:calcium ratio had normalised by 7 d but had
increased further by 10 d. In contrast, the relative
concentration of calcium to magnesium increased to
5 d after wounding but had declined to near normal
levels by 7 d. Although copper levels were low for the
wound healing period studied, the ratio of zinc to
382 A. B. G. Lansdown, B. Sampson and A. Rowe
Table. Trace metal concentrations in the middorsal skin of a
young adult Sprague–Dawley rat
Zinc 63.94³11.73 µg}g dry tissue weight
Copper 5.13³1.06 µg}g
Calcium 0.59³0.01 mg}g
Magnesium 0.46³0.03 mg}g.
copper appeared to decline in the early postwound
period (C 1–2 d). They were increased at later stages.
Wound healing in acute skin wounds is a well defined
sequence of events involving cell–cell contact and
cell–macromolecular interactions leading to haemo-
stasis, inflammation (granulation tissue formation),
fibroplasia, neovascularisation, contraction, and re-
epithelialisation of the wound site (Cai et al. 1991).
Repair processes and their associated biology have
traditionally involved the use of in vitro systems and
laboratory animal models which allow the constituent
processes to be studied under controlled conditions
with the avoidance of such variables as genetic
configuration, nutrition and environmental factors
(Mulder, 1991). With new technology and the avail-
ability of specific antisera, it now possible to in-
vestigate more closely the role of cytokines, growth
factors and other modulators in cellular morpho-
genesis, extracellular matrix protein synthesis, and
membrane receptor activity (Mast, 1992; Nickoloff et
al. 1991; Kilcullen et al. 1998). Although the par-
ticipation of metalloenzymes is appreciated, the
present work is the only evaluation to our knowledge,
where sequential changes in trace metal concen-
trations have been correlated with cytological events
in the healing cascade. We assume here that largest
part of the metals present will be as constituents of
metalloenzyme systems (Lansdown, 1995; and Table).
Through sequential histological observations over
the wound healing period, we can appreciate the
duration of phases of haemostasis, inflammation and
granulation tissue formation, proliferation and nor-
malisation (remodelling) in the rat model. However,
whereas we can identify periods of high}maximal
inflammatory activity, fibrogenesis or epidermal cell
proliferation, our model can only provide an approxi-
mate guide to the time of onset of and ‘cut-off’
points for the 4 main phases.
The complex sequence of events leading to repair in
the skin wound commences at the time the tissue is
damaged and the ‘negative feedback’ mechanism
induced (Gelfant & Candelas, 1972; Bullough, 1975;
Argyris, 1981). Vascular damage following trans-
epidermal wounding leads to haemorrhage with the
release of platelets (Wahl & Wahl, 1992). Aggregation
of the platelets promoted through contact with type
IV and V collagen molecules of the subendothelium
and the release of platelet derived growth factor
(PGDF), thromboxane A#, ADP, fibronectin, sero-
tonin, factor VIII (von Willebrand’s factor) (Lynch,
1991) achieves haemostasis. Calcium performs a key
role in this haemostatic process, probably as a primary
catalyst in platelet aggregation and in the production
of clotting factors VIII, IX and X (Born & Cross,
1964; Blair et al. 1990). Calcium is also necessary in
neutrophil activation and epidermal cell proliferation
at slightly later stages. In the rat wound model,
haemostasis is accomplished within 30 min but after
12 h minimal evidence is seen histologically of im-
pending repair mechanisms. Nevertheless, the im-
portance of calcium at these early stages in wound
healing is recognised by the 100% increase in
local concentrations of the metal in the 24 h after
wounding. The importance of calmodulin as a major
calcium binding protein in keratinocyte maturation is
demonstrated in this study by heavy deposits in the
regenerating epidermis at and near the wound margin.
Inflammation and granulation tissue formation are
promoted largely by secretions released from activated
platelets (Wahl & Wahl, 1992). In the rat, this phase
has commenced by 24 h after wounding and persists
until at least 5 d. There is considerable overlap
between this phase and the period of epithelial cell
proliferation and fibrogenesis, which also seems to
peak at about 5 d postwounding. By this time, local
concentrations of calcium, zinc and magnesium have
increased greatly over prewounding levels. Although
zinc metalloenzymes play a major role in immune and
inflammatory responses and are essential in tissues
subject to proliferation and metabolic activity (Lans-
down, 1996), our analyses show that demands for
calcium are notably higher at all stages in the wound
healing profile. Whereas calcium is required for the
activation of many intercellular reactions, zinc tends
to act as an inhibitor (Chvapil et al. 1975; Brewer
et al. 1979). Zinc is now known to regulate calcium
uptake and its contribution to the calmodulin-c-AMP
pathway. Zinc injected intradermally has been shown
to reduce calmodulin and calcium concentrations in
the skin, and raise cAMP levels (Heng et al. 1993).
We can appreciate that the critical ratio of zinc:
calcium will change as the inflammation}granulation
tissue formation progresses, and proliferative activity
gathers momentum in the epidermis at the wound
margin. Thus the zinc:calcium ratio declined from the
time of wounding until approximately 5 d. Thereafter,
Trace metals and skin wound healing 383
metal concentrations and ratios of zinc:calcium
tended to return to prewounding levels by 7 d.
The final phase in the wound healing cascade is
characterised by a marked decline in the cellularity of
the wound site with reduction in the inflammatory cell
infiltrate and granulation tissue stage, and consoli-
dation of the fibrous scar tissue. The re-epithelial-
isation process is complete and the collagen IV-rich
basement membrane reformed. Once this is achieved,
the epidermal ‘normalisation’ process involves a
realignment of postmitotic cells from a lateral or
semilateral migration pathway (as seen in the re-
epithelialisation process) to the vertical movement
(Pinkus, 1970). In the rat, this occurs over the 7–10 d
postwounding period. But, even after 10 d the wound
exhibited an epidermis 2 to 3 times thicker than
normal. By 10 d, we saw a fractionally higher zinc
concentration and increased zinc:calcium ratio. It
is conceivable that in this situation, zinc ions are
involved in suppressing calcium motivated epidermal
cell proliferation and maturation (Hennings et al.
1980; Yuspa et al. 1989).
The modulation of trace metal ions in normal
mammalian systems is not well documented but is
likely to involve carrier proteins, growth factors and
cytokines (Hembry et al. 1985; Winge & Nielson,
1985). In the present study, we have examined only
calmodulin and metallothionein as markers. Metallo-
thionein occurs as 2 major isoforms, MT-I and
MT-II, which may be expressed in a tissue-dependant
manner (Pauwels et al. 1994; Yang et al. 1994). There
are also 2 other isoforms, MT-III which occurs
primarily in neural tissue (Palmiter et al. 1992), and
MT-IV which has been identified in squamous
epithelial tissue (Quaife et al. 1994). The antibody
used in the present study is known to react with MT-
I and MT-II (DAKO, Copenhagen), but it is unclear
if it reacts with MT-IV also. Although we noted that
calmodulin was present in most tissues, it was highest
in differentiating keratinocytes reflecting the import-
ance of calcium in this process. Metallothioneins on
the other hand, related well to zinc and possibly
copper involvement in dermal fibrogenesis and scar
tissue formation. Thus we saw increases in metallo-
thionein reaction product in the papillary dermis at
about the time at a time of early inflammatory cell
infiltration and fibrogenesis (collagenesis) with a
decline at later stages in wound healing. These changes
are consistent with the demand for zinc dependant
RNA and DNA-polymerases in essential repair
processes (Lindeman & Mills, 1980). Metallothioneins
are induced by and bind copper, such that changes in
their distribution are likely to reflect areas of increased
copper metabolism and the requirement for copper in
lysyl oxidase in collagenesis and elastic tissue for-
mation (Chou et al. 1968; Madden & Peacock, 1968).
Therapeutically, advantage is taken of the need for
additional calcium in wound healing in the devel-
opment of calcium alginate products (Jarvis et al.
1987; Blair et al. 1990; Lansdown, & Payne, 1994).
Thus, when calcium alginate fibres contact the blood,
calcium ions exchange for sodium in the wound
exudate and are released into the wound environment
to be available for haemostatic and other calcium
dependant processes. Zinc-containing products have
been developed variously for treating diaper rash and
other minor skin lesions (Lansdown, 1993, 1996).
Experimental studies have demonstrated that high
concentrations of zinc applied to the skin in an
amphiphilic cream to aid absorption, can be det-
rimental in wound healing, presumably because the
excess zinc absorbed is inhibiting calcium dependant
pathways (Lansdown & Sampson, unpublished).
Although copper may have a role in stabilising
epidermal cells and as Solcoderm is claimed to be
successful in treating epidermal cysts (Ronnan et al.
1993), excess copper is contraindicated in wound
healing (Barranco, 1972; Sucvi et al. 1981; Fisher,
1986). Our studies show that demands for copper are
particularly low in normal wound healing in the rat,
suggesting that excess concentrations are likely to
prove toxic through upsetting critical balances with
other divalent trace metal ions.
The skin in the rat is overtly different from human
skin but the constituent tissues are of a similar origin.
They will undergo comparable morphogenetic
patterns and be subject to the influence of similar
chemotactic systems, growth regulators, cytokines,
and nutritional factors. In the presence of injury, we
can expect cellular responses and homeostatic mech-
anisms to respond in a similar fashion in rats and
humans, differences being more in the rate and
magnitude of reaction, than of type. In mechanistic
studies the rat wound model is presented as an initial
screen in which to evaluate the potential value of a
dietary or therapeutic means of advancing tissue
repair.
Toxicologically, we can predict that any situation
liable to impair the uptake or metabolism of trace
metals, or alter the critical ratios of one trace metal to
another will be detrimental in wound healing in the
skin. This will include the application of medicated
dressings and topical creams designed to deliver
increased levels of zinc or calcium. More importantly,
percutaneous absorption of environmental pollutants
containing lead, cadmium or other xenobiotic min-
384 A. B. G. Lansdown, B. Sampson and A. Rowe
erals which are known to interact with zinc or other
essential trace element are likely to be detrimental in
wound healing (Lansdown, 1995). On the other hand,
silver which is known to induce metallothionein and
through preferential binding releases zinc, has been
shown to advance wound healing in the skin.
(Lansdown et al. 1997).
We are grateful to Mr Obah Ojarikre for his assistance
with statistics and preparation of the tables used in
this paper.
ARGYRIS TS (1981) The regulation of epidermal hyperplastic
growth. CRC Critical Reviews in Toxicology 9, 151–200.
BARRANCO VP (1972) Eczematous dermatitis caused by internal
exposure to copper. New England Journal of Medicine 239,
386–388.
BLAIR SD, JARVIS P, SALMON M, MCOLLUM C (1990)
Clinical trialof calciumalginatehaemostatic swabs.BritishJournal
of Surgery 77, 568–570.
BORN GVR, CROSS MJ (1964) Effects of inorganic ions and of
plasma proteins on the aggregation of blood platelets by
adenosine phosphate. Journal of Physiology 17D, 397–414.
BREWER GJ, ASTER J, KNUTSEN CA, KRUCKEBERG WC
(1979) Zinc inhibition of calmodulin: a proposed molecular
mechanism of zinc action of cellular functions. American Journal
of Haematology 7 53–60.
BULLOUGH WS (1975) Mitotic control in adult mammalian
tissues. Biological Reviews 50, 99–127.
CAI J-P, HARRIS B, FALANGA V, EAGLESTEIN P, MERTZ
M, CHIN YH (1991) Recruitment of mononuclear cells into
wounded skin: mechanism and modulation. In Clinical and
Experimental Approaches to Dermal and Epidermal Repair:
Normal and Chronic Wounds (ed. Barbul A, Caldwell MD, Hunt,
TK, Marshall, D, Pines E, Skover G ), pp. 243–256. New York:
Wiley-Liss.
CHOU WS, SAVAGE JE, O’DELL BL (1968) Relation of
monoamine oxidase activity and cross-linking in copper deficient
and control tissues. Proceedings of the Society for Experimental
Biology and Medicine 128, 948–952.
CHVAPIL M, WELDY PL, STANKOVA L (1975) Inhibitory
effect of zinc ions on platelet aggregation and serotonin release
reaction. Life Sciences 16, 561–571.
FISHER AA (1986) Contact Dermatitis. Philadelphia: Lea and
Febiger.
FLECKNALL PA (1987) Laboratory Animal Anaesthesia, pp.
159–223. London: Academic Press.
GELFANT S, CANDELAS GC (1972) Regulation of epidermal
mitosis. Journal of Investigative Dermatology 59, 7–12.
GROTENDORST G (1992) Chemoattractants and growth factors.
In Wound Healing; Biochemical and Clinical Aspects (ed. Cohen
IK, Diegelmann RF, Lindblad WJ), pp. 237–246. Philadelphia:
W. B. Saunders.
HALSTEAD JA, SMITH JC, IRWIN MI (1974) A conspectus of
research on zinc requirements in man. Journal of Nutrition 104,
345–378.
HEMBRY RM, MURPHY G, REYNOLDS JJ (1985) Immuno-
localisation of tissue inhibitor of metalloproteinases (TIMP) in
human cells. Journal of Cell Science 73, 105–119.
HENG MK, SONG MK, HENG MCY (1993) Reciprocity
between tissue calmodulin and cAMP levels : modulation by
excess zinc. British Journal of Dermatology 129, 280–285.
HENNINGS H, MICHAEL D, CHENG C (1980) Calcium
regulation of growth and differentiation in mouse epidermal cells
in culture. Cell 19, 245–254.
HUNT TH, HUSSAIN Z (1992) Wound environment. In Wound
Healing: Biochemical and Clinical Aspects (ed. Cohen IK,
Diegelmann RF, Lindblad WJ), pp. 274–281. Philadelphia:
W. B. Saunders.
JARVIS PM, GALVIN DAJ, BLAIR SD, MCOLLUM C (1987)
How does calcium alginate achieve haemostasis in surgery?
Thrombosis and Haemostasis 58, 80.
KILCULLEN JK, QUAN P, CHANG TH, LEVENSON SM,
STEINBERG JJ (1998) Non-viable Staphylococcus aureus can
stimulate macrophage recruitment, angiogenesis, fibroplasia and
collagen accumulation in wounded rats. Wound Repair and
Regeneration 6, 149–156.
KLEVAY LJ (1975) The ratio of zinc to copper of diets in the
United States. Nutrition Reports International 11, 237.
LANSDOWN ABG (1985) Morphological variations in keratin-
ising epithelia in beagle dogs. Veterinary Record 116, 127–130.
LANSDOWN ABG (1993) The influence of zinc oxide on the
closure of open wounds: an experimental study. International
Journal of Cosmetic Science 15, 83–85
LANSDOWN ABG, PAYNE MJ (1994) An evaluation of the local
reaction and biodegradation of calcium sodium alginate (Kalto-
stat2) following subcutaneous implantation in the rat. Journal
of the Royal College of Surgeons, Edinburgh 39, 284–288.
LANSDOWN ABG (1995) Physiological and toxicological changes
in the skin resulting from the action and interaction of metal ions.
CRC Critical Reviews in Toxicology 25, 397–462.
LANSDOWN ABG (1996) Zinc in the healing wound. Lancet 347,
706–707.
LANSDOWN ABG, PATE P (1993) Animal models in wound
healing in the skin. In Proceedings of the 5th Symposium of
the Federation of European Laboratory Science Associations,
Brighton. pp. 395–397.
LANSDOWN ABG, SAMPSON B (1996) Dermal toxicity and
percutaneous absorption of cadmium in rats and mice. Lab-
oratory Animal Science 46, 549–554.
LANSDOWN ABG, SAMPSON B (1997) Trace metals in
keratinising epithelia in beagle dogs. Veterinary Record 141,
571–572.
LANSDOWN ABG, SAMPSON B, LAUPATTARAKASEM P,
VUTTIVIROJANA A (1997) Silver aids healing in the sterile
skin wound: experimental studies in the laboratory rat. British
Journal of Dermatology 137, 728–735.
LINDEMAN RD, MILLS BJ (1980) Zinc homeostasis in health
and disease. Mineral and Electrolyte Metabolism 3, 223–236.
LYNCH SE (1991) Interactions of growth factors in tissue repair.
In Clinical and Experimental Approaches to Dermal and Epidermal
Repair: Normal and Chronic Wounds (ed. Barbul A, Caldwell
MD, Eaglestein WH, Hunt TK, Marshall D, Pines E et al.),
pp. 3412–337. New York: Wiley-Liss.
MADDEN JW, PEACOCK EE (1968) Studies on the metabolism
of collagen during wound healing. I. Rate of collagen synthesis
and deposition in cutaneous wounds in the rat. Surgery 64,
228–294.
MAST B (1992) The skin. In Wound Healing: Biochemical and
Clinical Aspects (ed. Cohen IK Diegelmann RF, Lindblad WJ),
pp. 344–355. Philadelphia: W. B. Saunders.
MENON GK, GREYSON S, ELIAS PM (1985) Ionic calcium
reservoirs in mammalian epidermis : ultrastructural localisation
by ion-capture cytochemistry. Journal of Investigative Derma-
tology 84, 508–512.
Trace metals and skin wound healing 385
MOYNAHAN EJ (1974) Acrodermatitis enteropathica: a lethal
inherited human zinc deficiency disorder. Lancet 1, 399–400
MULDER GD (1991) If wounds could talk. In Clinical and
Experimental Approaches to Dermal and Epidermal Repair:
Normal and Chronic Wounds (ed. Barbul A, Caldwell WH,
Hunt TH, Marshall D, Pines E et al.), pp. 55–66. New York:
Wiley-Liss.
NICKOLOFF BJ, VARANI J, MITRA RS (1991) Modulation
of keratinocyte biology by gamma-interferon: relevance to cu-
taneous wound healing. In Clinical and Experimental Approaches
to Dermal and Epidermal Repair: Normal and Chronic Wounds
(ed. Barbul A, Caldwell MD, Eaglestein WH, Hunt TK, Marshall
D, Pines E, Skover G), pp. 141–154. New York: Wiley-Liss.
NWOMEH BC, LIANG H-X, DIGELMANN RF, COHEN IK,
YAGER DR (1998) Dynamics of the matrix metalloproteinases
MMP-I and MMP-8 in acute open human dermal wounds.
Wound Repair and Regeneration 6, 127–134.
PALMITER RD, FINDLEY SD, WHITMORE TE, DURNHAM
DM (1992) MT-III, a brain specific member of the metallo-
thionein gene family. Proceedings of the National Academy of
Sciences of the USA 89, 6333–6337.
PAUWELS M, VAN-WEYENBERGH J, SOUMILLION A,
PROOST P, DE LEY M (1994) Induction by zinc of specific
metallothionein isoforms in human monocytes. European Journal
of Biochemistry 220, 105–110.
PINKUS H (1970) The direction of growth of human epidermis.
British Journal of Dermatology 83 556–564.
QUAIFE CJ, FINDLEY SD, ERICKSON JC, FROELICK GJ,
KELLY EJ, ZAMBROWICZ BP et al. (1994) Induction of a new
metallothionein isoform (MT-IV), occurs during differentiation
of stratified epithelia. Biochemistry 33, 7250–7259
RONNAN M, SHUSTER S, KLIN B (1993) Treatment of
epidermal cysts with Solcoderm (a copper ion and acid solution).
Clinical and Experimental Dermatology 18, 500–503.
ROWE A, FARRELL AM, BUNKER CB (1997) Constitutive
endothelial and inducible nitric oxide synthase in inflammatory
dermatoses. British Journal of Dermatology 136, 18–23.
SUCVI I, PRODAN L, LAZAR V (1981) Research on copper
poisoning. Medical Law 3, 190.
WAHL GF, WAHL SM (1992) Inflammation. In Wound Healing:
Biochemical and Clinical Aspects (ed. Cohen IK, Diegelmann RF,
Lindblad WJ), pp. 40–62. Philadelphia: W. B. Saunders.
WINGE DR, NIELSON KB (1985) Distinct metal binding
configurations in metallothionein. In Trace Elements in Man and
Animals (ed. Mills CF, Bremner I, Chesters JK), pp. 307–311.
Slough: CAB.
YANG YY, WOO ES, REESE CE, BAHNSON RR, SAIJO N,
LAZO JS (1994) Human metallothionein isoform gene ex-
pression in cisplatin-sensitive and resistant cells. Molecular
Pharmacology 89, 453–460.
YUSPA SM, KILKENNY AE, STEINERT PM (1989) Expression
of murine epidermal cell markers is tightly regulated by extra-
cellular calcium concentrations in vitro. Journal of Cell Biology
109, 1207–1217.
386 A. B. G. Lansdown, B. Sampson and A. Rowe