Is there life in the horny layer? Dihydropyridine and ryanodine receptors in the skin of female and male chickens (Gallus domesticus)

By its primary function, skin is a protective barrier. It provides a first frontier of defence against a wide variety of physical and organic environmental threats and against the loss of body fluids. The protective characteristics of the skin are associated with its anatomical construction,the integrity of its proteinaceous cell envelope, the extracellular lipid barrier (Norlen, 1999; Elias, 2004), and its ability to withhold water in the horny layer (stratum corneum, SC), which seems to be essential for normal epidermal function(Segre, 2003; Rawlings and Harding, 2004). Skin also functions as a secretory organ by excreting fluid and sebum from specialized glands. In birds that are devoid of glands covering the whole body surface, secretion takes place by structural and functional plasticity of the interfollicular epithelium (Menon et al.,1981; Menon et al.,1996; Peltonen et al.,1998; Peltonen et al.,2000).

The skin may also have a role in calcium metabolism. It has been established that the amount of ionic Ca²⁺ is strictly regulated,and that blood, interstitial fluid and tissues, especially bone, form a functional unit that buffers the excessive change in the amount of biologically active Ca²⁺. In chickens, the major fluctuations in ionic calcium concentration take place in laying females within the daily cycle, the lowest values being displayed at the time of egg calcification(Simkiss, 1967; Luck and Scanes, 1979; Etches, 1996). At the same time, at the site of deposition the local uterine Ca²⁺ values may be increased four- to twelvefold, creating a local high-calcium microenvironment (Arad et al.,1989). In addition to the uterus, skin may also contain calcium microenvironments. It was shown that in a regular `layer's diet', the amount of ionic Ca<sup>2+</sup> in the cutaneous interstitial fluid of male chickens started to increase after the completion of growth, while in females the Ca<sup>2+</sup> remained in equilibrium between these compartments until they came out of lay (Peltonen et al.,2006). As shown in cardiomyocytes, calcium overload may cause enhanced calcium entry (Katra and Laurita,2007). If this holds for the skin, accumulation of ionic Ca<sup>2+</sup> may enhance the influx of Ca<sup>2+</sup> into the skin cells. During cell differentiation, Ca<sup>2+</sup> may be either `entrapped' in the corneocytes and lost via desquamation, or associated with some secretory pathway and released into the extracellular space at the transitional interface between the stratum transitivum and corneum (ST-SC). To test this hypothesis, we determined whether transport systems for plasmalemmal Ca²⁺ influx and Ca²⁺ release for possible intracellular signalling exist in the skin of chickens, and whether the densities of the receptors for these vary in relation to sex, the level of the nutritional Ca²⁺ input, the concentration of ionic Ca²⁺ outside skin cells, or the skin layer. Regarding Ca²⁺ influx, we examined the density and distribution of the dihydropyridine receptor (DHPR), the ion-conducting α₁-subunit of L-type calcium channels, by labelling it with a high-affinity enantiomer of DHP. As to the intracellular Ca²⁺ release, we examined the density and distribution of the ryanodine receptor (RyR), also by labelling it with a high-affinity enantiomer of ryanodine. Voltage-gated L-type Ca²⁺ channels typically mediate Ca²⁺ influx in response to depolarization of the plasma membrane. Coupled to or in close proximity with DHPRs, Ca²⁺-releasing RyRs form an apparatus for Ca²⁺-induced calcium release that regulates cell functions such as contraction, secretion, neurotransmission and gene expression in many different cell types(Catterall et al., 2005; Coronado et al., 1994).

We also measured the activity of the enzyme alkaline phosphatase (ALP) to indicate Ca²⁺ utilization. ALPs are a group of enzymes that are widely expressed within tissues and catalyse hydrolysis of monophosphate esters at alkaline pH. They are present in soluble and membrane-bound form,the latter anchored by glycosyl phosphatidylinositol to the outer leaflet of the plasma membrane (Moss,1997). Plasma ALP activity has been associated with bone-forming activity by osteoblasts, and with other tissues that have high cellular turnover (Bell, 1971). Since the tissue-non-specific ALP is synthesized in the skin(Yamashita et al., 1987; Crawford et al., 1995; Hui and Tenenbaum, 1995), and may be involved in physiological and pathological mineralization in tissues other than bone or cartilage (Hui et al.,1997), we also measured its activity in the cutaneous interstitial fluid.

All the experiments conformed to the guidelines for proper animal care and use and had been authorized by the local ethical committee of University of Helsinki, Finland.

Eight female and male White Leghorn chickens (Gallus gallus domesticus Linnaeus 1758) were raised from the age of 1 day in a temperature-controlled environment and later housed in separate 20 m² rooms at room temperature (20–23°C), and under a constant 12 h:12 h light:dark photoperiod. Animals were able to move freely and use roosts. Experiments were done in two successive stages: stage 1 and stage 2. At stage 1, the sexually mature, adult animals were fed ad libitum with food that contained Ca²⁺ at 3.5% of the total amount (dry weight) of nutrients. At stage 2, reduction of the available amount of Ca²⁺ in the gut (Ca²⁺ input) was accomplished in both sexes by changing to a low-calcium diet of 0.1% (dry weight). At stage 1, both sexes had reached their maximum body mass (BM), which was 1.86±0.142 kg in females and 2.62±0.135 kg in males. At stage 2,females came off lay and their BM decreased by approximately 360 g, whereas BM remained unaltered in males. At all times, chickens were provided with oyster shells ad libitum as a recognizable calcium source.

Skin samples, suction blister fluid (SBF) and blood were collected under injectional anaesthesia. The chickens were given intramuscular ketamine(Ketalar®, Pfizer, Espoo, Finland) in combination with xylazine(Rompun®, Bayer, Leverkusen, Germany) in doses of 20 mg kg^–1 and 5 mg kg^–1, respectively. SBF was collected by the method first developed for humans by Kiistala and Mustakallio(Kiistala and Mustakallio,1964) and modified for birds by Peltonen et al.(Peltonen et al., 2006). The suction cups were attached on the thoracic apterial skin. Whole blood and serum samples were taken from the brachial vein by aspiration into 2 ml lithium heparin syringes (Pico70, Radiometer Medical A/S, Copenhagen,Denmark). Blood samples for serum were removed to 500 μl vials through a short 21-22G needle. Serum was separated and collected after 1–2 h of incubation at room temperature and centrifugation.

Frozen skin samples treated with 2% paraformaldehyde and 2% glutaraldehyde were cut into 8 μm cryosections (N=8 per sample) at–20°C and incubated in 20 nmol l^–1 high-affinity(–)-enantiomer of dihydropyridine labelled with orange fluorophore and 0.5 μmol l^–1 high-affinity (–)-enantiomer of ryanodine labelled with green fluorophore (Molecular Probes, Leiden,Netherlands) for 90 min, and processed as previously described by Mänttäri et al.(Mänttäri et al.,2001). The control samples were preincubated for 10 min in 10μmol l^–1 of the DHPR blocker nifedipine (Sigma, St Louis,MO, USA) and 50 μmol l^–1 of the RyR blocker dandrolene(Sigma) prior to addition of labelling solution. Images of the sections were obtained with a confocal laser scanning microscope (LSM-5 Pascal, Zeiss, Jena,Germany) by using excitation at 543 nm for DHPRs and 488 nm for RyRs.

The distribution and the relative amount of membrane-bound calcium in the skin were determined in five chickens (N=5) with a fluorescent Ca²⁺-sensitive probe, chlorotetracycline (CTC). Samples were fixed in an ice-cold mixture of 2% paraformaldehyde and 2% glutaraldehyde, 90 mmol l^–1 potassium oxalate and 1.4% sucrose in 0.1 mol l^–1 sodium cacodylate buffer at pH 7.4. After rinsing in phosphate buffer (pH 7.4), the samples were immersed in 2% potassium pyroantimonate/potassium hexahydroxyantimonate at 4°C. After rinsing three times for 10 min with alkaline distilled water (pH 10), samples were frozen in liquid nitrogen. Cryosectioned samples (4 μm) were treated with 100 μmol l^–1 CTC. The CTC fluorescence was determined against the quenched fluorescence of samples treated with 10 mmol l^–1 of the calcium-chelating agent EGTA. Samples were examined with an Olympus fluorescence microscope (Olympus, Tokyo, Japan).

The concentration of ionized Ca²⁺ in adult chickens (stage 1)was measured with ISE (ion-selective electrodes) in fresh whole blood and in SBF as previously described by Peltonen et al.(Peltonen et al., 2006). In short, to maintain pH and ion concentration stability, fresh blood samples were stored in an ice bath without exposure to air until measured, within 2–4 h (Cao et al., 2001). SBF samples were also stored in an ice bath, but in 250 μl Eppendorf vials with an air space. Samples were analysed with a KONE Microlyte 3 device(Thermo Electron Inc., Clinical Chemistry and Automation Systems, Vantaa,Finland). In non-layers and coeval males (stage 2), we used our previous results for reference (Peltonen et al.,2006). Plasma and SBF activity of ALP were measured by Spotchem II analysing system (Arkray Inc., Kyoto, Japan).

One-way analyses of variance (ANOVA) were used for statistical analyses. As a post hoc test, Bonferroni's test or Tukey's test for multiple comparisons was used. The threshold P value for statistical significance was set at 0.05.

The density of DHPRs in the skin of adult female and male chickens was determined by labelling the ion-conducting α₁-subunit with a fluorescently labelled high-affinity enantiomer of dihydropyridine. DHPRs were localized throughout the epidermis and in the dermal blood vessels (Figs 1 and 2). However, the highest densities were measured at the outermost living cell layer and at the SC(Fig. 1A,C,E). The total skin fluorescence was higher in males than in laying females, indicating a higher density of DHPRs in the male skin (P<0.001, Fig. 1A). After reduction of the dietary Ca²⁺ input, male density values decreased down to the level of females' in the viable epidermis and dermis, and below the female values in the SC (Fig. 2A). In females, the selective fluorescence for DHPRs was unaffected by the nutritional Ca²⁺ input or the cessation of laying(Fig. 1A, Fig. 2A).

Similar to DHPRs, the highest fluorescence intensity of labelled RyRs was in the skin surface. However, RyRs were distributed more evenly within the epidermal layers (Fig. 1E,F, Fig. 2E,F). In the mid-epidermis, punctate and intense fluorescence was clustered in proximity to the apical and basal plasma membrane. These horizontal lines were interrupted by flattened nuclei that were often found vertically aligned(Fig. 1F). In the deeper layers, high intensities were observed on the basal plasma membrane of the basal cells and on the perinuclear endoplasmic reticulum(Fig. 2F). Semi-quantitative analysis showed that the total fluorescence intensity did not differ between sexes. However, there were differences among cell layers; the living epidermis and the upper dermis showed higher intensities in males on the basic calcium diet (Fig. 1B; P<0.001 and P<0.01, respectively). With the low-calcium diet, a marked decrease in the fluorescence intensity was observed in males (Fig. 2B).

The corresponding control samples, incubated with nifedipine or dandrolene,resulted in a loss of staining in all skin layers(Fig. 3).

The amount of biologically active Ca²⁺ in blood was the same in females and males irrespective of the physiological status associated with laying or the relative amount of calcium in the food(Fig. 4). The average Ca²⁺ concentration was 1.46±0.02 mmol l^–1. In contrast to its stability in blood, the amount of ionic Ca²⁺ in the tissue fluid increased significantly in females after the nutritional Ca²⁺ input decreased and they had come out of lay(Fig. 4A). In laying females,Ca²⁺ concentration was similar in SBF and blood. Males displayed higher-than-blood SBF values irrespective of the nutritional Ca²⁺input (Fig. 4B).

The local activity of ALP was shown to be lower in the skin than in the plasma. However, this difference was statistically significant in females only(Fig. 5). After decreasing the nutritional Ca²⁺ input, a significant sex-specific difference in the plasma and SBF activities was observed; females displayed significantly higher enzyme activity than males. In three females ALP activity of the plasma was given the value of 1500 U l^–1 since the measured activities exceeded the highest value of the linear range set for ALP activity.

CTC-labelled skin samples of adult male and female chickens suggest that Ca²⁺ is present in all skin layers, including SC(Fig. 6).

Our results suggest that both voltage-dependent L-type Ca²⁺channels and RyRs exist in the skin cells of adult female and male chickens. The existence of membrane transport systems for calcium influx and intracellular release implies a more versatile functional capacity of the avian epidermis than had previously been thought. The horny layer would not be a `dead moiety' since the Ca²⁺ entrapped within the corneocyte could be secreted via desquamation, or it may be secreted and recycled at the transitional interface by a secretory pathway including Ca²⁺ influx channels and peripherally located, subplasmalemmal Ca²⁺ release channels. In keeping with the physiology of the L-type channel, the epidermis may also serve as a superficial sensor for the electric potential in an epithelium practically devoid of sensory innervation(Saxod et al., 1995; Cahoon and Scott, 1999; Cahoon-Metzger et al.,2001).

Early evidence about the existence of L-type Ca²⁺ channels in the skin was obtained by Reverdin et al.(Reverdin et al., 1989) in the apical membrane of human keratinocytes by detecting inward-directed current that was activated by the DHP agonist Bay-K8644. Later, indirect evidence was obtained by using specific channel agonists, nifedipine and verapamil, which were shown to block the delaying effect of an increased extracellular concentration of Ca²⁺ on the barrier repair(Lee et al., 1992). Recently,the expression of the ion-conducting subunit of L-type channels has been demonstrated in the keratinocytes of mutant hairless mice and in cultured neonatal human keratinocytes (Denda et al., 2006). However, data on cutaneous Ca²⁺ channels are still scarce. There is no evidence of the presence of RyRs in the skin, or reports on the existence of either channel in vertebrate species other than mammals. Nevertheless, the functional role of these channels in the epidermal plasticity of the skin may be very important.

We hypothesized that, at the interface between the transitional and the cornified cell layer, Ca²⁺ may be secreted into the extracellular space or it may be deposited in the transitional cells and thereafter `lost'via corneocyte desquamation at the surface of the skin. To support this hypothesis, our results show that the density of DHPRs and RyRs increased at the transition interface, and the density in males was higher when their nutritional Ca²⁺ input was the same as in egg-producing females. Thus, a higher DHPR and RyR density, together with a higher SBF Ca²⁺ concentration ([Ca²⁺]_SBF), implies enhanced activity of the skin cells for Ca²⁺ influx and intracellular Ca²⁺ release. However, the amount of free Ca²⁺ in extracellular fluid per se does not seem to affect receptor density; a considerable decrease in nutritional Ca²⁺ input resulted in an increase in female [Ca²⁺]_SBF without an effect on receptor density (Fig. 2, Fig. 4A), and a sustained higher-than-blood [Ca²⁺]_SBF in males was associated with a lowered receptor density when nutritional Ca²⁺input was decreased (Fig. 2 and Fig. 4B).

The expression of DHPR and RyR genes is not tissue specific, even though they are predominantly expressed and their functional connection is best described in skeletal, cardiac and smooth muscle cells(Franzini-Armstrong, 2004). Until the present study, DHPRs had not been shown to coexist with RyRs in non-excitable epithelial cells. Previously, DHPRs have been visualized in the epidermis of mice and neonatal humans(Denda et al., 2006), and their functionality has been demonstrated in the pigment epithelium of the retina in rats, monkeys and humans (Ueda and Steinberg, 1993; Ueda and Steinberg, 1995; Mergler and Strauss, 2002). Expression of RyRs has been shown in the epithelial cells of the cornea (Socci et al., 1993) and in the epithelial cells of the kidney tubulus(Tunwell and Lai, 1996). Together, previous and our present findings suggest that intracellular signalling via RyRs is not exclusively confined to excitable cells,and that DHPRs and RyRs might be spatially and functionally connected in non-excitable skin cells.

According to our results in chickens, the possible channel units of DHPRs and RyRs in a sebokeratinocyte are peripherally located. This spatial relationship seems to resemble the arrangement of the smooth muscle cell in which the sarcoplasmic proteins, calsequestrin and RyRs colocalize with DHPRs in numerous, peripherally located sites within the caveolar domains(Moore et al., 2004; Pucovsky and Bolton, 2006). Due to the native arrangement of the stratified epidermis in our study, the exact array of DHPRs on the plasma membrane could not be revealed. However,RyRs were located in the proximity of the plasma membrane in horizontally aligned clusters, indicating the possible sites where the two channels might interact via spatial proximity. In a single smooth muscle cell of the urinary bladder, DHPRs have been shown to occupy the plasmalemma in longitudinal stripes that overlap almost entirely with the corresponding stripes formed by labelled RyR proteins(Moore et al., 2004). The authors estimated that a single cell may display approximately 100–200 overlapping patches, i.e. sites where the transduction of membrane depolarization into the release of Ca²⁺ from the sarcoplasmic reticulum (SR) takes place. In avians, cardiac cells also seem to display the pattern of peripheral location of DHPR/RyR units. Excitation–contraction units develop via docking of the SR membrane to the surface membrane due to the lack of intruding T-tubules in these cells; the appropriate locations and quantities of DHPRs and RyRs within the unit are established gradually during development (Sun et al.,1995; Protasi et al.,1996).

The avian epidermis is formed by columns of non-excitable epithelial cells that move and differentiate via a vertical pathway from the basement membrane to the surface of the skin. Its sensory functions seem to be limited since it is inhibitory, or at least non-permissive, to the growth of sensory nerves (Saxod et al., 1995; Cahoon and Scott, 1999; Cahoon-Metzger et al., 2001),and its outermost layer is generally categorized as a dead moiety of the skin. Nevertheless, according to our findings, the strongest specific fluorescence for the DHPRs is located in the surface of the skin, implying a voltage-sensing function. Previously, maintaining the surface electric potential at a negative value of about –3 mV was shown to be essential for the homeostasis of the skin in mice(Denda et al., 2001). It was suggested that L-type Ca²⁺ channels might be involved because specific channel blockers for the α₁-subunit helped to restore the ionic balance when it was disrupted(Denda et al., 2006).

The ionic environment of the avian epidermis might differ considerably from that of mammals. In pigeons (Columba livia Gmelin 1789),Ca²⁺ has been shown to accumulate in the corneocytes at the transitional interface and remain in the SC up to the outermost cell layers(Peltonen et al., 2003). Our preliminary results obtained using CTC, the fluorescent marker of calcium,support the idea that calcium is also preserved in the chicken SC and might colocalize with DHPRs and RyRs in the epidermis. Given that the composition of the extracellular fluid affects the function of channels sensitive to voltage,an increased concentration of monovalent cations outside the cell is likely to alter the open–closed probability of these channels, and thus the cellular response. To date, only little is known about the character, location and magnitude of the L-type Ca²⁺ currents in skin cells.

We measured plasma and SBF activity of ALP in order to estimate the level of ALP-mediated Ca²⁺ utilization in adult female and male chickens. Since it is known that bone remodelling is accompanied by local fluctuations of free extracellular Ca²⁺ concentration(Parfitt, 1987), we expected to see changes in the activity of ALP in relation to nutritional Ca²⁺ input and Ca²⁺ concentration. Our results show that there was a clear sex-specific difference in the response to the reduced nutritional Ca²⁺ input: the plasma and skin activity of ALP increased in females in association with the increased Ca²⁺concentration of SBF but not blood, whereas in males the activity of ALP was unaffected by the level of nutritional Ca²⁺ input and[Ca²⁺]_SBF values were displayed at a sustained,higher-than-blood level. It is clear from these results that both sexes are able to regulate the amount of free Ca²⁺ in blood at a normal level despite the change in diet. Because the chickens were feeding freely, the extent of the possible compensation for low-calcium feed by oyster shells remains unknown. However, the reduction in Ca²⁺ input was functionally effective since all females went out of lay. Together, our findings on ALP activity suggest that either physiological or pathological ALP-mediated mineralization could take place in the skin. These processes are most probably confined to the dermis because the activity measured in the epidermis was found to be less than 1% of the activity of the underlying dermis (Mier and van Rennes,1982). In general, the cutaneous ALP seems to be of a tissue-non-specific type, present in the dermal condensations of developing feather germs, hair follicles, capillaries and sweat glands(Mier and van Rennes, 1982; Crawford et al., 1995; Saga and Morimoto, 1995). Furthermore, it is present in both a soluble and a membrane-bound form(Mier and van Rennes, 1982). In the present study, the cellular location of ALP activity could not be assessed. However, feather follicles and sweat glands are excluded as the sites of enzyme activity because the experimental area is void of these structures.

Since the skeletal growth was already completed in both sexes, we expected that fluctuations in the plasma activities of ALP would be sex specific and probably associated with egg shell calcification. This suggestion of the association between ALP activity and egg shell calcification has been put forward by several authors (Wilcox et al.,1963; Singh et al.,1983; Carpenter et al.,2001; Harr, 2002),but opposed by others (Arad et al.,1989; al-Bustany et al.,1998). Our findings show no difference in plasma activity between laying females and males of the same age. Thus, our results are in line with studies indicating that the plasma activity of ALP is not associated with egg shell calcification but, instead, with the physiological state that ensues at the cessation of laying. The cause of this increased ALP activity in both plasma and SBF is not known. In avians, increases in the plasma level of ALP activity appear to be associated with osteoblast or bone-forming activity. Such activity is present especially during skeletal growth after hatching(Campbell and Coles, 1986; Vinuela et al., 1991; Lumeij, 1997; Tilgar et al., 2004a). In the present study, skeletal growth was complete, excluding somatic growth as a causative factor. After skeletal growth is complete, total ALP activity, and especially the activity of its bone-derived isoform(Tilgar et al., 2004a; Tilgar et al., 2004b),decreases, given that the amount of calcium is adequate for normal growth(Meluzzi et al., 1992; al-Bustany et al., 1998; Jurani et al., 2004). Since increased ALP activity has also been associated with osteoblast activity in general, the observed increase in the females that have gone out of lay might be caused by the reversal of the formation of the medullary and structural bone. When laying ceases, medullary bone disappears and the formation of structural bone recommences accompanied by increased osteoblast activity(Whitehead, 2004).

In conclusion, the presence of L-type Ca²⁺ channels for ion influx and RyRs for Ca²⁺ release in the skin cells of female and male chickens suggests that their skin has the functional capacity for accumulation, storage and secretion of Ca²⁺. Thus, the higher density of these channels in males receiving excessive Ca²⁺ implies an increased capacity for Ca²⁺ influx and intracellular processing. Whether there are structural and functional interactions between DHPRs and RyRs in the sebokeratinocytes is not known. However, the peripheral location and the high receptor density at the site of exocytosis of the lamellar bodies in the interface between the transitional and horny layers imply a role in a signalling pathway for secretion. Together, the horny layer seems to be more a live than a dead moiety of the skin! Variations in the ALP activity suggest that Ca²⁺ is utilized in the skin. In females, the increased activity may be connected with the reformation of structural bone after the cessation of laying.

We wish to thank the Animal Care and Management unit and the Department of Basic Veterinary Sciences of the Faculty of Veterinary Medicine, University of Helsinki, for providing facilities. We also thank M. Sc. Katja Anttila of University of Oulu, for technical support.