The cannabinoid receptors system in horses: Tissue distribution and cellular identification in skin

Abstract Background The endocannabinoid system (ECS) is composed of cannabinoid receptors type 1 (CBR1) and type 2 (CBR2), cannabinoid‐based ligands (endogenous chemically synthesized phytocannabinoids), and endogenous enzymes controlling their concentrations. Cannabinoid receptors (CBRs) have been identified in invertebrates and in almost all vertebrate species in the central and peripheral nervous system as well as in immune cells, where they control neuroimmune homeostasis. In humans, rodents, dogs, and cats, CBRs expression has been confirmed in the skin, and their expression and tissue distribution become disordered in pathological conditions. Cannabinoid receptors may be a possible therapeutic target in skin diseases. Objectives To characterize the distribution and cellular expression of CBRs in the skin of horses under normal conditions. Animals Fifteen healthy horses. Methods Using full‐thickness skin punch biopsy samples, skin‐derived primary epidermal keratinocytes and dermal‐derived cells, we performed analysis of Cnr1 and Cnr2 genes using real‐time PCR and CBR1 and CBR2 protein expression by confocal microscopy and Western blotting. Results Normal equine skin, including equine epidermal keratinocytes and dermal fibroblast‐like cells, all exhibited constant gene and protein expression of CBRs. Conclusions and Clinical Importance Our results represent a starting point for developing and translating new veterinary medicine‐based pharmacotherapies using ECS as a possible target.


| INTRODUCTION
The endocannabinoid system (ECS) includes cannabinoid receptors (CBRs) and their ligands. 1 The CBRs are represented by cannabinoid receptor type 1 (CBR1) and type 2 (CBR2), members of the rhodopsin-like superfamily of 7-transmembrane G protein-coupled receptors (GPCRs). 2,3 They are evolutionarily conserved and have been characterized in all vertebrates and in many invertebrate species. 4,5 Initially, it was assumed that CBR1 is predominantly distributed within different areas of the central nervous system (CNS), whereas CBR2 dominates on immune cells. 6 However, it is currently known that CBRs are present in both locations as well as in peripheral tissues, where their balanced expression regulates the neuroimmunological homeostasis of multiple tissues. 7 The CBRs activation and signal transduction pathways are initiated after binding of 3 general ligand subtypes. 8 First, naturally-synthesized endocannabinoids are represented by anandamide (N-arachidonoylethanolamine) and 2-arachidonoylglycerol (2-AG), which bind classical CBRs. In turn, another endogenous ligand, palmitoylethanolamide (PEA), has an affinity to non-CBRs restricted to peroxisome proliferator-activated receptor alpha (PPAR-α). Palmitoylethanolamide also may act on G protein-coupled receptor 55 (GPR55) and transient receptor potential cation channel subfamily V member 1 (TRPV1). 9,10 Non-CBRs were identified in experimental studies using CBR1 À/À and CBR2 À/À knockout mice and are represented by others including TRPV1, GPR55, PPAR-α, peroxisome proliferator-activated receptor gamma (PPAR-γ), transient receptor potential A member 1 (TRPA1), and serotonin 1A receptor or 5-HT1A receptor (5-HT1). Exogenous phytocannabinoids, cannabidiol (CBD), and the psychoactive (À)-Δ9-tetrahydrocannabinol (Δ9-THC) are natural compounds of Cannabis sativa, a plant used in medicine since ancient times. Finally, cannabinoid-based compounds deserve the most attention, featuring well-defined specificity and pharmacokinetics, which partially or entirely inhibit or activate CBRs signaling. 11 The CBR-based treatments have been applied successfully in several pathological conditions in humans and in animal disease models. 2 Further investigation has indicated that several common and chronic proinflammatory skin diseases, such as atopic dermatitis, psoriasis, fibrotic disorders, and skin cancers, can alter the expression of CBRs. [12][13][14][15] In turn, experimental studies have shown that disordered cutaneous homeostasis might be established in a CBRs-dependent manner. 13,16,17 Accordingly, peripheral application of cannabinoid-related compounds has shown therapeutic benefits, decreasing associated comorbidities and reconstituting proper skin tissue architecture and physiology. 18,19 These CBR-based treatments recently have attracted considerable attention in equine veterinary medicine after their use in other vertebrate species, including dogs, cats, and pigs. [20][21][22][23][24] Unfortunately, knowledge about the ECS in horses is still limited to dorsal root ganglion (DRG) neurons and equine genome sequencing analysis. [25][26][27][28] No studies on CBRs distribution and cellular expression in the cutaneous milieu are available. We present a detailed and advanced in situ and in vitro characterization of the equine ECS, using skin punch biopsy samples and primary keratinocyte and fibroblast cultures from horses cultures. Real-time PCR, Western blotting (WB), and confocal microscopy were used to estimate CBR mRNA transcripts and protein concentrations. These preliminary results may suggest strategies for developing and implementing peripheral CBR-based dermatological treatments for horses.

| Animals
Skin samples were collected under sterile conditions from the metacarpal area of healthy horses (n = 15). The characteristics of the horses are presented in the Table S1. All laboratory procedures described below are detailed in the Data S1.

| Sample collection and tissue processing
Skin samples were collected into stabilizing reagents and appropriate fixatives, depending on the procedure. For gene and protein expression analysis, skin samples were placed in molecular biology reagent (RNAstay, A&A Biotechnology, Gdansk, Poland) and frozen at À80 C. For histology, skin biopsy samples were placed in 10% buffered formalin for 48 hours, and formalin-fixed paraffin-embedded (FFPE). For frozen sections, biopsy samples were fixed in 4% paraformaldehyde (4% PFA, POCH S.A., Gliwice, Poland) and embedded in OCT medium (Thermo Fisher Scientific, Waltham, Massachusetts, USA) as described previously. 29 Finally, the skin was placed in culture medium to initiate skin-derived primary cell expansion. Moreover, the brains from 3 healthy horses were collected as reference material.

| Histology
The FFPE 5-μm paraffin sections were subjected to a standard procedure of deparaffinization. Each sample was stained by hematoxylin and eosin (H&E; Roth GmbH, Karlsruhe, Germany and POCH S.A., Gliwice, Poland).

| Confocal microscopy
The OCT 12-μm-thick frozen sections were prepared on glass slides (Ultra Superfrost Plus, Thermo Fisher Scientific, Waltham, Massachusetts, USA) and further subjected to an immunofluorescence protocol, with minor modifications. 29 After post-fixation and incubation in the T A B L E 1 Primary non-conjugated and HRP-conjugated antibodies as well as secondary antibodies fluorochrome-conjugated for immunofluorescence (IF) microscopy and HRP-conjugated for Western-blot (WB) analysis   Table 1.

| Histological and morphological analyses of skin
The skin of the studied horses was normal, without visible histopathological changes. The epidermis consisted of 4 to 7 layers of epithelial cells, covered by an outer keratinized layer ( Figure 1).

| In-situ distribution and expression analysis of CBRs in whole skin and brain tissue reference material
The CBRs distribution within whole equine skin tissue sections was determined using confocal microscopy. Considering the high expression of CBRs in the brain of other species, the cortex sections were used as reference material, and in the case of both receptors, positive immunoreactivity was confirmed. The CBR1 was present in different sets of neuronal cells and in perivascular regions of microcirculation within respective endothelial cells, in contrast to CBR2 immunoreactivity, in which lower expression was observed ( Figure S1).

| Epidermal expression of CBRs
The equine epidermis is a stratified squamous epithelium mainly com-

| Dermal expression of CBRs
The equine dermis has a heterogeneous cell population of which fibroblasts are the primary cells supported by skin dendrocytes, principal elements of the equine cutaneous immune system. 31 The dermis is

| Gene expression analysis
We introduced 3 common equine housekeeping genes (HKGs) for internal controls, namely glyceraldehydes 3-phosphate dehydrogenase (Gapdh), actin-β (Actb), and β2-microglobulin (β2m) ( Table 2) Our analysis indicated the presence of both Cnr1 and Cnr2 in equine brain, skin, keratinocytes, and fibroblasts. We noticed the significantly higher relative expression of CBRs mRNA in brain tissue for both genes compared to the remaining analyzed material, verified using the Kruskal-Wallis ANOVA test (P-value = .007 for Cnr1, Pvalue = .0005 for Cnr2), and results including significant differences are summarized in Table 3.
Although the transcript levels of Cnr1 and Cnr2 in skin, fibroblasts, and keratinocytes were low, they were still detectable, confirming mRNA presence of the investigated genes in those samples.
T A B L E 2 Equine forward and reverse sequences for target Cnr1 and Cnr2 genes and HKGs with their respective FAM-labeled probes Gene Primer sequences: forward (F: 5 0 -3 0 ) and reverse (R: 3 0 -5 0 ) UPL probes NCBI accession number F I G U R E 9 Comparison of the relative expression level of Cnr1 and Cnr2 in equine brain (Eq. Brain); skin (Eq. Skin); keratinocytes (Eq. Keratinocytes) and fibroblasts (Eq. Fibroblasts). The transcript level in each biological sample is displayed as the mean value of 2 ÀΔCT ± SEM

| DISCUSSION
The animal ECS has been identified in all vertebrates and many invertebrate species, where it participates in the regulation of many biological mechanisms. 5 The tissue distribution and expression patterns of CBRs are comparable and well characterized in several mammalian species, but they still need to be investigated in equine tissues. 7 Our strategy to identify CBRs in the equine cutaneous system has its roots in observations in human medicine and basic science investigations, where in vivo and in vitro approaches clearly show that CBRs are an exciting target for dermatological treatments in humans. 13,33 Moreover, considering their utilization and promising results as a therapeutic target in many medical aspects, their application in equine veterinary medicine also is of great interest. 5,33,34 In that field, treatments, which are still under development, are highly promising because their main advantage is the limited interaction of administrated cannabinoid-based compounds with the CNS. Unfortunately, many CBRs-targeted drugs cross the blood-brain barrier (BBB), resulting in many undesired adverse effects. 15 As in humans, expression of CBR1 and CBR2 in the normal skin of animals is balanced and controls cutaneous homeostasis. 13,35 Several studies show that the application of different cannabinoid-based ligands influences CBRs expression. Exogenous application of AEA significantly influences CBR1-expressing keratinocytes in situ as well as in the skin-derived cells via the downregulation of keratins 6 and 16 and inhibition of epidermal keratinocyte proliferation. 36 The endogenous release of AEA inhibits protein kinase C activity in a CBR1-dependent manner and modulates keratinocyte differentiation programs. 16 In mice with CBR1 keratinocyte-specific deletion, the experimental induction of contact hypersensitivity response occurs with significantly higher intensity than normal stable skin expressed CBR1.
Moreover, receptor deletion in keratinocytes results in the amplified secretion of C-X-C motif chemokine ligand 10 (CXCL10) and C-C motif chemokine ligand 8 (CCL8), key chemokines responsible for myeloid cell infiltration. 17 In turn, their infiltration may up-regulate CBR2, which has been confirmed in an experimental skin model of wound healing. 37 Using animal models, it was shown that neutrophil activity is increased in CBR2 À/À mice, and under normal conditions, CBR2-dependent recruitment is impaired after agonist treatment. 38 42 In the normal skin of dogs and cats, CBRs tissue expression patterns are comparable and concern DRG terminals and skin cells, in agreement with previous findings. 23,24 The expression and distribution of CBRs along DRG neurons, the longest axonal DRG projections in vertebrates, presents therapeutic advantages.
The DRG axons terminate in most epithelial tissues including the skin and are responsible for transmitting sensory-like signals where axonal transport plays an important role. 43,44 One study confirmed CBRs expression along the DRG sensory neurons. 25 Results of that study were consistent with previous observations in other species and those presented by other groups. 21,22,45,46 The DRG nerve terminals play a specific role in CBRs neurobiology and have been investigated in many aspects, mainly neuroinflammation. 47 Studies on rat peripheral tissues show that experimental sciatic nerve injury is an inducible factor for detecting CBRs at mRNA and protein levels. 44,48 There is strong evidence that the peripheral expression of CBRs might result from DRG sensory neurondependent axonal flow, and their delivery from CNS or production in situ is highly regulated by skin conditions. 49 We observed that detecting CBRs transcripts by real-time PCR was low in whole skin samples, whereas levels in primary keratinocyte and fibroblast cultures were even lower, albeit with acceptable detection. At the same time, all gene expression analyses in the whole-skin and primary cultures were performed in comparison to brain reference material, where expression was almost 200-fold higher. Moreover, the molecular biology of CBRs transcript expression and regulation seems to be more complex and dynamic.
The transcript cargo along DRG neurons depends on neurotransmission conditions, its short half-life, and several epigenetic factors, which should be considered with regard to CBRs mRNA regulation and expression. 50,51 These unusual expression patterns of Cnr1 and Cnr2 transcripts in the skin seem reliable and have excellent primer efficiency for target genes. In postnatal skin, PGP 9.5 is still expressed and led to identification of an epithelial neuro-immuno-endocrine cell situated in the basal layers of epidermis. For that reason, PGP 9.5 serves as an excellent connecting marker between neuronal and non-neuronal cells. 29,[57][58][59] Our studies observed heterogeneous PGP 9.5 immunoreactivity in different skin areas, but 2 main staining patterns correspond to cytoplasmic and nerve fiber-like structures. Cytoplasmic expression detects neuro-immuneendocrine cells of the epidermis and single cells within all dermal compartments, mainly of the inner root sheath of hair follicle cells, sebaceous glands, and sweat glands. In turn, single nerve terminals were detected almost across the entire skin. In normal skin, epidermal and dermal PGP 9.5 nerve terminals are rare, and their low quantities detected in our studies were consistent with studies on human and equine skin as well as horse intestine. 29,53,59 In contrast to tissue distribution in equine cells and tissues, antibody specificity using WB needs to be included, especially where analyzed target proteins belong to the large superfamily of GCPRs. 60 The anti-CBR1 antibody provided a single 50 kDa band in the whole equine skin and brain lysates, and those results correspond with those obtained in human endothelial cells, immune cells in mice, and macaque brain tissue. 6 The CBR2 post-translational modification has not been reported, but we suspected that the multiband we found might reflect glycosylated forms. 68,69 Only 1 study reported WB for CBRs using equine source tissue, namely the cervical DRG at high molecular masses for CBR1 (100 and 120 kDa) and 90 kDa for CBR2. Similar results for keratinocytes have been reported for normal human epidermal keratinocytes (NHEK) and a human spontaneously immortalized keratinocytes (HaCaT): approximately 50 kDa CBR1 and 46 kDa (or 60 kDa). 16,70 The immunogene peptides for equine and human CBR1 and CBR2 are summarized in a Table 4. A number of factors may affect the molecular mass of proteins found in the WB reactions, such as a source of the protein (tissue or cellular lysate origin) animal species, and producer of antibodies.
Our study had some limitations. Because samples were obtained from horses at euthanasia, our study group was not homogeneous, influencing our final results. At the same time, skin biopsy samples were obtained from only 1 side of the body, without comparative analysis of other regions. Therefore, it is necessary to conduct further research to assess the distribution of CBRs elsewhere in the body.