Combined Systemic and Cutaneous Ovalbumin–Aluminum Sensitization Triggers Thermal Hyperalgesia, Spinal Gliosis, and Nociceptive Nerve Fibers Sprouting in Mice

Abstract

Background/Objectives: This study investigates whether systemic ovalbumin–aluminum sensitization followed by cutaneous challenge induces thermal hyperalgesia and spinal cord gliosis in mice. Methods: Adult female ICR-CD1 mice received ovalbumin–aluminum salt solution intraperitoneally on days 0, 7 and 14, and subcutaneously with daily skin application via brush during days 15–21, 28–35 and 42–49. Control animals received saline. Plantar thermal hyperalgesia was assessed on days 21, 35 and 49. On day 49, tissues (dorsal skin, spinal cord, footpads) were harvested after perfusion and processed histologically to detect mast cells, astrocytes, microglia, afferent fibers and intraepidermal nerve profiles. Results: Ovalbumin-treated mice displayed thermal hyperalgesia, increased astrogliosis, and reactive microglia in the spinal cord, and expanded CGRP- and IB4-immunoreactive fiber areas. In footpads, CGRP-positive intraepidermal nerve profile density was elevated, and dorsal skin showed increased mast cell density compared to controls. Conclusions: Combined systemic and cutaneous ovalbumin–aluminum sensitization induces skin inflammation, spinal gliosis, and nociceptive fiber sprouting in both central and peripheral sites, which likely contribute to the observed thermal hyperalgesia.

1. Introduction

The cutaneous administration of ovalbumin solution is a chemical method used for inducing atopic dermatitis in mice [1,2], rats [3], and dogs [4]. In this chemical model of murine atopic dermatitis, like what is observed in humans, there is an increase in epidermal thickness (epidermal hyperplasia), an abundance of eosinophils, mast cells, and T-cell lymphocytes, overexpression of inflammatory cytokines and chemokines, sensory hyperinnervation, and scratching behavior [2]. These findings suggest that this murine model reproduces approximately 85–90% of the signs and symptoms observed in human atopic dermatitis. Likewise, it has recently been reported that subcutaneous injection and topical application (with a brush) of an ovalbumin and aluminum salt solution can induce dermatitis-like changes in mice [5]. The overall prevalence of atopic dermatitis is estimated to be 15–20% in children and up to 10% in adults [6,7].

Patients with atopic dermatitis experience skin pruritus or itching, leading to scratching that exacerbates the infectious and inflammatory processes of the condition [8,9]. However, between 55–60% of patients with atopic dermatitis also report experiencing skin pain [10,11,12]. The plantar, palmar, and thoracic skin areas are the regions where skin pain is most frequently observed in these patients [13]. Patients with atopic dermatitis describe the pain as burning, itchy, electric shock-like, or cold, often characterizing it as sharp and bothersome [12].

Compounds such as complete Freund’s adjuvant (CFA) [14,15], carrageenan [16,17,18,19], formalin [20], and histamine [21,22] are known to trigger inflammatory pain when administered cutaneously. In contrast, the intracutaneous injection of ovalbumin induces an inflammatory response in animals pre-sensitized with ovalbumin and aluminum salts [23,24]. Furthermore, animals sensitized with 200 µL of ovalbumin solution and aluminum salts in saline, administered subcutaneously into the skin of the dorsum, exhibit paw-licking responses when ovalbumin is injected into the footpads, suggesting nocifensive behavior [25]. These findings suggest that all these compounds, including ovalbumin injected into the skin, induce both inflammation and pain.

In experimental models of atopic dermatitis induced by treatment with ovalbumin and aluminum salts in phosphate-buffered saline (PBS) or saline solution, administered via epicutaneous application with a patch impregnated with the solutions [26,27,28,29], by subcutaneous injection and brush application [5], or by subcutaneous injection alone [30], an increase in mast cells has been observed in the treated skin areas. Mast cells, during degranulation, synthesize and release a wide variety of chemical mediators including: (i) biogenic amines (histamine, serotonin, dopamine), (ii) lipid metabolites (platelet-activating factor, prostaglandins, leukotrienes), (iii) peptides (bradykinin, VIP, CRH), (iv) growth factors (NGF, bFGF, VEGF), (v) peptidoglycans (heparin, chondroitin sulfate), (vi) cytokines (IL1, IL3, IL4, IL6, IL8, IL9, IL17, IL33, Interferon gamma, TNF-α, TGF-β, (vii) chemokines (MCP-1, eotaxin, TARC, RANTES), (viii) enzymes (chymase, tryptase, serine proteases, aspartic acid proteases, cysteine proteases, metalloproteinases), and (ix) nitric oxide [31,32,33,34,35,36].

Some of these chemical mediators interact with transmembrane proteins (membrane receptors) on cutaneous nociceptors in the dermis and epidermis. These mediator-receptor interactions alter ionic conductance in nociceptive nerve fibers, generating action potentials that propagate through the somatosensory pain system, and give rise to the sensation and perception of pain [37,38,39]. Through a similar mechanism, these mediators can also activate pruritogenic nerve endings in the epidermis and dermis, generating action potentials that lead to the sensation and perception of itch or pruritus [39,40,41]. These chemical mediators released by mast cells also induce peripheral sensitization of nociceptors and pruritogenic receptors in the skin [42,43,44,45]. Taken together, this evidence suggests that mast cells, through the release of chemical mediators during degranulation, play a role in activating and sensitizing nociceptors and pruritogenic receptors (specifically those on C- and A-delta fiber nerve endings) that innervate the skin.

Intraplantar injection of carrageenan induces hyperalgesia and microgliosis in the dorsal horn of the spinal cord [46,47]. Activated microglial cells have also been observed in the thalamus and parietal cortex in response to carrageenan-induced inflammation in the hind paw [48]. CFA-induced pain hypersensitivity and peripheral inflammation, along with strong microglial activation in the spinal cord, have also been reported [49]. CFA-injection into the plantar surface of the hind paw causes thermal hyperalgesia and mechanical allodynia, as well as microgliosis and astrogliosis in the spinal cord, midbrain and thalamus [50]. Subcutaneous formalin injection into the hind paw similarly produces mechanical allodynia, thermal hyperalgesia, and gliosis in the dorsal horn of the spinal cord [51]. Microgliosis has also been observed in the dorsal horn of the spinal cord and the gracile nucleus of the brainstem following formalin injection into the hind paw [52]. These findings suggest that peripheral skin inflammation leads to pain responses associated with gliosis in both the spinal cord and supraspinal structures.

Furthermore, in an experimental model of rhinitis induced by treatment with ovalbumin, microgliosis was observed in the olfactory bulb [53]. Intrathecal injection of an ovalbumin solution (20 µg/mL in PBS) has also been shown to induce microgliosis in the hippocampus [54]. In another study, adult rats were sensitized with ovalbumin and treated intranasally on gestational day 15 with either an ovalbumin solution or vehicle. The offspring of rats exposed to ovalbumin showed microgliosis in the hippocampus [55]. These experimental findings indicate that microglial changes can occur in the central nervous system following exposure to ovalbumin.

Taken together, this evidence suggests that the cutaneous application of ovalbumin with aluminum salts in saline solution induces skin inflammation, which can lead to pain and activation of glial cells in the spinal cord. This animal model is intended to reflect human dermatitis, a cutaneous inflammation that causes nociceptive sensitization. The objectives of the present study are (1) to determine whether cutaneous application of saline solution containing ovalbumin and aluminum salts induces pain, as assessed by the plantar thermal algesimetry (or plantar) test; (2) to investigate, using immunohistochemical techniques, whether changes occur in microglial and astroglial cells in the spinal cord; (3) to assess histological changes in nociceptive afferent fibers in the dorsal horn of the spinal cord; and (4) to evaluate whether this treatment alters skin innervation in the foot pads.

2. Results

In response to noxious thermal stimulation, OVA-treated animals exhibited significantly shorter withdrawal latencies on days 35 and 49 compared to the control (saline) group, indicating the development of thermal hyperalgesia (Figure 1).

Immunohistochemical analysis of the spinal cord tissue sections at day 49 of follow-up revealed slightly hypertrophic astrocytes stained with GFAP antibodies in both OVA and saline-treated groups. The astrocytes appeared slightly more reactive in the OVA group (Figure 2C). However, the GFAP immuno-reactivity area was significantly larger in the OVA group compared to controls (Figure 2A). Non-reactive microglial cells exhibit small somas with highly branched cytoplasmic processes, whereas reactive microglia display retracted processes, an enlarged soma, and an amoeboid shape, with phagocytic capacity. In the IBA1-stained spinal cord sections, reactive microglia were identified by white arrowheads (Figure 2D). The percentage of reactive microglia was significantly higher in OVA-treated animals compared to saline-treated mice (Figure 2B). These results suggest that OVA treatment induces gliosis in the spinal cord, including both astrogliosis and microgliosis.

In spinal cord histological sections, OVA treatment also promoted the sprouting of peptidergic (CGRP-positive) and non-peptidergic (IB4-positive) nociceptive afferent fibers in the dorsal horn of the spinal cord. This was evidenced by a significant increase in the immunoreactivity area for these two markers in the dorsal horn (Figure 3A,B). These peptidergic (CGRP) and non-peptidergic (IB4) afferent nerve fibers project in the dorsal horn of the spinal cord onto the most superficial laminae of Rexed of the dorsal horn (laminae I, II) (Figure 3C,D).

In skin histological sections, OVA treatment significantly increased mast cell density compared to the saline controls (Figure 4A). OVA-treated sections showed more mast cells throughout the dermis of the dorsal skin. (Figure 4B). Mast cells appear as bluish-violet cells distributed throughout the dermis of the dorsal skin (Figure 4B). Under higher magnification, many mast cells displayed degranulation (Figure 4C).

In addition, OVA treatment caused a significant increase in CGRP-positive intraepidermal nerve fiber density in the footpads, compared to saline-treated animals (Figure 5).

The relationship between the withdrawal latency to the thermal stimulus in the plantar test and the different histological parameters analyzed is shown in Figure 6. Animals treated with ovalbumin (OVA) had a shorter withdrawal latency and, at the same time, a larger area of GFAP immunoreactivity and percentage of reactive microglia in the spinal cord, as well as a larger area of CGRP and IB4 immunoreactivity in the dorsal horn of the spinal cord, and a higher density of mast cells in the dorsal skin, as well as a higher density of intraepidermal CGRP nerve profiles in the footpads, compared to animals treated with saline solution (saline). Significant differences were observed between these two experimental groups.

Likewise, the percentage of mast cells in degranulation was 37.83 ± 2.39% in the OVA group and 21.22 ± 1.28% in the saline group, observing significant differences (p < 0.0001) between both experimental groups.

3. Discussion

In this study, OV-AL-treated mice exhibited significantly higher mast cell density in the dorsal skin, along with gliosis in the spinal cord. Additionally, there was an increased density of peptidergic (CGRP) and non-peptidergic (IB4) nociceptive afferent fibers as well as elevated CGRP-positive intraepidermal profiles in the foot pads. Functionally, animals treated with OVA-AL developed thermal hyperalgesia. These findings suggest that subcutaneous OV-AL sensitization coupled with topical application induces thermal hyperalgesia, associated with inflammation in the dorsal skin and gliosis in spinal cord.

Previous studies using similar OVA protocols have also reported increased mast cells in the skin [1,28,58,59,60]. There is evidence that cutaneous injection of ovalbumin promotes mast cell degranulation [61] and intraperitoneal injection of ovalbumin also facilitates mast cell degranulation [62]. Furthermore, in OVA-induced allergic airway inflammation, mast cell degranulation has been observed [63]. Collectively, these findings suggest that ovalbumin treatment promotes mast cell degranulation. Mast cell granules contain a wide variety of chemicals including biogenic amines (histamine, serotonin, dopamine), eicosanoids (prostaglandins, leukotrienes, thromboxanes), cytokines/chemokines (IL1, IL3, IL4, IL6, IL8, IL9, IL17, IL33, TNF-α, Interferon gamma, TGF-β, MIP-1, GM-CSF, eotaxin, TARC, RANTES, MCP-1, CXCL8, CCL2, CCL5), peptides/neuropeptides (bradykinin, VIP, CRH), growth factors (NGF, NT3/NT4, bFGF, VEGF, BDNF), enzymes/proteases (chymase, tryptase, serine proteases, aspartic acid proteases, cysteine proteases, metalloproteinases) and nitric oxide [31,32,33,34,35,36,64,65,66,67]. Some of these mediators activate and depolarize the nociceptors and pruritoceptors that innervate the skin, generating action potentials in these nerve fibers that propagate toward the spinal cord and brain, giving the sensation/perception of pain and itch-scratching [8,37,38,39,40]. These mediators enhance afferent fiber excitation by interacting with membrane receptors of nociceptors/pruritoreceptors, such as serotonin by binding to 5-HT3 receptors [68], histamine by binding to H1 receptors [69], and bradykinin by binding to B1 receptors [70]. In addition, other chemical mediators also contribute to the sensitization of nociceptors/pruritoreceptors including bradykinin, prostaglandins, and histamine [42,43,44,45,71]. Finally, other chemical mediators, such as prostaglandins (PGE2), contribute to the hyperexcitability of nociceptive afferent fibers by promoting ionic currents through voltage-gated sodium ion channels [72].

Peripheral sensitization, excitation, and hyperexcitation of nociceptors/pruritoreceptors that innervate the skin leads to the release of neurotransmitters and neuromodulators in the spinal cord dorsal horn affecting both neurons and glial cells. Nociceptive afferent fibers primarily release glutamate while A-delta afferent fibers also release CGRP and ATP, whereas peptidergic C fibers release CGRP, SP and ATP, while non-peptidergic C fibers also release ATP [73,74,75,76,77,78]. Itch-sensitive afferent nerve fibers release SP, natriuretic polypeptide B, gastrin releasing peptide (GRP), and glutamate [79,80]. These mediators activate second-order neurons, spinal interneurons, and glial cells, including microglia. It is well known that CNS microglial cells express a variety of receptors on their membrane including glutamate receptors [81,82], CGRP receptors [83,84], substance P receptors [85,86,87], purinergic receptors [88,89,90,91], and natriuretic peptide receptors [92]. This triggers microglial reactivation, shifting them from a quiescent state to a reactive state marked by the release of cytokines (IL-1β, TNF, IL-6), chemokines (CXCL10, CXCL12), and neurotrophins (BDNF, NGF, NT3) [93,94,95,96,97,98]. Similarly, astrocytes express glutamatergic, purinergic, and natriuretic peptide receptors [92,99], enabling them to respond to neurotransmitters released by nociceptive/pruritogenic afferent fibers, transforming into reactive astrocytes that secrete cytokines/chemokines (IL1-beta, IL6, TNF-alpha, MCP1/CCL2, CXCL8, MIF) [100], and neurotrophins (GDNF, NGF) [101,102]. These cytokines and chemokines released by reactive glial cells contribute to central sensitization of nociceptive neurons in the dorsal horn of the spinal cord [103,104,105] and promote long-term potentiation of synapses between nociceptive afferent fibers and spinal nociceptive neurons [103,106,107]. Certain neurotrophins such as BDNF released by reactive glial cells also promote long-term potentiation of synaptic transmission in the dorsal horn of the spinal cord [108]. Other neurotrophins, like NGF, induce the sprouting of nociceptive afferent fibers in the spinal cord [109]. In addition, CGRP-positive afferent nerve fibers express the TRKA receptor [110], while IB4-positive afferent nerve fibers express the GDNF receptor [111]. Thus, CGRP-positive peptidergic afferent fibers are sensitive to the action of NGF [112,113], while IB4-positive non-peptidergic afferent fibers are sensitive to GDNF [114,115,116], and both neurotrophins favor the sprouting of nerve fibers [117,118,119,120]. These findings suggest that the sensitization and excitation of cutaneous nociceptors and pruritoreceptors by chemical mediators from degranulated mast cells lead to an increased release of neurotransmitters in the dorsal horn. This, in turn, results in enhanced excitation of spinal nociceptive neurons and reactivation of glial cells. Chemical mediators facilitate central sensitization of nociceptive neurons, enhance synaptic transmission and sprouting of nociceptive afferent fibers in the dorsal horn, causing pain and hyperalgesia, as observed in the present study.

The preceding paragraphs indicate that ovalbumin treatment causes mast cell degranulation, which triggers activation and sensitization of nociceptors and pruritoreceptors in the skin. Understanding how ovalbumin induces degranulation is key. Mast cells present several receptors on their plasma membrane that are involved in their activation and degranulation, such as high-affinity IgE receptor (FcεRI), low-affinity IgG receptors (FcγRs), mast-related G-protein-coupled receptor X2 (MRGPRX2), P2X Purinoceptor 7 (P2RX7), and adhesion G-protein-coupled receptor E2 (ADGRE2) [121]. Treatment with ovalbumin induces the generation of IgE antibodies [122,123,124,125,126], which bind to high-affinity FcεRI on mast cells, triggering their activation and degranulation. FcεRI’s extracellular alpha chain binds IgE, while the beta and gamma chains are responsible for intracellular transduction. These intracellular chains present an immunoreceptor tyrosine-based activation motif (ITAM). When IgE binds to the receptor, tyrosine residues in the ITAM are phosphorylated by the Src family protein tyrosine kinase (SfK). This recruits another tyrosine kinase, Syk, and all of this allows the phosphorylation of adaptors and signaling enzymes in mast cells, such as PLC, which generates IP3 and DAG, which in turn activate PKC and mobilize calcium from the endoplasmic reticulum. This calcium outflow from the reticulum is detected by the STIM1 and 2 sensors, allowing STIM1/2 to bind to the plasma membrane Ca²⁺ channel Orai-1 (calcium release-activated channel [CRAC]), leading to activation of the Orai1 channel and the subsequent influx of Ca²⁺ into the mast cell. This increase in calcium in the mast cell promotes fusion of the granules with the membrane, i.e., mast cell degranulation [127]. The granule-membrane fusion machinery involves the N-ethylmaleimide-sensitive factor-binding protein receptor (SNARE) protein cluster. The first SNARE identified in mast cells is SNAP-23. During stimulation, SNAP-23 relocates from the plasma membrane to form cytoplasmic degranulation channels. A second identified t-SNARE is STX3, which reverse-localizes from secretory granules to the plasma membrane. Together, these recognition proteins between secretory granules and the plasma membrane allow the release of granule contents by exocytosis [127,128,129].

Treatment with ovalbumin induces the generation of IgG antibodies [130,131,132,133,134], which can bind to the low-affinity IgG receptor on mast cells. Activation of these receptors by IgG antibodies triggers an intracellular signaling pathway involving ITAM tyrosine phosphorylation, followed by dynamic protein phosphorylation cascades, including the SRC family of tyrosine kinases (Lyn), Syk, raft-associated transmembrane adaptors (LAT and NTAL), and phosphatidylinositol 3-kinase, among others [135]. There is evidence to suggest that activation of these low-affinity IgG receptors also induces calcium ion influxes [136]. Consequently, the cytosolic increase in calcium ions favors the activation of protein complexes that help fuse the granules with the plasma membrane, promoting the release of their contents through exocytosis.

On the other hand, peptides derived from human albumin could cause mast cell degranulation by binding to the MRGPRX2 receptor [137,138]. Other peptides (e.g., Substance P, VIP) are also ligands for MRGPRX2 receptors [121,138]. Nociceptive terminals in the skin can secrete substance P [139,140], and cutaneous nociceptors have high-affinity IgE receptors [141], as well as low-affinity IgG receptors [142]. Therefore, the IgE and IgG antibodies generated after ovalbumin administration can induce an increase in calcium ions in the nociceptive nerve terminal, which favors the fusion of vesicles loaded with substance P with the nociceptor plasma membrane, releasing this neuropeptide into the skin [139,140]. Now, this release of substance P in the skin can interact with the MRGPRX2 receptors of mast cells, facilitating their degranulation [121,138]. Like what has been described above, the binding of the ligand to the mast cell MRGPRX2 receptor activates the intracellular SRC, LYN, and PLC-gamma cascade. The latter factor favors the activation of Orai channels, allowing an influx of calcium ions, which activate protein complexes that promote the fusion of granules with the mast cell plasma membrane [138,143].

Overall, mast cells have different membrane receptors that, when activated by ligands that appear after treatment with ovalbumin, promote the entry of calcium ions into these cells, which in turn allows the fusion of the granules with the mast cell membrane and the release of their contents into the extracellular space.

On the other hand, hyperalgesia and pain responses have also been observed in animals treated with OVA. In an animal model of allergic inflammation induced by inhalation of OVA after sensitization with this same product, animals develop mechanical allodynia, categorized as a neuropathic pain response, due to the reactivation of microglial cells in the spinal cord [144]. In the experimental model of ocular allergy induced by OVA, increased scratching behavior is accompanied by heightened electrophysiological activity in ocular nociceptive fibers, indicating that OVA promotes nociceptor activation [145]. Additionally, the intraplantar subcutaneous injection of OVA and CFA induces mechanical hyperalgesia, due to elevated levels of inflammatory cytokines (IL1beta, TNF-alpha) in the plantar skin [146]. Mice sensitized with OVA plus aluminum also exhibit a significant increase in licking behavior in response to intraplantar injection [25]. Consistent with these findings, in the present study OVA-treated ICR-CD1 mice displayed significant plantar thermal hyperalgesia.

Mast cell degranulation releases biogenic amines (e.g., histamine, serotonin, dopamine), cytokines (e.g., TNF-α, IL-4, TGFβ, bFGF-2, IL-15), growth factors (e.g., NGF, VEGF), lysosomal enzymes (e.g., cathepsins, galactosidase, glucuronidase, hexosaminidase), and proteases (e.g., chymase, tryptase, cathepsin G, MMP9) [147]. These mediators sensitize and nociceptive terminals in the skin [44,148,149], contributing to hyperalgesia [150,151,152].

In OVA-treated animals, the density of CGRP-positive intraepidermal nerve profiles in footpad skin was significantly increased compared to saline controls. The increased density of intraepidermal profiles in the plantar skin could also be explained by the sprouting changes induced by the release of NGF by reactive glial cells in the spinal cord of OVA-treated mice, that is, the sprouting of CGRP-positive fibers in the dorsal horn of the spinal cord also causes sprouting of these same nerve terminals in the plantar skin, with an increased number and density of intraepidermal CGRP-positive nerve profiles. In this sense, CGRP-positive intraepidermal nerve fibers are also sensitive to the action of NGF [153,154]. These intraepidermal CGRP-positive afferent fibers located in mouse footpads or human skin have been identified as nociceptors [155,156,157]. Preclinical and clinical studies link cutaneous nerve fiber sprouting with increased pain sensitivity [158,159,160], while local injection of capsaicin triggers a loss of CGRP-positive epidermal nerve fibers that is accompanied by a reduction in pain [156,161]. Taken together, these mechanisms likely underlie the plantar thermal hyperalgesia observed in OVA-treated animals in this study.

All the previous paragraphs have presented scientific evidence that, in part, explains the results obtained in this study. However, it should be noted that the present study aims to reflect human dermatitis: a cutaneous inflammation related to allergens and adjuvants that causes nociceptive sensitization. In this context, there are different animal models of active sensitization to induce dermatitis, including repeated exposure to trinitrochlorobenzene (TNCB), dinitrofluorobenzene (DNFB), dinitrochlorobenzene (DNCB), ovalbumin (OVA), oxazolone (OXA), and calcipotriol (MC903) [2,162,163,164,165,166]. In the DNFB-induced dermatitis model, animals were initially systemically sensitized with intravenous injections of IgE antibodies. Subsequently, the skin was painted with DNFB 48 h post-sensitization. This caused IgE- and mast cell-dependent skin swelling. This is known as the first and second phases [167]. A third phase of skin swelling occurs after 8 days, characterized by an accumulation of eosinophils, mast cells, and T cells [168,169]. Repeated epicutaneous application of 2,4,6-trinitrochlorobenzene (TNCB) induces epidermal hyperplasia, accumulation of CD4-positive T cells and mast cells, and elevated plasma levels of IgE. Type 2 helper cells are responsible for the cutaneous response and IgE production. TNCB triggers a cytokine response from type 2 helper cells [170,171]. Topical application of dinitrochlorobenzene (DNCB) causes modification of macromolecules that are internalized by antigen-presenting cells in the skin (e.g., Langerhans cells, dermal dendritic cells, macrophages), processed, and presented to T lymphocytes, triggering their activation. Twenty-four hours after exposure to DNCB, macrophage accumulation is observed in the exposed area, contributing to the development of skin swelling and inflammation. Animals treated with DNCB show increased serum IgG1 levels, much higher than other immunoglobulins, as well as activation of type 2 helper lymphocytes with the production of cytokines (e.g., IL4, IL5, IL10) [172,173,174]. Repeated application of oxazolone (OXA) to the skin causes edema and the influx of inflammatory cells, especially mononuclear and polynuclear cells, into the area of application, which persists for 15 h. At this time, edema persists, but the cellular infiltrate, especially mast cells, decreases. After 48 h, the edema disappears, and the infiltrated cells decrease [175]. It is now known that OXA induces the influx of T and B lymphocytes, as well as neutrophils, and increases the levels of cytokines/interleukins (e.g., IL1β, TNF-α, IL4, IL10, IL13) [176,177]. Animals treated with MC903 exhibit dermatitis, with type 2 inflammation, skin barrier dysfunction characterized by increased trans-epidermal water loss (TEWL), scratching associated with pruritus, and increased serum IgE levels [178,179]. Histological analysis of MC903-treated skin reveals infiltration of granulocytes (e.g., eosinophils) and CD3-positive T cells, as well as an increase in dermal mast cells. Increased expression of interleukins (IL4, IL13, IL33) and serum IgE levels are also observed. Many of these changes persist long-term [178,179,180]. Finally, ovalbumin (OVA)-induced dermatitis triggers a type 2 helper T-cell-dependent immune response, with production of interleukin (IL4, IL5, IL13) and elevated serum IgE levels. Histological analysis of OVA-treated skin shows lymphocytes (CD4 positive), eosinophils, mast cells, and thickening of the dermis and epidermis. The animals exhibit scratching behavior in response to itching or pruritus [25,26,27,28,29,181]. In most of these animal models of dermatitis, an elevated presence of granulocytes (e.g., eosinophils, neutrophils), mast cells, and T lymphocytes is observed in the treated skin. These cells synthesize and secrete a variety of chemical mediators, which in addition to inducing swelling, skin redness, and inflammation [182,183,184,185], also trigger the excitation and sensitization of cutaneous nociceptors and pruritoreceptors [186,187,188,189,190,191], triggering itching and scratching behaviors of the skin [184,185,191,192,193]. The increased excitability of skin nociceptors/itch receptors causes these sensory nerve fibers to release increased numbers of neurotransmitters in the spinal cord. These neurotransmitters not only excite second-order spinal nociceptive neurons [194,195,196] but also activate spinal cord microglial cells, which are also capable of responding to these neurotransmitters [197]. In turn, these reactive microglial cells secrete increased chemical mediators into the spinal cord parenchyma, which facilitate chemical neurotransmission through sensory afferent fibers reaching the dorsal horn and central sensitization of second-order nociceptive neurons [198,199]. These chemical mediators released by reactive microglia cells diffuse through the parenchyma and can reach the lumbosacral regions, where more microglia cells can also be reactivated [200,201], contributing to more chemical mediators that can sensitize lumbosacral nociceptive neurons, and this in turn can promote hyperalgesia [202,203].

Although animal models for human diseases are very important for examining the mechanisms involved, as well as for developing treatment strategies for diseases, there are increasingly attempts to use alternative methods to the use of laboratory animals, with the aim of reducing their use. In this context, 2D cell cultures, 3D cells cultures, cell co-cultures, ex vivo skin organ culture, “skin on a chip” and skin tissue engineering can be used to study dermatitis or other dermatological diseases. In this context, it has been shown that 2D co-cultures between immune cells (e.g., eosinophils, basophils, T cells) and structural cells of the skin (e.g., keratinocytes, fibroblasts) stimulated by pruritogenic cytokines (e.g., IL31, IL33) mimic the environment of atopic dermatitis, with secretion of cytokines (IL1β, IL6) and chemokines (CXCL1, CXCL8, CCL2, CCL18) [204,205,206,207]. There are two alternative models to laboratory animals that adequately mimic the characteristics of the human epidermis: the reconstituted human epidermis (RHE) model and the full-thickness human skin equivalent (FTHE) model. These models reproduce epidermal morphology, differentiation, and barrier function [208,209,210,211,212]. Using the FTHE model, the application of interleukins (e.g., IL4, IL13) triggers changes like atopic dermatitis, allowing the testing of the effects of various treatments, such as indole-3-lactic acid [213] and Tofacitinib [214]. In the RHE model, the effect of various metal allergens (K₂Cr₂O₇, CoCl₂, NiCl₂) has been tested [215]. This model has also been used to study the influence of cytokines on the organization and differentiation of the epidermis [216], as well as to study the changes in the skin barrier under the influence of Th2-derived cytokines [217]. The ex vivo skin organ culture (EVSOC) model consists of using small pieces of skin explants to perform various experimental studies. This model has been used to evaluate the effects of ultraviolet radiation on the skin [218], test the effects of various skin creams [219], and analyze hair growth [220], among others. The EVSOC model has been used to induce dermatitis upon contact with Staphylococcus aureus, and to analyze the production of cytokines [221]. The “skin-on-a-chip” (SoC) model has been used for a wide variety of dermatological studies including the development of strategies to promote hair follicle regeneration [222], to study the degree of absorption of creams [223], and to assess inflammation and edema, and the use of treatments [224]. An epidermis-on-a-chip model has also been tested to evaluate the effects of various irritants [225], and this technology has been used to test treatment with Propionibacterium acnes as an acne inducer [226]. More recently, SoC has been used to test skin cancer treatments [227]. Taken together, these findings suggest that there are in vitro alternatives to the use of animal models that mimic the structure and function of the skin, and that they can be used to study cutaneous physiological and pathophysiological processes, as well as for the screening of cutaneous treatments.

To our knowledge, this is the first report showing that systemic OVA-aluminum sensitization followed by cutaneous challenge in mice induces (1) thermal hyperalgesia, (2) sprouting of nociceptive endings in the foot pads, (3) increased peptidergic and non-peptidergic fiber density in the spinal dorsal horn and (4) astrogliosis and microgliosis, when compared to saline-treated controls. Moreover, these findings not only demonstrate the robust effects of combining systemic and cutaneous ovalbumin–aluminum sensitization but also provide a solid foundation for future studies aimed at elucidating the specific inflammatory and neuro-immune mechanisms underlying the observed functional and histological changes.

This study has some limitations that should be acknowledged. First, the number of animals per experimental group was relatively small. Although this warrants caution in data interpretation, the consistent trends observed clearly support the conclusions presented. Future studies with larger sample sizes will help further validate these findings. Second, only plantar thermal hyperalgesia was assessed, whereas mechanical sensitivity (e.g., von Frey filaments) was not evaluated. Thus, it cannot be excluded that the animals developed mechanical allodynia in addition to thermal hyperalgesia. Including additional behavioral assays in future studies could provide a more comprehensive assessment of pain-related responses. Finally, the estrous cycle of female mice was not monitored. Because hormonal fluctuations may influence neuroimmune interactions and pain responses, monitoring the estrous cycle in future studies could add valuable insight. Importantly, none of these limitations undermine the main conclusions of this study but rather highlight useful directions for further research.

4. Materials and Methods

4.1. Animals, Study Design, and Ethical Considerations

Adult female mice of the ICR-CD1 strain, aged 8-weeks and weighing 25–30 g, were used (Janvier Laboratories, Le-Genest-Saint Isle, France). The sample size was calculated using the GRANMO software for comparison of two independent means. Accepting an alpha risk of 0.05 and a statistical power of 0.5 in a one-tailed test, six animals per group were required to detect a statistically significant a difference greater than or equal to 1 unit. In this study, six mice per experimental group were assigned to each experimental group.

Upon arrival at the Animal Care Unit of the University of Barcelona (Bellvitge campus), the mice were randomly allocated to different cages. Each cage was designated as an experimental group. It is important to note that the experimental unit was each individual animal.

A maximum of six mice were housed in each standard mouse cage (#1145T; 369 × 165 × 132 mm, floor area: 435 cm²; Tecniplast, Madrid, Spain). All cages were maintained in the same animal facility room, at a stable temperature of 21 ± 1 °C, with a humidity between 40 and 60%, and with a 12:12 h light–dark cycle. In these cages, the animals had access to food and drink ad libitum. Only the study animals were housed in the animal room, with no other types of animals or mouse strains other than those used in this experiment. All experiments were performed between 8:00 a.m. and 18:00 p.m., under light conditions, in silence, and with the minimum possible ambient noise.

This study employed a model of atopic dermatitis based on the subcutaneous injection of ovalbumin and aluminum salts in saline solution, combined with topical application of the same solution using a brush, as previously described [5]. Sensitization occurred on days 0, 7, and 14 via intraperitoneal injection of the ovalbumin–aluminum solution (OVA-AL). On day 14, after the intraperitoneal injection of OVA-AL and shaving and depilating the dorsal skin, the animals also received skin impregnation with OVA-AL solution using a brush. Daily, on days 15–21, 28–35, and 42–49, the animals received OVA solution intraperitoneally (i.p.), as well as OVA-AL solution subcutaneously (s.c.), and the dorsal skin was impregnated with OVA-AL solution using a brush. Plantar thermal hyperalgesia was assessed on days 21, 35, and 49. At the end of the study, dorsal skin, plantar pads, and spinal cord were collected and processed for histological analysis [5]. The experimental design of this study is illustrated in Figure 7.

All in vivo experimental procedures were performed at the Bellvitge Animal Facility of the University of Barcelona, in accordance with the ARRIVE 2.0 guidelines, the ethical principles of the International Association for the Study of Pain (IASP) for the use of animals in pain research animals [228], and the Directive of the European Parliament and of the Council of 22 September 2010 (2010/63/UE). The Animal Experimentation Ethics Committee of the University of Barcelona (CEEA, ref: 50/19; approved on 11 April 2019) and the Ethics Committee of the Department of Agriculture, Livestock, Fisheries, Food, and the Natural Environment of the Government of Catalonia (DAAMM, ref: 10672, approved on 22 November 2019) approved the animal experimentation procedures for this study.

4.2. Preparation of Ovalbumin Solutions and Skin Treatment

The animals were treated with two types of ovalbumin (OVA) solutions: one containing aluminum salts (OVA-AL solution) and one without (OVA solution). The OVA-AL solution consisted of 0.5 g of ovalbumin grade V (#A5503 Sigma-Aldrich-Merck, Darmstadt, Germany) and 0.2 g of aluminum hydroxide hydrate, (#A1577 Sigma-Aldrich-Merck, Germany) dissolved in 500 mL of 0.9% saline solution (Vitulia Physiological Serum, Barcelona, Spain) The OVA solution contained 0.5 g of ovalbumin dissolved in 500 mL of saline solution. Both solutions were filtered through 0.22 µm syringe filters (#SFNY-122-100; Labbox, Barcelona, Spain), aliquoted into 50 mL tubes (# PTSP-E50-025; Labbox) and stored at 4 °C until use.

Mice were injected intraperitoneally (i.p.) with 0.2 mL of OVA-AL solution on days 0, 7, and 14. On day 14, after the i.p. administration of the OVA-AL solution, the mice were anesthetized with sodium pentobarbital (#Y000219450 Sigma-Aldrich-Merck; 50 mg/kg, 10 mg/mL, i.p.). The dorsal skin was shaved using an electric shaver and then depilated using a depilatory cream (Silky fresh sensitive skin; Veet, Reckitt Benckiser Healthcare, Slough, UK). The depilated area was subsequently treated with OVA-AL solution using a bristle brush (n° 10; Pelikan 721431; Hannover, Germany).

From days 15 to 21, 28 to 35, and 42 to 49, the mice received daily treatments consisting of (i) 0.2 mL of OVA solution intraperitoneally, (ii) 0.2 mL of OVA-AL solution subcutaneously, and (iii) topical application of OVA-AL solution to the dorsal skin using a brush. These mice constituted the OVA group. Control animals (saline group) were treated exclusively with saline solution (0.9% Vitulia Physiological Serum, Barcelona, Spain), following the same schedule and procedures as the OVA group. For repeated application of OVA or saline solutions, the procedure described above was followed with minor modifications [29,229]. Two experimental groups were used in this study: (i) the saline group, treated with saline solution only, and (ii) the OVA group, sensitized with OVA-AL (i.p.) and subsequently treated with OVA (OVA i.p.+ OVA-AL s.c. + OVA-AL with a brush) during the periods of days 15–21, 28–35, 42–49, as described above.

4.3. Thermal Hyperalgesia Evaluation

Thermal hyperalgesia was assessed on days 21, 35, and 49 following the administration of OVA or saline solutions, using a plantar-test apparatus (#37370; Ugo Basile, 21036 Gemonio VA, Italy). Mice were placed in methacrylate boxes with glass floors and allowed to acclimatize for 60 min. A beam of incandescent light from a 100 W bulb was then directed at the plantar surface of the hind paws, and the withdrawal latency to this painful stimulus was determined. The exposure time was limited to a maximum of 30 s to prevent burns. The mean withdrawal latency for both hind paws was calculated as the average of three separate trials, conducted at 5-min intervals. It should be noted that the animals stand directly on the glass floor, which is 2 mm thick, and immediately below, also touching the glass, is the incandescent light outlet [230,231,232].

4.4. Histological Evaluation

On day 49, after completing the functional assessments, the animals were deeply anesthetized with sodium pentobarbital (90–100 mg/kg; 10 mg/mL; i.p.) and perfused intraventricularly with 4% paraformaldehyde solution in phosphate-buffered saline (PBS; 0.1 M, pH = 7.4). Subsequently, the dorsal skin was then excised and transferred to a jar containing the same fixative. Following a dorsal laminectomy, the spinal cord was removed and placed in Eppendorf tubes with fixative. Finally, the plantar pads (B and C) of both hind legs were also excised and placed in Eppendorf tubes with fixative. All tissues were fixed at 4 °C for 15 days. After fixation, samples were transferred to a cryo-protective solution (30% sucrose in PBS) and stored at 4 °C for an additional 15–20 days.

4.4.1. Skin Processing

Dorsal skin samples were stretched on cork boards and secured with needles. From the center, a 2 cm² section was excised and embedded in tissue freezing medium (Ref: 0201-08-926; Leica, Barcelona, Spain), by using a special mold. Blocks were frozen in a cryostat (CM1520, Leica, Barcelona, Spain) at −24 °C. Skin was cryosectioned at 20 µm, and 8–10 sections per animal were mounted on pregelatinized slides. Sections were stained with Giemsa (azur-eosin-methylene blue; #EOMB-MSD-1K0; Labbox, Barcelona, Spain) diluted 1:4 in distilled water (60 mL Giemsa solution and 240 mL distilled water) for 90 min. Afterwards, the sections were rinsed in 0.1% acetic acid for 15–20 s, followed by 96% ethanol for another 15–20 s, and then washed in methanol (three baths, 2 min per bath). Finally, the skin sections were rinsed three times in xylene (2 min per bath) and coverslipped with DPX (# 1.01.979.500; Merck, Germany).

4.4.2. Spinal Cord Processing

Spinal cord segments (2–3 mm, thoracic region) were embedded in the same freezing medium, frozen at −24 °C, and sectioned at 60 µm. Sections (8–10 per animal) were collected in 6-well porcelain plates (#SPPC-006-001; Labbox, Barcelona, Spain) and washed at room temperature with agitation in PBS (10 min), PBS + 0.3% Triton-X-100 (10 min), and PBS-Triton + 1% fetal bovine serum (30–45 min).

Sections were incubated for 48 h at 4 °C in a humid chamber with constant agitation in the following primary antibodies: (i) Rabbit anti-glial fibrillary acidic protein (GFAP; 1:200; #ab7260, ABCAM, Cambridge, UK) for astrocytes, (ii) Rabbit anti-ionized calcium-binding adapter molecule type 1 (Iba1; 1: 200; # 019-19741; WAKO, Richmond, VA, USA) for microglia, (iii) Goat anti-calcitonin gene-related protein (CGRP; 1:200; #ab36001, ABCAM, UK) for peptidergic nociceptive afferent fibers, and (iv) Biotinylated Bandeiraea simplicifolia Isolectin (IB4; 1:200; # L2140, Sigma-Aldrich-Merck, Germany) for non-peptidergic nociceptive afferent fibers.

After the primary antibody incubation, sections were washed three times for 10 min each with PBS-Triton and then incubated for 24 h at 4 °C in a humidified chamber with constant agitation in the following secondary antibody solution: donkey anti-goat and donkey anti-rabbit, and avidin, all conjugated to cyanine 3.18 (Cy3; 1:200; Jackson Immunoresearch, West Grove, PA, USA). After several additional washes with PBS-Triton and PBS, the sections were mounted on pregelatinized slides, dehydrated through graded in ethanol (70°, 96°, then absolute) and coverslipped with DPX [233].

Footpads were similarly embedded in tissue freezing medium, frozen at −24 °C and cryosectioned at a thickness of 60 µm. All sections from each footpad were collected in the same well of a 6-well porcelain plate and processed using the same immunohistochemical protocol with goat anti-CGRP as primary antibody and Cy3-conjugated donkey anti-goat as secondary, to visualize intraepidermal CGRP-positive nociceptor profiles [155,234]. To verify antibody specificity, some spinal cord and footpad samples) were processed as described, except the primary antisera was omitted.

Spinal cord and footpad sections were examined under a Leica DMRXA epifluorescence microscope (Leica Microsistemas S.L.U., Hospitalet de Llobregat, Spain) equipped with a digital camera (FMVU-13S2C-CS). Using Image-J software (NIH, Bethesda, MD, USA; version 2.14.0/1.54f), CGRP and IB4 immunolabelling areas were quantified from spinal cord images at ×100 magnification. GFAP immunoreactivity was quantified at ×200 and served as an index of astrogliosis [166]. Iba1-immunostained sections (×200) were used to count reactive versus non-reactive microglial cells expressed as a percentage to quantify microgliosis [232,235].

Dorsal skin sections stained with Giemsa were analyzed under the Leica DMRXA optical microscope (without epifluorescence), and images were captured using the FMVU-13S2C-CS camera. Mast cells were manually counted in Image-J (same version) and density was expressed as cells/mm². The percentage of degranulating mast cells was also determined.

4.5. Statistical Analysis

Functional and histological analyses were performed blinded, with each animal assigned a numerical code. The researcher assessing thermal hyperalgesia was unaware of the experimental groups. Histological analyses were independently performed by two other researchers who were also blinded to group assignment.

Statistical comparisons between the different experimental groups were conducted using Kruskal–Wallis and Mann–Whitney U non-parametric statistical tests. Results are presented as mean ± standard error of the mean (SEM), and a significance level of p < 0.05 was applied. All analyses were conducted using GraphPad Prism 9.0 for Macintosh.