Evaluation of Hypoxia-Inducible Factor-1 Alpha (HIF-1α) in Equine Sarcoid: An Immunohistochemical and Biochemical Study

Background: equine sarcoids are the most frequent skin tumors in equidae worldwide. It is well known that delta bovine papillomaviruses are their causative agents. We have recently shown the presence in equine sarcoids of abnormal vessel structures, which could cause a hypoxic condition. The aim of this study was to analyze the expression of hypoxia-inducible factor-1 alpha (HIF-1α) in a subset of BPV positive equine sarcoids and explore the relationship with vascular endothelial growth factor (VEGF) expression. Results: 80% of equine sarcoids showed strong cytoplasmic staining in >60% of neoplastic fibroblasts, while 20% of samples showed a moderate cytoplasmic staining in 40–60% of neoplastic fibroblasts for HIF-1α. Results of Western blotting (WB) were consistent with immunohistochemistry (IHC). Moreover, a positive correlation between HIF-1α and VEGF expression (r = 0.60, p < 0.01) was observed. Conclusion: we have shown that HIF-1α was strongly expressed in equine sarcoid. The upregulation of HIF-1α has been described in numerous tumors and can be modulated by many proteins encoded by transforming viruses. Thus, it is also possible that BPV could have a relevant role in HIF-1α pathway regulation, contributing to the development of equine sarcoids by promoting HIF-1α/VEGF mediated tumor angiogenesis.


Introduction
Equine sarcoids are the most frequent fibroblastic benign skin tumor in equidae worldwide, with prevalence rates of 0.5-2.0% [1,2]. They often form from scar tissue or sites of a previous wound or trauma, on the head, the limbs, the ventrum, and the paragenital region [3]. Sarcoids often persist, can be locally invasive, and rarely regress [3,4], although it has been recently reported that a high percentage of equine sarcoids spontaneously disappeared without therapy [5]. It is widely known that these lesions do not always respond to therapy and are often correlated with a high recurrence rate after surgical excision [3,4,6], leading to morbidity and impacting the function and aesthetics of affected equids, which decreases their economic value [7,8]. Equine sarcoids have a multifactorial cause [9]; however, it has been proved that there is a strong correlation between permanent infection with delta bovine papillomaviruses (BPV type 1, 2 and 13), and persistent and progressive sarcoids [7,10,11]. The evidence that BPV has a main role in the pathogenesis of sarcoids may be partly explained by the detection, in naturally occurring equine sarcoid, of BPV DNA in an episomal form and in multiple copies [11][12][13][14]. In addition, BPV oncogenes and capsid gene transcripts

Histological Features
The examined sarcoids (n = 35) showed the classic histological features of the lesion: epidermal hyperplasia with rete pegs; hyperkeratosis; proliferation of neoplastic fibroblast in the dermal layer, oriented in a 'picket fence' perpendicular to the basilar epidermal layer; exuberant extracellular matrix; presence of many small vessels irregular in shape [26,27].

Immunohistochemical Results
All results of HIF-1α expression pattern in 35 equine sarcoids and 10 normal skin samples are shown in Table 1.

Control Samples
Equine kidney used as positive control showed strong immunostaining for HIF-1α in tubular epithelial cells ( Figure S1). Equine normal skin and sarcoid, used as the control, showed staining for Purified Rabbit IgG (P120-201-Bethyl Laboratories, Inc.) in cells at the base of the epidermis but not in fibroblasts ( Figures S2 and S3). All 10/10 normal skin samples showed immunostaining for HIF-1α, in 10% of basal epidermal cells. Fibroblasts were negative ( Figure 1).

Statistical Results
The rank correlation analysis showed that the percentage of HIF-1α positive cells was positively correlated with the percentage of VEGF positive cells found during a previous study by our group (r = 0.60 p < 0.01) [27].

Biochemical Results
By Western blot, a band of the expected molecular size for HIF-1α (120 kDa, Figure S4) was identified in the tested samples, as well as in Hela and K562 cell lines used as positive control as suggested by antibody datasheet and literature data, confirming the specificity of the antibody ( Figure 3A) [54]. HIF-1α was expressed at higher levels in sarcoid samples with respect to normal skin samples, where the band was detected at low to undetectable levels, as confirmed by densitometric analysis (Figure 3A,B).

Statistical Results
The rank correlation analysis showed that the percentage of HIF-1α positive cells was positively correlated with the percentage of VEGF positive cells found during a previous study by our group (r = 0.60 p < 0.01) [27].

Biochemical Results
By Western blot, a band of the expected molecular size for HIF-1α (120 kDa, Figure S4) was identified in the tested samples, as well as in Hela and K562 cell lines used as positive control as suggested by antibody datasheet and literature data, confirming the specificity of the antibody ( Figure 3A) [54]. HIF-1α was expressed at higher levels in sarcoid samples with respect to normal skin samples, where the band was detected at low to undetectable levels, as confirmed by densitometric analysis (Figure 3A,B).

Statistical Results
The rank correlation analysis showed that the percentage of HIF-1α positive cells was positively correlated with the percentage of VEGF positive cells found during a previous study by our group (r = 0.60 p < 0.01) [27].

Biochemical Results
By Western blot, a band of the expected molecular size for HIF-1α (120 kDa, Figure S4) was identified in the tested samples, as well as in Hela and K562 cell lines used as positive control as suggested by antibody datasheet and literature data, confirming the specificity of the antibody ( Figure 3A) [54]. HIF-1α was expressed at higher levels in sarcoid samples with respect to normal skin samples, where the band was detected at low to undetectable levels, as confirmed by densitometric analysis (Figure 3A,B).

Discussion
The insufficient levels of cellular oxygen, a condition also known as hypoxia, were demonstrated in many tumors [55] and were associated with a structural and functional abnormality of vessels or to an increase of oxygen consumption caused by the rapid proliferation of neoplastic cells [56]. HIF-1α has a relevant role in oxygen homeostasis, and experimental evidence has indicated that it is a major regulator of normal and tumor cell adaption to hypoxic stress [52,53,55,57]. HIF-1 is a heterodimeric protein composed of a constitutively expressed HIF-1β subunit and an O2-regulated HIF-1α subunit [58]. HIF-1α is degraded by the ubiquitin-proteasome pathway [59] under normoxic conditions, while it is protected from ubiquitination and proteasomal degradation under hypoxic-conditions [42]. After PHD inhibition, HIF-1α dimerizes in HIF-1β to form HIF-1, which is responsible for the transcription of genes encoding glucose transporters, glycolytic enzymes, and VEGF [40][41][42]. HIF-1α and VEGF are major regulators of angiogenesis [60] in the tumor microenvironment and have a crucial role in tumor progression [60][61][62]. As VEGF [27,29,32,34,60], HIF-1α is overexpressed in a large variety of tumors [60,63], and its association with unfavorable prognosis has been reported, as it activates genes that play a relevant role in angiogenesis, invasion, and metastasis [57,59,64].
In this study, we have observed, by immunohistochemistry and biochemical analysis, HIF-1α expression levels in BPV positive equine sarcoids, located in different body regions [27], and we have evaluated the correlation between HIF-1α and VEGF expression, previously analyzed in a study of ours.
In our samples, surprisingly, HIF-1α showed a cytoplasmic expression, while the antibody used by us (#ab114977, Abcam) was reported to have a nuclear expression. We hypothesize that the abnormal upregulation and accumulation of HIF-1α in the cytoplasm could be related to the inhibition of prolyl-hydroxylation (PHD) under hypoxia and to the consequent suppression of HIF-1α degradation, leading to its rapid accumulation in the cytoplasm [65]. HIF-1α shuttling between cytoplasm and nucleus is a complex process regulated by numerous factors [65], and it was already reported its cytoplasmic expression in a broad spectrum of tumors [66,67].
All normal skin samples showed negative immunostaining for HIF-1α in fibroblast, while a weak immunostaining was observed in the basal epidermal cells, where HIF-1α is known to be constitutively expressed. Moreover, 80% of sarcoid samples showed a strong and finely granular cytoplasmic staining for HIF-1α in >60% of sarcoid fibroblasts and endothelial cells, while in remaining samples (20%) the intensity of immunostaining was moderate and observed in 40-60% of neoplastic fibroblasts and endothelial cells.
Although the samples located on the neck, paragenital region, and limb showed higher intensity staining and percentage positive score (see Table 1), no correlation could be demonstrated.
In a previous study [27] including the same samples, we have reported VEGF overexpression in most keratinocytes, sarcoid fibroblasts, and endothelial cells. Moreover, we have recently shown that, even if small blood vessels were numerous, they showed irregularity in shape, and their lumina appeared indistinct. Taken together, these data strongly suggest that in sarcoid tissue, there could be a hypoxic condition in which HIF-1α would have a crucial role, leading to the upregulation of VEGF and an increase in the vessels' number. However, despite the increase of vessels, they didn't appear sufficiently mature, possibly inducing worse or persistent hypoxia, which in turn could induce the upregulation of HIF-1α and then of VEGF. This strict relationship between HIF-1α and VEGF was also, in part, evidenced in our statistical analysis results, which showed a positive correlation between HIF-1α and VEGF expression (r = 0.60; p < 0.01). Our results seem to suggest that HIF-1α not only could regulate VEGF expression but also contributes to the formation of a complex proangiogenic microenvironment in equine sarcoids, thereby affecting vessels' morphology and, ultimately, the vessels' function [68].
Numerous studies reported that HIF-1α synthesis, or its decreased degradation, can be modulated by different tumor suppressor genes and oncogenes, among which there are many proteins encoded by transforming viruses [58,69]. There is strong evidence that human papillomaviruses (HPV) oncoproteins can promote tumor angiogenesis via the upregulation of HIF1/VEGF pathways, specifically manipulating aspects of the cellular hypoxic response [70,71]. HPV E6 and E7 were shown to increase independently the induction of HIF-1α [65] or to interfere with HIF-1α degradation, leading to the inactivation of proteasomal degradation and to HIF-1α stabilization [39,72,73]. Moreover, built upon these evidences, and in light of the common biological functions of papillomavirus oncogenes [74], we may speculate that BPV oncoproteins could play a relevant role in regulation of HIF-1α pathway, contributing, at least in part, to the development of equine sarcoids by promoting HIF1α/VEGF mediated tumor angiogenesis.
Further investigations are needed to clarify the specific role of BPV in the regulation of HIF-1α/VEGF pathway, and to evaluate if there could be any correlation in equine sarcoid between HIF-1α and glucose transporter 1 (GLUT1), which is known to be the rate-limiting enzyme for glycolysis [75]. In hypoxic conditions, neoplastic cells have been reported to increase GLUT1 expression under the positive regulation of HIF-1α, leading to increase cellular glucose uptake, and support the aerobic glycolysis of cancer cells [75].
We believe that new knowledge of equine sarcoid pathogenesis would be necessary in order to gain new insights into the development of novel therapies for this pathology.

Tumor Samples
We analyzed 10 normal skin samples and 35 equine sarcoids. The normal skin samples, located on head (n = 3), abdomen (n = 2), neck (n = 2), limb (n = 2), and pectoral regions (n = 1), were obtained during necropsy from healthy horses. The equine sarcoids, located on the limbs (n = 10), pectoral region (n = 8), head (n = 6), abdomen (n = 6), paragenital (n = 3) and neck (n = 2) (Table 1), were surgically excised from affected horses, using best practice of veterinary care, according to Directive 2010/63/EU (art. 1), and processed for routine diagnosis and treatment. Each owner consented to use tissues for research purposes, according to the ethical guidelines of the Anatomic Pathology Diagnostic Service of the Department of Veterinary Medicine and Animal Production (University of Naples Federico II). All samples were 10% formalin-fixed, paraffin-embedded for routine histological processing, and stained with hematoxylin and eosin (HE). Four sarcoids (S26, S27, S28, S29) and 2 normal skin samples (N6-N7) were immediately frozen at −80 • C and analyzed by Western blotting. Al sarcoid samples, the same as those previously used [27], were BPV positive, while normal skin samples were BPV negative [24]. Moreover, no previous treatments with topical or intra-tumoral therapy prior to excision were used.

Immunohistochemistry
Sections (5 µm) were processed for immunohistochemistry using the streptavidin-biotinperoxidase method. All sections were deparaffinized in alcohol decreasing solutions, and endogenous peroxidase activity was blocked by incubation in 0.3% H 2 O 2 in methanol for 20 min. Antigen retrieval was performed by pre-treating with microwave heating (twice for 5 min each at 750 W) in citrate buffer, pH 6.0. The slides were washed three times with phosphate buffered saline (PBS, pH 7.4, 0.01 M), then incubated for 1 h at room temperature with normal goat serum (Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted at 20% in PBS. As primary antibody, a polyclonal rabbit anti-human (predicted to cross-react with horse) to anti-HIF-1α (#ab114977, Abcam; 775-826 amino acids at C terminal) diluted 1:100 in PBS was used and applied overnight at 4 • C. Control sections (equine normal skin and sarcoid) were incubated with PBS instead of the primary antibody and with rabbit IgG (Purified Rabbit IgG P120-201-Bethyl Laboratories, Inc.) at the same concentration as the primary antibody. Equine kidney sections were used as positive control and incubated with anti-HIF-1α antibody (#ab114977, Abcam; 775-826 amino acids at C terminal).
Then, sections were incubated with MACH 1 probe (Biocare Medical, LLC, Concord, CA, USA) for 20 min at room temperature and with MACH-1 Universal HPR-Polymer (Biocare Medical, LLC, Concord, CA, USA) for 30 min at room temperature. Sections were counterstained with hematoxylin, and the immunolabelling was revealed with diaminobenzidine tetrahydrochloride.

Statistical Analysis
Pearson correlation test was used to correlate the percentage of HIF-1α positive cells with the percentage of VEGF positive cells reported in a previous study [27].
Following four washing steps of 10 min in TTBS, donkey anti-rabbit secondary antibody conjugated with peroxidase was employed at 1:2000 dilution for 1 h at room temperature. After additional washing steps, bound antibody was visualized by enzyme chemiluminescence with Clarity ™ Western ECL Blotting Substrate (Bio-Rad Laboratories, Milano, Italy). The blots were stripped and reprobed for β-actin (CP01, Calbiochem, San Diego, CA, USA) (1:500) as a loading control in order to perform normalization. Protein quantization and normalization were performed as reported in a previous study [27].

Conclusions
Finally, in our study, we have demonstrated for the first time the increase of HIF-1α expression in equine sarcoid, and we hypothesized that HIF-1α, together with VEGF, could have a role in sarcoid development.
Recent advances in cancer biology at the cellular and molecular levels highlighted the HIF-1α pathway as a crucial survival pathway for which novel strategies of cancer therapy could be developed [76]. We hope that in the future, hypoxia-inducible factor-1 (HIF-1) could be an important cancer drug target for equine sarcoid, since not all sarcoids are responsive to therapy, despite the numerous treatment choices available.
Supplementary Materials: The following are available online at https://www.mdpi.com/2076-0817/9/1/58/s1. Figure S1: Equine kidney used as positive control showed strong immunostaining for HIF-1α in tubular epithelial cells; Figure S2: Equine normal skin showed staining for Purified Rabbit IgG (P120-201-Bethyl laboratories-INC) in cells at the base of the epidermis but not in fibroblasts; Figure S3: Equine sarcoid showed staining for Purified Rabbit IgG (P120-201-Bethyl laboratories-INC) in cells at the base of the epidermis but not in fibroblasts; Figure S4: Full length membrane of Western blotting for HIF-1 α from Figure 3