RACK1, a clue to the diagnosis of cutaneous melanomas in horses

Melanocytic tumors are common in horses. Up to 18% of all skin tumors are melanocytic [1]. The true incidence may be even higher since a number of epidemiological studies do not include a histological report. Equine melanocytic tumors may occur in any body area regardless of age, sex or coat color. Most of these tumors are clinically benign at initial presentation, but two-thirds are thought to progress to malignancy and metastasize [2]. Accordingly, setting a diagnosis on histopathological analysis alone can be challenging [2]. The term “melanoma” is used for malignant melanocytic tumors, whereas “melanocytoma” refers to the benign forms, with the corresponding restrictions. Indeed morphological criteria are not always predictive of clinical features [1]. In human melanoma, morphology-based melanoma classification has presented limited clinical relevance. Nevertheless, Bastian and colleagues have recently shown that clinical and morphologic features associated with known mutations could be used to identify biologically related disease groups [3, 4]. Coat color genetic studies in horses identified genes responsible for associated pathologies [5, 6], like melanoma in gray horses [7]. More studies are needed to determine the mutation status of horse melanocytic proliferations.

At present, a molecular marker of malignancy would be of great interest to distinguish benign from malignant melanocytic tumors. We found that immunodetection of RACK1 (Receptor for activated C kinase 1) protein deserves consideration as such a marker. Indeed RACK1 is strongly detected in melanoma cells of primary tumors and metastases developed in MeLiM minipigs as well as in human patients. In contrast, RACK1 is not detected in normal skin melanocytes or in benign tumoral proliferations [8].

RACK1 is a 36 kDa scaffold protein containing seven internal WD40 repeats, originally identified as an anchoring protein for protein kinase C (PKC) [9]. It is now well established that RACK1 is ubiquitous, with a tightly regulated expression, and that it interacts with a large number of proteins. Through its ability to coordinate the interaction of key signaling molecules, RACK1 is widely perceived as playing a central role in critical biological responses both in normal cell physiology and in tumorigenesis [10]. Several in vitro studies have shown that RACK1 could be implicated in cancer hallmarks [10–14]. Particularly, there is evidence for a role of RACK1 in the pathogenesis of melanoma. In the MeWo human melanoma cell line, RACK1 serves as an adaptor protein for PKC-mediated JNK (c-Jun NH2-terminal kinase) activation and increases the survival to UV induced-apoptosis [15]. RACK1 may allow cross-talks between several pathways involved in melanoma development through the orchestration of protein-protein interactions.

In this study, we tested the value of RACK1 detection in the diagnosis of horse melanoma.

Early identification of malignant melanocytic lesions is crucial for patient survival in human and veterinary medicine. RACK1 is a scaffold protein found to integrate various metabolic pathways involved in tumorigenesis [22, 23]. It was proposed as a marker of malignancy in pig and human melanomas [8]. We extend these observations to melanomas in horses which displayed an overexpression of RACK1 when compared to normal cutaneous melanocytes.

In an attempt to find a more refined morphological classification to distinguish benign from malignant lesions, we used the criteria defined for human melanoma in equine melanocytomas and melanomas. Although none of the morphological criteria taken separately were powerful enough to distinguish between benign and malignant tissue, we highlight specific histological features of equine melanoma that recapitulate those seen in atypical rare forms of human malignant nævi and melanomas such as pigment synthesizing melanoma, desmoplastic melanoma, primary dermal melanoma, and malignant blue nævus [24–27]. This makes the study of equine melanoma a source of information to understand development of atypical melanoma in mammals.

We show that PAX3 and MITF immunolabeling can be used to identify melanocytes as well as melanocytic transformed cells within a tumor bulk. MITF is known to mark melanocytic proliferations in humans and pigs [8]. More specifically, we show that PAX3 is expressed by mature melanocytes in horses, as in humans [21]. In every sample from healthy and tumoral tissues, the melanocytic cells displayed a specific nuclear MITF labeling. Achromic tumoral samples were also positively stained for MITF. We therefore propose MITF immunolabeling as a diagnosis marker for melanocytic tumors in horses.

All the samples identified as melanocytomas by histopathologic analysis stained heterogeneously for RACK1. On the other hand, a homogeneous staining for RACK1 only appeared on melanomas. Two samples identified as melanomas also stained heterogeneously for RACK1. Both had morphological characteristics of low aggressiveness as a low mitotic index. This is a common pitfall in histological analysis of equine melanoma, and reports on antigens related to the cell cycle gave controversial results [17, 28]. However, no classification based on detailed staging [29] is available for equine melanomas. We extended previous observations on RACK1 expression in melanoma from humans and pigs to melanocytic lesions in horses. We show here that the cellular distribution of RACK1 reflects melanoma progression and aggressiveness: the more homogenous and diffuse the cytoplasmic RACK1 labeling is, the more aggressive the melanoma. Thus, we propose RACK1 as a marker of malignancy in equine melanocytic proliferations.

Furthermore, our data show that the cellular distribution pattern of RACK1 on melanocytes in tumors, i.e. homogenous if malignant and heterogeneous if benign, is constant throughout the whole lesion. RACK1 distribution pattern analyzed in a punch biopsy appears as informative for diagnosis and less invasive than a complete biopsy. This is a particularly interesting element since skin biopsies in horses are not that easy to carry out. A quick grading could influence on the decision of treating melanomas more urgently, even if they are less suited for surgical therapies.

In domestic animals, veterinary pathologists use the term melanocytomas to describe benign melanocytic tumors [30]. In humans, such benign melanocytic tumors are designated as nævi and the term melanocytoma is seldom used. It defines melanocytic tumors with uncertain malignancy status [31, 32]. Our data show that the distribution of RACK1 labeling is heterogeneous in horse melanocytomas while it is absent or faint in human nævi [8]. This confirms the difference between melanocytomas and nævi at the molecular level. In horses, the second most common clinical presentation of melanoma is the malignant transformation of a melanocytoma [33]. Based on RACK1 distribution, we hypothesize that horse melanocytomas may be premalignant entities ready to switch to malignancy. RACK1 detection would reveal this switch to malignancy, if any.

RACK1 protein was detected with intense, diffuse and homogenous staining in MITF-positive cells of equine melanomas. This observation is consistent with stainings in human melanomas and MeLiM minipig melanocytic tumors. We now confirm the usefulness of RACK1 labeling as a diagnosis marker for melanoma. The homogenous distribution pattern of RACK1 signal shared by human, pig, and horse melanomas strongly suggests a function for RACK1 in melanoma progression in mammalian skin.