Melanocytic tumor diagnosis comprises one of the most challenging areas in pathology due in part to morphologic heterogeneity. Histopathologic features range from bland-appearing nevoid melanoma to variants with highly pleomorphic morphology. When the architectural and cytologic features are beyond threshold criteria, consensus on the diagnosis may be relatively uniform. However, when more than 1 morphology is present within a single lesion, the Breslow thickness may be in question and accurately distinguishing melanoma from an associated dermal nevus can be extremely challenging, leading to the use of ancillary stains and expert consultation. No immunohistochemical stain can reliably distinguish benign versus malignant and instead must be interpreted in the context of all the available information, both clinically and histopathologically.
Primary cilia (PC) are cell surface microtubular organelles that are often compared with antennae due to their involvement in sensing and transducing various extracellular signals to initiate cellular signaling.1 They are composed of the extracellular portion termed the axoneme and the intracellular base termed the basal body (formed from the mother centriole), which anchors the cilium to the cell surface. The PC senses the extracellular environment through resident receptor and effector molecules, which include Hedgehog, canonical and noncanonical Wnt and Notch signaling.2–5 It is therefore not surprising that ciliary dysfunction can lead to altered cell cycle dynamics, loss of response to growth signals, loss of cell polarity control, and altered interaction with the extracellular milieu, as characterized in classic ciliopathies.6 In fact, studies have shown the loss of PC in pancreatic cancer, renal cell carcinoma, breast cancer, cholangiocarcinoma, and melanoma.7–11
As the centrioles that would otherwise anchor the PC are diverted for use in the cell cycle, the lack of a cell surface PC has close correlations with cell cycle progression. However, studies have shown that an increased proliferation rate does not completely explain the close to complete lack of PC in melanoma. Likewise, inducing cell cycle arrest in melanoma cell lines when compared with control melanocytes does not result in complete recovery of PC formation.12 Therefore, we hypothesize that the cumulative changes in melanoma cells, regardless of the precise unique set of genetic changes, culminate in inhibiting or disrupting PC formation. To further our understanding of the status of PC in melanocytic lesions, we collected a set of combined melanocytic lesions in which a nevus component is identified next to an invasive melanoma. In these cases, the ciliation index of the nevus portion can be used as an internal control for the relative loss of PC formation in the melanoma portion.
MATERIALS AND METHODS
Tissue samples were procured from the archives of the University of California, San Francisco, and all cases were reviewed, and histopathologic diagnosis was reverified (J.K., T.H.M., and U.E.L.).
Microscopy and Analysis
Immunostaining and microscopy was performed, as previously described.8 Briefly, 4-μm-thick formalin-fixed paraffin-embedded sections were deparaffinized in xylene, rehydrated through graded ethanol dilutions, and subjected to antigen retrieval according to standard procedures. These specimens were costained with antibodies against SOX10 to detect melanocytes,13 acetylated alpha-Tubulin to detect the modified tubulins enriched in the ciliary axoneme,14 and gamma-Tubulin to detect the basal bodies of cilia15 (gamma-Tubulin also marks the centrosome when it is not elongating a ciliary axoneme). To assess the abundance of PC in SOX10-positive melanocytes, we defined an elongated acetylated alpha-Tubulin–positive structure in association with a gamma-Tubulin–positive structure to be a PC (as previously described).16 The counting was performed using a Keyence BZ-X710 (Keyence Corporation of America, Itasca, IL) fluorescence microscope and its accompanying software at ×60 magnification.
Antibodies and Reagents
Immunofluorescence staining was performed using the following antibodies: mouse anti-SOX10 IgG1 (catalog #API 3099; Biocare, Pacheco, CA), rabbit anti–gamma-Tubulin (catalog #T5192; Sigma, Burlington, MA), and mouse anti–acetylated-alpha-Tubulin IgG2b (catalog #T6793; Sigma) antibodies. The following secondary antibodies were used: Alexa Fluor 594 donkey anti-mouse IgG1 (catalog #A-21203; Life Technologies, Carlsbad, CA), Alexa Fluor 647 chicken anti-rabbit (catalog #A-21443; Life Technologies), and Alexa Fluor 488 goat anti-mouse IgG2b (catalog #A-21141; Life Technologies). Primary antibodies were used at 1:2000 and secondary antibodies at 1:300 dilution. Slides were mounted using ProLong Gold Antifade Mountant (catalog #P36930; Life Technologies) and stored at room temperature.
Results were obtained using Microsoft Excel software and expressed as mean ± SE. P values (Student 2-tailed, unpaired t test). Statistical significance is considered for data with P ≤ 0.05.
Clinical, Histopathologic, and Immunohistochemical Findings
Invasive melanomas with associated melanocytic nevi from 10 patients were studied. Patients had a median age of 61 years (range 35–78 years). The lesions were located on the back, arm, abdomen, and shoulder. There were 4 women and 6 men (Table 1). Characteristic architectural and cytomorphologic features were used to distinguish the invasive melanoma component from the dermal nevus within the lesions (Figs. 1A, B). In addition, HMB-45 staining was performed and demonstrated strong staining within the melanoma portion and loss of staining within the nevus in all cases (Table 2 and Fig. 1C). The nevus portion in all samples retained p16 expression, whereas the melanoma portion showed either a partial or a total loss of staining in 9/10 samples (Fig. 1D). Also, a combined Melan-A/Ki-67 stain was performed and showed a mild increase in the number of positive melanocytes within the melanoma portion (Fig. 1E, arrow-heads show positive cells) and no staining in the nevus portion.
Primary Cilia Status
An average of 192 melanocytes (range: 150–250) were counted per nevus region, and an average of 174 melanocytes (range: 150–250) were counted per melanoma region. A positive melanocyte was defined by the presence of centrioles (dot-like structures) along with an elongated ciliary axoneme (hair-like structures). Immunofluorescence results revealed that a mean of 4% ciliation (SD: 7%) in the melanoma component, whereas the associated nevus component had an average of 59% ciliation (SD: 17%). There was a significant decrease in the number of PC in the melanoma versus nevus components in each of these cases (P < 0.00001) (Table 1 and Figs. 2 and 3).
Although we counted the number of PC containing melanocytes in each component to be quantitative, qualitative assessment was interpreted as being “high” in the nevus portion and “low” in the melanoma portion for each case. There was fairly uniform ciliation in each component of these lesions; therefore, random areas were selected for counting cells. These results are similar to previous reports of melanoma having very low percentage ciliation.8 A benign congenital melanocytic nevus was used as a positive control with a 98% ciliation index.
Previous work has demonstrated that PC are lost in melanoma in situ, primary invasive melanoma, and metastatic melanoma, whereas retained in melanocytic nevi.8 Our current study demonstrates an additional utility of using the relative PC loss within combined melanocytic lesions to distinguish benign from malignant. This technique requires minimal tissue in the form of 1 unstained formalin-fixed paraffin embedded slide, and the turnaround time can be as little as 1–2 days. However, the lack of automation and requirement of an immunofluorescent microscope for the actual analysis may be restrictive for use in some laboratories. As our study is limited to 10 cases, a larger cohort for validation is also still necessary.
We performed immunohistochemical stains for HMB-45 and p16 and found that the nevus component demonstrated retention of p16 staining in all cases, and the melanoma portion showed either partial or total loss in 9/10 cases. HMB-45 stained strongly in the melanoma and was lost in the nevus component. These additional stains further support the original diagnosis made by histopathologic examination. The 1 case (#3) that showed retention of p16 in both melanoma and nevus highlights the limits of usefulness that a single immunohistochemical stain has in any diagnosis. A combined Melan-A/Ki-67 stain demonstrated a mild increase in the proliferative index in the melanoma portion compared with the nevus but did not come close to the relative change in the ciliation index between the 2 components.
Compared with the benign congenital melanocytic nevus used as the control with a 98% ciliation index, the average 59% ciliation within the nevoid component of combined cases raises the question of where on the spectrum of malignant potential these associated nevi belong. One could hypothesize that these associated nevi are in a more permissive state of additional transformation due to partial dysregulation of the PC. As the PC is involved in many biological processes including cell cycle, structural influences of the cytoskeleton, cellular proteostasis, and signal transduction, it is equally possible that alterations to any of these pathways could have effects on PC status. Although analyzing the genetic landscape of these combined lesions was beyond the scope of this project, it will be a critical next step in further understanding and characterizing the various factors that lead to melanoma progression.
Although most melanomas are believed to arise de novo and not in association with a precursor nevus, our ability to reliably distinguish an existing precursor within a malignancy generally depends on there being distinct morphologic or architectural differences between the 2 components. Although the cases used in this study demonstrated morphologically distinct regions of melanoma versus nevus with immunohistochemical staining patterns further supporting the diagnosis, this is not always the case. Using the presence or absence of an additional morphologic feature, such as the PC, to further analyze a given lesion may be helpful in challenging cases. The study of the PC in this context is one example of a biologic state the melanocyte may have reached after integrating genetic mutations, epigenetic changes, posttranslational modification, and signaling from the extracellular environment.
The underlying mechanisms for PC loss are undoubtedly multifactorial, reflecting the complexities of a dynamic cellular process, and it is currently unclear whether the loss of PC plays a role in cancer progression or is a consequence of transformation. Ultimately, these findings raise limitless questions, which must be addressed experimentally to further define our understanding of how melanocytes and their respective PC organelle coexist. Until that point, we can still leverage the potential of the ciliation index as a novel marker of malignant potential, to be used within the context of the overall clinical and microscopic features.
1. Satir P, Pedersen LB, Christensen ST. The primary cilium at a glance. J Cell Sci. 2010;123:499–503.
2. Goetz SC, Anderson KV. The primary cilium: a signalling centre during vertebrate development. Nat Rev Genet. 2010;11:331–344.
3. Hassounah NB, Bunch TA, McDermott KM. Molecular pathways: the role of primary cilia
in cancer progression and therapeutics with a focus on hedgehog signaling. Clin Cancer Res. 2012;18:2429–2435.
4. Berbari NF, O'Connor AK, Haycraft CJ, et al. The primary cilium as a complex signaling center. Curr Biol. 2009;19:R526–R535.
5. Wong SY, Seol AD, So PL, et al. Primary cilia
can both mediate and suppress hedgehog pathway-dependent tumorigenesis. Nat Med. 2009;15:1055–1061.
6. Oh EC, Katsanis N. Cilia in vertebrate development and disease. Development. 2012;139:443–448.
7. Hassounah NB, Nagle R, Saboda K, et al. Primary cilia
are lost in preinvasive and invasive prostate cancer. PLoS One. 2013;8:e68521.
8. Kim J, Dabiri S, Seeley ES. Primary cilium depletion typifies cutaneous melanoma
in situ and malignant melanoma
. PLoS One. 2011;6:e27410.
9. Seeley ES, Carrière C, Goetze T, et al. Pancreatic cancer and precursor pancreatic intraepithelial neoplasia lesions are devoid of primary cilia
. Cancer Res. 2009;69:422–430.
10. Menzl I, Lebeau L, Pandey R, et al. Loss of primary cilia
occurs early in breast cancer development. Cilia. 2014;3:7.
11. Razumilava N, Gradilone SA, Smoot RL, et al. Non-canonical hedgehog signaling contributes to chemotaxis in cholangiocarcinoma. J Hepatol. 2014;60:599–605.
12. Snedecor ER, Sung CC, Moncayo A, et al. Loss of primary cilia
cells is likely independent of proliferation and cell cycle progression. J Invest Dermatol. 2015;135:1456–1458.
13. Shin J, Vincent JG, Cuda JD, et al. Sox10 is expressed in primary melanocytic
neoplasms of various histologies but not in fibrohistiocytic proliferations and histiocytoses. J Am Acad Dermatol. 2012;67:717–726.
14. Piperno G, LeDizet M, Chang XJ. Microtubules containing acetylated alpha-tubulin in mammalian cells in culture. J Cell Biol. 1987;104:289–302.
15. Mahjoub MR, Stearns T. Supernumerary centrosomes nucleate extra cilia and compromise primary cilium signaling. Curr Biol. 2012;22:1628–1634.
16. Schmidts M, Vodopiutz J, Christou-Savina S, et al. Mutations in the gene encoding IFT dynein complex component WDR34 cause jeune asphyxiating thoracic dystrophy. Am J Hum Genet. 2013;93:932–944.