Background
Macrophage migration inhibitory factor (MIF), so named because it inhibited the random migration of macrophages, was first discovered as a cytokine product of T lymphocytes [
1,
2]. It is now known that a variety of other cells types produce MIF, including other immune cells, endocrine, endothelial and epithelial cells [
3,
4]. High levels of MIF have also been reported
in vivo in several cancer types including breast [
5], lung [
6] and gastric cancers [
7] and the work of several groups points to a correlation between MIF expression and cancer prognosis, e.g. head and neck cancer and glioblastoma [
8‐
10]. Moreover, findings that MIF is involved in critical pro-survival signalling pathways together with cell cycle control has provided interest in possible associations with the development and progression of cancer.
MIF protein is stored in pre-formed, cytoplasmic pools and is rapidly released in response to stimuli such as microbial products, proliferative signals and hypoxia [
3,
4,
11] through a nonconventional ABC transporter pathway [
12]. It is considered to be atypical of the conventional classes of cytokines with known functions extended to roles as both a hormone and an enzyme (reviewed in [
3,
13]). MIF has also been shown to play a role in cell proliferation where it has been suggested to be involved in the development and progression of cancer, acting as an extracellular, pro-tumourigenic factor [
14,
15].
The transmission of MIF signals occurs through a number of receptor systems, the first identified being the cell surface receptor CD74 [
16]. CD74 itself lacks intracellular signalling domains [
17] but it has been shown that CD44 acts as a co-receptor for CD74 to provide the means whereby MIF signals are transmitted [
18]. More recently, the CXC chemokine receptors CXCR2 and CXCR4 were also identified as MIF receptors and CD74 has also been shown to form functional heteromeric receptor complexes with CXCR2 and CXCR4 [
19,
20]. Depending on the cellular context, binding of MIF to its known cell surface receptors can lead to activation of two fundamental signalling axes, namely the mitogen-activated protein kinase (MAPK) pathway and PI3K/Akt signalling [
14,
21‐
23]. In addition, the cytoplasmic pool of MIF has also been shown to exert other signalling actions.
MIF expression has also been shown to be of significance in a limited number of studies investigating melanoma biology. Higher levels of MIF mRNA were identified within isogenic variants of the human A375 melanoma selected for higher metastatic potential in nude mice [
24]. Inhibition of MIF expression in the G361 human melanoma cell line resulted in inhibition of proliferation, migration and tumour-induced angiogenesis [
25]. MIF production was also shown in human uveal melanoma cell lines whereby MIF prevented their lysis by NK cells [
26]. Additionally, in the B16-F10 mouse melanoma model, inhibition of MIF by RNAi significantly delayed tumour establishment when injected into mice [
27]. Collectively these results implicate MIF in melanoma progression, but despite this evidence, little is known on the downstream signalling pathways regulated by MIF signalling, nor has this concept been evaluated in patient studies.
In the present study, we sought to establish the primary downstream signalling pathways activated by MIF in a panel of human melanoma cell lines in vitro using specific knock-down studies and determine the prognostic significance of MIF expression in metastatic melanoma. Our data demonstrates that MIF is involved in melanoma proliferation and anchorage-independent growth, mediated through the activity of the PI3K/Akt pathway. We also establish that in clinical melanoma samples, MIF expression increases with metastatic progression and is correlated with survival for metastatic melanoma patients. Taken together, these results highlight the importance of the MIF-signalling axis with implications for targeted treatment approaches in melanoma.
Discussion
To date, apart from two studies each using a single cell line [
24,
25] the role of MIF in the context of human cutaneous melanoma has not been intensively studied. In the present report we adopted an siRNA-based strategy to examine the function of endogenous MIF expression in multiple human melanoma cell lines. In MelCV and Me1007 cell lines, MIF knockdown resulted in significantly reduced cell number and viability over 6 days, indicating that endogenous MIF expression could be generally required for the growth of melanoma cells. The reduced cell numbers corresponded to the increased accumulation of cells in G0/1 and a decrease of cells in S-phase. Moreover, accounting for the successive reduction in the number of viable cells during the experiment, there was an increased proportion of apoptotic cells following MIF depletion. Similar findings were also obtained when considering anchorage-independent growth where it was shown that MIF siRNA transfection significantly compromised the number and size of colonies formed by melanoma cells.
To better understand the role of MIF expression in melanoma cells, further quantitative assays were employed on six different melanoma cell lines. Cell proliferation after MIF knockdown was further explored using the Click-iT assay, a sensitive and quantitative assay which measures the cell cycle. In particular the assay provides an accurate measure of the number of cells entering S-phase in a fixed time period. This analysis showed that MIF knockdown significantly reduced cells transitioning to the S-phase in four of the six melanoma cell lines suggesting the proliferative capacity of the majority of the melanoma cell lines studied have some degree of reliance on MIF expression. In agreement with these findings, work from several authors have shown that MIF is involved in cell cycle regulation in different cancer cells [
14,
23,
40], and MIF knockdown can cause G1 arrest by inhibiting G1/S transition [
41]. At least for the MelCV and Me1007 lines examined in detail, MIF depletion was clearly cytostatic but also compromised cell viability. Collectively this reinforces the idea that MIF signalling displays potential as a pathway that could be targeted for melanoma treatment.
Leading on from these findings the question is raised as to how MIF functions in this setting. We could not establish that the sensitivity of individual melanoma cell lines to MIF depletion resulted from the differential expression of known MIF receptors (i.e. CD74/CD44 and/or CXCR4). We also considered the responsiveness of cells lines in the context of known downstream signalling pathways. It is well established that MIF function is associated with two major pro-survival pathways, namely the MAPK and PI3K/Akt signalling pathways, each known to be important in melanoma. Indeed, oncogenic MAPK signalling through ERK (RAS-RAF-MEK-ERK) is constitutively activated in the majority of melanomas [
42,
43] with aberrant activation frequently stemming from activating mutations in BRAF. Of these, the most common BRAF mutation occurring in ~50% of all melanomas comprises a glutamic acid to valine substitution at position 600 (V600E) [
43‐
45]. Pharmacological inhibition of the RAF/MEK/ERK pathway, in particular, via inhibition of mutated and activated BRAF, has therefore appeared as a promising strategy for treatment. This has led to the development of mutant-BRAF-specific inhibitors [
46‐
49] that have shown promising results in clinical trials [
47,
50,
51]. In our study, of the melanoma cell lines tested for the effects of MIF inhibition, three express wildtype BRAF (Me1007, MelRM and MelFH) while the others bear the BRAF V600E mutation (MelCV, MelRMu and MM200 [
52]). Two of the three lines most sensitive to MIF depletion are BRAF mutants (MelCV and MelRMu) indicating the effects of MIF signalling in melanoma were likely outside this pathway. This observation has important therapeutic implications in patients that are resistant to mutant BRAF inhibitors (e.g. [
53]) whereby MIF depletion/targeting could be used as an alternative strategy.
Alongside the MAPK pathway, constitutive activation of the PI3K/Akt pathway is also important in a high proportion of melanoma cases [
54‐
56]. Such activation is frequently associated with down-regulation or loss of PTEN phosphatase that antagonises PI3K signalling. Absent or decreased PTEN expression occurs in up to 90% of primary melanomas with mutations of PTEN or loss of heterozygosity at the PTEN locus accounting for this deficiency [
57,
58]. Transcriptional repression of the gene by promoter methylation also occurs in a high proportion of metastatic melanomas [
59,
60]. In addition to PTEN, other major elements of the PI3K/Akt pathway are found to be amplified or mutated in melanoma. Akt3 expression is increased as a result of elevated gene copy number in ~50% and ~70% of primary melanomas and metastatic melanomas, respectively [
36]. PI3K/Akt pathway activation in melanoma can also occur through mutational activation of PI3KCA along with the mutational activation of upstream receptor tyrosine kinases such as c-KIT and EGFR [
61,
62]. Given the previous links established between MIF and Akt signalling as described in the Introduction, we focussed our efforts on investigating Akt as the likely major pathway downstream of MIF in melanoma.
Akt activation can stimulate proliferation through multiple downstream targets that affect cell-cycle regulation [
63]. For example, Akt can phosphorylate p27 (Kip1) cyclin-dependent kinase inhibitor, preventing its localisation to the nucleus and attenuating its cell-cycle inhibitory effects [
64‐
66]. In addition, Akt also serves to phosphorylate and inactivate GSK3. As GSK3 phosphorylates cyclin D and cyclin E and targets them for proteasomal degradation [
67,
68], inhibition of GSK3 by Akt thereby acts to stabilise key cyclins involved in cell cycle entry. In the current study, MIF knockdown resulted in decreased Akt phosphorylation in all melanoma cell lines tested, albeit to varying degrees. This effect was accompanied by a reduction in expression of cyclin D1 and cyclin dependent kinase 4 (CDK4), and an increased expression of p27. Moreover there was a correlation between the effects of MIF knockdown and the degree of Akt activation amongst individual cell lines. Collectively this suggests that activation of the Akt pathway is one of the major mechanisms whereby MIF contributes to cell cycle regulation in melanoma. Since the overall importance of the PI3K/Akt pathway to melanoma biology cannot be understated, these findings imply clinical significance of MIF signalling in this disease.
Supporting this notion, a previous study of melanocytic tumours showed MIF mRNA and protein levels were high in malignant tumours while expression was significantly lower in benign naevi [
69]. Here we verified these data using independent expression microarray datasets where collectively these findings support the general concept that MIF is differentially expressed between non-malignant and malignant lesions with increased expression during melanoma progression. Despite these observations and previous work associating increased MIF with enhanced melanoma growth and metastasis in nude mice [
24], the clinical significance of MIF tumour levels has surprisingly not been previously examined. Our
in silico analyses of GEO datasets now reveal significant correlations between MIF expression and patient outcome. While MIF levels in primary tumours had no bearing on patient outcomes there was a clear indication that high MIF expression levels in metastatic lesions were significantly associated with shorter survival times. Indeed, in the GSE19234 cohort, ~70% of patients whose tumours has lower MIF expression remained alive approaching 40 months of clinical follow-up.
One caveat to consider when linking our
in vitro findings to the clinical setting is the inherent complexity of tumour tissues
in vivo. It seems probable that MIF expression in melanoma cells has an impact upon their proliferative capacity
in vivo but whether the MIF gene expression detected in clinical samples is wholly tumour derived is not entirely clear. Tumour tissue comprises a non-homogenous mixture of tumour and stromal cells including variable amounts of infiltrating immune cells. In breast cancer, MIF is expressed in both tumour cells and stromal cells, including tumour-associated macrophages [
5]. Indeed, MIF is a key cytokine in both innate and adaptive immune cells [
3,
4] and thus infiltrating immune cells must also be considered as an intra-tumoural source of MIF. Some breast cancer cells respond to exogenous MIF by triggering a massive burst of MIF secretion suggesting auto- or paracrine regulation of MIF [
70]. Finally, it is already well established that interactions between the tumour and its microenvironment play an important role in influencing the behaviour of tumour cells. Again in breast cancer, it was found that MIF was highly upregulated in tumours cells when they were co-cultured with macrophages. In turn, increased MIF secretion by tumour cells contributed to metalloproteinase production by the macrophages and this augmented the invasive potential of the tumour cells [
70]. Similarly it is known that tumour-associated macrophages can enhance melanoma growth though secreted factors (e.g. [
71]) and equally there are other infiltrating cells such as lymphocytes [
72] which are also potential sources of MIF. However the relative importance of MIF production in melanoma tumour cells versus stromal cells remains to be established.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
Conceived and designed the experiments: CSO, CEDB, RFT. Performed the experiments: CSO, TJM, ES, XYG, RFT. Contributed clinical samples and expertise: PH, XDZ. Analyzed the data: CSO, CEDB, RFT, TJM, ES. Contributed reagents/materials/analysis tools: XDZ, PH. Wrote the manuscript: CSO, CEDB, RFT. All authors read and approved the final manuscript.