Macrophage migration-inhibitory factor is associated with chromosome 8q gain and Vasculogenesis in UM
Zahra Souri1 *, Martine. J. Jager2
- Department of Cell and Molecular Biology and Microbiology, Biological Science and Technology, University of Isfahan, Iran
- Department of Ophthalmology, LUMC, Leiden, The Netherlands
Abstract: Uveal Melanoma (UM) is a rare type of intra-ocular malignancy that is highly associated with inflammation. Tumour-infiltrated leukocytes play a major role in the development of metastasis in UM. Macrophage migration-inhibitory factor (MIF), is known as a pro-inflammatory cytokine associated with the prevalence of leukocytes in the tumour microenvironment of some cancers. As chromosome 8q gain influences macrophage infiltration, we investigate whether MIF shows any association with chromosome 8q gain in UM and evaluate the possible association of the MIF signalling cascade with vascular factors.
Methods: The expression of MIF, CD74, CD34, ICAM1, ICAM3, PECAM, HIF1-α and VHL was determined using an Illumina HT12V4 array in 64 primary UM. In order to detect Chromosome aberrations, SNP analysis was performed using the Affymetrix 250K_NSP and Affymetrix SNP 6.0 array. Immunohistochemical staining was performed for the detection of BAP1 status.
Results: We found that high MIF was related to worse survival, ciliary body involvement and metastasis in our study cohort (P= 0.007; P= 0.003, P= 0.008).
MIF was positively correlated with vasculogenic markers CD34 (P= 0.02; P= 0.05), ICAM1 (P= 0.02), ICAM3 (P= 0.006), PECAM (P= 0.003) and HIF-1α (P= 0.002), while negatively correlated with VHL (P= 0.02; P= 0.001, P= 0.01).
When we looked at all 64 UM tumours, high MIF was related to loss of chromosome 3, loss of BAP1 expression and extra copies of chromosome 8q (P= 0.001; P= 0.002, P= 0.001); additionally MIF showed a higher expression pattern in tumours which had extra copies of chromosome 8q compared to tumours with normal chromosome 8q when we only looked at BAP1 positive tumours (P= 0.005).
Moreover we found that although MIF was not correlated with CD68 and CD163 markers (P= 0.82; P= 0.50) when BAP1 was normal, it still positively correlated with CD34 (P= 0.04; P= 0.02), ICAM1 (P= 0.03) and ICAM3 (P= 0.005).
Conclusion: MIF is a pro-inflammatory soluble factor inside the aqueous humour; we find that MIF is overexpressed in high-risk tumours which are monosomy 3 and BAP1-negative and also tumours which are BAP1 positive with 8q gain. Moreover, as we find an association between MIF and vasculogenic markers, we suggest that targeting MIF signalling cascade could dampen inflammation and vasculogenesis in UM.
