Insights into high-risk multiple myeloma from an analysis of the role of PHF19 in cancer

Abstract

Despite improvements in outcome, 15-25% of newly diagnosed multiple myeloma (MM) patients have treatment resistant high-risk (HR) disease with a poor survival. The lack of a genetic basis for HR has focused attention on the role played by epigenetic changes. Aberrant expression and somatic mutations affecting genes involved in the regulation of tri-methylation of the lysine (K) 27 on histone 3 H3 (H3K27me3) are common in cancer. H3K27me3 is catalyzed by EZH2, the catalytic subunit of the Polycomb Repressive Complex 2 (PRC2). The deregulation of H3K27me3 has been shown to be involved in oncogenic transformation and tumor progression in a variety of hematological malignancies including MM. Recently we have shown that aberrant overexpression of the PRC2 subunit PHD Finger Protein 19 (PHF19) is the most significant overall contributor to HR status further focusing attention on the role played by epigenetic change in MM. By modulating both the PRC2/EZH2 catalytic activity and recruitment, PHF19 regulates the expression of key genes involved in cell growth and differentiation. Here we review the expression, regulation and function of PHF19 both in normal and the pathological contexts of solid cancers and MM. We present evidence that strongly implicates PHF19 in the regulation of genes important in cell cycle and the genetic stability of MM cells making it highly relevant to HR MM behavior. A detailed understanding of the normal and pathological functions of PHF19 will allow us to design therapeutic strategies able to target aggressive subsets of MM.

Keywords: Cancer progression; EZH2; Epigenetic; Multiple Myeloma; PHF19; PRC2; Polycomb Repressive Complex 2.

Fig. 1

Human PHF19 isoforms and Top 20 TFs and chromatin regulators that regulate PHF19 expression in cancer. A Schematic representation of PHF19 locus, gene and isoforms. B The Cistrome DB Toolkit (http://dbtoolkit.cistrome.org/) was used to identify what TF likely regulates PHF19 expression spanning a region of ∼10 kb upstream the transcription start site (TSS). The Y axis represents the regulatory potential score which were calculated by Cistrome DB Toolkit [23, 24]. The x-axis represents the different TFs

Fig. 2

PHF19 expression in normal human and mouse hematopoiesis in normal human and mouse B-cell differentiation. PHF19 expression in human (A) and mouse (B) expression profiles in normal hematopoiesis as reported in the Blood Spot database [54] (http://servers.binf.ku.dk/bloodspot/, accessed on 27 July 2021). A the Normal human hematopoiesis (HemaExplorer) dataset was used. B the Mouse normal hematopoietic system (upper) and Mouse Normal RNA-Seq (Lower) datasets were used. LT-HSC: Long Term Hematopoietic Stem Cell, ST-HSC: Short term Hematopoietic stem cell, HSC: Hematopoietic Stem Cell, LMPP: Lymphoid-primed multipotential progenitors, MPP: Multipotent Progenitor, CLP: Common Lymphoid Progenitor, CMP: Common Myeloid Progenitor, GMP: Granulocyte Monocyte Progenitor. PHF19 expression in mouse (C) and human (D and E) expression profiles in normal human and mouse B-cell differentiation. C Phf19 expression as reported in the Blood Spot database [54] (http://servers.binf.ku.dk/bloodspot/, accessed on 27 July 2021) using the Mouse immgen B cells dataset (upper) and as reported in Immgen dataset (http://rstats.immgen.org/Skyline/skyline.html, accessed on 27 July 2021) using RNA-seq Gene Skyline tool (Lower). D and E PHF19 expression as reported in the GenomicScape database [55] (http://genomicscape.com/microarray/expression.php, accessed on 27 July 2021) using the Human B cells to plasma cells GCRMA dataset (D) and Human B cells to plasma cells (in-vitro) dataset (E). NBC: Naive B cells (n=5), CB: Centroblasts (n=4), CC: Centrocytes (n=4), MBC: Memory B cells (n=5), prePB: preplasmablasts (n=5), PB: Plasmablasts (n=5), PC: Early plasma cells (n=5), BMPC: Bone marrow plasma cells (n=5), act.BC: Activated B-cells (n=5)

Fig. 3

Chromatin marks and transcription factors at the PHF19 locus and PHF19 expression in human MM cell lines. A Snapshot from UCSC genome browser showing binding events within the PHF19 locus in the MM1S cell line. B PHF19 expression as reported using RNA-seq in the Cancer Cell Line Encyclopedia (CCLE) database [104] (https://depmap.org/portal/download/, accessed on 25 July 2021) (upper) and as reported in Keats lab dataset (https://www.keatslab.org/data-repository) (Lower). C PHF19 expression as reported using microarray in the CCLE database

Fig. 4

Schema showing suggested mechanism of action of PHF19 overexpression in MM cell lines. Ren Z et al. showed that PHF19 promotes PRC2 activity and represses cell cycle inhibitor genes. Yu T et al. showed that PHF19 impedes PRC2 activity by promoting the phosphorylation of EZH2 via AKT pathway leading to EZH2 inactivation leading to an increase in expression of genes that play an important role in MM. Schinke C et al. showed that PHF19 repressed the expression of tumor suppressor protein (TSP) and upregulated the expression of pro-survival and proliferation genes

Fig. 5

High PHF19 level is associated with high-risk and PHF19 transcriptional signature in MM suggests that it might regulate cell cycle progression. A and B PHF19 is expressed in all MM subtypes with higher expression levels seen in the non-hyperdiploid subgroups (X2=38, p=5.7e^-10) and HR patients defined by GEP70. Interestingly, PHF19 expression is significantly higher at relapse (t=2.8, df=34, p=0.006). C Using a logrank test we identify a PHF19 expression level of 9.65 as an optimal cut point for overall survival (OS) splitting the population into high and low PHF19 expressing groups. The OS of patients with elevated PHF19 expression is significantly shorter than patients with lower PHF19 expression (HR=2.98 (2.2-4), p=3.68e^-13. D PHF19 expression is significantly higher at relapse (t=2.8, df=34, p=0.006). E PHF19high and low groups were defined using an elbow test. F Volcano plot showing genes differentially expressed between PHF19high and PHF19low MM samples. Analysis identified 835 differentially expressed genes (DEG) (Fold change >2, FDR < 0.05), with 547 (65%) upregulated and 288 (35%) downregulated genes. G Gene set enrichment analysis (GSEA) analysis of the differentially expressed genes between PHF19 high and low MM samples (Gene ontology (GO)). H Venn diagram showing the overlap of: 1- DEG between PHF19high and PHF19low MM samples, 2- Downregulated genes between PHF19-KD and PHF19-WT in MM1S cell line (i.e. genes upregulated by PHF19), 3- Upregulated genes between PHF19-overexpression (rescue) and PHF19-KD in MM1S cell line (i.e. genes upregulated by PHF19) and 4- upregulated genes between PHF19high and PHF19low MM samples. I Venn diagram showing the overlap of: 1- the 294 genes defining PHF19 signature in MM, 2- DEG between centroblasts (CB) and bone marrow plasma cells (BMPC) and 3- DEG between preplasmablasts (PrePB) and BMPC

Fig. 6

Single cell RNA-seq data from germinal centers links PHF19 expression modulation to cell cycle. A PHF19 expression is the highest in cycling GC B-cells and preplasmablasts and plasmablasts. Upper panel shows a UMAP projection and cluster identification from single cell RNAseq profiles of human B cells maturation in tonsils [120]. Lower panel shows UMAP plot showing expression of PHF19 in the different B cells clusters as shown in the upper panel (www.tonsilimmune.org/, accessed on 16 June 2021). B Heatmap displaying the relative expression fold change (log2) of PHF19 and selected genes in clusters of dark zone B cells representing different stages of the cell cycle (data are from [119]. Ten clusters were identified by PhenoGraph in dark zone GC B cells based on the expression of genes associated with the S-G2-M stages of the cell cycle: three clusters of cells transitioning from G1 to S phase (C1, C2 and C3), two clusters in the S-phase (C4 and C5), three clusters of cells transitioning from G2 to M phase (C6, C7 and C8), one cluster in the M-phase (C9) and one cluster of cells transitioning from M to G1 phase (C10). The differential expression analysis is performed by comparing each cluster to all the others. PHF19 expression peaks in the G2-M phase of the cell cycle and this is followed by the modulation of several genes of the PHF19 transcriptional signature

Fig. 7

PHF19 expression is associated with cell proliferation in MM. A PHF19 expression in molecular subtype of MM, proliferation subgroup (PR, n=47), low bone disease (LBD, n=58), MMSET (n=68), hyperdiploidy (HD, n=116), Cyclin D-1 (CD-1, n=28), Cyclin D-2 (CD-2, n=60) and MAF (n=37). B Coexpression analysis of PHF19 using GenomicScape database [55] (http://genomicscape.com/microarray/expression.php, accessed on 03 September 2021). Heat maps showing the top 30 genes positively correlated with PHF19 in a UAMS cohort [121]. C Gene set enrichment analysis (GSEA) analysis was performed using g:Profiler web tool (Biological Process ontology, GO:BP, https://biit.cs.ut.ee/gprofiler/gost). D Venn diagram showing the overlap between the 294 genes identified as the PHF19 transcriptional signature of MM and the 209 genes that positively correlate with expression of PHF19