Background: A critical goal of precision oncology is the identification of deregulated signaling pathways that serve as disease-specific biomarkers and vulnerabilities that may be exploited as therapeutic targets. Cutaneous T cell lymphomas (CTCLs) are heterogeneous neoplasms with variable clinical outcomes. Whereas the genomic landscapes of CTCLs are being characterized, a large scale analysis of their proteomic signatures and post-translational modifications (PTM) has not been performed. In this study, we sought to gain insights into the pathogenesis of CTCLs using high mass accuracy tandem mass spectrometry (HMA-MS/MS)-based phosphoproteomics and glycoproteomics.

Methods: CTCL-derived cell lines representative of Sézary syndrome (SS, HUT78), mycosis fungoides (MF, HH), and ALK-negative anaplastic large cell lymphoma (ALCL) with PCM1-JAK2 or NPM1-TYK2 translocations (Mac1, Mac2a, and Myla) were analyzed in technical and biologic triplicates to perform unbiased proteome characterization. Phosphoproteins were enriched using metal oxide affinity chromatography and immunoprecipitation using 3 anti-phosphotyrosine antibodies and N-glycoproteins were enriched using solid phase extraction and hydrazide resin and an Orbitrap Fusion or LTQ OrbitrapXL in-line with a Paradigm MS2 HPLC. Precise site-mapping of the PTMs was assigned and quantitated using Maxquant. Hierarchical clustering was performed using Perseus. Whole exome sequencing (WES) was performed using an Illumina HiSeq 2500. Functional studies using tyrosine kinase inhibitors were performed to assess selective vulnerabilities.

Results: Phosphoproteomic analysis identified 1899 tyrosine (Y) phosphoproteins and 6029 serine/threonine (S/T) phosphoproteins. Unsupervised hierarchical clustering of the Y-phosphoproteome data accurately grouped the CTCLs to those with and without underlying tyrosine kinase translocations. Unique driver kinase phosphosignaling signatures were identified in MYLA (NPM1-TYK2) and MAC1/2 ( PCM1-JAK2) representing the ALCL subgroup, as well as close similarity between the signatures of MF and SS-derived cells (HH and HUT78). Notably, phosphoproteomic analysis of HUT78 (SS) yielded high spectral counts of Y-phosphorylated residues at IL2RG p.Y325, JAK3 p.Y980/Y981, and STAT5 p.Y699 indicative of constitutive activation of the JAK/STAT5 pathway corroborated by WES showing activating mutations in JAK3 p.A573V and JAK1 p.Y654F.

Beyond driver kinases, oncogenic kinase-specific substrates with distinct site-specific phosphorylated residues were identified for each disease class. Phosphoproteins selectively expressed in MF/SS, (46 p-S/T and 25 p-Y proteins) included members of the killer-cell immunoglobulin-like receptor (KIR) family implicated in immune modulation through immune tyrosine-based inhibitor or TYRO protein tyrosine kinase binding domains for signal transduction. Additionally, proteins not previously associated with MF/SS including NKG2-D type integral membrane protein family members were discriminative of this subgroup. Phosphoprotein members of lysine demethylases (KDM family) and Aryl Hydrocarbon Receptor families were distinctively identified in MF/SS cells, while components of the T-cell receptor signaling pathways were common in all groups of CTCLs. By comparison, N-glycoproteomic analysis identified 811 unique N-glycoproteins. A total of 51 N-glycoproteins were found solely in HUT78 of which 46 were novel in their specificity and included 13 with a role in cell-adhesion while N-glycoproteins unique to HH revealed 12 novel candidate biomarkers (e.g. MERTK, CD97 and SLC44A2). Integrative phosphoproteomic and N-glycoproteomic analysis strikingly revealed convergent pathways implicating a prominent role for the KIR family proteins in SS. Orthogonal immunophenotyping by flow cytometry has shown high sensitivity (96%) detection of KIR3DL2 in primary human SS cases. Significantly, anti-KIR3DL2 inhibition using a humanized antibody revealed activity against SS cells. By comparison, TYK2-driven CTCL (MYLA) was sensitive to inhibitors of TYK2 and JAK.

Conclusions: Integrated phosphoproteomic and N-glycoproteomic analysis identifies known and novel candidate biomarkers, unique PTM signatures, and therapeutic vulnerabilities implicating disease-specific pathogenetic mechanisms in CTCL.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution