Differential Phosphoproteomics Provides New Insights into Role of DYRK1A in Lymphopoiesis

Dual Specificity Tyrosine-Regulated Kinase 1A (DYRK1A) is a serine/threonine (S/T) kinase that is encoded within the Down syndrome Critical Region of chromosome 21(Abbassi, Johns et al. 2015), underlying its potential role in neurologic disorders associated with Down syndrome (DS). Given that children with DS are at a significantly increased risk for developing leukemia, we are studying the role of DYRK1A in normal and malignant hematopoiesis. We previously reported a critical role for DYRK1A in lymphopoiesis through its phosphorylation of Cyclin D3, thereby mediating cell cycle exit and lymphocyte maturation(Thompson, Bhansali et al. 2015). In the absence of Dyrk1a, both B and T cell development is stalled at the transition from highly proliferative progenitors to maturing quiescent cells. Although Dyrk1a -deficient lymphocytes fail to exit the cell cycle, they also cease proliferating.

In order to further delineate the role of DYRK1A in lymphopoiesis, we employed a phosphoproteomics-based approach to identify specific pathways affected by its activity. In brief, we treated primary murine pre-B cells with a novel DYRK1A inhibitor or DMSO for 2 and 9 hours. Treated cells were then collected, lysed and isolated proteins were digested with trypsin. The resultant peptides were phospho-enriched using TiO₂in order to yield S/T phospho-peptides, which were then labeled with tandem mass tags (TMT). These labeled peptides were analyzed using mass spectroscopy. We included both unfractionated and fractionated lysates in our analysis in order to account for low abundance peptides. A cutoff of 1.5-fold change in phosphorylation with a false discovery rate of ≤1% upon inhibition of DYRK1A was implemented to devise a list of phospho-peptides that may be affected by DYRK1A activity.

At 2 hours of DYRK1A inhibitor treatment, 101 proteins were significantly dephosphorylated and 37 proteins were significantly hyperphosphorylated. By contrast, at 9 hours of DYRK1A inhibitor treatment, 41 proteins were significantly dephosphorylated, while 784 proteins were significantly hyperphosphorylated, indicating that global phosphorylation trends downstream of DYRK1A are a time-dependent process. We also analyzed the GO Biological Process enrichment of the 2 hour timepoint and observed clear differences in the pathways associated within the dephosphorylated and hyperphosphorylated groups. The dephosphorylated proteins were enriched in processes such as cell cycle, mitosis, JAK-STAT signaling and metabolism, while the hyperphosphorylated proteins were enriched in chromatin assembly/disassembly and DNA packaging. Next, using the STRING database(Szklarczyk, Morris et al. 2017), we found that the dephosphorylated set of proteins map into 5 clusters based on their roles in cell biology. For example, Cyclin D3 clustered with Geminin, SKA3 and PLK1, which are involved in the regulation of mitosis. Another potential DYRK1A target is STAT3, which clustered with BCR and TdT, both of which are known to play a role in leukemia. In the hyperphosphorylated set of proteins, the only major cluster included several histone subunits. Notably, DYRK1A has been implicated in maintaining open chromatin for transcription of cytokine genes in DS-leukemia through phosphorylation of H3 at T45 and S57(Jang, Azebi et al. 2014). Upon DYRK1A inhibition, we observed hyperphosphorylation of alternate S/T residues of H1, H3 and H4 suggesting that chromatin regulation by DYRK1A may play a role in lymphopoiesis as well.

We selected several proteins from the dephosphorylated cohort as potential DYRK1A substrates based on criteria including documented interactions, the presence of the DYRK1A consensus sequence and roles in cell biology. We generated phospho-deficient alleles of these putative targets and performed in vitro kinase assays to confirm they are direct substrates. This study includes analysis of several potential DYRK1A B-cell substrates such as STAT3, geminin, BCL7C, SKA3 and several members of the RNA Polymerase II complex, including MED15 and GRINL1A. Data on their contributions to B cell development will be presented. Our study provides a global view of the contribution of DYRK1A to phosphorylation events related to lymphopoiesis. The identification of novel substrates and their associated pathways may explain the paradoxical phenotype we observe in lymphocytes upon loss of Dyrk1a.