Triad1 Regulates Myelopoiesis through Different Ubiquitin Ligase Activities.

The modification of cellular proteins with poly-ubiquitin chains plays an essential role in hematopoiesis. Different types of ubiquitin chains may have opposite effects on the marked proteins. Chains linked through lysine 48 of ubiquitin are recognized by the proteasome resulting in progressive degradation of the ubiquitylated proteins. Ubiquitin chains linked through lysine 63 are not recognized by the proteasome. Instead, these chains can bind proteins that regulate signal transduction and gene transcription. Although it is known that ubiquitylation is essential for hematopoiesis and that alterations in ubiquitylation have been implicated in malignant hematopoiesis the ubiquitin ligases that catalyze protein ubiquitylation remain largely unknown. Triad1 is an ubiquitin ligase that inhibits the proliferation of myeloid progenitor cells through its ligase activity. Triad1 belongs to a unique class of ligases that harbor two RING finger protein domains. This domain specifically binds ubiquitin conjugating enzymes (Ubcs). Together with the Ubcs the ligases determine which type of ubiquitin chain is catalyzed. To understand how Triad1 regulates myelopoiesis we screened a panel of Ubcs for Triad1 interaction and found that Triad1 binds UbcH7 through its N-terminal RING domain and Ubc13 through its C-terminal RING domain. UbcH7 catalyzes the formation of ubiquitin chains linked through lysine 48 that are recognized by the proteasome. Importantly, ubiquitin chains catalyzed by Ubc13 are linked through lysine 63 and are not recognized by the proteasome. In agreement with these interactions, in vitro ubiquitylation assays using different ubiquitin mutants containing only one lysine residue showed that Triad1 can catalyze the formation of both types of ubiquitin chains. The relevance of these findings in myelopoiesis was studied by generating a panel of Triad1 deletion mutants that lack the coiled coil, DRIL or RING domains. Next, the growth inhibitory effect of these mutants was tested in clonogenic assays by retroviral transduction of U937 cells. Like in primary cells, wild type Triad1 inhibited U937 colony formation by over 60% compared to empty vector transduced cells. Three Triad1 mutants lacking the DRIL domain or either one or two of the coiled coil domains inhibited clonogenic growth at a comparable rate as wild type Triad1. In contrast, deletion of either the N- or C-terminal RING finger completely abrogated the inhibitory effect of Triad1 in clonogenic growth. Thus, loss of either the UbcH7 or Ubc13 binding domain of Triad1 affects its inhibitory function in myeloid cell proliferation. Recently, we reported that Triad1 binds the transcription factor Gfi1. Gfi1 plays an important role during many hematopoietic developmental stages and is essential for neutrophilic differentiation. Remarkably, Triad1 inhibited Gfi1 ubiquitylation and proteasomal degradation. The finding here that Triad1 can catalyze the formation of different ubiquitin chains might suggest that Triad1 may modify Gfi1 with ubiquitin in a functional way, rather than marking it for proteasomal degradation. Indeed, in vivo ubiquitylation experiments showed that Gfi1 can be modified with ubiquitin chains not linked through lysine 48. The relevance of this finding and the role of Triad1 in this process is currently studied. Together, these data indicate that the dual ubiquitin ligase activity that results in the formation of different poly-ubiquitin chains is crucial to the central role of Triad1 in myelopoiesis.