Ubiquitin-SUMO tag-team wrestles acute promyelocytic leukemia

The reciprocal chromosomal translocation t(15;17) fuses the promyelocytic leukemia (PML) gene to the retinoic acid receptor alpha (RARA) gene. The resulting PML-RARA oncogene is the causative agent that drives acute promyelocytic leukemia (APL) by altering the RARA transcriptional program, blocking the differentiation of promyelocytes (1, 2). APL patients had a dismal prognosis prior to the introduction of differentiation therapy, consisting of arsenic trioxide (arsenic) combined with all trans retinoic acid (3, 4). These agents induce degradation of the PML-RARA fusion protein, restoring RARA-driven differentiation of promyelocytes, curing the patients. A small subset of patients relapses due to arsenic resistance, frequently caused by missense mutations in PML.

Interestingly, different posttranslational modifications are involved in arsenic-induced degradation of PML-RARA and PML: ubiquitin and small ubiquitin-like modifiers (SUMOs). These modifiers team up to form complex polymers that lead to extraction of PML by the segregase p97/VCP and its subsequent degradation by the proteasome (5). The mixed polymers are built by SUMO-targeted ubiquitin ligases, known as STUbLs. The STUbL RNF4 recognizes poly-SUMO1 and poly-SUMO2/3 and plays a key role in arsenic-induced PML degradation (6, 7, 8) (Fig. 1). In the current study from the Hay lab, a second STUbL TOPORS is uncovered that plays an equally important role in arsenic-induced PML degradation (9). TOPORS has a remarkable preference for poly-SUMO1 over poly-SUMO2/3, indicating that not only SUMO2/3 but also SUMO1 is required for efficient degradation of PML (Fig. 1). The study builds on and extends previous findings from the teams of Profs. Mailand and Hay, who first uncovered TOPORS as a STUbL cooperating with RNF4 and required for repair of DNA-protein cross-links to maintain genome stability (10).

Arsenic-resistant missense mutations in PML map in the region between L211 and S220. Here, Prof. Hay and his team have explored the mechanism underlying arsenic resistance of two of these PML mutants A216T and L217F (9). Mechanistically, the A216T mutant of PML was poorly SUMO modified, thereby failing to recruit both STUbLs TOPORS and RNF4, blocking ubiquitination, and subsequent degradation by the proteasome. In contrast, the L217F mutant of PML was efficiently conjugated to SUMO2/3, allowing ubiquitination by the STUbL RNF4 in response to arsenic. Subsequent investigation showed that SUMO1 modification of L217F was poor, and as a consequence, TOPORS was poorly recruited, limiting overall ubiquitination. This revealed the remarkable preference of TOPORS for SUMO1. Clearly, both STUbLs were required to reach a threshold of ubiquitination needed to recruit the p97/VCP segregase, essential to translocate PML from nuclear bodies. Overall, this reveals a remarkable cross talk between the posttranslational modifiers SUMO1, SUMO2/3, and ubiquitin.

Whereas this is a major step forward to boost our understanding of arsenic-induced signaling intricacies, more work is required for full understanding of the process. It is not fully clear how the STUbLs TOPORS and RNF4 collaborate in PML degradation. Results suggest that RNF4 and TOPORS act in a nonredundant manner. What are the individual contributions of these STUbLs? Do they act sequentially? Do they activate each other? Can we completely rule out any roles for the STUbL RNF111, also known as Arkadia (11)?

Furthermore, the detailed architecture of the physiological SUMO polymers and ubiquitin polymers is unclear. In vitro, TOPORS efficiently ubiquitinates SUMO1 tetramers (9), but SUMO1 is not expected to polymerize efficiently in cells. Do TOPORS and RNF4 recognize hybrid SUMO1/SUMO2/3 chains? How are these hybrid SUMO chains linked? Are these hybrid SUMO polymers primarily linked via K11 of SUMO2/3, situated in a strong SUMOylation consensus motif? Or are other SUMO lysines involved as well? The study shows that eight different SUMO–SUMO linkages can be found, revealing major complexity of the hybrid chains. Is the formation of longer SUMO2/3 polymers capped by SUMO1 possible in cells (11)? What is the length and architecture of these polymers? Do TOPORS and RNF4 build complementary types of ubiquitin polymers that are both required for activation of the segregase p97/VCP? Answering these questions in a comprehensive manner will require detailed follow-up studies.

So far, the main disease associated with TOPORS mutations is retinitis pigmentosas, a group of rare eye diseases that affect the retina (12). The consequences of mutations of these STUbLs for PML patients likewise require attention. Would mutations in the STUbLs TOPORS and RNF4 or other ubiquitin E3 ligases lead to arsenic resistance in APL patients? To answer this question, a large, genome-wide study of APL patients needs to be conducted.