The ubiquitin proteasome system (UPS) is a major degradation machinery for clearance of unwanted and misfolded proteins in eukaryotes, which is vital for maintaining cellular homeostasis (Goldberg, 2005, Livneh et al., 2016). Proteasome inhibitors have been shown to block the development of malaria parasites at multiple stages, indicating a critical role of the UPS during parasite development (Gantt et al., 1998, Lindenthal et al., 2005, Czesny et al., 2009). Hence, the proteasome of protozoan parasites, including Plasmodium, has emerged as a potential drug target (Khare et al., 2016, Li et al., 2016, Ng et al., 2017, Wyllie et al., 2019). The UPS includes ubiquitin, enzymes involved in tagging the substrate with ubiquitin and the 26S proteasome wherein ubiquitinated proteins are degraded (Hershko and Ciechanover, 1982, Ciechanover, 2009). Several non-proteasomal ubiquitin-binding proteins also contribute to proteasome function. One such class of proteins are the shuttle proteins, which include Rad23, Dsk2 and DDI1 (Finley, 2009, Saeki, 2017). These proteins contain a ubiquitin-like (UBL) domain and a ubiquitin-associated (UBA) domain, which mediate their interactions with the proteasome and ubiquitin chains of the ubiquitinated protein, respectively. DDI1 also contains an aspartyl protease domain, similar to the retroviral protease (RVP) of HIV (Sirkis et al., 2006).

DDI1 was first discovered as one of the over-expressed proteins upon treatment of Saccharomyces cerevisiae with DNA-damaging agents, hence, it was named as the DNA damage inducible 1 protein (DDI1) (Liu and Xiao, 1997). A later study named it v-SNARE master 1 (VSM1) based on its interaction with Snc2 v-SNARE proteins, and showed that knockout of VSM1 increased protein secretion (Lustgarten and Gerst, 1999). VSM1/DDI1 inhibits assembly of Sso t-SNAREs with Sec9 t-SNARE, which is required for exocytosis (Marash and Gerst, 2003, Gabriely et al., 2008). Deletion analysis demonstrated that both UBL and RVP domains of S. cerevisiae DDI1 (ScDDI1) are necessary for inhibition of protein secretion. However, the ScDDI1 C-terminal that contains the Sso t-SNARE-binding region and UBA domain was found to be dispensable for suppression of protein secretion (White et al., 2011a), which indicated that suppression of protein secretion by VSM1/DDI1 is independent of its interaction with Sso t-SNARE. Over-expression of ScDDI1 rescued the S-phase checkpoint defect of a temperature-sensitive Pds1 mutant, which required the UBA but not UBL domain (Clarke et al., 2001). Another study showed that UBL, catalytically competent RVP and UBA domains of ScDDI1, are required for the rescue of S-phase checkpoint defect (Gabriely et al., 2008). ScDDI1 has been shown to shuttle ubiquitinated Ho endonuclease to the proteasome for degradation, which requires interaction of UBA and UBL domains with ubiquitinated Ho endonuclease and the Rpn1 subunit of 26S proteasome, respectively (Kaplun et al., 2005, Voloshin et al., 2012). Interaction of the ScDDI1 UBL domain with the ubiquitin interaction motif (UIM) of the F-box protein Ufo1 has been shown to be critical for proteasomal degradation of Ufo1 (Ivantsiv et al., 2006). The UBA domain of ScDDI1 homolog in Schizosaccharomyces pombe, Mud1, has also been shown to bind Lys48-linked polyubiquitin chains (Trempe et al., 2005). Interestingly, DDI1 itself is a substrate of E3 ligase UBE3A in Drosophila melanogaster, which does not target ubiquitinated DDI1 for degradation, and the biological significance of ubiquitination of DDI1 is not known yet (Ramirez et al., 2018). Recent studies have demonstrated important roles of ScDDI1 in removing the replication termination factor RTF2 from stalled replication forks and DNA-topoisomerase complexes (Svoboda et al., 2019, Serbyn et al., 2020), which require catalytically competent ScDDI1, indicating that ScDDI1 functions as a protease to remove DNA-protein crosslinks.

The DDI1 homolog of Drosophila melanogaster, Rngo, is essential for oocyte formation (Morawe et al.). As has been shown for ScDDI1, Rngo also dimerizes through the RVP domain, binds ubiquitin via the UBA domain and interacts with the Rpn10 subunit of the proteasome via the UBL domain. The Homo sapiens DDI2 (HsDDI2) and Caenorhabditis elegans DDI1 (CeDDI1) have been shown to cleave the endoplasmic reticulum-associated Nrf1/Skn-1A, which results in the release and translocation of the cleaved form to the nucleus wherein it upregulates the proteasome genes (Koizumi et al., 2016, Lehrbach and Ruvkun, 2016). Despite multiple studies indicating the requirement of catalytically competent DDI1, a direct demonstration of the protease activity of DDI1 proteins has been elusive until recently. Two recent studies directly demonstrated that HsDDI2 and ScDDI1 cleave ubiquitinated Nrf1 and a peptide substrate corresponding to the Nrf1 cleavage site, respectively (Dirac-Svejstrup et al., 2020, Yip et al., 2020). These two studies also indicated that ScDDI1 and HsDDI2 do not cleave ubiquitin chains, and hence are unlikely to function as deubiquitinases.

As described above, multiple studies highlight contributions of UBL, RVP and UBA domains in functions of DDI1 proteins. How these domains crosstalk with each other in the full-length DDI1 is not known due to the unavailability of a full-length DDI1 structure. Structural studies of the RVP domains of ScDDI1, Leishmania major (LmDDI1) and HsDDI2 revealed that the RVP domain exists as a dimer in which aspartyl protease motifs from both the monomers contribute to the formation of an active site (Sirkis et al., 2006, Sivá et al., 2016, Trempe et al., 2016, Kumar and Suguna, 2018). The active site of ScDDI1 has identical geometry with that of HIV protease, but has a wider substrate binding groove than HIV protease, suggesting that ScDDI1 can accommodate bulkier substrates (Sirkis et al., 2006). Interestingly, in addition to the UBA domain, the UBL domains of ScDDI1 and HsDDI2 also contain a ubiquitin-interaction motif (UIM) that mediates interaction with ubiquitin (Nowicka et al., 2015, Sivá et al., 2016, Trempe et al., 2016).

As a therapeutic target, not much is explored on the role of DDI1 in many economically important protozoan parasites, despite a significant adverse impact of these on human and domestic animal health. The DDI1 proteins of Leishmania major and Toxoplasma gondii are the only reported representatives. The LmDDI1 has been proposed to be the major target of the anti-leishmanial effect of HIV protease inhibitors (White et al., 2011b). Recombinant LmDDI1 has also been shown to degrade peptides and BSA at pH 4-5, which appears to be contrary to the presence of DDI1 proteins in a cytosolic pH environment (Perteguer et al., 2013). Knock-out of Toxoplasma gondii DDI1 (TgDDI1) has been shown to cause accumulation of ubiquitinated proteins and a loss of virulence (Zhang et al., 2020). HIV protease inhibitors have been shown to block the development of malaria parasites at multiple stages (Skinner-Adams et al., 2004, Parikh et al., 2005, Hobbs et al., 2009, Nsanzabana and Rosenthal, 2011, Hobbs et al., 2013). Furthermore, a study in humans showed that treatment with HIV protease inhibitor-based antiretroviral therapy reduced the incidence of malaria by 41% compared with the group treated with non-protease inhibitor-based anti-retroviral therapy, and the lower incidence was attributable to decreased recurrence of malaria (Achan et al., 2012). As DDI1 is important for several cellular processes and the presence of a RVP domain makes it a potential target of HIV protease inhibitors, we undertook cellular, genetic and biochemical approaches to investigate Plasmodium DDI1. Our study revealed that Plasmodium DDI1 is critical for parasite development, associated with chromatin and contributes to DNA-protein crosslink (DPC) repair, and it could be a target of HIV protease inhibitors.