TI - M4A8 ubiquitinates BCL10 via its RING domain . AB - The RING domain of API2 enables it to function as an E3 ubiquitin ligase , capable of auto-ubiquitination [9] , [39] and ubiquitination of interacting proteins [10] , [40] -[42] , like BCL10 via its interaction with the first BIR domain of API2 [29] . We thus wondered whether also MALT1-API2 possessed E3 ubiquitin ligase activity . The different bio-tagged MALT1-API2 fusion variants were therefore transiently expressed in HEK-293T cells together with HA-tagged ubiquitin ( Ub ) and subsequently purified via streptavidin-mediated pull-down . Western blot analysis with alpha-Flag ( for API2 and MALT-API2 ) and alpha-HA ( for HA-Ub ) clearly showed that all MALT1-API2 fusion variants were ubiquitinated , similarly as API2 (Fig 3A) . Furthermore , mutation of the critical His and the two flanking Cys residues of the RING domain ( RING mutant ; RM ) abolished most ubiquitination ( Fig 3A ) , indicating auto-ubiquitination of MALT1-API2 . Interestingly , cellular BCL10 co-purified with bio-tagged M4A8 , and to a much lesser extent with M7A8 , showed a similar pattern of ubiquitination , which again was completely dependent on a functional RING domain in MALT1-API2 (Fig 3A) . Remarkably , we were not able to co-precipitate BCL10 together with bio-tagged full-length API2 (Fig 3A) , in contrast to a recent report on the BIR -mediated interaction of BCL10 with API2 [29] . To exclude that the slower migrating BCL10 bands observed in Fig 3A resulted from BCL10 PHOSphorylation , protein complexes co-purified with bio-tagged wild-type ( WT ) M4A8 were treated with lambda-PPase , but this did not affect the migration pattern of BCL10 (Fig 3B) . To further investigate these BCL10 modifications by M4A8 , we transiently expressed HA-bio-tagged BCL10 in HEK-293T cells , alone or in combination with Flag-tagged WT - or RM-M4A8 . While no effect of the RING mutant M4A8 on BCL10 was seen in the cell lysate , co-expression of BCL10 with wild-type M4A8 again induced BCL10 modifications ( indicated with asterisks , alpha-BCL10 and alpha-HA panel Fig 3C ) . When the cells were treated with the proteasomal inhibitor MG132 prior to harvesting , the BCL10 modifications became more pronounced , further suggesting that M4A8 induces BCL10 ubiquitination (Fig 3C) . This was even more obvious when BCL10 was purified via streptavidin pull-down in denaturing conditions to destroy existing protein interactions ( alpha-BCL10 panel Fig 3D , left part ) . Again , PHOSphorylation could be excluded , since treatment of the purified BCL10 proteins with lambda-PPase did not affect the slower migrating BCL10 bands ( Fig 3D , right part ) . In contrast , these bands could be revealed with an alpha-Ub antibody (Fig 3D) . Altogether , these data indicate that MALT1-API2 can function as an E3 ubiquitin ligase via the RING domain of API2 . While all fusion variants auto-ubiquitinate , only M4A8 is able to induce significant BCL10 ubiquitination in overexpression experiments .