Reply to Willis et al.: Casein kinase II phosphorylation of Maf1 triggers RNA polymerase III activation

Abstract

We recently reported (1) that casein kinase II (CK2)-mediated phosphorylation of Maf1 stimulates polymerase III (Pol III) transcription upon transfer of yeast from repressive conditions in glycerol medium to medium with glucose. Willis et al. (2) have taken issue with our conclusion that Maf1 is a target for CK2 and brought to our attention the possibility that activation of Pol III in our experimental conditions may be mediated by PKA or Sch9 kinases independent of any involvement of CK2.In response to Willis's objection, we note the following: first, the CK2 phosphoacceptor sites in Maf1 were determined by mass spectrometry, mutated to alanine, and shown in vivo to ablate the mobility shift caused by phosphorylation. The resultant mutant remains fully functional in vivo, as shown by genetic complementation after Maf1 deletion. This functionality seems incompatible with the structural disruption of Maf1 suggested by Willis et al. (2) as leading to failure to exhibit a mobility shift after phosphorylation elsewhere. Second, we reported that endogenous Maf1 coimmunoprecipitates with CK2, establishing a physical association (Fig. 2 in ref. 1 ), which was ignored by Willis et al. (2).Maf1 mediates several signaling pathways; PKA and Sch9 were previously identified (3, 4) but are not the only Maf1 kinases. Maf1 mutated at the potential phosphorylation sites for Sch9 and PKA is still partially functional in transmitting signals to Pol III (4). We believe that Maf1 is inactivated by the sequence of phosphorylation events and claim that CK2 operates at the very beginning, because of its association with Pol IIItranscribed genes (1). Moreover, in contrast to CK2, PKA is not involved in physiological Maf1-mediated Pol III regulation by carbon source (5), and its role in Maf1-mediated Pol III repression by nutritional stress is controversial (4).The second objection they raise is that the responses we see after CK2 inactivation in vivo might be indirect. They observe quite correctly that CK2 has many targets and that secondary effects that are indirect can be expected, including loss of viability. We are aware of this and accordingly assayed changes after only 1 h of treatment. Our data show that Maf1 phosphorylation and activity are altered within 1 h of CK2 inactivation by specific chemical inhibitor or temperature shift of a ts allele (with appropriate controls). Secondary effects are minimized by the brevity of the treatment, and viability is unlikely to be compromised so rapidly. Furthermore, our evidence for physical association clearly supports our contention that Maf1 responds directly to CK2.We believe that the sum of our data builds a strong case for CK2 being a physiological regulator of Maf1. We are confident that future studies will substantiate this discovery.