The Saccharomyces cerevisiae genome is interspersed with binding sites for relatively abundant DNA-binding proteins collectively known as general regulatory factors (GRF) because of their involvement as global regulators in diverse chromosomal functions. GRFs include Rap1, Reb1, Tbf1 and Abf1 proteins, all containing Myb-related motifs capable of sequence-specific DNA binding (Azad and Tomar 2016; Brigati et al. 1993; Chasman et al. 1990; Ju et al. 1990; Shore 1994). GRF-binding sites are scattered throughout the genome, with a higher frequency within promoter regions, where the presence of bound GRFs influences promoter activity, but also within telomeric and subtelomeric regions, where GRFs contribute to telomere structure and regulation (Fourel et al. 1999; Grunstein 1997), and at replication origins whose function can be influenced by bound GRFs (Raychaudhuri et al. 1997). It is thought that a common role shared by these DNA-binding proteins at different chromosome locations might be the establishment and maintenance of well-defined chromatin structures (Ganapathi et al. 2011; Hartley and Madhani 2009).

At promoter regions, GRFs may additionally act by recruiting the basal transcription machinery (Papai et al. 2010). Interestingly, promoter regions marked by GRFs generally belong to heavily transcribed genes, many of which are involved in growth-related functions such as ribosome biogenesis (Bosio et al. 2011). Indeed, it has long been known that in S. cerevisiae most of the 138 ribosomal protein (RP) genes contain Rap1-binding sites in their promoter regions, with the exception of a small subset of RP genes for which Rap1 is replaced by Abf1 (Lascaris et al. 1999). These two GRFs appear to be interchangeable in this case, as they both act as constitutive transcriptional activators (Mager and Planta 1990).

Given the high number of binding sites located across the genome and the ability of GRFs to influence nucleosome occupancy/chromatin structure, GRFs have the potential to act as master regulators of genome expression in response to nutritional cues or other environmental perturbations. Suggestively with respect to this possibility, the four GRFs mentioned above have all been reported to be phosphoproteins (Albuquerque et al. 2008; Francesconi and Eisenberg 1991; Morrow et al. 1990). In particular, the balance between Abf1 hyper- and hypo-phosphorylated forms was reported to vary in response to nutrient availability (Silve et al. 1992), and Abf1 was included among candidate proximal targets of TORC1 by a recent phosphoproteomic study (Oliveira et al. 2015). However, until recently no evidence was reported for regulated changes in GRF association with cognate promoter elements, nor for any other functional alteration in their activity at promoters, in response to environmental stimuli. This consolidated the notion of GRFs as permanent “placeholders” constitutively acting at promoter regions by counteracting nucleosome deposition and, thus, favoring regulatory transactions due to other, more tunable transcription factors (Bhattacharya and Warner 2008).

Recently, a study by our laboratory challenged this notion by showing that the association of Abf1 with at least some of its target promoters is responsive to nutritional status (Fermi et al. 2016). In particular, we found that inhibition of the nitrogen-sensitive TORC1 pathway by rapamycin, a perturbation known to cause quick repression of ribosomal protein (RP) gene transcription, unexpectedly entails a large increase in Abf1 association with the small cohort of RP genes whose promoter is demarcated by Abf1 instead of Rap1 binding sites. Even though the dependence of this phenomenon on changes of Abf1 phosphorylation state is unknown at present, this observation points to Abf1 as the first GRF whose association to DNA is modulated by nutritional status, and, at the same time, raises at least three orders of questions: (a) why should a repressive stimulus (TORC1 inactivation) entail an increased recruitment of an activator to transcriptionally repressed promoters? (b) what is the physiological reason (if any) for the existence in S. cerevisiae of two distinct subsets of RP gene promoters, one characterized by nutrient-unaffected Rap1 binding, the other by nutrient-dependent fluctuations in Abf1 binding? (c) considering other genes bound by Abf1 in their promoter regions, how does Abf1 behave at these promoters in response to nutritional stress, and which are the possible regulatory interconnections between the different categories of Abf1-demarcated genes?

As to the first question, we provided evidence that increased Abf1 association with RP gene promoters observed in response to TOR pathway inactivation might help the transcriptional rescue of these repressed promoters, once more favorable nutritional conditions are established (Fermi et al. 2016). But quick changes in promoter occupancy by Abf1, which were also observed, albeit to lower extents, at ribosome-unrelated promoters in response to TORC1 inhibition (Fermi et al. 2016), might also be related to extra-transcriptional roles of Abf1 (Reed et al. 1999; Yu et al. 2009).

The second and third questions are related to each other, because the different behaviors of Abf1 at its target promoters [which have already been noted in association with different chromatin organization propensities (Paul et al. 2015)] might deal with still unexplored levels of co-regulation between different groups of Abf1-demarcated genes. An intriguing possibility is that the Abf1-dependent subset of RP genes shares some specific regulatory features with the genes required for ribosome biogenesis belonging to the so-called Ribi regulon. According to such hypothesis, the promoters of these genes are enriched in Abf1-binding sites (Yarragudi et al. 2007; Bosio et al. 2016), and their association with Abf1 increases in response to TORC1 inactivation exactly as it does in the case of Abf1-dependent RP genes (Fermi et al. 2016).

In a systematic search for other S. cerevisiae gene regulons whose promoter regions tend to be enriched in Abf1 binding sites, we noticed that such cis-acting elements appear to be overrepresented also in the promoter regions of the nuclear genes coding for mitochondrial ribosomal proteins (MRPs). As shown in Table 1, among the 74 S. cerevisiae genes coding for MRPs (according to the Saccharomyces Genome Database), those displaying in the promoter region one or more Abf1-binding sites are 24, 13 of which have been shown to be bound by Abf1 in vivo and/or to be transcriptionally dependent on Abf1 (Chang et al. 2011; Schlecht et al. 2008; Yarragudi et al. 2007). Interestingly, 15 more mitochondrial ribosomal protein (MRP) gene promoters contained one or more Reb1-binding sites, and 3 more MRP promoters displayed one or more Rap1-binding sites (Chang et al. 2011). Therefore, the majority of MRP promoters are demarcated by GRFs, in particular by Abf1. In light of these observations, it is tempting to speculate that Abf1 might orchestrate a network of subtle regulatory interconnections between genes involved in the biogenesis and structure of both cytoplasmic and mitochondrial ribosomes. With this respect, it is worth noting that Abf1 has been suggested by a previous study to be involved in the intergenomic signaling pathway of mitochondrial–nuclear communication (Woo et al. 2009).

Table 1 Mitochondrial ribosomal protein (MRP) genes displaying one or more Abf1-binding sites in their promoter region according to YPA
Full size table

Given that there are several hundreds of Abf1 target promoters in the budding yeast genome, with many of them regulating growth-related genes, this transcription factor might represent one of the key regulators of genome expression reprogramming which allows for rapidly alternating growth arrest and growth burst phases in response to changing nutrient availability (Ho and Gasch 2015; Soontorngun 2016).