Translation is subject to multiple layers of regulation

Translation offers many levers for regulation. The pool of ribosomes that can enter translation is determined by the production, processing and modification of rRNA and ribosomal proteins and their assembly in the nucleus. The actual synthesis of a translation- competent ribosome, followed by nuclear export and maturation in the cytoplasm, are long and complex processes and sensitive to many influences (Karbstein, 2011; Nierhaus, 1991; Wilson and Nierhaus, 2007). Their regulation controls the cell’s basic disposition to perform translation. Similarly crucial to the synthesis of ribosomes is the synthesis of factors that are involved in translation.

The actual accessibility of both ribosomes and factors is the next layer of control. For example, 80S monosomes were found to bind eEF2, but not to be translationally active (Khatter et al., 2015; Voorhees et al., 2014); the 40S subunit in these complexes is thus not accessible for translation initiation. Stability of an 80S•eEF2 complex has also been observed in vitro (Budkevich et al., 2014; Davydova et al., 1993). An example for the control of protein factors is the eukaryotic initiation factor eIF4E, which can be bound and thus inactivated by eIF4E-binding protein (4E-BP), whose affinity is regulated by post- translational phosphorylation; initiation is reduced then by the decreased availability of eIF4E (Bah et al., 2015; Marcotrigiano et al., 1999). Also eEF2 is known to be inactivated by phosphorylation (Ling and Ermolenko, 2016).

Besides such quite simple mechanisms that are based on the accessibility of substrates, there might exist a more complex machinery that is aimed to enhance the production of specific proteins and down-regulate others. Whereas in long-term, the proteome of a cell is controlled by transcription, dynamic, rapid adaption mechanisms could base on differential translation. How this works exactly is not clear; discussed is among others the existence of ‘specialized ribosomes’ (Xue and Barna, 2012). An indicator for a possible regulation mechanism is that ribosomal proteins can be modified post-translationally. The ribosomal protein with the longest history of investigations of its posttranslational modification is ribosomal protein eS6, which is phosphorylated and whose putative impact on translation is one of the questions raised in this thesis (see below). Also, the topology of translation in the cell and the composition and architecture of polyribosomes (see below) might be a dynamic mechanism to alter the translatome.

Actively translating ribosomes are organized into polysomes

It is the normal case that on one mRNA, multiple ribosomes are translating the message at the same time and that depending on the length of the mRNA, multiple initiation events happen before the first ribosome reaches the stop codon.

Such assemblies of multiple ribosomes on the same mRNA are called polyribosomes or polysomes. Each ribosome that is part of a polysome group is at some stage of translation (Figure 13). Elongation intermediates are the by far most often sampled states, because the elongation cycle must take place for each codon between start and stop in contrast to initation, which takes place only at the start codon, and termination, which takes place only at the stop codon.

By isolating polysomes ex vivo, Behrmann et al. were able to reconstruct ten different states of mammalian elongation (Behrmann et al., 2015). The differential population of the states reflects the distribution of functional states in polysomes. Although the time between purification of the polysomes used for analysis and the freezing of the sample might allow for changes, that work shows the in vivo energy landscape of translation as close to the native state as possible. Of special interest for studying the regulation of translation is the distribution of functional states (Figure 13). The captured states by far don’t represent the complete spectrum of possible conformational and functional states of the mammalian ribosome during elongation, but only local minima. Referring to the nomenclature of the ‘dynamics‘ section, the states represent fluctuations in the macro- to mesoscopic range (Munro et al., 2009).

The shapes of polysomal assemblies have been extensively studied (Afonina et al., 2013, 2014; Brandt et al., 2010; Myasnikov et al., 2014; Viero et al., 2015). The relative orientations of the ribosomes in a polysomal assembly raise the question of how ribosomal proteins or RNA on the solvent side might promote these orientations via specific inter-ribosomal contacts. Moreover, it was observed that the abundance of rectangular shapes is associated with a low translation activity (Myasnikov et al., 2014; Viero et al., 2015) and that it can be increased by treatment of cells with rapamycin or by serum- withdrawal (Viero et al., 2015). Change in the ratio of ribosomes to mRNA can also modify the pattern of proteins synthesized (Lodish, 1974). Polysome patterns after post mitochondrial supernatant (Duncan and McConkey, 1982) as well as investigations of the shapes of polysomes (Myasnikov et al., 2014; Viero et al., 2015), suggest that polysomes react to different conditions, like serum withdrawal, by changing their shape and/or composition.

In eukaryotic cells, polysomes can be additionally differentiated by the cellular compartment they are attached to. Essentially, there are cytosolic polysomes and endoplasmatic reticulum (ER)-bound polysomes. So far, it is agreed upon that every ribosome has the potential to join either class and that it is the signaling peptide encoded in the mRNA that is responsible for recruiting the ribosome to the ER membrane (Meister, 2011).

Phosphorylation of ribosomal protein eS6: a possible switch for translation regulation?

Among the many discussed possibilities of how translation might be regulated, there are post-translational modifications of ribosomal proteins (Xue and Barna, 2012). Some ribosomal proteins undergo phosphorylation, e.g. ribosomal protein eS6 (eS6), a eukaryote- specific protein of the 40S subunit located at the foot of the 40S body/platform (Figure 3A). eS6 is 249 residues long in humans and consists of a globular part and an alpha helical part with a flexible C-terminus (Figure 14).

eS6 is a necessary protein for ribosome biogenesis. Its conditional deletion has been shown to lead to a p53-dependent inhibition of cell cycle progression (Panić et al., 2006; Sulić et al., 2005; Volarevic, 2000) in multiple tissues.

The alpha helical, C-terminal part of eS6 reaches into the vicinity of expansion segment 6 of the 18S rRNA and displays five serine residues (S235, S236, S240, S244, S247) at its C-terminus, which are the substrates of four independent kinases (RSK, Protein kinase A, S6K and CK1) with partially overlapping specificities (Bandi et al., 1993; Krieg et al., 1988; Wettenhall and Morgan, 1984) (Figure 15).

Phosphorylation happens in an ordered fashion, such that first, the residues serine 235 and serine 236 are phorsphorylated, then serine 240 and serine 244 and last serine 247 (Martin-Pérez and Thomas, 1983; Wettenhall et al., 1992). Phosphorylation of serine 247 requires prior phosphorylation of serines 240/244, and phosphorylation of serine 247 in turn promotes/stabilizes phosphorylation of serines 240/244 (Hutchinson et al., 2011). The modification is counteracted by the action of phosphatase (PP) 1 which is responsible for dephosphorylation at all five sites (Barth-Baus et al., 2002; Belandia et al., 1994; Hutchinson et al., 2011; Li et al., 2012).

The phosphorylation of the five serine residues at the C-terminus is massively increased in regenerating liver cells by partial hepatectomy, which was the first model for studying eS6-phosphorylation (Gressner and Wool, 1974). In the context of growth, two pathways lead to eS6 phosphorylation: The PI3K/Akt/TCS/Rheb/mTORC1/ S6K pathway and the Ras/Raf/MEK/ERK/RSK pathway (Figure 16) (reviewed in (Meyuhas, 2008, 2015)). A large number of stimuli have been shown to induce eS6 phosphorylation via these pathways, including stimulation of serum-starved cells with serum (Meyuhas, 2008, 2015; Roux et al., 2007).

As to the consequences of this modification, many observations have been made that are thought to be associated with eS6 phosphorylation, but none of them can be explained mechanistically. In 2005, Ruvinsky and colleagues found that global protein synthesis is higher in knockin mouse embryonic stem cells (MEFs) (P-/-) compared to wild type MEFs, but these observations could not be reproduced by using S6K-deficient mice (S6K1 -/-, S6K2 -/-) (Chauvin et al., 2014; Mieulet et al., 2007; Ruvinsky et al., 2005). Many cell- types derived from eS6P-/- mice are significantly smaller than in the wild type, among them pancreatic beta cells, IL7-dependent cells from fetal livers, MEFs, and muscle myotubes (Granot et al., 2009; Ruvinsky et al., 2005, 2009). However, some cells have normal size, for example acinar cells from pancreas (Pende et al., 2004; Ruvinsky et al., 2005). eS6P-/- mice have impaired renal hypertrophy after unilateral nephrectomy (Xu, 2015). Interestingly, there seems to be a link to initiation, as phosphorylation of serine 247 promotes association with mRNA Cap-binding complex in vitro (Hutchinson et al., 2011), and phosphorylation of serines 240/244 is required for Cap binding (Roux et al., 2007). Further, by covariance analysis of a cryo-EM map of a 43S initiation complex, density variance in the eS6 region was shown to be related to density variance of initiation factors (Liao et al., 2015) (the sample analyzed was (Hashem et al., 2013)). Nevertheless, if there is a clear structural role of eS6 phosphorylation is not known.