In this study, we have identified domain of S6 kinase 1(S6K1) as a novel regulator of the actin cytoskeleton and found that it is pivotal for the cross linking of F actin (filamentous actin) and induction of filamentous actin bundles. Several aspects of this function are worth highlighting. We showed for the first time the domain specific interaction of S6K1 with filamentous actin (F actin) with a binding potential inclined more towards the catalytic domain truncated version of S6K1 [∆NH2 − 146 / ∆CT104]. The observation fits with the model preposed for S6K1 activation [14]. According to which the kinase exists in two conformations, inactive and active state. In the inactive state of S6K1, the carboxyl-terminal autoinhibitory domain, which has sequence similarity to the substrate region of the S6 protein, may act as a pseudosubstrate and interacts with the N-terminus According to this model, S6K1 activation is initiated by the release of the autoinihibition exerted by the autoinhibitory domain [15]. This is then followed by a series of phosphorylation of eight or more serine or threonine residues at the autoinhibitory domain, the linker region, and then the catalytic domain, to obtain full kinase activation [14, 16–20].
Therefore the increased activity associated with the different truncated versions of S6K1 as against to full length S6K1 may therefore, be simply due to steric freedom achieved by the truncated versions of the enzyme ordinarily accomplished by activating phosphorylations. In the case of catalytic domain truncated version of S6K1 [∆NH2 − 146 / ∆CT104] the binding site becomes fully exposed to the substrate which makes it to follow the higher binding kinetics with its interacting partner F actin, which doesn’t remain the case with its full length version when this binding site for actin remains less exposed and hidden within the core of the protein showing marginally lesser binding kinetics with F actin as governed by its protein confirmation.
The actin cross linking assay of the preclear lysates of different versions of S6K1 expressed in SF9 (Insect cell line) and bacterial (Prokaryotic) system have shown that interaction is primarily with the filamentous actin rather than with monomeric actin (G actin) well supported by a similar kind of study by AST Wong et al., group [13]. The binding of S6K1 and F actin is not facilitated by any other protein rather it is direct interaction as revealed by the actin cross linking assay carried out in bacterial system. The S6K1 was always found in the pellet, indicating that the pelleting is because of binding to F-actin. The fact that no other proteins were involved indicates a direct association between S6K1 and actin.
The possibility of a physical interaction between S6K1 and actin suggests that S6K1 may affect the kinetics of actin polymerization. The S6K1 did not change the rate or the extent of polymerization suggesting that unlike other common activities of these proteins, actin polymerization is not a generic property of S6K1. It seems polymerization of G actin is a very fast event reaching almost steady state within 5 minutes of its induction with polymerization buffer indicating that polymerization and bundling of actin is a synchronous event as observed by us and well supported by light scattering experiment carried out by AST Wong et al., group [13]. Actin filaments are in a continuous state of assembly/disassembly. As a consequence, in the steady state, a small but finite concentration of actin monomers will be present in any filament population as obvious from the time course experiment [21, 22].
Due to the higher binding potential of [∆NH2 − 146 / ∆CT104] S6K1 and due to its predominant α helical structure [23] has made us to focus on this kind of domain so as to elucidate its functional significance in the binding kinetics of S6K1 and actin. We could show that the [∆NH2 − 146 / ∆CT240] S6K1 is the main binding domain. It also narrowed down the minimal region in S6K1 which is responsible for binding to actin. Mutagenesis analysis [24] suggests that interaction of one of the neuronal protein neurabin with S6K1 is through its PDZ domain located in the middle portion of the protein. Because S6K1 does not share homology with PDZ domain containing proteins, and because deletion of the C-terminal five amino acids of S6K1 abrogates binding, this interaction is an example of heterotypic binding of a PDZ domain to the C terminus of a substrate protein S6K1 (amino acids 332–502). Nakanishi et al., [25] independently identified neurabin as an F-actin cross linking protein. Neurabin binds to F-actin through a unique N-terminal domain (amino acids 1-144) that targets the protein to the cytoskeletal compartment. This actin-binding domain is some distance from the PDZ domain of neurabin, so actin and S6K1 may bind independently providing a supportive literature evidence to the kind of novel interaction we have found between F actin and (amino acids 147–285) of S6K1.
Bioinformatics study was carried out in order to further confirm this novel region; it was observed that this region contains predominantly hydrophobic amino acids and secondary structure prediction of this region has revealed predominant α helical and coiled coil regions which serves as a structural basis for some of the known actin binding proteins. The hydrophobic cleft between actin sub domains 1 and 3 appears to be a ‘hot spot’ for both G-actin- and F-actin-binding proteins. The conformation of this cleft is such that it preferentially binds an α helix of the binding partner, which is characterized by the presence of some exposed and conserved hydrophobic side chains [26]. In addition to the α helices of gelsolin, DBP and ciboulot – which bind this cleft in their respective structures with actin [27–29] – the α helix in the D-loop of actin and ADF/cofilin α-helix 3 are proposed to also bind in this cleft and the coiled-coil domain of HS1 has been observed to bind to F-actin [30].
In vitro interaction between S6K1 and F actin was not regulated by the phosphorylation of S6K1 but happens in a manner independent of phosphorylation suggesting a phosphorylation independent mechanism for S6K1 and F actin binding.
We also used electron microscopy to visualize the F-actin bundling. In the absence of [∆NH2 − 146 / ∆CT240] S6K1, actin filaments formed a uniform meshwork of fine filaments, but did not show bundling. In contrast, robust multifilament bundles were clearly seen in its presence. These closely apposed bundles were often slightly curved, suggesting flexible cross-linking. An intriguing aspect concerns the biological function of such an interaction. The finding that S6K1 both cross-links and stabilizes actin filaments suggests a role for S6K1 in regulating actin dynamics. Several key actin-binding proteins, such as α-actinin, fimbrin and fascin also function to regulate actin bundling [31, 32]. As S6K1 has a well established role in protein synthesis, this interaction may also have significance in synthesizing local proteins important for propagating the migratory response. For example, β-actin mRNA has been shown to localize to the leading lamella and its active translation there is important for cell migration [33, 34].
In summary, the present study reveals novel identification of S6K1 domain and its function in regulating actin cytoskeleton dynamics. The S6K1 is frequently active in ovarian and a wide range of cancer types, and it has a crucial role in several processes considered hallmarks of cancer like metastasis. Targeting this S6K1 actin binding domain through the use of molecular inhibitors, SiRNAs and natural compounds may represent a beneficial new avenue for cancer therapy and opens new areas of investigation in S6K1 biology.