|Home | About | Journals | Submit | Contact Us | Français|
The function of nuclear actin is poorly understood. It is known to be a discrete component of several chromatin-modifying complexes. Nevertheless, filamentous forms of actin are important for various nuclear processes as well. Nuclear actin is often associated with nuclear actin-related protein Arp4 and other actin-related proteins like Arp8 in the INO80 chromatin remodeler. We recently determined the crystal structure of S. cerevisiae Arp4 that explains why Arp4 is unable to form actin like filaments and shows that it is constitutively bound to an ATP nucleotide. More interestingly, in vitro activities of Arp4 and Arp8 seem to be directed towards stabilizing monomeric actin and to integrate it stoichiometrically into the INO80 complex. Based on this activity, we discuss possible roles of nuclear Arps in chromatin modifying complexes and in regulating more general aspects of nuclear actin dynamics.
The dynamic nature of chromatin and its epigenetic information content are regulated by numerous chromatin modifying enzymes. Among those are ATP dependent remodelers, which slide or evict nucleosomes, or exchange histone variants, as well as enzymes that add, alter or remove posttranslational modifications on histones and DNA.1 Several of these chromatin remodeling and modifying complexes including INO80, SWR1, RSC, NuA4 and SWI/SNF contain actin and actin-related proteins (Arps) as subunits.2–6 Arps and actin are often integrated into these complexes as pairs like for example the Arp4—actin pair and bind to a domain present in the Swi2/Snf2 ATPase subunit of the remodelers, the helicase—SANT associated (HSA) domain.7 Several nuclear Arps have been shown to interact with histones, suggesting a function in targeting their remodelers to the chromatin substrate.8–11 The mammalian INO80 complex for example is targeted to DNA damage sites in an Arp8 dependent manner.12 Nevertheless, additional functions of Arps within the chromatin modifying complexes and the functional role of actin and the actin-Arp interactions remain elusive. Figure 1A depicts a model of the INO80 chromatin remodeling complex that contains an Arp4-Arp8-actin subcomplex.
The functional understanding of nuclear Arps, which appear to be specific for chromatin modifying complexes, is further hampered by a general lack of understanding of the properties of actin in the nucleus. In fact, roles of nuclear actin and its polymerization state are still rather controversial, as phalloidin stainable actin filaments are not observed in the nucleus.13 Nevertheless, nuclear actin has been implicated to function in long-range chromatin movement and is required for proper transcriptional activity of all three RNA polymerases.14,15 In particular, the inhibition of polymeric actin by latrunculin in vivo inhibits RNA transcription.16 Likewise, inhibition of RNA polymerase I mediated by an antibody directed against actin can only be overcome by providing polymerization competent actin but not by actin mutants that do not support filament formation.17 Several studies have also shown that the movement of nuclear loci exhibits the characteristics of an active transport process and is dependent on polymeric nuclear actin.14,18,19 In short, some polymeric forms of actin seem to play important roles in nuclear processes and there is presumably a need for regulation of nuclear actin polymerization.
To initiate structural work on nuclear Arps and to investigate a possible connection between nuclear Arps and nuclear actin dynamics, we first determined the crystal structure of S. cerevisiae Arp4.20 Arp4 is structurally very similar to actin, exhibiting the classical actin fold with four subdomains centered around a central nucleotide.21 Small angle x-ray scattering suggests that Arp4 remains monomeric to rather high concentrations and does not appear to form filaments by itself. This is due to the presence of specific amino acid insertions at the barbed and pointed end side of Arp4 as well as a deletion at the DNase binding loop—all positions which are crucial for contact formation within the actin filament (see Fig. 1B and C).
An interesting feature of the Arp4 structure is the stable binding of an ATP nucleotide that has been copurified from the expression host. The apparent lack of ATPase activity is in contrast to actin and can be explained by examining the ATP binding site of Arp4. Several amino acid changes combined with the positional shift of an α-helix lead to a reduced accessibility of the bound nucleotide to the solvent (see Fig. 1B and C). Together with the exchange of catalytically important residues (His 161 and Gln 137 of actin are exchanged by Ser 166 and Thr 142 in Arp4 respectively) this might lead to a stable ATP bound state of the protein.22–24 This does not rule out the possibility that Arp4's ATPase activity might be triggered in the cellular context, e.g., by binding to other interaction partners. On the other hand, ATP might simply be a structural component to stabilize the fold of Arp4. The structure now offers a framework to address the role of ATP binding and hydrolysis of Arp4 by mutagenesis studies.
Arps are prominently involved in regulating actin dynamics in the cytoplasm. In particular, the Arp2/3 complex, which contains seven subunits including Arp2 and Arp3 nucleates actin filaments and regulates branching of preexisting filaments in the cytoplasm.25 We therefore tested if the nuclear Arps of the INO80 complex have an influence on actin filament formation in vitro as well. Indeed, we found that Arp4 is able to inhibit actin polymerization and to depolymerize actin filaments.20 Our biochemical data suggest that it interacts with actins' barbed end and seems to prefer ADP bound actin over ATP bound actin. Interestingly, Arp4 does not form a polymerization incompetent complex with actin, making its properties similar to the well known actin binding protein profilin.26 The mode of binding of Arp4 to actin remains to be established. Preliminary data indicate that the barbed end side of Arp4 is important to mediate its interaction with actin which is surprising at first glance as one would expect a filament-like pointed to barbed alignment of actin with the conserved actin-fold of Arp4 (unpublished). If this is true, the interaction of Arp4 and actin via their barbed ends could be reminiscent of the so called antiparallel lower actin dimer that is implicated in the early stages of filament formation.27,28
Arp8 on the other hand is not able to inhibit actin polymerization on its own but slowly depolymerizes actin filaments with a preference for ADP actin as well. Most interestingly, Arp4 and Arp8 together seem to reinforce their individual effects. Arp8 increases the inhibition of actin polymerization by Arp4 while the depolymerization of actin filaments by both proteins seems to be additive leading to increased depolymerization. In other words, both proteins stabilize actin in its monomeric state by inhibiting polymerization and promoting depolymerization of filaments in vitro. Those activities are consistent with the possible function of Arp4 and Arp8 to integrate actin as a stochiometric component into the INO80 complex.
Hence, the general properties of Arps in being able to influence actin dynamics seem to also hold true for nuclear Arp4 and Arp8 which were previously only implicated in histone binding. The nuclear Arps might therefore be macromolecules with a dual functionality both in chromatin association and actin dynamics.
Some interesting questions that have not been answered so far are if the activities of Arp4 and Arp8 are merely tailored to recruit monomeric actin into the INO80 complex, or if both proteins have a more general role in regulating nuclear actin dynamics outside of the INO80 complex. Regarding the first hypothesis, the question remains what exactly the function of actin in the remodeler is, or whether it is merely a structural component. Furthermore, the role of the assembled Arp4-Arp8-actin complex within the INO80 remodeler is unclear. Finally, since Arp4, actin and possibly Arp8 bind to ATP and are potential ATPases it would be interesting to study the influence of the ATP/ADP state of those proteins on the assembly state and conformation of the subcomplex.
It has been shown that a portion of nuclear Arp4 occurs in a monomeric pool separate from the high molecular weight fraction that is usually associated with nuclear complexes.29 Moreover, a role of Arp8 in chromosome alignment during mitosis has been shown to be independent of other components of the INO80 complex.30 This supports a more general role of Arp4 and Arp8 in actin metabolism as both proteins can perform functions independently from larger complexes. Specifically, the close resemblance of Arp4s' in vitro properties to profilin could indicate that it has a role in maintaining the pool of monomeric actin in the nucleus. Additionally, a knockdown of mammalian Arp4 by RNAi increases chromosome territory size and knockout of Arp8 in yeast leads to increased cell volume and irregular cell morphology.31,32 Although not conclusive, these observations are consistent with a possible role of Arp4 and Arp8 in actin metabolism.
Likewise it might be possible that once actin has been properly assembled within INO80 it gains additional functions in actin polymerization, reminiscent of the Arp2/3 complex. According to our data, Arp4 interacts with actins' barbed end whereas Arp8 neither seems to bind to the barbed nor the pointed end. This gives rise to a possible complex model where the pointed end of actin, if not masked by either the HSA domain or another INO80 subunit is free to accept the binding of an additional actin molecule. The complex could thus stimulate the formation of new actin filaments from the pointed end, or bind to existing filaments and stabilize them. In support of this, the BAF complex, a chromatin remodeler that also contains Arp4 and actin binds to actin filaments in the presence of phosphatidylinositol-4,5-bisphosphate in vitro and stabilizes them.33 A stimulation of pointed end filament growth however, would be quite unusual and in contrast to known actin nucleators like the Arp2/3 complex which usually mediate barbed end elongation of filaments.25
Monomeric actin has a rather poor ATPase activity that is stimulated about 40,000-fold upon integration of the monomer into the actin filament.34 Although much has been learned from electron microscopic and fiber diffraction studies of actin filaments there are no high resolution atomic structures of filaments available and the exact rearrangements of amino acid side chains at the nucleotide binding site upon ATPase activation remain unclear.35,36 It might be possible that a similar stimulation of ATPase activity occurs for the Arps once they are integrated into the INO80 complex and/or the complex is in a particular functional state. The resulting change in ATP binding state of Arp4 or Arp8 could then modulate the interaction of those proteins with each other and with other binding partners. It has been shown that both Arp4 and Arp8 interact with histones8–10 and could possibly function as histone chaperones. The nucleotide state of the Arp complex could therefore also regulate its interaction with histones or nucleosomes.37 The preferred interaction of Arp4 and Arp8 with ADP-actin might indicate that actin itself acts as a switch to control the assembly state of the complex with ATP binding triggering a dissociation of the complex.
Our publication on the INO80 subunits Arp4 and Arp8 and their impact on actin dynamics possibly connects two previously unrelated fields of research and provides another hint towards a sophisticated and highly regulated actin network within the nucleus. Despite the lack of phalloidin staining of nuclear actin, more and more evidence emerges that filamentous actin does indeed exist in the nucleus and is involved in several important processes. Nuclear Arps might contribute to altered properties of actin dynamics compared to the cytosol and this could occur dependent or independent of Arp and actin harboring chromatin remodeling complexes.
It should be kept in mind that the possible activities of Arp4 and Arp8 as well as the Arp4-Arp8-actin subcomplex of INO80 in regulating actin dynamics are speculative up to this point. There are hints in the literature as well as from our biochemical data that would support such roles. However, it will require further in vivo testing to really prove or disprove such functions. Such studies will be inherently difficult as the probing of nuclear actin and especially its polymerization state are tricky and not well established. Furthermore, the roles of Arp4 and Arp8 in interacting with histones need to be separated from their potential roles in actin metabolism. Nevertheless, it would be an exciting finding if the INO80 complex connects chromatin remodeling on the one hand with the regulation of nuclear actin dynamics on the other hand via its Arp4-Arp8-actin subcomplex.
Work on Arps in K.P.H.'s laboratory is supported by a scholarship from the Boehringer Ingelheim Fonds to S.F. and grants from the DFG (SFB/TR5, SFB646 and SFB684) and the German Excellence initiative (Center for Integrated Protein Science and Munich Center for Advanced Photonics) to K.P.H.