Ssu72 Discussion

Function

In attempting to infer the function of the Drosophila Ssu72 protein, two broad sources of information were used, each leading to distinct conclusions (the compatibility of which will be considered in due course). The value of the first data set as a proxy for experimental information about the functionally uncharacterised Drosophila protein is predicated on the high level of similarity in amino acid sequence that were found to exist between this protein and the human Ssu72 protein and what we infer must be a very close match in both secondary and 3D structure. This data comprises both in vivo and in vitro functional analyses of the latter protein, which tend to suggest that the protein in Drosophila similarly functions as a serine phosphatase – presumably at the C-terminal domain of RNA polymerase II (Krishnamurthy et al., 2004) – and has a role in 3' end mRNA processing that is independent of its activity as a phosphatase. The second piece of guidance comes from the results generated by the use of computational and bioinformatic tools in this paper, which prima facie reveal the Drosophila protein to be a protein tyrosine phosphatase. Therefore, these two seemingly conflicting conclusions must now be evaluated on the basis of the limited evidence available to determine which is more likely to be true, or whether one necessarily precludes the other.

Protein Tyrosine Phosphatase?

An evaluation of the Drosophila protein as a potential Protein Tyrosine Phosphatase (PTPase) is a logical place to begin this discussion as this is what Meinhart, Silberzahn, & Cramer (2003) concluded about the human protein, based on their initial characterization of its primary, secondary and 3D structures, as well as a number of important pieces of experimental data. They began by identifying, within the amino acid sequence of the Ssu72 gene, the presence of a PTPase signature motif at the N-terminal end of the protein. This motif is a short sequence, the consensus sequence being (H/V)C(X5)R(S/T), which forms the phosphate-binding site (P-loop) in the mature protein and is where, invariably, members of the PTPase superfamily exert their catalytic action. This motif is similarly present in the Drosophila Ssu72 protein and, what is more, shares almost the precise sequence as that of the human protein (VCSSNQNRS), but for an unknown amino acid in place of the glutamine (Q). Meinhart, Silberzahn, & Cramer (2003) inferred, based on this motif, that if the human Ssu72 were a PTPase, it would be one of the low molecular weight family. This inference was based on the presence of a characteristic asparagine (N) in the motif, which the Drosophila protein shares. Further analysis of the amino acid sequence of the human protein found the presence of two important residues, which form part of an 'aspartate loop' that is important for the binding of substrate at the P-loop. Those residues are aspartates 140 and 143, both of which are present in the Drosophila protein and which were found, by mutational analysis, to contribute to the human protein's activity (Meinhart, Silberzahn, & Cramer, 2003).

In the functional results presented in this report, ProFunc was used to, among other things, identify potential, functionally-important triad template motifs from a database. The only such motif that was found to be present in the Drosophila protein was centered around the three important residues of the PTPase signature motif - valine 8, cysteine 9 and serine 16 - the latter two of which are essential for PTPase activity. The 3-dimensional coordinates that those residues occupy in the protein were found to match those of a low molecular PTPase (myobacterium tuberculosis low molecular weight tyrosine phosphatase) with an RSMD of 0.15Å (see Figure 1). The degree to which this motif is conserved must, at least in part, be indicative of its physiological importance. This high degree of evolutionary conservation across a range of organisms (see Figure 2), combined with the fact that mutation of the essential cysteine residue to a serine abolishes human Ssu72 activity and is embryonic lethal (Meinhart, Silberzahn, & Cramer, 2003), is extremely strong evidence that this motif is functionally important. According to Zhang (1998), 'amino acid sequence comparisons of the catalytic domains of PTPases with the catalytic subunits of protein Ser/Thr phosphatases have shown no sequence similarity'. It is therefore unlikely that the serine phosphatase activity reported by Krishnamurthy et al. (2004), among others, occurs at this catalytic site; or, if it does, is by a mechanism that has not been characterised thus far.

Meinhardt, Silberzahn, & Cramer (2003) then went on to examine in vitro whether the human protein did indeed function as a tyrosine phosphatase. Their finding that it did was based on it acting in vitro to cleave a phospho-tyrosine analogue, pNPP, and that PTPase inhibitors caused an abolition of this phosphatase activity. It is likely, given the conservation of the PTPase catalytic domain in the Drosophila protein, that this experimental result could be replicated with it. However, it now becomes important to consider how the broad structure of the Drosophila protein compares with that of known PTPases. In the structural results presented in this report, DaliLite was used to superimpose the 3D form of the Drosophila Ssu72 protein over inter alia a representative low molecular weight tyrosine phosphatase - Bovine Low Molecular Weight PTPase (see Figure 3). The match is, on the face of it, quite poor. However, in the absence of more relevant information about which particular regions are shared between the two proteins, it is impossible to determine what the significance of such a difference is. It is possible that it is of no functional consequence as far as potential PTPase activity is concerned, the catalytic motif being present, and that the difference accounts for additional activity lacking in traditional PTPases - such as serine/threonine phosphatase activity. However, more analysis is required to elucidate this more completely.

Serine Phosphatase?

Based on the very high similarity in primary structure between the human and Drosophila Ssu72 proteins (see Figure 2) and in the absence of structural information about the former, due to there being no PDB entry for it, we may suppose, for the purposes of this analysis and in the absence of further experimentation, that the two have equivalent functional characteristics. On the basis of that proposition, it is worthwhile to briefly discuss the evidence which supports their role as a serine phosphatase. Krishnamurthy et al. (2004) report that human Ssu72 is responsible for the dephosphorylation of serine 5-phosphate (S5-P) of the C-terminal domain of RNA Polymerase II, as well as mediating the 3' end processing of pre-mRNA. However, only the former will be discussed here, due to the latter being brought about in a manner that is independent of any Ssu72-mediated phosphatase activity.

In their paper, Krishnamurthy et al. (2004) induced a depletion of Ssu72 in vivo and observed a corresponding increase in the amount of phosphorylated S5 of the CTD of RNAPII. Secondly, the researchers introduced a form of Ssu72 in vitro that had a C15S mutation and observed no decrease in the relevant S5-P. Lastly, they inhibited the action of Ssu72 and, once again, observed an increase in CTD S5-P. These results certainly do indicate that Ssu72 is capable of causing the dephosphorylation of phospho-serine. However, it is significant to note that this activity appears to be dependent upon cysteine 15, one of the essential PTPase motif residues. Therefore, Ssu72 is either a 'new class of protein phosphatases', as the researchers suggest, or that dephosphorylation of S5-P is an indirect result of Ssu72 PTPase activity. This latter explanation would certainly provide a more reasonable explanation for why Ssu72 appears to be capable of two catalytic activities (tyrosine and serine dephosphorylation), using one site that has only ever been associated with specific PTPase action. However, it is clear that further experimentation is needed in order to gain a more clear understanding about how the human Ssu72 protein functions, as well as whether the apparent functional equivalency of the Drosophila Ssu72 that has been discussed in this paper can be demonstrated or falsified experimentally.

References

Krishnamurthy, S., He, X., Reyes-Reyes, M., Moore, C., and Hampsey, M. (2004) Ssu72 Is an RNA Polymerase II CTD Phosphatase. Molecular Cell 14(3), pp. 387-394.

Meinhart, A., Silberzahn, T., and Cramer, P. (2003) The mRNA Transcription/Processing Factor Ssu72 Is a Potential Tyrosine Phosphatase. J. Biol. Chem., Vol. 278, Issue 18, 15917-15921.

Zhang, Z-Y. (1998) Protein-Tyrosine Phosphatases: Biological Function, Structural Characteristics, and Mechanism of Catalysis. Critical Reviews in Biochemistry and Molecular Biology, 33(1):1–52.

Evolution

Homologues to the target sequences are present in a wide variety of eukaryotes. The protein sequence is well conserved in this variety of organisms, particularly in regions corresponding to secondary structure features (figures 20 and 21). Some organisms have multiple copies of the gene, and these have diverged. However, the phylogenetic tree indicates that the functional copy of the gene has accumulated mutations at a steady and low rate – occasionally, too low to distinguish confidently between closely related organisms such as the tetrapods (figure 24). This indicates that the protein is vital for cell function. Also, the correspondence between the phylogenetic tree and the overall taxonomy (see figure 23) suggests that the protein was first present in Eukaryotes' last common ancestor.

Tyrosine phosphatase?

The target protein shares the catalytic residues characteristic of tyrosine phosphatases (H/V)C(X5)R(S/T), and also has an identifiable aspartate loop, making up the tyrosine phosphatase active site (figure 22). However, the overall sequence similarity is low. This indicates that the protein is not in the tyrosine phosphatase family, but is likely to have tyrosine phosphatase activity.

Serine phosphatase?

The family of ssu72 proteins and homologues shares the phosphatase binding/catalytic residues, a particular D(I/V)(Q/K/R)D aspartate loop, and many conserved areas of secondary structure (figure 22). The aspartate loop is different to that of similar tyrosine phosphatases (D(A/S)D): it is longer, the residues it contains are larger, and it contains an additional polar (neutral or acidic) residue. It is probable that this domain is responsible for allowing ssu72 to take serine as a substrate.

The pattern of conserved residues is consistent with the known similarities between the secondary structures of ssu72 and tyrosine phosphatases. However, ssu72 has a number of additional regions, including the region with the new aspartame loop. The other regions contribute to secondary structure, but do not alter it drastically: the first region is an additional strand in a beta sheet, the second is random coil, and the third forms part of an alpha helix. There is no clear change to the secondary structure which could be responsible for the ability of ssu72 to work on serine. However, it may be important that extra space is given to the aspartate loop. This would allow the protein to accommodate the long, charged residues on the loop. It could also make the loop more flexible and accessible.

Structure

Through structure comparison, similarities were founded between the protein being researched (PDB:3fdf) and a 8-chain protein serine phosphatase, a tyrosine phosphatase, and Ssu72 RNA polymerase II CTD phosphatase homolog [Homo sapiens]. It is important to point that although tyrosine and 3fdf protein are not strongly similar similar in the sequence, they have similarities in the structure. Also, the homo sapiens protein structure is just a predicted one using the pdb database for searching template that actually is 3fdf). Only one domain is found in 3dfd: Ssu72 according to Pfam.

Heading

Abstract | Introduction | Results | Discussion | Method | References