Discussion 5

From MDWiki
Jump to navigationJump to search

ProFunc returned a number of promising results including Interpro, PDB, SSM and DALI and 3D functional template searches. These results while not being significant did support our hypothesis. The result of the UniProt search however was significant. A multiple sequence alignment search returned 10 significant matches 2 of which had a percentage identity higher then 80%. Both of these sequences were listed as YlqF proteins. All of the proteins featured in these results were either GTPases or GTP-binding proteins. Therefore this sequence search supports our hypothesis of 1pujA being a GTPase.

The SymAtlas results show the expression levels of ortologs for 1pujA in different tissues of both mice and humans. The results indicated that these proteins are widely distributed in all tissues of both animals. This is consistent with a protein essential for fundamental biological processes. Though the results do not provide a greater level of detail than this, a GTPase would fall into this category, further supporting our hypothesis.

ProKnow assigns function by extracting and interpreting protein features from sequences and structure. It uses metaserver strategy through a knowledgebase of annotation profiles coupled with Bayesian scoring. (REF= http://www.doe-mbi.ucla.edu/Services/ProKnow/proknow.png) The results of our ProKnow search indicated that GTP binding was the molecular function and to a lesser degree nucleotide binding, methyltransferase activity and GTPase activity. The biological process was small GTPase mediated signal transduction, which refers to any series of molecular signals in which a small monomeric GTPase relays one or more of the signals. These ProKnow results also support our hypothesis that 1pujA is a GTPase.


Structural analysis shows that YlqF is likely to be a GTPase. A significant finding was seen with Dali search analysis. The analysis of Dali showed that many of the proteins with similar structrues to YlqF to be GTPases. However, Consensus protein fold classifiation determined by SCOP, CATH, and Dali showed different views of protein fold space. This is due to the fact that they use different methods to define and categorize protein folds. Pfam analysis also revealed two domains, MG442 and MMR_HSR1, which are both GTP binding proteins. The secondary structure analysis revealed that YlqF is 50% helical which is made up of 13 helices containing 142 amino acids and 10% beta sheet which is of 6 strands containing 31 amino acids.

Other studies have shown that YlqF, the smallest cpGTPases in domain composition possess a CPG domain and a C-terminal alpha helical domain (Anand et al., 2006). They also suggest that YlqF lacks an N-terminal domain (Anand et al., 2006). Therefore, proteins like YlqF are the prototypes of cpGTPases (Anand et al., 2006). They are not likely to exist as single GTP binding domains (Anand et al., 2006). Studies suggest that cpGTPases require the presence of C-terminal domain in order for it to bind GTP (Anand et al., 2006).




P-loop NTPases are the most abundant proteins in cellular organisms, constituting 10-18% of all gene products. They are distinguished by the Walker A motif (consensus GxxxxGK[ST]), Walker B motif (consensus hhhDxxG, where h = hydrophobic residue), and the [NT]KxD motif which is unique to P-loop NTPases. The Walker B and [NT]KxD motifs indicate specificity towards GTP (Leipe, D. et. al. 2002). MSA of known and suspected GTPases displayed high conservation of these characterising motifs (Figures 8 & 9) indicating that 1pujA is likely a GTPase (Anand, B. 2006). Furthermore, all sequences aligned had the [NT]KxD motif circular permutation, indicative that they are all part of the YlqF/YwaG family within the P-loop NTPase superfamily (Leipe, D. et. al. 2002). It has previously been hypothesised that the Last Universal Common Ancestor (LUCA) to all extant life forms possessed several GTPases. If a particular GTPase family is widely represented in the three primary kingdoms (Archaea, Bacteria, and Eukaryota), this is evidence for presence in LUCA. This is supported if the phylogenetic tree conforms to the “standard model” topology, having bacterial and archeo-eukaryotic primary clades. Conversely, a different topology such as a bacteria-eukaryote grouping could indicate presence in LUCA but ancestral form in eukaryotes displaced by horizontal gene transfer (HGT) (Leipe, D. et. al. 2002). 1PUJA homologs/orthologs were found to be widely represented in the three primary kingdoms and conformed to the “standard model” topology (Figure 11). All of which is suggestive of the presence in LUCA. Ylqf proteins such as 1pujA may have been transferred from bacteria to eukaryotes once during the early stages of eukaryotic evolution (probably from the proto-mitochondrion), and a second time from chloroplasts to plants. It has been stipulated that GTPase activity is a result of adaptation. GTP is more constant within a cell and not subject to the same fluctuations as ATP. Hence specificity for use of GTP as a substrate was recruited in crucial functions such as translation (Leipe, D. et. al. 2002). The splitting of eukaryotes into two groups (Figure 11) is due to an addition N-terminal domain (Figure 10). This domain is a suspected coiled coil region associated with RNA and ribosmal attachment, distinguishes proteins of the YwaG family. It has been shown that YlqF and YwaG share GTP related domains (Figures 8 & 9) and C-terminal domains however YlqF lacks the N-terminal coiled coil regions of YwaG. Dispite this YlqF proteins are still expected to associated with RNA or ribosomes due to the shared C-terminal domains (Anand, B. et. al. 2006). The Phylogentic tree also suggests that YlqF is ancestral to YwaG.



Structural, functional and evolutionary analyses collectively indicate that 1pujA is a GTPase of the YlqF family. High conservation of YlqF over large spans of evolutionary time (Figure 11) indicate strong stabilising selection and suggest a role in one or several crucial biological processes (Leipe, D. et. al. 2002). Functional results indicate YlqF has a very important function inferred from the results of expression in all tissues of humans and mice. Although the current study can not directly show association of 1pujA with ribosomes, it is concluded that YlqF proteins are GTPase likely to be involved in a fundamental biological process such as translation. This study support of the work by Matsuo et. al. (2002) in which it was hypothesised that YlqF is needed for correct assembly of the 50S ribosomal subunit.