Structure analysis showed that the NUBP2(MinD homolog, E.Coli) is like to belong to the membrane associated ATPase family, a significant finding was seen with Dali search analysis. Due to the structure of NUBP2 was not solved, as the structure of 2ph1 from Achaeoglobus fulgidus is 39.66% similar to NUBP2, it was used to search in Dali. Most of the proteins with similar structures to 2ph1 are from this family with similar function.
1.1Summary of 2ph1 structure
The nucleotide binding protein 2ph1 from Achaeoglobus fulgidus is a 240 residue protein with a molecular weight 26087.there are 8 alpha helices 7-14,32-45,61-65,111-122,143-151,167-178,211-217,230-237,and 12 beta strands 83-84,91-93,50-54,131-135,20-24,155-160,184-189,222-225,74-76,79-81,194-195,202-203.. the resolution of 2ph1 is 2.7 Å. The PDB file observed in Pymol shows that the structure of 2ph1 is alpha+beta. The P-loop is started from about residue 9 to 16 “[GA]-X2-(G)-X-G-K-[ST]”. The walker B motifs are around residue 34-49 and residue 117-130. the 3-D structure shows that these motifs form a pocket in the middle of the protein which is the ATP binding site.( Nakashima, 1999;Boer, 1991)
1.2 Dali analysis
From the Dali result, the protein 1g3q has the highest Z-score (20) and 21% similarity with 2ph1. The alignment from Pymol of these two protein shows the RMS of 8.55. The high similarity of structure suggested the high similarity of function as a cell division inhibitor. 1g3q also belong to the MinD ATPase family.
1.3 P-loop in 2ph1
In 2ph1 the P-loop presents as GKGGVGKST. The walker B motif is presents as LDADFLG. These motifs are characteristic features of ATPase as they comprise important nucleotide binding and intramolecular signaling domains in P-loop ATPase. (Hayashi, 2001)
2.1 Protein-interaction annotation:
Exploration of existing database on UnitProtKB/Swiss-Prot annotated the relationship of NUBP2 and MinD. This data similarity search detected closely related sequences from other species, and their annotation give the first clue regarding a possible function inferred from the known function of related sequences. Similar sequences originated from different species- and in an order consistent with evolutionary distance may be considered potentially orthologous.
2.2 Sequence analysis :
UniProt database searches with 2ph1 (chain A) amino acid sequences indicated the sequence homology between MinD (MRP, in prokaryote) and NUBP (NBP, in eukaryote). Sequence similarity searches showed a likely relationship to the E. coli MinD gene in 48-45% of the amino acid position (Table 3). MinD gene is a membrane-associated ATPase that inhibits cell division at the poles and consequently induces normal cell division. Based on the similarity, the function of NUBP2 is predicted to involve in cell division. The remarkable alignment score illustrated the evolutionary distance between E. coli and human. Due to the similarity between NUBP2 and MinD, we favor the hypothesis that NUBP2 participates in regulating cell division and poses the characteristic of ATP-binding protein.
Research conducted by Shahrestanifar in 1994 investigated the expression of rat homolog of NUBP2 in several cell lines; it was found that NUBP2 was presence in all samples. Highest expression of NUBP2 was found in human adult’s lung, testis and skeletal tissues followed by kidney, brain, spleen and heart (Nakashima et al., 1999). Another research conducted by Unger and Hartwell in 1976 demonstrated mutation of the gene was found to be lethal, indicating the NUBP2 plays a vital role in cell division.
2.3 Sequence motif (domain) analysis:
Results returned from InterProScan indicate this nucleotide binding protein (NBP) shares a characteristic motif with the ATPase superfamily (Table 2.3). Additionally, a characteristic sequence motif “[GA]-X2-(G)-X-G-K-[ST]” called the phosphate binding group (p-loop) was identified (Table 2.4). Many biologically processes are thermodynamically unfavorable, and therefore cannot occur without the use of an extra source of energy. In many cases, this source of energy comes from the hydrolysis of adenosine triphosphate (ATP) molecules. Based on its ATP binding motif, this gene was called nucleotide-binding protein (NBP). Therefore, NBP is also known as ATP-binding domain.
2.4. Protein-protein interaction:
Regardless of function, proteins seldom act alone and are usually assembled into complexes and dynamic macromolecular structures to perform their task in the cell. The majority of the interacting proteins (up to 70-80%) share at least one function (Lane et al., 2007). This property is used to assign a function to the given protein based on the functions of its characterized binding partners. In-silico analysis using String database showed that Methionyl-tRNA synthetase function is conserved between prokaryotes and eukaryotes (as highlighted in Figure 1 and 2). This annotated function is supported by the studies done by Unger and Hartwell. Their experiments concluded the necessity of Methionyl-tRNA synthetase to control cell division in yeast. Hence, it supported our hypothesis that NUBP2 has a vital role in regulation of cell division (cytokinesis).
2.5 Subcellular location:
The returned results from GNF SymAtlas database showed that human and mouse NUBP2 protein is expressed in all range of cell. Highest expression of this protein is found in lung, liver, spleen, testis and skeletal tissues as they involve in cell repair and cell division. NUBP2 belongs to nucleotide-binding proteins family; it is required to hydrolyse ATP to drive an energetically unfavorable reaction into thermodynamically favorable one. The minimum expression of NUBP2 in other tissues explains that other nucleotide-binding protein family might be up-regulated in certain tissues, for example NUBP1 which is highly expressed in spleen and eye organs.
By using bioinformatic methods, we identified four conserved motifs of NBP (ATPase) across three domains. Besides the ATP-binding domain (p-loop, 23~34 a.a. of NUBP2), we also found the alpha and beta NBP/MRP domains, which is consistent to Nakashima et al. (1999). While varying results were shown in Archae and Bacteria tree, unifying pattern was observed in Eukaryota from the consense tree. Close-up of the eukaryotic tree showed two distinct subgroups each represented by human NUBP1 and NUBP2.
3.1 Prokaryotic MRP proteins might be paralogs to NUBP(1, 2):
MRP proteins are expressed in most Archae and Bacteria for multiple resistance (e.g. salts, ions) and pH adaptations. They are a group of Na+/H+ antiporters that share little similarities with other prokaryotic antiporters. Upon primary energization (e.g. electrical gradient imposed by respiration), they undergo conformational changes that lead to Na+ efflux and H+ influx. (Swartz et al., 2005) Meanwhile, human NUBP2 is expressed in tissues that continually divide, i.e. embryo, brain, ovary and teratocarcinoma, and may regulate the cycle of centrosome replication in concert with NUBP1 and KIFC5A (Christodoulou et al., 2006). By psi-BLAST human NUBP2, we repeatedly found sequences containing MRP-like regions (p-loop). In a word, MRP proteins and NUBP2 share sequence similarity but possess different functions, which indicates they are paralogs that share a common ancestral gene. Nonetheless, how the prokaryotic MRP proteins evolved into such functionally distinct proteins like NBP remains unclear. Further analysis such as looking at point mutations (e.g. gene duplication) and comparing protein 3D structures of the two proteins might reveal its molecular basis.
3.2 MinD converged with NUBP2 to regulate cell division:
Surprisingly, weak (<75%) sequence similarity was found between bacterial MinD and mammlian NUBPs (Nakashima et al., 1999). Also, from our blast and msa results, NUBP2 only showed high sequence similarity to proteins containing the MRP-like domain (p-loop) which is not present in MinD. However, MinD and NUBP2 have been intensively studied on their functional similarity in regulating cell division (Amos et al., 2003). Therefore, we proposed that NUBP2 is the functional 'homolog' to E.coli MinD, but they did not come from the same ancestral gene. Instead, NUBP2 and MinD converged from different ancestors.
3.3 P-loop is a highly conserved structure performing ATPase functions during nucleotide binding:
P-loop folding relies on at least two motifs, S-G-K-G(2)-V-G-K-[ST] and D-X-D-[VFLIM]-X-G. Our ClustalX alignment showed these two motifs are highly conserved across Archae, Bacteria and Eukaryota. Although the molecular events that created p-loop remain unknown, the results strongly argued that p-loop is structurally and functionally preferred by the force of natural selection. Meanwhile, consistent to Nakashima et al. (1999), we determined another two conserved alpha and beta NUBP/MRP, which are likely to associate with the folding of p-loop components.
3.4 Nucleotide binding protein 1 (long form) might appear earlier than nucleotide binding protein 2 (short form) during evolution:
Firstly, based on the psi-BLAST results, nucleotide binding protein 1 was found in all eukaryotes, while nucleotide binding protein 2 was mainly found in yeast and higher eukaryotes. Besides, our phylogenic tree showed high similarities between NUBP1 and NBP from plants and fungi, NUBP2 and NBP from yeast and higher eukaryotes, respecitvely. Thus, NUBP1 group is more widely expressed than NUBP2 group in tissues and across domains. Secondly, by in vitro studies, NUBP1 knockout gave more severe phenotypic defects than NUBP2 knockout, and showed equal phenotypic defects to NUBP1/NUBP2 double knockouts, which indicates a fine-tuning rather than an essential role of NUBP2 function during mammlian cell divisions (Christodoulou et al., 2006). Therefore, we conclude that NUBP1 group exists from the lower to higher eukaryotes, while NBP2 originated from NBP1 and mostly appears in higher eukaryotes. The origination of NBP2 was probably completed by a deletion of the cystein-rich N-terminals from NBP1. How this deletion affected the location, structure and function of NBP1 would be intriguing to know for further studies.