We have first started by looking at similar proteins in the Databank. We started with a Blast Search against all PDB proteins. We found two main homologs. The Sulfide: Quinone Reductase of Acidothiobacillus ferrooxidans and that of the hyperthermophile Aquifex Aeolicus.
We have then used MODELLER Software to align the protein from A. Aeolicus as the template to our target sequence from Leptolyngbya Hensonii. (please see data in the supporting information). Upon comparing the DOPE (Discrete Optimized Protein Energy) profiles, we noticed that there are regions of clear discrepancies. We have chosen the best five models to compare, this is the comparison for the best model we have.
Figure 2. A. Aeolicus and L. Hensonii comparison of DOPE profiles.
The energy of the model is a lot higher especially in the region between 200 and 250. This region is not far from the FAD-binding site (the site of electron transfer). Moreover, the sequence responsible for anchoring the A. Aeolicus into the membrane is very different in our case. Based on the homology model, there should be a turn in this region of 376 to 412. However, the turn is incorrectly predicted as there is no PROLINE GLY sequence found in our target protein. Instead, a VAL GLY TRP sequence is found. However, sites for the binding of S8 and the binding of Quinone are the same.
Figure 3. A turn in a sequence responsible for anchoring the A. Aeoulicus into the membrane.
We have then decided to try the use of A. ferrooxidans SQR as a template. While both A. Aeolicus and A. ferrooxidans share an identity of 42%. In the case of A. ferrooxidans, the coverage of the query is 5 percent higher, leading us to believe that the template might be better. The DOPE profiles are compared below.
Figure 4. A. Aeolicus and A. ferrooxidans comparison of DOPE profiles.
A first glance shows that the comparison is better. However, regions between 210 to 300 are still problematic. As in the case with the A. Aeolicus, the CYS disulfide bridges appear not conserved between the models and the template. A very important note is that our sequence does not share the C-terminal sequences known for anchoring the SQR of A. Aeolicus and A. ferrooxidans into the membrane. It is possible that our target SQR associates with the membrane through another protein. In that case, we have decided to add an anchoring sequence to the protein for expression in Cyanobacteria. As with the A. Aeolicus we were able to find the Cysteine residues responsible for electron transport in the active site. The three Cysteine residues responsible for electron transport are CYS 127, CYS 159, and CYS 345.
Figure 5. Cysteine residues responsible for electron transport in the active site of A. Aeolicus.
In A. ferrooxidans, the amino equivalent amino acids are CYS 128, CYS 160, CYS 356.
Figure 6. Cysteine residues responsible for electron transport in the active site of A. ferrooxidans
A. ferrooxidans SQR showing the Pentasulfur intermediate in the active site homologous to that of SQR from L. Hensonii . Unlike A. Aeolicus, there are no disulfide bridges in neither SQR of A. ferrooxidans nor in SQR of L. Hensonii. That is expected as disulfide bridges in A. Aeolicus are associated with a hyperthermophilic activity of SQR in that species .
Figure 7. Protein structure of SQR from L. hensonii.
The image was generated using Visual Molecular Dynamics (VMD) software using the Internal Rendering Tachyon algorithm of the whole protein structure of SQR from L. hensonii based on the template from A. ferrooxidans.
1. Cherney, M. M., Zhang, Y., Solomonson, M., Weiner, J. H., & James, M. N. (2010). Crystal structure of sulfide: quinone oxidoreductase from Acidithiobacillus ferrooxidans: insights into sulfidotrophic respiration and detoxification. Journal of molecular biology, 398(2), 292-305.
2. Marcia, M., Ermler, U., Peng, G., & Michel, H. (2009). The structure of Aquifex aeolicus sulfide: quinone oxidoreductase, a basis to understand sulfide detoxification and respiration. Proceedings of the national academy of sciences, 106(24), 9625-9630.