It is a great day today and I am calling it “Albumin Day”, because that is what we are discussing in this post. As mentioned earlier, human serum albumin is the most abundant protein in the plasma. This makes it an ideal carrier of various steroid hormones and therapeutic drugs. Albumin is involved as a carrier protein of therapeutics that are used in treatment of diabetes, cancer, rheumatoid arthritis and infectious diseases. HSA also plays a vital role in the diagnosis of diabetes, breast cancer and rheumatoid arthritis. (Elsadek and Kratz 2012).
A little detail on Albumin’s Structure
When the structure of albumin is looked at closely, what we see is that there are two main sites where a drug can potentially bind namely, site 1 and site 2. These sites are located in separate domains of the molecule’s quaternary structure. It has been observed that site 1 ligands appear to be Hulk-like large heterocyclic compounds with a negative charge present in hydrophobic centre of the molecule, while site 2 preferably accommodates aromatic carboxylic acids with a negative charge centered on the alpha carbon, distant from the hydrophobic region of the molecule..
Albumin in Althotas
The wonders that Althotas Virtual Laboratory can do for testing substrates against HSA site 1 and site 2 are going to be discussed now. We have done an exhaustive training of our software so that it recognizes patterns in the existing substrates that will help in predicting experimental ligands or non-ligands for HSA. The Althotas server uses combined SVM-docking based method to specify site of HSA that the substrate can bind to, if any (Zsila et al., 2011).
Click on the Create new job button, enter the job name, choose SVM predictions, specify the structure you want to use for docking and select the ligands for your experiment. Your results will be uploaded to your dashboard in no time.
Results for substrate prediction for Albumin using the Althotas server against three ligands; Allopurinol, Azapropazone and Diiodosalicylic acid. Only the lowest free energy of docking is shown.
What makes us better than the other available methods for albumin substrate prediction? First of all, our system qualitatively separates the albumin ligands from the non-albumin ligands which is due to the thorough training of the software with 163 example molecules. Secondly, our approach discriminates between site 1 and site 2 ligands which is not done by the existing techniques and thirdly… well I am not revealing the third one just yet. Explore it yourself! Try Althotas!
B. Elsadek, F. Kratz, Impact of albumin on drug delivery — new applications on the horizon, J. Control. Release 157 (2012) 4–28.
Evaluation of drug-human serum albumin binding interactions with support vector machine aided online automated docking (2011) Ferenc Zsila, Zsolt Bikadi, David Malik, Peter Hari, Imre Pechan, Attila Berces, and Eszter Hazai. Bioinformatics. doi:10.1093/bioinformatics/btr284