Raunak Shrestha

Source: Mészáros B, Tóth J, Vértessy BG, Dosztányi Z, Simon I. Proteins with complex architecture as potential targets for drug design: a case study of Mycobacterium tuberculosis. PLoS Comput Biol. 2011 Jul;7(7):e1002118.

29th September 2011

A brief overview

• Tuberculosis (TB) – a high burden disease

• Causative agent – Mycobacterium tuberculosis (MTB)

• TB: Ancient disease

• co-habitation of MTB and humans : ~ 9000 years

• 1950s – 1970s : Massive use of Antibiotics

• 1980s – significant reduction in TB

• 1990s : Re-emergence of TB cases • Multi Drug-Resistant TB (MDR-TB)

• Extensively Drug-Resistant TB (XDR-TB)

• STRONG NEED OF NEW DRUG TARGETS AGAINST MTB

Evolutionary Approach in Drug Target Screening

Human Proteins

Human Micro-floras/Eukaryotes

Mycobacteria

Tuberculous Non - Tuberculous

MTB Other MTB-complex Mycobacteria

Comparative sequence analysis of the MTB proteome

Domain composition

• Domains: evolutionary building blocks of proteins

• Pfam database (http://pfam.sanger.ac.uk/) used to determine domain composition in MTB• 3948 MTB proteins had 2099 different domains

• Proteome dataset:• 1,904,578 protein sequences from 467 (20 eukaryotic including

human and 447 bacterial) known complete proteomes

• PSI-BLAST to search for sequence similarities of the domains in MTB proteome with the Proteome dataset from 467 organisms.• Significant hits : E-values < 10E-4

Domain composition

Figure : Occurrences of domains of M. tuberculosis in other organisms.

Pfam domains only cover about 63% of all

residues

Complex Domain Architecture

Domain 1 Domain 2

Disordered regions : domain linkers

disorder-to-order transition

Other protein or molecule

Disordered protein : high free energy

Ordered protein : stable conformation

Disordered Proteins

• IUPred (http://iupred.enzim.hu/): prediction of protein disorder

• Scores between 0 to 1 for every amino-acid residues for its tendency towards disorderness

• >= 0.5 : more disordered

• < less disordered

• Disordered proteins have distinct sequence properties

• ANCHOR (http://anchor.enzim.hu/): recognizes specific sequence properties of the disordered proteins that can undergo disorder-to-order transition

• Compositional Bias

• Low complexity regions

Disordered Proteins

• 11.8 % residues – disordered segments

• 5.7% residues - disordered binding region

• Relatively high values for the disordered proteins as compared to most bacteria

• 7.2% residues in disordered protein overlap with Annotated Pfam domains

• 28% of the protein residues in MTB not covered by Pfam or by disordered and disordered binding regions

• MTB specific proteins?

• Limitations of current methods !!!

Categorization of MTB proteins based on their specificity and function.

• large-scale sequence similarity search for all proteins in MTB with wide-range of proteins from other organism

• Categorization of proteins at various evolutionary levels

• Proteins specific to MTB or highly similar with M. bovis

• Proteins at the level of Mycobacteria

• Proteins found in other bacteria

• Proteins that are ubiquitous from mycobacteria to eukaryotes

• Functional grouping of proteins

• TubercuList (http://genolist.pasteur.fr/TubercuList/)

Categorization of MTB proteins based on their specificity and function.

1. virulence, detoxification, adaptation2. lipid metabolism3. Information pathways4. cell wall and cell processes5. insertion sequences and phages;

6. PE/PPE7. intermediary metabolism and respiration8. regulatory proteins9. conserved hypotheticals

Similarity based clustering of MTB proteins• enrichment and

depletion of protein segments in the MTB proteome across different species.

Figure: Clusters of MTB proteins based on local protein similarities. Hierarchical tree representing the clustering of the 3,948 MTB proteins

Similarity based clustering of MTB proteins

Similarity based clustering of MTB proteins

Similarity based clustering of MTB proteins

pkn family

PE/PPE family

Conclusion

• pkn and PE/PPE, that showed unusual species-specific enrichment of domains

• Probable candidate for future deeper level drug target screening and drug design

Limitations

• Inclusion of only MTB-H37Rv strain (well characterized)

• Failure to include other strains like “Beijing Genotype” which is widely reported MTB genotype strain