1. ©CMBI 2003 MUTANT DESIGN BIO- INFORMATICS QUESTION ‘MOLECULAR BIOLOGY’ BIOPHYSICS

2. ©CMBI 2003 AbstractProtein folding, structure, stability AppliedProcess optimization MUTANT DESIGN BIO- INFORMATICS QUESTION ‘MOLECULAR BIOLOGY’ BIOPHYSICS

3. ©CMBI 2003 MUTANT DESIGN Three strong warnings and disclaimers: 1.I know nothing about MAKING mutants 2.Most times ‘evolutionary’ (that is grant- writing terminology for smart trial-and- error) beat design approaches. 3.Mutants are not always the best way to answer questions. Often good old- fashioned protein chemistry, spectroscopy, or even literature searches get you the answer more quickly.

4. ©CMBI 2003 WHY MUTATIONS 1.Understand protein folding, structure, stability (against many different things); 2.Atomic model validation (homology models, drug binding), or abstract model validation (functional hypotheses); 3.Disrupting interactions, or make them permanent; 4.Protein activity is very hard to engineer; 5.Support for structure determination, e.g. Selenomethionine for SAD or MAD, Cysteine for heavy-metal binding, solubility for NMR; introduce fluorophore; 6.Humanization (normally more than just mutations); 7.Delete, or sometimes add post-translational modifications; 8.Purification tags, e.g. his-tag, flag-tag (not really mutations); 9.Temperature sensitive mutants; 10.Alanine or cysteine scan, or variants thereof; 11.‘Mutate away’ metal binding sites; Many mutations belong in more than one category…..

5. ©CMBI 2003 PROTEIN STRUCTURE helix strand turn Alanine 1.42 0.83 0.66 Arginine 0.98 0.93 0.95 Aspartic Acid 1.01 0.54 1.46 Asparagine 0.67 0.89 1.56 Cysteine 0.70 1.19 1.19 Glutamic Acid 1.39 1.17 0.74 Glutamine 1.11 1.10 0.98 Glycine 0.57 0.75 1.56 Histidine 1.00 0.87 0.95 Isoleucine 1.08 1.60 0.47 Leucine 1.41 1.30 0.59 Lysine 1.14 0.74 1.01 Methionine 1.45 1.05 0.60 Phenylalanine 1.13 1.38 0.60 Proline 0.57 0.55 1.52 Serine 0.77 0.75 1.43 Threonine 0.83 1.19 0.96 Tryptophan 1.08 1.37 0.96 Tyrosine 0.69 1.47 1.14 Valine 1.06 1.70 0.50 AbstractApplied

6. ©CMBI 2003 PROTEIN STABILITY Δ G = Δ H - T Δ S Δ G = -RT ln(K) K = [Folded] / [Unfolded] So, you can interfere either with the folded, or with the unfolded form. Choosing between Δ H and Δ S will be much more difficult, because Δ G is a property of the complete system, including H 2 O….

7. ©CMBI 2003 PROTEIN STABILITY Hydrophobic packing Helix capping Loop transplants

8. ©CMBI 2003 PROTEIN STABILITY A whole series of tricks can be applied: Gly -> Any; Any -> Pro; Introduce hydrogen bonds; Hydrophobic packing; Cys-Cys bridges; Salt bridges; β-branched residues in β- strands; Pestering water from the core; etc. The main thing is that you should first know WHY the protein is unstable. Abstract: F UApplied: F LU I

9. ©CMBI 2003 MUTATIONS ‘SHOULD’ ADD UP

10. ©CMBI 2003 BUT THEY DON’T….

11. ©CMBI 2003 LOCAL UNFOLDING

12. ©CMBI 2003 WEAK SPOTS IN PROTEINS

13. ©CMBI 2003 WEAK SPOT PROTECTION

14. ©CMBI 2003 SUPPORT FOR EXPERIMENTS 1.Selenomethionine for Xray; 2.Solubility (i.e. for NMR); 3.Tags for purification (His-tag, Flag-tag, etc); 4.Addition or removal of post-translational modification sites; 5.‘Mutate away’ metal binding sites; 6.Introduce fluorophore; 7.Block binding, or make binding irreversible; 8.Etcetera.

15. ©CMBI 2003 PREDICT MUTATIONS FROM ALIGNMENTS It is rapidly becoming apparent that multiple sequence alignments are the most powerful tool in bioinformatics. And that is also true for mutation design. If you can predict something that nature has done already, success is almost guaranteed.

16. ©CMBI 2003 CONSERVED, VARIABLE, OR IN-BETWEEN QWERTYASDFGRGH QWERTYASDTHRPM QWERTNMKDFGRKC QWERTNMKDTHRVW Gray = conserved Black = variable Green = correlated mutations

17. ©CMBI 2003 CORRELATED MUTATIONS SHAPE TREE 1 AGASDFDFGHKM 2 AGASDFDFRRRL 3 AGLPDFMNGHSI 4 AGLPDFMNRRRV

18. ©CMBI 2003 CORRELATION = INFORMATION 1, 2 and 5 bind calcium; 3 and 4 don’t. Which residues bind calcium? 1 ASDFNTDEKLRTTYI Ca+ 2 ASDFSTDEKLKTTYI Ca+ 3 LSFFTTDTKLATIYI 4 LSHFLTDLKLATIYI 5 ASDFTTDEKLALTYI Ca+