Universidade de Aveiro-RMN

7/23/2019 Universidade de Aveiro-RMN

http://slidepdf.com/reader/full/universidade-de-aveiro-rmn 1/39

Elucidação Estrutural e Metabolómica na Indústria Farmacêutica

Elucidação Estrutural e

Metabolómica na IndústriaFarmacêutica

Brian J Goodfellow

1




The Pharmaceutical Industry

We have a challenging and inspiring mission: toimprove the quality of human life by enablingpeople to do more, feel better and live longer.

GlaxoSmithKline

As a global biopharmaceuticalcompany, our activities touch many

people’s lives

DreamHealth is our dream. Because we believe thatdreams should be pursued, we are researching anddeveloping new drugs to offer a longer and betterlife. In Portugal and worldwide. Caring for yourHealth. And there are dreams that come true...

2




The Pharmaceutical Industry

$

2




How do we get new drugs ?

What is a drug molecule?

! A drug molecule has one or more functional groups positioned in 3D spaceon a structural framework that enables the molecule to bind to a targetedbiological macromolecule, the receptor

biological response

+“druggable” target“drug-like” molecule

Properties: low Mwt, not toolipophilic, not too hydrophilic, needsfunctional groups

Properties: usually protein, leads tobiological response, does not causetoxicity

3




New Drugs - recent trends

Reasons:! fundamentally difficult

! high failure rate (4% of compounds become a drug and of these 89 % failclinical trials)

! expensive > 800 million dollars and can take 15 years

! cost minimization in industry leads to fewer drugs

! Mainly based on high throughput screening (HTS) or structure based drugdesign (SBDD)

4




http://www.kubinyi.de/lectures.html

New Drugs - SBDD and HTS

! Requires 3D structure of protein target to design drugs to bind to activesite, therefore need rapid methods for structure determination of targets

Chemical Microarrays (Graffinity, Heidelberg)

! ca. 100,000 fragments linked onto chip

! parallel detection of weak, specific interactions

! confirmation of hits by functional testing and X-ray/NMR

! Corporate databases (100,000s of compounds) are screened automaticallyby robots - identify high-affinity binders but produce false leads - need toidentify lead compounds that interact with the target in a biologicallyrelevant mechanism

5




The diversity of NMR enables the techniqueto be applied at multiple stages along the

drug discovery pathway

New Drugs - How can NMR help ?

NMR is being used for:

! protein structure determination

! lead discovery and optimization

! evaluating in vivo selectivity andefficacy

! analyzing drug toxicity profiles andidentifying new drug discoverytargets

6




NMR can:

! rapidly determine protein and protein-ligand structures

! efficiently screen fragment-based libraries to identify biological relevantligand interactions and identify new therapeutic targets

! monitor changes in the metabolome from biofluids and cell lysates toexplore in vivo drug activity


7




Protein structure determination - NMR

C

H

C

H5Å

NOE

1 Å = 1x10-10 m

Input: intercity distance table Output: map of USA

9





C

H

C

H5Å

NOE

1 Å = 1x10-10 m

Input: 1H-1H distance table Output: structure of protein

100 VAL H 100 VAL HB 3.32 #peak 288 #SUP 0.96 #QF 0.96 21 LEU H 21 LEU HG 3.97 #peak 289 #SUP 0.98 #QF 0.93189 LYS H 189 LYS HB2 3.71 #peak 290 #SUP 0.86 #QF 0.86100 VAL H 100 VAL QG2 3.97 #peak 292 #SUP 1.00188 VAL QG1 189 LYS H 4.32 #peak 293 #SUP 0.84 #QF 0.84 45 GLU H 46 VAL H 4.60 #peak 297 #SUP 0.96 #QF 0.96131 TYR QD 189 LYS H 4.55 #peak 300 #SUP 0.79 #QF 0.79 45 GLU H 45 GLU QG 3.96 #peak 321 #SUP 0.98 #QF 0.98 45 GLU H 45 GLU HB2 3.30 #peak 323 #SUP 0.91 #QF 0.91 45 GLU H 45 GLU HB3 3.30 #peak 324 #SUP 0.95 #QF 0.95 71 ASP H 114 ALA H 3.83 #peak 330 #SUP 0.99 #QF 0.99 72 LYS H 114 ALA H 4.37 #peak 337 #SUP 0.82 #QF 0.82138 TYR QD 139 ALA H 4.10 #peak 339 #SUP 0.96 #QF 0.96 32 GLU HB2 33 GLU H 4.52 #peak 467 #SUP 0.94 #QF 0.94 23 THR QG2 33 GLU H 3.83 #peak 470 #SUP 0.31 #QF 0.31 99 LYS HB3 101 TRP HE1 5.13 #peak 482 #SUP 0.44 #QF 0.44 84 PHE HB2 85 ALA H 4.01 #peak 493 #SUP 0.98 #QF 0.98 85 ALA H 101 TRP HD1 4.05 #peak 497 #SUP 0.99 #QF 0.99 85 ALA H 168 TYR QD 4.95 #peak 499 #SUP 0.94 #QF 0.94………..

9




amide

aromatic

methyl

Lehninger, Nelson & Cox, 3rd ed.

aliphatic

! But to obtain a structure we need toidentify every peak

! get a ca. peak from each 1H in theprotein

! but 150 residue protein has ~1000 1Hsignals

! incomplete dispersion of frequencies

! need to separate resonances inanother frequency dimension


many peaks overlapped

10




2D

t 1

!M

1D

FourierTransform

2D FourierTransform

Protein structure determination - 2D NMR

11




Protein structure by NMR - going to nD

! faster structure determination needs higher dimensions (and isotopicenrichment of the protein)

! But a 5D experiment for a 30 sec FID would take (128x32x16x8x4).......around 2 years !

12




a mathematical calculation that converts a table of

distances into a map or structure

Backbone atoms of the final

family of 20 conformers 173-180

Dias et al. J Biol Chem (2006) vol. 281 (42) pp. 31553-31561

Dias et al. J Biomol NMR (2005) vol. 32 (4) pp. 338-338

Protein structure by NMR - example using up to 3D

! p22HBP a 22 kDa protein

! Used 1851 distances and took ca. 1 year fromstart to finish

13




! Robotic can be used for automated production of protein samples for NMRand x-ray

! New NMR pulse sequences shorten the time required to collect a set ofNMR spectra: automated projection spectroscopy (APSY); G-matrix Fouriertransform NMR (GFT-NMR) are two examples

! Dimensionality of traditional multi-dimensional triple resonance experimentsis reduced = shorter acquisition time. Works by joint sampling of multiple

chemical shifts in a single indirect dimension

Protein structure by NMR - speed up data collection

14




! Robotic can be used for automated production of protein samples for NMRand x-ray

! New NMR pulse sequences shorten the time required to collect a set ofNMR spectra: automated projection spectroscopy (APSY); G-matrix Fouriertransform NMR (GFT-NMR) are two examples

! Dimensionality of traditional multi-dimensional triple resonance experimentsis reduced = shorter acquisition time. Works by joint sampling of multiple

chemical shifts in a single indirect dimension

Protein structure by NMR - speed up data collection

14




Protein structure by NMR - Structural genomics

15




NMR can:

! efficiently screen fragment-based libraries to identify biological relevantligand interactions and identify new therapeutic targets


16




Ligand binding by NMR

! where/how does the ligand bind ?

! which ligand of a series binds ?

! NMR techniques for ligand bindinguse the fact that the ligand, whenbound, has properties similar to the

large target or use magnetizationtransfer between target and ligand

17




! Saturation Transfer Difference isb a s e d o n a t r a n s f e r o fmagnetization from the protein tothe bound molecule

[ligand] = 1 mM

[protein] = 35!M

expt time 1-4 hours

Ligand binding by NMR - which ligand(s) binds ?

STD experiment for p38 kinase domain

saturate baseline

saturate protein

difference

18





[ligand] = 0.1-0.5 mM

[protein] = 25 !M

expt time 20 hours

INPHARMA

! LA binds so that HA to HT transferoccurs

! LA dissociates and LB enters! HT to HB transfer now occurs! Result is a NOE peak between HA and

HB mediated by HT

19




! Diffusion Ordered SpectroscopY (DOSY)

DOSY

" 1Hrápido

lento

Desvio químico

Difusão

Cromatografia em camada fina TLC

Polaridade

Edição espectral por coeficiente de difusão


Chromatograpy, TLC

P o l a r i t y

D i f f u s i o

n

fast

slow

chemical shift

Spectra editing by diffusion constantAdapted from EJ Cabrita, Presentation, UA 2008

20




D i f u s ã o

Desvio químico

lento

rápido

D i f u s ã o

lento

rápido

Host (H)

Guest (G)

H + G HG


D i f f u s i o n

D i f f u s i o n

slow

slow

fast

fast

chemical shift Adapted from EJ Cabrita, Presentation, UA 2008

21




! HSQC = Heteronuclear SingleQuantum Correlation

! one peak for each NH gives a“fingerprint” of the protein

! need to label the protein with 15Nand or 13C

! fast experiment 5-20 mins

Ileal bile acid binding protein, 14 kDa

Ligand binding by NMR - where/how does the ligand bind ?

! NMR chemical shift mappingidentifies ligand binding sites

! binding site obtained rapidly andthe process can be automated(robotic sample-changers and flow-probes)

22




[ligand] = 0.5-1 mM

[protein] = 0.5-1 mM

expt time 1-4 hours


23




PPIX-induced shift perturbations

small negative shiftslarge negative shifts

small positive shifts (0.05-0.1ppm)

large positive shifts (>0.1ppm)


hydrophobic interaction

between p22HBP and heme

24



Eluciadçao Estrutural e Metabolomica na Indústria Farmacêutica

Summary of some NMR methods for binding studies

Approach Observe Use Description

Shift mapping target1º screeningHit validationBinding site

Perturbs chem shifts oftarget near binding site

STD ligand1º screeningHit validation

Identifies compounds thatbind weakly.

Bulid up curves identifyFGs

waterLOGSY ligand 1º screeningIdentifies compounds thatbind via water mediated

NOEs

DOSY ligand 1º screeningHit validation

Mesures differences indiffusion rates for ligandsin bound versus free state

SAR by ILOE ligand-to-ligand Fragment based DDLead optimization

Detects protein mediatedligand-ligand interactions(occupying adjacent sites)

INPHARMA ligand-to-ligand Lead characterizationDetects protein mediatedligand-ligand interactions

(competition for samesite)

25




NMR can:

! monitor changes in the metabolome from biofluids and cell lysates toexplore in vivo drug activity


26




! Basically, a 1D 1H NMR spectrum of a lysed cell or biofluid captures itsmetabolic state (NMR only observes the most abundant metabolites)

Metabolomics by NMR

Proteins

Genomics (DNA)25,000 genesTranscriptomics (RNA)100,000 mRNAs

Proteomics (proteins)

1,000,000 proteins

The metabolome reflects the

general cellular state

Metabolomics (metabolites)

2,500 metabolites (small

molecules)

27




Metabolomics by NMR

Metabolomics definition

“the quantitative measurement of the dynamic multiparametric metabolicresponse of living systems to pathophysiological stimuli or geneticmodification” (Imperial College, London, 1999)

Perturbedsystem

- Disease- Xenobiotics- Genetic modification- Diet- Environmental effects

Fingerprint ofbiochemical perturbation

Metabolic adjustmentsto mantain homeostasis

! Changes in metabolic profiles

NMR of:Biofluids (plasma, urine, etc.)TissuesCells

Non-perturbedsystem

28




Metabolomics by NMR - multivariate analysis

! Comparing 1 or 2 spectra by eye is easy - 40 is harder…..! Use multivariate analysis (PCA) to identify differences between spectra

! Always need many spectra to filter out external/experimental variables

expt time = 5 mins x Nº samples

some samples have a lot moretrehaolse than others

other differences harder to see!

29




!"#%"#&"#'"#("##"#)"#*"#+"#,"#

!"#$%&' )* +

! Scores plot identifies grouping of samples - similar spectra group together! Loadings plot identifies differences between grouped spectra

expt time = 5 mins x Nº samples

Metabolomics by NMR - multivariate analysis

!"#"$

!"#""&

!"#""'

!"#""(

!"#"")

"

"#"")

"#""(

"#""'

"#""&

"#"$

!"#"( !"#"* !"#") !"#"$ " "#"$ "#") "#"* "#"(

!"# %&'(

!" ' )*&+(

+,-./,0 123,-4-5670

8** 123,-4-5670

+,-./,0 9.7:,-7/;

8** 9.7:,-7/;

cells frozen in N2(l) then extractedusing chlorofrm/methanol/water,freeze dried and buffer added

30




! The effect of the drug 8-azaxanthine on the target urateoxidase can be seen

! The drug specifically inhibitsurate oxidase

Metabolomics by NMR - in vivo models

cells frozen in N2(l) thenground and buffer added

31




Rat urine

Metabolomics by NMR - in vivo, biofluids

phosphate buffer added to urine

32




The COMET project: 5 pharmaceutical companies + Imperial College (UK)

! Generation of a comprehensive database of NMR spectra of biofluids

treated with various toxicants (147 studies; >30000 samples)

Achievements

! Predictive screening methodologies that both detect abnormal metabolicstates and classify target organ toxicity (mainly liver and kidney)

! Animals (rodents) subjected to treatments generating typical toxicologicallesions and their biofluids analysed by NMR.

Approach

Metabolomics by NMR - in vivo, toxicity

33




! Differences in metaboliteprofiles for Aito Town,Chicago, and Guangxip o p u la t i o n re ve a le ddifferences related togenetic, dietary, and gutmicrobial factors

Metabolomics by NMR - in vivo, know your subject

34




Metabolomics by NMR - disease

! Can distinguish malformations from controls in diffusion edited spectra of AF

! 22 metabolites were found to changesignificantly in the AF of malformed fetuses,compared to controls, together with fourunassigned resonances

!

Results confirmed that malformed fetusessuffer altered energy metabolism and kidneyunderdevelopment

35




MetabolicPhenotype

Nutritionalstatus

Gut microbiota

Disease Co-/pre-administration of

other drugs

P h a r m a c o - m

e t a b o n o m i c s

P h a r m a c

o - g e

n o m i c

s

Genotype

Metabolomics by NMR - the future

! Personalized medicine

36




The diversity of NMR enables the technique to be applied at multiple stagesalong the drug discovery pathway

! lead discovery and optimization helped:

! by rapidly determining protein and protein-ligand structures

! efficiently screening fragment-based libraries to identify biologicalrelevant ligand interactions and new therapeutic targets

! Evaluation of in vivo selectivity and efficacy helped by:

! monitoring changes in the metabolome from biofluids and cell lysates toexplore in vivo drug activity and by analyzing drug toxicity profiles andidentifying new drug discovery targets

Conclusion

! Protein structure (UAveiro, ITQB, FCT-UNL, UCoimbra)

! Ligand binding (UAveiro, ITQB, FCT-UNL, UCoimbra)

! Metabonomics (UAveiro - diseases)

37

Documents

Universidade de Aveiro-RMN