Membrane Protein Roadmap 03/26/09 Slide #1 of 27
Biophysical dissection of ABC Transporter mechanism
John RamosPaul “The Man” SmithNathan KarpowichBo ChenOksana Martsinkovich
Funding: NIH, MOD, CFF
Linda Millen
Jonathan Moody
Phillip ThomasUT Southwestern Physiology
Membrane Protein Roadmap 03/26/09 Slide #2 of 27
ATP-Binding Cassette (ABC) Transporters
ABC domains ABC dimer?
Membrane Protein Roadmap 03/26/09 Slide #3 of 27
F1
F0
www.sanken.osaka-u.ac.jp
ABC domains are “F1-like” ATPases.
Walker A B
Membrane Protein Roadmap 03/26/09 Slide #4 of 27
The “A-Protein” paradigmfor (F1-like) mechanical ATPases
• ATP binds at a domain-domain interface.
• The flanking domains act as an ATP-dependent mechanical clamp.
• ATP encapsulation in the interfacial active site drives the mechanochemical “power-stroke” which is closure of the clamp formed by the flanking domains.
• Non-hydrolyzable analogues typically have much lower binding affinity than ATP … and therefore often fail to drive the powerstroke of ATPase motors.
• Same principles apply to at least some non-F1-like mechanical ATPases (definitely GroEL, DnaK/Hsp70, perhaps even myosin).
Mg++
++
+++
--
• Therefore, ATP binding (NOT HYDROLYSIS) drives the mechanochemical powerstroke.Walker A B
Membrane Protein Roadmap 03/26/09 Slide #5 of 27
ATP-Binding Cassette (ABC) Transporters
ABC domains ABC dimer?
Membrane Protein Roadmap 03/26/09 Slide #6 of 27
But the ATPase active site was not located at an interdomain interface in 5 different ABC crystal structures…
No consistent pattern of oligomerization of ABC domains!
?Completely solvent-exposed ATPase active site --
no where near an interdomain interface!
Membrane Protein Roadmap 03/26/09 Slide #7 of 27
Concluded that the problem was the inability to observe the inherently transient complex with ATP
E171
E171Q
(combined with the fact that AMP-PNP and ATP--S are lousy analogs)
Membrane Protein Roadmap 03/26/09 Slide #8 of 27
Using enzymological subterfuge to block ATP hydrolysis yields hyper-stable ABC dimerization!
MJ07962•ATP2: Rfree= 25.1% @ 1.9 Å
E171
E171Q
Membrane Protein Roadmap 03/26/09 Slide #9 of 27
The -helical subdomain rotates away from the core in the absence of the -phosphate of ATP
• Up to 20˚ rotation of -helical subdomain observed in some non-ATP-form ABC domain structures.
Membrane Protein Roadmap 03/26/09 Slide #10 of 27
ABC motor domain mechanism is fairly well understood … but how do the associated TM domains drive transport?
MJ07962•ATP2: Rfree= 25.1% @ 1.9 Å
E171
E171Q
Membrane Protein Roadmap 03/26/09 Slide #11 of 27
Divergent transmembrane domain structures within the ABC Transporter superfamily
Membrane Protein Roadmap 03/26/09 Slide #12 of 27
FRET … to vitamin B12!
B12 ABSORBANCE
- - - 488 Alexa Fluor Excitation___ 488 Alexa Fluor Emission
- - - 546 Alexa Fluor Excitation___ 546 Alexa Fluor Emission
Wavelength (nm)
BtuCD-Ftransports
vitamin B12(structures
from Locher, Rees, et al.)
Membrane Protein Roadmap 03/26/09 Slide #13 of 27
Purification of BtuCD & Alexa-Fluor546-labeled BtuF*
BtuF+CysBtuCD
E142QCD
29kD
Fluorescecnescan
Coomasiestain
A C-terminal cys engineered into BtuF
Membrane Protein Roadmap 03/26/09 Slide #14 of 27
FRET provides a ruler measuring B12 movement relative to a fixed point in the transporter
Membrane Protein Roadmap 03/26/09 Slide #15 of 27
Anisotropy can monitor molecular associationalthough it also (weakly) influenced by quenching/lifetime effects
Perrin Equation:
1/r = (1/r0)(1 + (/rot) = (1/r) (1 + (RT•/V) )
Membrane Protein Roadmap 03/26/09 Slide #16 of 27
Avoid topological problems by using bicelles
QuickTime™ and a decompressor
are needed to see this picture.
Monitoring Bicelle formation usingLight Scattering (Ex.:297-Em.:315nm)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 200
5.0×10 4
1.0×10 5
1.5×10 5bicellerange~1.0-2.5
SolublizedRegion
Long to Short Chain Lipid Mass Ratio(diC6PC/3Polar-1PC)
(0Intensity
o)Detector
Membrane Protein Roadmap 03/26/09 Slide #17 of 27
Model for the conformational reaction cycle of BtuCD(partially, but not fully, validated)
Membrane Protein Roadmap 03/26/09 Slide #18 of 27
FRET / anisotropy provide rich information concerning transport reaction mechanism
Membrane Protein Roadmap 03/26/09 Slide #19 of 27
Model for the conformational reaction cycle of BtuCD(partially, but not fully, validated)
Membrane Protein Roadmap 03/26/09 Slide #20 of 27
The E-to-Q active-site mutation (E142Q) in the active traps ATP in BtuD (just like in the isolated MJ0796 ABC domain)
Membrane Protein Roadmap 03/26/09 Slide #21 of 27
BtuF binds to BtuCD with higher affinity whenATP is locked in the active site
Membrane Protein Roadmap 03/26/09 Slide #22 of 27
Model for the conformational reaction cycle of BtuCD(partially, but not fully, validated)
Membrane Protein Roadmap 03/26/09 Slide #23 of 27
The E142Q mutation in BtuD traps B12 in BtuCD just like the wild-type transporter -- but does not release it
Membrane Protein Roadmap 03/26/09 Slide #24 of 27
Model for the conformational reaction cycle of BtuCD(partially, but not fully, validated)
Membrane Protein Roadmap 03/26/09 Slide #25 of 27
A non-BtuCD-interacting mutant variant of BtuF (E50R/E180R) can monitor free B12 concentration
Membrane Protein Roadmap 03/26/09 Slide #26 of 27
Model for the conformational reaction cycle of BtuCD(partially, but not fully, validated)
Membrane Protein Roadmap 03/26/09 Slide #27 of 27
Correspondence of crystal to functional states is obscure -- except for E-to-Q MalEFG-K
Membrane Protein Roadmap 03/26/09 Slide #28 of 27
Biophysical dissection of ABC Transporter mechanism
John RamosPaul “The Man” SmithNathan KarpowichBo ChenOksana Martsinkovich
Funding: NIH, MOD, CFF
Linda Millen
Jonathan Moody
Phillip ThomasUT Southwestern Physiology
Membrane Protein Roadmap 03/26/09 Slide #29 of 27
Update on E. coli IMP overexpression physiology project
• Expressed 5 IMP’s (4 ABC Transporters, 1 MFS) in MG1655 via pQE60/pRep4 plasmids; compare 2 soluble proteins (enolase & the cytoplasmic domain of one of the ABC Transporters).
• Characterize cell growth rates, morphology +/- membrane stains, transcriptome (via microarray), expression of selected reporter genes (for factors), and protein expression level / physical state.
Funding: NIH R21!
• Expressing cells suffer from “gigantism” -- when expressing either soluble or membrane proteins.
• IMP expressing cells are growth-inhibited -- & toxicity level correlates with amount of detergent solulizable IMP produced.
• Evidence of intracellular lipid-rich inclusions in overexpressing cells, whether or IMP is recoverable.
• No activation of s, E, or Cpx systems (or s32).• ~200 genes reduced in expression -- most shared by soluble proteins &
IMPs, many annotated to be related to acid stress response.• ~50 genes increased in expression -- most specific to IMPs, 45 part of
flagellar biosynthesis / chemotaxis regulon.• ~5 overexpressed genes may be specific for IMP overexpression.
Observations (conclusions?):