Transcript No Slide Title

A Genetic Approach to Analyzing
Membrane Protein Topology
Colin Manoil and Jon Beckwith
Science, Vol. 233, 1403-1408, September 26, 1986
Question:
After reading this article we decided to create some of our very own alkaline phosphatase
fusions to investigate the topology of a “fictional” membrane protein we have named,
BADH, which we discovered recently from an “unique” bacterium known as B. anseli.
Our new protein, like the E. coli Tsr protein, seems to be involved in chemotaxis as a
chemoreceptor and so we hypothesize that it may closely resemble Tsr in its membrane
protein topology as well.
After careful screening of various fusions we identified one, named pBA100, with high
alkaline phosphatase activity of 330 units/OD600 compared to other fusions with varying
activities of between 5-9 units/OD600.
Based on the results of the “Beckwith and Manoil” paper, where on the schematic diagram
below of the BADH protein, would you expect that the TnphoA inserted itself onto the
BADH gene to produce the fusion pBA100 (use arrows to indicate)?
Answer:
Background
• The study of the three-dimensional structure of proteins is a key to
unlocking the mystery of their interactions in vivo
• One of the most informative methods for discerning protein
structure is x-ray crystallography
• If a protein can be crystallized, then the diffraction of x-rays by the
crystals can be used to determine the position of every atom in the
molecule
• Membrane proteins are difficult to crystallize and structurally
analyze
• Genetic approaches to analyzing membrane protein structure offer
a useful alternative to traditional diffraction studies
Introduction Part I
• Integral membrane proteins have easily identified long contiguous stretches of
hydrophobic amino acids
• Structural analyses of photosynthetic reaction center polypeptides have shown
that such long hydrophobic sequences correspond to transmembrane alpha-helical
stretches of a membrane protein (J. Mol. Biol. 163, 451 (1983))
• Using the plot of average hydrophobicity along the sequence of a membrane
protein, one can determine possible 2-D membrane topologies for the polypeptide
(J. Mol. Biol. 157, 105 (1982))
• Previous methods of studying membrane protein topology included:
1. Studying or identifying sites that interact with proteins of known
cellular location (Methods Enzymo. 98, 91 (1983))
2. Reaction with small molecules, proteases or antibodies added from one
side or other of the membrane (Methods Enzymo. 125, 453 (1986))
3. NMR spectroscopy of purified membrane proteins (J. Biol. Chem. 249,
8019 (1974))
Introduction Part II
• In this paper the authors propose a gene fusion approach to
studying membrane protein toplogy
• The chemorececptor protein Tsr, in E. coli, has a tetrameric
transmembrane structure, organized in such a way that each
polypeptide chain has a periplasmic domain between two
transmembrane sequences, and a large cytoplasmic domain at the
carboxyl terminus of the protein
• Alkaline phosphatase needs to be exported to the periplasm in
order to show enzymatic activity (J. Bacteriol. 154, 366 (1983))
• For this study, fusions of alkaline phosphatase to Tsr protein were
made by random insertion of transposon TnphoA into a plasmid
carrying the tsr gene and screened on media containing the alkaline
phosphatase indicator, 5-bromo-4chloro-3-indolyl phosphate
Hypothesis
The alkaline phosphatase activity of genetically engineered
fusions at different positions on the E. coli Tsr protein
should reflect the normal membrane topology of the
protein.
Figure 1- Scheme for using alkaline phosphatase fusions to
identify membrane protein topology
Table 1- Properties of tsr-phoA fusion plasmids
Figure 2-Fusions of alkaline phosphatase to chemoreceptor
proteins
Fusions to a Tsr Protein
Deletion Mutant
• Analyzed fusions of alkaline phosphatase to the
cytoplasmic end of a mutant Tsr protein lacking
second transmembrane domain (tsrΔ1)
• Without this domain, normally cytoplasmic regions
might pass into the periplasm
• pCM234, pCM235 have high enzymatic activity
Figure 3-Fusions of Alkaline Phosphatase to a Deletion
Derivative of tsr Protein
Fusions to a Tsr Protein
Deletion Mutant (cont.)
• Used Tnpho1 to insert the wild-type second
transmembrane sequence back into the plasmids that
express fusion proteins with high AP activity
• These transmembrane seuqences are expected to
translocate alkaline phosphatase back into the
cytoplasm, and the resulting fusion plasmids should
show low activity
• pCM251, pCM252
Table 1- Properties of tsr-phoA fusion plasmids.
Activation of a Low Activity
Fusion Protein
• Hypothesized that any manipulation involving the
deletion of the second transmembrane sequence
increases activity
• pCM211
• Expect plasmids without RV1-RV3 and RV2-RV3
sequences to show high activity
Figure 4-Activation of the Alkaline Phosphatase in a
Cytoplasmic Domain Hybrid Protein
Activation of a Low Activity
Fusion Protein (cont.)
•
Most transformed colonies were Pho-
•
Examined Pho- and Pho+ colonies through restriction
analysis
•
Pho- colonies had precise RV1-RV3 deletions (ex.
Plasmid pw1)
•
Pho+ colonies had:
1) RV2-RV3 deletions (ex. plasmid pb1)
2) RV1-RV3 deletions with a small loss of DNA (ex.
plasmid pb4)
Sequence of tsr Protein in This Study
Differs From Published Sequence
• Precise RV1-RV3 deletions lead to TnphoA sequence
being out of frame and low enzymatic activity
• Comparison of the sequences of pb1 and pb4 with
pw1 identified a difference in the translational reading
frame of EcoRV1 between the sequence of this plasmid
and the published tsr sequence
• Any precise deletions involving RV1 would put the
phoA DNA out of frame and lead to low AP activity
Fractionation of tsr-phoA Hybrid Proteins
Grow cells with
labeled amino acids
Osmotically shock
cells and centrifuge
Supernatant
Periplasmic fraction
Resuspend pellet and
incubate cells with lysozyme
Centrifuge
Supernatant
Pellet
Cytoplasmic fraction
Membrane fraction
Fractions analyzed via SDS gel electrophoresis
Fractionation of tsr-phoA Hybrid Proteins
• β-lactamase – alkaline phosphatase hybrids secreted
to periplasm give fragments the size of alkaline
phosphatase1
• Full-length hybrid proteins fractionated primarily
with the membrane
• Hybrid proteins degraded to size of alkaline
phosphatase in periplasmic fractions (ex. pCM203 and
pCM235), consistent with a periplasmic location of
alkaline phosphatase in these fusions
1 Proc.
Natl. Acad. Sci. U.S.A. 82, 5107 (1985)
Table 2-Fractionation of tsr-phoA Hybrid Proteins
Potential Problems
• Potential problems with this approach identified by the
authors:
• The protein sequence carboxyl terminal to the fusion
junction must be nonessential to protein localization
• Alkaline phosphatase must not dominate localization of the
hybrid
• Need to combine traditional crystallographic techniques with
this approach
• Can’t apply this technique if alkaline phosphatase is toxic to
the host cell
• Need to show a negative control for alkaline phosphatase
activity
Conclusion
• The transmembrane sequences of tsr are responsible
for how the protein inserts into the membrane
• In-frame fusion proteins lacking the second tsr
transmembrane domain show decreased alkaline
phosphatase activity
• Replacement of the second transmembrane domain,
recovers the alkaline phosphatase activity
• TnphoA insertion into the periplasmic domain of Tsr
protein results in high alkaline phosphatase activity whereas
insertion into the cytoplasmic domain results in low alkaline
phosphatase activity
Current Trends
Topology prediction
PSORT - Prediction of protein subcellular localization
TargetP - Prediction of subcellular location
DAS - Prediction of transmembrane regions in prokaryotes using the Dense Alignment Surface method (Stockholm
University)
HMMTOP - Prediction of transmembrane helices and topology of proteins (Hungarian Academy of Sciences)
PredictProtein - Prediction of transmembrane helix location and topology (Columbia University)
SOSUI - Prediction of transmembrane regions (TUAT; Tokyo Univ. of Agriculture & Technology)
TMAP - Transmembrane detection based on multiple sequence alignment (Karolinska Institut; Sweden)
TMHMM - Prediction of transmembrane helices in proteins (CBS; Denmark)
TMpred - Prediction of transmembrane regions and protein orientation (EMBnet-CH)
TopPred 2 - Topology prediction of membrane proteins (Stockholm University)
Protein Secondary
& Tertiary
Structure
Predictions Online
3-D Model of Tsr
For more fun with proteins, visit http://us.expasy.org/!