Transcript The presentation part II

Yeast
•A sampling of the
yeast proteome.
Futcher B, Latter GI, Monardo P, McLaughlin
CS, Garrels JI.
•Correlation between protein and
mRNA abundance in yeast.
Gygi SP, Rochon Y, Franza BR, Aebersold R
Objectives
• Gather quantitative data for protein abundance.
=> Create database for yeast proteins.
• Correlation between mRNA level to
corresponding proteins level.
• Correlation between codon bias* and protein
levels.
• Protein expression patterns under various
environmental conditions (i.e. ethanol/glucose).
Motivation
why claculate mRNA and protein correlation?
• Quantitative analysis of global mRNA levels
currently is a preferred method for the analysis of
the state of cells and tissues.
mRNA level <= ? => protein level
• Several methods which either provide absolute
mRNA abundance or relative mRNA levels in
comparative analyses are easy to apply.
* Fast
* Very Sensitive
Why Yeast?
• Low complexity(relative lack of introns),
perfect for lab work, unicellular , well
understood physiology, etc..
• The genome of the yeast was sequenced.
• The number of mRNA molecules for each
expressed gene was recently (1999)
measured. (SAGE*)
• Codon bias tables are well known.
* SAGE – Serial Analysis of Gene Expression
SAGE* – mRNA frequency tables.
1.
2.
3.
Generating a single unique sequence tag (15 bp) of
each mRNA’s 3’-most cutting site for NlaIII of the
Yeast Cell.
Concatenation into a single molecule and then
sequencing, revealing the identity of multiple tags
simultaneously.
Computer software was used to calculate mRNA
abundance, and creating the frequency tables.
* 20,000 transcripts were made.
* Estimated 15,000 mRNA
molecules per cell.
a 1.3-fold coverage even for mRNA
molecules present at a single copy per
cell.
(a 72% probability of detecting single
copy transcripts)
* SAGE – Serial Analysis of Gene Expression
Codon bias
• Definition: A given codon is used more (less)
often to code for an amino acid over different
other codons fot the same a.a.
• Highly biased mRNAs may use only 25 of the
61 codons.
• Different ways to measure C.B exist.
• The larger the codon bias value, the smaller
the number of codons that are used to encode
the protein.
Codon bias - continued
• Use of these codons may
make translation faster or
more efficient and may
decrease misincorporation.
• Codon bias is thought to
be an indicator of protein
expression, with highly
expressed proteins having
large codon bias values.
Experiment Synopsis
• Label all Proteins with [35S] methionines &
cysteines (pulse).
• wait . . .X min (chase).
• Separate Proteins via:
- Centrifugation
- 2D Gels
• Identify (various MS methods and more)
• Quantify Protein Amounts. (use radioactivity)
phosphorimaging, scintillation counting,
autoradiography.
Cells extract in log phase in glucose.
Results present new problems
• 1400 spots were visualized (1200 proteins).
3.1 <pI < 12.8 ; 10kDa < Mr < 470kDa
• Problem: One gel => poor resolution.
Think McFly, Think…
• Solution: Use 3 different gels with different pH ranges.
• Problem: Comigration & coverage – weak spots can be
seen only when they are well separated from strong spots.
• No real solution yet.
Results
• 169 spots representing 148 proteins were
identified using:
peptide sequencing, MS , amino acid
composition and gene overexpression.
• Pulse-chase experiments were made to
determine protein turnover (half lives).
=> all spotted proteins are very stable proteins.
Results – protein quantitation
• Effectively same half life.
=> radioactivity is proportional to protein
abundance.
• The number of methionine and cysteine per
identified protein is known.
=> the number of protein molecules can be
calculated.
Results – some numbers
• Protein abundance range of 300 fold (!).
• Less than a 100 proteins account for half of
the total cellular protein.
Correlation of protein abundance
with mRNA abundance
• mRNA abundance
– SAGE.
– hybridization of cRNA to oligonucleotide arrays.
• Both methods give broadly similar results.
• An adjusted mRNA ratio was calculated
combining the two.
• Elaborate correlation statistics were made.
(Don’t Worry, I will not elaborate today… )
Correlation of protein abundance with adjusted
mRNA abundance.
• Spearman rank correlation
coefficient, rs, was
0.74 (P < 0.0001).
• Pearson correlation
coefficient, rp, on log
transformed data was
0.76 (P < 0.00001).
• A 10-fold range of protein
abundance, f or mRNAs of a
given abundance. (why?)
Correlation of codon bias with protein
abundance
• The rs for CAI versus
protein abundance is
0.80 (P < 0.0001).
(a strong correlation)
• When some abundant
proteins were removed
from consideration,
The rs was essential
unchanged.
Additionl experiments.
• Changes in protein abundance on glucose and
Ethanol were quantified as well.
Gluconeogenesis enzymes more abundant on ethanol.
Heat shock proteins more abundant on ethanol.
Protein synthesis enzymes were more abundant on
glucose.
• Phosphorylation of proteins.
• And more.
Discussion - numbers
• 1200 proteins were quantified.
1/3 – 1/4 of total proteins expressed.
• 148 IDed.
others can be IDed using gene overexpression.
But There is always a (_|_) …
• The remaining proteins will be difficult to see and
study with these methods.
(weak spots are covered by strong spots).
2nd research - Correlation between
protein and mRNA abundance in yeast.
• Similar experiments were made by Gygi et al.
• Similar methods (MS) were used to identify 156
proteins (products of 128 genes).
• Correlation Analysis between mRNA and codon
bias to protein abundance levels were made.
• Genes with missing data were excluded.
×
×
×
×
×
no SAGE data.
ambiguous tags.
no Met’s.
comigration.
pI did not match Mr.
106 genes
Codon bias to protein Correlation.
• No genes were
identified with codon
bias values less than
0.1 even though
thousands of genes exist
in this category.
something’s fishy!?
who said bias?
mRNA protein correlation
rp = 0.93
total:
Lets take
a closer look…
inner set:
rp = 0.35
including progressively more, and higherabundance, proteins in each calculation
Discussion - conclusions
• Codon bias, an indicator of the boundaries of
current 2D gel proteome analysis technology.
• A promising approach is the use of narrow-range
focusing gels.
• Current proteome technology is incapable of
analyzing low-abundance regulatory proteins
without employing an enrichment method.
• For higher eukaryotes the detection of lowabundance proteins would be even harder.
Discussion – words of the wise.
Gygi et al: “This study revealed that
transcript levels provide little predictive
value with respect to the extent of protein
expression.”
Futcher et al: “there is a good correlation
between protein abundance and mRNA
abundance for the proteins that we have
studied”.
Discussion – biases
• Codon Bias.
• Long half lives.
• Low abundance proteins
were not found.
(T.Fs, kinases etc.)
• SAGE data.
• Met’s processed away.
• Comigration.
• Different statistical manipulations.
Why Proteomics – revised





quantity of large scale protein expression.
the subcellular location.
the state of modification.
the association with ligands.
the rate of change with time of such
properties.
 GO HOME ! 