Yeast

1
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

2
Objectives

• Gather quantitative data for protein abundance.
• gt 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).

3
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 lt ? gt protein level
• Several methods which either provide absolute
  mRNA abundance or relative mRNA levels in
  comparative analyses are easy to apply.
• Fast Very Sensitive

4
But But But

• We worked so hard on micro arrays

5
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
6
SAGE mRNA frequency tables.

• Generating a single unique sequence tag (15 bp)
  of each mRNAs 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.
  

a 1.3-fold coverage even for mRNA molecules
present at a single copy per cell. (a 72
probability of detecting single copy transcripts)
20,000 transcripts were made. Estimated
15,000 mRNA molecules per cell.
SAGE Serial Analysis of Gene Expression
7
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.

8
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.

9
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.

10
Cells extract in log phase in glucose.
11
Results present new problems

• 1400 spots were visualized (1200 proteins).
• 3.1 ltpI lt 12.8 10kDa lt Mr lt 470kDa
• Problem One gel gt 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.

12
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).
• gt all spotted proteins are very stable
  proteins.

13
Results protein quantitation

• Effectively same half life.
• gt radioactivity is proportional to protein
  abundance.
• The number of methionine and cysteine per
  identified protein is known.
• gt the number of protein molecules can be
  calculated.

14
Results some numbers

• Protein abundance range of 300 fold (!).
• Less than a 100 proteins account for half of the
  total cellular protein.

15
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.
• (Dont Worry, I will not elaborate today )

16
Correlation of protein abundance with adjusted
mRNA abundance.

• Spearman rank correlation coefficient, rs, was
  0.74 (P lt 0.0001).
• Pearson correlation coefficient, rp, on log
  transformed data was 0.76 (P lt 0.00001).
• A 10-fold range of protein abundance, f or mRNAs
  of a given abundance. (why?)

17
Correlation of codon bias with protein abundance

• The rs for CAI versus protein abundance is
  0.80 (P lt 0.0001).
• (a strong correlation)

• When some abundant proteins were removed from
  consideration, The rs was essential unchanged.

18
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.

19
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).

20
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 Mets.
• comigration.
• pI did not match Mr.

106 genes
21
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.
• somethings fishy!?
• who said bias?

22
mRNA protein correlation

• rp 0.93

total
Lets take a closer look
23
including progressively more, and
higher-abundance, proteins in each calculation
24
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
  low-abundance proteins would be even harder.

25
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.
26
Discussion biases

• Codon Bias.
• Long half lives.
• Low abundance proteins were not found.
• (T.Fs, kinases etc.)

• SAGE data.
• Mets processed away.
• Comigration.

• Different statistical manipulations.

27
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 ! ?