Uses of microarrays and related methodologies in

animal breeding

Bruce Walsh, jbwalsh@u.arizona.eduUniversity of Arizona

(Depts. of Ecology & Evolutionary Biology, Molecular & Cellular Biology, Plant Sciences, Animal Sciences, and Epidemology &

Biostatistics)

QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture.

The basic idea behind gene expression arrays

• With a complete (or partial) genome sequence in hand, one can array sequences from genes of interest on small chip, glass slide, or a membrane

• mRNA is extracted from cells of interest and hybridized to the array

• Genes showing different levels of mRNA can be detected

Types of microarrays• Synthetic oligonucleotide arrays

– Chemically synthesize oligonucleotide sequences directly on slide/chip/membrane (e.g., using photolithography)

– Affymetrix, Agilent

• Spotted cDNA arrays– PCR products from clones of genes of interest

are spotted on a glass slide using a robot– Extracted cellular mRNA is reverse-

transcribed into cDNAs for hybridization

Cell type 1 Cell type 2

Extract mRNA

Label mRNA with redfluorescent dye (Cy5)

Label mRNA with Greenfluorescent dye (Cy3)

mRNA mainlyfrom Cell Type1

mRNA equal mixfrom cell Types1 and 2

mRNA mainlyfrom Cell Type2

Hybridize mRNA to array

Each spot (or feature)corresponds to a differentgene

The color of the spotcorresponds to the relative concentrationsof mRNAs for that genein the two cell types

CellType 1

Cell type 2

mRNAs for thesegenes more abundantin cell type 2

mRNAs from thesegenes more abundantin cell type 1

mRNAs from theseGenes of roughly equalAbundance in both celltypes

Analysis of microarray data

• Image processing and normalization• Detecting significant changes in

expression• Clustering and classification

– Clustering: detecting groups of co-expressed genes

– Classification: finding those genes at which changes in mRNA expression level predict phenotype

Significance testing-- GLM

Yklijk = u + Ak +Rkl + Ti + Gj + TGij +elkijk

Array kReplicate l in array k

Treatment iGene jInteraction betweengene i and treatment j

k-th spotting of gene j undertreatment i on replicate l of arrayk

Problem of very many tests (genes) vs. few actual data

vectors• Expectation: A large number of the GxT

interactions will be significant– Controlling experiment-wide p value is very

overly conservative (further, tests may be strongly correlated)

• Generating a reduced set of genes for future consideration (data mining)– FDR (false discovery rate)– PFP (proportion of false positives)– Empirical Bayes approaches

Which loci control array-detected changes in mRNA expression?

• Cis-acting factors– Control regions immediately adjacent to the

gene

• Trans-acting factors– Diffusable factors unlinked (or loosely

linked) to the gene of interest

• Global (Master) regulators– Trans-acting factors that influence a large

number of genes

David Treadgill’s (UNC) mouse experiment

• Recombinant Inbred lines from a cross of DBA/2J and C57BL

• The level of mRNA expression (measured by array analysis) is treated as a quantitative trait and QTL analysis performed for each gene in the array

Genomic location of mRNA level modifiers

CIS-modifiersTRANS-modifiers MASTER modifiers

Gen

om

ic locati

on

of

gen

es o

n a

rray

Distributionof >12,000

geneinteractions

Candidate loci : Differences in Gene Expression between

lines• Correlate differences in levels of

expression with trait levels• Map factors underlying changes in

expression– These are (very) often trans-acting factors

• Difference between structural alleles and regulatory alleles

Expanded selection opportunities offered by

microarrays• G x E

– Candidate genes may be suggested by examining levels of mRNA expression over different major environments

– With candidates in hand, potential for selection of genes showing reduced variance in expression over critical environments

• Breaking (or at least reducing) potentially deleterious genetic correlations– Look for variation in genes that have little (if any)

trans-acting effects on other genes

Towards the future

• Selection decisions using information on gene networks / pathways

• Microarrays are one tool for reconstructing gene networks

• Tools for examining protein-protein interactions– Two hybrid screens– FRET & FRAP

Analysis and Exploitation of Gene and Metabolic

Networks

• Graph theory• Most estimation and statistical

issues unresolved• Major (current) analytic tool:

Kascer-Burns Sensitivity Analysis

Gene networks are graphs

Kascer-Burns Sensitivity Analysis(aka. Metabolic Control Analysis)

“No theory should fit all the facts because some ofthe facts are wrong” (N. Bohr)“All models are wrong, although some models areUseful” (Box)

A B D E Fe1 e2 e3 e4

Flux = production rate of a particular product, here F

How best to increase the flux through thispathway?

Perhaps we increase the concentration of e1

However, it may be more efficientTo increase the concentration of e4

The flux control coefficient, introduced byKascer and Burns, provides a quantitative solutionto this problem

Flux Control Coefficients, Cji=@Fi@EjEjfi=@lnFi@lnEjThe control coefficient for the flux at step j ina pathway associated with enzyme j,

CjiRoughly speaking, the control coefficient is the percentage change in flux divided by percentage change in enzyme activity

.

Activity

When the activity of E is near zero,C is close to 1When the activity of E is large,C is close to zero

Why many mutations are recessive: a 50%reduction in activity (the heterozygote)results in only a very small change in the flux

Kacser-Burns Flux summation theorem:XiCji=1

• Truly rate-limiting steps are rare• Coefficients are not intrinsic properties of an enzyme, but rather a (local) system property

• If a control coefficient is greatly increased in value, this decreases the values of other control coefficients

• While most values of C for proteins are positive,negative regulators (repressors) give negative values,allowing for C values > 1.

“rate-limiting” steps in pathways

.

1.00.80.60.40.20.00

2

4

6

8

10

12

14

16

18

20

Control Coefficient C

Hence, the limiting increase in f isf=11°CjEf=11°r°1rCjESmall-Kacser theorem: the factor f by which flux isincreased by an r-fold increase in activity of E is

Using estimated Control Coefficients as selection aids

• Loci with larger C values should respond faster to selection

• Such loci are obvious targets for screens of natural variation (candidate loci)

• Selection with reduced correlations– Tallis or Kempthorne - Nordskog restricted selection

index – Select on loci with large C for flux of interest,

smallest C for other fluxes not of concern– Positive selection on C for flux of interest, selection to

reduce flux changes in other pathways

A B D E Fe1 e2 e3 e4

G

H

e5

e6

Flux we wish to increase

We wish this flux toremain unchanged

The initial approach might be totry either e3 or e4, rather than e1 or e2

A more correct approach, however is toPick the step(s) that maximize CF while minimizing CH

Index selection on pathways

• The elements of selection include both phenotype and C, and (possibly) marker markers as well

• Problems:– C is a local estimate, changing as the

pathway evolves– Still have all the standard concerns with a

selection index (e.g., stability of inverse of genetic covariance matrix)

– These are important caveats to consider even under the rosy scenaro where all C’s are know

What to call it?

MAS = Marker Assisted Selection

CAS = Control Coefficient Assisted Selection

CASH $ = Control Activity Selection Helper

Summary• Microarray analysis = data mining• Potential (immediate) useage:

– Suggesting candidate loci– More efficient use of G X E– Reducing/breaking deleterious correlations

• Cis (easy) vs. trans (hard) control of expression levels

• Future = analysis of pathways– Index selection (and all its problems)

QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture.

U of A Campus

Farewell from the “desert”