Limits...
Developmental constraints on vertebrate genome evolution.

Roux J, Robinson-Rechavi M - PLoS Genet. (2008)

Bottom Line: We show that, in both species, genes expressed early in development (1) have a more dramatic effect of knock-out or mutation and (2) are more likely to revert to single copy after whole genome duplication, relative to genes expressed late.We determine the pattern of these constraints, which differs from the model used to describe vertebrate morphological conservation ("hourglass" model).While morphological constraints reach a maximum at mid-development (the "phylotypic" stage), genomic constraints appear to decrease in a monotonous manner over developmental time.

View Article: PubMed Central - PubMed

Affiliation: Université de Lausanne, Département d'Ecologie et d'Evolution, Quartier Sorge, Lausanne, Switzerland.

ABSTRACT
Constraints in embryonic development are thought to bias the direction of evolution by making some changes less likely, and others more likely, depending on their consequences on ontogeny. Here, we characterize the constraints acting on genome evolution in vertebrates. We used gene expression data from two vertebrates: zebrafish, using a microarray experiment spanning 14 stages of development, and mouse, using EST counts for 26 stages of development. We show that, in both species, genes expressed early in development (1) have a more dramatic effect of knock-out or mutation and (2) are more likely to revert to single copy after whole genome duplication, relative to genes expressed late. This supports high constraints on early stages of vertebrate development, making them less open to innovations (gene gain or gene loss). Results are robust to different sources of data -- gene expression from microarrays, ESTs, or in situ hybridizations; and mutants from directed KO, transgenic insertions, point mutations, or morpholinos. We determine the pattern of these constraints, which differs from the model used to describe vertebrate morphological conservation ("hourglass" model). While morphological constraints reach a maximum at mid-development (the "phylotypic" stage), genomic constraints appear to decrease in a monotonous manner over developmental time.

Show MeSH

Related in: MedlinePlus

Variation across zebrafish development of the expression of essential genes compared to non-annotated genes.At each time point, the ratio of the number of essential genes expressed on the number of non-annotated genes expressed is plotted. A gray box on the x-axis indicates the phylotypic period. (A) Gene expression as reported using in situ hybridization data. The x-axis is proportional to time. A weighted linear regression was fitted to the data and the regression line plotted. (B) Gene expression as reported by “present” calls from Affymetrix array data. The x-axis is in logarithmic scale. A Spearman correlation was computed (coefficient ρ).
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2600815&req=5

pgen-1000311-g001: Variation across zebrafish development of the expression of essential genes compared to non-annotated genes.At each time point, the ratio of the number of essential genes expressed on the number of non-annotated genes expressed is plotted. A gray box on the x-axis indicates the phylotypic period. (A) Gene expression as reported using in situ hybridization data. The x-axis is proportional to time. A weighted linear regression was fitted to the data and the regression line plotted. (B) Gene expression as reported by “present” calls from Affymetrix array data. The x-axis is in logarithmic scale. A Spearman correlation was computed (coefficient ρ).

Mentions: First, we used the phenotypes of gene loss-of-function as an indicator of selective pressure on genes. We extracted genes essential for the viability of the zebrafish, giving a lethal phenotype when non functional [16]. We expect that the loss of a gene should be more deleterious if this gene is expressed at a developmental stage with strong constraints. Thus we estimated whether genes were expressed or not at each stage, and computed the ratio of expressed essential genes to expressed reference genes (no reported loss of function phenotype). We then plotted the variation across development of this ratio. We used two different types of data to evaluate the presence of gene expression: (i) expression patterns from in situ hybridizations (Figure 1A), and (ii) “present” or “absent” calls from an Affymetrix microarray experiment (Figure 1B). Results are consistent for both data types: the proportion of essential genes is higher among genes expressed in early development, with a significant negative correlation. For the in situ hybridizations (Figure 1A), a linear regression is significant, but a parabola is not. The parabola has been suggested as the quantitative expectation of an hourglass-like model 3,17. These results indicate a continuous trend over developmental time, with stronger constraints on early development.


Developmental constraints on vertebrate genome evolution.

Roux J, Robinson-Rechavi M - PLoS Genet. (2008)

Variation across zebrafish development of the expression of essential genes compared to non-annotated genes.At each time point, the ratio of the number of essential genes expressed on the number of non-annotated genes expressed is plotted. A gray box on the x-axis indicates the phylotypic period. (A) Gene expression as reported using in situ hybridization data. The x-axis is proportional to time. A weighted linear regression was fitted to the data and the regression line plotted. (B) Gene expression as reported by “present” calls from Affymetrix array data. The x-axis is in logarithmic scale. A Spearman correlation was computed (coefficient ρ).
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2600815&req=5

pgen-1000311-g001: Variation across zebrafish development of the expression of essential genes compared to non-annotated genes.At each time point, the ratio of the number of essential genes expressed on the number of non-annotated genes expressed is plotted. A gray box on the x-axis indicates the phylotypic period. (A) Gene expression as reported using in situ hybridization data. The x-axis is proportional to time. A weighted linear regression was fitted to the data and the regression line plotted. (B) Gene expression as reported by “present” calls from Affymetrix array data. The x-axis is in logarithmic scale. A Spearman correlation was computed (coefficient ρ).
Mentions: First, we used the phenotypes of gene loss-of-function as an indicator of selective pressure on genes. We extracted genes essential for the viability of the zebrafish, giving a lethal phenotype when non functional [16]. We expect that the loss of a gene should be more deleterious if this gene is expressed at a developmental stage with strong constraints. Thus we estimated whether genes were expressed or not at each stage, and computed the ratio of expressed essential genes to expressed reference genes (no reported loss of function phenotype). We then plotted the variation across development of this ratio. We used two different types of data to evaluate the presence of gene expression: (i) expression patterns from in situ hybridizations (Figure 1A), and (ii) “present” or “absent” calls from an Affymetrix microarray experiment (Figure 1B). Results are consistent for both data types: the proportion of essential genes is higher among genes expressed in early development, with a significant negative correlation. For the in situ hybridizations (Figure 1A), a linear regression is significant, but a parabola is not. The parabola has been suggested as the quantitative expectation of an hourglass-like model 3,17. These results indicate a continuous trend over developmental time, with stronger constraints on early development.

Bottom Line: We show that, in both species, genes expressed early in development (1) have a more dramatic effect of knock-out or mutation and (2) are more likely to revert to single copy after whole genome duplication, relative to genes expressed late.We determine the pattern of these constraints, which differs from the model used to describe vertebrate morphological conservation ("hourglass" model).While morphological constraints reach a maximum at mid-development (the "phylotypic" stage), genomic constraints appear to decrease in a monotonous manner over developmental time.

View Article: PubMed Central - PubMed

Affiliation: Université de Lausanne, Département d'Ecologie et d'Evolution, Quartier Sorge, Lausanne, Switzerland.

ABSTRACT
Constraints in embryonic development are thought to bias the direction of evolution by making some changes less likely, and others more likely, depending on their consequences on ontogeny. Here, we characterize the constraints acting on genome evolution in vertebrates. We used gene expression data from two vertebrates: zebrafish, using a microarray experiment spanning 14 stages of development, and mouse, using EST counts for 26 stages of development. We show that, in both species, genes expressed early in development (1) have a more dramatic effect of knock-out or mutation and (2) are more likely to revert to single copy after whole genome duplication, relative to genes expressed late. This supports high constraints on early stages of vertebrate development, making them less open to innovations (gene gain or gene loss). Results are robust to different sources of data -- gene expression from microarrays, ESTs, or in situ hybridizations; and mutants from directed KO, transgenic insertions, point mutations, or morpholinos. We determine the pattern of these constraints, which differs from the model used to describe vertebrate morphological conservation ("hourglass" model). While morphological constraints reach a maximum at mid-development (the "phylotypic" stage), genomic constraints appear to decrease in a monotonous manner over developmental time.

Show MeSH
Related in: MedlinePlus