Embedded Files
Current Research
Why don't we observe all possible trait combinations in nature?
Evolution has generated a wide variety of traits, but not all possible combinations. For example, we never see creatures that resemble the imaginary animal depicted in the figure below in the natural world. One possible explanation is that some combinations are possible but have low fitness, so they are eliminated by natural selection. Another possibility involves bias in trait variability, whereby certain combinations are less likely to appear than others due to bias in availability of traits induced by genetic (mutational) and non-genetic (environmental) perturbations.
Potential impact of variability on evolution
Biased variability can constrain adaptive evolution if it limits variation in the direction of selection (left), but it can accelerate adaptive evolution if it increases variation in dimensions aligned with the direction of selection (right). Even environmentally induced phenotypic variation has the potential to initiate and direct adaptive evolution through processes such as genetic assimilation. Accordingly, without a thorough understanding of the nature of phenotypic variability, our understanding of why evolution has progressed in the manner that it has–or why some phenotypes evolve repeatedly while others are one-offs–remains incomplete.
Is biased variability a special feature?
It is widely known that animals exhibit biased phenotypic variability in traits such as body shape and coloration, with certain variants emerging more frequently than others in response to both genetic and non-genetic perturbations. To investigate whether such bias is a general biological principle rather than a peculiarity of sophisticated traits in multicellular organisms, I examined its presence at the molecular level, focusing on gene expression. Gene expression—the process by which genes are turned on or off—can vary due to genetic mutations or environmental changes. Even in genetically identical cells, gene expression can differ due to random processes, known as "stochastic cellular noise." These variability can play a significant role in evolution because it influences how traits develop in organisms. Even stochastic noise can facilitate evolution, such as seen in partial penetrance. Therefore, understanding the organization of variability in the gene expression is of great importance in evolutionary biology.
Stochastic cellular noise: Temporal montage of a single channel of a microfluidic chemostat device. Escherichia coli , a model bacterium, expressing GFP from a constitutive promoter on the chromosome is cultured in the narrow channel (a few microns wide). A representative snap shot of the channel is indicated by red, showing cell-to-cell heterogeneity in GFP expression among a clonal population at the same time point. A representative single lineage of E. coli is outlined by yellow/cyan dashed lines, where the cells outlined by cyan corresponds to the cells immediately after cell divisions, showing temoral fluctuations in GFP expression of the same cell. See Past Research.
A common mechanism for transcriptional variability among perturbations?
Cells consist of numerous interacting molecules (right figure), with their quantities constantly changing. In such a complex system, different perturbations are expected to produce distinct responses. For example, changes in gene expression caused by environmental perturbations (light blue pathway from outside of the cell to a focal gene) are expected to differ from those caused by mutations (light red pathway from a mutated gene to the focal gene).
Recently, I observed that both genetic and non-genetic perturbations in gene expression often exhibit similar patterns in terms of sensitivity (the bottom left schematic diagram) and direction (the bottom right schematic diagram). For instance, genes that are susceptible to genomic mutations–particularly trans-acting mutations–also tend to be sensitive to environmental influences. Moreover, the major axes of genetic variation across many genes closely resemble those of environmental variation.
Despite arising from different sources, these patterns suggest a potentially shared underlying mechanism. The similarities imply that gene regulatory networks—which govern gene activity—may play a central role in shaping how traits evolve in response to diverse types of change.
Similarity in sensitivity: Genes that respond strongly to environmental changes tend to do the same with trans-acting genomic mutations. For example, genes essential for basic cell functions show little variability in expression levels, while those involved in nutrient processing show more. See Tsuru and Furusawa, Nature Commun, 2025; Tsuru and Furusawa, Sci Rep, 2025.
Similarity in directionality: Expression levels among genes may vary together in the same direction, regardless of whether the trigger is genetic or environmental. This pattern suggests there may be a common underlying process influencing variability. See Tsuru and Furusawa, Nature Commun, 2025.
The role of transcriptional regulatory networks on transcriptional variability
What molecular mechanisms underlie the biased variability in gene expression levels across different types of perturbations? In living organisms, traits result from interactions within networks of genes. Accordingly, variability in gene expression levels is expected to be influenced by the properties of the regulatory networks, regardless of the underlying causes of perturbations. Recently, I identified genetic properties of transcriptional regulatory network underlying transcriptional variability across different types of perturbations in E. coli (Tsuru and Furusawa, Nature Commun, 2025). The sensitive genes with higher transcriptional variability were often regulated by densely interconnected influential global regulators (red nodes), forming network structures capable of sensing and propagating diverse perturbations. These regulators also induced correlated transcriptional changes among genes, resulting in biased directionality in transcriptional variability shared across genetic and non-genetic perturbations (Tsuru and Furusawa, Nature Commun, 2025; Baumstark, Hanzelmann, Tsuru et al., Nature Commun, 2015).
The role of cis-regulatory regions on transcriptional variability for cis-acting mutations
Since genomic motions occurred all over the genome, the sensitivity governed by transcriptional regulatory network should reflect sensitivity to trans-acting mutations, not cis-acting mutations. To confirm this, I constructed mutant libraries expressing GFP from randomly mutated cis-regulatory regions in E. coli. Flow cytometry measured the variance of GFP distribution as a measure of sensitivity to cis-acting mutations. There was no correlation between sensitivity to cis-acting and trans-acting mutations, confirming their distinction. Interestingly, cis-regulatory regions controlling essential genes exhibited a pronounced bias in mutational variability: mutants with reduced expression were more frequent than those with increased expression. Our evolutionary simulations on a rugged fitness landscape provide a rationale for this vulnerability. These results suggest that past selection has shaped cis-regulatory regions to exhibit nonuniform mutational variability, biased toward lower expression levels in random mutants. See Tsuru, Hatanaka and Furusawa, Mol Biol Evol, 2024.
Past Research
Deciphering noise in gene expression and its role
Under construction.
Approach: Engineer to understand
Using techniques in synthetic biology, we design, construct, and analyze gene regulations to understand how gene regulatory networks contribute to patterns of variability in gene expression levels. For instance, we have created synthetic operons, regulatory links and promoters. These synthetic approaches identify unique features of natural regulations which are distinct from intuitive expectations.
Current research foci
Promoters constrain evolution of expression levels of essential genes in E. coli.
Congruence between noise and plasticity in protein expression.
Genetic properties underlying transcriptional variability across different perturbations.
Laboratory evolution of the bacterial genome structure through insertion sequence activation.
Google Sites
Report abuse