One of the physiologist's goals is to understand how animals cope with fluctuations in their environment. Andrew Gracey discusses how scientists can relate physiological responses to environmental change to shifts in gene expression (p. 1584). Of the many techniques available to analyse changes in gene expression, cDNA microarrays remain the most powerful technique for screening nonmodel organisms, because of the large numbers of genes that can be analysed. Highlighting work carried out in the carp, Gracey and his colleagues have created a cDNA microarray to investigate the gene regulatory mechanisms underlying cold and hypoxia acclimation. The challenge is to integrate all the genetic and physiological mechanisms with the ultimate goal of predicting how an organism will respond to environmental and physiological perturbations.

Again focussing on non-model species, Dietmar Kültz(p. 1593) writes about the challenges and techniques used to understand osmoregulation and coping with salinity in creatures such as tilapia, sharks and sponges. While researchers rely on gene databases for model organisms such as the stenohaline zebrafish to find out which genes are involved in physiological responses,there could be problems relying on these data to learn about the equivalent responses in the euryhaline tilapia. Not only do these fish operate over different salinity ranges, there will also be a high degree of plasticity in the physiological response and in the underlying gene and protein networks in response to salinity. The solution, Kültz suggests, is not to be too ambitious too soon but to focus on one biological process at a time rather than trying to understand the whole non-model organism at once.

According to Bradley Buckley(p. 1602), one technique available to researchers to compare broad scale patterns of gene expression between species is heterologous hybridization, where a microarray from one species, such as the eurythermal goby, is used to probe for gene expression in another, such as species of cold-adapted Antarctic fish. While this approach saves researchers the effort of having to construct a new cDNA microarray for each new species they wish to study, there are caveats that have to be addressed. One issue is that the two species need to be closely related enough for the genes from one species to accurately identify the same genes in the other species. The length of the DNA probes on the array also affects success: longer probes reduce the chance of random mismatching.

It's not just comparisons between species that benefit from microarray technology; the technique can also be used to look at the well-studied problem of aging. Stuart Kim writes that aging is a complex process, which results in cumulative changes in the expression of many genes(p. 1607). By using microarrays to perform genome-wide scans, researchers can define the aging process, by comparing young and old organisms, such as worms and flies, and tissues in mice and humans. While many differences in gene expression between young and old are specific to a particular species, there are some common features. For example, the 95 genes that encode components of the electron transport pathway in mitochondria `show common age regulation from worms to humans', says Kim; their expression decreases about twofold in older animals.

The final paper in this section, by Douglas Crawford and Marjorie Oleksiak,offers a word of caution (p. 1613). They stress the importance of measuring individual variation,because pooled samples can hide important physiological information. Levels of gene expression can differ greatly, even between closely related individuals. Using microarray analysis, they compared gene expression in isolated heart ventricles from killifish (Fundulus heteroclitus). They found that hearts from different individuals had differences in their metabolism: for example, which metabolic substrate was preferred by the tissues. 81% of this variation was explained by altered patterns of gene expression in various sets of genes coding for proteins in different parts of the metabolic pathway in separate groups of individuals. The implications of this are that results from inbred strains of animals should be interpreted with caution.