Tuesday, March 13, 2012

Principles of life, in action

We presented a list of principles last week that we think are generic and help explain much of what goes on in life -- during development, between organisms, and over evolutionary time.  They won't tell you which gene is turned on when, or where the enhancer is that controls its expression -- to answer questions such as these, you need different sets of principles.  But, they do, we think, comprise a list of basic observations about life that are shown over and over again to be true.  The proof of that is the gold standard in science: the principles are, implicitly and explicitly, routine underpinnings and assumptions of daily research in the life sciences.

A couple of papers in last week's Science describing how bacteria adapt to changing environments illustrate this nicely. In their paper, Nicholas et al. undertook to explore changes in the 'transcriptome' (suite of genes expressed at any given time) when the bacterium, Bacillus subtilis, underwent environmental changes, 104 in all, in controlled conditions in their lab.  That is, they set out to characterize 'transcriptome plasticity' in this critter.  They catalogued the genes expressed in every set of conditions, and found 'highly correlated changes in expression', in response to conditions.
Of the previously annotated coding sequences (CDSs), only 186 (4.4%) were not expressed under any condition. Most of these CDSs were of unknown function and predicted to originate from horizontal transfer (SOM 3 and table S3). The 30% of the CDSs most highly expressed under each condition were defined as “highly expressed” (SOM 3). Eighty-five percent of all CDSs were highly expressed in one or more conditions (fig. S3A), but only ~3% (144) of all CDSs were highly expressed under all conditions, indicating that most B. subtilis genes are differentially expressed. Genes in the latter group encode proteins with essential functions and enzymes involved in glycolysis, iron sulfur metabolism, and detoxification pathways.
That is, a small percentage of genes were always expressed, another small percentage of genes never expressed, and the large majority of genes were highly expressed in at least one environment.  These facts show the modularity of genomes and differential combinatorial context-specific usage,  which implies various kinds of internal compartmentalization (effectively, the sequestration of components), even in  simple bacteria.

They also analyzed promoter regions, and discovered that about 46% of all genes can be transcribed from more than one promoter.  And:
We comprehensively mapped transcription units (TUs) and grouped 2935 promoters into regulons [reglators of groups of genes] controlled by various RNA polymerase sigma factors [transcription initiation factors in bacteria that facilitate RNA polymerase, the enzyme that enables the stringing together of nucleotides into RNA, to bind to gene promoters], accounting for ~66% of the observed variance in transcriptional activity. This global classification of promoters and detailed description of TUs revealed that a large proportion of the detected antisense RNAs arose from potentially spurious transcription initiation by alternative sigma factors and from imperfect control of transcription termination.
That is, a considerable amount of transcription is imprecise (what we've called 'slop').  For example,  transcription of coding regions often extended beyond the boundary of the gene, and this produced antisense RNA (asRNA), rather than protein-coding RNAs.  But, they propose that these 'spuriously' produced asRNA's may actually have a biological role, sometimes in gene regulation.  Thus, slop turns out to be an important part of life.  And gene regulation is a combinatorial phenomenon that involves some sorts of context-specific balance among factors.  Combinatorial causation (one can say via 'Boolean' logic if one is familiar with computerese) is fundamental even to bacterial life and eology.

Among other things, the authors conclude that their analysis "revealed that asRNAs generated by inefficient control of transcriptional events might be a drawback of [transcriptional plasticity], though they might contribute to the creation of previously unknown regulatory functions."

Buescher et al. also analyzed changing conditions on Bacillus subtilis, in their case, food -- glucose and malate, sources of carbon for these bacteria -- to try to understand the interaction of regulatory and function networks in the cell.  As described in the commentary to these two papers, Buescher et al found that:
Most genes were differentially expressed, and 127 out of 154 transcription factors changed their activity during one or both shifts. Changes in carbon metabolism during both shifts were mediated largely by altering the abundance of a small number of proteins. Although both nutrients are preferred carbon sources for B. subtilis, adaptation to glucose availability was slow and largely controlled transcriptionally, whereas adaptation to malate was fast and primarily regulated posttranscriptionally.
To achieve adaptation, B. subtilis makes some compromises. The observations of Nicolas et al.suggest that transcriptional plasticity is often associated with imperfect control, leading to the generation of antisense RNAs. This may arise due to aberrant termination of transcription, or spurious transcription initiation while using alternative sigma factors. Similarly, the results from Buescher et al.suggest that, depending on the prevailing environmental condition, the preferential uptake of one carbon source over the other might confer condition-specific evolutionary advantages in growth. This is achieved by active regulation or constitutive expression of several genes—two distinct strategies, neither of which is advantageous per se. Thus, to adapt to changing environments, B. subtilis makes a trade-off between the implementation of complex regulatory programs and imprecise regulation.
So, how does this illustrate our principles?  Here's the list again, briefly, but the annotated list is here.

1. Inheritance with memory
2. Modularity
3. Sequestration
4. Coding and interaction
5. Contingency
6. Chance
7. Adaptability
8. Cooperation

Inheritance with memory is why B. subtilis, as any organism, are what they are generation after generation, and why they have predictable responses to environmental changes. Modularity is obvious throughout; regulatory regions, genes, cells themselves are all modules.  And cells are sequestered, but not completely, which is why they can detect the environment and respond to changes.  Coding and interaction are fundamental aspects of gene expression,  of course, and which genes are expressed is contingent upon environmental conditions.  Chance comes into the picture in how reliably gene transcription takes place (not always reliably), and indeed, in what environment the bacteria must respond to when.  And it all involves cooperation among genes, regulatory elements, and so on.

We think it's important and worthwhile to try to itemize principles like these, not because we feel we have made any sort of 'discovery', but because in the search for principles of life, there are twin thrusts, one being molecular biological reductionism, and the other being overly simplistic deterministic effectively one-gene or gene-for causal and evolutionary thinking.  The latter leads to a focus on competition when the principles in the format that we try to enunciate them show how much more life depends on cooperative (coordinated, jointly functional) rather than competitive interactions.

No comments: