THURSDAY, 5 MAY 2022
Since the development of cell theory in the 19th century, the definition of cells still holds up firmly: they are the structural and functional units of life. Not only do cells store, repair, and pass on the molecule of DNA, but they can also activate the genes written in the DNA, expressing them as observable cellular traits. Therefore, if two cells activate the same genes, they acquire the same traits. Historically, scientists used this principle to group cells into populations, hence the conventional view of cell populations as ensembles of cells that activate the same genes and so share the same features.
However, technological advancements have revealed a much more complex picture. We now know that individual cells within a population can activate the same genes to different degrees or can even activate different genes. This distribution is referred to as single cell variability and results in cells having slightly (or very) different traits, challenging the traditional idea that cell populations are homogenous. At the heart of these cell-to-cell variations lie both regulated and random events involving epigenetic marks, errors in DNA replication or rearrangements of chromosomes, to name a few. Certainly, no two cells can ever be truly identical.
The great scale at which single cell variability can manifest has therefore led researchers to ask whether it might play a role in regulating the collective behaviour of the cell population. Indeed, many examples of the functional importance of single cell variations have been discovered. When subjected to a hostile environment, versatile cell populations that activate a diverse set of genes have a better chance of survival than their uniform counterparts, because they display a wider range of features. Although some cellular traits might decrease the overall fitness of the population, whether or not to carry such diversity is an easy gamble: bet hedging provides the cell population with greater chances of surviving in varied environmental conditions rather than expressing a unique trait specialized for only one environment.
Variation across single cells may also allow a population to give a stratified response to external stimuli. In predator-prey dynamics, if all predators decided to hunt simultaneously, all prey would be eaten at once, causing the predator population to quickly go extinct. Similarly, homogeneous decision-making of single cells would cause a sudden switch in the population’s behaviour, which could be detrimental: imagine if all of our cells that produce insulin uniformly programmed their death in response to an external signal — our bodies would altogether run out of insulin. Instead, fluctuations in the activation of genes within the cell population allow for fractional decision-making processes to cushion the blow of any external stimulus.
There are two additional interesting cases of population-level properties emerging from single cell variability. One is termed crowd control, whereby a subset of cells responds uniquely to a perturbation, producing molecules that reprogramme the entire population’s behaviour. The other considers that single cell variability enables the coding and transfer of information. Since the set of activated genes within a cell is like a barcode, a population that is not made of identical cells carries a higher information content than a homogenous one. Therefore, single cell variation endows a population with higher data capacity. This is evident in the brain, where extensive cell heterogeneity within populations may increase the complexity of neural circuits, enhancing the brain’s ability to transfer information.
One approach to examining the idea that single cell variability in a given gene is functionally important is to determine whether it is evolutionarily conserved between species, known as the comparative approach. Suppose we measure how much the activation of a particular gene varies across single cells in species A and, after comparing this with the same measurement taken in species B, we note that there is a difference. How do we know that this difference is significant? To address this, one could first assess the degree of cell-to-cell variability of that gene in relation to the degree of variability of other genes within the same species, and only then use this corrected measurement for comparisons among species.
To complicate things more, the use of the comparative approach faces several other significant issues. One problem is that there is no robust model to predict the single cell variability of activation associated with each gene. Such a model would be useful to rectify the variability that one measures experimentally. One further issue is biological: a gene may have conserved high cell-to-cell variability of activation because it may be directly responsible for some of the population-level functions discussed above, or, conversely, it may have conserved low single cell variability because its activation must be tightly regulated. But if so, how can we be sure that a low variability of activation is not functional to population behaviour? We cannot exclude the possibility that population-level properties result from a non-specific degree of activation of random genes, rather than being dependent on the precise level of activation of specific genes. To distinguish between these two possibilities is not at all easy.
However, one thing that we can directly test is how deactivating genes in the cell population affects the population’s behaviour. Several methods now exist to achieve this, one of which is using short RNA molecules called small interfering RNAs (siRNAs). These are purposely designed to stop the activation of genes. Importantly, siRNAs can be designed to target genes either specifically or randomly. If a particular cell population’s behaviour is not altered following deactivation of random genes, then that property is gene-independent: it would exist regardless of the specific genes whose activation varies across single cells. On the other hand, by suppressing specific genes, one could pinpoint the molecular players whose activation must vary from cell to cell for the population to acquire a given behaviour.
All in all, we now think of individual cells as existing in a population, just as individual organisms exist in a community: each cell’s traits and behaviours contribute to the population’s properties, just as traits and behaviours of different organisms serve community dynamics in an ecosystem. Some speculate that single cell variation may follow a kind of Darwinian mechanism for driving population-level decisions and might therefore offer a platform through which natural selection can occur. Even more broadly, single cell variability could represent a way of integrating environmental cues into the functioning of organisms.
Still, the hypothesis that single cell variability has evolved for certain functions is complementary to the more conservative idea that biological mechanisms evolve to allow function despite variability. Surely, cohesive behaviour of individual cells is important for the existence and survival of cell populations. At the same time, if the examples above have taught us anything, it is that single cell variability per se is indispensable for a population’s survival. Either way, there is no doubt that understanding the population-level functions that originate from single cell variability will be key to unravelling the complexities of multicellular systems.
Roberta Cacioppo is a third-year PhD student in molecular biology at Corpus Christi College. Illustration by Sumit Sen.
