Microbial Stowaways: Inimitable Survivors or Hopeless Pioneers?

Consider the Ubiquity and Resiliency of Earth Life

We know from the rock record that complex life is amazingly resilient (Escarguel et al., 2011). Despite repeated near annihilations, complex life has never failed to adapt to new environments through natural selection, although at the expense of previous forms and strategies (Kutschera and Niklas, 2004). The events spurred morphological novelty and unprecedented diversity. While we don't actually know the mechanism that allowed this to occur, there is a theme of sorts in the solution. Once complex biology discovered bilateralism, it became a successful solution. HOX genes are the basic genetic solution for the mapping problem in assigning directionality for cell differentially in metazoans today (Gehring et al., 2009; Pick and Heffer, 2012). HOX gene strategy leads to a quite elegant explanation for the diversity we frequently see in fossils. While circumstantial and without any basis for “why” HOX genes might have bridged some innovative divide that was fixed in metazoans, it's satisfyingly consistent that the genetics as we understand them bolster the argument. One might say that HOX genes were to bilateral entities what wheels were to anything that rolls. Variation on a theme was a powerful engine of diversity given catastrophic events.

We believe that once life got started on this planet it thrived, one way or another. This statement reflects in part on that biology we know can leave a relatively robust history in fossils, but it rests upon the assumption that this is true also of the earlier period in Earth's history when microbial life dominated (Mata and Bottjer, 2012). This is a biology that is largely unable to leave its morphological record of extinctions or even its presence in rocks for us to consider. We do have some. Stromatolites, or the more general term microbialites, are a result of unique conditions that have provided us with the most robust fossil data so far, although even these aren't without detractors for validity (Allwood et al., 2006, 2009; Kershaw et al., 2012; Mata and Bottjer, 2012). Interestingly, even the very old Earth rocks indicate that these preserved community structures were individually and collectively quite complex (Allwood et al., 2006, 2009). If these are indeed real, which we believe they are, the importance of them is the fact that they speak to a very early, very diverse microbiota and community life-form as a common strategy.

The Adaptive World of Prokaryotic Lifestyles

Because microbial life today is integral to all forms of life (consider especially their elemental recycling) (Newman and Banfield, 2002), they are the earliest Earth life; and from the fact that they harbor the very same adaptive machinery at their core as the rest of life on our planet, we can likely assume that they too have been resilient when exposed to catastrophic events. The question was how? Being clonal entities meant identical genetics, and certainly without something like the ingenious toolbox of HOX genes central to more complex cellular organizations, the default evolutionary strategy was that diversity occurred piecemeal through mutational happenstance: a strategy requiring large populations, lots of time, and strict vertical inheritance. Even pioneers in the field of bioinformatics, such as Margaret Dayhoff, dismissed the likelihood that whole sections of genes moved willy-nilly though microbial genetic landscapes (Dayhoff et al., 1974; Dayhoff and Schwartz, 1981). We assumed we could count on strict vertical inheritance, even if it did seem a bit plodding and mildly tedious.

However, once sequencing technology of nucleic acids became practical and cheap, the database of most prokaryotic genome content was telling a different story than what our intuition had supposed. There were indications that there was a great deal more going on in microbial genomes that was as promiscuous and as likely to produce diversity and adaptive potential as eukaryotic sex; it was just that the mechanism(s) were different (Lawrence and Ochman, 2002). And these mechanisms were fast (Popa and Dagan, 2011). The methodology of antibiotic resistance is a stellar example of bacterial adaptation fueled in real time (Andersson and Hughes, 2011; Martinez, 2011; Sommer and Dantas, 2011). The fact that previously innocuous gut bacteria could become pathogenic horrors through genetic islands (Arnold and Jackson, 2011; Jackson et al., 2011; Partridge, 2011) further directed our thinking on just how fluid bacterial genomes could be (Toussaint and Merlin, 2002; Toussaint and Chandler, 2012). Bioinformatics evidenced a whole suite of mobile elements that served as environmental vectors for bringing in or taking out strips of genetic material to neighboring prokaryotic entities (Siefert, 2009; Aminov, 2011), be they related or not. It was a brave new world in our understanding of prokaryotic ability to diversify. Horizontal gene transfer (HGT) in all its ramifications was at the core (Andam et al., 2011; Wiedenbeck and Cohan, 2011). Rapid adaptive response to a changed ecology was providential when the bank of genetic resource available was contained in the hundreds of billions of constituents in a community. This bank was global, considering that prokaryotes and their parasitic viruses had the potential for travel abroad (Thurber, 2009; Lennon and Jones, 2011; Newton et al., 2011; Veesler and Cambillau, 2011). As long as there was the possibility to eke out a living in any little crevice or water spot that a microbe might happen to land, the potential for a one-in-a-billion chance of accumulating the right genetic material to allow adaptation was not only a possibility but almost a guarantee.

As the estimations of just how many prokaryotes and viruses were around in the oceans and lakes (Whitman et al., 1998; Zinger et al., 2011), in the dirt (van der Heijden et al., 2008), and in our guts (Desikan, 2010; Qin et al., 2010; Dominguez-Bello et al., 2011; Esteve et al., 2011), were being published, it was clear that there was a desperate need to understand microbiology in terms of populations if we were to get a handle on how obligatory HGT might be to the adaptive process.

As a matter of fact, the incidence of HGT impacted our hope that prokaryotic taxonomy had a shot at being based upon phylogeny. Was the small subunit ribosomal RNA a valid indicator of the organismal phylogeny, or did we need more data (Lozupone and Knight, 2008)? And if gene transfer in the horizontal direction was rampant in prokaryotes without regard to which genes were being transferred, what could we possibly use as valid, nontransferred genes to augment the data set? What exactly constituted a species in the prokaryotic kingdom (Cohan, 2006; Klenk and Goker, 2010; Lukjancenko et al., 2010; Mira et al., 2010; Andam and Gogarten, 2011; Vos, 2011)?

In the last two decades, several scientific activities have come to bear upon the mechanisms that provide adaptation in the microbial world and our understanding of the evolution of diversity in prokaryotes. They bear directly on the chances of survival of forward contamination to wherever we should travel.

On February 24, 1988, Richard Lenski began an experimental evolution study using 12 populations of genetically identical E. coli that reproduced only asexually (without bacterial conjugation) (Lenski and Travisano, 1994; Lenski et al., 2003). The idea was to sample each population each day and inoculate fresh media with the previous day's population. The experiment is still tracking, and the populations reached the milestone of 50,000 generations in February 2010. At various points, analyses of the evolutionary trajectory have been undertaken (Schneider and Lenski, 2004; Ostrowski et al., 2008; Barrick et al., 2009). Specifically, the questions addressed have been to know how rates of evolution varied over time, how repeatable evolutionary events were in identical environments, and how phenotypic changes are reflected in the genotype. To summarize to date the results of this elegant experiment, each population is thought to have generated hundreds of millions of mutations over the first 20,000 generations, but Lenski estimates that only 10–20 beneficial mutations were fixed in each population and less than 100 total point mutations (including neutral ones) were fixed. Between 31,000 and 31,500 generations, one population evolved a citrate-using variant (a characteristic that is often used to distinguish E. coli from Salmonella), and there was evidence that this ability could re-evolve from earlier time points in this lineage, at a rate of occurrence of once per trillion cells. Lenski explains this through Stephen J. Gould's argument that “historical contingency can have a profound and lasting impact” on the course of evolution (Blount et al., 2008).

These experiments are certainly evidence that point mutations provide adaptive potential in pure culture situations. In summary, though, we note that for a mutation to become fixed in the population, it required tens of thousands of generations and trillions of cells and the nutrient availability in which to conduct the experiment.

Once bioinformatics moved out into the field, the microbiology community strived to identify bacterial species, or ecotypes, by what they did in their environment (Cohan, 2002; Scanlan et al., 2009; Aller et al., 2010). This seemed sensible because organismal identity in prokaryotes was part historical inheritance and part horizontal. We accepted as fact that diversity was rapidly sparked by HGT and was at play everywhere. Even if the genetic machinery that provided the adaptive innovation was borrowed, for microbes it was indicative of that organism's role in its environment. This was important because even though the tenets of population biology don't mesh well with entities that don't have barriers that can define species precisely in the classic sense, for bacteria it aided in our ability to identify community members in a particular environment.

For bacterial communities then, one needed to appreciate the diversity within the community (expressed in this combination of vertical and horizontal genetic ecotype terms), the community as a whole, and the broader environment itself: what each has to offer in the way of commodity and challenge, respectively. For each individual in the community, the intra- and intercommunity environment had to be utilized to sustain growth and reproduction while adapting. Under these considerations, microbial population biology has made some good progress. The work at Cuatro Ciénegas Bolson (CCB) is an example of that progress (see articles in this issue). Briefly, the CCB combines geographic isolation, long-term continuity, and strong local selection pressures (especially from phosphorus limitation) creating high levels of endemic microbial biodiversity (in the form of locally unique microbiota) (Elser et al., 2005; Souza et al., 2006, 2008; Cerritos et al., 2008; Desnues et al., 2008; Escalante et al., 2008; Breitbart et al., 2009). Recognizing that extensive HGT can provide a means for evolutionary innovation and adaptation, we have come to understand that it can also obliterate local diversification by homogenizing the populations, especially when outside genetic resources are not obtainable due to geographic isolation. While the unique factors of the CCB are good at producing locally adapted endemic species and a lot of diversity while doing so, they require a long-term stable environment in which to do it.

What Is Our Potential for Seeding Mars?

So given what we know about the efficiency and the occurrence of the adaptive processes for microbial entities on Earth, how does this speak to our concern for contamination?

Crisler et al. (2012) recently conducted an experimental regime exposing certain microbial populations to conditions that simulated the surface of Mars. Martian surface environmental parameters include extremes in salinity, temperature, desiccation, radiation, a diurnal cycle that includes alternate freezing and thawing, low water activity, and a thin, anoxic atmosphere (Clark, 1998). If water ice in permafrost regions should melt or water vapor condense, brines would likely result that are nearly saturated in MgSO₄ (McEwen et al., 2011). Using bacterial populations from the Great Salt Plains, Crisler and colleagues lightly inoculated (below 0.05 optical density at 600 nm) the most halotolerant members to a variety of experimental regimes. [A back-of-the-envelope calculation can estimate that the number of cells in a light inoculation is about 55 million (Volkmer and Heinemann, 2011).] They demonstrated that aerobic, halotolerant bacteria could grow in the high concentrations of MgSO₄ estimated for martian soils. However, they found that exposure to cyclic drying and rewetting or freezing and thawing was a limiting factor for the growth of these terrestrial organisms. They could not survive.

Let's estimate that, in our efforts to clean spacecraft, we leave 20,000 Bacillus spores aboard. (Spacecraft clean room requirements are much more stringent than this; the Goddard clean room has a Class-10,000 rating, which indicates that any cubic foot of air has no more than 10,000 particles floating around in it larger than 0.5 microns. A bacterium is roughly 1.0 micron; Bacillus spores are 0.8 microns.) Once on Mars (or really anywhere else that is not Earth-like), they must at once continue to reproduce and adapt. We imagine that, in Earth terms, Mars is even more barren of nutrients than places like the CCB; therefore, the adaptive challenge for microbial stowaways would be huge, much more challenging than bringing citrate across a membrane, utilizing an alternate sugar, or conserving resources by scavenging components of community members. Given the results from the Crisler et al. study, we know that the community would have to endure and survive several challenging ecological parameters, which are not limited to repeated thawing and freezing in the occasional, near-saturated brine, in an unfamiliar atmosphere that was not oxidizing. If the reliance for adaptive progress relies on point mutations, we have some idea of how long they must survive if challenged minimally—years. We have no estimates if the challenges are multiple and simultaneous as the martian surface would necessitate. Without a diverse community, gene transfer, possibly the primary engine of bacterial diversity for real-time adaptive evolution on Earth, would be a non sequitur. The genetic repertoire that makes Earth microbiology so responsive to environmental change is not available extraterrestrially, if the only stowaways are a few other nearly identical clones. Add to that geographic isolation and fewer resources to sustain any amount of evolution, and Earth-based life on Mars would go extinct, if indeed it ever was able to colonize.

What If the Impossible Happens?

The curious question might be, though, that if the impossible did happen, what would it actually mean? Given that we understand the unique strategies that microbial Earth life use to diversify and adapt, it's almost certain that, if some part of 20,000 clonal Bacillus survived, they would likely have devised a whole other strategy than anything that is currently employed on Earth. We are more and more certain that microbial evolution on Earth works because it's a big numbers game in which a global genetic resource is utilized. If neither of these attributes were available to the stowaways, we'd have to consider what unintentional and successful seeding of the surface of Mars meant. We'd have to decide whether or not that event constituted a new origin of life, a window into the way life could seed and succeed, some kind of historical contingency recovered from a time when Mars was full of life, or a rewind of that mysterious time in Earth's history when life indeed came from a common ancestor pool of nondiversified entities. [For an interesting take on how one might differentiate potential seeding sources based on ribosome evolution, see Fox (2011).] It might not be an experiment that we would ethically want to conduct, but it would be an unparalleled experimental result.