Microbiology, like most disciplines of biology, is replete with an extensive vocabulary, with many synonyms and, of course, much ambiguity. Nonetheless, there exist a number of core terms and concepts, knowledge of which will make it much easier to understand microbial evolution, both in the abstract and in terms of reading (and understanding) the literature. Note, however, that the applicability of these microbiology terms to some degree will be dependent on the degree of microbiological training of the authors of the publication you might be reading. That is, there exist a number of terms that are somewhat equivalent to those employed in microbiology that various authors have borrowed instead from, e.g., ecology or evolutionary biology. The resulting multiplicity of vocabulary is generally a problem when taking on interdisciplinary work. Here I provide a number of terms and concepts that are applicable to microbiological systems, not all of which necessarily will be found in general microbiology texts.

Growth media

Organisms have specific requirements for replication and these requirements are met by the physical, chemical, and sometimes even biological aspects of the media in which growth is attempted. In microbiology, growth media is typically described in terms of the chemical characteristics, i.e., as a list of ingredients, though more physical aspects such as osmolarity and pH are important characteristics of media as well. Presence or absence of oxygen, temperature, and presence or absence of light can also be important components, and in the case of symbiotic organisms, so too can be the presence of proper host organisms.

Microbiological media can be described as liquid (broth), semi-solid, or solid. Degree of solidity is usually though not quite always controlled by varying levels of added agar, a gelling agent isolated from seaweed. Broth media has no or, sometimes, very little agar added. Semi-solid media has enough agar to prevent the media from behaving as a liquid, in terms of flow, but not so much that the media has much resistance to shear (i.e., semi-solid media generally lacks structural integrity). Solid media, such as that found as the agar layer in Petri dishes, not only inhibits flow but also is resistant to shear, i.e., it doesn't tear as easily as semi-solid media. Broth cultures can be shaken, stirred, or not actively mixed. Shaking or stirring results in an environment that mostly lacks spatial structure. If mixing is not actively encouraged, then this can approximate a static microcosm. There diffusion can occur, as it also can occur in semi-solid and solid media. Mixing as a consequence of active motility can also occur. The latter to a degree can also result in organism movement in semi-solid media and, though less so, in solid media. Static microcosms along with solid or semi-solid media thus lack mixing except to the extent that organisms move themselves about. The consequence of inhibition of mixing as well as inhibition of diffusion (as in agar-based media) is a tendency for especially non-motile organisms to replicate in the vicinity of their birth.

Growth media can be complex or chemically defined. Chemically defined media is such that its ingredients can be purchased in relatively pure form from chemical supply companies. The advantage of using chemically defined media is better control over growth, potentially better reproducibility, and an ability to slow down organism physiologies in well-defined or diverse manners, e.g., by employing different carbon sources. Complex media, by contrast, contains ingredients that are not chemically pure. Often complex media is quite rich (i.e., eutrophic), though this is not necessarily the case, and especially not if the media has been diluted or obtained from an oligotrophic source (e.g., pond or stream water). Complex media are often easy to produce (i.e., they often can be purchased in a dehydrated form) and can be used to approximate the chemistry of natural environments. On the other hand, natural environments in most cases are more oligotrophic than eutrophic, so highly rich, complex media can be a terrible model for a microorganism's normal, i.e., in-the-wild life styles. In general, what type of growth media is employed will depend on the experiments being conducted as well as their historical context, such as how others have done similar experiments or how one was otherwise trained. It often isn't such a bad idea to spend at least a little time thinking about the suitability of media employed, either as the experimentalist or as a reader of published studies.

Culturing Approaches

Cultures, which generally are populations of microorganisms, can be grown continuously, in batch, or can be serially transferred. The most common approach is batch culture using broth media, though batch culture also can be done using solid media. Generally culturing in batch involve an exhaustion of nutrients or a buildup of waste that at some point terminates organism population growth. In broth, cultures typically become turbid if an organism is replicating at the expense of dissolved nutrients, or less turbid if the organism is replicating at the expense of suspended particles, such as bacteriophages replicating at the expense of intact bacteria. On solid media, organisms grow as colonies, which are piles of cells; as plaques, which are local absences of cells; or as diffuse swarms in the case of more motile organisms. Serial transfer can be from broth to broth, solid media to solid media, or indeed any combination of the two. Continuous growth usually involves broth media, though there are a few instances where organism populations can grow outward seemingly indefinitely on solid media. The most common continuous culture devise is the chemostat, where fresh media is added continuously and at the same rate that culture media is removed. There exist, however, more complex means of establishing continuous cultures, such as the turbidostat, where turbidity is held constant via a more complex monitoring and feedback scheme than what is employed in the chemostat.

In most cases microorganism study involves pure cultures that are generated via a process known simply as pure culture technique. Typically solid media is employed to isolate single cells and their replicative products (i.e., isolated bacterial colonies or virus plaques). Effort needs to be undertaken to avoid contaminating cultures on the one hand (addressed by employing what is known as aseptic technique, as well as by using growth inhibitors against which one's focus organism is resistant). On the other hand, effort is made to minimize culture evolution, especially during stock propagation. By contrast, serial transfer or continuous culture may be employed to encourage evolution, though for these long-term cultures avoiding contamination can be even more relevant. Batch culture, too, may be used to effect evolution, especially when selection is strong, such as for host range shift in parasites when otherwise permissive hosts are unavailable.

Evolution also can be achieved using enrichment techniques that have the effect of culling from community samples those organisms possessing unwanted characteristics (e.g., such as an inability to employ crude oil as a carbon source in the case of isolating bacteria to employ in anti-oil spill bioremediation). Such enrichment can also be achieved starting with clonal populations, especially if mutation rates are substantial. The latter can serve as a means of fine-tuning organism properties for industrial application, e.g., such as toward enhanced production of valuable byproducts of microbial metabolism. It also serves as the basis for biopanning, where organisms with specific binding properties are selected among otherwise isogenic individuals displaying randomly generated protein motifs (e.g., phage display). In short, not only is microbial evolution important, it can be useful as well.

Standard Bacterial Growth Curve

Though not a perfect way to view microorganisms, nonetheless well-mixed, broth, batch growth, especially as it occurs in fairly rich media, is the most common way that growth is handled in the laboratory. In following such growth it is typical to consider what is often described as the standard bacterial growth curve. If nothing else, there are a number of important terms that derive from these growth curves that at least pedagogically that will pop up again and again when discussing issues of microbial growth, ecology, and evolution. These are lag, log (or exponential), stationary, and death (or decline) phases.

Lag phase is a time of biochemical gearing up for a bacterium. During this time protein synthesis, cell repair, and cell increase in size can occur. By definition cell division does not occur, however. Thus, lag phase is a lag in cell division. Lag phase will result, typically, when bacteria are transferred from one environment to another, such as from media that are less rich to media that are more rich, or from storage to conditions that instead are conducive to growth. Experimentally, the lag phase can represent an important lag in the growth of bacterial cultures. That is, there are two phenomena that can serve to delay bacteria from reaching suitable densities for experimental manipulation: the length of their lag phase (which generally is longer the longer bacteria have been kept in storage) and their rate of replication once they have exited their lag phase.

Log phase, a.k.a., exponential growth or exponential growth phase, occurs immediately post lag, assuming that bacterial densities are sufficiently low that further population growth is possible. Log phase is called log phase because log-transformed density increases linearly as a function of time. During this phase bacteria are considered to be both most active metabolically and most susceptible to anti-bacterial agents. Note that typically death is assumed to not occur during log phase. This is not necessarily 100% the case, however, plus within continuous cultures loss to outflow very much occurs (by design!) during log phase. Thus, more generally, log phase can be viewed as a time of bacterial division where, especially, division rates, at a minimum, exceed death rates.

Stationary phase is a time either of greatly reduced bacterial replication rates, bacterial death (or loss) balancing bacterial replication rates, or both. More specifically, stationary phase refers to a time when change in bacterial numbers, independent of predation of bacteria by other organisms, equals zero. Since this change typically is measured in terms of viable rather than total cell counts, stationary phase is a time when any increases in viable bacterial numbers are balanced by decreases in viable bacterial numbers. The simplest stationary phase, especially as seen in batch culture, is a cessation of bacterial replication. In this case bacterial births and deaths can perfectly balance based solely on both quantities equaling zero. In this form of stationary phase, cell susceptibility to anti-bacterial agents tends to be fairly low. Another form of stationary phase is seen in continuous culture. Here bacterial replication can be perfectly balanced by media and therefore bacteria outflow from the bacterial growth chamber. This stationary phase, like as may be seen in batch culture, coincides with an approximate minimum of bacterial nutrient densities and represents that bacterial density that depletes the density of these nutrients to the point where replication, nutrient gain, nutrient loss to consumption, and outflow together establish a steady-state equilibrium. Note nonetheless that bacteria are displaying ongoing replication in this continuous-culture stationary phase and therefore should remain susceptible to the aforementioned anti-bacterial agents. Bacteria may also enter stationary phase due to a buildup of metabolic wastes. In addition, the default metabolic state of bacteria in the wild likely is more stationary phase-like than not, especially absent bacterial predators. This occurs because densities of bacteria will tend to increase in size until resource densities are reduced to the point that bacterial densities cannot, absent further resource availability, increase any further.

Bacteria that are in stationary phase, in batch culture, are less metabolically active than they are in either lag or log phases. Consequently, stationary phase bacteria are less capable of keeping up with the repair of cells and eventually will succumb to death, defined as an inability to resume growth given exposure to fresh media. This "death" may be delayed via various mechanisms of enhanced bacterial durability, the most notable of which is the formation of the bacterial endospore. Bacterial death, like bacterial population growth, typically occurs exponentially. This resemblance is a consequence more of mathematical similarity than to physiological parallels. That is, in both cases numbers at later times are functions of numbers at earlier times. Another way of putting this is that numbers in growing populations are a function of previous numbers along with per-capita birth (division) rates whereas in a declining population numbers are a function of previous numbers along with per-capita death rates. Interestingly, a corollary to considerations of bacterial death phase is the question of just what is bacterial death, since metabolism may continue onward despite "death" plus cells may be able to be more readily revived under certain circumstances versus others. An analogy is that survival of a car accident may differ greatly given one’s distance (measured temporally) from a state of the art emergency room.

There are additional considerations vis-à-vis the standard bacterial growth curve. The first is that bacterial growth, as well as evolution can continue onward even following death phase, with this growth sustained by nutrient release from dead bacteria. A second consideration is spatial structure, as observed in bacterial colonies, microcolonies, and biofilms. There, bacteria may display different phases of growth that are a function not of (or not just of) time but also of position, where bacteria that are more buried beneath other bacteria will have less access to nutrients, may be exposed to greater levels of metabolic wastes. These bacteria therefore may be more likely to be in stationary or death phase than bacteria that are located closer to the surface and therefore closer to fresh media, with these latter bacteria more likely to be replicating. Interestingly, bacteria growing under these conditions, like bacteria in stationary phase, tend to be less susceptible to anti-bacterial agents such as antibiotics than are bacteria that are more activity growing in fresh media.

Koch's Postulates

Part of either doing or studying microbial evolution is an appreciation that microbiology is an experimental science and that experiments can be done well, but also can be executed poorly. The study of microbial evolution consists also of the use of microbiological systems to answer questions that are not necessarily traditionally microbiological. The result is that many of the individuals who study microbial evolution did not necessarily receive their primary training as microbiologists. In fact, even among microbiologists, training varies and often seems to emphasize cutting edge techniques rather than more traditional but hugely important experimental approaches. Thus, in this section I speak specifically to how to do, and interpret, experiments. I start with a more general discussion and then employ what are known as Koch's Postulates along with Molecular Koch's Postulates as examples.

The general discussion focuses specifically on the concept of controls. In experimental science one typically hears of positive and negative controls. A positive control consists of employing conditions and materials that ought to give rise to an observable change in the experimental outcome. This is done so that one can tell that experiments are capable to achieving positive results under at least some circumstances. An experiment that reports negative results, i.e., a failure to obtain an observable change in an experimental outcome, is potentially valid only to the extent that a positive control is included that is capable of showing that an observable change in fact could have occurred under different circumstances. Otherwise without a proper positive control, one must assume that negative results are ambiguous, that is, that little or no meaningful information may be obtained. Note that designing positive controls is not necessarily straightforward and that there certainly are circumstances in which it is valid to employ more than one positive control in an experiment. Bottom line: If you don't see a positive control, then disregard any negative results as they are, effectively, not scientifically valid.

The negative control is employed to show that observable changes in experimental outcomes in fact can fail to occur. That is, positive experimental results need to be validated as more than just what would have happened under any circumstances. Very often one finds that negative controls are performed more often than are positive controls since typically researchers design protocols so that experiments will provide positive results, i.e., changes in experimental outcomes that are a consequence of specific, intentional perturbations. Just as with positive controls, it can be perfectly valid to employ more than one negative control. Furthermore, and perhaps key to effective experimental technique, it is very important in designing negative controls to make sure that the specific phenomenon one is exploring is individually being withheld from the negative control. That is, if one is attempting to show that Factor A is responsible for result X, but employ as an experiment a specimen containing Factors A, B, and C while one's negative control lacks all three of those factors, then that experiment alone cannot be employed as evidence that Factor A is responsible for Result X (since factors B and C could very well be responsible as well). In this case, B and C are considered to be controlled variables, and an experiment's validity is highly dependent on making sure that controlled variables really are controlled, that is, that they remain constant between experimental and negative control treatments.

Koch's Postulates, named for the famed microbiologist, Robert Koch, are a means of demonstrating that a specific disease is caused by a specific pathogen. In this case Factor A would be the specific pathogen and Result X would be the disease. Koch's Postulates formally employs the following steps: A diseased organism is presented and potentially pathogenic microorganisms isolated from it, i.e., into pure culture. These pure cultures are then individually inoculated into healthy organisms. If the organisms become ill with the disease in question, and the same pathogen can be isolated from those organisms as were inoculated into them, then one concludes, "By Koch's Postulates", that the pathogen in question is the cause of the disease. Note in this case the importance of the negative control, which are organisms to which the potentially pathogenic microorganisms were not applied. These control organisms should be treated identically to the test organisms, except for that one difference. Indeed, it can be useful also to employ mock inoculations that are identical to the actual inoculation, except lacking in the test pathogen. Importantly, it can be relevant how similar that mock inoculation is to the actual inoculation, just as one wants to make sure that negative controls and experiments differ solely in terms of Factor A and not also in terms of Factors B and C (etc.). With Koch's Postulates the isolation of the inoculated microorganism from so-inoculated organisms, but not from the negative controls, at least, provides evidence that a disease state correlates with the presence of a specific pathogen.

In this modern, molecular age, a derivative of Koch's Postulations has been developed that is termed Molecular Koch's Postulates. In Molecular Koch's Postulates the goal is not to demonstrate that a specific pathogen is responsible for a specific disease but instead to show that a specific aspect of a specific pathogen is responsible for the ability of that pathogen to cause the disease in question. In other words, once by Koch's Postulates a pathogen has been found to be responsible for a specific disease, then one's next step is to demonstrate how, molecularly, this disease is effected. Note that we can replace the word "disease" with that of "phenotype" and what we have is a very basic description of how especially reductionist experimental biology is done, i.e., that which constitutes the bulk of modern, experimental molecular microbiology.

Molecular Koch's Postulates, not surprisingly, somewhat resembles traditional Koch's Postulates which, in turn, simply represents an ideal of well-controlled experimental and, in this case, molecular microbiology. The steps are simple: Remove a factor from an organism in some manner (e.g., a presumptive virulence factor), show that the consequence is a decline in some manner of phenotypic expression (e.g., decline in pathogenicity), and then show that with reestablishment of the factor the phenotype returns (such as restoration of pathogenicity). Note that for this to be done properly, it is important to show that only the factor in question has been removed (e.g., Factor A). That is, while it is important to show that the specific factor has been removed, it is crucial also to show that only that specific factor has been removed (i.e., and not also Factors B and C). In fact, one can view the last step, adding back the factor, as a means of making sure that the only the factor in question has been altered.

By way of example, let's say that one is attempting to demonstrate that a specific prophage is responsible for a specific phenotype associated with its host such as lysogenic conversion of a bacterium. (A prophage is a bacterial virus that has genetically integrated into a host bacterium, rather than immediately produce new virus progeny, and lysogenic conversion is the display by a bacterium of a phenotype that in fact is encoded by a prophage.) To show this it might be convenient to employ, as negative controls, bacterial strains from which a prophage has been spontaneously lost. To characterize the resulting strain, one might be tempted to show that the prophage really has been lost. That, however, is not sufficient, by Molecular Koch's Postulates, to demonstrate the suitability of these strains for further experimentation, and this is because also crucial is demonstration that only the prophage in question (i.e., Factor A) has been lost from the bacterial host. Making that determination might be accomplished, for example, through various levels of bacterial genome sequencing. One should remain skeptical that only Factor A has been lost, however, since, for example, the prophage might have been lost via deletion of more than just the prophage DNA rather than a more clean excision of just the prophage DNA. This uncertainly thus points to the need for adding back of Factor A, in this case the prophage, where the strain lacking the added-back prophage would serve as the phage-negative control or, without contradiction, the now prophage-containing strain can be viewed as a phage-positive control. It is relatively easy, in other words, to destroy existing phenotypes, while the specific restoration of a specific phenotype typically will require some knowledge of just what previously had been lost.

Generally it is crucially important, in experimental biology, to limit differences between experimental procedures and negative controls. The art of doing science well, however, is to both properly design controls and otherwise to realize that more than one control, or experimental result, may be necessary to effectively link Factor A with Result X. To the student of experimental microbiology, it is crucial to learn to distinguish between claims that experiments are being properly performed, and otherwise demonstrating what is being claimed, and what experiments actually show. Above all, has the experimenter identified and otherwise incorporated the appropriate controls? Indeed, are there possible alternative explanations for results that have not been successfully ruled out? Particularly, when delving deeply into a paper, keep in mind that conclusions are only as valid as underlying logic and experimentation. A study, furthermore, is only complete once all other possible explanations for experimental results have been reasonably eliminated as possibilities. In no instance should science focus exclusively on procedures that serve solely to validate hypotheses rather than the hugely more important need to attempt to falsify hypotheses, which can be accomplished only through the design of properly controlled experimentation.