The vast majority of species on Earth are single-celled. Most of these languish in obscurity – many have never even been named – but some of the relatively few species that have been studied exhibit remarkable abilities.
Many of these are physical: some micro-organisms are amazingly strong; others can hibernate for hundreds of thousands of years or thrive in environments so extreme that they would kill off most other life forms in a flash.
But many bacteria and protists also exhibit behaviour that looks remarkably intelligent. This behaviour isn't the result of conscious thought – the sort you find in humans and other complex animals – because single-celled organisms don't have nervous systems, let alone brains.
A better explanation is that they're "biological computers" with internal machinery that can process information (see our review of Wetware: A Computer in Every Living Cell). Here are some of the most striking examples of this "intelligent" behaviour from the New Scientist archive.
Communication
Bacteria talk to each other with chemicals. They do so for a host of reasons, some of them hard to understand unless you are another bacterium (or a dedicated bacteriologist), but one of the most straightforward is demonstrated by Bacillus subtilis.
If B. subtilis individuals are growing in a food-poor area, they release chemicals into their surroundings. These essentially tell their neighbours: "There's not much food here, so clear off or we'll both starve."
In response to these chemical messages, the other bacteria set themselves up further away, completely changing the shape of the colony.
Decision-making
Many single-celled organisms can work out how many other bacteria of their own species, are in their vicinity – an ability known as "quorum sensing".
Each individual bacterium releases a small amount of a chemical into the surrounding area – a chemical that it can detect through receptors on its outer wall. If there are lots of other bacteria around, all releasing the same chemical, levels can reach a critical point and trigger a change in behaviour.
Pathogenic (disease-causing) bacteria often use quorum sensing to decide when to launch an attack on their host. Once they have amassed in sufficient numbers to overwhelm the immune system, they collectively launch an assault on the body. Jamming their signals might provide us with a way to fight back.
City living
Not only can bacteria be talkative and co-operative, but they also form communities. When they do, the result is a biofilm, most familiar as the thin layers of slime that coat the insides of water pipes, or kitchen surfaces in student residences. They're also found in biological refuges, like the inner linings of human digestive systems – anywhere, in fact, where there is plenty of water.
Many different species live side by side in these "bacterial cities", munching one another's wastes, cooperating to exploit food sources, and safeguarding one another from external threats – such as antibiotics.
Accelerated mutation
Many microbes can accelerate the rate at which their genes mutate. This allows them to obtain new abilities that may be helpful when conditions get tough. This is a risky strategy, since many of the new mutations will be harmful or even fatal and is, in effect, a last-ditch tactic when there's little left to lose.
Examples are legion: Escherichia coli mutates more rapidly when under stress (Science, DOI: 10.1126/science.1082240), and yeast has also been shown to perform the same trick (Critical Reviews in Biochemistry and Molecular Biology, DOI: 10.1080/10409230701507773).
During the early 1990s, researchers suggested that bacteria might have a way to "choose" mutations that would be particularly useful. This idea of directed mutation was extremely controversial, and by 2001 the evidence was stacked against it (Nature Reviews Genetics, DOI: 10.1038/35080556).
Navigation
It's common knowledge that many animals can navigate across vast distances, migrating birds and honeybees being among the best-known examples. But microbes are also pretty good at it.
The single-celled algae collectively called Chlamydomonas swim towards light, but only if it is of a wavelength that they can use for photosynthesis.
Similarly, some bacteria move according to the presence of chemicals in their environment – a behaviour called chemotaxis. E. coli, for example, move like sharks on the trail of blood if a few molecules of food are dropped into their environment.
Another group of bacteria align themselves to the Earth's magnetic field, allowing them to head directly north or south (Science, DOI: 10.1126/science.170679). Known as magnetotactic bacteria, their special ability comes from specialised organelles loaded with magnetic crystals.
But perhaps the most striking feat of microbial navigation is performed by the slime mould Physarum polycephalum. This colony of amoeba-like organisms always finds the shortest route through a maze.
Learning and memory
When the amoeba Dictyostelium searches the surface of a Petri dish for food, it makes frequent turns. But it does not do so entirely randomly.
If it has just turned right, it is twice as likely to turn left as right on its next turn, and vice versa. In some way, it "remembers" which direction it last turned. Human sperm have the same ability.
E. coli goes one better. This bacterium spends part of its life cycle travelling through the human digestive system encountering different environments as it goes. In the course of its journey, it encounters the sugar lactose before it finds the related sugar, maltose. At its first taste of lactose, it switches on the biochemical machinery to digest it – but it also partially activates the machinery for maltose, so that it will be ready for a feast as soon as it is reached.
To show that this was not simply hard-wired, the researchers from Tel Aviv University grew E. coli for several months with lactose, but without maltose. They found that the bacteria gradually changed their behaviour, so that they no longer bothered to switch on the maltose-digesting system (Nature, DOI: 10.1038/nature08112).
Remarkable though these behaviours are, we have probably only scratched the surface of what single-celled organisms can do. With so many still entirely unknown to science, there must be plenty more surprises in store
The secret language of bacteria
16 September 1995 by Elizabeth Pennisi
WHEN Karl von Frisch reported 60 years ago that bees communicate their foraging successes through figure-of-eight dances, he was met with criticism and disbelief. In 1973, this German biologist won a Nobel prize for that work, and it is now clear that bees rely on chemical signals as well as body movements to get their messages across. Chemical communication enables bees to distinguish cousins from strangers, coordinate attacks on hive intruders and, in their own sort of way, be as social as the best human host or hostess.
The idea of bees "speaking" with chemicals raised questions about animal language and cognition that are hotly debated to this day. But at least bees have a rudimentary nervous system and a multiplicity of different cells in their bodies. Never before has the debate spilled over into the world of the humble microbe. Until now.
The chemicals bacteria secrete are no longer being dismissed as uninteresting byproducts of metabolism, as "waste". "Bacteria are spitting things out so they can talk to each other," insists Douglas Kell, a microbiologist at the University of Wales in Aberystwyth. In soil, in the spaces between our teeth, along the stinky edges of hot springs, bacteria are using chemical signals to network with relatives, negotiate with allies and deter enemies - and why shouldn't they? After all, birds, mammals and insects emit pheromones to guide mating, aggression and other social behaviours.
The difference is that bacterial communication is so subtle it has succumbed to scientific investigation only in the past few years, with the advent of better methods for growing and studying bacteria. One key technique is flow cytometry, a technology which allows researchers to tease apart different types of bacterial cells. Another is the use of genetic engineering to identify the genes in bacteria that are crucial to the ability to emit or respond to chemical signals.
Understanding the chemical chatter of bacteria is just the start. Under attack is the whole, deep-seated notion of bacteria as dumb cells that act robotically on the information in their DNA. Chemical crosstalk, claim today's microbiologists, enables bacterial cells to cooperate in ways approaching the complexity of animal communities; to specialise like those of multicellular organisms; to behave socially. At the annual meeting of the American Society of Microbiology, held in May in Washington DC, an entire symposium was devoted to the way bacteria aggregate into specific structures and then take on special roles in this new organisation. Such ventures - once thought nonexistent, or at least rare - may be the norm for many kinds of bacteria, says Julian Davies from the University of British Columbia in Vancouver.
What makes such discoveries all the more newsworthy is the burgeoning problem of drug-resistant infections. As supposedly thwarted microbial pathogens return with a vengeance, producing virulent cholera epidemics and a resurgence of tuberculosis, research into the genetic flexibility which underlies drug resistance has taken on new importance. To date, the focus has been mainly on DNA, and for good reason. Bacteria are unusually adept at swapping and rearranging genes in response to environmental changes and attack by antibiotics. But now researchers are taking their questions beyond this genetic crosstalk. Could bacterial versatility also be caused by their ability to converse using chemicals?
To answer such questions, you first need to know how the chemical language of bacteria influences their genetic language. And to know that, you need to translate the chemical language. As with any social situation, bacterial chatter involves a mix of antagonistic and friendly gestures, and identifying and decoding those gestures has been a long job. Sixty years ago, microbiologists discovered that bacteria can emit proteins - known as bacteriocins - to repel or kill other bacteria. Then in the 1960s, Julius Adler, of the University of Wisconsin at Maddison, discovered that bacteria will move towards or away from a variety of nutrients. But only now are the broader implications of bacterial communication, as well as many of the molecular details, becoming clear.
Microbiologists recognise, for example, that bacteriocins come in all shapes and sizes. Yet most of them work in the same way, penetrating the target organism and disrupting its membrane.
Territorial tendencies
Armed with bacteriocins, in theory a microbe species could stop other species muscling in on its territory and resources. And in practice, this is what seems to happen. Last year, Daniel Smith and Martin Dworkin of the University of Minnesota, Minneapolis, described a vivid example involving two species of Myxococcus. Place these bacteria in the same culture dish, the researchers found, and they will stake out territories and grow as separate colonies. In liquid media, the antipathy is even more striking. One microbe, Myxococcus virescens, ends up greatly outnumbering the other. In both cases, argue the researchers, bacteriocins are at work, but they have yet to isolate the substances used to define the territories.
Bacteria that cannot produce bacteriocins run the risk of being wiped out by competitors. Last year, a team of Spanish and British investigators reported the quick demise of one such strain of lactic acid bacteria. To better understand the fermentation process involved in producing green olives for export, a group led by Jose Luis Ruis-Barba of the Institute of Oils and Fats in Seville, Spain, added several strains of lactic acid bacteria to olive brine and allowed them to ferment. All these strains persisted except one that was unable to produce any bacteriocin. It disappeared within seven weeks.
It may take a crisis to get bacteria talking. In a nutrient-rich medium, for example, Bacillus subtilus will divide in an apparently aloof manner, radiating out from a starting point to create a circular colony. But reduce the nutrient supply and cooperative behaviour kicks in. The growing cells seem to sense, and obligingly move away from, chemicals emitted by their neighbours. The result is reduced competition for resources and a colony that takes on a spiralling pattern, notes Eshel Ben-Jacob of Tel-Aviv University. As the amount of food decreases, long branching arms radiate and spiral away from a central spot. When the nutrients are very low, the branches become quite fine and the colony structure appears to become very well organised.
Ben-Jacob has developed mathematical models that mimic this pattern formation. The patterns generated by his models look most realistic when he introduces data describing the chemical cues produced by the bacteria themselves. This has convinced him that the bacteria are communicating with one another and not just responding to nutrients in the dish.
Other bacteria are constantly on the lookout for a get-together. Vibrio fischeri, the bacteria that make fish bioluminesce, and landlubbers such as Pseudomonas, will emit a chemical - dauntingly entitled beta-keto-caproyl homoserine lactone - to attract compatriots. At a certain point, and because of the signal, such bacteria "sense" a quorum and respond by altering their internal biochemistry and behaviour. For V. fischeri, the result is bioluminescence. But other bacteria respond by turning into highly social colonies akin to multicellular organisms.
Here individual bacteria give up their ability to act independently and take on specialised roles within the colony. At Stanford University in California, Dale Kaiser studies this kind of social role-play in myxobacteria. Myxobacteria gather together for feeding - they prey on dead, decaying or dying microorganisms - and for surviving tough times. When food runs low, they gather in millions, rearranging themselves into fruiting bodies once there are enough of them. Some species form small blobs, others grow into finely branched stalks, and several develop bright colours. Outer cells become a protective, often slimy or cartilaginous, coat for inner cells that transform into dormant myxospores.
The point of the exercise is to acquire collective bulk. "The microorganisms become a macroscopic object that can be moved in the soil," explains Kaiser. Unlike small, undifferentiated colonies, myxobacterial clumps are liable to be wafted by air, or washed by water, to new food sources.
Harder to understand is the biochemistry driving this kind of bacterial specialisation. But even here progress is being made. Over the past five years, Kaiser and his colleagues have identified the chemicals that command the bacteria first to aggregate, then to form fruiting bodies, and in some cases to become spores. The key was to create a series of gerietic mutants, each lacking a gene needed to produce one of the command chemicals. The behaviour of these mutant bacteria told the researchers what each of the command chemicals did.
The next step will be to work out how bacterial cells "know" when to issue chemical commands and how those commands succeed in influencing the behaviour of neighbouring bacterial cells. Tracing the genes involved will be crucial. Myxobacteria have about 6000 genes, few enough, explains Kaiser, to be able to work out what each of those genes is doing.
In practice, however, things can get complicated. The way a bacterial cell responds to a particular chemical signal may depend on much more than just the nature of the chemical. Take the case of Bacillus subtilisis in which a single cell divides to produce two new daughter cells of different sizes. The bigger of the two engulfs the smaller, which eventually becomes a spore, encased and dormant. Why the difference? Richard Losick of Harvard University believes that the position and size of the two cells influences the way they respond to chemical commands. In one cell, chemical signals activate the "spore" genetic program. But Losick still can't explain for sure how it all happens.
But far from being dismayed by the complexity of bacterial chatter, microbiologists are profoundly excited by it. What they believe they are discovering is a chemical and genetic language reminiscent of the one that guides the embryonic development of multicellular organisms.
In embryos, all the signs are that cells discover what kind of tissue to grow into by "tuning in" to chemical signals emitted by other cells. Each cell's location determines how it reacts to a succession of chemical signals, and ultimately whether it switches on genes that turn it into muscle tissue or brain tissue or something quite different. That, at least, is the hypothesis. Discovering which combination of chemical and spatial cues stimulate embryo cells to switch on certain genes and not others is one of modern biology's biggest crusades - and microbiologists are keen to join. Cellular specialisation is "the essence of a multicellular organism", says Losick, and the message of microbiology in the 1990s is that bacteria do it too.
Sometimes that specialisation even takes the form of a "partnership" between species. In hot springs, microbial mats develop as different types of bacteria come together. The cells grow and divide, eventually building up tiny mushroom-shape bodies separated by channels of flowing water. Bacteria on the edges of these channels use the water's dissolved oxygen. A few also have whip-like structures, possibly for circulating water. But at the core of the "mushrooms" lies a different species, one that can survive without oxygen. Nutrients, but not oxygen, are transported to these inner cells, according to research by Richard Castenhols at the University of Oregon in Eugene. Set up like this, he says, the mat "begins to resemble a primitive multicellular organism".
Some microbiologists push the case for bacterial sophistication further. "What goes on in bacteria is not fundamentally different from what goes on in human beings," says Adler. "Social interactions, multicellularity, it's more the rule than the exception with bacteria," says James Shapiro, from the University of Chicago. "Everywhere we look for it, we see it going on." Shapiro even uses the term "sentient" to describe the way some bacteria mutate in an apparently nonrandom manner.
But at this point, sceptics tend to weigh in. Social behaviour includes emotional responses, something bacteria lack, argues Mitchell Sogin, a molecular evolutionist at the Marine Biological Laboratory in Woods Hole, Massachusetts. How, he asks, can one seriously think of bacteria as social creatures when they lack the ability to recognise - and consequently look after - kin? When they have neither the nervous systems nor the cognitive ability to "respond" to cues of genetic relatedness?
The answer, it seems, is to give bacteria the benefit of the doubt, and plenty of it. Microbes do, after all, prefer to form colonies with their own kind: could not that be seen as a function of kin recognition of some sort? And bacteria do seem "aware", in a chemical sense, of the world around them: could not that be construed as evidence of a "nervous system"?
At the University of Wisconsin at Madison, Adler, clearly believes so. He has developed a way to grow bacterial cells that are big enough to have their membranes probed by electrodes. With these giant cells, he hopes to show that bacteria function like one-cell nervous systems. Davies even considers bacteria "emotional", in the sense that like people they have needs which they act to satisfy.
Repressed memories?
The idea of microbial emotions is probably not one whose time has yet come. The same, though, may not be true of microbial memories and learning.
Social animals, whether they are birds, humans or other mammals, pass traditions and skills from one generation to the next through culture, education and imitation. Kell suggests that similar kinds of information transfer can occur between generations of bacteria. This happens when certain chemicals released from a "mother" cell are absorbed by a "daughter" cell. The antibiotic streptomycin is one example. As well as killing off competitor bacteria, streptomycin galvanises its own bacterial colony by activating genes needed for growth and producing more streptomycin. A daughter cell that absorbs streptomycin is in a sense absorbing the information it needs to "work out" how to make this potent chemical. "There's your memory," says Kell, "in one molecule."
When von Frisch's work hinted that bees can communicate in an intelligent manner, it took decades of experiments for others to begin to agree. Achieving that kind of recognition for microorganisms may prove even more difficult. "Most people are really hung up on this idea that intelligence is limited to humans," says Shapiro. "We assume that bacteria are rather simple, rather primitive and rather limited. I think that's to our detriment.
Microbes on the move
03 March 1990 by PHILIP POOLE
Imagine trying to swim through quick-setting concrete. For a bacterium, swimming through water is a feat of similar proportions. To an organism less than a thousandth of a millimetre long, water is extremely thick, and at this size, viscosity is the biggest obstacle to a microbe's movements. Despite living in a world of treacle, bacteria are not sluggish: in our laboratory in Oxford, we have seen them sprint through water, covering more than 100 times their own length in no more than a second.
Scientists have been fascinated with the Olympian swimming abilities of bacteria since the 17th-century Dutch microscopist Antonie van Leeuwenhoek first looked down his home-made microscope and discovered 'little animalcules'. Many have since been lured into looking for the answers to two questions: how bacteria swim and why.
Why bacteria swim is a question of the need to seek out nutrients and a suitable environment in which to grow and multiply. Like any other organism, a bacterium must respond to its environment and adjust to the stresses it encounters there: to do this it must be able to sense its surroundings.
Although bacteria are among the simplest autonomous forms of life, they show a remarkable ability to sense their environment. They are attracted towards materials they need and are repelled by harmful ones. The ability of bacteria to swim towards or away from a particular stimulus is called taxis. Two renowned German scientists, Theodore Engelmann and Wilhelm Pfeffer, first demonstrated the phenomenon in the 1890s with a series of classic experiments. Engelmann worked with the bright red, spiral-shaped bacterium Rhodospirillum rubrum, which traps light energy to power photosynthesis. The bacterium's colour is the result of carotene pigments, which work alongside chlorophyll in the process of photosynthesis. In his experiments, Engelmann shone spots of light on the bacteria and found that they congregated in the light. Such movement towards light, or phototaxis, ensures that the bacteria move to an environment with the optimum conditions for photosynthesis.
In another set of experiments, Engelmann showed that aerobic species of bacteria, which need oxygen for respiration, accumulated around air bubbles on a microscope slide - so exhibiting positive 'aerotaxis'. He went on to discover its opposite, negative aerotaxis, by looking at a species of Spirillum that is harmed by high concentrations of oxygen. This Spirillum was clearly repelled by air bubbles.
While Engelmann was testing responses to light and air, Pfeffer watched bacteria swim along a gradient (from a weaker solution to a stronger one) of a chemical attractant, and so demonstrated 'chemotaxis'. His technique was disarmingly simple: he dipped capillary tubes full of solutions of various chemicals into a small beaker of bacteria and watched the response of the microbes. When the chemical was sugar, the bacteria accumulated around the tip of the tube where sugar was diffusing out.
From then on, few scientists took much interest in taxis until the 1960s, when Julius Adler, of the University of Wisconsin, Madison, rediscovered the German experiments. One of Adler's main interests was butterflies. He wanted to study them as a model of neuronal behaviour, but a butterfly has around 105 neurons and is extremely difficult to work on. Adler decided to work on the chemotactic behaviour of bacteria instead, because he thought it might serve as a model of how a single nerve cell responds to its environment. He chose to work on the gut bacterium Escherichia coli because it was the best known of all bacteria. This still applies today and much of what we now know about chemotaxis comes from studies of E. coli and the closely related bacterium Salmonella typhimurium.
Adler made rapid progress once he had devised techniques to measure chemotaxis reliably. He improved on Engelmann's capillary assay and made it quantitative by counting the number of bacteria that swam into the capillary full of attractant. He could then compare chemotactic responses and correlate them mathematically, rather than simply scoring a response as strong or weak. In addition, he developed a simple plate assay for chemotaxis. This involved placing bacteria in a soft agar suspension containing a low concentration of attractant. If the bacteria are capable of chemotaxis, they swim out in a ring as they consume the attractant closest to them. Although the assay gives only a qualitative answer to whether the bacteria are chemotactic, it allows rapid screening of mutant bacteria whose chemotactic response is impaired. Isolating mutants is often a crucial step in microbiology because the mutant can be compared with the parent, helping scientists to identify the nature of the mutation.
With these techniques, and with microscopes more advanced than anything Engelmann or Pfeffer had, researchers soon discovered that bacteria swim very erratically; short smooth runs are followed by brief tumbles, after which the bacteria shoot off randomly in a new direction. They showed none of the smooth-flowing movements of the cells of higher animals, such as the oozing of an amoeba, and it was difficult to see how the apparently random swimming of bacteria could result in a directed movement. The question researchers were left with was: How can a swimming pattern that consists of only two modes, smooth runs and tumbles, lead to movement towards a specific goal?
An essential clue was the observation that when a chemical was added to a suspension of bacteria, the bacteria swam smoothly for much longer periods. This means that when bacteria encounter a gradient of an attractant they swim in long straight runs and the tumbles are suppressed. Whenever bacteria begin to move away from the chemical, there is a greater chance of a tumble, which results in a random change of direction. The net effect is to make runs towards the chemical longer and runs away from it shorter. Attractants give direction to a bacterium's movement simply by biasing the length of the run. Repellents produce the opposite response: as bacteria detect an increasing concentration of the repellent, their tendency to tumble increases, leading to longer runs as they move away from the chemical. Phototaxis and aerotaxis follow the same pattern of swimming behaviour.
Understanding how the swimming pattern of a bacterium changes in response to a gradient of a chemical attractant does not tell us anything about how the bacterium detects the gradient. There are two possible mechanisms. First, a cell might compare the concentration of a chemical at the front and back of its body simultaneously. If the concentration is higher at the front then the cell is moving up the concentration gradient. The second way might be to measure the concentration at one instant and again after a brief interval. In this case a cell must retain a memory of the initial concentration.
Robert Macnab, of Yale University, and Dan Koshland Jr, of the University of California at Berkeley, reasoned that if bacteria do compare concentrations at different times, then when they quickly mixed bacteria with an attractant, the cells would behave as if swimming up a concentration gradient and swim in long smooth runs. If, on the other hand, a bacterium detects a chemical gradient by measuring it simultaneously at two points on the cell body it would not respond to the jump in concentration because it is surrounded by a uniformly higher concentration of the chemical. When Macnab and Koshland did the experiment they found that the bacterium did respond by swimming smoothly. They had fooled the cell into 'thinking' it was going up a concentration gradient of attractant and had shown that bacteria detect a gradient by comparing the concentrations at different times.
The most spectacular recent advances in understanding taxis have been in unravelling the molecular mechanism by which bacteria detect chemicals. Chemoattractants do not need to enter the cell to produce a response. Instead, they bind to protein receptors, called methyl-accepting chemotaxis proteins (MCPs), which lie within the cell membrane. There are four different MCPs, each specific to a certain group of attractants. The proteins span the cell membrane, so they communicate with both the outside world and the inside of the cell. This enables them to function as antennae that detect an attractant or repellent outside the cell and then signal this to the inside of the cell via a change in their shape. Because the proteins are anchored in the membrane, however, they cannot transmit the message across the inside of the cell. This task falls to another set of small soluble proteins simply known as 'Che' (short for chemotaxis) proteins. Only in the past two years have researchers discovered some of the details of how these messengers work.
Molecular biologists at several laboratories have discovered that at least two of the Che proteins, called CheA and CheY, can be phosphorylated - that is, a phosphate group can be added to or detached from them. This is crucial to their function as messengers. These proteins can also interact with the MCPs in the membrane of the cell. Exactly how they interact is an area of intense research, particularly because scientists would like to know which, if any, of the Che proteins bind to the MCPs. A strong candidate is the CheA protein, which can phosphorylate itself as well as the CheY protein. One possibility is that the MCPs regulate the phosphate level on CheA, and CheA in turn regulates the phosphate level on CheY. This is almost certainly an oversimplification of a very complex process and several groups of researchers are racing to try to work out precisely how the process works.
What is clear is that when the MCP binds an attractant it signals its presence across the membrane and reduces the phosphorylation of the CheY protein. The CheY protein then diffuses rapidly throughout the inside of the cell and binds to the 'motor' that powers the swimming of the cell. This tiny motor is attached to a spiral-shaped propeller, called a flagellum, which is rotated outside the cell to create thrust . The presence of CheY makes the motor rotate smoothly so that the cell moves in long smooth runs, characteristic of an encounter with an attractant. In contrast, when the bacterium encounters a repellent, CheY would tend to be phosphorylated, interrupting the rotation of the motor and causing the cell to tumble. So, depending on the degree of phosphorylation, the CheY protein signals to the motor either to rotate smoothly or to tumble.
MCPs are not simply senders of signals. They have a second function: adaptation. Bacteria adapt to the presence of an attractant: after two or three minutes of smooth swimming the cells return to their normal tumbling pattern. There is a good reason why a cell adapts to the presence of an attractant such as a nutrient. If it does not, then the cell will be stimulated indefinitely by just one compound and will not respond to other attractants. Bacteria need a range of nutrients to grow and so they must respond to a wide range of attractants. MCPs make this possible: as the MCP binds an attractant, it is chemically modified by a process called methylation, in which a methyl (-CH3) group is added to the protein. As methylation proceeds, the strength of the signal from the MCP to the Che proteins fades; essentially, methylation dampens the signal from the MCP to the Che proteins. Only an increase in the concentration of the attractant will produce a continued signal from the MCP to the Che proteins and promote movement of the bacterium up a concentration gradient.
The degree of methylation is the 'memory' of Macnab and Koshland's experiment. It is a record of the past concentration of attractant. By comparing the current level of binding of attractant to the MCP with the level of methylation, bacteria can determine whether they are travelling up or down a chemical gradient. It also means that the bacteria will not be trapped by any single attractant; instead they will respond to it for two or three minutes and then adapt, leaving them free to respond to other attractants.
This is the mechanism of the classic form of chemotaxis. There are other mechanisms by which bacteria appear to show at least limited chemotactic responses. Judy Armitage and her colleagues at the University of Oxford have shown that the photosynthetic bacterium Rhodobacter sphaeroides responds to several organic acids and simple ions even though it does not have MCPs and probably also lacks Che proteins. For this bacterium to respond, the attractants must enter the cell and be metabolised, but it is still unclear which metabolic products provide the signal to the flagellar motor. The important feature of this mechanism is that the cell may be sensing its own metabolic state without the need for complex receptors, signalling systems or adaptation. Just what role this simple form of chemotaxis has is unknown, but it may occur in many bacteria that live in environments short of nutrients. It provides a mechanism by which energy-starved bacteria can respond to the nutrients they require without the energetic burden of producing the MCP system.
Bacteria employ chemotaxis in the way that higher animals use sight and smell, to sense their environment. They actively avoid some heavy metals and solvents, which are poisonous, while seeking out amino acids and sugars. Exactly what attracts or repels a bacterium depends on the individual species. The movements of bacteria can have a large impact on other organisms. For instance, the ability of nitrogen-fixing Rhizobium bacteria to form nodules on legumes may be influenced by chemotaxis. The plants, and the animals that eat them benefit from the symbiosis between Rhizobium and legume. On the other hand, the ability of spirochaetes and Salmonella to invade their hosts seems to be enhanced by their chemotaxis. It is a sobering thought that when you eat a piece of chicken infected with Salmonella, it is the attraction to the bacteria of compounds released from the wall of your gut that helps to bring on the bout of food poisoning.
[top]
The microscopic motor that moves cells
The bacterial flagellum is the spiral 'tail' that generates thrust and pushes the bacterium along. It consists of a single type of protein called flagellin, which spontaneously assembles to form a rigid helix. The flagellum is only 20 to 40 nanometres across but can be several micrometres long. Some bacteria have only one but others, such as the gut bacterium Escherichia coli, have as many as 8 or 10 distributed around the cell body. In the case of E. coli, the flagella are brought together as a single bundle when the bacterium is swimming smoothly, but when one or more of the flagella interrupt the smooth rotation, the whole bundle flies apart and the cell tumbles.
The flagellum lies outside the cell, but it is powered by a motor that sits in the cell membrane. The word 'motor' is used deliberately: this is the world's smallest motor, at around 20 nanometres across. Unlike the more familiar motors of human technology the bacterial motor is powered by protons rather than electrons. The bacterial flagellar motor is remarkable in another respect - it causes the flagellum to rotate at around 100 hertz. Many other organisms have flagella or cilia, but these whip them back and forth, rather than rotate them as E. coli does.
Experiments with E. coli tethered to cover slips showed that the motor rotates both clockwise and counterclockwise. The direction of rotation is not random: during periods of smooth swimming, the flagella rotate counterclockwise, and during tumbles they rotate clockwise. This explains the switch from smooth swimming to tumbling in E. coli. Just before a tumble, one or more of the individual flagella that compose the bundle switch to clockwise rotation and the bundle flies apart.
The tendency of each individual flagellum to switch direction is determined by the relative levels of the CheY proteins and the phospho-CheY proteins. Phospho-CheY is sometimes called the 'tumble generator', because the higher its concentration, the greater the chance of clockwise rotation and a switch to tumbling motion.
Now you see it, now you don't
Bacteria swim so fast and so erratically that measuring their movement is a formidable problem. Watching them under a microscope provides a general picture of movement but provides limited information about how fast the bacterium moves, and how often it stops or changes direction.
The advent of computerised image analysis in the past decade has solved many of the problems. In this process the image from the microscope is transferred via a video camera to a computer. The computer receives the video signal and transforms it into a set of numbers; the image can then be represented by a series of two-dimensional coordinates in the computer's memory.
Individual bacteria are identifiable as darker shapes in a lighter field, so the computer software searches for an area of defined size and darkness in its memory map. Some image analysis systems search the entire image and identify all the bacteria and then draw an outline around each one.
Another approach, which we have used in Oxford, is to identify and lock on to one bacterium. The video camera sends a new image to the computer about 50 times a second, allowing us to track the bacterium as it moves in each successive frame. The history of the cell's movement for every fiftieth of a second can be stored in memory, and provides a wealth of information - on its speed, turning angles, stopping frequency and even the curvature of the path it has swum.
With such a tool, it is possible to compare the movements of bacteria under different conditions, analysing a cell's path after introducing a chemoattractant, for example. This allows scientists to understand what happens to the swimming pattern of cells when they are stimulated by a chemical.
Dr Philip Poole researches in the microbiology unit at the University of Oxford.
