X. Of mice and men – why men are better (models)
We tried to test the original hypothesis, that Toxo positives give up earlier than Toxo negatives in a struggle, not only by using questionnaires given to people, but also by experimenting on animal models. However, when planning the latter, we committed a basic error. We planned to put together two male mice, one Toxo positive and the other Toxo negative, and observe which one of them first surrendered the fight. When two male mice, thus far kept separate, are put in the same cage, they begin fighting almost immediately. If we don’t separate them in time, the stronger will literally stress out the other male to death in a matter of days. We, however, selected male mice which had not been raised separately, but in large groups. If a mouse’s group is large enough, no hierarchy forms between the males. The males don’t fight among themselves, not even when we pick out two and place them in a separate cage. Hence, when we put these Toxo positive and negative mice together, they pretty much ignored each other. As a result of this unpleasant mistake, our otherwise carefully prepared experiment didn’t work out. Seeing as it was only one of many parts of Štěpánka Zitková-Hrdá’s undergraduate work, we never repeated it. Sustaining 60 males, each in a separate cage, is an expensive matter. It’s another of the topics we should get back to. Especially since it’s a fairly simple experiment, so long as we avoid the mistake we made the first time.
Aside from the experiment with the hot plate, we also used the so-called tail-flick test to study the effect of latent toxoplasmosis on the reaction time of infected mice. This standard test observes how quickly a mouse pulls away his tail when its base is subjected to heat. The mouse is kept in a short polypropylene tube, and a ray from a heated light bulb is aimed at the base of its tail with a magnifying glass. As soon the mouse feels the heat, it quickly jerks away its tail. Using a photocell shadowed by the base of its tail, we can quite accurately determine how soon after the heat stimulus (when we switched
Box 51 Why isn’t the Earth covered by a continuous blanket of mice?
Really, it’s a mystery. Females of all known species – except, perhaps, for humans in most modern, developed countries – have an average of two (though usually more) offspring in their lifetime. It follows that the population should continuously increase. So how come the Earth isn’t covered by a single continuous blanket of mice (or platypuses, or even tapeworms, in which the female can produce several thousand eggs a day over a period of five years)? A species population can temporarily increase or decrease, but in the long term it stays about the same. Maintaining the same population size in an environment of ceaseless, unpredictable change, requires regulating either the growth or mortality rate in relation to population size. Nature applies both types of regulation, based on two kinds of negative feedback (see Box 52 The feedback loop, and why scientists cherish it). The first type, called chemostatic regulation,occurs in species whose population growth is restricted by the availability of a limited resource, such a food or shelter. If the population happens to increase, each individual will get less of the resource, which limits the speed of the population’s increase. Meanwhile, the mortality rate remains fairly constant, so the population gradually decreases. The second type of regulation is turbidostatic.Turbidostatic regulation acts on species whose mortality rate depends on the activity of predators or parasites. If a chance fluctuation increases a species population, then the predators and parasites of that species will also multiply, and once more decrease its population. When the species population is decreasing, its predators and parasites (who are also subject to chemostatic regulation) begin to die out, which lowers their effect on the size of the regulated population. The existence of two types of population regulation has a number of serious ecological and evolutionary implications. For example, it enables two species which require the same resources to live side by side over long periods of time (if one is regulated turbidostatically and the other chemostatically). The evolution of species that are regulated chemostatically aims towards more efficient use of resources (a so-called ; for example, scrounging more energy out of a molecule of sugar). In contrast, turbidostatically regulated species steer towards a faster use of resources (an r-strategy; the ability to utilize more molecules of sugar per minute) (37).
on the light bulb) the mice jerked away its tail. Aren’t we natural scientists just like playful children? From a technical aspect, we conducted this experiment very well, but it turned out that there was no statistically significant difference between the Toxo positives and negatives in withdrawal reflex.So why is there suddenly a box about the reproduction of mice, which has nothing to do with the previous discussion of reaction times? Is there a reference to it in the next experiment – was the box misplaced by a printer’s error? Far from it. This box is unrelated to anything else – I just needed somewhere to brag about my very first discovery in the field of evolutionary biology, the theory of turbidostatic and chemostatic selection (37). I came upon the theory in my first year of college, when originally trying to illuminate the cause of aging in multicellular organisms. Unfortunately, I did not succeed – a shame, since pretty soon the information could come in handy.
Box 52 The feedback loop, and why scientists cherish it
When a nuclear physicist, a molecular biologist and a psychologist are discussing their work (and don't happen to be trash-talking crazy bosses, incompetent lab workers or stingy tax payers), they often hard-pressed to find a topic of common interest. The goal of any scientific field is to understand the workings of the world we live in – in other words, to create a model, the behavior of which corresponds with the behavior of the real world. Each scientific field has selected part of the world to study, and each field has a distinct, limited arsenal of elements for creating its models (see also Box 27 Why to model in science, and what can and can’t be modeled How are hypotheses tested in science) Scientists treat these elements like black boxes.
. They know how the elements behave; therefore they can deduce what set of ouput signals they will obtain from a set of input signals. But the reason behind the transfer, the mechanism which makes input into output signals (i.e., the contents of the black box) does not interest them. Indeed, this approach is key to the success of science: what one scientist sees as a black box, another pursues as the object of his research. The research topic of a psychologist is aggression; his black box is the human brain. The neurophysiologist's topic is the brain; his black box contains each type of nervous cell. These cells are studied by the cellular biologist; his black box is an arsenal of various molecules and macromolecules. These molecules are dissected by the chemist, whose black box includes atoms and atomic particles – which, in turn, are studied by the physicist. Scientists of every field can simultaneously research on different levels. Chemists don't have wait for the questions of physics to be resolved; biologists don't rely on the discoveries of chemistry; and so on. If the physicists find that the atom works differently than they had thought, it most likely won't have too big of an impact on chemistry. The chemists may switch out some of the contents of their black boxes, but their actual theories and hypotheses generally remain unaffected. Scientists are used to occasionally changing some more or less important contents of their black box. Truth be told, they often don't even register the change, or it doesn't really impact them on the professional level. (Sometimes it does: changing the contents of a black box can change the transfer of input to output signals of a particular element, which may affect that element's behavior and role in a system.) Therefore, a scientist should, at least out of the corner of his eye, follow the progress of other scientific fields, and time to time look through the most important multidisciplinary journals, like Science or Nature
Something which is prominent in all of science, and happily discussed among scientists of seemingly unrelated fields, is the feedback loop. Every scientist is familiar with it, and looks for it in his systems. A feedback loop is the direct or indirect effect of a certain element’s output signals on its input. There are two types of feedback loops: negative and positive. The negative feedback loop forms the basis of any kind of regulation. In a negative feedback loop, the signals from the output damp down the input. Let’s say that the output signal happens to grow stronger. For example, when a steam engine speeds as it rolls down a hill, then the speed of its centrifugal governor increases. The axis of the centrifugal governor is aligned with the axle(s) of the wheels. As the centrifugal governor spins, momentum moves the attached flyballs out and up, a mechanism which opens the valve on the steam cylinder, lowering the pressure of the steam within. This slows the turning of the wheels, along with the speed of the locomotive and of the rotating balls. As the momentum of the flyballs decreases, the balls move back inwards and downwards, reducing or closing off the throttle valve from the steam cylinder. Now the pressure in cylinder begins to decrease, etc., etc. A negative feedback loop works in nature on myriad levels, and scientists of all fields encounter it. A positive feedback loop involves a positive effect of the element’s output on its input. While a negative feedback loop gives a system stability, a positive one causes instability, and triggering its self-increasing cycle often leads to the destruction of the system. For example, in one round of the experimental game Public goods (see chapter XII), several players randomly contribute little to the public pool. Many of the other players feel cheated, and in the next round they also contribute little to the public pool . Therefore, in the third round even more players will feel cheated, so these will also stop contributing. Over the course of several rounds, cooperation may completely fall apart, with nobody contributing anything to the common bank. A technical application of the positive feedback loop is the atomic bomb; in contrast with the negative feedback loop which operates in the nuclear reactor of a power station. Scientists like to look for negative and
positive feedback loops, for the discovery of any kind of feedback loop can ofteexplain the existence of interesting phenomena. In and of itself the feedback loop is so interesting that one is almost always able to publish its discovery in a high-ranking journal. Again, scientists of any field – as well as the reviewers of the respective manuscript – generally agree that discovering a new feedback loop is valuable and worthy of being published.
Another result of our mouse studies was our discovery that at least some behavioral changes of Toxo positive mice, such as a delayed reaction in jumping off a hot plate, decrease over time after the infection. Eventually, perhaps two months after infection, the changes disappear completely (18). This brings us to a crucial general conclusion. The majority of hitherto published research on the effect of toxoplasmosis on the behavior of lab mice does not study the effects of long-term latent infection, but rather observes the passing behavioral changes which accompany acute toxoplasmosis, or the phase just after. This, of course, is an important distinction. When a mouse is sick, it’s not surprising that it behaves differently than when healthy. About a dozen works published by British authors in the 80s may not actually describe the manifestations of Toxoplasma’s manipulatory activity, but rather the late symptoms of acute toxoplasmosis. When testing the manipulation hypothesis, one must wait until the manifestations of acute toxoplasmosis pass, and then take note primarily of those signals which grow stronger with time after infection, not weaker.
The model of the lab mouse or practically any other small rodent isn’t particularly suitable for studying the manipulatory activity of Toxoplasma. A much better model for the study of latent toxoplasmosis is the brown, or Norwegian rat, which Joanne Webster and Manuel Berdoy began using for their studies in the 90s in Britain (Box 53 How Oxford scientists study the effect of toxoplasmosis on the behavior of brown rats), and
Box 53 How Oxford scientists study the effect of toxoplasmosis on the behavior of brown rats
The same year that we published our first work describing the effect of Toxoplasma on human behavior, a British team also published the first article from their series of studies detailing the effect of Toxoplasma on the behavior of brown rats (38) (39). The subject of their study was a colony of brown, or Norwegian rats (Rattus norvegicus) raised in part-natural conditions, not the lab rats (white laboratory brown rats) that are used in most studies. Joanne P. Webster, the main author of recent publications from their studies, revealed to me that their team began researching the manipulatory activity of Toxoplasma by accident, like we did. Originally she wanted to study a completely different parasite found in brown rats, the bacterium Leptospira icterohaemorrhagiae. It turned out, however, that only a minimal number of the animals were infected with this parasite. In contrast, a large number could be diagnosed with toxoplasmosis. In her first studies, Webster showed that the infected brown rats were more easily and sooner caught in traps than the uninfected rats, because they are not as afraid of unfamiliar objects and are more active. The most famous finding of the Oxford group is the phenomenon of “fatal attraction,” in which infected brown rats seek out the scent of cats (see Box 31 What our British competitors discovered on animals, and how we will soon beat them).
which have a much greater lifespan than mice. Another preferable model, which we use in our experiments, is the human. For brown rats and humans, acute toxoplasmosis is only a brief episode, considering their lifespan. If there are behavioral changes after this period, they will likely be manifestations of latent toxoplasmosis, not enduring manifestations of acute toxoplasmosis. In this sense, man is a particularly useful model organism. While he is not the natural host of Toxoplasma, from which the parasite could get into the stomach of its final felid host – at least not today – the human has a long lifespan, and his period of acute toxoplasmosis lasts only about a month. Over the rest of his fairly long life, perhaps another 70 years, he should not exhibit the effects of his past mild illness. Any behavioral changes we observe about two years after infection will most likely be the manifestations of latent toxoplasmosis and the possible manipulatory activity of the parasite, rather than the side-effects of his past bout of acute toxoplasmosis.
Nevertheless, we must realize that it makes no difference for Toxoplasma, that the behavioral changes it impacts on an infected mouse are transitory. The probability that the mouse would live more than two months in the wild is small anyways. So maybe it wasn’t that great an error when, in the 80s, Hutchinson and his colleagues described the behavioral changes of mice infected with Toxoplasma, changes which, in the light of our current knowledge, seem to be the manifestations of acute rather than latent toxoplasmosis. Still, when testing the manipulation hypothesis or researching the possible effects of latent toxoplasmosis on humans, it is better to use a long-lived host as a model. Testing on an organism with a long lifespan means that there is less of a risk that we could confuse transitory behavioral manifestations of acute toxoplasmosis with the specific manifestations of the parasite’s manipulatory activity in its latent phase.
A few more words on the course of toxoplasmosis in mice. Generally it is presumed that acute toxoplasmosis has two phases. The mouse begins to show signs of sickness about a week after infection. At this point the disease has its first peak: the mouse quickly loses weight, its fur puffs up, it grows listless, it stops eating, and, in the case of an infection which is stronger or of a more virulent strain, the creature dies. Ten days after this phase, survivors begin to gain back weight, and seem to get well. However, about 30 days after infection comes the second peak of the disease, and once again some of the mice die. Deaths in the first peak are associated with the reproduction of tachyzoites in the tissues of infected mice, whereas the second peak is associated with the point of greatest reproduction rate of tissue cysts in the brain. Mice which are weaker or infected with more parasites, die in the second peak of encephalitis.
Fig. 27 An approximately 20 month-old mouse infected with Toxoplasma. The control mice did not show any signs of deteriorated health.
It’s interesting, and perhaps important for us long-lived Toxoplasma hosts, that later there appears a third phase of toxoplasmosis. The following result we, again, obtained by accident. It all began when I was loath to kill the mice after finishing the ethological experiments. Even though it’s against the rules for working with laboratory animals, I sometimes let them live out the rest of their lives in the mouse husbandry (Box 54 How and why I don’t always follow the rules for working with lab animals). It wasn’t entirely a matter of my softhearted nature – what if I could still use them at a later date, maybe to infect other experimental mice. At worst, my colleagues could take the control mice and feed them to their snakes or monitor lizards. So it happened that I kept some of the mice in the husbandry for a year or more. After this period, it turned out that the mice that had seemed to be completely healthy – which, two months after infection, had even regained the weight they had lost during the acute phase – in reality were far from healthy. The older ones had skin lesions, shed fur, and, in a number of cases, became paralyzed or blind. No such defects were observed in the control, uninfected mice (Fig. 27). In nature, a mouse doesn’t usually get the chance to grow old, so only in lab husbandries could it become apparent that toxoplasmosis is far from harmless – as is often thought.
Box 54 How and why I don’t always follow the rules for working with lab animals?
Once an experiment is finished, lab animals are supposed to be killed in a humane way. They may not be used again, even when they were previously in the control group and nothing was really done to them. Clearly one cannot test on animals which have already been subjected to pharmaceutical or surgical procedures, given, perhaps, some unusual food. It might not affect the new experiment, and it is even possible that we could prove the existence of a certain phenomenon more effectively; but no one (including ourselves) would be able to easily replicate the results. If the animals in question are part of an ethological study, and only their behavior was observed, the rule that they cannot be used in another experiment becomes fairly absurd. So if the animals had not been subjected to anything, I often broke the rule on having to immediately sacrifice them, and always tried to find a way to use both the uninfected and infected animals. I don’t believe that my toxoplasmosis – which causes lowered strength of the superego (willingness to comply with rules and social norms) – was responsible for this. Rather, if lab animals must die for my experiments, let it at least be for the greatest possible avail.
It would certainly be important to see if something similar happens in humans. Doctors presume that toxoplasmosis acquired as an adult poses no health problem, at least not Toxoplasma from Europe or North America. On the other hand, in South America, where several atypical strains of Toxoplasma circulate, toxoplasmosis is often accompanied by serious symptoms (Box 46 There’s no Toxoplasma like Toxoplasma). But in the light of our chance observations on aged mice it seems that the negative effects of toxoplasmosis could manifest even in species in which this was hitherto unexpected, maybe even in man. Based on the symptoms of the aged mice, I’d guess that an autoimmune disease was to blame. I don’t wish to foretell evil, but indirect proof of a negative effect of toxoplasmosis on the health of older organisms exists even for humans. When studying the prevalence of toxoplasmosis in different age groups, one often finds that the percentage of infected people at first rapidly increases, but later, in the older age groups, the growth slows. In the Czech Republic, a large number of children becomes infected before the age of nine. It’s almost certainly related to their playing in sandboxes or putting various unsanitary objects in their mouth, eating dirt and so on. After this sharp increase during early childhood, the prevalence of toxoplasmosis in the population grows more slowly. In women, there occurs one more sharp increase between the ages of 24 and 34. It’s likely that this second peak is related to the fact that women are establishing families during this time, and begin cooking more – they may, for example, become infected through scratches in their skin when preparing raw meat, or perhaps by tasting a mixture for meatballs. After the age of thirty, the frequency of latent toxoplasmosis in the population grows more or less constantly until the age of 60. After this, growth slows or comes to a stop, and in the oldest age group the prevalence of toxoplasmosis is lower than in the younger age groups. Of course it is possible that older people are less likely to become infected. Older people are generally more conservative in their habits, so if they have not yet become infected, they are unlikely to do so now (unless in their old age they decide to return to playing in the sandbox). In some of the infected people, the level of Toxoplasma antibodies decreases, so they appear to us as Toxo negative. That is a more optimistic scenario. A more pessimistic scenario says that Toxo positive persons never reach this older age. I’m not sure which is more likely. I do know, however, which scenario would have more serious medical implications, and so I believe that greater attention (some, at least!) should be paid to the effect of latent toxoplasmosis on health.
Once more I’d like to return to the question of whether it’s suitable to use man as a model. For Toxoplasma, man is basically a dead end – when the parasite infects a mouse, it has a good chance of getting into a cat, but when it infects a human, the parasite’s path to its definitive host is likely forever blocked off. It is therefore relevant to ask (and the reviewers of our manuscripts the vexing fault-finders that they are, often do like to demand) whether it isn’t better to experiment on the natural hosts of Toxoplasma, on mice and brown rats (or other rodents). It may seem strange, but I think that even in this aspect man may be a better host model than a mouse or brown rat. The lifecycle of today’s Toxoplasma gondii works so reliably primarily because human settlements are surrounded by a large number of felines, from tame tabbies to feral cats. The populations of these domestic and “domestic” cats are so large, that the parasite easily finds its definitive host, and the prevalence of the parasite in the definitive and in the intermediate host stays at a high level.
But one must realize two things. Firstly, the domestic cat isn’t Toxoplasma’s optimal definitive host. Usually the cat becomes infected as a kitten, then for some time it releases cysts, mostly for only a couple of days to weeks. After this time, it stops releasing them, and for the rest of its life it likely does not spread the parasite. Of course there are known cases, in which the cat releases cysts several times over its lifespan, but these are probably exceptions which may be related to a dysfunction of the immune system, induced, perhaps, by a different infection. For the spread of Toxoplasma it would be much more useful if the infected cat shed cysts for the entirety of its life, as does the definitive host of the tapeworm or fluke.
Secondly, one must realize that, until recently, the domestic cat was almost never found in our settlements. In Europe, this cohabitant became more widespread after the Middle Ages. I often tell students that before 1800 cats were but briefly kept by witches not yet burned at the stake. That may be something of an exaggeration, but cats, for example, show up in pictures only starting in the 19th century, so it is likely that before this time, cats in houses and around human settlements were not as numerous as they are today. I suspect that for our ancestors, cats were also welcomed as a source of animal protein, which effectively regulated their populations surrounding the human settlements, as well as the prevalence of toxoplasmosis. I’m not an expert in the delicacies of Chinese cuisine, but I have a similar explanation for the low prevalence of toxoplasmosis in China. I believe that toxoplasmosis long did not exist in Europe, and appeared only recently, with the spread of domestic cats. Later, in chapter XV, we’ll return to this theory, but within a different context.
So who was the primary host of Toxoplasma before man domesticated the cat – or, as any cat owner knows, before cats domesticated us? If we restrict our hypotheses to the Felidae, and don’t presume that the definitive host could be a member of, for example, the civet family (which thus far has not been proven, but then again, it’s doubtful that anyone has really tested it), or perhaps of the hyena family (as the preliminary results of our scent experiments suggest, see page xxx), we are left primarily with various small and large members of the Felidae. I don’t know about the prevalence of toxoplasmosis in small cats, which are found more in Africa than in Europe, but in large cats, whether it be leopards or lions, toxoplasmosis is very widespread. For example, in 116 lions of four African reservations, the prevalence of toxoplasmosis was usually found to be 100% (only in Serengeti was it a “mere” 92%). And out of 8 studied leopards, 7 were found to be infected (40). And it is clear than neither lions nor leopards catch mice very often – they do, however, frequently prey on monkeys, including apes. One can therefore imagine that, in the past, our evolutionary ancestors formed a natural part of Toxoplasma’s life cycle, and therefore
Box 55 Are species evolutionarily adapted to current conditions?
Most biologists would say that unless the conditions have recently, radically changed, then generally species are. Today's biology textbooks write that species are constantly subject to natural selection, which weeds out the individuals not adapted to the environment they live in. If the environmental conditions change, then, due to natural selection, the species will adapt over several generations to the new conditions. As I explained in Box 35 How I refuted Darwin?, this classic idea of all-powerful evolution may not apply to sexually reproducing organisms. According to the theory of frozen plasticity (24) (25), a sexually reproducing species can change only just after its formation; for the rest of its existence it waits passively until environmental conditions have changed so much, that the species dies out. This would mean that species aren't adapted to the conditions in which they find themselves, but to the conditions which governed their environment as they were forming (or, more precisely, when they lost their evolutionary plasticity). Among other things, this explains why most of the species we find in nature are rare. The theoretical ecologist would expect that most species would have an average prevalence; there should be few abundant and rare species, and even fewer highly abundant and extremely rare species. In reality, every environment has but a couple of highly abundant species, but many rare species. This basic ecological axiom can be explained by the theory of frozen plasticity. Young, still plastic species have the greatest populations, for they are now ideally adapted to their living conditions. Statistically, it is likely that (on average) the older the species, the more its current conditions differ from its original conditions. Older species are poorly adapted to current conditions, so their populations are not very large.
that our experimental human-Toxoplasma model for studying the manipulation hypothesis is actually more natural than the often-used model of mouse-Toxoplasma. Toxoplasma definitely had more time to learn to manipulate the behavior of hominids for its own benefit than that of house mice (Box 55 Are species evolutionarily adapted to current conditions?). It is thus possible that today the Toxoplasma found in rodent hosts is merely trying to do what once paid of in the bodies of our evolutionary ancestors.