Remember all

The Research Institute of Physics of the Russian Academy of Medical Sciences discovered the ability of rats to transfer knowledge to new generations. Even after death and even if descendants have not yet been conceived. At least, this conclusion suggests itself after reading the publication in the June issue of the Bulletin of Experimental Biology and Medicine.

Last week, a fellow journalist found an amazing article in the library, presented in the pages of the journal Bulletin of Experimental Biology and Medicine [1]. The article said that rats trained to find a way out of a maze after death somehow transferred this knowledge to other animals. And not to their neighbors in the cage, but to the rat pups, who did not even exist at the time of the animals’ death. The rats were conceived after the death of the trained rodents, right on the cage with headless carcasses: you must agree, this is a very unusual discovery in a very unusual experiment.

I decided to contact the authors directly - to get the full text of the article and at the same time find out the details.

The first author was Sergei Konstantinovich Sudakov (Director of the P.K. Anokhin Institute of Normal Physiology of the Russian Academy of Medical Sciences), and he kindly sent a copy of the article. We will describe the experiment as is usually done in scientific publications, omitting only the details that are completely uninteresting to the general public.

Method and result

The rats were allowed to swim in a round pool, where in a certain place under the water there was an area invisible to them. This is a standard test of spatial memory, the Morris water maze. Sudakov’s group took as a basis the standard protocol for training animals - 4 days in a row, 4 times every day the rats were allowed to swim in the pool. The animals learned to quickly swim to the only place where there was at least some support under their paws: they were guided by the objects standing around the pool.

Then the most interesting thing happened. Three days after training, the animals—20 in total—were killed and the carcasses were placed on the first floor of a two-story cage with a lattice ceiling. On the second floor there was already a pair of a male and a female in a state of estrus: the rats, despite the rather gloomy appearance of the cage, successfully mated and became pregnant. Having killed 2 times 10 trained rats, the experimenters received two offspring, conceived in such an extreme environment. The authors called these two offspring “group A.”

For control, we also killed those rats that were allowed into the pool on the first day of training and in which they did not find a platform under water. Two more females and two males mating above their bodies gave rise to two more offspring - group B.

When groups A and B grew up, they themselves began to be trained in the Morris maze, and it was discovered that on the third day of training, animals in different groups had different times to reach the platform: group A showed the best results. Moreover, analysis of variance and Student's t-test showed high (p<0.05) reliability of these differences. Even sexual dimorphism was revealed: males of group A studied worse, and females - better. After another three months, the same animals were again trained to find a platform in the maze and again found differences - although this time on different days.

From this the conclusion is drawn: “The results of the study indicate the possibility of contactless transfer of information about previous learning from a dying organism to a nascent one, which at this time does not have any organs of perception of known types of information.”

Comments from the study author

Of course, both the setup of the experiment and the conclusions drawn from it look extraordinary. As you know, extraordinary claims require extraordinary proofs (“extraordinary claims require extraordinary evidence”), so you should very carefully consider the setup of the experiment itself.

Here is the text of my conversation with Sergei Sudakov:

S.S.: What causes the greatest doubts?

A.T.: The animals in the experimental group are from only two pairs of parents, i.e. it is possible that they simply have a different phenotype and because of this they learned differently.

S.S.: Two males and four females participated in the experiment. The males were the same for the production of groups A and B. The females, naturally, were different - two in each group. Selection of males and females for reproduction - animals that did not find the platform in 60 seconds.

A.T.: Why wasn’t standard testing carried out after training? When is the animal put into the pool again?

S.S.: The last training session can also be considered as such a test, what is the problem? In addition, we were interested in the dynamics of learning.

A.T.: Are you planning to conduct video tracking and record the trajectory of animals?

S.S.: We have video recording, but we have not analyzed it yet. The experiment generally went without any plan, we just at some point decided to return to the topic started back in 1985.

A.T.: How could such information be transmitted? After all, a map of the area in an animal’s brain is probably a certain graph of synaptic connections or some other changes; how could it pass to another animal without contact, and even after death?

S.S.: We don’t know and don’t want to speculate yet. But rats somehow convey information that you should not eat poison! Even if it's a dying animal?

Questions for the experiment

Suppose that our picture of the world is in some way incorrect and we have missed a place for a miracle in it. In this case, can the data be considered sufficiently convincing?

Vladimir Volokhonsky, psychologist, specialist in the field of statistical methods

The actual sample consists of two cases in one group and two cases in the other. The case is the entire offspring of a mating, and not each individual individual, because they are related by a common genetic set. In the experiment, there were, in fact, only two pairs of parent rats for each subgroup.

As I understand it, in the given example there were 16 measurements, each of which was compared, and differences were found in some of them. Well, in this example there is no correction for multiple comparisons. Some overview of these amendments can be found in my colleague Andrey Chetverikov [2]. And also here [3].

Inga Poletaeva, Dr. biol. Sciences, Leading Researcher, Laboratory of Physiology and Genetics, Department of Higher Nervous Activity, Faculty of Biology, Moscow State University

Groups of animals are small in size. This is the first one. Second, they did not use a significant difference test for the process, such as analysis of variance for repeated measures.

Third, there was no group of intact rats that would mate whenever they wanted, without contact with the slaughtered rats. There was also no control for the effect of simply euthanizing the rats, without cutting off the heads. There is no information that the females were immediately separated and a vaginal test was taken for fertilization!

In short, if I were the editorial board of the Bulletin, I would not have missed this article for purely professional reasons: the data are inconclusive due to the lack of the necessary control group and poor design of the experiment. You can look at papers on the Morris test and see that such differences may be found in other works, but they are not treated as differences, but as differences in the course of training. In addition, in training according to the Morris test, the decisive moment is the final test, when the platform is either removed or placed in a new place and the strength of the developed spatial skill is assessed by how long the animal spins in the old place of the platform. The article does not talk about such a test; if it was not done, this is experimental incompetence, and if they did it and did not provide the data, this is dishonest.

* * *

On my own behalf, I will add that it would be interesting to look at what the trajectory of the animals was and from what place they were released. The fact is that when training rats in the Morris water maze, they are launched from different places each time, so that the animal cannot learn a simple program for finding a platform like “swim straight, then slightly to the left,” but learns to navigate by objects standing around the experimental setup. If the rats of group A were at some point launched from a place that was simply closer to the platform than the place from which group B started, the difference in the time of reaching the platform is easily explained without involving the transfer of information from deceased animals.

Alexey Timoshenko

* * *

A comment

Researchers who take their work seriously sometimes obtain useful results even with a seemingly unscientific formulation of the problem. The German psychiatrist Hans Berger, in search of the material basis of telepathy, recorded electrical oscillations of brain origin - an electroencephalogram. But it is interesting that Berger decided to publish the first article about EEG only five years after the discovery of this phenomenon - after he had repeatedly confirmed his observations in new experiments [4].

However, while the formulation of a problem may sometimes be a personal matter for the researcher, the same cannot be said about the research methodology. You cannot work in science without trusting the results of other researchers. The more unusual the results obtained, the less they fit into the existing picture of the world, the greater changes in the system of scientific views they require. Obviously, the responsibility of a researcher when publishing breakthrough results is especially high, and the requirements for the methodological level of such work should be much higher than usual.

All of this seems to be a truism. But in the article under discussion, absolutely fantastic results were obtained at such a low methodological level, which is also unacceptable in student work. What makes it meaningless is the fact that both the experimental and control samples actually included only two cases each—the offspring of two females. Differences in learning could well arise, for example, if the X chromosomes of at least one of the two litters conceived from the corpses of trained rats accidentally had a predominance of genes that promote better learning.

In addition, the statistical tools used by the authors are designed for single testing of pre-formulated hypotheses, and not for conducting multiple tests without additional adjustments. Such tools cannot be directly used for so-called post hoc analysis, which should be used to find out whether the variations identified in the analysis could have arisen by chance. A good illustration of the danger of inaccurate use of statistics was given several years ago by American scientists in a widely known comic work. Using approximately the same statistical methods, they “discovered”... the perception of human emotions by dead salmon [5, 6].

It remains a mystery how S.K.’s not at all humorous, but obviously illiterate article. Sudakov and his colleagues could be published in a completely reputable scientific journal, one of the very few Russian physiological journals that has a Thomson Reuters impact factor (albeit a rather modest one - 0.305).

Even more surprising is that the first author of this article is the director of one of the leading specialized physiological institutes and a corresponding member of the Russian Academy of Medical Sciences. In this regard, questions cannot but arise about what are the prospects for Russian biomedical science if a person occupying such a responsible position in it can have such an attitude towards the quality of research.

Sergey Shishkin,
Ph.D. biol. sciences

1. S.K. Sudakov, G.A. Nazarova, E.V. Alekseeva. Contactless transfer of acquired information from a dying subject to a nascent one. Experimental study on rats. Bulletin of Experimental Biology and Medicine. 2012, volume 153, no. 6, p. 788-790.

2. http://chetvericov.ru/nauka/problema-mnozhestvennyx-sravnenij/

3. http://pubhealth.spb.ru/COPC/STATSH/anova.htm

4. D. Millett. Hans Berger: From psychic energy to the EEG. Perspectives in biology and medicine. 2001. Vol. 44. Pp. 522-542.

5. C.M.Bennett et al. Neural correlates of interspecies perspective taking in the post-mortem Atlantic Salmon: An argument for multiple comparisons correction. http://prefrontal.org/files/posters/Bennett-Salmon-2009.pdf

6. A. Madrigal. Scanning Dead Salmon in fMRI Machine Highlights Risk of Red Herrings. Wired, September 18, 2009. www.wired.com/wiredscience/2009/09/fmrisalmon/