Revealed genetic changes accompanying selection for "good behavior" in foxes Belyaev

Belyaev's famous long-term experiment in breeding domesticated (as well as aggressive) foxes continues and is gaining momentum. Researchers are tapping into all the possibilities that today's research technologies provide. In 2018, several articles were published with the results of sequencing fox genomic DNA and RNA from their brain tissues. It was possible to identify many genes involved in changes and undergoing positive selection in different lines. Among them were genes related to hormonal regulation, differentiation of neural crest cells, the formation of intercellular contacts and synaptic signaling in the brain, as well as immunity genes.

The experiment on fox domestication, which was started in 1959 by Dmitry Konstantinovich Belyaev and Lyudmila Nikolaevna Trut at the fur farm of the Novosibirsk academic town of the Siberian Branch of the USSR Academy of Sciences, is widely known today not only among biologists, but also among the lay public. Many popular articles have been written about him and his interim results (see links at the end of the text).

The experiment began with the formation of a sample of silver-black foxes taken from the farm (foxes were raised there for skins for fur coats, etc.). The idea was to reproduce on foxes the same domestication process that wolves went through in the past, giving rise to domestic dogs. For this purpose, among the offspring of silver-black foxes, they began to select fox cubs that demonstrated loyalty and friendliness towards humans.

To carry out the selection, a method was selected that made it possible to determine the extent to which each fox is characterized by a manifestation of fear of a person or curiosity in relation to a person. This simple technique consists of analyzing the behavior of foxes (about 6 months old) for the following situations:
1) the experimenter stands near a closed cage, not trying to attract the attention of the animal;
2) the experimenter opens the cage door, stands nearby, but does not initiate communication;
3) the experimenter reaches out and tries to touch different parts of the animal's body;
4) the experimenter closes the cage door and stands calmly near the cage.

The video recordings of the trial are then analyzed to evaluate the animal's behavior against a number of trait criteria (see R. M. Nelson et al., 2016. Genetics of Interactive Behavior in Silver Foxes ( Vulpes vulpes)).

The offspring of the next generation was obtained from the least shy foxes, and then the testing and selection procedure was repeated again. Already in the fifth generation, individual individuals began to appear, which demonstrated an attraction to communication with humans, comparable to that of dogs. Over time, these became more and more, the sign of "good nature" intensified. Now all foxes in this line exhibit such doggy loyal and playful behavior (including even barking and "protecting" the owner) that some of them are sold as pets.

What was surprising about this experiment was not only the surprisingly rapid response to selection for behavior, but also the concomitant changes that began to appear in the phenotype of the foxes subjected to selection. These changes concerned signs that, at first glance, had nothing to do with behavior: white and red spots began to appear on the skin, foxes became more variable in metric characteristics (in some animals, a shortening of the length of the muzzle and paws was observed), in some animals, the tail began to curl, and violations appeared. occlusion, delayed hardening of the ear cartilage, discoloration of the iris. Moreover, foxes began to experience disruptions in the seasonality of reproductive behavior, an important trait for wild foxes that guarantees the appearance of puppies in the most favorable season of the year.

Taking into account the increase in variability in phenotype traits under experimental conditions, Belyaev introduced the concept of "destabilizing selection" - in contrast to the more typical natural evolutionary process "stabilizing selection" (this term was introduced in the first half of the XX century by I. I. Shmalgauzen), which, on the contrary , makes the phenotype more resistant. Belyaev admitted that the increase in variability observed in this experiment could have occurred in the process of domestication of wolves, and that this could give a good start for the formation of all that variety of breeds among dogs, which cannot but surprise, given that they all lead to the beginning from one common ancestor - the wolf, and this diversification of breeds began, apparently, no more than 15 thousand years ago.

It should be added that some time after the start of the experiment (namely, from 1970) a second line of foxes was added. On the contrary, they were selected for maximum aggressiveness and distrust of people. Despite the fact that the behavior of foxes in response to selection changed accordingly, some of the external phenotypic traits in this line began to converge with the corresponding traits in the line of good-natured foxes, although not so noticeably. At the same time, a control line of foxes is also conducted in parallel, in which no selection is made - and in this line, no special deviations from the initial phenotype of farmer silver-black foxes are noted. Parallel maintenance of three lines allows comparative analyzes, crossing experiments aimed at finding genetic loci associated with changes. The population size of each line is constantly maintained at about 200 individuals. The organization of the experiment implies taking measures to avoid excessive inbreeding between animals (this could lead to distortion of the results due to the increased effects of gene drift and a decrease in the viability of the offspring).

There are quite a few options for explaining the accompanying changes in traits that are not directly related to behavior. For example:
1) The effects of matching linked polymorphisms (this mechanism is also called genetic hitchhiking, see Genetic hitchhiking).
2) Pleiotropic action of the selected genes. In particular, there are genes that regulate the state of chromatin (working or non-working) through DNA methylation or histone modification - such genes can alter the work of a large number of other genes. A similar effect is expected for genes involved in alternative splicing or intracellular signaling.
3) Adaptive compromises, which are expressed in the fact that direct selection in some characters indirectly creates a new selection vector for other characters, which are functionally related to the first in ontogenesis.
4) Accidental appearance and preservation of new traits due to the increased role of gene drift (for example, due to the relatively small size of populations). However, this explanation hardly carries much weight here - after all, no significant changes were observed in the control line.
5) It is impossible to exclude an increase in the overall frequency of mutations, due, for example, to the fixation under the influence of the selection of mutations, which reduces the accuracy of replication or DNA repair.

Belyaev offered his original explanation for the observed phenomenon. His hypothesis was that intense selection for behavior perpetuated multiple mutations that alter the balance of hormones in the body. It is widely known that hormones play a huge role in determining temperament and emotional state in both humans and animals. These mutations are likely to have a pleiotropic effect, affecting, among other things, the provision of morphogenetic processes in the course of individual development. For example, the thyroid hormone system has a wide range of effects. It is possible that these mutations disable the mechanisms that normally provide stability (canalization) of morphogenesis, leading to the effect of destabilizing the phenotype. This hypothesis is supported by the weak heritability of some of the listed phenotypic abnormalities. Puppies from one pair of foxes are very diverse in appearance and character.

The hypothesis suggests that the mutations fixed in the course of selection affect those genes that control the maturation of neural crest cells in vertebrates (see: "The fourth germ layer" of vertebrates originated in lower chordates, "Elements", 02/04/2015). These cells, differentiating, firstly, participate in the formation of the adrenal cortex, where hormones such as adrenaline are produced, which, in particular, affect the triggering and realization of fear reactions in animals. Secondly, cells of the ear cartilage and some bones of the skull, including the jaw, pigment cells in the skin of the animal, cells of the iris, and sensitive cells of the inner ear, also originate from the neural crest. It is logical that the same mutations in the genes that control the development of neural crest cells can have a complex effect on all these traits. In this case, it is assumed that mutations lead to inhibition of differentiation or migration of neural crest cells and their lack in those tissues where they should eventually work. Getting into different combinations when crossing selected foxes, these mutations give rise to the observed diversity of phenotypes.

The genetic basis of the observed behavioral changes in foxes was confirmed by experiments with embryo transfer or puppy substitutions between females of different lines (“bad” and “good”) - such exchanges do not eliminate the differences in behavior developed during selection (AV Kukekova et al., 2008. Measurement of segregating behaviors in experimental silver fox pedigrees). And in a recent work, scientists have identified a large number of genetic loci associated with 98 behavioral criteria-signs, and have shown that these associations are complicated by epistatic influences depending on the combinatorics of allelic variants (HM Rando et al., 2018. Construction of Red Fox Chromosomal Fragments from the Short-Read Genome Assembly).

There is something remarkable about this whole story: the experiment began when the technology for molecular research was still very primitive. It was impossible to fully test certain hypotheses. But the experiment, thanks to Lyudmila Nikolaevna Trut and other employees of the Institute of Cytology and Genetics of the SB RAS, continued even after Belyaev's death in 1985 and continues to this day. Throughout all these years, the experiment has borne fruit in the form of regular publications, invariably attracting the attention of not only Russian, but also foreign specialists working in the field of genetics, developmental biology, and evolutionary biology. With the advent of new sequencing technologies, which are becoming more effective and accessible every year, scientists have the opportunity to investigate the molecular genetic basis of the observed phenotypic changes in animals. And this, of course, was done. The expansion of the research was also facilitated by the cooperation established since 2011 with foreign laboratories.

During 2018, as part of this study, as many as three articles were published in leading scientific journals. We will discuss the results presented in these works below.

Tatiana Romanovskaya