Commentary on ‘Comparative ’ Victor H. Denenberg Professor Emeritus University of Connecticut Comparative Psychology and Visiting Professor University of Washington Is Still Alive but May Be Losing Relevance

ABSTRACT: Greenberg et al., in their perspective on the current state and fate of comparative psychology, present convincing data that the field is viable and that comparative are making important contributions to the research literature. The central feature of the field is its emphasis upon . This is also its weakness since advances in genetic techniques permit researchers to create laboratory animals that have no counterpart in the natural world, and thus have no evolutionary history. These ‘‘unnatural’’ animals are widely used in behavioral, biological, and medical studies, but the findings cannot be interpreted within a comparative psychology framework. As the use of these preparations expand, the relevance of comparative psychology diminishes. ß 2003 Wiley Periodicals, Inc. Dev Psychobiol 44: 21–25, 2004.

Keywords: anagenesis; general psychology; inbred strains; gene knockouts; gene insertions; Y chromosome; sex reversal

Greenberg et al. are to be commended for their spirited COMPARATIVE PSYCHOLOGY ¼ GENERAL perspective on comparative psychology for the 21st PSYCHOLOGY? century. Their view of comparative psychology is built upon the concepts of anagenesis and integrative levels, Greenberg et al. advocate a larger role for comparative the mathematical framework of dynamic systems theory, psychology in the pantheon of psychological sciences. and the many contributions of the animal behavior re- Indeed, they imply that comparative psychology should searchers from the American Museum of Natural History. be first among equals by equating comparative psychol- Overall, they succeed in establishing that comparative ogy with general psychology. The difficulty with the claim psychology is viable, is making important contributions to that comparative psychology equals general psychology the research literature, and that its survival prognosis is is that when two different terms are used to specify one excellent. However, in their enthusiasm, I feel that they field, there will competition for labeling rights, and gene- have overextended their vision of comparative psycho- ral psychology is certain to win (a form of survival of the logy, and ignored some issues that need to be addressed. fittest). This would be a shame since comparative psycho- More important, there is current genetic research which, logy is one of the gods, albeit a lesser one, in the pantheon, I believe, falls outside the purview of comparative and its place and contributions need to be recognized. psychology and thus reduces its importance for future Comparative psychology may be defined as the study contributions. of the evolution of behavior and behavior patterns, includ- ing species-specific behaviors (e.g., maternal behavior, aggression) and psychological processes (e.g., , ). In many instances, the focus is upon beha- viors that occur in a natural setting (Dewsbury, 1992a,b). Received 30 July 2003; Accepted 30 July 2003 At times, the behaviors are associated with biological Correspondence to: V. H. Denenberg characteristics, usually involving the nervous or endocrine Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/dev.10156 systems. A developmental perspective is often espoused. The core feature of comparative psychology is the ß 2003 Wiley Periodicals, Inc. emphasis upon evolution. Without an evolutionary focus, 22 Denenberg there is no comparative psychology. Since genes are one social good, but the processes involved in doing science of the major mediators of evolutionary events, there is, are very different from the decisions involved in attem- at least implicitly, a focus upon genetics and other bio- pting to apply scientific findings, and I think it is important logical factors. None of these features—evolution, genes, to maintain that distinction. biology—are required or needed for the broader field of general psychology, and this is why I cannot accept equat- MISAPPROPRIATION OF CONCEPTS ing comparative psychology with general psychology. Many of the concepts that Greenberg et al. have associated ANAGENESIS with the field of comparative psychology do not have a unique relationship to the field. Thus, researchers in many In Greenberg et al.’s discussion of anagenesis, defined as scientific disciplines use a developmental perspective, ‘‘directional trends in evolution, trends which reflect manipulate genetic factors and are aware of epigenesis, change from the simple to the ’’ is the following are not reductionistic in their philosophy (though they statement: ‘‘Animals less behaviorally plastic function at may well be in their methodology), recognize levels of lower behavioral levels at which biological processes are organization and emergent properties, acknowledge the of great significance; more behaviorally plastic organisms necessity of taking environmental—including ecologi- function at higher behavioral levels at which psycho- cal—context into account in interpreting findings, and logical processes (e.g., mediation) direct the course of use nonlinear systems analyses. The feature unique to behavioral development’’ (p. 5). This strikes me as a comparative psychology is the evolutionary emphasis. ‘‘proportion of variance’’ statement commonly seen in heritability studies, which parse behavior into genetic and nongenetic (i.e., environmental) components that are TWO ISSUES FOR COMPARATIVE statistically independent. The phrase ‘‘less behaviorally PSYCHOLOGY TO ADDRESS plastic’’ implies that heredity accounts for the greater proportion of variance whereas the environmental com- ponent is larger for ‘‘more behaviorally plastic organ- The laboratory is biologically very different from its isms.’’ This is a static interpretation that is antithetical to wild ancestors (Richter, 1959). How does this animal fit an epigenetic–developmental perspective and appears into a comparative analysis? Is it to be put on the same inconsistent with the major theses of the article. scale as recently captured avian species and nonhuman The attempt to keep separate biological and psycho- that have been bred in a laboratory for only a logical processes also is seen in Greenberg et al.’s later few generations? Or should the laboratory rat be better discussion of anagenesis, where they restrict use of the compared with domesticated farm animals? Further, how term to psychological processes, excluding biological does comparative psychology view humans—as wild, ones because they acknowledge that the concept is of domesticated, or in between? limited use in biology. Then they go on to state that these two processes, biological and psychological, are compa- Animal Models tible with each other; this strengthens their argument to apply anagensis only to psychological processes. But This usually starts with a human disease or disturbance for compatability is a form of equivalence, so I find it difficult which an animal model is sought. Cancer in mice is one of to see how one can use an equivalence argument to support the earliest examples. Diabetes is another example. Such a nonequivalence conclusion. models are most easily validated when there are structural similarities in the two species (e.g., presence of cancers), but are more dubious when trying to study functional MULTIPLE ROLES FOR COMPARATIVE disturbances (e.g., schizophrenia). What contribution, if PSYCHOLOGY any, can comparative psychology make when trying to find an appropriate model system? Greenberg et al. list four roles for comparative psychol- ogy: managing resources, increasing activity in social and political issues, studying relationships between animals CAN COMPARATIVE PSYCHOLOGY DEAL and changing environments, and providing a historical WITH ‘‘UNNATURAL’’ ANIMALS? perspective on human evolution. The last two items represent scientific objectives, but the first two do not. The central concept in comparative psychology is that it Instead, they may be characterized as social services. is the study of the evolution of behavior, with particular I strongly favor using scientific findings and theory for emphasis upon the natural behavior of a species. However, Comparative Psychology: Losing Relevance? 23 recent advances in biology have created a number of gene. As one example, Thomas and Palmiter (1997a) ‘‘unnatural’’ animal preparations that are eagerly being knocked out the dopamine beta-hydroxylase gene, result- investigated by biologists and psychologists. These pre- ing in a mouse unable to produce the neurotransmitter parations raise challenging issues for a comparative norepinephrine. These mutants were equally competent as analysis of behavior, as discussed next. controls in passive avoidance learning, but were less effective in active learning (Thomas & Palmiter, 1997a), and their maternal behavior also was found to be impaired Inbred Strains (Thomas & Palmiter, 1997b). Inbred strains of animals do not exist in nature; they are The targeted gene of a knockout mouse is permanently purely a product of laboratory manipulations. The vast inactivated. It also is possible to create a conditional majority of inbred animals are found in laboratory knockout where the experimenter can turn the gene ‘‘on’’ populations of fruit flies, mice, and . It will be useful or ‘‘off’’ at will; it also is possible to activate the gene in to describe how an inbred strain is produced. certain brain regions and not in others. Thus, Gross et al. (2002; see also Snyder, 2002) were able to knock out the Producing an Inbred Strain. To simplify, if the dominant serotonin1A receptor (5-HT1AR) gene (a) during early form of a gene is labeled A and the recessive form a,then development or in later life and (b) in the hippocampus the distribution of these genes in the progeny will be one and cortex, or in the raphe nucleus. They found that the AA, two Aa, and one aa. Extend this analysis forgenes B, C, mice without 5-HT1AR receptors were more anxious than and so on for a total of approximately 30,000 genes, which controls, as measured by three different behavioral tests. is one of the current estimates of the number of genes in Normal levels of anxiety were restored in mice when the humans and mice. The consequence is that each individual 5-HT1AR gene was turned on early in development in the animal has, on average, half of its genes in the heterozy- forebrain region. No other combination was effective. gous state (e.g., Aa, Bb, etc.) and half in the homozygous states (e.g., AA, BB, etc., or aa, bb, etc.). The purpose of A Down Syndrome Mouse inbreeding is to eliminate all the heterozygous gene combinations. This is accomplished by successive Broth- Three copies of Chromosome 21 is a feature common to er Sister mating. Start with a pair of animals from a all humans with Down Syndrome (DS). Mice with three natural population. When young are produced and are copies of Chromosome 16 have been used as a model of sexually mature, a brother and sister are mated. From their DS, but full trisomy results in death. However, mice that offspring, a brother and sister are mated, and so on. The are trisomic for only that portion of Chromosome 16 also offspring from the first mating are maximally hetero- found in humans with DS will survive and grow to zygous. Thereafter, each successive Brother Sister maturity (Davisson et al., 1993). These animals have been mating reduces genetic variability by 50%. Twenty genera- found to have cognitive deficits on several learning tasks tions of successive Brother Sister matings are required to (Hyde, Frisone, & Crnic, 2001; Reeves et al., 1995). define an inbred strain. Assuming one started with 30,000 genes, after 20 generations only about 0.03% of the genes Inserting Human Genes into the Mouse are still in a heterozygotic state (Denenberg, 2002). All A number of human genes have been introduced into the other pairs are either double dominant (AA, BB, etc.) or mouse genome. Lamb et al. (1993, 1997) inserted the double recessive (aa, bb, etc.). Further, the gene patterns beta-amyloid precursor protein (APP) gene, which has are identical for all animals within the strain. If one animal been associated with Alzheimer’s disease and Down is double dominant for Genes A, X, and Q, then all other Syndrome. Hwang et al. (2001) transplanted the human animals also are double dominant for Genes A, X, and Q. Cytochrome P450 (CYP) gene into a mouse with the Thus, the amount of genetic variance in the population has objective of studying environmental chemical toxins in been reduced to essentially zero, and any two animals of this ‘‘human-gene’’ preparation. the same sex picked at random from the population are identical twins whereas a male and female have identical Changing the Y Chromosome genes on all their autosomes (all chromosomes except the and Sex Reversal sex chromosomes). We see, then, that an inbred population is in essence a population of clones (Denenberg, 2002). Eicher & Washburn (1983) mated males with the Mus domesticus poschiavinus Y chromosome to C57BL/6JEi females, and followed this with repeated backcrosses and Deleting Genes inbreeding. This resulted in a strain of C57 mice that It is possible to functionally ‘‘delete’’ a gene by preventing differed from the normal C57 strain only in having a it from being activated, a process called ‘‘knocking out’’ a different Y chromosome. The progeny of these males 24 Denenberg included a phenotypically and chromosomally normal Thomas and Palmiter (1997a,b) found that the gene XX female, a phenotypically normal but chromosomally involved in producing norepinephrine impairs both active sex-reversed XY female, and an XY hermaphrodite male. avoidance learning (but not passive learning) and The cause of these anomalies appears to be an interaction maternal behavior, behaviors which are not commonly between the sex-determining gene on the Y chromosome, associated together. And the introduction into the mouse’s Sry, and one or more of the C57’s autosomal genes. genome of human disease genes or extra chromosomes Behavioral studies using tests that distinguish males from allow us to ‘‘custom design’’ animal models of human females found that the two female groups did not differ ailments. on open-field behavior, the Lashley III maze, or active Even though each gene these animals carry has both avoidance learning, indicating that these behaviors were phylogenetic and ontogenetic histories, the animal as an under hormonal (i.e., gonadal) control; however, the XY integrated functioning organism does not have an evolu- females were superior to the XX females on the hidden tionary history. Thus, it does not appear possible to platform Morris maze, suggesting that males may have evaluate such populations within a comparative psycho- both a hormonal and a genetic mechanism to insure logy context. Advances in genetic technology have visuospatial superiority (Stavnezer, Hyde, Bimonte, provided, and will continue to provide for a long time to Armstrong, & Denenberg, 2000). come, large numbers of new animal populations. These Another way to manipulate sex chromosome effects is populations are actively being used for behavioral studies. to delete the Sry gene from the Y chromosome and insert As one index, I entered the phrase ‘‘inbred AND mice an Sry transgene onto an autosome (De Vries et al., 2002). AND behavior’’ into PubMed (PubMed. National Library Hormonal factors were found to correlate with male of Medicine [http://www.ncbi.nlm.nih.gov:80/entrez/ copulatory behavior, social exploration, and several sexu- query.fcgi?CMD¼&DB¼PubMed]) and got more than ally dimorphic structures; however, male and female mice 9,000 citations. As these preparations become better with XY chromosomes were more masculine than XX known and more readily available, I expect there will be mice in the density of vasopressin-immunoreactive fibers. increasing numbers of investigators taking advantage of Thus, both studies agree that many of the commonly these new research opportunities. One consequence is observed sex differences derive from hormonal factors a lessening interest and emphasis upon comparative attributable to the ovaries and testes, but other sex dif- psychology. Thus, though comparative psychology is ferences are related to the sex chromosome, leading to clearly alive and viable, it does not provide a relevant the conclusion ‘‘that sex chromosome genes contribute framework for those working with ‘‘designer gene popu- directly to the development of a sex difference in the lations,’’ and I expect its influence to diminish over time. brain’’ (De Vries et al., 2002, p. 9005).

IMPLICATIONS OF ‘‘UNNATURAL’’ ANIMALS REFERENCES FOR A COMPARATIVE PSYCHOLOGY Davisson, M. T., Schmidt, C., Reeves, R. H., Irving, N. G., The preparations listed earlier—genetically homozygous Akeson, E. C., Harris, B. S., & Bronson, R. T. (1993). Segmental trisomy as a mouse model for Down syndrome. mice, animals with one or more genes knocked out, mice Progress in Clinical Biological Research, 384, 117–133. with three copies of a chromosome, animals with human Denenberg, V. H. (2002). Cloning: A tool for behavioral genes inserted into their genome, subjects with foreign sex research. APS Observer, 15, 11. chromosomes—are all ‘‘unnatural’’ in the sense that they De Vries, G. J., Rissman, E. F., Simerly, R. B., Yang, L.-Y., are ‘‘designer gene populations’’ created by human re- Scordalakes, E. M., Auger, C. J., Swain, A., Lovell-Badge, searchers in a laboratory and have no evolutionary history R., Burgoyne, P. S., & Arnold, A. P. J. (2002). A model of adapting to environmental challenges. system for study of sex chromosome effects on sexually Yet, for those of us interested in behavior, these dimorphic neural and behavioral traits. Journal of Neu- ‘‘unnatural’’ animal preparations are yielding valuable roscience, 22, 9005–9014. scientific information, advancing knowledge, and causing Dewsbury, D. A. (1992a). Comparative psychology and etho- us to rethink some of our theories and concepts. Gross et al. logy: A reassessment. American , 47, 208– 215. (2002) showed that the 5HT1AR gene must be activated at Dewsbury, D. A. (1992b). Triumph and tribulation in the history a certain developmental time in a specific brain region to of American comparative psychology. Journal of Compara- have an effect on anxiety level. Eicher and Washburn’s tive Psychology, 106, 3–19. (1983) studies revealed that determination of phenotypic Eicher, E. M., & Washburn, L. I. (1983). Inherited sex reversal sex is not due solely to the Sry gene on the Y chromosome, in mice: Identification of a new primary sex-determining but may be conditional upon the presence of other genes. gene. Journal of Experimental Zoology, 228, 297–304. Comparative Psychology: Losing Relevance? 25

Gross, C., Zhuang, X., Stark, K., Ramboz, S., Oosting, R., Gearhart, J. D. (1993). Introduction and expression of the 400 Kirby, L., Santarelli, L., Beck, S., & Hen, R. (2002). Seroto- kilobase amyloid precursor protein gene in transgenic mice nin 1A receptor acts during development to establish normal [corrected]. Nature Genetics, 5, 22–30. anxiety-like behaviour in the . Nature, 416, 396–400. Reeves, R. H., Irving, N. G., Moran, T. H., Wohn, A., Kitt, C., Hwang, D. Y., Chae, K. R., Shin, D. H., Hwang, J. H., Lim, Sisodia, S. S., Schmidt, C., Bronson, R. T., & Davisson, M. T. C. H., Kim, Y. J., Kim, B. J., Goo, J. S., Shin, Y. Y., Jang, I. S., (1995). A mouse model for Down syndrome exhibits learning Cho, J. S., & Kim, Y. K. (2001). Xenobiotic response in and behaviour deficits. Nature Genetics, 11, 177–184. humanized double transgenic mice expressing tetracycline- Richter, C. P. (1959). Rats, man, and the welfare state. American controlled transactivator and human CYP1B1. Archives of Psychologist, 14, 18–28. Biochemistry and Biophysics, 39, 32–40. Snyder, S. H. (2002). Serotonin sustains serenity. Nature, 41, Hyde, L. A., Frisone, D. F., & Crnic, L. S. (2001). Ts65Dn mice, 377–380. a model for Down syndrome, have deficits in context discri- Stavnezer, A. J., Hyde, L. A., Bimonte, H. A., Armstrong, mination learning suggesting impaired hippocampal function. C. M., & Denenberg, V. H. (2002). Differential learning Behavioural Brain Research, 118, 53–60. strategies in spatial and nonspatial versions of the Morris Lamb, B. T., Call, L. M. M., Slunt, H. H., Bardel, K. A., Lawler, water maze in the C57BL/6J inbred mouse strain. Beha- A. M., Eckman, C. B., Younkin, S. G., Holtz, G., Wagner, vioural Brain Research, 133, 261–270. S. L., Price, D. L., Sisodia, S. S., & Gearhart, J. D. (1997). Thomas, S. A., & Palmiter, R. D. (1997a). Disruption of the Altered metabolism of familial Alzheimer’s disease-linked dopamine beta-hydroxylase gene in mice suggests roles for amyloid precursor protein variants in yeast artificial chromo- norepinephrine in motor function, learning, and memory. some transgenic mice. Human Molecular Genetics, 6, 1535– , 111, 579–589. 1541. Thomas, S. A., & Palmiter, R. D. (1997b). Impaired maternal Lamb, B. T., Sisodia, S. S., Lawler, A. M., Slunt, H. H., Kitt, behavior in mice lacking norepinephrine and epinephrine. C. A., Kearns, W. G., Pearson, P. L., Price, D. L., & Cell, 91, 583–592.