BRAIN, BEHAVIOR, AND IMMUNITY 12, 83–89 (1998)ARTICLE NO. BI980523

COMMENTARY

A Genetic Basis forNeuroendocrine–Immune Interactions

Robert H. Bonneau,*,1 Pierre Mormede,† George P. Vogler,‡

Gerald E. McClearn,‡ and Byron C. Jones‡

*Department of Microbiology and Immunology, College of Medicine, The Pennsylvania StateUniversity, Milton S. Hershey Medical Center, Hershey, Pennsylvania 17033; †Laboratories for

Neurogenetics and Stress, INSERM-INRA, Universite Victor Segalen, Institut Francois Magendie deNeurosciences, 33077 Bordeaux cedex, France; and ‡Department of Biobehavioral Health, College

of Health and Human Development, The Pennsylvania State University, University Park,Pennsylvania 16802

Psychoneuroimmunology is an exciting, complex field that elucidates interactions amongthe nervous, endocrine, and immune systems. The contribution of psychosocial factors andbehavioral processes to these interactions has been the focus of numerous studies designedto investigate the intricate pathways that are involved in the ‘‘mind-body connection.’’ Inaddition, the effects of this connection on the development and progression of variousdisease conditions are of considerable interest. Although efforts have been made to identifythe cellular and molecular mechanisms underlying these relationships, the impact of geneticmakeup on the communication among these systems has yet to be fully realized. Thedevelopment of sophisticated genetic analytical methods and gene mapping techniques nowprovide the ‘‘tools’’ to determine the influence of genetics on behavior–neuroendocrine–immune interactions—an area of study that may represent the next frontier in psycho-neuroimmunology. 1998 Academic Press

Key Words: psychoneuroimmunology; genetics; immunology.

INTRODUCTION

Early studies in psychoneuroimmunology had sought to establish a scientific basisfor the association among emotion, behavior, and susceptibility to disease. Usingvarious experimental models, this association was shown to be mediated by the func-tional integration of the immune system with the nervous and endocrine systems.Such research in psychoneuroimmunology not only has defined the role of each ofthese systems in the etiology and resolution of disease but also has provided thecatalyst to identify the molecular mechanisms that underlie neuroendocrine–im-mune–disease interactions.

Although rigorous scientific investigations have provided compelling evidence thatbehavior can influence the onset and magnitude of neuroendocrine activity, a compre-hensive analysis of the genetic factors that contribute to this relationship has yet tobe conducted. Furthermore, the contribution of these genetic factors both to neuroen-docrine-mediated immune modulation and to susceptibility to disease has not been

1 To whom correspondence should be addressed at the Department of Microbiology and Immunology,H107, College of Medicine, M. S. Hershey Medical Center, 500 University Drive, Hershey, Pennsylvania17033. Fax: (717) 531-6522.

830889-1591/98 $25.00

Copyright 1998 by Academic PressAll rights of reproduction in any form reserved.

84 COMMENTARY

investigated. Research to identify genetic components underlying relationshipsamong the nervous, endocrine, and immune systems should test the hypothesis thatbehavior-associated modulation of the immune response is mediated by genetic fac-tors that dictate the composition and degree of neuroendocrine activation, which inturn influence the degree of immune protection that can be achieved. Baseline levelsof neuroendocrine activity are also likely to be subject to gene-based regulation. Suchregulation therefore may dictate the degree of natural resistance to disease withoutregard for external stimuli (e.g., psychological stress). More importantly, modernmeans of multivariate genetic analysis permit the testing of hypotheses regardinggenetic influences not only on individual variables but also on the relationshipsamong related traits. As discussed below, these analytical methods have been usedin other biological models to map complex behavioral traits to particular genetic loci.Similar approaches applied to psychoneuroimmunology could result in the identifica-tion of a genetic basis for behavioral–neuroendocrine–immune relationships.

A GENETIC BASIS FORNEUROENDOCRINE–IMMUNE–DISEASE

INTERACTIONS?

A simple understanding of the etiology of diseases caused by pathogenic microor-ganisms has resulted in the design of immune-based strategies in which the effectsof such pathogens can be largely controlled. Despite our ability to ward off mostinfections, it is still quite evident that the immune response to and the severity ofdisease are complex entities that show large inter-individual differences. By under-standing various ways in which the immune system interacts with the nervous andendocrine systems, and together, how they impact upon the course and outcome ofdisease, researchers have attempted to attribute some of these differences to variabil-ity among individuals in their neuroendocrine responsiveness to environmentalstimuli.

Genetic factors have long been known to contribute to the expression of variousbehavioral (Royce, Carran, & Howarth, 1970) and neurochemical (Oliverio,Castellano, & Puglisi-Allegra, 1979; Ingram & Corfman, 1980, Siegfried, Frisch-knecht, & Wasner, 1984; Shanks, Griffiths, Zalcman, Zacharko, & Anisman, 1990)phenotypes. Such phenotypes have the potential to modulate overall immune functionand resistance to infectious disease. Indeed, studies in mice have shown that strain-associated differences in various measures of immune function correlate with patternsof social behavior (Petitto, Lysle, Gariepy, Clubb, Cairns, & Lewis, 1993), avoidancebehavior (Sandi et al., 1991), housing conditions (Petitto, Lysle, Gariepy, & Lewis,1994), application of a conditioned aversive stimulus (Shurin, Kusnecov, Riech-man, & Rabin, 1995), chemical sympathectomy (Kruszewska, Felten, & Moynihan,1995), electric footshock (Brenner & Moynihan, 1997), and restraint stress (Hermann,Tovar, Beck, Allen, & Sheridan, 1993). Furthermore, studies using recombinant in-bred (RI) mice have demonstrated a genetic basis for the differential susceptibilityto infection and pathogenesis associated with Salmonella typhimurium (O’Brien, Ro-senstreich, & Taylor, 1980; Benjamin Jr., Turnbough, Posey, and Briles, 1986),Leishmania donavi (O’Brien et al., 1980), Trypanosoma cruzi (Trischmann andBloom, 1982), and Sendai virus (Brownstein and Winkler, 1987).

However, does one’s genetic makeup exert control over the composition, magni-tude, and duration of one’s neuroendocrine response to a particular stimulus? And

COMMENTARY 85

do differences in susceptibility to disease actually stem from a combination of bothenvironmental cues and genetic factors unique to each individual? The identificationof a genetic basis for these relationships represents an uncharted area in the studyof the interactions that exist among the nervous, endocrine, and immune systems.Such studies would allow one to estimate the contribution of genetic factors that arecommon to these immunological, physiological, and behavioral phenotypes in thevarious stages of the disease process.

THE GENETIC APPROACH—‘‘THE TOOLSAND THEIR USE’’

The key to determining the impact of genetics on behavioral–neuroendocrine–immune–disease interactions lies in the use of various methods of quantitative geneticanalysis including genetic correlational analysis, structural equation modeling, andquantitative trait loci (QTL) analysis. These interactions are likely to be controlledby more than one gene and thus fit the definition of a complex or multigenic trait.Although such genes cannot be identified by standard genetic techniques, QTL analy-sis is able to track down those genes which, together, govern the expression of asingle trait. QTL analysis can be done on any animal or plant species for which thereare inbred strains. In QTL analysis, F1 progeny derived from two parental strains(which differ in a given trait) are selectively bred back to one of these strains. Thesesecond generation progeny are then ‘‘scored’’ for the degree of the particular trait thatthey exhibit relative to each of the parentals. By knowing the locations of genomicpolymorphism between the parentals, regions in the genome that contribute to thetrait in question can be identified. QTL analysis thus is a correlational technique inwhich a phenotype which varies between genetically defined organisms is ‘‘mapped’’through linkage analysis to a region on a chromosome. These regions would likelycontain candidate gene(s) which influence the phenotype.

The use of such analyses to define a genetic basis for complex traits is not a recentinnovation. For example, QTL analysis originated in agricultural studies nearly 75years ago (Sax, 1923) and, together with genetic correlational analysis, has allowedscientists in numerous disciplines to map countless phenotypes. Such methods con-tinue to be employed by geneticists to assess the influence of genetics on complextraits in a wide array of organisms. More recently, QTL analysis has been applied tomurine models to study the genetic basis of both epilepsy and hypertension (Hilbert,Lindpainter, Beckmann, Serikawa, Soubrier, Dubay, Cartwright, De Gouyon, Julier,Takahasi, Vincent, Ganten, Georges, & Lathrop, 1991). Such an approach has alsoprovided a better understanding of the genetic basis underlying individual variationin diverse traits such as cancer susceptibility, drug sensitivity, aggressive behavior,and resistance to infection.

Quantitative genetics has been used in some of our own laboratories to provide anovel approach to identifying the relationships among the brain, behavior, and geneticcomposition as they apply to drug-induced behaviors. Such studies have utilized RIstrains of mice which provide powerful tools for assessing the genetic impact onwell-defined behavioral traits. For example, cocaine-induced stereotypy significantlycorrelates with the density of dopamine D2 receptors in the striatum. Although thestructural gene for this receptor is located on chromosome 9 in the mouse, the QTLfor this relationship is located on chromosome 15. This suggests that chromosome15 contains a gene which influences the expression of the D2 receptor gene (Jones

86 COMMENTARY

et al., unpublished data). Other studies have shown that the corticotropin-releasinghormone (CRH) response to stress correlates with alcohol self-selection (Erwin andJones, 1993) and that mesolimbic neurotensin (NTR) plays a role in mediating thebehavioral and hypnotic effects associated with alcohol consumption (Erwin andJones, 1993). QTL analysis has also determined that ethanol-induced anesthesia(DeFries, Wilson, Erwin, & Petersen, 1989), hyperlocomotion, and hypothermia (Er-win, Jones, & Radcliffe, 1990) are under polygenic influence. More recent studiesdemonstrate that a series of QTL are common between ethanol-induced behaviors andCNS NTR levels (Erwin, Markel, Johnson, Gehle, & Jones, 1997a; Erwin, Radcliffe,Gehle, & Jones, 1997b). These latter studies provide direct genetic evidence that NTRitself may regulate ethanol-induced behavior to some extent. Other studies (Sandi,Castanon, Vitiello, Neveu, & Mormede, 1991; Berton, Ramos, Chaouloff, &Mormede, 1997; Ramos, Berton, Mormede, & Chaouloff, 1997) have utilized strainsof rats, derived from bidirectional breeding programs, that diverge in their behavioralresponse to a given situation, such as stress. Such strains have been used to assessthe links among behavior and physiological parameters that are related to a particularresponse. Such an approach may be useful in the search for genes that contribute tointer-individual variations in emotional reactivity. Together, these studies illustratethe potential complexities that underlie genetic-based effects on behavior and neuro-endocrine function.

The application of quantitative genetic analysis to behavioral responses to environ-mental challenges in humans has typically relied on the availability of large cohortsof families and twins from which data on numerous behavioral, physiological, andgenetic parameters have been collected. Studies from some of our own laboratorieshave successfully used quantitative genetic analysis to determine the impact of genet-ics and environmental factors on depression (Gatz, Pedersen, Plomin, Nesselroade,& McClearn, 1992), Type A-like behaviors (Pedersen, Lichtenstein, Plomin, DeFaire,McClearn, & Matthews, 1989), cognitive ability (Finkel, Pedersen, McGue, &McClearn, 1995; Plomin, Pedersen, Lichtenstein, & McClearn, 1994), serum lipidmeasures (Heller, Pedersen, de Faire, & McClearn, 1994; Heller, de Faire, Pedersen,Dahlen, & McClearn, 1993), pulmonary function (McClearn, Svartengren, Pedersen,Heller, & Plomin, 1994) and an array of risk factors for cardiovascular disease(Vogler, McClearn, Snieder, Boomsma, Palmer, de Knijff, & Slagboon, 1997). To-gether, these studies demonstrate the applicability of quantitative genetics to studiesof human behavior and physiology.

APPLYING ‘‘THE TOOLS’’ TO PSYCHONEUROIMMUNOLOGY

An investigation of the role of genetics in the regulation of behavioral, neuroendo-crine, and immune responses represents a unique dimension of psychoneuroimmunol-ogy that has yet to be explored. The application of genetic analysis to studies inpsychoneuroimmunology could ascertain the existence of a genetic component inneuroendocrine–immune relationships. The use of genetic correlational analysis andstructural equation modeling could determine the genetic and environmental natureof the complex relationships among behavioral, physiological, and immunologicalparameters. In addition, the use of QTL analysis could be effective in locating chro-mosomal markers which point to suspect candidate genes that influence the singleor multiple quantitatively measured phenotypes of behavior, neuroendocrine activa-tion, and immune responsiveness. More importantly, single and tandem selection for

COMMENTARY 87

one and two QTLs could be conducted to demonstrate the nature of additive andgene–gene (epistatic) interactions that may be operative in behavior–neuroendo-crine–immune model systems.

The identification of regions containing loci of interest is becoming substantiallyeasier with the continually growing library of markers. Given the mapping of approxi-mately 1500 polymorphic markers (with more likely to be identified in the next fewyears), QTL analysis in recombinant inbred mice provides a powerful technique foridentifying genes which mediate neuroendocrine effects on immune function. Usingvarious murine models, several QTL have already been mapped for complex behav-ioral phenotypes such as open-field behavior (Gershenfeld, Neumann, Mathis, Craw-ley, Li, & Paul, 1997), hyperactivity (Moisan, Courvoisier, Bihoreau, Gauguier, Hen-dley, Lathrop, James, & Mormede, 1996) and emotionality (Flint, Corley, DeFries,Fulker, Gray, Miller, & Collins, 1995).

It is important that the use of murine models in these studies not be overlooked.As described above, the use of RI strains of mice provide a powerful set of tools forassessing the genetic impact on well-defined behavioral traits by systematic QTLanalysis. On the surface, such animal models may appear to be somewhat far removedfrom the human condition and thus not pertinent to neuroendocrine–immune interac-tions in humans. This, however, is not the case. As the description of genomes hasadvanced in recent years, it is clear that the order of a series of genes on the chromo-somes may indeed be the same in different species. Indeed, there is considerableoverlap between mouse and human genomes (Copeland, Jenkins, Gilbert, Eppig, Mal-tais, Miller, Dietrich, Weaver, Lincoln, Steen, Stein, Nadeau, & Lander, 1993). Thus,if a mouse QTL is located in a region with known loci which is also known in humans,and if these loci constitute such a ‘‘syntenic’’ series, an immediate guide is providedfor finding the human locus. As additional syntenic regions are identified, the abilityto associate a mouse QTL with that of a QTL in humans increases. Thus, the use ofgenetically defined mice for such studies is appropriate considering the extensiveoverlap between the mouse and human genomes.

Although the identification of a chromosomal region in which may reside a QTLis a long way from its precise localization, QTL analysis is likely to become a newstandard method in many domains of genetic research, including psychoneuroimmu-nology. The application of the knowledge gained from such a research endeavor maywell be the identification of genetic-related markers which identify individuals at riskfor disease states whose initiation and progression is under significant control of theneuroendocrine response following exposure to various environmental stimuli.

REFERENCES

Benjamin, W. H., Jr., Turnbough, C. L., Jr., Posey, B. S., & Briles, D. E. (1986). Salmonella typhimuriumvirulence genes necessary to exploit the Itys/s genotype of the mouse. Infect. Immun. 51, 872–878.

Berton, O., Ramos, A., Chaouloff, F., & Mormede, P. (1997). Behavioral reactivity to social and nonso-cial stimulations: A multivariate analysis of six inbred rat strains. Behav. Genet. 27, 155–166.

Brenner, G. J., & Moynihan, J. A. (1997). Stressor-induced alterations in immune response and viralclearance following infection with herpes simplex virus-type-1 in BALB/c and C57BL/6 mice.Brain Behav. Immun. 11, 9–23.

Brownstein, D. G., & Winkler, S. (1987). Genetic determinants of lung virus titers and resistance tolethal Sendai virus pneumonia. Arch. Virol. 96, 201–214.

Copeland, N. G., Jenkins, N. A., Gilbert, D. J., Eppig, J. T., Maltais, L. J., Miller, J. C., Dietrich,W. F., Weaver, A., Lincoln, S. E., Steen, R. G., Stein, L. D., Nadeau, J. H., & Lander, E. S. (1993).

88 COMMENTARY

A genetic linkage map of the mouse: Current applications and future prospects. Science 262, 57–66.

DeFries, J. C., Wilson, J. R., Erwin, V. G., & Petersen, D. R. (1989). LS 3 SS recombinant inbredstrains of mice: initial characterization. Alcohol Clin. Exp. Res. 13, 196–200.

Erwin, V. G., & Jones, B. C. (1993). Genetic correlations among ethanol-related behaviors and neuroten-sin receptors in long sleep (LS) and short sleep (SS) recombinant inbred strains of mice. Behav.Genetics 23, 191–196.

Erwin, V. G., Jones, B. C., & Radcliffe, R. (1990). Further characterization of LS 3 SS recombinantinbred strains of mice: Activating and hypothermic effects of ethanol. Alcohol Clin. Exp. Res. 14,200–204.

Erwin, V. G., Markel, P. D., Johnson, T. E., Gehle, V. M., & Jones, B. C. (1997a). Common quantitativetrait loci for alcohol-related behaviors and CNS neurotensin measures: hypnotic and hypothermiceffects. J. Pharm. Exp. Therapeut. 280, 911–918.

Erwin, V. G., Radcliffe, R. A., Gehle, V. M., & Jones, B. C. (1997b). Common quantitative trait locifor alcohol-related behaviors and CNS neurotensin measures: Locomotor activation. J. Pharm. Exp.Therapeut. 280, 919–926.

Finkel, D., Pedersen, N. L., McGue, M., & McClearn, G. E. (1995). Heritability of cognitive abilitiesin adult twins: Comparison of Minnesota and Swedish data. Behav. Genet. 25, 421–431.

Flint, J., Corley, R., DeFries, J. C., Fulkner, D. W., Gray, J. A., Miller, S., & Collins, A. C. (1995). Asimple genetic basis for a complex psychological trait in laboratory mice. Science 269, 1432–1435.

Gatz, M., Pedersen, N. L., Plomin, R., Nesselroade, J. R., & McClearn, G. E. (1992). Importance ofshared genes and shared environments for symptoms of depression in older adults. J. Abnorm.Psychol. 101, 701–708.

Gershenfeld, H. K., Neumann, P. E., Mathis, C., Crawley, J. N., Li, X., & Paul, S. M. (1997). Mappingquantitative trait loci for open-field behavior in mice. Behav. Gen. 27, 201–210.

Heller, D. A., deFaire, U., Pedersen, N. L., Dahlen, G., & McClearn, G. E. (1993). Genetic and environ-mental influences on serum lipid levels in twins. New Engl. J. Med. 328, 1150–1156.

Heller, D. A., Pedersen, N. L., deFaire, U., & McClearn, G. E. (1994). Genetic and environmentalcorrelations among serum lipids and apolipoproteins in elderly twins reared together and apart. Am.J. Human Genet. 55, 1255–1267.

Hermann G., Tovar, C. A., Beck, F. M., Allen, C., & Sheridan, J. F. (1993). Restraint stress differentiallyaffects the pathogenesis of an experimental influenza virus infection in three inbred strains of mice.J. Neuroimmunol. 47, 83–94.

Hilbert, P., Lindpainter, K., Beckmann, J. S., Serikawa, T., Soubrier, F., Dubay, C., Cartwright, P., DeGouyon, B., Julier, C., Takahasi, S., Vincent, M., Ganten, D., Georges, M., & Lathrop, G. M.(1991). Chromosomal mapping of two genetic loci associated with blood-pressure regulation inhereditary hypertensive rats. Nature (London) 353, 521–529.

Ingram, D. K., & Corfman, T. P. (1980). An overview of neurobiological comparisons in mouse strains.Neurosci. Biobehav. Rev. 4, 421–435.

Kruszewska, B., Felten, S. Y., & Moynihan, J. A. (1995). Alterations in cytokine and antibody productionfollowing chemical sympathectomy in two strains of mice. J. Immunol. 155, 4613–4620.

McClearn, G. E., Svartengren, M., Pedersen, N. L., Heller, D. A., & Plomin, R. (1994). Genetic andenvironmental influences on pulmonary function in aging Swedish twins. J. Gerontol. 49, 264–268.

Moisan, M.-P., Courvoisier, H., Bihoreau, M.-T., Gauguier, D., Hendley, E. D., Lathrop, M., James,M. R., Mormede, P. (1996). A major quantitative trait locus influences hyperactivity in the WKHArat. Nature Genet. 14, 471–473.

O’Brien, A. D., Rosenstreich, D. L., & Taylor, B. A. (1980). Control of natural resistance to Salmonellatyphimurium and Leishmania donovani in mice by closely linked but distinct genetic loci. Nature(London) 287, 440–442.

Oliverio, A., Castellano, C., & Puglisi-Allegra, S. (1979). A genetic approach to behavioral plasticityand rigidity. In J. R. Royce and L. P. Mos (Eds.), Theoretical advances in behavioral genetics,pp. 139–165. Sijhtoff & Noordhoff, Alphen aan de Rijn.

COMMENTARY 89

Pedersen, N. L., Lichstenstein, P., Plomin, R., DeFaire, U., McClearn, G. E., & Matthews, K. A. (1989).Genetic and environmental influences for type A-like measures and related traits: A study of twinsreared apart and twins reared together. Pychosom. Med. 51, 428–440.

Petitto, J. M., Lysle, D. T., Gariepy, J. L., Clubb, P. H., Cairns, R. B., & Lewis, M. H. (1993). Geneticdifferences in social behavior: Relation to natural killer cell function and susceptibility to tumordevelopment. Neuropsychopharmacology 8, 35–43.

Petitto, J. M., Lysle, D. T., Gariepy, J. L., & Lewis, M. H. (1994). Association of genetic differencesis social behavior and cellular immune responsiveness: effects of social experience. Brain Behav.Immun. 8, 111–122.

Plomin, R., Pedersen, N. L., Lichtenstein, P., & McClearn, G. E. (1994). Variability and stability incognitive abilities are largely genetic later in life. Behav. Genet. 24, 207–215.

Ramos, A., Berton, O., Mormede, P., & Chaouloff, F. (1997). A multiple-test study of anxiety-relatedbehaviours in six inbred rat strains. Behav. Brain. Res. 85, 57–69.

Royce, J. R., Carran, A., & Howarth, E. (1970). Factor analysis of emotionality in ten inbred strains ofmice. Multivar. Behav. Res. 5, 19–48.

Sandi, C., Castanon, N., Vitiello, S., Neveu, P. J., & Mormede, P. (1991). Different responsiveness ofspleen lymphocytes from two lines of psychogenetically selected rats (Roman high and low avoid-ance). J. Neuroimmunol. 31, 27–33.

Sax, K. (1923). Association of size differences with seed-coat pattern and pigmentation in Phaeseolusvulgaris. Genetics 8, 552–560.

Shanks, N., Griffiths, J., Zalcman, S., Zacharko, R. M., & Anisman, H. (1990). Mouse strain differencesin plasma corticosterone following uncontrollable footshock. Pharmacol. Biochem. Behav. 36, 515–519.

Shurin, M. R., Kusnecov, A. W., Riechman, S. E., & Rabin, B. S. (1995). Effect of a conditioned aversivestimulus on the immune response in three strains of rats. Psychoneuroendocrinology 20, 837–849.

Siegfried, B., Frischknecht, H.-R., & Wasner, P. G. (1984). Defeat, learned submissiveness, and analgesiain mice: Effect of genotype. Behav. Neural. Biol. 42, 91–97.

Trischmann, T. M. & Bloom, B. R. (1982). Genetics of murine resistance to Trypanosoma cruzi. Infect.Immun. 35, 546–551.

Vogler, G. P., McClearn, G. E., Snieder, H., Boomsma, D. I., Palmer, R., de Knijff, P., & Slagboom,P. E. (1997). Genetics and behavioral medicine: Risk factors for cardiovascular disease. Behav.Med. 22, 141–149.