C H A P T E R

26

The Marmoset as a Model in BehavioralNeuroscience and Psychiatric Research

Jeffrey A. FrenchNeuroscience Program and Callitrichid Research Center, University of Nebraska at Omaha, NE, United States

INTRODUCTION

In its mission statement, the US National Institutes ofHealth provides a clear statement of its focus: “. to seekfundamental knowledge about the nature and behaviorof living systems and the application of that knowledgeto enhance health, lengthen life, and reduce illness anddisability” [emphasis added, www.nih.gov/about-nih/what-we-do/mission-goals, 2015). Disorders associatedwith brain or behavioral dysfunction represent the lead-ing disease burden and highest source of lifetime yearsliving with disability on a global basis (YLD: [1]) andtogether these disorders represent one of the leadingcontributors to disease-associated mortality worldwide[2]. Clearly, then, there is a premium on understandingboth normative behavioral states and their relationshipto brain function and the nature of brain dysfunctionas is relates to pathological behavioral states.

As in all fields of biomedical science, animal modelsplay an important role in characterizing the basicbiology of brainebehavior relationships, in exploringthe potential etiology of disorders (genetics, develop-ment, environmental perturbations), and in developingand refining treatments in a number of modalities thattarget dysfunction on brain processing. Nonhuman pri-mates (NHPs) constitute an attractive animal model forbehavioral and psychological states for a variety ofimportant reasons. Given the relatively recent commonevolutionary origin of primates [3], NHPs share com-monalities in complex brain structure and function (ho-mologies), most importantly represented by an increasein brain size and elaboration of the neocortex, especiallythe prefrontal cortex (PFC) [4]. As a consequence ofthese morphological adaptations, NHPs are capable ofcomplex cognitions that involve multisensory integra-tion, complex and conditional decision-making, and

top-down as well as bottom-up regulation of affectand emotion. Finally, the changes in NHP brain struc-ture and function can facilitate the mediation of chal-lenges associated with group living, includingaggression, affiliation, and the establishment and main-tenance of long-term complex social relationships thatdistinguish these species from nonprimate animals [5].

The Utility of Marmosets in Behavioral Modelsin Neuroscience and Psychiatric Research

From the perspective of a biomedically orientedfocus, research on behavioral states (both normativeand atypical) is of interest to the extent that it can pro-vide useful information regarding the developmentalfactors that lead to normative neuropsychological func-tion, as well as the underlying neurobiological proper-ties that contribute to atypical or pathological states.Animal models, primarily rodents, have been widelyused in exploring both developmental and neurobiolog-ical functions in the behavioral realm [6,7], but progresshas been limited with rodent models, relative to primatemodels, for multiple reasons. First, the complexity of thecentral nervous system, primarily in cortical regionsimportant for complex thought, cognition, affect, social-ity, and decision-making, is not as elaborate in rodents asit is in primates, including humans [8]. The 80 millionyears since the last common ancestor between rodentsand primates has led to divergence in the relative func-tional morphology and connectivity of the brain, inparticular the expansion of the PFC in the primate line-age [9]. Second, the cognitive, social, and behavioralphenotypes of primates are much richer and more com-plex than those of rodents [10], presumably a result ofthe elaboration and patterns of connectivity among

477The Common Marmoset in Captivity and Biomedical Research

Copyright © 2019 Elsevier Inc. All rights reserved.https://doi.org/10.1016/B978-0-12-811829-0.00026-1

cortical regions of the primate brain. As a consequence,there is a premium in developing suitable NHP modelsfrom the perspective of face validity (behavioral pheno-types in primates are more likely to be similar tohumans) and construct validity (the genetics, ontogeny,and neural architecture are likely to be homologous inprimates and humans).

The common marmoset (Callithrix jacchus) representsan NHP species of growing interest in biomedical andbehavioral research. Among the reasons for theincreased use of this species are both practical and man-agement issues. These include their small size(350e500 g), thereby reducing the need for large enclo-sures and animal facilities. Marmosets are also highlyfecund, capable of producing two twin (or larger) littersper year. Relative to cercopithecine models (e.g., rhesusmacaques) or hominoid primates (e.g., chimpanzees),the zoonotic risk posed by marmosets is low.

A more compelling case for the marmoset as a modelin biobehavioral and neuroscience research derives fromthe scientific relevance of a host of features associatedwith the natural history and associated phenotypes ofthe species. Marmosets (and most other callitrichinaeprimates) are obligate twinning species, producing dizy-gotic twins. Twins are dizygotic or fraternal, andbecause of unique features of the development ofplacental and shared embryonic vasculature, twin fe-tuses exchange embryonic stem cells and multiplesignaling molecules across these vascular anastomoses.As a consequence, marmoset twins are chimeric in hem-atopoietically derived tissues [11] and perhaps in othertissues as well [12]. Twinning is thus potentially impor-tant from the perspective of experimental design (e.g.,differential twin phenotypes that develop in spite of ashared intrauterine environment [13] and in the explora-tion of immuneebrain interactions [14e16]). From adevelopmental perspective, marmoset life span trajec-tories are considerably accelerated relative to otherNHP models and humans, passing through the infantstage in several months, reaching sexual and socialmaturity by 18 months of age, and displaying morpho-logical and behavioral signs of senescence within6e8 years [17,18]. From the perspective of social neuro-science, the marmoset represents an exceptional modelfor human sociality [19]. The marmoset shares many fea-tures in the realm of social phenotypes with humans,including the fundamental social unit (small nuclear orextended family social groups [20]), offspring rearingdynamics (cooperative breeding, which entails sharedinfant care by mothers, fathers, and older offspringwithin the family group [21]), and the nature of adultheterosexual social relationships (consisting of manyfeatures associated with social monogamy [22]).

From the perspective of biobehavioral model devel-opment, there is a wealth of normative information on

marmoset brain anatomy and function, including awell-annotated brain atlas that is continuously updated[23], critical information regarding the structure andconnectivity of frontal cortices [24,25], a growing knowl-edge of sensory processing in a number of modalities[26e29], and details on motor control circuitry [30]. Inaddition, in the past 5 years a remarkable suite of toolsin the neurosciences have been developed with and/orapplied to the marmoset central nervous system. Silva[31] recently reviewed advances in imaging of themarmoset brain, and the methodologies include high-resolution structural MRI, functional connectivityamong multiple brain regions during processing in ahost of sensory modalities, and two-photon laser cap-ture microscopy for monitoring neural activity at thelevel of individual neuronal cells. A host of techniqueshave been utilized to alter gene function in themarmoset brain (see review in Ref. [32]), including theproduction of transgenic marmosets relevant for a num-ber of pathologies associated with the brain, lentiviralretrograde vectors for tract tracing in the brain, andshRNA silencing of gene expression in targeted regionsof the brain. Finally, the “proof of concept” for an opto-genetic preparation in the marmoset, via the inductionof channelrhodopsin into multiple cortical regions ofthe brain [33], anticipates the potential to manipulateneuronal function in the marmoset brain within tightspatial and temporal boundaries. Thus, within the neu-rosciences the marmoset may well deserve its recentdesignation as a “supermodel” in the biomedical senseof the word [34].

Organization of this Review

This review of marmosets as behavioral models inbiomedical and behavioral research is organized inline with the Research Domain Criteria (RDoC) taxon-omy. The RDoC movement at the NIH began in lightof decreasing rates of morbidity and mortality from ahost of disease states (e.g., cardiovascular disease, can-cer) as a consequence of investments in and knowledgegained in basic and clinical science, but mortality rateshave remained unchanged for mental illness, anddepending on the disorder, prevalence rates acrossrecent decades have remained stable or have increased[35]. The RDoC as a research tool represents a moveaway from the standard clinical diagnostic protocolsfor psychological disorders [the American PsychiatricAssociation’s DSM-V (www.dsm5.org) and the WorldHealth Organization’s ICD-10, www.who.int/whosis/icd10]. These protocols dictate the diagnosis of mentaldisorders based on clusters of behavioral symptoms, aprocedure that likely poses a potential problem forfundamental research and translational insights intothe underlying mechanisms of these disorders.

26. THE MARMOSET AS A MODEL IN BEHAVIORAL NEUROSCIENCE AND PSYCHIATRIC RESEARCH478

III. RESEARCH APPLICATIONS

Two major problems have been identified with theseapproaches [35e38]. First, the breadth of the diagnosticcriteria for any given disorder is wide, and any givendiagnostic classification includes multiple symptoms,not all of which need be presented by a patient for adiagnostic decision. It is possible therefore that two indi-viduals diagnosed with the same “mental” disorderwould share no symptomology in common, and treatingthe underlying cause therefore becomes problematic. Asan analogy to this issue from a nonpsychiatric context, apatient that presents at the clinic with the symptom“shortness of breath” would not be provided with a sin-gle treatment regimen without further diagnostic evalu-ation. A physician would need to identify theunderlying cause of dyspnea. Once diagnosed, differen-tial therapeutic treatments would be created, dependingon the cause: heart failure, myocardial infarction, lungcongestion, a broken rib, or a state of anxiousness. Sec-ond, there are multiple symptoms that are common indramatically different diagnostic states. For instance,dysfunction in social communication constitutes one ofthe DSM-V diagnostic criteria for schizophrenia spec-trum and other psychotic disorders, social anxiety disor-der, and autism spectrum disorder. Given that the brainregions and the perceptual and motor circuitry underly-ing communication in humans is fairly well established,it would seem logical to look across diagnoses for theunderlying etiology of the symptom(s) that are commonacross psychiatric diagnoses.

The RDoC construct is premised on the notion thatlooking for underlying pathology as a function of thebroad DSM and ICD classifications has not proven tobe useful. Instead, an inversion of the process may bemore fruitful, that is, utilize the knowledge about

normative brain function derived from the breadth ofmodern neuroscience (from genetics to cognition) andidentify the degree to which these multiple mechanismsare disrupted in individuals suffering from “brainillness” [37]. While the utility of the RDoC in a diag-nostic context remains unresolved and is a matter ofconsiderable debate (e.g., Ref. [39]), there is a clear heu-ristic value of the RDoC approach in the context of basicresearch into the neural mechanisms underlying thesymptoms associated with psychopathology. Beginningin 2009, the US National Institutes of Mental Health, un-der the guidance of the director of the institute, ThomasInsel, has developed and elaborated the RDoC constructto guide basic and clinical research in this area. Theoutline of the RDoC construct is shown in Table 26.1.The five major research “Domains,” identified by multi-ple experts over several years, constitute important,broad areas of psychological and behavioral processes.Within each Domain are multiple constructs that nestwithin each Domain. What is critical from the RDoCperspective is that each of the constructs can be opera-tionally defined in multiple model systems, can bemeasured in normative or in perturbed conditions,and can lead to testable, hypothesis-driven scienceregarding the brain circuitry underlying each constructand can be revised and validated according to standardscientific methods.

The literature review that follows will highlight sig-nificant or innovative behavioral protocols that havebeen developed or modified for use in marmosets andtheir relevance for RDoC domains. An exhaustivereview of the all of the findings in each of the RDoCDomains is beyond the scope of this contribution, andthe examples provided within each Domain are meant

TABLE 26.1 RDoC (Research Domain Criteria), NIMHa

Negative Valence

Domain

Positive Valence

Systems Cognitive Systems

Systems for Social

Processes

Arousal/Modulatory

Systems

Acute threat (“fear”) Approach motivation Attention Affiliation andattachment

Arousal

Potential threat(“anxiety”)

Initial responsivenessto reward

Perception Social communication Biological rhythms

Sustained threat Sustainedresponsiveness toreward

Working memory Perception andunderstanding of self

Sleepewake

Frustrative nonreward Reward learning Declarative memory Perception andunderstanding ofothers

Habit Language behavior

Cognitive control

Major domains (neurobehavioral systems) are listed in bold, and constructs (potentially operationalized and measured features of domains) are

listed underneath each domain.aCuthbert and Insel [35]; Insel [36]; Insel and Cuthbert [37].

III. RESEARCH APPLICATIONS

INTRODUCTION 479

to highlight the ways in which the unique attributes ofthe marmoset model can be utilized to explore theneurobiological mechanisms underlying the constructswithin each Domain. The RDoC approach is inspiredby a recent review by Oikonomidis et al. [38], whoapplied the RDoC organization to marmoset models ina few of the RDoC Domains (especially Valence andCognition Domain). This chapter will cover NegativeValence, Social Processes, and Arousal in some detail,and readers interested in a fuller discussion of PositiveValence and Cognitive Systems are referred to Oikono-midis et al. [38], for a more complete review. One finalcomment: considerable information on normative func-tion in marmosets within each Domain is accessiblefrom observations of routine behavior in undisturbedconditions. However, experimental manipulations inthe realm of learning, motivation, cognition, sociality,and arousal and other aspects of marmoset behavioralbiology yield greater confidence in the inference of cau-sality, and hence the focus in this review will be primar-ily on experimental models.

MARMOSET BEHAVIORAL MODELSWITHIN THE CONTEXT OF RDOC

Negative Valence Systems

This domain is characterized by the processing ofstimuli and generating responses to aversive stimuliand events, and the antecedents (anticipation) and con-sequences (long-term function) of these stimuli andevents. The constructs within this domain include theactivation of fear circuits associated with adaptive defen-siveness from real or perceived danger and can be eli-cited by both interoceptive and exteroceptive stimuli.Furthermore, constructs also entail the psychologicaland physiological processing associated with potentialaversive events in which the events are distant, ambig-uous, or have a low or uncertain probability of occurring(anxiety). Further constructs focus on emotional statesthat result from a sustained or uninterrupted exposureto contexts or stimuli that are normally avoided (sus-tained threat), situations involving deprivation or separa-tion from a significant physical or social resource (loss),and responses to the lack of positive outcomes in theface of efforts to produce them (frustrative nonreward).

Two common behavioral models are utilized in theexploration of the responses of marmosets to fear- andanxiety-inducing stimuli. The first involves exposureof marmosets to actual or simulated models of preda-tors, hereafter referred to as the Marmoset PredatorConfrontation Test (MPCT) [40,41]. In the northeasternBrazilian coastal forests, marmosets of the genus Calli-thrix are at predation risk from multiple predators,

including aerial raptors, terrestrial felines, and terres-trial and tree-climbing snakes [42]. As such, these classesof predators constitute significant naturalistic threatsand reliably elicit a cluster of responses even incaptive-house marmosets, including modification inlocomotory patterns like movement along substrate orjumps from location to location, piloerection, movementof the upper body back-and-forth on a stable substrate,and spatial proximity to the predator stimulus. A prom-inent feature of predator exposure is the production ofspecies-specific vocalizations that are observed in wildmarmosets [43]. Among those typically elicited by pred-atory stimuli are vocalizations labeled tsik, egg, peep,and twitter, all of which are carefully characterizedand rigorously quantified by Pistorio et al. [44]. The sec-ond widely used model for fear and anxiety also in-volves confronting marmosets with an actual orpotential stimulus, in this case a human standing inclose proximity to the marmosets’ cage (HITdHumanIntruder Test). While less ecologically relevant than theMPCT, this assessment tool relies on the notion that mar-mosets show fear to the presence of humans, eitherinnately or as a consequence of routine captive husband-ry or testing. Similar behavioral measures are employedin the HIT as in the MPCT.

From the perspective of test validation of these com-mon methods, an explicit comparison of the responsiv-ity of marmosets to the HIT and MPCT revealed thatthe two tests induce similar behavioral responses [45].This investigation demonstrated a strong, positive corre-lation between the two methodologies. Furthermore, in-dividual variability among marmosets in their responseprofiles remained highly stable in independent assess-ments separated by as much as 3.5 years. Neural andneurochemical correlates of high-responsive (low prox-imity to stimulus, high rates of tsik and tsik-egg calls)versus low-responsive marmosets were documented.High-responsive marmosets had lower levels of seroto-nin (5-HT) in the amygdala as measured by microdialy-sis and smaller structural volume of the anteriorcingulate cortex (ACC), an important region for the inte-gration of cognitive and emotional information. Whilethe two methodologies are correlated, another side-by-side comparison of the two protocols revealed greaterbehavioral responsiveness of marmosets to the HITthan to MPCT [46].

The MPCT has provided insight into both the behav-ioral components associated with exposure to fear- andanxiety-inducing stimuli and further details on the neu-ral mechanisms that underlie these behavioral elements.A multivariate principal component analysis (PCA) ofdifferential behavioral responding by marmosets whenconfronted with predator stimuli or human intruders[47] revealed two significant components that accountfor significant variance: emotionality, a PCA dimension

III. RESEARCH APPLICATIONS

26. THE MARMOSET AS A MODEL IN BEHAVIORAL NEUROSCIENCE AND PSYCHIATRIC RESEARCH480

that loads heavily on visual avoidance of the threateningstimulus, reduced locomotion, and greater distancefrom the stimulus, and coping, which loads primarilyon tsik and tsik-egg vocalizations. Support for the roleof predator-induced vocalizations as a coping mecha-nism for marmoset derives from two findings. First,high rates of tsik vocalizations during predator exposureare associated with low baseline cortisol (as measuredby hair cortisol [48]). Second, playback of recorded tsikcalls to marmosets during social isolation (a procedurethat typically elevates cortisol [49e51]) eliminatesseparation-induced cortisol responses [52]. A third po-tential behavioral component relevant to RDoC con-structs (anxiety) derives from the consequences ofrepeated exposure to predatory stimuli, which reliablyelicits elevated rates, relative to baseline or acute expo-sure to predatory stimuli, of scratching, self-grooming,and scent-marking behavior [53].

A third paradigm common in this arena is traditionalaversive Pavlovian (classical) conditioning, in which aninitially neutral conditioned stimulus (CS, e.g., a light, asoft tone, or an environmental context) is paired with anaversive unconditioned stimulus (US, e.g., a loud noise,an exteroceptive shock, a predatory stimulus), and withsufficient associative pairings, the CS acquires the abilityto elicit conditioned fear or anxiety response. Marmosetseasily acquire conditioned responses under these con-texts, as measured by increased autonomic output (heartrate and blood pressure) and enhanced vigilance elicitedby the previously neutral or irrelevant CS [54e57].

Additional behavioral models in the marmoset rela-tive to the Negative Valence Domain have not receivedas much attention as the HITandMPCT but have the po-tential to contribute to the development of unique mea-sures of behavioral phenotypes and their underlyingneurobiology. Among these methods are open-fieldtesting, a modification of the rodent model in whichbehavioral indices of fear/anxiety as well as spatial nav-igation within the open field are assessed [58]. Inanother modification of a standard rodent model forfear/anxiety (the elevated plus maze), Wang et al. [59]tested location preferences of marmosets in a multiple-chambered testing box, in which some compartmentshad opaque walls and others had transparent walls. Ingeneral, marmosets showed a preference for the trans-parent chambers, and this technique may prove usefulin protocols that induce anxiety-like states, either behav-iorally or pharmacologically.

Most of these behavioral protocols have yielded sig-nificant insights into the basic neuroscience of NegativeValence constructs in the marmoset, including structuralregions of the brain that are important regulators ofresponsiveness and in the neurotransmitter systemsthat modulate these brain circuits. Structurally, the orbi-tofrontal cortex (OFC) and PFC in marmosets appear to

independently play an important role in modulating re-sponses to threatening stimuli because bilateral lesionsof either area enhanced emotional components andreduced coping components in an MPCT [47]. In thecontext of aversive Pavlovian conditioning, marmosetswith lesions of the OFC and PFC maintain elevatedconditioned HR responses during extinction trials (CSonly, no US), especially in PFC-lesioned marmosets[54]. In the same study, animals lesioned in either OFCor PFC exhibited elevated emotional responses in theHIT, but reduced coping responses were noted only inthe marmosets with PFC lesions, suggesting a structuralseparation of function for these components of emotionalresponding. Marmosets that are highly responsive ineither HIT or MPCT have reduced volume of the ACC,an important region for the integration of cognitive andemotional information [45]. Finally, localized pharmaco-logical manipulations in restricted brain regions high-light important regulatory circuit nodes. Localdeactivation via a GABA agonist of Area 25 in the PFCreduced the strength of aversive Pavlovian conditioning,whereas deactivation of Area 32 enhanced some auto-nomic measures of conditioned responding [57].

Pharmacological manipulations or measurements inthe context of this Domain have also produced findingsof basic and potentially therapeutic importance. TheGABAA agonist diazepam, a widely used anxiolytictreatment, clearly impacts behavioral and physiologicalmeasures in these paradigms and reveals the importanceof the GABA system in regulating fear/anxiety. Marmo-sets given anxiolytic diazepam in the HIT exhibitreduced threat postures toward the intruder andincreased proximity to the intruder, but administrationof anxiogenics (GABAA inverse agonist or amphetamine)does not produce augmented behavioral reactivity in theHIT [60]. Diazepam also reduces behavioral indices ofanxiety (phee calls, vigilance) and eliminates the normalpreference for spending time in the periphery in theopen-field test [58]. Simultaneous testing of marmosetsin multiple behavioral paradigms has revealed a differ-ential sensitivity of response characteristics to varyingdoses of diazepam. High doses of diazepam reduceresponsiveness of marmosets in both the HIT andMPCT tests, whereas lower doses of diazepam are effec-tive only in the MPCT model [46]. Treatment of marmo-sets with the inverse GABAA agonist FG-7142 increasestsik and tsik-egg vocalizations to comparable levels asthose noted in novelty exposure and MPCT [61]. Otherneurotransmitters no doubt play a role in the manifesta-tion of fear/anxiety phenotypes in marmosets, includingCRH [62], dopamine (DA) [63], and 5-HT [64].

A recent study has provided evidence of a geneticsubstrate for variation in responding in the behavioralprotocols that tap into constructs within the NegativeValence RDoC domain. Santangelo and colleagues [64]

III. RESEARCH APPLICATIONS

MARMOSET BEHAVIORAL MODELS WITHIN THE CONTEXT OF RDOC 481

genotyped marmosets from three sourcesdthe Cam-bridge University colony, the NIH colony, and a wildpopulation in Brazil. They documented variable numbertandem repeat polymorphisms in several loci of the pro-moter region of the 5-HT transport allele, SLC6A4. On apopulation basis, a dinucleotide polymorphism in thethird repeat and two single nucleotide substitutions inthe fourth and 23rd repeat yield the haplotypes AC/C/G in 49.6% of the target colony and CT/C/C in42.4% of a colony, with roughly similar proportions inthe second colony and in wild marmosets. These poly-morphisms have meaningful consequences, as qPCRrevealed the highest SLC6A4 expression in the CT/C/Chaplotype, the lowest in AC/C/G, and intermediateexpression in heterozygotes for this polymorphism. Re-sponses in the HIT revealed behavioral consequences ofthese polymorphisms as well, with AC/C/G marmosetsdisplaying the highest anxiety and lowest coping scoresand CT/C/C marmosets exhibiting low anxiety andhigh coping. Clinical relevance for the use of genetic in-formation for tailoring “personalized” pharmacologicaltreatments for patients on the basis of genotype wasdemonstrated in the marmoset model by the observationthat the selective serotonin uptake inhibitor citalopramenhanced HIT scores in AC/C/G marmosets, butreduced HIT scores in CT/C/C marmosets.

Positive Valence and Cognitive Systems

This section describes behavioral models in two Do-mains because the distinction among them is not as clearin animal models as they are in the context of humanphenotypes. The Positive Valence System addresses con-structs associated with unconditioned and learned/adaptive responses to positive motivational stimuli orcontext. Among the constructs are processes associatedwith approach tendencies to innate or learned stimuli(approach motivation), reward-seeking behavior and he-donic responses to rewards (initial responsiveness toreward attainment), and the biobehavioral consequencesof attaining reward (sustained/longer-term responsivenessto reward attainment). Learning and cognitive processesare reflected by capacities in reinforcement learningand differentiating among stimulus-reinforcement out-comes (reward learning), and the persistence of behav-ioral responses or cognitive processes associated withreward without excessive cognitive resources and/orin the absence of changes in reinforcement outcomes(habit). The Cognitive Systems Domain addresses thecontinuum of processes associated with detecting envi-ronmental stimuli (attention), peripheral and centralcomputational processing of these stimuli in singleand multiple sensory domains (perception), the abilityto integrate, via cognitive and affective systems,

environmental contexts and select adaptive responsesto these contexts (cognitive control), and the encoding,maintenance, and recall of relevant information for spe-cific tasks or goal outcomes (working memory). The con-structs of declarative memory and language (symbolicrepresentation) will not be considered in this review. Acommon protocol for generating experimental protocolsassessing learning and cognition in marmosets, derivedfrom the Cambridge Neuropsychological Test Auto-mated Battery (CANTAB), can be found in Spinelliet al. [65]. Details on training marmosets to provideresponse selections using touch-sensitive computerscreens can be found in Takemoto et al. [66].

Among the most straightforward tasks for assessingthe reinforcing property of stimuli for an organism isthe conditioned place preference (CPP). Typically a spe-cific environmental context (chamber, room, or otherenvironmental cue) is repeatedly paired with the pre-sentation of a potential rewarding stimulus, whereas asecond distinct context does not lead to the stimulus.The subsequent development of a preference for thecontext that predicts the delivery of the stimulus is agauge of the positive rewarding property of the stim-ulus, whereas avoidance of the context is taken as evi-dence that the stimulus is a negative reinforcer [67].Marmosets form a CPP for several reinforcing stimuli,including cocaine [68] and sweet/high-fat food (choco-late) [69].

Reversal learning is a common paradigm for assess-ing generalized discrimination learning, and in partic-ular to assay response persistence in the face of alteredreward consequences. Typically marmosets are pre-sented with two stimulus objects, one associated withreward and the other nonrewarded. Once criterion per-formance (e.g., 75% choice of the correct stimulus) onthat set of stimuli is established, reward contingenciesare reversed; the responding to the previously unre-warded stimulus now leads to reward, and vice versa.This protocol has identified both structural and neuro-chemical correlates of performance in the task. Differentneurotransmitters in the medial caudate nucleus appearto be critical for successful reversal learning in marmo-sets because depletion of DA in the caudate disruptssuccessful reversal learning, whereas depletion of5-HT does not [70]. Lesions of the medial striatum andOFC also disrupt rapid reversal learning, whereasamygdala lesions do not affect the transition to newstimulusereward contingencies [71]. Lacreuse and col-leagues have addressed the role of sex steroid hormonesin modulating learning processes in marmosets in thecontext of the reversal task. Estradiol appears to disruptreversal learning performance because relative to un-treated ovariectomized (OVX) females, OVX femalesgiven estrogen replacement therapy (ERT) took moretrials to reach criterion and exhibited a higher number

III. RESEARCH APPLICATIONS

26. THE MARMOSET AS A MODEL IN BEHAVIORAL NEUROSCIENCE AND PSYCHIATRIC RESEARCH482

of errors in the task [72]. With repeated exposures to setsof reversed reward contingencies, control females madefewer errors to reach criterion performance, whereasestrogen-treated females made more errors. Hormonereplacement appears to have sex-specific effects becauseuntreated and testosterone-treated male marmosets didnot differ on reversal learning performance [73].

Marmosets also perform well in the more complexreversal paradigm, probabilistic discrimination learningand reversal. In this task, two stimuli are not associatedwith reward certainty (e.g., selection of stimulus A leadsto reward [or nonreward] 100% of the time). Rather, bothstimulus cues are associated with differential probabili-ties of reward, in which responding to the “correct” stim-ulus leads to reward in 80% of the trials and nonrewardin 20%, and to the “incorrect” stimulus leads to reward in20% of trials and nonreward in 80% of trials. Perfor-mance on this task is related to serotonergic activity asdemonstrated by site-specific depletion of 5-HT by thelocal administration of 5,7-DHT [74]. Reducing 5-HT ineither the amygdala or the OFC inhibited the acquisitionof the initial task and performance in the reversal phasesof the task. Reducing 5-HT in the amygdala altered themarmoset’s sensitivity to reward contingency, reducingresponding to the “correct”’ stimulus and enhancingresponding to the incorrect stimulus, which results inoverall lower reward density for the marmosets. Thesedata suggest that the amygdala and OFC are critical forlearning and adjusting to changing reward outcomes,and that lower 5-HT in the amygdala alone serves tomediate complex probabilistic reinforcement contin-gencies within the context of reversal learning.

A second common behavioral task within these Do-mains is the barrier-reach task. In the typical experi-ment, food is made available to the marmoset that isavailable by performing a fixed motor reach response(e.g., reaching from the right or left side). This is accom-plished in a number of ways, but placing food inside atransparent or opaque cube with a single open face to ac-cess the food is common. Once a motor response isestablished by the marmoset in food acquisition, thereaching response required to access the food is alteredby modifying the orientation of the cube (e.g., shiftingthe open face from the left to the right). This task isparticularly important for assessing response flexibility(good performance) and perseverative tendencies(poor performance). Performance on this task in marmo-sets is reduced by drugs known to affect neurotrans-mitter systems involved in schizophrenia in humans(the DA D3 receptor agonist PD-128,907 and theNMDA-receptor antagonist ketamine) and enhancedby treatment of marmosets with blonanserin, a pharma-cological agent that antagonizes DA D2/3 receptors andblocks 5-HT2A receptors [75] but is insensitive to estro-gen replacement in OVX female marmosets [72].

In many cases marmosets are useful for experimentsthat explore the interstices between Negative, Positive,and Cognitive Domains. Shiba et al. [56] is a good exem-plar. Marmosets were initially classified as low versushigh reactive to a fear/anxiety stimulus in the standardPavlovianmodel described above. Marmosets were thenpresented with two behavioral tasks: one uses thereward value of a food as a discriminative stimuluswith unexpected reward outcomes (an OFC-dependenttask) and the second task was the barrier-reach taskdescribed earlier in this section, which is a PFC-dependent task. In the first task, marmosets were pre-sented with two transparent boxes, within which werelocated either an unreachable high- or low-preferencefood item. The reward contingencies for touching eitherbox yielded reversed reward outcomesdtouching thebox containing the highly preferred was not rewarded,whereas touching the box with the low-preferencefood item led to the delivery of highly rewarding syrupbread. Thus, the high probability response of touchingthe box containing the highly preferred food had to beinhibited to receive a reward. The magnitude of condi-tioned fear and anxiety in the Pavlovian phase of theexperiment predicted performance on the two tasks.More perseverative error in the first cognitive task(touching the high-preference food box) was associatedwith poorer conditioned vigilance in the Pavloviantask, whereas more perseverative error in the barrier-reach task was associated with lower baseline bloodpressure in the Pavlovian trials.

Systems for Social Processes

The scope of this Domain is self-evident from its titleand includes multiple components of features associ-ated with conspecific sociality. Among these are thequality and patterning of affiliative social interactionsand the consequent development of social bonds (affilia-tion and attachment) and the production, perception,interpretation, and responses to the social stimuli thatmediate these interactions and sustain the social bonds(social communication). Awareness of one’s role in socialinteractions is reflected in the constructs associatedwith self-awareness and self-agency (perception and un-derstanding of self) and in the ability to recognize theidentity of partners, their capacity for agency, and to pre-dict or interpret a partner’s mental state (perception andunderstanding of others).

Affiliative processes in marmosets have beenexplored, including caregivereoffspring attachmentand “pair-bonding” in adult males and females. Withinthe context of caregivereoffspring affiliation, the focushas been on two broad areas: features associated withregulatory processes for caregiver interactions with

III. RESEARCH APPLICATIONS

MARMOSET BEHAVIORAL MODELS WITHIN THE CONTEXT OF RDOC 483

infants and the consequences of variation in the qualityand quantity of caregiving on subsequent offspringdevelopment. With regard to the regulation of parentalcare, correlational studies have implicated a suite ofneuroendocrine modulators of parental care, includingestradiol, testosterone, glucocorticoids, prolactin, andoxytocin (OT) [76e79]. Experimental models to exploreparental motivation include instrumental learning inwhich the reward state is access to infant stimuli [80]and exposure and response to infants or infant-relatedstimuli [81e84]. These tasks have revealed that sensi-tivity to, and interest in, infants and their stimuli areregulated by multiple neuroendocrine systems, and thatthe behavioral measures are often contingent on age orprevious exposure to infants within the social group.

The prominent experimental model to explore theneurobiological and behavioral sequelae to caregivereinfant attachment is the caregiver deprivation modelchampionedbyPryceand colleagues. Thismodel involvesshort-term, 30e120-min separations of infant marmosetfrom caregivers from 2 to 28 days of age [85], and theconsequence of this disrupted caregivereinfant affiliationis widespread across multiple physiological systems,including behavioral, reward, and cognitive processing,responses to stressors, and gene expression in multipleneurotransmitter systems [86,87]. Less-invasive modelsinclude those that examine the impact of normative vari-ation in offspring care on subsequent biobehavioraldevelopment. These models demonstrate that “adverse”patterns of early caregiving (e.g., high rates of infant rejec-tion and transfer among caregivers) also produce long-lasting changes in neurobiology and behavior [88e90].

While there is considerable academic argumentregarding whether marmosets are strictly “sociallymonogamous” (e.g., Ref. [91], and articles therein), thereis no argument that the close social relationships be-tween adult male and female marmosets, characterizedby coordinated activities, high levels of grooming andproximity, cosharing of infant care responsibilities, andjoint defense of territories [92], are distinct from otherprimates considered “nonmonogamous” [53]. Severalbehavioral protocols have been used to evaluate theestablishment, persistence, and strength of this affilia-tive relationship. The first paradigm is partner prefer-ence testing in which marmosets can choose to spendtime and/or interact with partners or opposite-sexstrangers in a simultaneous choice task [93e95]. With re-gard to translational potential, the sensitivity of partnerpreference to environmental and social context shouldbe considered a strength, given the complexities ofhuman social relationships. Partner preference andpair-directed affiliation in marmosets is sensitive to OTmanipulations, in that treatment with an OT antagonistreduces affiliative behavior in newly paired marmosets[95] and reduces interest and interactions with an

unfamiliar opposite-sex partner in well-establishedpairs [94]. In addition, correlational studies haverevealed that closely bonded family members, includingbreeding pairs, exhibit highly correlated levels of uri-nary OT [96], providing additional support for nonapep-tide involvement in social affiliation in marmosets.

The second task for assessing pair-related affiliationand attachment involves measuring the behavioral andphysiological responses to temporary separation frompartners [49,51,97,98]. In heterosexual pairs, partner sep-aration and housing in a novel environment is associ-ated with elevated cortisol, and the presence of thepartner [51], but not an opposite-sex stranger [98], canreduce or buffer the stress response in this context. Infact, presentation of phee calls from the partner butnot calls from an opposite-sex stranger during separa-tion also buffers the stress response [49]. In long-termsame-sex pairs of marmosets, separation stress is buff-ered by both partners and same-sex strangers in males,but the presence of neither partners nor strangers servedto buffer cortisol responses to separation in female mar-mosets [97]. This would seem an appropriate model sys-tem to explore dynamic changes in the socially mediatedchanges in the HPA axis, and the role of the hippocam-pus, temporal cortex, and PFC in translating these socialbuffering effects into physiological and behavioralregulations.

Marmosets have a rich vocal repertoire that mediatesa host of intra- and intergroup social interactions.Knowledge of the information content of these socialsignals and the ways in which they are centrally evalu-ated in the brain can serve as a useful base to explorenormative and atypical vocal communication processes.A standard protocol for assaying information content inmarmoset calls involves the playback of recordedmarmoset vocalizations and documenting behavioraldifferences in response to these calls. Information con-tent regarding signalers in marmoset phee calls wasexperimentally evaluated by Smith et al. [99]. Normativeacoustic characteristics of adult phee calls were quanti-fied, and sex differences were noted in numerous acous-tic parameters. Synthesized vocalizations that includedthe important differences in frequency components ofthe calls were generated, and the responses of marmo-sets to playback of synthetic and natural calls did notdiffer; both calls elicited enhanced vigilance, and vigi-lance was greater for male-typical than female-typicalcalls. Systematic manipulation of individual compo-nents of call structure eliminated differential respond-ing to synthetic phee calls, suggesting that perceptionand interpretation of the sex of the caller requires aholistic, multivariate assessment of call characteristics.Similar work has been conducted on natural versussynthetic calls in multiple call types, with neural confir-mation that these two classes of stimuli are processed

III. RESEARCH APPLICATIONS

26. THE MARMOSET AS A MODEL IN BEHAVIORAL NEUROSCIENCE AND PSYCHIATRIC RESEARCH484

similarly [100]. The potential for studying the behavioraland neural processing of acoustic social signals in realtime in marmosets has been enhanced by the develop-ment of a sophisticated behavioral model of communica-tive “conversations” [101,102]. In this model, recordedexemplars of marmosets are played back to subjectswith varied probabilities and timing, as “virtual marmo-sets” in an interaction. The patterning of the vocalizationby the real marmoset is contingent on these manipula-tions in calls from the virtual marmosets [103]. Giventhat vocal exchange patterns appear to be learned duringontogeny in marmosets [104], this behavioral techniquehas implications for both adult vocal interactions andvocal development.

Several groups have utilized the expression of imme-diate early gene (IEG) products, signals of early tran-scriptional activity within neurons that are responsiveto stimuli, to identify brain regions associated with theperception and processing of species-specific acousticsignals in marmosets. Marmosets listening to species-specific vocalizations have differential IEG expressionin regions of the brain that differ by task (listening tocalls, producing calls, and engaging in antiphonal call-ing). Perception of calls is associated with IEG expres-sion (cFOS) in the ventrolateral PFC and regions of theauditory cortex: producing calls is associated withdistinct activation in the premotor cortex, and vocal ex-changes activate the perirhinal cortex in addition to theareas previously mentioned [105]. Using a different IEGmarker (Egr-1), Simoes et al. [106] demonstrated thatvocal exchanges also enhanced IEG expression notonly in the PFC but also in the ACC. Single-cell record-ings in the premotor cortex also support the role of thisregion in vocal production in marmosets [107]. For amore complete review of the neurobiology of marmosetvocal communication see Refs. [108,109].

Given that marmosets utilize scent-marking behaviorextensively in social interactions and have individual-and sex-specific olfactory “signatures” [110,111], theneurobiology underlying olfactory communicationwould appear ripe for exploitation. fMRI studies haverevealed enhanced BOLD signaling in hypothalamic re-gions ofmale marmosets exposed to the odors of sexuallyreceptive females [112] and in cortical and subcortical re-gions associated with reward and emotional processes[113], suggesting a complex integration of informationacross the brain. The development of a behavioral condi-tioning model demonstrating that neutral stimuli can beconditioned with sexually arousing female stimuli inmale marmosets [114] anticipates future sophisticatedexperimental assessments of social olfaction and its pe-ripheral and central processing in marmosets.

Evidence for the ability of marmosets to perceive andunderstand others derives from several behavioral pro-tocols. A fundamental demonstration that marmosets

can perceive that other conspecifics have agency andgoals derives from a paradigm in which marmosetswatch video presentations of one of three actorsspending time with and exploring one of two distinctobjects [115]. After a short gap, marmosets are presentedwith a second video, with the actor either interactingwith the same object but in a different location (expectedoutcome) or interacting with the second object that isnow in the same location as the object of interest in thefirst video (unexpected). As with looking time at visualstimuli in human infants, marmosets spend more timelooking at the unexpected than the expected object.The critical manipulation in this experiment is the na-ture of the model in the video: a marmoset, a quadru-pedal marmoset-ish robot, or a black box. As expected,marmosets spent more time looking at the unexpectedoutcome, but only when the actor was a marmoset orthe robot. When given access to the actual objects por-trayed in the video, marmosets spent more time withthe “correct” object but again only when the video actorwas a marmoset or robot, demonstrating social learningonly when the model possessed “agency.”

Marmosets provide an excellent opportunity to studythe perception of others in paradigms that require two ormore partners to acquire or distribute food rewards,which speaks to the importance of social life in many as-pects of the behavioral biology of this species. Severaltasks have been employed that require cooperation oraltruism among interacting marmosets, in which bothmarmosets must respond cooperatively to access fooditems [116] or act “altruistically” to provide food toothers, while not receiving rewards themselves, eitherin a dyadic context [117,118] or in a context in whichan individual provides support to multiple group mem-bers [119,120]. Marmosets easily learn to perform coop-erative tasks, although like many features of marmosetbiology, the ease with which cooperation occurs variesas a function of relatedness and status of cooperatingpartners [116]. Marmosets do exhibit “other-regarding”sharing of food in the altruistic food-sharing paradigms,although again the nature of social relationships be-tween the donors and recipients, and the social role ofthe “donor” in the family group determines the proba-bility of altruistic responding [117,118,120]. OT is relatedto these measures of prosociality. The degree of OT syn-chrony among dyad members predicts levels of proso-ciality, with higher dyadic synchrony in OT levelscorrelated with a higher likelihood of provisioningfood to the partner [120]. Experimental manipulationof OT paints a different story; OT reduces sharing inboth adult males and females with unfamiliaropposite-sex recipients but does not alter altruisticsharing with long-term pairmates [118]. It is certainlythe case that OTsynchronymeasured in urinary samplesprovides a different index of OT activity than

III. RESEARCH APPLICATIONS

MARMOSET BEHAVIORAL MODELS WITHIN THE CONTEXT OF RDOC 485

pharmacological manipulation of OT, but nonethelessboth recent papers point to an important role for OT incooperative social behavior. These paradigms have thepotential to serve as springboards for more detailed ex-plorations of neuroendocrine mediation of complex so-cial traits such as cooperation and altruism and thecortical decision-making circuits that regulate complexdecision-making in a social context.

The ability to perceive and understand self-awareness and agency is a difficult concept to study inpre- and nonverbal organisms, and this certainly appliesto marmosets. In spite of its faults (of which there aremany, e.g., Ref. [121]), the predominant test in NHPfor self-awareness remains the “mirror test” [122]. Whileno studies have been published regarding the perfor-mance of common marmosets on this task, two papershave appeared on other closely related callitrichinaespecies. Pygmy marmosets with lengthy exposure tomirrors in their home cages showed no evidence ofmirror-mediated self-directed behavior, although theyappeared to use the mirrors to gain visual access toneighboring groups and exhibited aggressive displaystoward individuals in these groups [123]. Thus, whilepygmy marmosets show no evidence of self-awareness, they “understand” or at least utilize visualinformation provided by mirrors. Likewise, cotton-toptamarins with long-term exposure to mirrors spentconsiderable time looking into the mirrors, displayedrepeated actions in front of the mirrors, but showed noevidence of self-directed behavior that appeared to bemirror-guided [124]. These negative findings andabsence of findings point to two potential interpreta-tions: (1) like most nonhominoid primates [122], marmo-sets do not possess the cognitive complexity to expressself-awareness or (2) we as investigators lack the clever-ness and sophistication to assess this question.

Arousal/Modulatory Systems

The final Domain covers neural systems associatedwith the coordination of responses to environmentalstimuli to achieve a homeostatic state or to effect appro-priate responses to environmental change and hence anappropriate allostatic state. Constructs in this Domaininclude generating an appropriate physiological statefor the current interoceptive or exteroceptive context(arousal), maintaining appropriate 24-h patterns acrossmultiple organ systems and behavior (circadianrhythms), and the organization of sleep/wake states inthe service of optimizing physiological and behavioralfunctions (sleep and wakefulness).

Marmosets have proven to be excellent models forphotic and nonphotic regulation of the circadian rhythmand represent a model species that is diurnal or

crepuscular, as opposed to the commonly usednocturnal rodent models. The free-running cycle (s) inthis species under constant 24-h dim lighting conditionshas been estimated in multiple studies (e.g.,Refs. [125e127]), and with little variation it has beenestablished as 23.3 h. Like other species, entrainmentof the circadian cycle under constant 24-h illuminationcan be accomplished by photic stimulation with a vari-ety of frequency and timing manipulations [125,126].What is particularly compelling from an RDoC perspec-tive is the sensitivity of circadian rhythmicity to devel-opmental and social variables. There are important ageand sex differences in circadian activity profiles in mar-mosets, with juveniles more active than adults throughthe active phase, adults showing earlier activity onsetthan juveniles, adult males showing the earliest phase-onset and -offset for the active period, and a shifting tolater activity onset in postpubertal than prepubertalmarmosets [128]. Daily rhythms appear to be exquisitelysensitive to social contexts. Entrainment of activity pro-files is more tightly coupled in individuals within a fam-ily group than between family groups, and the strongestcoupling occurs between sibling twin pairs, followednext most closely by maleefemale breeding pairs[129]. While direct and preferential social interaction,cofeeding, and coincidental allogrooming could beamong the mechanisms by which tight activity couplingis mediated, there is also evidence that vocal communi-cation signals can also play an important role. Rates ofphee calling in marmosets exhibit a bimodal distributionduring the light phase, with the highest rates occurringshortly after light onset. Entrainment to “light onset”can occur in marmosets housed indoors under constant24-h dim light conditions, under conditions where theycan hear phee calls from marmosets exposed to normallight cycles [127].

Marmosets also appear to be a useful model forexploring the circadian and thermoregulatory changesthat are associated with menopause in women [130].Gervais and colleagues [131] utilized implanted biote-lemetry to monitor sleep quality (assessed by EEG)and core body temperature in OVX female marmosetsbefore and after ERT. Both high and low doses of estro-gen (6 or 12 mg/kg day) significantly reduced core bodytemperature during the night phase. EEG analysesrevealed fewer nighttime arousals and higher deltawave power, both indices of better-quality sleep.

Circulating plasma cortisol concentrations in marmo-sets follow the circadian pattern typical for diurnalmammals, with rising concentrations as the light phaseapproaches, highest concentrations in samples collectedshortly after activity onset in the morning, with fallingconcentrations throughout the day until the nadir2e4 h after the onset of the dark phase [132]. Similar pat-terns derive from samples that can be collected

III. RESEARCH APPLICATIONS

26. THE MARMOSET AS A MODEL IN BEHAVIORAL NEUROSCIENCE AND PSYCHIATRIC RESEARCH486

noninvasively, including urine [50], saliva [133], andfecal samples [134]. The timing and patterning of circa-dian cortisol rhythms appear to be sensitive to socialcontext [89,90]. Morning cortisol concentrations in iso-lated, singly housed marmosets and marmosets in un-stable groups with high aggression are higher thanthose in marmosets housed in stable breeding pairs orfamily groups. Furthermore, rather than exhibiting thenormative reduction in cortisol in the afternoon, cortisolconcentrations remain elevated in those marmosetsliving in unstable or isolated social contexts, and mea-sures of behavioral arousal are strongly correlated withafternoon cortisol levels. In light of the fact that of themultiple parameters that can be measured in the HPAaxis, the biggest predictor of risk for depression andother disorders is elevated afternoon cortisol, alteredcircadian cortisol in marmosets may therefore be a use-ful proxy measure in models for depression andpsychopathology.

Assessing circadian rhythms in marmosets demon-strates the ways in which behavioral models can cutacross RDoC domains in meaningful ways, includingArousal and Positive Valence. Marmosets not surpris-ingly form a CPP for a context and location in whichfood is available [135]. In an interesting twist, thestrength of the CPP was contingent on a match betweenthe time at which associative learning trials were con-ducted (either in the early morning or late afternoon)and when CPP was assessed. CPP performance wasrobust when the time of assessment matched the timeof training but was effectively eliminated when thetimes did not match. This result suggests that the phaseof the circadian cycle forms an important part of thebroad stimulus properties (interoceptive and extero-ceptive) to which reward learning in marmosets issensitive.

SUMMARY

The utility of the marmoset model in behavioralneuroscience is borne out by an NCBI PubMed searchon Marmoset AND (Brain OR Behavior), which yieldsover 1900 citations, 1300 of which have been publishedsince the year 2000. Interpreting this vast literature is adaunting task, but one that is given structure and be-comes more heuristically valuable by adopting theRDoC classification matrix, as illustrated by Oikonomi-dis et al. [38] and this chapter. This review indicatesthat explorations of the nature of brainebehavior rela-tionships in the marmoset have the potential to speakto all Domains and most constructs in the RDoC frame-work. This is especially so in the case of the Social Pro-cesses Domain, given the analogous similarities in thesocial systems of the marmoset with those of humans

and the homologies between the human and thenonhuman primate brain. The unique position of themarmoset in this regard has been recognized by others[19,136e138].

This review has identified experimental behavioralmodels that have been employed in the service of under-standing the ways in which the brain generates behav-ioral and social patterns in marmosets and also theways in which brain structure and function in turn arealtered by environmental and social experiences. Thedegree of sophistication of the neurobiological methodsthat have been employed vary primarily as a function ofthe degree to which the neurobiological measuresrequire invasive or restrictive protocols (e.g., single-cellrecording, restraint within an fMRI) or can be conductedwith minimally or noninvasive protocols (e.g., measure-ment of neuropeptide metabolites in urine; oral or nasaladministration of neuropeptides). Obviously, the degreeof resolution with regard to neural function and organi-zation varies as a consequence of the method used, butso, too, does the ability to study complex, interactivebehavior in a complex, interactive, and dynamic socialgroup. In a real sense, then, a form of Heisenberg’s Un-certainty Principle applies in this context. Simply para-phrased, Heisenberg pointed out that two keycomponents were required to understand some featuresof atomic physicsda particle’s location and its mo-mentum. By measuring either location or momentum,the ability to measure the other was compromised. Tocomplete the analogy as it applies in the current context,methodologies that provide greater precision, resolu-tion, and timing of neural activity that mediate complexaffective, cognitive, and social processes require lesscomplexity and sophistication in the behavioral systemsthat are under study. This conundrum echoes the cogentarguments of Krakauer et al. [139] that generating mean-ingful basic and translational research in neurosciencerequires that experiments be designed and carried outin light of species-specific evolutionary adaptations,with an eye toward species-specific sensoryemotorcapacities, and an emphasis on appropriately complexsocial contexts.

The new methodological innovations that were high-lighted at the outset of this chapter may provide a way toreduce the Uncertainty Principle as it applies tomarmoset brainebehavior relationships. These includethe potential to remotely activate selective neural cir-cuits in freely behaving marmosets via optogenetic stim-ulation or to alter gene expression during developmentvia transgenic models or in adulthood via shRNAsilencing. In that sense, the utility of the marmosetmodel in behavioral research will likely grow in its stat-ure as an important transitional and translationalbiomedical model in the neurosciences increasesda ‘su-permodel’, indeed!

III. RESEARCH APPLICATIONS

SUMMARY 487

Acknowledgments

The preparation of this chapter and the research by the authordescribed herein were supported in part by funds from the NationalInstitutes of Health (HD04882, HD089147).

References

[1] Whiteford HA, Degenhardt L, Rehm J, Baxter AJ, Ferrari AJ,Erskine HE, Charlson FJ, Norman RE, Flaxman AD, Johns N,Burstein R, Murray CJ, Vos T. Global burden of disease attribut-able to mental and substance use disorders: findings from theGlobal Burden of Disease Study 2010. Lancet 2013;382:1575e86.https://doi.org/10.1016/S0140-6736(13)61611-6.

[2] Walker ER, McGee RE, Druss BG. Mortality in mental disordersand global disease burden implications: a systematic review andmeta-analysis. JAMA Psychiatry 2015;72:334. https://doi.org/10.1001/jamapsychiatry.2014.2502.

[3] Perelman P, Johnson WE, Roos C, Seuanez HN, Horvath JE,Moreira MA, Kessing B, Pontius J, Roelke M, Rumpler Y,others. A molecular phylogeny of living primates. PLoS Genet2011;7:e1001342.

[4] Geschwind DH, Rakic P. Cortical evolution: judge the brain by itscover. Neuron 2013;80:633e47. https://doi.org/10.1016/j.neuron.2013.10.045.

[5] Dunbar RI, Shultz S. Evolution in the social brain. Science 2007;317:1344e7.

[6] Nestler EJ, Hyman SE. Animal models of neuropsychiatricdisorders. Nat Neurosci 2010;13:1161e9. https://doi.org/10.1038/nn.2647.

[7] Stewart AM, Kalueff AV. Developing better and more valid ani-mal models of brain disorders. Behav Brain Res 2015;276:28e31.https://doi.org/10.1016/j.bbr.2013.12.024.

[8] Belmonte JCI, Callaway EM, Caddick SJ, Churchland P, Feng G,Homanics GE, Lee K-F, Leopold DA, Miller CT, Mitchell JF,Mitalipov S, Moutri AR, Movshon JA, Okano H, Reynolds JH,Ringach DL, Sejnowski TJ, Silva AC, Strick PL, Wu J, Zhang F.Brains, genes, and primates. Neuron 2015;86:617e31. https://doi.org/10.1016/j.neuron.2015.03.021.

[9] Wise SP. Forward frontal fields: phylogeny and fundamentalfunction. Trends Neurosci 2008;31:599e608. https://doi.org/10.1016/j.tins.2008.08.008.

[10] Kaiser T, Feng G. Modeling psychiatric disorders for developingeffective treatments. Nat Med 2015;21:979e88. https://doi.org/10.1038/nm.3935.

[11] Sweeney CG, Curran E, Westmoreland SV, Mansfield KG,Vallender EJ. Quantitative molecular assessment of chimerismacross tissues in marmosets and tamarins. BMC Genom 2012;13:98.

[12] Ross CN, French JA, Ortı G. Germ-line chimerism and paternalcare in marmosets (Callithrix kuhlii). Proc Natl Acad Sci 2007;104:6278e82.

[13] French JA, Smith AS, Gleason AM, Birnie AK, Mustoe A,Korgan A. Stress reactivity in young marmosets (Callithrix geof-froyi): ontogeny, stability, and lack of concordance among co-twins. Horm Behav 2012;61:196e203.

[14] Genain CP, Hauser SL. Experimental allergic encephalomyelitisin the New World monkey Callithrix jacchus. Immunol Rev2001;183:159e72.

[15] Neehus A-L, Wistuba J, Ladas N, Eiz-Vesper B, Schlatt S,Muller T. Gene conversion of the major histocompatibilitycomplex class I Caja-G in common marmosets (Callithrixjacchus). Immunology 2016;149:343e52. https://doi.org/10.1111/imm.12652.

[16] Rolleke U, Flugge G, Plehm S, Schlumbohm C, Armstrong VW,Dressel R, Uchanska-Ziegler B, Ziegler A, Fuchs E, Czeh B,Walter L. Differential expression of major histocompatibilitycomplex class I molecules in the brain of a New World monkey,the common marmoset (Callithrix jacchus). J Neuroimmunol2006;176:39e50. https://doi.org/10.1016/j.jneuroim.2006.04.015.

[17] Abbott DH, Barnett DK, Colman RJ, Yamamoto ME, Schultz-Darken NJ. Aspects of common marmoset basic biology andlife history important for biomedical research. Comp Med2003;53:339e50.

[18] Schultz-Darken N, Braun KM, Emborg ME. Neurobehavioraldevelopment of common marmoset monkeys. Dev Psychobiol2016;58:141e58.

[19] Burkart JM, Hrdy SB, Van Schaik CP. Cooperative breeding andhuman cognitive evolution. Evol Anthropol Issues News Rev2009;18:175e86.

[20] Nievergelt CM, Digby LJ, Ramakrishnan U, Woodruff DS. Ge-netic analysis of group composition and breeding system in awild common marmoset (Callithrix jacchus) population. Int J Pri-matol 2000;21:1e20.

[21] Erb WM, Porter LM. Mother’s little helpers: what we know (anddon’t know) about cooperative infant care in callitrichines: ERBand PORTER. Evol Anthropol Issues News Rev 2017;26:25e37.https://doi.org/10.1002/evan.21516.

[22] French JA, Cavanaugh J, Mustoe AC, Carp SB, Womack SL. So-cial monogamy in nonhuman primates: phylogeny, phenotype,and physiology. J Sex Res 2018;55:410e34 (in press).

[23] Okano H, Mitra P. Brain-mapping projects using the commonmarmoset. Neurosci Res 2015;9:3e7. https://doi.org/10.1016/j.neures.2014.08.014.

[24] Burman KJ, Rosa MGP. Architectural subdivisions of medial andorbital frontal cortices in the marmoset monkey (Callithrixjacchus). J Comp Neurol 2009;514:11e29. https://doi.org/10.1002/cne.21976.

[25] Roberts AC, Tomic DL, Parkinson CH, Roeling TA, Cutter DJ,Robbins TW, Everitt BJ. Forebrain connectivity of the prefrontalcortex in the marmoset monkey (Callithrix jacchus): an antero-grade and retrograde tract-tracing study. J Comp Neurol 2007;502:86e112. https://doi.org/10.1002/cne.21300.

[26] Eliades SJ, Wang X. Neural substrates of vocalization feedbackmonitoring in primate auditory cortex. Nature 2008;453:1102e6.

[27] Mitchell JF, Leopold DA. The marmoset monkey as a model forvisual neuroscience. Neurosci Res 2015;93:20e46. https://doi.org/10.1016/j.neures.2015.01.008.

[28] Moriya-Ito K, Tanaka I, Umitsu Y, IchikawaM, Tokuno H. The ol-factory bulb and the number of its glomeruli in the commonmarmoset (Callithrix jacchus). Neurosci Res 2015;93:158e63.https://doi.org/10.1016/j.neures.2014.12.007.

[29] Solomon SG, Rosa MGP. A simpler primate brain: the visual sys-tem of the marmoset monkey. Front Neural Circuits 2014;8.https://doi.org/10.3389/fncir.2014.00096.

[30] Bakola S, BurmanKJ, RosaMGP. The cortical motor system of themarmoset monkey (Callithrix jacchus). Neurosci Res 2015;93:72e81. https://doi.org/10.1016/j.neures.2014.11.003.

[31] Silva AC. Anatomical and functional neuroimaging in awake,behaving marmosets: neuroimaging in Awake. Behaving Mar-mosets Dev Neurobiol 2017;77:373e89. https://doi.org/10.1002/dneu.22456.

[32] Watakabe A, Sadakane O, Hata K, Ohtsuka M, Takaji M,Yamamori T. Application of viral vectors to the study of neuralconnectivities and neural circuits in the marmoset brain. DevNeurobiol 2016.

[33] MacDougall M, Nummela SU, Coop S, Disney A, Mitchell JF,Miller CT. Optogenetic manipulation of neural circuits in awakemarmosets. J Neurophysiol 2016;116:1286e94.

III. RESEARCH APPLICATIONS

26. THE MARMOSET AS A MODEL IN BEHAVIORAL NEUROSCIENCE AND PSYCHIATRIC RESEARCH488

[34] Cyranoski D. Marmoset model takes centre stage. Nature 2009;459:492e492. https://doi.org/10.1038/459492a.

[35] Cuthbert BN, Insel TR. Toward the future of psychiatric diag-nosis: the seven pillars of RDoC. BMC Med 2013;11. https://doi.org/10.1186/1741-7015-11-126.

[36] Insel T. The NIMH research domain criteria (RDoC) project: pre-cision medicine for psychiatry. Am J Psychiat 2014;171:395e7.

[37] Insel TR, Cuthbert BN. Brain disorders? Precisely Sci 2015;348:499e500. https://doi.org/10.1126/science.aab2358.

[38] Oikonomidis L, Santangelo A, Shiba Y, Clarke H, Robbins TW,Roberts AC. A dimensional approach to modelling symptomsof neuropsychiatric disorders in the marmoset monkey. DevNeurobiol 2016.

[39] Lilienfeld SO, Treadway MT. Clashing diagnostic approaches:DSM-ICD versus RDoC. Annu Rev Clin Psychol 2016;12:435e63.https://doi.org/10.1146/annurev-clinpsy-021815-093122.

[40] Barros M, Major RS, Huston JP, Tomaz C. Predatory stress as anexperimental strategy to measure fear and anxiety-related be-haviors in non-human primates. Rev Neurosci 2008;19:157.

[41] Barros M, Tomaz C. Non-human primate models for investi-gating fear and anxiety. Neurosci Biobehav Rev 2002;26:187e201.

[42] Ferrari SF, Ferrari MAL. Predator avoidance behaviour in thebuffy-headed marmoset, Callithrix flaviceps. Primates 1990;31:323e38. https://doi.org/10.1007/BF02381104.

[43] Bezerra BM, Souto A. Structure and usage of the vocal repertoireof Callithrix jacchus. Int J Primatol 2008;29:671e701. https://doi.org/10.1007/s10764-008-9250-0.

[44] Pistorio AL, Vintch B, Wang X. Acoustic analysis of vocal devel-opment in a New World primate, the common marmoset (Calli-thrix jacchus). J Acoust Soc Am 2006;120:1655e70.

[45] Mikheenko Y, Shiba Y, Sawiak S, Braesicke K, Cockcroft G,Clarke H, Roberts AC. Serotonergic, brain volume and atten-tional correlates of trait anxiety in primates. Neuropsychophar-macology 2015;40:1395e404. https://doi.org/10.1038/npp.2014.324.

[46] Cagni P, Goncalves I, Ziller F, Emile N, Barros M. Humans andnatural predators induce different fear/anxiety reactions andresponse pattern to diazepam in marmoset monkeys. PharmacolBiochem Behav 2009;93:134e40.

[47] Shiba Y, Kim C, Santangelo AM, Roberts AC. Lesions of eitheranterior orbitofrontal cortex or ventrolateral prefrontal cortexin marmoset monkeys heighten innate fear and attenuate activecoping behaviors to predator threat. Front Syst Neurosci 2015;8.https://doi.org/10.3389/fnsys.2014.00250.

[48] Clara E, Tommasi L, Rogers LJ. Social mobbing calls in commonmarmosets (Callithrix jacchus): effects of experience and associ-ated cortisol levels. Anim Cogn 2008;11:349e58.

[49] Rukstalis M, French JA. Vocal buffering of the stress response:exposure to conspecific vocalizations moderates urinary cortisolexcretion in isolated marmosets. Horm Behav 2005;47:1e7.

[50] Smith TE, French JA. Psychosocial stress and urinary cortisolexcretion in marmoset monkeys. Physiol Behav 1997;62:225e32.

[51] Smith TE, McGreer-Whitworth B, French JA. Close proximity ofthe heterosexual partner reduces the physiological and behav-ioral consequences of novel-cage housing in black tufted-earmarmosets (Callithrix kuhli). Horm Behav 1998;34:211e22.

[52] Cross N, Rogers LJ. Mobbing vocalizations as a coping responsein the common marmoset. Horm Behav 2006;49:237e46.

[53] Barros M, de Souza Silva MA, Huston JP, Tomaz C. Multibeha-vioral analysis of fear and anxiety before, during, and afterexperimentally induced predatory stress in Callithrix penicillata.Pharmacol Biochem Behav 2004;78:357e67.

[54] Agustın-Pavon C, Braesicke K, Shiba Y, Santangelo AM,Mikheenko Y, Cockroft G, Asma F, Clarke H, Man M-S,Roberts AC. Lesions of ventrolateral prefrontal or anteriororbitofrontal cortex in primates heighten negative emotion.

Biol Psychiatry 2012;72:266e72. https://doi.org/10.1016/j.biopsych.2012.03.007.

[55] Clarke HF, Horst NK, Roberts AC. Regional inactivationsof primate ventral prefrontal cortex reveal two distinctmechanisms underlying negative bias in decision making. ProcNatl Acad Sci 2015;112:4176e81. https://doi.org/10.1073/pnas.1422440112.

[56] Shiba Y, Santangelo AM, Braesicke K, AgustAn-PavA3n C,Cockcroft G, Haggard M, Roberts AC. Individual differences inbehavioral and cardiovascular reactivity to emotive stimuli andtheir relationship to cognitive flexibility in a primate model oftrait anxiety. Front Behav Neurosci 2014;8. https://doi.org/10.3389/fnbeh.2014.00137.

[57] Wallis CU, Cardinal RN, Alexander L, Roberts AC, Clarke HF.Opposing roles of primate areas 25 and 32 and their putativerodent homologs in the regulation of negative emotion. ProcNatl Acad Sci 2017;114:E4075e84. https://doi.org/10.1073/pnas.1620115114.

[58] Cagni P, Komorowski M, Melo GC, Lima T, Barros M. Diazepam-induced decrease in anxiety-like behaviors of marmoset mon-keys exposed to a novel open-field. Pharmacol Biochem Behav2012;100:518e21.

[59] Wang Y, Fang Q, Gong N. A modified light-dark box test for thecommon marmoset. Neurosci Bull 2014;30:394e400. https://doi.org/10.1007/s12264-013-1426-8.

[60] Carey GJ, Costall B, Domeney AM, Jones DNC, Naylor RJ.Behavioural effects of anxiogenic agents in the commonmarmoset. Pharmacol Biochem Behav 1992;42:143e53.

[61] Kato Y, Gokan H, Oh-Nishi A, Suhara T, Watanabe S,Minamimoto T. Vocalizations associated with anxiety and fearin the common marmoset (Callithrix jacchus). Behav Brain Res2014;275:43e52.

[62] French JA, Fite JE, Jensen H, Oparowski K, Rukstalis MR, Fix H,Jones B, Maxwell H, Pacer M, Power ML, others. Treatment withCRH-1 antagonist antalarmin reduces behavioral and endocrineresponses to social stressors in marmosets (Callithrix kuhlii). Am JPrimatol 2007;69:877e89.

[63] Rogers DC, Costall B, Domeney AM, Gerrard PA, Greener M,Kelly ME, Hagan JJ, Hunter AJ. Anxiolytic profile of ropinirolein the rat, mouse and common marmoset. Psychopharmacology(Berl) 2000;151:91e7.

[64] Santangelo AM, Ito M, Shiba Y, Clarke HF, Schut EH,Cockcroft G, Ferguson-Smith AC, Roberts AC. Novel primatemodel of serotonin transporter genetic polymorphisms associ-ated with gene expression, anxiety and sensitivity toantidepressants. Neuropsychopharmacology 2016;41:2366e76.https://doi.org/10.1038/npp.2016.41.

[65] Spinelli S, Pennanen L, Dettling AC, Feldon J, Higgins GA,Pryce CR. Performance of the marmoset monkey on computer-ized tasks of attention and working memory. Cogn Brain Res2004;19:123e37.

[66] Takemoto A, Izumi A, Miwa M, Nakamura K. Development of acompact and general-purpose experimental apparatus with atouch-sensitive screen for use in evaluating cognitive functionsin common marmosets. J Neurosci Methods 2011;199:82e6.https://doi.org/10.1016/j.jneumeth.2011.04.029.

[67] Prus AJ, James JR, Rosecrans JA. Conditioned place preference.In: Buccafusco JJ, editor. Methods of behavior analysis in neuro-science, frontiers in neuroscience. Boca Raton (FL): CRC Press/Taylor & Francis; 2009.

[68] Barros M, Dempster EL, Illott N, Chabrawi S, Maior RS,Tomaz C, De Souza Silva MA, Huston JP, Mill J, Muller CP.Decreased methylation of the NK3 receptor coding gene(TACR3) after cocaine-induced place preference in marmosetmonkeys: cocaine CPP in monkeys. Addict Biol 2013;18:452e4.https://doi.org/10.1111/j.1369-1600.2011.00409.x.

III. RESEARCH APPLICATIONS

REFERENCES 489

[69] Duarte RBM, Patrono E, Borges AC, Cesar AAS, Tomaz C,Ventura R, Gasbarri A, Puglisi-Allegra S, Barros M. Consump-tion of a highly palatable food induces a lasting place-conditioning memory in marmoset monkeys. Behav Processes2014;107:163e6. https://doi.org/10.1016/j.beproc.2014.08.021.

[70] Clarke HF, Hill GJ, Robbins TW, Roberts AC. Dopamine, but notserotonin, regulates reversal learning in the marmoset caudatenucleus. J Neurosci 2011;31:4290e7. https://doi.org/10.1523/JNEUROSCI.5066-10.2011.

[71] Clarke HF, Robbins TW, Roberts AC. Lesions of the medial stria-tum in monkeys produce perseverative impairments duringreversal learning similar to those produced by lesions of the orbi-tofrontal cortex. J Neurosci 2008;28:10972e82. https://doi.org/10.1523/JNEUROSCI.1521-08.2008.

[72] Lacreuse A, Chang J, Metevier CM, LaClair M, Meyer JS,Ferris CM. Oestradiol modulation of cognition in adult femalemarmosets (Callithrix jacchus). J Neuroendocrinol 2014;26:296e309.

[73] LaClair M, Lacreuse A. Reversal learning in gonadectomizedmarmosets with and without hormone replacement: are malesmore sensitive to punishment? Anim Cogn 2016;19:619e30.

[74] Rygula R, ClarkeHF, Cardinal RN, Cockcroft GJ, Xia J, Dalley JW,Robbins TW, Roberts AC. Role of central serotonin in anticipa-tion of rewarding and punishing outcomes: effects of selectiveamygdala or orbitofrontal 5-HT depletion. Cereb Cortex 2015;25:3064e76. https://doi.org/10.1093/cercor/bhu102.

[75] Kotani M, Enomoto T, Murai T, Nakako T, Iwamura Y, Kiyoshi A,Matsumoto K, Matsumoto A, Ikejiri M, Nakayama T, Ogi Y,Ikeda K. The atypical antipsychotic blonanserin reverses(þ)-PD-128907- and ketamine-induced deficit in executive func-tion in common marmosets. Behav Brain Res 2016;305:212e7.https://doi.org/10.1016/j.bbr.2016.02.031.

[76] da Silva Mota MT, Franci CR, de Sousa MBC. Hormonal changesrelated to paternal and alloparental care in common marmosets(Callithrix jacchus). Horm Behav 2006;49:293e302.

[77] Finkenwirth C, Martins E, Deschner T, Burkart JM. Oxytocin isassociated with infant-care behavior and motivation in coopera-tively breeding marmoset monkeys. Horm Behav 2016;80:10e8.

[78] French JA, Fite JA, Ross CN. Family life in marmosets: causesand consequences of variation in offspring care. In: Bridges RS,editor. The parental brain. New York: Elsevier Press; 2008.p. 461e78.

[79] Storey AE, Ziegler TE. Primate paternal care: interactions be-tween biology and social experience. Horm Behav 2016;77:260e71. https://doi.org/10.1016/j.yhbeh.2015.07.024.

[80] Pryce CR. The regulation of maternal behaviour in marmosetsand tamarins. Behav Processes 1993;30:201e24.

[81] Sanchez SM, Ziegler TE, Snowdon CT. Both parents respondequally to infant cues in the cooperatively breeding commonmarmoset, Callithrix jacchus. Anim Behav 2014;97:95e103.https://doi.org/10.1016/j.anbehav.2014.09.002.

[82] Taylor JH, French JA. Oxytocin and vasopressin enhance respon-siveness to infant stimuli in adult marmosets. Horm Behav 2015;75:154e9.

[83] Zahed SR, Prudom SL, Snowdon CT, Ziegler TE. Male parentingand response to infant stimuli in the common marmoset (Calli-thrix jacchus). Am J Primatol 2008;70:84e92. https://doi.org/10.1002/ajp.20460.

[84] Ziegler TE, Peterson LJ, SosaME, BarnardAM. Differential endo-crine responses to infant odors in common marmoset (Callithrixjacchus) fathers. Horm Behav 2011;59:265e70. https://doi.org/10.1016/j.yhbeh.2010.12.001.

[85] Dettling AC, Feldon J, Pryce CR. Repeated parental deprivationin the infant commonmarmoset (Callithrix jacchus, primates) and

analysis of its effects on early development. Biol Psychiatry 2002;52:1037e46.

[86] French JA, Carp SB. Early-life social adversity and develop-mental processes in nonhuman primates. Curr Opin Behav Sci2016;7:40e6. https://doi.org/10.1016/j.cobeha.2015.11.004.

[87] Pryce CR, Ruedi-Bettschen D, Dettling AC, Weston A, Russig H,Ferger B, Feldon J. Long-term effects of early-life environmentalmanipulations in rodents and primates: potential animal modelsin depression research. Neurosci Biobehav Rev 2005;29:649e74.https://doi.org/10.1016/j.neubiorev.2005.03.011.

[88] Birnie AK, Taylor JH, Cavanaugh J, French JA. Quality ofmaternal and paternal care predicts later stress reactivity in thecooperatively-breeding marmoset (Callithrix geoffroyi). Psycho-neuroendocrinology 2013;38:3003e14.

[89] Johnson EO, Kamilaris TC, Calogero AE, Gold PW, Chrousos GP.Effects of early parenting on growth and development in a smallprimate. Pediatr Res 1996;39:999e1005. https://doi.org/10.1203/00006450-199606000-00012.

[90] Johnson EO, Kamilaris TC, Carter CS, Calogero AE, Gold PW,Chrousos GP. The biobehavioral consequences of psychogenicstress in a small, social primate (Callithrix jacchus jacchus). BiolPsychiatry 1996;40:317e37.

[91] Dıaz-Munoz SL, Bales KL. “Monogamy” in primates: variability,trends, and synthesis. Am J Primatol 2016;78:283e7.

[92] Agmo A, Smith AS, Birnie AK, French JA. Behavioral character-istics of pair bonding in the black tufted-ear marmoset (Callithrixpenicillata). Behaviour 2012;149:407e40.

[93] Cavanaugh J, Huffman MC, Harnisch AM, French JA. Marmo-sets treated with oxytocin are more socially attractive to theirlong-term mate. Front Behav Neurosci 2015.

[94] Cavanaugh J, Mustoe AC, Taylor JH, French JA. Oxytocinfacilitates fidelity in well-established marmoset pairs byreducing sociosexual behavior toward opposite-sex strangers.Psychoneuroendocrinology 2014;49:1e10.

[95] Smith AS, Agmo A, Birnie AK, French JA. Manipulation of theoxytocin system alters social behavior and attraction in pair-bonding primates, Callithrix penicillata. Horm Behav 2010;57:255e62. https://doi.org/10.1016/j.yhbeh.2009.12.004.

[96] Finkenwirth C, van Schaik C, Ziegler TE, Burkart JM. Stronglybonded family members in common marmosets show synchro-nized fluctuations in oxytocin. Physiol Behav 2015;151:246e51.

[97] Galvao-Coelho NL, Silva HPA, De Sousa MBC. The influence ofsex and relatedness on stress response in common marmosets(Callithrix jacchus). Am J Primatol 2012;74:819e27. https://doi.org/10.1002/ajp.22032.

[98] Norcross JL, Newman JD. Effects of separation and novelty ondistress vocalizations and cortisol in the common marmoset(Callithrix jacchus). Am J Primatol 1999;47:209e22.

[99] Smith AS, Birnie AK, Lane KR, French JA. Production andperception of sex differences in vocalizations of Wied’s black-tufted-ear marmosets (Callithrix kuhlii). Am J Primatol 2009;71:324e32. https://doi.org/10.1002/ajp.20656.

[100] DiMattina C. Virtual vocalization stimuli for investigating neuralrepresentations of species-specific vocalizations. J Neurophysiol2005;95:1244e62. https://doi.org/10.1152/jn.00818.2005.

[101] Miller CT, Thomas AW. Individual recognition during bouts ofantiphonal calling in common marmosets. J Comp Physiol A2012;198:337e46.

[102] Miller CT, Wang X. Sensory-motor interactions modulate a pri-mate vocal behavior: antiphonal calling in commonmarmosets. J Comp Physiol A 2006;192:27e38.

[103] Toarmino CR, Wong L, Miller CT. Audience affects decision-making in a marmoset communication network. Biol Lett 2017;13:20160934. https://doi.org/10.1098/rsbl.2016.0934.

III. RESEARCH APPLICATIONS

26. THE MARMOSET AS A MODEL IN BEHAVIORAL NEUROSCIENCE AND PSYCHIATRIC RESEARCH490

[104] Chow CP, Mitchell JF, Miller CT. Vocal turn-taking in anon-human primate is learned during ontogeny. Proc R Soc BBiol Sci 2015;282:20150069e20150069. https://doi.org/10.1098/rspb.2015.0069.

[105] Miller C, DiMauro A, others. Vocalization induced cFos expres-sion in marmoset cortex. Front Integr Neurosci 2010;4:128.

[106] Simoes CS. Activation of frontal neocortical areas by vocal pro-duction in marmosets. Front Integr Neurosci 2010;4. https://doi.org/10.3389/fnint.2010.00123.

[107] Roy S, Zhao L,Wang X. Distinct neural activities in premotor cor-tex during natural vocal behaviors in a new World primate, thecommon marmoset (Callithrix jacchus). J Neurosci 2016;36:12168e79. https://doi.org/10.1523/JNEUROSCI.1646-16.2016.

[108] Eliades SJ, Miller CT. Marmoset vocal communication: behaviorand neurobiology: marmoset vocal communication. Dev Neuro-biol 2017;77:286e99. https://doi.org/10.1002/dneu.22464.

[109] Ghazanfar AA, Eliades SJ. The neurobiology of primate vocalcommunication. Curr Opin Neurobiol 2014;28:128e35.

[110] Smith T. Individual olfactory signatures in common marmosets(Callithrix jacchus). Am J Primatol 2006;68:585e604. https://doi.org/10.1002/ajp.20254.

[111] Smith TE, Tomlinson AJ, Mlotkiewicz JA, Abbott DH. Femalemarmoset monkeys (Callithrix jacchus) can be identified fromthe chemical composition of their scent marks. Chem Senses2001;26:449e58.

[112] Ferris CF, Snowdon CT, King JA, Duong TQ, Ziegler TE,Ugurbil K, Ludwig R, Schultz-Darken NJ, Wu Z, Olson DP,Sullivan Jr JM, Tannenbaum PL, Vaughan JT. Functional imagingof brain activity in conscious monkeys responding to sexuallyarousing cues. Neuroreport 2001;12:2231e6.

[113] Ferris CF, Snowdon CT, King JA, Sullivan JM, Ziegler TE,Olson DP, Schultz-Darken NJ, Tannenbaum PL, Ludwig R,Wu Z, Einspanier A, Vaughan JT, Duong TQ. Activation of neuralpathways associated with sexual arousal in non-humanprimates. J Magn Reson Imaging 2004;19:168e75. https://doi.org/10.1002/jmri.10456.

[114] Snowdon CT, Tannenbaum PL, Schultz-Darken NJ, Ziegler TE,Ferris CF. Conditioned sexual arousal in a nonhuman primate.Horm Behav 2011;59:696e701. https://doi.org/10.1016/j.yhbeh.2010.10.009.

[115] Burkart J, Kupferberg A, Glasauer S, van Schaik C. Even simpleforms of social learning rely on intention attribution in marmosetmonkeys (Callithrix jacchus). J Comp Psychol 2012;126:129.

[116] Werdenich D, Huber L. Social factors determine cooperation inmarmosets. Anim Behav 2002;64:771e81.

[117] Burkart JM, Fehr E, Efferson C, van Schaik CP. Other-regardingpreferences in a non-human primate: commonmarmosets provi-sion food altruistically. Proc Natl Acad Sci 2007;104:19762e6.

[118] MustoeAC, Cavanaugh J, HarnischAM, Thompson BE, French JA.Do marmosets care to share? Oxytocin treatment reduces prosocialbehavior toward strangers. Horm Behav 2015;71:83e90.

[119] Burkart JM, Allon O, Amici F, Fichtel C, Finkenwirth C,Heschl A, Huber J, Isler K, Kosonen ZK, Martins E, others. Theevolutionary origin of human hyper-cooperation. Nat Commun2014;5:4747.

[120] Finkenwirth C, Burkart JM. Long-term-stability of relationshipstructure in family groups of common marmosets, and its linkto proactive prosociality. Physiol Behav 2017;173:79e86.https://doi.org/10.1016/j.physbeh.2017.01.032.

[121] Heyes CM. Reflections on self-recognition in primates. AnimBehav 1994;47:909e19. https://doi.org/10.1006/anbe.1994.1123.

[122] Anderson JR, Gallup GG. Mirror self-recognition: a review andcritique of attempts to promote and engineer self-recognitionin primates. Primates 2015;56:317e26. https://doi.org/10.1007/s10329-015-0488-9.

[123] Eglash AR, Snowdon CT. Mirror-image responses in pygmymarmosets (Cebuella pygmaea). Am J Primatol 1983;5:211e9.https://doi.org/10.1002/ajp.1350050305.

[124] Inoue-Nakamura N. Mirror self-recognition in nonhuman pri-mates: a phylogenetic approach. Jpn Psychol Res 1997;39:266e75. https://doi.org/10.1111/1468-5884.00059.

[125] Erkert HG. Characteristics of the circadian activity rhythm incommon marmosets (Callithrix j. jacchus). Am J Primatol 1989;17:271e86. https://doi.org/10.1002/ajp.1350170403.

[126] Glass JD, Tardif SD, Clements R, Mrosovsky N. Photic and non-photic circadian phase resetting in a diurnal primate, the com-mon marmoset. Am J Physiol Regul Integr Comp Physiol 2001;280:R191e7.

[127] Silva CA da, de Pontes ALB, Cavalcante J de S, deAzevedo CVM. Conspecific vocalisations modulate the circadianactivity rhythm of marmosets. Biol Rhythm Res 2014;45:941e54.https://doi.org/10.1080/09291016.2014.939441.

[128] Melo PR, Goncalves BSB, Menezes AAL, Azevedo CVM. Circa-dian activity rhythm in pre-pubertal and pubertal marmosets(Callithrix jacchus) living in family groups. Physiol Behav 2016;155:242e9. https://doi.org/10.1016/j.physbeh.2015.12.023.

[129] Melo P, Goncalves B, Menezes A, Azevedo C. Socially adjustedsynchrony in the activity profiles of common marmosets inlight-dark conditions. Chronobiol Int 2013;30:818e27. https://doi.org/10.3109/07420528.2013.767823.

[130] Pines A. Circadian rhythm and menopause. Climacteric J IntMenopause Soc 2016;19:551e2. https://doi.org/10.1080/13697137.2016.1226608.

[131] Gervais NJ, Viechweg SS, Mong JA, Lacreuse A. The middle-aged ovariectomized marmoset (Callithrix jacchus) as a modelof menopausal symptoms: preliminary evidence. Neuroscience2016;337:1e8.

[132] Bertani S, Carboni L, Criado A, Michielin F, Mangiarini L,Vicentini E. Circadian profile of peripheral hormone levels inSprague-Dawley rats and in common marmosets (Callithrixjacchus). Vivo Athens Greece 2010;24:827e36.

[133] Cross N, Rogers LJ. Diurnal cycle in salivary cortisol levels incommon marmosets. Dev Psychobiol 2004;45:134e9.

[134] Sousa MBC, Ziegler TE. Diurnal variation on the excretion pat-terns of fecal steroids in common marmoset (Callithrix jacchus)females. Am J Primatol 1998;46:105e17.

[135] Valentinuzzi VS, Neto SPD, Carneiro BTS, Santana KS, Araujo JF,Ralph MR. Memory for time of training modulates performanceon a place conditioning task in marmosets. Neurobiol LearnMem 2008;89:604e7. https://doi.org/10.1016/j.nlm.2007.08.002.

[136] Burkart JM, Finkenwirth C. Marmosets as model species inneuroscience and evolutionary anthropology. Neurosci Res2015;93:8e19.

[137] Miller CT, Freiwald WA, Leopold DA, Mitchell JF, Silva AC,Wang X. Marmosets: a neuroscientific model of human socialbehavior. Neuron 2016;90:219e33.

[138] Saito A. The marmoset as a model for the study of primateparental behavior. Neurosci Res 2015.

[139] Krakauer JW, Ghazanfar AA, Gomez-Marin A, MacIver MA,Poeppel D. Neuroscience needs behavior: correcting a reduc-tionist bias. Neuron 2017;93:480e90. https://doi.org/10.1016/j.neuron.2016.12.041.

III. RESEARCH APPLICATIONS

REFERENCES 491