How strain differences could help decipher the neurobiology of mammalian playfulness: What the less playful Fischer 344 rat can tell us about play

Abstract

Play is common among the young of many mammalian species. How that play is exhibited results from a dynamic interplay between genetic framework and experiential influences that, in turn, operate on hard-wired brain systems. One approach towards understanding how genes and environment interact with brain substrates to yield a particular playful phenotype is to take advantage of inbred strains of rats that come with a known genetic identity and assess the effects of varying early social experiences and targeted neurobiological interventions on rats of these strains. This paper primarily summarizes research utilizing the F344 inbred strain, a rat that consistently plays less than most other strains.

Introduction

When first beginning to work on my PhD under the mentorship of Jaak Panksepp the idea of studying play behavior in rats had never actually entered my mind. At around the time I began work on my PhD, Jaak’s lab was beginning to pivot from studying the brain mechanisms of energy balance to a more encompassing view of how the brain modulates the variety of emotions experienced by most mammals. His lab had recently published a few papers looking at the role of opioids in social emotions (Panksepp, Herman, Vilberg, Bishop, & DeEskinazi, 1980; Panksepp, Meeker, & Bean, 1980; Panksepp, Najam, & Soares, 1979) and it was this work that caught my attention in the first place. With the recent discovery of opioid receptors in the brain (Pert & Snyder, 1973) and the corresponding endogenous ligands for these receptors (Hughes et al., 1975; Li & Chung, 1976) there was considerable excitement as labs began searching for putative opioid involvement in a variety of behavioral processes. The idea of studying brain mechanisms that could be modulating social emotions was quite unusual at that time but I found it intriguing and assumed that this was what I would be working on over the next 4 years or so. What I was unaware of at the time was that Jaak had also published a couple of smaller papers looking at play in rats (Panksepp, 1981; Panksepp & Beatty, 1980) and was just about to embark on a full-scale assault on the neurobiology of play. It didn’t take much convincing to get me on board with this research program and I’ve been somewhat consumed by it ever since.

During those early years of studying play in rats I never gave much thought to the strain of rat that was being used. My guess is that most labs would use whatever strain had always been used in that lab. While working in Jaak’s lab we used Long-Evans derived rats that were bred and born locally. When setting up my own lab a few years later I chose to use Sprague-Dawley rats, partly because they were relatively inexpensive, I had a very limited budget, and the questions I was asking at the time seemed to simply require a rat that readily plays and the Sprague-Dawley fit that bill nicely. It didn’t seem that the choice of strain would have much of an impact on the work that I was planning to do. This naïve thought would soon change.

A historical perspective on studying play in the rat

Rats are an incredibly social and playful species. Playful interactions begin prior to weaning and continue throughout the juvenile period, peaking around 35 days of age, and then steadily decreasing as the animals reach puberty (Panksepp, 1981). Play in the rat primarily takes the form of “rough-and-tumble” activity; rats will vigorously chase each other, pounce on each other’s dorsal surface, nuzzle and nip at the nape, and pin each other (Panksepp, Siviy, & Normansell, 1984; Pellis & Pellis, 2009; Siviy & Panksepp, 2011; Vanderschuren, Niesink, & Van Ree, 1997; Vanderschuren & Trezza, 2014). There are a number of easily identifiable and stable behavioral measures that have been used to reliably quantify playfulness. We, and others, initially focused on two major postures that are commonly seen during play but rare in other non-playful social encounters – nape contacts and pins. A nape contact is defined as occurring when one rat vigorously pounces at the nape of the other rat with either its snout or front paws. A pin occurs when one rat rolls onto its back with the other rat on top. Pins often, but not always, occur in response to a nape contact and are thought to prolong a play bout and/or signal playful intent (Panksepp et al., 1984; Pellis, Field, Smith, & Pellis, 1997; Vanderschuren & Trezza, 2014).

Rats also exhibit high frequency (50–55 kHz) ultrasonic vocalizations (USVs) when playing and these vocalizations have been suggested to be markers of a positive affective state (Burgdorf et al., 2008; Burgdorf, Panksepp, & Moskal, 2011; Knutson, Burgdorf, & Panksepp, 2002). These USVs may serve a communicative function, as the playback of 50 kHz vocalizations will readily elicit approach (Seffer, Schwarting, & Wöhr, 2014; Willadsen, Seffer, Schwarting, & Wöhr, 2013). USVs are also more likely to occur shortly before playful contact is made and this would be consistent with rats using USVs to communicate playful intention (Himmler, Kisko, Euston, Kolb, & Pellis, 2014; see also the paper by Burke, Euston & Pellis, in this Special Issue).

While nape contacts and pins are an efficient and highly reliable set of measures for quantifying overall levels of playfulness, how rats respond to nape contacts have been further characterized and can provide additional insight into the responsiveness of rats to playful overtures (Pellis, 1988; Pellis & Pellis, 1987). According to this protocol, rats exhibit several tactics by which they “defend” their nape when “attacked”, with most nape attacks leading to some type of defensive response. Rats can run away or evade an attack, or execute a facing defense where the rat will rotate along their longitudinal axis to remove access to the nape and face their play partner. During the juvenile period the most common type of facing defense is a complete rotation and this is the measure that my lab tends to use most often. With this range of possible body positions it is quite straightforward to quantify both play solicitation and responsiveness to that solicitation in two playful partners.

Although one has to wonder how much playfulness has been a beneficiary of domestication and selective breeding (e.g., Lockard, 1968), a recent series of studies compared the play of standard laboratory strains to that of a wild strain (WWCPS strain collected from Warsaw, Poland). The results of these studies suggest that the playfulness of the standard laboratory rat is not a total aberration (Himmler et al., 2013; Himmler et al., 2014). While domesticated strains tend to be more playful than the wild strain, all of the components seen during play are comparable between domesticated and wild strains. All strains, domesticated and wild, target the nape and defend the nape in a similar manner. However, what seems to differ is the relative frequency of these measures and the tactics used to defend the nape. There appears to be as much variability between the various domesticated strains as there is between domesticated strains and a wild strain. Considering that both domesticated strains and wild rats incorporate comparable elements (e.g., attack and defense of the nape) into their play this suggests that the play of common laboratory rats can provide a good estimation of rat playfulness and, presumably, presents a valid animal model for understanding the neurobiological substrates of playfulness.

The neurobiology of play and the road to the Fischer 344 rat

One of the first set of pharmacological studies using play as the behavioral measure showed that acute treatment with psychomotor stimulants such as amphetamine and methylphenidate resulted in robust reductions in play (Beatty, Costello, & Berry, 1984; Beatty, Dodge, Dodge, White, & Panksepp, 1982). These data, along with some other work done around this time looking at noradrenergic and serotonergic involvement (Normansell & Panksepp, 1985a, 1985b) suggested that monoamines are likely to have a prominent role in the modulation of play so this is where I initially set my sights when setting up my own lab. Our first set of studies demonstrated that blocking alpha-2 noradrenergic receptors reliably increased play (Siviy, Atrens, & Menendez, 1990; Siviy, Fleischhauer, Kuhlman, & Atrens, 1994). This complemented the earlier work of Normansell & Panksepp (1985a) where they showed that the alpha-2 agonist clonidine was potent at reducing play and that this effect could be partially blocked by the alpha-2 antagonist yohimbine. Having some level of pharmacological specificity, especially with an agonist reducing play and an antagonist increasing play, led us to believe that norepinephrine acting at alpha-2 receptors modulates play. There were still a number of issues that would not be worked out until much later by Louk Vanderschuren’s group (Vanderschuren et al., 2008) but we felt confident enough at this point to move on to the next logical monoamine – dopamine.

There were, and still are, many reasons to suppose that dopamine should have a significant impact on play, especially the rough-and-tumble activity typically seen in rats. It was not unexpected when dopamine antagonists such as haloperidol were found to consistently reduce play in a dose-dependent manner. In addition, there were a few reports showing modest increases in play with low doses of the dopamine agonist apomorphine, although this effect was not very robust nor easily replicable (Beatty et al., 1984; Holloway & Thor, 1985; Niesink & Van Ree, 1989). This may be the result of multiple receptor subtypes that recognize dopamine and drugs such as haloperidol and apomorphine are not especially sensitive to any particular receptor subtype. As selective agonists and antagonists became more readily available we began to look at whether the use of these drugs would shed additional light on the role of dopamine in play. Much like the earlier work reported for apomorphine, low doses (e.g., 10 μg/kg) of the more selective dopamine D2 receptor agonist quinpirole modestly increased pinning but the most prominent effect of quinpirole was a dose-dependent reduction in play (Siviy, Fleischhauer, Kerrigan, & Kuhlman, 1996). In hindsight, and in light of more recent work showing that dopamine is more likely to be involved in motivational factors associated with play rather than the production of the behavior itself (Achterberg et al., 2016), this should not have been that unexpected. Nevertheless, this level of pharmacological inconsistency was disconcerting at the time and I began to wonder if there might be a better strain of rat for further investigating a role for dopamine in play.

The use of Fischer 344 rats in play research

From my initial reading of the literature, rats of the Fischer 344 (F344) strain appeared to be particularly sensitive to the behavioral effects of dopamine. When compared to rats of the Buffalo strain, F344 rats are more sensitive to the stereotypy-inducing effects of apomorphine (Helmeste, Seeman, & Coscina, 1981), are more sensitive to the apomorphine-induced inhibition of locomotor activity (Helmeste, 1983), and are reported to have a higher density of D2 receptors in the striatum (Kerr, Unis, & Wamsley, 1988). Therefore, it seemed that this strain might be better suited for assessing a role for dopamine in the modulation of play. To the best of my knowledge there were no reports at that time describing the social behavior of this strain but I had no a priori reason to think that F344 rats would play any differently than the Sprague-Dawley rat that we had been using up to that point. Nevertheless, it seemed prudent to conduct an initial study assessing play after various periods of social isolation in F344 rats. Since the papers that originally pointed me to the F344 strain used the Buffalo strain as the comparison strain (Helmeste, 1983; Helmeste et al., 1981) and since it seemed desirable to compare two inbred strains, this was the strain comparison we initially opted for. What we found was that F344 rats were less playful than Buffalo rats (Siviy, Baliko, & Bowers, 1997). In particular, isolation-induced increases in nape contacts and pins were less pronounced in F344 rats when tested in same-strain pairings. When tested in cross-strain pairings (i.e., F344 paired with Buffalo), F344 rats had fewer nape contacts, were less likely to defend a nape attack, and were less likely to respond to nape contacts with either a complete rotation or evasion.

Putting any further pharmacological assessment temporarily on hold, and to get a better handle on play in this strain, we proceeded to further characterize play in F344 rats (Siviy, Love, DeCicco, Giordano, & Seifert, 2003). This also led to a deeper reading of the literature regarding the F344 strain (Camp & Robinson, 1994; Kosten, Miserendino, Chi, & Nestler, 1994; Self & Nestler, 1995; Stöhr et al., 2000) and switching to the inbred Lewis strain for our comparison strain. When tested after 24 hours of isolation in same-strain pairings, our earlier work was replicated, in that F344 rats had fewer nape contacts and were less likely to respond to a nape contact with a complete rotation than were Lewis rats. In order to determine whether any of these effects were an artifact of same-strain pairings, F344 and Lewis rats were then paired with Sprague-Dawley rats so that each strain would be tested with a comparably playful rat. Again, F344 rats continued to be less playful as indicated by delivering fewer nape contacts and being less likely to rotate completely when contacted. In order to tease apart play solicitation from responsiveness, animals were tested with a Sprague-Dawley partner that was made unresponsive to playful overtures with scopolamine (Thor & Holloway, 1985). By allowing rats to interact with an unresponsive partner any confound that might be associated with active responses made by that partner can be minimized. When tested in this protocol, F344 rats delivered fewer nape contacts but spent more time in social investigation (e.g., anogenital sniffing) than Lewis rats. This suggests that while F344 rats are consistently less playful, they may be more socially curious.

Together, these data can be used to make a reasonably strong case for genetic differences accounting for the phenotypic differences in play between these strains. The differences seen in play are not unique to this behavior. F344 mothers tend to spend less time engaged in active maternal behaviors, such as crouching, licking and grooming, nursing, and pup retrieval (Gomez-Serrano, Sternberg, & Riley, 2002; Moore, Wong, Daum, & Leclair, 1997) and early maternal care has been shown to impact later playfulness (Parent & Meaney, 2008; van Hasselt et al., 2012). So it’s also possible that strain differences in play may reflect specific strain differences in other behaviors such as maternal behavior towards the pups during the first few weeks of life. One way to parse the relative involvement of genetics and early maternal behavior towards pups is through a cross-fostering approach where F344 rats would be raised by Lewis mothers and vice versa. Earlier work had shown that genetic differences could not explain all of the phenotypic differences between F344 and Lewis rats (Gomez-Serrano, Tonelli, Listwak, Sternberg, & Riley, 2001) so we performed a small-scale cross-fostering study as well. This allowed us to look at environmental contributions to behavioral differences above genetic ones. We found that cross-fostering had no impact on the levels of play in F344 rats, suggesting that the differences in play between these two strains could not be easily accounted for by differences in maternal care (Siviy et al., 2003). Since that study was somewhat under-powered, we conducted another more extensive cross-fostering study and obtained similar results (Siviy, Eck, McDowell, & Soroka, 2017), suggesting that the overall level of playfulness is somewhat resistant to changes in the early postnatal environment.

This is not to say that interventions during the first few weeks of life have no effect on subsequent play behavior. For example, pups that have experienced brief (e.g., 1–15 minutes) daily periods of separation from the mother, a procedure commonly referred to as “neonatal handling”, tend to play more than those from undisturbed litters (Aguilar, Carames, & Espinet, 2009; Siviy & Harrison, 2008). In order to determine whether F344 rats might be sensitive to the effects of brief periods of maternal separation, litters of F344 and Lewis rats were either separated from the mother for 15 minutes a day over the first 2 postnatal weeks or were undisturbed during this time. After weaning, play was assessed by pairing inbred rats with untreated Sprague-Dawley partners and overall strain differences in play remained relatively unaffected. When compared to Lewis rats, F344 rats still directed fewer nape contacts to their partner and while handling led to a modest, but significant, increase in the likelihood of rotating completely to supine in response to contacts directed to the nape in both strains, F344 rats were still less likely to rotate completely to supine (Siviy, 2018).

In our hands, the effects of neonatal handling on play are subtle and this may be due to the minimally intrusive nature of the manipulation. Therefore, we conducted a pilot study where we assessed the effects of adding tactile stimulation to the daily periods of separation in order to amplify the amount of stimulation received by handled rats (Garliss, Gentes, & Siviy, unpublished observations). For this study, pups were separated from the mother as detailed in our earlier studies (Siviy, 2018; Siviy & Harrison, 2008) and each of the pups within the handled litters were individually stroked with a paintbrush for approximately 1 minute over the course of a 10 minute separation period. This continued from postnatal days 1 through 14. Control rats were undisturbed during the first 2 postnatal weeks. Rats were weaned at 21 days of age and tested for play with an untreated Sprague-Dawley partner of the same sex and same age after 4 and 24 hours of isolation. As can be seen in Figure 1A, Lewis rats receiving tactile stimulation during the separation period directed more contacts to the Sprague-Dawley partner than the untreated control Lewis rats. As expected, F344 rats directed fewer nape contacts overall than did Lewis rats and no effect of tactile stimulation was observed. Tactile stimulation had no effect on the likelihood of responding to a nape contact with a complete rotation in either strain and, as expected, F344 rats were less likely to respond to a nape contact with a complete rotation (Figure 1B). While these data suggest that early postnatal experiences can have a modest impact on overall playfulness of rats, the relative lack of play among F344 rats cannot be readily overcome by early postnatal manipulations such as cross-fostering, handling, or early postnatal tactile stimulation. It also suggests that inbred strains may be differentially sensitive to the effects of early postnatal stimulation.

Figure 1.

Effects of brief maternal separation (handling) with added tactile stimulation of pups on nape contacts (panel A) and the probability of responding to a nape contact with a complete rotation (panel B). Pups from the control group were undisturbed during the first two postnatal weeks. As there was no main effect of sex of the rat nor did sex contribute to any of the interactions, the data depicted are collapsed across sex. For nape contacts, there was a significant main effect of strain, F(1,22) = 30.96, p < .001, with F344 rats directing fewer nape contacts than Lewis rats. There was also a strain x group interaction, F(1,22) = 6.85, p = .016, with handling + tactile stimulation increasing nape contacts in LEW rats but not F344 rats. For complete rotations, there was only a significant effect of strain, F(1,22) = 16.96, p < .001, with F344 rats less likely to respond to a nape contact with a complete rotation. * p < .05 comparing tactile stimulation to control. # p < .05 comparing F344 to Lewis.

In our standard protocol for assessing strain differences in play, rats are housed in same-strain and same-sex groups upon weaning or immediately upon arrival to the laboratory. This means that rats are being housed with other rats of comparable playfulness and that different strains are likely experiencing different levels of play throughout the developmental period when play is being assessed. Few studies have assessed the behavioral consequences of systematically varying the post-weaning social environment in rats. Schneider and colleagues (Schneider, Bindila, et al., 2016; Schneider et al., 2014; Schneider, Pätz, Spanagel, & Schneider, 2016; see also the paper by Stark & Pellis in this Special Issue) conducted a series of studies where the consequences of social rejection during adolescence were modeled by housing a single Wistar rat with multiple same-aged F344 rats and comparing their behavior at a later age to that of Wistar rats housed with other Wistar rats during the post-weaning period. In these experiments, the Wistar rat was the target of interest whereas the F344 rat was used to limit social interactions during adolescence and, hence, modeling social rejection. Flipping this approach around, we (Siviy, Campbell, & Gentes, unpublished observations) asked what would happen if a single non-playful F344 rat is placed in a chronic housing situation along with two playful Sprague-Dawley rats? Would living with more playful rats make the F344 rat a bit more playful?

To test this hypothesis, F344 and Lewis rats were housed with either 2 rats of the same strain (our standard housing protocol) or 2 Sprague-Dawley rats from weaning, at 22 days of age, until tested for play behavior between 36 and 40 days of age. After 2 weeks of acclimating to the new housing conditions rats were tested in a neutral testing chamber with an unfamiliar Sprague-Dawley rat. F344 rats from both housing conditions continued to direct fewer playful nape contacts to their play partner than Lewis rats from either housing condition (Figure 2A). In other words, F344 rats housed for two weeks with rats of a more playful strain continued to solicit less play than Lewis rats. While strain differences in nape contacts were unaffected by differential housing after weaning, responsiveness to playful solicitations among F344 rats was quite sensitive to post-weaning housing (Figure 2B). In particular, F344 rats housed for two weeks with two Sprague-Dawley rats were more likely to rotate completely to a supine position than F344 rats housed with other F344 rats. Furthermore, the responsiveness of these rats did not differ significantly from Lewis rats either housed with other Lewis rats or Sprague-Dawley rats.

Figure 2.

Effects of cross-housing on nape contacts (top 2 panels) and probability of responding to a nape contact with a complete rotation (bottom 2 panels). Rats were either housed with 2 rats of the same strain (LEW/LEW; F344/F344) or 2 Sprague-Dawley (SD) rats (LEW/SD; F344/SD) from weaning until testing. Rats were tested for 10 minutes with an unfamiliar Sprague-Dawley rat after 4 and 24 hours of isolation housing. For nape contacts (top 2 panels), there was a significant effect of strain, F(1,46) = 24.07, p < .001, with F344 rats directing fewer nape contacts to the Sprague-Dawley partner. For probability of responding to a nape contact with a complete rotation (bottom 2 panels), there was a significant effect of strain, F(1,46) = 10.26, p = .002, with F344 rats less likely to respond with a complete rotation than F344 rats. There was also a significant strain X housing interaction, F(1,46) = 5.61, p = .022 with F344 rats housed with 2 SD rats more likely to rotate completely to a supine position than F344 rats housed with 2 F344 rats. * p < .05 comparing housing condition

Several tentative conclusions can be drawn from these studies. Play solicitation in F344 rats seems to be relatively impervious to postnatal variations in the immediate social environment, whether that be from the mother prior to weaning or from peers after weaning. On the one hand, how an animal responds to playful overtures may be more sensitive to early pre-weaning social experiences. We saw a hint of this in both strains with cross-fostering and handling (Siviy, 2018; Siviy et al., 2017) but these were subtle effects so slight variations in licking and grooming may only yield modest changes and, therefore, strain differences in both indices of play were quite resistant to manipulations prior to weaning. On the other hand, housing F344 rats with two playful Sprague-Dawley rats for an extended period of time after weaning provided more opportunities for greater variations in social experiences and this led to a pattern of playful responsiveness similar to that of the more playful Lewis rat. Not only do the above studies tell us something about the behavioral differences of the F344 strain, it has also resulted in a baseline to measure the relative playfulness of the F344 rat compared to other strains. By learning more about the behavioral differences associated with the F344 rat we can begin to link these differences to known differences in their neurobiology.

Using strain differences to study the underlying neurobiological substrates of play

With well-defined, robust, and consistent behavioral differences between strains as a starting point, several methodological approaches can be utilized to further our understanding of the neurobiological substrates of play. For example, strain differences in responsiveness to targeted pharmacological manipulations would suggest differential sensitivities in the modulation of play by these systems. This strategy has been successfully employed in comparing the relative sensitivity of Wistar and Sprague-Dawley rats to opioids and cannabinoids (Manduca, Campolongo, et al., 2014; Manduca, Servadio, et al., 2014). In particular, rats of the Wistar strain are particularly sensitive to the play-stimulating effects of opioid agonists and compounds that prolong the synaptic availability of endogenous cannabinoids.

Another approach would be to look for structural and/or functional differences between strains in potentially relevant neural systems. Once behavioral differences can be coupled to neurobiological differences, specific hypotheses can be generated and tested regarding the role of those systems in play. For example, rats of the F344 strain seem to be particularly vulnerable to mitochondrial distress and this is reflected in altered corticostriatal DA functioning in an in vitro striatal slice preparation (Akopian et al., 2008). Collaborating with John Walsh, a former post-doctoral colleague, we asked whether there were parallels between the corticostriatal dopamine (DA) dysfunction his lab was seeing and the dysfunctional play of the F344 rat that we were seeing. Using fast-scan cyclic voltammetry of brain slices in the dorsal striatum and nucleus accumbens, we found that F344 rats release less DA in response to electrode stimulation (Siviy, Crawford, Akopian, & Walsh, 2011). Using high performance liquid chromatography F344 rats had higher DA content in the striatum and prefrontal cortex while showing less DA turnover relative to Sprague-Dawley rats at both sites. Together, these data suggested that F344 rats may have a deficit in the packaging of DA into vesicles, leading to an accumulation of DA in the cytoplasm and a corresponding dysfunction in vesicular release. If this is the case, and if F344 rats are sequestering DA in the cytoplasm with less available to be released, we thought it possible that some of the behavioral differences observed in F344 rats may be attributable to problems associated with synaptic transmission of DA.

In a subsequent study (Siviy et al., 2015) we tested this hypothesis by further exploring the neurochemical differences in DA functioning in brain slices obtained from F344 and Sprague-Dawley rats and using these data to inform behavioral hypotheses. If, as suggested by our earlier work, there is an accumulation of cytoplasmic non-vesicular DA then this should be detectible by treating brain slices with amphetamine, due to enhanced cytoplasmic DA release through reverse transport at dopamine transporters. Consistent with this hypothesis, we found that amphetamine led to a more rapid and greater release of DA in striatal slices from F344 rats. To further test this hypothesis in a behavioral model, we assessed the effectiveness of amphetamine to increase locomotor activity in an open field. As predicted, amphetamine was found to be more effective at increasing activity in F344 rats when strain differences in baseline activity were taken into account. This suggests that F344 rats are more susceptible to the stimulant properties of amphetamine and is consistent with the hypothesis that rats of this strain are more susceptible to non-vesicular release of DA through amphetamine-mediated reverse transport.

Finding support for our hypothesis that F344 rats have deficits in the packaging and delivery of vesicular DA, we then sought to determine if the dysfunctional play of F344 rats may be related to reductions in vesicular release of DA. If F344 rats are less playful because of impaired vesicular release of DA within the striatum it seemed reasonable to hypothesize that increased synaptic availability of DA through amphetamine-induced non-vesicular release of DA may lead to an increase in play in rats of this strain. This hypothesis was not supported; amphetamine reduced play in a dose-dependent manner to a comparable extent in both F344 and Sprague-Dawley rats (Siviy et al., 2015). So while there is a clear difference in how these two strains handle DA in the pre-synaptic terminal and that these differences can have behavioral consequences, it is still unclear as to whether impaired DA functioning within the striatum can readily account for the relative lack of play in the F344 rat.

While dysfunction in DA functioning may not easily account for the low levels of play in the F344 rat, recent studies have bolstered the notion that DA acting within the ventral striatum provides an important modulatory influence over playfulness. Infusions of either amphetamine or the DA agonist apomorphine into the nucleus accumbens increases play in Wistar rats, with the increases seen with amphetamine blocked by co-infusion of either D1 or D2 antagonists (Manduca et al., 2016). There are also several lines of converging evidence suggesting that DA may be particularly important for enhancing motivation to play (Achterberg et al., 2016) and this may be due to increased DAergic activity within the nucleus accumbens. As mentioned earlier, rats will emit high frequency 50 kHz ultrasonic vocalizations when playing (Burgdorf et al., 2008; Manduca, Campolongo, et al., 2014) and playback of 50 kHz vocalizations leads to approach in juvenile rats (Wöhr & Schwarting, 2007) as well as increased release of DA in the nucleus accumbens (Willuhn et al., 2014). Similarly, social interactions can increase DA release in the NAc of both adolescent and adult rats (Robinson, Zitzman, Smith, & Spear, 2011). A recent study has also found a functional relationship between social dominance and DA-mediated mitochondrial respiration in the nucleus accumbens (van der Kooij, et al., 2018) which is particularly interesting in light of F344 rats being more sensitive to mitochondrial toxins (Akopian et al., 2008). Incorporating inbred strains, such as the F344 rat, into these types of experimental paradigms would be very informative.

Some concluding thoughts

While this paper has focused on the F344 rat it is important to note that play has also been assessed in other strains and in rats selected for other traits. For example, the Spontaneously Hypertensive Rat, a strain used to model symptoms of ADHD (Sagvolden et al., 1992; Sagvolden, Russell, Aase, Johansen, & Farshbaf, 2005), is less playful when paired with either a normotensive Wistar-Kyoto rat or a Sprague-Dawley rat (Ferguson & Cada, 2004). Rats that have been bred for high rates of tickling-induced ultrasonic vocalizations (USVs) solicit more play than those bred for low rates of tickling-induced USVs (Webber et al., 2012) while rats bred for either high or low rates of USVs based on rates of separation-induced USVs play less than random control lines (Brunelli et al., 2006). Lastly, rats that are more likely to display seizures in response to priming stimulation of the amygdala (amygdala kindling) tend to be more playful than rats selected for resistance to amygdala kindling (Reinhart, McIntyre, Metz, & Pellis, 2006; Reinhart, Pellis, & McIntyre, 2004). These models could also provide an informative framework to further pursue the neurobiological substrates of play.

Play is a robust phenotype that is present in the behavioral repertoire of many mammalian species, with the overall playfulness of a single individual reflecting a complex and dynamic interplay between genetic and postnatal environmental influences. Working with common outbred strains of rat has yielded a trove of data that has been indispensable for providing an overall picture of how the brain may be modulating play and beginning to delineate some of the putative functions of play (Pellis & Pellis, 2007, 2009; Vanderschuren & Trezza, 2014). Use of selected inbred strains and/or strains that have other unique characteristics have provided an added dimension to the empirical toolbox, allowing for targeted hypotheses to complement those obtained with other strains and gain further clarity on gene and environment interactions.

Play is important for the overall development of those animals that readily engage in play, an obvious theme in this Special Issue. The young mammalian brain is likely hard-wired to engage in playful behaviors, with adverse consequences resulting when opportunities for play are thwarted (Baarendse, Counotte, O’Donnell, & Vanderschuren, 2013; Van den Berg et al., 1999; Von Frijtag, Schot, van den Bos, & Spruijt, 2002; see also the paper by Stark and Pellis in this Special Issue). The importance of play in normal childhood development has been highlighted on several occasions by the American Academy of Pediatrics, most recently in 2018 when it was recommended that all pediatricians “prescribe” hefty doses of play when children come in for scheduled exams (Ginsburg, 2007; Yogman, Garner, Hutchinson, Hirsh-Pasek, & Golinkoff, 2018). Jaak would often say that play is a source of pure joy in the brain and by studying play we can get to an actual science of joy. Gaining a solid foot-hold on how the brain modulates play in young rats, a task that will be aided by targeted use of selected strains of rats, will help us to better understand how we can have more joyful brains and reap the full benefits that play may bestow on our own children as they navigate their way to adulthood.