The Emotional Foundations of Personality

by Kenneth L Davis

  The Meaney group extended their groundbreaking research by further demonstrating that these maternally induced effects could also be reversed with fifty days of either postweaning social and environmental enrichment for offspring of low LG-ABN mothers or corresponding social isolation for offspring of high LG-ABN mothers (Champagne & Meaney, 2007). Meaney’s group has hypothesized that these epigenetic alterations may provide a means for regulating genomic expression in response to current environmental conditions reflecting different states of environmental adversity (for further discussion, see Meaney, 2010). Thus, this research group demonstrated major differences in the maternal “personalities” of rat mothers, which had real effects on how their offspring reacted to stressors in their environments. The possible implications of what this might mean for human mothering and the corresponding impact on human personalities remain to be more fully demonstrated (more on this in Chapter 15).

  SEPARATION DISTRESS ENGENDERED BY THE PANIC/SADNESS EMOTIONAL SYSTEM

  It is thought that rats vocalize in ultrasonic ranges that humans and most other mammals cannot hear in order to avoid detection by predators. Rat neonates emit a type of ultrasonic vocalization centered around 40 kHz that occurs in response to social isolation or cold stress and correspondingly elicits pup retrieval in mother rats (Panksepp & Burgdorf, 2000). There remains some ambiguity whether this call is evolutionarily related to the separation distress that has been more extensively studied in dogs, guinea pigs, and infant chickens, because it tends to disappear if testing is conducted in warm environments where fetal rat pups do not get cold (for critique, see Panksepp, 2003), a temperature effect that is also seen in young dogs before they are mature enough to thermoregulate.

  Clearly investigators need to make better distinctions between the aversion of physical distress, such as getting cold, and the emotional distress from being separated from their mothers. True separation distress vocalizations have been associated in several species with the PANIC/Sadness system, which has been linked to depression in humans in part by the finding that opioids reduce both separation distress and human depression (Panksepp, 1998a, 2015, 2016). Infant rat data remain ambiguous on this issue. In any case, to demonstrate inherited influences on the distress vocalizations of very young infant rats, Susan Brunelli and colleagues at Columbia University selected lines of rats based on their high or low rates of such ultrasonic vocalizations in response to social isolation at ten days of age and selectively bred these lines for several generations (Brunelli, 2005). Of course, laboratory rats, because of traditional laboratory practices, are inadvertently bred to endure living in isolation (one rat per cage after weaning, typically done at three weeks of age, has long been standard practice). So Brunelli’s group started with a highly heterogeneous strain of rats from the National Institutes of Health, hoping to maximize their chances of selecting rats experiencing high levels of separation distress. In the 20th generation, there were about four times as many isolation-induced 45-kHz vocalizations in the high-vocalization line of animals compared to a randomly selected control line, which likewise exhibited substantially more 45-kHz calls than the low vocalization line, which exhibited 45-kHz calls that averaged close to zero (Brunelli & Hofer, 2007).

  In the Porsolt Swim Test, an animal model for depression, the high vocalization line showed more depression-related immobile floating than did the low vocalization line, suggesting the high vocalization line represented a depressive phenotype and heightened susceptibility to stressors. However, adult rats from the high-vocalization line were also less active in the center of an open field test, suggesting higher anxiety and increased activity of the FEAR system as well. In addition, males from the low vocalization line when paired together after social isolation exhibited fighting in 70 percent of the cases, compared to 35 percent for random line males. (The high vocalization line was not included in this study.)

  Brunelli and Hofer (2007) concluded that the high-vocalization line was characterized by anxiety/depression and the low-vocalization line by aggressive/impulsive behavior. Is it possible that the RAGE, FEAR, and PANIC/Sadness systems were all inadvertently influenced by their selective breeding program? As we describe in Chapter 10 with a fish selection study for aggression (Bakker, 1994), in general it seems important for investigators to have better and more comprehensive behavioral assays for the various negative emotional systems, to determine if multiple systems were changing in response to the selection procedures. For example, lowering FEAR sensitivity in the low separation distress group may contribute to higher levels of aggressiveness (as we describe with Huntingford’s fish studies in Chapter 10). Overall, it seems likely that Brunelli and her colleagues have genetically selected for high and low PANIC/Sadness sensitivity in their high and low distress vocalization lines, but it will be interesting to see in future research whether the high separation distress line would respond similarly to separation distress vocalization manipulations studied in other species, such as reduction by opioids and direct brain stimulation manipulation of key brain sites, as further evidence for the actual involvement of the PANIC/Sadness system (see Panksepp, 1998a).

  Still, there are many other threads to the genetics of the personality story. Another group headed by Eva Redei at Northwestern University selectively bred two lines of rats to be either genetically susceptible or resistant to depression. Specifically, they selected sexually mature animals (seventy days old) based on their behavior in the forced swim test mentioned above. Rats that quickly gave up swimming and floated in the water were used to propagate their WMI (Wistar-Kyoto most immobile) line, with those swimming the longest selected as parents for the WLI (Wistar-Kyoto least immobile) line. By the second generation of progeny (F2s), significant differences emerged from the different parental pools (Will, Aird, & Redei, 2003). By generation twenty-two, the two lines no longer differed in their levels of anxiety as measured, for example, by activity in the center of an open field test, behavior in an elevated plus-maze, and blood levels of corticosterone after stressful restraint, all suggesting that the “behavioral differences between the two substrains of WKYs [the Wistar-Kyoto parental rats] are not fear or anxiety driven, but rather related to depressive state” (Andrus et al., 2012, p. 52).

  This group then used their two selected strains to study the inborn genetic tendencies to become depressed, in contrast to depression induced by chronic stress, which they had extensively studied in various nonselected strains (see Pajer et al., 2012). In the high endogenous depression group, they identified potential biomarker transcripts (a transcript is an RNA copy of a particular DNA segment, which is copied as the first step of gene expression) in the hippocampus or amygdala, as well as in the blood samples of the WMI and WLI strains (more on similar techniques in Chapter 18 on psychopathology). They also identified transcript differences in blood samples of humans with early-onset major depression disorder versus subjects with no disorder.

  In a news-grabbing finale, the Redei group was able to extend what had begun as a search for a genetic model of depression in rats—what we might call endogenous sensitivity to depression as a likely function of the PANIC/Sadness brain emotion system—into a human blood screening test for sensitivity to adult major depression disorder and possibly even the likelihood of responding to psychotherapy (Redei et al., 2014). In short, they identified nine genetic transcripts, which can be assessed in human blood samples, that distinguished depressed from nondepressed control subjects. Three of these transcripts distinguished control subjects from those with major depression even after the depression subjects had recovered. There were even gene candidates that correlated with whether depressed patients were treated successfully with therapy. Clearly, the Redei group’s research represents the value of animal research and a bottom-up approach to understanding human emotional problems. This diagnostic breakthrough is comparable to another therapeutic breakthrough that the study of animal primal emotional systems has fostered (GLYX-13), (Panksepp, 2015, 2016), which is disc
ussed in Chapter 18.

  RAT PLAY AND SEEKING SYSTEMS

  If fear, anger, and social distress tendencies contribute to negativistic temperamental tendencies, the evolutionary emergence of play behavior in mammals adds a positive balance and social complexity to the domain of personality. The stable variability in playfulness among mammals suggests it is a temperamental variable, but it is a hard one to selectively breed for because it takes two animals to play, plus selection for fearfulness always reduces playfulness (Panksepp, unpublished observations). However, a new positive-affect assay for studying positive playful feelings in rats involves a human hand taking the place of a partner rat pup and simulating a “play bout” while simultaneously monitoring positive affective vocalizations that are especially common during rat pup social play (Panksepp & Burgdorf, 2000). We know these 50 kHz ultrasounds reflect positive affect because anywhere in the brain one can evoke 50 kHz ultrasonic calls with brain stimulation (mostly in subcortical sites running along the SEEKING system), animals demonstrate these states are rewarding, because they always work (self-stimulate) to obtain this positive shift in emotional state (Burgdorf et al., 2007).

  Using 50-kHz ultrasonic vocalizations as an objective measure of positive social affect in these simulated play (tickling) sessions, Knutson, Burgdorf, and Panksepp (2002) selectively bred for high and low levels of this positive affective trait, yielding breeding lines that highlighted the genetic underpinning of social play (Panksepp & Burgdorf, 2003). After four generations of selective breeding, mean differences between high and low ultrasonic vocalization lines were already appearing. The high positive vocalization breeding line exhibited stronger social motivation to “play,” with shorter approach latencies and less avoidance time, than either the low vocalization or the randomly bred line of rats. The high vocalization line also exhibited more play behavior with other rats, and these rats were preferred as play partners compared to low vocalization line rat pups (Panksepp et al., 2001; Panksepp & Burgdorf, 2003).

  This simulated play with a human hand—sometimes referred as “tickling”—is very rewarding to rat pups and can be used to train them to approach a hand for a tickling bout. In addition, a Pavlovian conditioning procedure demonstrated the positive reinforcement value of play tickling in that a neutral signal with no initial power to elicit 50-kHz ultrasonic “chirping”—this gleeful vocalization rat pups emit when playing—elicited the chirping after being paired with hand tickling. Highlighting the appetitive nature of this positive social affect, play deprived (socially isolated) subjects exhibited stronger conditioning than did socially housed juveniles (Panksepp & Burgdorf, 2003).

  This play model has also been replicated generating new lines of high and low 50-kHz vocalizing rats (Burgdorf, Panksepp, Brudzynski, & Moskal, 2005). Again, after four generations of selection for high or low levels of the 50-kHz chirping, the high-vocalization line exhibited significantly more 50-kHz vocalizations than the other lines. In addition, research showed that negative affect 22-kHz vocalizations diverged in the opposite direction as well—happy rats complained less. Namely, they exhibited lower levels of 22-kHz distress vocalizations that have been associated with anxiety evoked by pain (Tonoue, Ashida, Makino, & Hata, 1986), addictive drug withdrawal (Mutschler & Miczec, 1998), predatory threat (Blanchard, Blanchard, Agullana, & Weiss, 1991), and social defeat (Kroes, Burgdorf, Otto, Panksepp, & Moskal, 2007). These two distinct types of ultrasonic vocalizations—50-kHz chirping versus 22-kHz distress calls—were also significantly negatively correlated in large groups of selected animals (r = –0.59, p < 0.0001; Burgdorf, Knutson, Panksepp, & Ikemoto, 2005), suggesting that the affectively positive 50-kHz vocalizations and affectively negative 22-kHz vocalizations represent polar opposite affective states in rats, confirming that these two forms of vocalizations are reciprocally related to each other (Burgdorf et al., 2001).

  A variety of other emotional phenotypic differences have been evident in these replicated lines. In generation fourteen, in addition to continued playfulness and ultrasonic vocalization differences, high 50-kHz animals also exhibited diminished aggression and biting when confronted by an intruder. Also, the low 50-kHz animals behaved more like “introverts” and spent less time in contact with each other when placed in the same cage (Burgdorf et al., 2009). Perhaps the cutest finding was that young animals preferred to spend time with adults who “laughed” (chirped) a lot compared to those that laughed little (Panksepp, 2007c).

  Overall, the selective breeding of highly playful rats supports the genetic basis of the PLAYful emotions within the brain. PLAY, which is the most evolutionarily recent of the six blue ribbon affective-neuroscience emotions related to personality, may also be the most complex of the six. PLAY integrates many psychological and behavioral elements providing young mammals rich social interactions and an opportunity to explore social limits in a relative safe context.

  The PLAY system is indeed complex, and additional work has indicated that the 50-kHz ultrasonic vocalizations can be broken into two types. There are frequency-modulated 50-kHz calls—a kind of “trill” that covers a broader sound spectrum—that are a better measure of positive social affect than “flat” 50-kHz calls (which may be a social-sampling “hello, is anyone out there” type of call). The ability to use these frequency-modulated rat vocalizations to differentially measure positive affect has also been validated in positive anticipatory SEEKING-type situations, including food anticipation, precopulatory mating situations, and anticipation of both natural play as well as the heterospecific hand play (tickling by an experimenter). In contrast, a large variety of negative affective situations promote the 22-kHz alarm calls. Overall, this suggests these vocalizations can be used to index temperamental positive and negative affects in rats. The rewarding and punishing nature associated with these vocalizations has also been supported by operant nose poking (rat investigating activity) being increased by playback of frequency-modulated 50-kHz vocalizations and decreased by 22-kHz calls (Burgdorf et al., 2008). Again, the scientific conclusion about these calls reflecting positive affective states in animals is derived from the fact that electrical stimulation of all brain sites that generated the frequency-modulated 50-kHz trills proved to be rewarding in self-stimulation tests (Burgdorf et al., 2007). Without that kind of prediction and confirmatory data, scientists would need to keep silent about (could not draw conclusions regarding) the affective states of nonspeaking animals.

  Related studies confirmed the frequency-modulated 50-kHz trills can be evoked by promoting dopamine activity, another link to the SEEKING system. Injections of amphetamine (a drug that simulates dopamine activity in the brain) into the shell of the nucleus accumbens, which is part of the mesolimbic dopamine pathway in the brain, significantly and robustly elevated the 50-kHz trills, especially in lines of rats selected for high positive affect compared to random and high negative affect lines of rats (Brudzynski et al., 2010). To further validate the neurochemical specificity of this effect and show that the increase of 50-kHz trills was mediated by dopamine, the coadministration of the dopamine antagonist raclopride attenuated the amphetamine effects, while cholinergic control drugs had no effects on these positive ultrasonic vocalizations. Indeed, because amphetamine and cocaine are addictive, it is noteworthy that the 50-kHz call can even be used as a spontaneous indicator of rats eagerly anticipating the receipt of such addictive drugs (Browning et al., 2011).

  Altogether, the above findings demonstrate how easily rats can be bred to be more playful and happy, to increase play and promote high positive affect, and that dopamine has a role in enhancing this positive affect. Moreover, this work has shown that rats can indicate their affective state through emotional vocalizations provoked under standardized rat-personality testing conditions, with the 50-kHz trills reflecting a positive affective state and the 22-kHz vocalizations a negative state.

  SUMMARY

  In sum, yes, rats have personalities. Indeed, they exhibit complex personalities, which includ
e expressing different levels of CAREing maternal behavior, which in turn have a direct impact on the stress tolerance of their offspring. Rats also share with cats, dogs, and other carnivorous mammals the capacity for two types of attack behavior: a predatory, quiet bite attack, which is linked to the SEEKING system, and attack behavior appropriate for defending their various resources as well as themselves, which is linked to the RAGE/Anger system.

  Rats possess a complex FEAR system that psychologists have discovered is easily activated and is interwoven with all other emotional brain systems: SEEKING (curiosity), RAGE/Anger (defense), CARE (maternal nurturing), PANIC/Sadness (separation distress), and PLAY (joyful social interaction). Many times rats have been selected for high and low levels of fear expression, confirming the genetic basis of the FEAR system. Likewise, at least two labs have selected for high and low levels of PANIC/Sadness behavior, reflecting depressive tendencies. The second of these (Redei’s group; see Andrus et al., 2012) recognized the importance of ensuring that elements of FEAR did not confound the behavioral differences in their high and low depression (PANIC/Sadness) lines, which allowed them to analyze gene expression in their rat subjects and ultimately generate blood screening tests for human depression. Along similar lines, Panksepp’s group on three separate occasions has selected rats for high and low levels of PLAY behavior, which has resulted in the development of a drug to treat depression (Burgdorf et al., 2011; Panksepp, 2014; discussed further in Chapter 18).