The further that research in the neurochemistry of emotionality advances, the more that neurochemical systems are being linked to emotional regulation. In fact, a dysregulation in practically all neurochemical families – monoamines (MAs), acetylcholine (ACh), GABA, hormones, neuropeptides, opioid receptors (ORs), transcription factors, etc. – appears to contribute to mood and anxiety disorders. However, the same neurochemical systems were implicated in the regulation of non-emotionality aspects of behaviour (attention, planning, habits, plasticity, and, as discussed below – empathy). There is, therefore, a functional overlap of neurochemical systems regulating aspects of behaviour related and unrelated to emotionality. As discussed elsewhere (Trofimova et al., 2018; Trofimova, 2021a,b; Trofimova and Netter, 2021), sorting through the neurochemical biomarkers of specific aspects of behavioural regulation is a very challenging task, requiring not only experimental research and statistical models but also detailed meta-analysis. The challenges arise from the fact that neurochemical systems regulate each other, have multiple functionalities, and represent multiple neurotransmitters and mediators. Each of these neurotransmitter systems has several types of receptors, each having different functionality and locations in the brain. Since all of these systems are contingent on each other’s activities, a one-to-one mapping between neurochemical systems and specific functional aspects of behaviour, temperament traits or symptoms of psychiatric disorders would be, therefore, not a very efficient way to present their interplay. Instead, an ensemble (multi-marker) presentation might be more fruitful.

In sorting out neurochemical systems of emotionality, one more complicating factor is that emotionality, despite the old age of this concept, is still poorly defined and often defined circularly (for example, “emotionality is how people express and feel their emotions”). There is no consensus or clear definition of emotionality even among psychologists. The American Psychological Association (APA) defines emotionality mostly based on its arousal component, as “the degree to which an individual experiences and expresses emotions, irrespective of the quality of the emotional experience.” Yet, there is another, a well-documented component of emotionality related to emotional valence (i.e., emotions having either positive, secure-to-happy or negative, dysphoric types) (Barrett and Bliss-Moreau, 2009; Barrett, 2017). In fact, without emotional valence (i.e., not having negative or positive affectivity), a solo component of arousal would not be experienced as emotion. This is what happened in classic experiments in the 1920–1940s when participants were asked to report their feelings after they were injected with epinephrine (adrenalin) (reviewed in Dror, 2016). Although the majority of participants reported feeling aroused, only a few experienced something emotional. Those who did report an emotional experience described it as having “cold,” and not genuine emotion. For example, in fearful contexts, they reported feeling “as if I am afraid” rather than being really afraid. The role of epinephrine and other hormones in behavioural arousal justified Schashter proposal to include arousal to components of emotionality, in addition to the cognitive appraisal component (Schachter and Singer, 1962).

One of the approaches that we found beneficial in sorting through neurochemical biomarkers of consistent behavioural patterns (CBPs) (temperament traits and symptoms of psychopathology) is the functional constructivism (FC) approach.

The Constructivism approach has been known for about a 100 years in general psychology (Bartlett, 2009; see for review Trofimova, 2017, 2021a,b) and especially in the psychophysiology of emotions (Russell, 2003; Vuilleumier, 2005; Ghashghaei et al., 2007; Barrett, 2009, 2017; Pessoa, 2010). This approach suggests that all behaviour is not reactive but constructive; all actions are constructed anew as the integration of behavioural capacities, needs, and alternatives that were selected and sequenced for their relevance across multiple levels of behavioural regulation. Even when consistent behaviour looks like cycles, habits, and repetitions, it is actually never repeated but generated anew based on an individual’s current capacities, intentions, and situational context (Bernstein et al., 1996). A descendent of Constructivism is the FC approach that classifies the functions of neurobiological systems by their contribution to action construction. Neurobiological systems can have multiple and overlapping functionalities, and it is our position that this functionality was reinforced in evolution by its role in the sustainability of both individual and species functioning. To classify the functionality of human neurotransmitter systems, the FC suggests, therefore, looking at the stages and functional aspects of action construction identified in the sciences of kinesiology and functional neurophysiology that specifically analyse these stages. The pioneers of this analysis were Bernstein, who is credited as the father of kinesiology (Bernstein et al., 1996), and Anokhin (Anokhin and Serzhantov, 1973; Anokhin, 1984), the author of the functional systems theory.

The FC approach to bio-behavioural taxonomies was implemented in the neurochemical framework Functional Ensemble of Temperament (FET) that summarised the most conservative findings regarding the functionality of neurotransmitters, hormones, and OR systems (Trofimova, 2016, 2018, 2021a,b; Trofimova and Robbins, 2016). The temperament of healthy individuals (including animals) and symptoms of psychiatric disorders can be presented as a continuum of weak vs. strong dysregulation within neurochemical systems generating CBPs (Sulis, 2018; Trofimova and Sulis, 2018; Rezaei et al., 2021; Trofimova, 2021b). This conception of the continuum is supported by decades of research in psychopharmacology and neurochemistry, showing the impact of neurochemical dysregulation (Sulis, 2018). The FET uses a “multimarker” approach, suggesting that there is no one-to-one correspondence between a specific CBP and any neurochemical system. Instead, every psychiatric symptom or temperamental (bio-behavioural) trait is associated with a team of neurochemical systems (Figure 1; Trofimova, 2016, 2018, 2021a,b; Trofimova and Robbins, 2016).

People face a vast number of behavioural alternatives, and so they must select which is most appropriate to a given situation for behaviour to be organised. Before this selection can be completed, however, individuals have to consider and reject multiple alternatives (degrees of freedom, d.f.) that are less suitable. This expansive consideration of alternatives is called here behavioural orientation. Selection of the most suitable d.f., sequencing them and suppressing the remaining d.f. is called integration, or programming. Generating similar actions using chosen integrations is called maintenance of behaviour. The FET framework follows Luria’s insights from clinical neuropsychology and insights from kinesiology, which identify three functional aspects of action construction: expansion of d.f. (orientation), selection/integration of a programme of action, and energetic maintenance (including decomposition of unneeded d.f.) (three columns of Figure 1). The FET also follows the activity-specific approach to taxonomies of temperament that differentiates between biomarkers regulating physical, verbal–social, and mental aspects of behaviour (Trofimova and Sulis, 2011; Trofimova, 2016; Trofimova and Robbins, 2016; Rusalov, 2018) (three top rows of Figure 1). The validity of this approach is supported neuroanatomically (by the functional specificity of brain areas regulating physical, verbal, and probabilistic aspects of behaviour), neurochemically (by the specificity of hypothalamic hormones differentially regulating social and physical aspects of behaviour) (Vazquez-Borrego et al., 2018; Swaab et al., 2021), and clinically (by differentiating between major psychiatric diagnoses) (Trofimova and Christiansen, 2016; Trofimova and Sulis, 2016a,b, 2018).

Moreover, analysis of neurochemical biomarkers of universal functional aspects of behavioural regulation showed that in terms of timing, they can be separated to those that regulate “here and now,” fast, explicit changes in actions and those that relate to eventual (“ever”), implicit and probabilistic aspects of events (Trofimova, 2021a). Our review focuses on this distinction, further analysing the roles of hormones in fast-changing situations and the roles of ORs in dispositional emotionality.

Due to space limitations, we will mention the role of OR systems only briefly since this role is discussed in greater detail in other reviews (Leknes, 2008; Trofimova, 2016, 2018, 2021a,b; Trofimova and Robbins, 2016; Bodnar, 2018).

Opioid receptor systems are well-documented players in emotionality (Bodnar, 2018). Moreover, more and more evidence is emerging suggesting the role of gut microbiota and the immune system in the regulation of dispositional emotionality (Cryan and Dinan, 2012; Fung et al., 2017). OR action is differentially complemented or contra-influenced by the action of microbiota and cytokines of the immune system. Therefore, whenever there is a reference to the action of the OR system, we refer to the action of OR, microbiota and immune systems as one complex. In contrast to microbiota, which rarely has direct access to brain structures, influencing them indirectly, the opioid peptides have a more direct way to regulate brain systems. We appreciate that these three systems (OR, microbiota, and immune cells) are entirely different neurochemically and biologically, ranging from peptides to complex cellular organisms. Thus combining them into one “body-biased” system is, therefore an over simplification and is done here temporarily for conceptual purposes.

There is good consensus in neurochemistry on the key role of endorphins binding to mu-opioid receptors (MORs) in positive emotional valence, physical and emotional pain relief (Leknes, 2008; Bodnar, 2018), relaxation, feeling of security, and affiliation (Barr et al., 2008; Curley, 2011). The action of MOR, activated by endorphins, suppresses the activation of the Hypothalamic-pituitary-adrenal (HPA) axis (i.e., suppresses mobilisation for an action) and induces the release of dopamine (DA) (Werling et al., 1988; Carr and Gregg, 1995; Degli Uberti et al., 1995; Yamauchi et al., 1997; Bodnar, 2018). There are several sites in the body and the brain that manufacture endorphins, and we can classify these sites into three groups (Figure 2).

The other locations of MOR in the brain, which the endorphins bind to, were linked to behavioural homeostatic maintenance (Golgi cells in the cerebellum and in reticular formation) as well as motivational and programming processes [nucleus accumbens (NAc), ventral tegmental area (VTA), prefrontal cortex (PFC), amygdala (AM), HT, PAG] (Hamel and Beaudet, 1984; Nieuwenhuys, 1985; Peckys and Landwehrmeyer, 1999; Carlezon et al., 2009; Le Merrer et al., 2009). MOR, and also delta-opioid receptors (DORs), are G protein-coupled receptors (GPCRs) regulating the release of MAs, including the facilitation of dopamine release (Werling et al., 1988; Carr and Gregg, 1995; Degli Uberti et al., 1995; Yamauchi et al., 1997; Bodnar, 2018). Considering the “approval” effect of endorphins and their links to positive emotional valence, there is no surprise that these brain structures were noted for their involvement in motivation, planning, and anticipation. Limbic and cortical processing of the context of situations supplies the information about environmental needs and alternatives, and this information is always compared to the MOR-based summary on the homoeostasis of the body.

Emotional valence is seen here as the multi-“voting” summary from several endorphin locations managing the homeostatic balance of the body state. Positive emotional valence is experienced when the amount of endorphins corresponds to demands of MOR density, and negative (dysphoria, irritability) when this amount is insufficient, whether due to underproduction of endorphins or upregulation (increased density) of MOR receptors.

Kappa-opioid receptors (KORs), despite belonging to the same neurochemical family as MOR, have significantly less effect on the valence of emotions in comparison to MORs. Instead of emotional valence, it is more accurately to associate KOR systems with sensory mobilisation processes activated to expand behavioural d.f. and to search for alternatives. Thus, KOR activation was linked to aversion, chronic anxiety, hallucinations and malaise, and activation of noradrenalin (NE) during chronic stress, but not to acute, situational stress, or specific phobias (Reyes et al., 2008; Schwarzer, 2009; Wittmann et al., 2009; Tejeda et al., 2012). There is significantly less KOR than MOR in the brain in general; however, the thalamus (anterior and lateral nuclei, i.e., the brain area integrating sensory information), has a significant presence of KOR in comparison to other brain structures (Hamel and Beaudet, 1984; Delay-Goyet et al., 1987; Mansour et al., 1987; Sharif and Hughes, 1989). Dynorphin that binds to KOR was found in GABA neurons in the central nucleus of the AM and lateral and medial hypothalamus (Tejeda et al., 2012), and high KOR density was also found in the lateral AM, HT, ventral striatum, and PFC (especially layer VI) (Mansour et al., 1987) (i.e., in the structures that, in addition to the thalamus, strongly contribute to perceptual and orientational mobilisation) (Hamel and Beaudet, 1984; Delay-Goyet et al., 1987; Mansour et al., 1987; Sharif and Hughes, 1989). KORs modulates NE release from the locus coeruleus (LC), contributing to the alertness in attentional and novelty-processing function of the NE system (Schwarzer, 2009; Bodnar, 2018), and, therefore, to perceptual sensitivity. This explains why in chronic stress (i.e., when KOR activation is above the optimal limit), people report an inability to focus and discomfort in novel situations, as an excess of NE release was also linked to “above optimal arousal” and inability to focus.

It has been shown that the same NE axons that respond to dynorphin have a co-localisation with excitatory glutamate (Glu) as well as stress-related corticotropin-releasing factor (CRF) (Reyes et al., 2007, 2008; Wee and Koob, 2010). Dynorphins that bind to KOR (and not other ORs’ peptides) appeared to be able to stimulate HPA axis activity, to increase adrenocorticotropic hormone (ACTH) via action on CRF (Bruchas et al., 2010) and cortisol release (Williams et al., 2003; Pascoe et al., 2008; Miller et al., 2018). This explains the involvement of KOR in alertness and (in clinical cases) dispositional anxiety.

The three groups of locations, including dispositional emotional valence, described earlier, also have neurochemical systems inducing dispositional alertness:

Despite the association of KOR with NE and HPA arousal (Hauser et al., 2005; Tejeda et al., 2012; Bodnar, 2018), KOR action is very specific to dispositional chronic anxiety and stress but not to acute stress or specific phobias (Schwarzer, 2009). Thus, studies on animal models reported that KOR antagonists decrease avoidance and anxious dispositions (Schwarzer, 2009; Van’t Veer and Carlezon, 2013; Bodnar, 2018); however, KOR agonists don’t produce a transient elevation of anxiety. Instead, they produce perceptual distortions of sensory stimuli, depersonalisation, speech processing problems, and thought disorganisation (i.e., features of perceptual over-arousal) (Hauser et al., 2005; Tejeda et al., 2012; Van’t Veer and Carlezon, 2013). Moreover, only high doses of KOR agonists induce anxiety, while low doses act as an analgesic, and very low doses may induce positive mood states (Schwarzer, 2009; Bodnar, 2018). It would be, therefore, wrong to consider KOR as a system of negative emotional valence even though they are often in rival relationships with MOR and activate the HPA axis.

The production of opioid peptides and even the existence of their receptors appeared to be a very plastic process, opening a window for individual variability. The density of ORs can be increased (upregulation) or decreased (downregulation and desensitisation) depending on the supply of endorphins binding to them and the sensitivity of the receptors themselves. A single administration of opiates often triggers a set of changes that usually is restored by a chain of recovery mechanisms (Waldhoer et al., 2004; Christie, 2008; Carlezon et al., 2009; Schwarzer, 2009; Nagi and Piñeyro, 2011). Downregulation of receptors is observed mostly after chronic overuse/overproduction of these receptors’ agonists as a protective feedback mechanism, and upregulation develops in cases of consistent deficit of needed peptides (due to their under-production) (Waldhoer et al., 2004; Christie, 2008; Carlezon et al., 2009; Schwarzer, 2009; Nagi and Piñeyro, 2011). Moreover, there are factors other than just chronic overuse of opiates affecting up–down regulation of ORs, such as genetic dispositions or exposure to toxins, etc. (Waldhoer et al., 2004; Christie, 2008; Carlezon et al., 2009; Schwarzer, 2009; Nagi and Piñeyro, 2011). Individual differences in the production of opioid peptides and/or in the density of their receptors could, therefore, produce consistent individual differences in dispositional emotionality. These consistent individual differences are colloquially named in the FET as emotionality-related traits of the temperament of dispositional Satisfaction (based on MOR), Impulsivity (not discussed here), and Neuroticism (based on KOR).

The difference between dispositional emotionality and situational emotional experiences is that emotional dispositions are experienced before or regardless of the triggering events, whereas commonly known emotions are described as a type of reaction to specific situations (Trofimova, 2018). In contrast to that, hormonal and MA–ACh systems are involved more in the regulation of situational selection of d.f.: hormones as fast-acting regulators, more in control of the somatic systems that have to adjust to explicit and immediate features of events, and MA–ACh systems as regulators of behaviour related to not immediate, distant, and/or abstract features of events (probabilistic).

In line with the idea of a neurochemical continuum between temperament traits and symptoms of psychopathology (Sulis, 2018; Trofimova and Sulis, 2011; Trofimova, 2021a,b), weak biases in the ratio between the supply and demand of endogenous peptides binding to OR can be expressed as emotionality-related traits of temperament whereas strong biases – as psychiatric disorders. Figure 2 illustrates several scenarios of emotional dispositions based on MOR and KOR systems of emotional valence and alertness. Indeed, upregulation of MOR was linked to extreme dysphoria and irritability, as seen in borderline personality disorder and attachment disorders (Cacioppo et al., 2000; Christie, 2008; Leknes, 2008; Bandelow et al., 2010; Curley, 2011). In contrast, MOR downregulation can be experienced as dispositional, often negligent satisfaction with “life as is,” low productivity, high agreeableness, and dispositional positive mood. As noted above, dysregulation in KOR systems can lead to chronic alertness (anxiety, neuroticism) and, on the opposite pole – to dispositional indifference. Indeed, clinical levels of KOR dysregulation were linked to agitation, chronic anxiety, chronic stress, inability to relax, and behavioural arousal seen in generalised anxiety disorder (Hauser et al., 2005; Reyes et al., 2008; Schwarzer, 2009; Wittmann et al., 2009; Schindler et al., 2010; Wee and Koob, 2010; Tejeda et al., 2012).

Traditionally, hormones are viewed as neurochemical systems of emotionality. Studies of the structure and functionality of the pituitary–hypothalamic complex, however, showed that this complex has well-defined morphological and functional segregation linked to body homoeostasis, largely unrelated to emotional regulation. The pituitary’s segregation, and hypothalamic projections to parts of the pituitary, support the idea of the differential regulation of physical vs. social aspects of behaviour, known as the activity-specific approach in temperament, which is one of the principles of the FET structure (Trofimova, 2016; Trofimova and Robbins, 2016; Trofimova and Sulis, 2016b; Rusalov, 2018). Indeed, there are groups of cells in the anterior pituitary generating hormones and neuropeptides which regulate the body’s physical endurance and metabolism (such as thyroid, growth hormone, and somatostatin), whereas so-called “social hormones,” OXT and arginine–VSP are manufactured by the magnocellular neurons hypothalamic supraoptic (SON) and paraventricular (PVN) nuclei projecting to the posterior pituitary (Vazquez-Borrego et al., 2018; Swaab et al., 2021).

Gonadal hormones regulated by the pituitary–hypothalamic complex (called the hypothalamic–pituitary–gonadal axis) have a distinct set of secretion sites and also represent an example of a system regulating non-emotional (sexual and reproductive) aspects of behaviour. Gonadal hormones, especially estrogen, regulate not only glands associated with sexual and reproductive functions but also the brain’s neurotransmitters and associated memory functions (noticeable in post-menopausal women) (Chavez et al., 2010; Lu et al., 2019) and behavioural plasticity (Almey et al., 2015; Lu et al., 2019) (i.e., also non-emotionality aspects of behaviour).

Finally, the HT PVN releasing corticotropin-releasing hormone/factor (CRF) as the component of the HT–pituitary complex was, for a long time, linked to emotional regulation due to its well-documented involvement in stress response. However, its involvement contributes more to the preparation of the body for sudden dramatic changes in behaviour and not even necessarily negative emotional valence of events. For example, HPA axis activation occurs during a sexual act between loving partners in very safe and trusting relationships, largely due to dehydroepiandrosterone (DHEA) produced by adrenal glands (Pluchino et al., 2015). Adrenaline/epinephrine is a substance acting on sympathetic ANS, and its effects are very diffuse, producing a generalised state of arousal. Contrary to the common view of adrenalin as primarily a stress hormone, studies from the 1960s show that the action of adrenalin is not specific to the fear reaction: adrenalin increases the emotion of anger in the presence of an anger-provoking stimulation, and a euphoric emotion in the presence of a euphoria-provoking stimulation (Schachter and Singer, 1962; Dror, 2016). As noted above, participants in these experiments reported “cold,” and not genuine emotions largely influenced by the context (“knowing” that they are probably afraid rather than being afraid in fearful situations and “knowing” that they are probably angry in anger-provoking situations). Adrenaline has since been recognised as a hormone preparing the body for an intense and immediate behaviour and providing behavioural arousal rather than inducing specific emotions (Schachter and Singer, 1962).

In this sense, here we point out that all parts of the HT–pituitary complex are more involved in the homeostatic regulation of the body’s state in fast-changing contexts rather than in mainly emotional regulation.

If we want to progress with the mapping of biomarkers of emotionality, traditional assignment of behavioural regulatory systems to brain structures and neuro-somatic locations should be likely reconsidered. It is important to note that brain structures differ in their functionality only because they consist of different chemical compositions. Functional roles of neurochemical systems, therefore, provide perspectives for analysing the involvement of brain structures in behavioural regulation. The action of opioid peptides is a clear example that it is not the location but the neurochemical system (presented in multiple locations) that has a distinct functionality in behavioural regulation. The examples are diffuse action of endorphins binding to MOR inducing dispositional emotional valence and action of dynorphins binding to KOR in dispositional alertness. All locations [gut microbiota, HT–pituitary complex, and ascending reticular activating (ARAS)-cortical complex] manufacturing and/or binding opioid peptides contribute their “votes” to emotional dispositions, each using different mechanisms and neurochemical partners (Figure 2).

The three groups of location for endorphin production all induce a sense of satisfaction, “approval” of current behavioural choices but with slightly different functionality. Guts endorphins give the dispositional sense of body’s comfort (or discomfort needing to be addressed) unrelated to the current events; HT–pituitary endorphins suppress hormones-based reactivity to fast-changing events, inducing the dispositional approval of current behavioural choices; finally, the brain MOR activity contributes to approval d.f. in actions or perceptions related to future, distant, or abstract events, i.e., more implicit elements of behavioural regulation. The last type of effect is seen in the improved mood while planning or dreaming of possible positive outcomes. The subjective feeling of satisfaction, approval, and acceptance can be seen in all activities linked to endorphins release – exercising, having meals, setting up plans, and following plans.

In contrast, KOR locations and entanglement with HPA arousal and NE release indicate KOR different functionality unrelated to emotional valence. The FET considers KOR systems associated with emotional dispositions for perceptual and behavioural alertness associated with expansion/search for alternative d.f. in behaviour. Such alertness works as an emotional, dispositional amplifier of orientation aspects of behaviour. Even in addiction studies, KOR systems have been implicated in ethanol-seeking behaviour, often comorbid with chronic anxiety (Wee and Koob, 2010; Tejeda et al., 2012), supporting the idea of their key role in perceptual arousal, amplifying behavioural orientation.

There are, therefore, functional differences in the effects of these two OR systems, which we can colloquially call MOR-based “satisfaction–dissatisfaction” (or “approval–disapproval”) and KOR-based “Alertness (expansion of d.f.)-indifference,” noticed in Neuroticism. MOR and KOR systems often act in opposite directions and suppress each other; however, they sometimes act in unison or independently of each other on other neurotransmitters (Pan, 1998; Bodnar, 2018). As noted, KOR do not induce emotional valence, and MOR system can add “sweeteners or bitterness” to both behavioural alertness and inactivity (as in relaxation) (Figure 2). These systems are, therefore, are functionally distinct and do not represent two polarities of one axis.

The roles of brain neurotransmitter systems – MA, ACh, Glu–GABA (GG) [not discussed here but reviewed earlier (Trofimova and Robbins, 2016; Trofimova, 2021a,b) differ from those played by OR and hormonal systems, as they provide a probabilistic assessment of events and generation of potential programmes of actions. In this context, it might be beneficial to separate at least three aspects of behavioural regulation that currently are considered as part of emotionality, but only one aspect can be viewed as true emotionality (Figure 4):

All three classes of neurochemical systems contribute to the selection of d.f. in actions. This selection is an interactive process, influenced by biases from OR-based emotional dispositions, the hormone-based situational status of the body and context-processing managed by brain neurotransmitters. The feedback at the preparation and execution stages of action affects the hormonal and brain neurotransmitter’s regulation of behaviour (middle and upper groups) but has a less immediate impact on emotional dispositions.

Also, as reflected in the title of the FET framework, all three classes of neurochemical biomarkers work in ensembles.

First of all, OR receptors represent the class of GPCRs regulating the release of all main neurotransmitters and many hormones. Therefore, they affect non-emotionality aspects of behavioural regulation regulated by brain neurotransmitters. As summarised in other FET-related reviews (Trofimova, 2021a,b), MORs induce DA release in VTA–NAc and suppress NE release, whereas KOR action can lead to the opposite pattern (Degli Uberti et al., 1995; Carlezon et al., 2009; Wittmann et al., 2009; Tejeda et al., 2012; Bodnar, 2018). Functionally, this means that the KOR system can pause the DA-based integration of actions but boost the NE-based search for novel alternatives, whereas MORs can suppress this new search (also the contextual analysis) and “approve” the existing alternatives in actions. There is also a strong neurochemical cooperation between the MOR system and the serotonin system, which is also prominent in the homeostatic maintenance of behaviour. It has been shown that MOR (but not other ORs) agonists inhibit 5-hydroxytryptamine (serotonin) (5-HT) in the hippocampus (HC) (Yoshioka et al., 1993). In our experiments, when pregnant rats were chronically exposed to opium that caused downregulation of MOR receptors, they also had a significant decrease in endurance in four types of maternal behaviour, a decrease of 5-HT in HC, a decrease of BDNF and an increase of corticosterone (stress hormone) (Rezaei et al., 2021). Also, out of all OR, only MOR develop heteromeric complexes with the 5-HT1 receptors (Fujita et al., 2014).

If OR and microbiota induce dispositions unrelated to the features of situations, hormonal systems are more tuned to the features of situations, regulating the body’s state related to fast-changing and explicit features. Although the OR and hormonal systems interact, they can be still seen as functionally distinct. As the main trend, OR systems suppress hormonal release, whether it is gonadal (Liu et al., 2011; Ho, 2019), social (Smith and Wise, 2001; Loth and Donaldson, 2021), or stress hormones. Gonadal hormones, in turn, have a way to modify OR binding and density, also in a very specific pattern. Estrogen can internalise MOR in cell groups of the limbic-hypothalamic complex, and this modulates sexual receptivity (Micevych et al., 2003; Huhn et al., 2018). There are, however, exceptions to this trend, indicative of more specific functionality of OR systems than just universal “suppressors.” Thus, KOR can activate VSP release (Zhao et al., 1988; Bernstein et al., 1996; Bodnar, 2018). Also, MOR density is higher in females than in males, indicative of MOR-estrogen interactions (Huhn et al., 2018). Moreover, while estrogen decreases KOR-binding dynorphins expression in the arcuate nucleus of the HT, testosterone increases dynorphins in the supraoptic nucleus of HT and bed nucleus of stria terminals (Becker and Chartoff, 2019).

These interactions between three classes of neurochemical systems implicated in emotionality can explain the observations of emotional dispositions being capable of affecting all non-emotional aspects of behavioural regulation – orientation, integration of a programme of actions, energetic maintenance (green and red arrows in Figures 1, 3, 4). As seen above from the general impact of MOR action and also from its cooperation with 5-HT, MOR-based Satisfaction/Approval systems amplify the sense of homeostatic maintenance balance; KOR-based systems of Alertness-Expansion (known as Neuroticism) amplify orientation aspects of behaviour. Impulsivity (not discussed here) can be seen as an amplification of the integration and initiation of actions.

Finally, cognitive appraisal (emerging from the activity of brain neurotransmitter systems involved in contextual processing) is often seen as part of emotionality (Schachter and Singer, 1962; Dror, 2016). Indeed, social emotions are experienced differently in those contexts where an individual was either an agent or a victim of harm (Kanen et al., 2021), illustrating the impact of the upper group of neurochemical systems in Figure 4. Moreover, evaluative biases stemming from emotional dispositions have been shown to affect all kinds of perceptual processes. This led to the theory of somatic markers (Damasio, 2000) and embodiment in cognition, which we first named as the “phenomenon of projection through capacities” when we found this effect in the late 1990s (Trofimova, 1997, 1999). These evaluative biases impact even the highest levels of cognition, such as attribution of meaning to emotionally neutral, abstract words (Trofimova, 1999, 2014). In our experiments, we demonstrated the existence of cognitive biases associated with temperament traits and sex (i.e., neuro-chemically based individual differences) of healthy participants affecting their highest form of cognition (meaning attribution). Temperament traits of high physical and social endurance, faster tempo in motor and verbal tasks were found to be associated with a positivity bias, whereas the traits of neuroticism, high emotionality were associated with a negativity bias in estimations of neutral abstract and very common concepts (Rusting and Larsen, 1998; Trofimova, 1999, 2014; Zelenski and Larsen, 1999). Interestingly, participants with high intellectual endurance (sustained attention) and with high probabilistic thinking had a significant negativity bias in their estimations of abstract common words. This is consistent with reports that MORs suppress brain ACh release, which, in turn, is the major neurotransmitter regulating sustained attention and probabilistic processing (Trofimova, 2019). Moreover, in line with the ETS theory, males in three different cultures had a positivity bias toward sensational concepts and evaluative criteria, whereas females favoured ordered and predictable objects (Trofimova, 2012, 2013, 2014).

These effects point to the impact of hormonal systems regulating social perception and underlying sex differences on perception in general, which can or cannot be associated with emotional valence.

The analysis of neuroimaging studies investigating neuroanatomic biomarkers of positive and negative affects showed no significant correlations between brain areas and positive affectivity (Kaufmann et al., 2019). Moreover, studies of psychiatric disorders associated with negative affect showed that practically all brain structures were involved in negative emotions. This lack of specificity of brain structures in emotional valence highlights the need to focus more on their neurochemical rather than neuroanatomic biomarkers.

Neurochemical hypotheses, however, can also have flaws, and one of the examples of common misconceptions is the attribution of positive emotional valence to the direct action of DA. Meanwhile, the release of DA in ventral striatum (so-called “rewards circles”) is controlled and facilitated by MOR activation as part of a GPCR mechanism dominating MA release. The “reward circuits” include the VTA and the NAc, which indeed are rich in DA and are highly involved in addiction. To disentangle the functional roles of MOR and DA in positive emotionality we can look at the effects on DA release when MOR systems are not involved.

First of all, the release of DA was more consistently associated with a reaction to the significance, saliency, of stimuli rather than to positive emotions. Thus, activation of DA neurons in the VTA, was found to be similar whether it was induced by negative experiences (prolonged stress) (prolonged stress) (Ortiz et al., 1996; Reynolds and Berridge, 2001; Carlezon and Thomas, 2009), the anticipation of aversive stimuli (Ortiz et al., 1996), or, as well-known, pleasurable experiences (psychoactive drugs). Some studies report that treatments that increase NAc excitability depress mood, whereas treatments that reduce this excitability appear to elevate mood (Carlezon and Thomas, 2009). Second, the DA hypothesis of positive emotional valence was confronted with non-supportive reports (Jensen et al., 2003; Saal et al., 2003). Appetitive stimuli appear to enhance activity in the mesocortical DA system to a lesser and not higher degree and for a more transient period than did aversive stimuli (Robbins and Everitt, 1996; Jensen et al., 2003; Saal et al., 2003; Seamans and Yang, 2004). Higher DA release was reported not only in positive but also in negative circumstances, such as a defeat, aversive stimuli, stress, and foot shock (Robbins and Everitt, 1996). Third, “reward circuits” were reported to be activated during anticipation of getting a reward but not at the receipt of the reward (O’Doherty et al., 2002), suggesting that they process the projection of future events rather than being responsible for inducing specific (positive) emotional valence. Finally, addiction itself appeared to be not driven by the positive affect but by the damaged habit system (drug addicts often report that they don’t like their drugs but just can’t stop themselves) (Everitt and Robbins, 2016).

Earlier, we pointed out the more consistent roles of DA in the regulation of the assignment of priorities, salience and sequencing to behavioural elements necessary for behavioural plasticity and integration of actions (Trofimova and Robbins, 2016; Trofimova, 2021a,b). This is in consensus with the position of other authors (Seamans and Yang, 2004; Yin and Knowlton, 2006; Tepper et al., 2007; Seamans and Robbins, 2010; Graybiel and Grafton, 2015; Haber, 2016) and specialists in DA research pointing to the flaws of the “dopamine theory of extraversion and positive affect” (Robbins and Everitt, 1996; Seamans and Yang, 2004; Robbins, 2018). Dysregulation within the DA system was indeed found less associated with changes in emotionality and more so with a compromised ability to integrate behaviour adequately to a situation (i.e., impulsivity, rigidity, OCD) (Szechtman et al., 1998; Seamans and Yang, 2004; Robbins, 2010, 2018). The remarkable role of DA in “salience-labelling” can be seen in its association with a pathological attachment of importance to irrelevant stimuli in schizophrenia (Gray, 1998; Coyle, 2006) and psychoticism (Kapur, 2003), both linked to an excess of DA but both associated with negative, and not positive, affectivity. Meanwhile, some studies of DA D4 receptor genes found no associations with Extraversion (Munafo et al., 2008). Moreover, the early opinion that Extraversion (or general arousal) was based on the cortical-ARAS system was confronted by experiments in neurochemistry that showed that the ARAS system has at least four different sub-systems of arousal, diverging in their functionality, and directionality of neurotransmitter release (Robbins and Everitt, 1996; Trofimova and Robbins, 2016). In line with other researchers (Ruhé et al., 2007), the FET framework suggests that neither DA nor other MAs (NE, serotonin, 5-HT) directly regulate moods. Instead their release during emotional experience and commonly described MA roles in emotionality-based psychopathology (for example, Liu et al., 2018), are, therefore, secondary to the activation of OR systems that regulate MA release. Positive or negative emotional valence is so far more consistently associated with the effect of mu-ORs acting on MA as part of a GPCR mechanism (Trofimova and Robbins, 2016; Trofimova, 2021a,b), rather than the effect of release of monoamines per se.

Brain neurotransmitter (NT) systems work in ensembles with hormonal and OR systems on the selection and composition of a behavioural act (as depicted in Figure 4). OR systems provide the body’s bias for decision making on “what’s next” in behaviour (“accept vs. not-accept,” “change-search vs. not-change”); hormonal systems provide the tuning of the body to immediately present and fast-changing aspects of events, whereas brain neurotransmitters deal with not immediately present, abstract, and probabilistic aspects of events. What is called “emotionality” is, therefore, presented here as a multi-component process resulting from the “voting” of many body and brain systems on the prognosis of events. Future research, therefore, can focus on disentangling the roles of these three neurochemical families in behavioural regulation by controlling for biases stemming from the action of OR and hormones on brain NTs during emotionality-related experiments.

Another trait that relates to behavioural orientation to specific reinforcers and that is often considered as an emotionality trait is EMP. Neurochemically, empathy is linked to hormonal systems, namely “social hormones,” such as OXT and arginine–VSP, that are manufactured by the magnocellular neurons in the hypothalamus and project to the posterior pituitary (Vazquez-Borrego et al., 2018; Swaab et al., 2021). Unlike “stress hormones,” reliably linked to the arousal component of emotionality, “social hormones” by themselves showed significantly less association with an emotional valence or emotional arousal. OXT was shown to induce a calming effect on HPA axis arousal (Quintana et al., 2016), including in association with estrogen (Ochedalski et al., 2007; van Anders et al., 2011; Gabor et al., 2012; McCall and Singer, 2012), but not necessarily positive emotions.

More consistently, OXT has been reported to have an association with pro-affiliative perception (van Anders et al., 2011; McCall and Singer, 2012; Decety et al., 2014; Shamay-Tsoory and Abu-Akel, 2016; Kanat et al., 2017; Jones and Sloan, 2018; Declerck et al., 2020). The neurochemical framework FET (Trofimova, 2016, 2021b; Trofimova and Robbins, 2016) summarised the consensus in the literature on the key role of OXT–VSP (in specific combination with gonadal hormones) in empathy (i.e., social type of behavioural orientation to other’s people motives and psychological states). Other authors also support the “social salience hypothesis of oxytocin” (reviewed in Shamay-Tsoory and Abu-Akel, 2016; Liu et al., 2019), underlying the tendency for OXT administration to increase pro-social perception, including trust (Declerck et al., 2020), face recognition (Decety et al., 2014; Kanat et al., 2017), facilitation of mother-infant interactions (Jones and Sloan, 2018), processing of social value (Liu et al., 2019), empathy, and perceived attachment (Heinrichs and Domes, 2008; Bartz et al., 2011, 2015, 2019; Meyer-Lindenberg et al., 2011; Olff et al., 2013; Kanat et al., 2014). The specificity of OXT in social orientation was demonstrated in cooperative games, showing increased cooperation after OXT administration when social information was present but reduced cooperation when it was not (Declerck et al., 2010). In support of the view on empathy as a social perception trait, OXT receptors are most present in those brain structures related to sensory-informational processing and behavioural orientation, [i.e., in the central nucleus of the solitary tract (NTS), central amygdala (cAM), septal nuclei, HC, insula, and prefrontal (both, medial and orbital) and cingulate cortex (Huber et al., 2005; Karelina and Norman, 2009; De Dreu et al., 2010; Viviani et al., 2011; Knobloch et al., 2012; Boccia et al., 2013; Baribeau and Anagnostou, 2015; Mitre et al., 2016; Quintana et al., 2016; Sabihi et al., 2017)]. Behavioural orientation is the consideration of d.f. in behaviour and the choice of reinforcers that an individual would use as feedback during the construction of subsequent actions. The presence of OXT receptors in these brain structures make them responsive to the social/affiliative features of situations in addition to other information processing.

Brain NE, the monoamine neurotransmitter that is consistently linked to orientational aspects of behavioural regulation (especially attention to novelty), also has a way to regulate OXT. Indeed, OXT and VSP magnocellular neurons have direct projections from the NTS, the centre of sensory integration of the state of somatic systems, and 80% of NTS to SON projections are NE-gic, putting the OXT system right at the core of social perception (Karelina and Norman, 2009). These associations can explain why high-NEU people have a higher sensitivity to social approval/support, in comparison to low-NEU, indicative of a possible dysbalance between (insufficient) endogenous OXT production and (likely high) OXT receptor density. In terms of situational emotions, neuroticism as negative dispositional expectations of outcomes, including expectations of low social support, can induce higher readiness to admit guilt and develop shame. As noted, OXT functionality is consistently associated with orientational (socially relevant) aspects of behaviour.

Oxytocin and VSP often down-regulate each other to balance their effects (Legros, 2001) and show opposite action in the central AM: OXT calms HPA arousal by acting on GABA neurons in the cAM (Huber et al., 2005; Viviani et al., 2011; Knobloch et al., 2012; Quintana et al., 2016) and suppresses VSP-responsive neurons (Meyer-Lindenberg et al., 2011; Viviani et al., 2011; Liu et al., 2019; Grinevich and Neumann, 2021). VSP is also involved in social perception but as a support of self-image in a social context. It is likely that their differences in the regulation of the orientation to others vs. orientation to self were reinforced in evolution by coupling with gonadal hormones: OXT has mutually supportive relations with estrogen, whereas VSP has with testosterone. There are consistent reports that the locations of OXT and VSP receptors in brain areas do not overlap, suggestive of their different functionality (Stoop, 2012). Moreover, VSP increases alertness for external stimuli by activating the central AM, increases sympathetic output, and decreases parasympathetic output (Huber et al., 2005; Knobloch et al., 2012). This contrasts with the calming action of OXT on central AM, increasing parasympathetic and decreasing HPA arousal.

Oxytocin and VSP are not, however, always rivals, coordinating their common task of social perception. They use each other’s receptors (Song and Albers, 2018), and VSP was also linked to empathy (Tabak et al., 2015). This might explain why, despite being called the “love hormone,” intranasal OXT administration reportedly also produces “antisocial” effects, such as elicitation of ethnocentrism (Zhang et al., 2019), racial bias (Sheng et al., 2013), or increased dishonesty toward outgroup members (Radke and de Bruijn, 2012; Shalvi and De Dreu, 2014; De Dreu et al., 2016; Daughters et al., 2017).

There is a balancing interaction between social and gonadal hormones, so a pro- or anti-social orientation shouldn’t be attributed to the OXT–VSP pair alone. It has been shown that a testosterone–oxytocin imbalance can induce either empathy (and social sensitivity, when OXT release is more active than testosterone release) or autism (and self-centred perception, in the opposite case) (Montoya et al., 2012; Crespi et al., 2016; Procyshyn et al., 2020) and even aggression (van Anders et al., 2011). Variations in the OXT gene indeed were linked to antisociality in adolescent boys (Hovey et al., 2016); however, this sex-age group is also known for having high testosterone, low cortisol, and dopamine (Rosenblitt et al., 2001; Sinclair et al., 2014). Also, VSP is released more in male-typical behaviour, such as pair-bond formation and competition (seen in aggressive actions) (Heinrichs and Domes, 2008; de Vries et al., 2012), as well as protective aggression (Heinrichs and Domes, 2008; Kanat et al., 2014). Maternal aggression is also correlated positively with OXT release in PVB and cAM (de Vries et al., 2012).

The ability of central OXT to inhibit HPA axis activity appeared to be dependent on circulating estrogen (Ochedalski et al., 2007; van Anders et al., 2011; Gabor et al., 2012; McCall and Singer, 2012) whereas central VSP appeared to be more dependent on testosterone (van Anders et al., 2011; Gabor et al., 2012; McCall and Singer, 2012). In fact, the role of VSP in self-centred and competitive behaviour was found in males and not females (de Vries et al., 2012; Uzefovsky et al., 2012). The imbalanced ratio of high testosterone combined with low cortisol was also found to be a factor in socially aggressive behaviour (Popma et al., 2007; Terburg et al., 2009; Montoya et al., 2012; Ponzi et al., 2016; Knight et al., 2017; Grotzinger et al., 2018). Also, as highlighted by Crespi (2016), testosterone exhibits opposite effects from OXT on diverse aspects of cognition and behaviour, most generally by favouring self-oriented, egocentric and non-social attention and information processing. Too much OXT can be associated with mental disorders such as social psychosis and schizophrenia if this (high OXT) is combined with low testosterone (Dinsdale et al., 2013; Crespi, 2016; Dinsdale and Crespi, 2017); an opposite pattern (high testosterone and low OXT) is also associated with psychopathology (autism) (Crespi, 2016; Crespi et al., 2016).

Based on the evidence in neurochemistry and the work of other scholars, the FET views EMP as a trait of behavioural orientation rather than an emotionality trait as it describes preferential behavioural orientation and perceptual sensitivity to other people’s motivation and states. As summarised here, EMP is most likely associated with higher activity OXT, in comparison to VSP and testosterone, with the direct involvement of noradrenaline that likely helps to maintain attention to changeable social aspects of situations.

The analysis of the functionality of neurochemical systems can also benefit from an evolutionary perspective. Sex differences in SS and EMP might be a reflection of species’ self-regulatory systemic factors as outlined in the ETS developed in the 1960s by the mathematical biologist Geodakyan (see Trofimova, 2015 for review). Geodakyan conducted an analysis of sex differences in dispersion patterns, mutation rates, phenotypic and genotypic diversity, birth rates, mortality rates and, susceptibility to new diseases in various species and noted higher variability, higher mortality rates, and higher mutation rate in males in comparison to female phenotypes (Geodakyan, 2000; Trofimova, 2015). In his ETS, he suggested that the benefits of sexual reproduction as a method of reproduction won over other methods because sexual dimorphism allows a species to have two functional partitions:

The ETS, therefore, explains the paradox of sexual reproduction by systemic tendencies for species’ self-regulation. Trofimova pointed to the correspondence of sex differences in verbal-affiliative and exploratory abilities and disabilities with the ETS theory (Trofimova, 2015). Indeed, there is a well-documented superiority in risk- and sensation seeking, physical abilities, higher rates of psychopathy, dyslexia, autism, higher birth and accidental death rates is a reflection of the systemic variational function of the male sex. In contrast, there is female superiority in verbal abilities, lawfulness, socialisation, empathy, and agreeableness, presented as a reflection of the systemic conservational function of the female sex. Figure 1 highlights the sex differences in temperament traits of young males and females, and Figure 3 summarises the interplay between gonadal and social hormones leading to these differences. As could be seen from Figure 3, oxytocin and testosterone indeed act in opposite directions on VSP and the HPA axis (especially adrenaline). OXT pairs its action with estrogen and (not discussed here) serotonin, suppressing VSP, whereas testosterone and KOR activate VSP, which, in turn, activates the HPA axis and NE release (Zhao et al., 1988; Bernstein et al., 1996; Bodnar, 2018). Gonadal hormones, therefore, contribute to sex differences in social perception and self-image, as well as HPA axis arousal (especially the dynamic of adrenaline).

From this perspective, psychological sex differences in communicative and exploratory dis/abilities might not be just an accidental result of sexual selection or labour distribution in early humans. It might be a reflection of a global functional differentiation tendency within a species to expand its phenotypic diversity and at the same time to conserve beneficial features in the species’ behaviour (see Trofimova, 2015 for review).

In summary, the interactions between OR, hormonal and brain neurotransmitter systems in deriving a “emotional verdict” don’t mean that these systems are solely involved in the regulation of emotionality. Similarly, behavioural regulation (including orientation aspects) provided by brain neurotransmitters goes beyond emotional regulation, and, therefore, all of these systems should not be blended into one pot of “emotionality biomarkers.” Heart rate also changes during the intense emotional experience, but we don’t consider the cardiovascular system as a biomarker of emotionality. By analogy, the contribution of vision and visual processing is crucial in behavioural regulation, and it interacts with all other sensory modalities (auditory, proprioceptive, etc.). Yet, conceptually we identify these sensory modality systems as distinct. The same separation should be done with emotionality, and there are benefits in the conceptual differentiation between biomarkers of emotionality. For example, the analysis of contributions from these three of neurochemical systems in social emotions under the condition of tryptophan depletion demonstrated a potential for a more careful and transparent presentation of the interactions between these systems, including an explanation of paradoxical “loss of remorse and shame” in highly neurotic people under tryptophan depletion (Kanen et al., 2021).

Space doesn’t permit us to describe all possible scenarios of psychopathology related to the named systems. We included the notes on them throughout the text, illustrated OR-related psychopathology in the Figure 3, and included several useful references in the last column of Table 1. As Figure 3 shows, there is a “passive approval” case that is often goes under a radar of psychiatrists because people with this condition do not complain. Yet, as seen in our own clinical practice, they have limited functionality (not taking care of their meals, appearance, hygiene, social and personal life, employment) and, therefore, this condition should be included in future classifications of psychopathology. We also want to add that in psychiatric settings, clinical studies using a compact temperament test for screening for the 12 FET-associated temperament traits in people with psychiatric disorders showed the benefits of differentiating between traits related to emotional dispositions and non-emotionality aspects of behavioural regulation. This differentiation was in line with differential symptoms of major depression, generalised anxiety, and other psychiatric disorders (Trofimova and Christiansen, 2016; Trofimova and Sulis, 2016a,b, 2018).