Microbiome and immuno-metabolic dysregulation in patients with major depressive disorder with atypical clinical presentation

Depression is highly prevalent (6% 1-year prevalence) and is the second leading cause of disability worldwide. Available treatment options for depression are far from optimal, with response rates only around 50%. This is most likely related to a heterogeneous clinical presentation of major depression disorder (MDD), suggesting different manifestations of underlying pathophysiological mechanisms. Poorer treatment outcomes to first-line antidepressants were reported in MDD patients endorsing an "atypical" symptom profile that is characterized by preserved reactivity in mood, increased appetite, hypersomnia, a heavy sensation in the limbs, and interpersonal rejection sensitivity. In recent years, evidence has emerged that immunometabolic biological dysregulation is an important underlying pathophysiological mechanism in depression, which maps more consistently to atypical features. In the last few years human microbial residents have emerged as a key influencing variable associated with immunometabolic dysregulations in depression. The microbiome plays a critical role in the training and development of key components of the host's innate and adaptive immune systems, while the immune system orchestrates the maintenance of key features of the host-microbe symbiosis. Moreover, by being a metabolically active ecosystem commensal microbes may have a huge impact on signaling pathways, involved in underlying mechanisms leading to atypical depressive symptoms. In this review, we discuss the interplay between the microbiome and immunometabolic imbalance in the context of atypical depressive symptoms. Although research in this field is in its infancy, targeting biological determinants in more homogeneous clinical presentations of MDD may offer new avenues for the development of novel therapeutic strategies for treatment-resistant depression.

Depression is highly prevalent (6% 1-year prevalence) and is the second leading cause of disability worldwide. Available treatment options for depression are far from optimal, with response rates only around 50%. This is most likely related to a heterogeneous clinical presentation of major depression disorder (MDD), suggesting different manifestations of underlying pathophysiological mechanisms.
Poorer treatment outcomes to first-line antidepressants were reported in MDD patients endorsing an "atypical" symptom profile that is characterized by preserved reactivity in mood, increased appetite, hypersomnia, a heavy sensation in the limbs, and interpersonal rejection sensitivity. In recent years, evidence has emerged that immunometabolic biological dysregulation is an important underlying pathophysiological mechanism in depression, which maps more consistently to atypical features.
In the last few years human microbial residents have emerged as a key influencing variable associated with immunometabolic dysregulations in depression. The microbiome plays a critical role in the training and development of key components of the host's innate and adaptive immune systems, while the immune system orchestrates the maintenance of key features of the host-microbe symbiosis. Moreover, by being a metabolically active ecosystem commensal microbes may have a huge impact on signaling pathways, involved in underlying mechanisms leading to atypical depressive symptoms. In this review, we discuss the interplay between the microbiome and immunometabolic imbalance in the context of atypical depressive symptoms. Although research in this field is in its infancy, targeting biological determinants in more homogeneous clinical presentations of MDD may offer new avenues for the development of novel therapeutic strategies for treatment-resistant depression. Table 1 Glossary.

Introduction
Depression is a widespread condition with major personal and public health implications. Overall, nearly one in five people (15-18%) will experience a depressive episode in their lifetime, severely limiting psychosocial functioning and diminishing quality of life (Malhi and Mann, 2018). Moreover, a large body of evidence has documented a shortened life expectancy of about 7-17 years in major depressive disorder (MDD) patients (Korhonen et al., 2021;Laursen et al., 2016). Although suicide mortality is known to be particularly high in depression (Chesney et al., 2014), several studies revealed that natural causes of death account for the majority of premature deaths in MDD patients (Swaraj et al., 2019;Walker et al., 2015) (see Table 1).
Cardiovascular disease (CVD) is the leading cause of death in western societies (Lozano et al., 2012). Depression and obesity, both carrying an increased risk for CVD (Penninx et al., 2001), tend to co-occur within individuals (Blasco et al., 2020;Luppino et al., 2010). The relationship between both conditions is bidirectional: depression increases the risk of obesity and the related metabolic syndrome and vice versa (Patsalos et al., 2021;Schachter et al., 2018). This link between both conditions is particularly evident in patients showing an adverse metabolic profile (e. g., insulin resistance (IR), C-reactive protein (CRP) or dyslipidemia) (Jokela et al., 2014). There are good reasons to believe that these conditions are linked through a vicious cycle of adverse physiological adaptations (Milaneschi et al., 2019). Alterations in inflammatory and energy-regulating pathways may represent fundamental shared underlying biological mechanisms (Milaneschi et al., 2020). Indeed, a growing body of data suggests that chronic low-grade inflammation contributes substantially to the pathophysiology of depression and metabolic disorders (Berk et al., 2013;Kawai et al., 2021). In addition, neuroimaging evidence suggests a central role for specific brain regions in the regulation of body mass and energy homeostasis in obese patients that overlap with those involved in mood regulation (Finucane et al., 2015;Kroemer et al., 2022;Opel et al., 2017;Pigeyre et al., 2016).
Emerging evidence suggests that metabolic disturbances and chronic low-grade inflammation are more consistently associated with a symptom profile, frequently labeled as "atypical" depression (Lamers et al., 2016;Milaneschi et al., 2021;Penninx et al., 2013). Atypical depression (AD) shares many of the typical symptoms of MDD. However, unlike melancholic depression, which lacks mood reactivity, AD is characterized by improved mood in response to positive events. In addition, AD is often featured by increased appetite or weight gain, hypersomnia, a feeling of heaviness leading to difficulty moving the arms and legs, and a long-standing pattern of interpersonal rejection sensitivity, whereas classic definitions of MDD include lack of reactivity to pleasurable stimuli, loss of appetite, and insomnia (Diagnostic and statistical manual of mental disorders, 4th ed, 1994;Łojko and Rybakowski, 2017). The prevalence of patients diagnosed with AD according to the Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) varies from 15% to 30% (Diagnostic and statistical manual of mental disorders, 4th ed, 1994; Gold and Chrousos, 2002). Epidemiological studies revealed AD to be four times more common in females (Blanco et al., 2012), more often associated with early-life stress (Matza et al., 2003), poor treatment response to first-line antidepressants (Uher et al., 2014), younger age of onset (Rodgers et al., 2016), and a chronic course of the disorder (Gili et al., 2012;Stewart et al., 1993).
Microbiome research has evolved enormously during the past two decades and represents a new paradigm from which to approach many of the common diseases including depression and obesity (Butler et al., 2019;Cruz-Pereira et al., 2020;Maruvada et al., 2017). In the same way the gut-brain axis has become a focus in psychiatric research, the human oral microbiota is increasingly recognized as a new area that may promote novel diagnostic and therapeutic biomarkers for depression (Maitre et al., 2020;Simpson et al., 2020;Wingfield et al., 2021). Accumulating evidence identified human-associated microbes as a major player in the interplay between immune and energy dysregulation, linking atypical depressive symptoms to comorbid metabolic disorders. However, despite clear associations with immunometabolic dysregulation, the clinically derived atypical MDD subtype has been largely neglected in microbiome research.
In this review we focus on the underlying link between the microbiome, immunometabolic alterations, and clinical features of AD. First, we introduce the involvement of inflammatory processes in developing atypical depressive symptoms and related impaired metabolic status. We then discuss dysregulations in the energy metabolism linking both conditions. Furthermore, we review the role of the human microbiome in the pathophysiology of immunometabolic dysregulations in depression. Finally, we will summarize the implications of these potential biological links for the treatment of depression. This will provide an upto-date assessment of the current state of knowledge on immunometabolic dysfunction in MDD patients with a particular focus on the microbiome and raise questions that need to be answered by future research.

Heterogeneity of immuno-metabolic dysregulations associated with depressive symptoms
A large body of evidence suggests that dysregulated immunoinflammatory responses and imbalanced energy homeostasis play a role in the pathogenesis of not only cardiometabolic disorders, but also depression (Luppino et al., 2010;B. W. Penninx, 2017). However, despite convincing pathological linkages between MDD and metabolic disorders, the overall association effect sizes between higher biomarkers of immunometabolic pathways and depression were relatively small in the majority of meta-analyses (Milaneschi et al., 2020). This can be due to several factors: antidepressant treatment, environmental factors such as diet, sedentary lifestyle, smoking, sociodemographic indicators, and exposure to stressful events, as well as poor specificity of these markers since these are elevated in a plethora of common disease including psychiatric disorders (for review, see (Milaneschi et al., 2020)). However, the heterogeneity in the data is most likely primarily due to the heterogeneity of MDD itself. Thus, not all MDD patients exhibit immunometabolic dysregulations, as depicted below.
Clinicians are well aware that patients with the same diagnosis of MDD may endorse very different symptom profiles, potentially representing manifestations of different underlying pathophysiological

Tryptophan metabolism
Tryptophan is an essential amino acid that cannot be synthesized by the human body and must be obtained through dietary sources.Tryptophan metabolism begins with its uptake by cells in the body, primarily in the intestine and liver. Once inside the cells, tryptophan can be metabolized through two pathways: the kynurenine pathway and the serotonin pathway.In the kynurenine pathway, tryptophan is converted into a series of metabolites, including kynurenine, which can further be metabolized into quinolinic acid and kynurenic acid. These metabolites have a variety of functions in the body, including immune modulation and neurotransmitter regulation. In the serotonin pathway, tryptophan is converted via an intermediate into serotonin. Tryptophan metabolism is tightly regulated, and imbalances in the pathway can lead to a variety of health problems.

VSL#3
A probiotic dietary supplement that contains a combination of 8 different strains of live bacteria, including Bifidobacterium, Lactobacillus, and Streptococcus. It is marketed to support gut health and immune function, and has been studied for its potential benefits in various health conditions. processes (Milaneschi et al., 2019). An important line of research on depression heterogeneity has classically focused on the distinction between two major clinical subtypes, melancholic and atypical. An overview of the AD concept and criticism of its validity has been previously discussed elsewhere and is beyond the scope of the present review (Łojko and Rybakowski, 2017;Parker and Thase, 2007;Stewart et al., 1993). For the majority of immunometabolic markers such as abdominal obesity, circulating levels of glucose, insulin, leptin, and proinflammatory cytokines strong associations emerged only when contrasting control subjects with patients with an atypical symptom profile. When considering all patients with an overall MDD diagnosis or other symptom profiles, immunometabolic dysregulations were weaker and statistically non-significant (for review, see (Brydges et al., 2022;Milaneschi et al., 2020)). These findings are further emphasized by genetic studies demonstrating that only MDD patients with atypical features were strongly associated with a polygenic risk for increased body mass index (BMI) and high circulating levels of CRP and leptin (Milaneschi et al., 2017). It is noteworthy that different studies have applied the same label of "atypical" to subsets of MDD patients based on different definitions (Milaneschi et al., 2020). However, evidence suggests atypical depressive symptoms that reflect altered energy balance like hyperphagia and weight gain, hypersomnia, and fatigue to be the major drivers in the relation of AD and immunometabolic dysregulations (Milaneschi et al., 2020). This hypothesis is supported by a recent case-control study demonstrating higher levels of insulin, IR, leptin and ghrelin and inflammatory markers in MDD patients with hyperphagia compared to healthy controls, but not in MDD patients with a loss of appetite . In addition, new insights into functional neuroimaging revealed differences in functional connectivity of the nucleus accumbens to be associated with changes in appetite and body weight in MDD patients (Kroemer et al., 2022). Moreover, inflammation and oxidative stress have been linked to persistent fatigue (Lacourt et al., 2018). Indeed, increased circulating cytokines following immune activation in the context of infections are often accompanied by a combination of behavioral symptoms like fatigue, weakness, malaise and sleep and appetite alterations, referred to as sickness behavior, which is an important coping mechanism in the acute phase, but may evolve depressive syndromes in long-term (Capuron and Miller, 2011;Miller and Raison, 2016). Similarly, disruption of insulin signaling and resulting glycemic excursion leads to weakness and fatigue (Fritschi and Quinn, 2010). Taken together, these findings suggest that immunometabolic dysregulations are most likely to manifest in atypical depressive features related to energy homeostasis. As outlined in the following paragraphs, underlying immunometabolic dysregulations are closely linked to alterations in the microbiome. Because by far the most data in this regard are available on the gut microbiome, the term "microbiome" will be used throughout this article to refer to the gut unless another compartment is explicitly referenced.

Microbiome and the immune system in atypical features of MDD patients
Ever since Nobel laureate Julius Wagner-Jauregg observed psychopathological signs accompanying the onset and duration of malaria infection, we have known about a possible link between immune activation and psychiatric disorders (Cruz-Pereira et al., 2020). Nowadays, a variety of infectious diseases (e.g., gastroenteritis-related virus, influenza virus, herpes virus, Epstein-Barr virus, cytomegalievirus, and Borna disease virus) are known to be associated with symptoms resembling MDD Yirmiya et al., 2015).
Furthermore, it has long been noticed that resident microbes serve as an important first-line defense against invading pathogens. Thus, without a balanced composition of the microbiome, the risk of infection with pathogenic microbes increases, and their rapid proliferation can cause severe local damage due to the subsequent immune response. In addition, many other aspects of host physiology can be affected by pathogen invasion, including important brain processes, such as neuroinflammation, hypothalamic-pituitary-adrenal (HPA) axis activation, neurotransmission and neurogenesis, which may manifest in MDD, particularly with an atypical profile (Anderson et al., 2017).

Bidirectional relationship between the development of the immune system and the microbiome in the pathogenesis of depressive symptoms
Human-associated microbes are vital for the early development of immune cells and immune system regulation. By constantly responding to different structural components of bacterial cells at the intestinal epithelium, a dynamic physiological low-grade inflammatory tone is maintained in homeostatic conditions, continually communicating with the brain, and enabling the immune system to mature (Cruz-Pereira et al., 2020). Thus, in germ-free (GF) mice, which are raised under sterile conditions or are depleted of their intestinal microbiota with oral broad-spectrum antibiotics, certain toll-like receptors (TLRs) are not fully expressed (Grasa et al., 2015), suggesting certain host-specific bacterial species need to be present for complete immune system development (Butler et al., 2019;Chung et al., 2012). Moreover, the gut microbiota influences the relative populations, migration, and functions of various subsets of immune cells, including Th cells and regulatory T cells (Dorrestein et al., 2014;Rooks and Garrett, 2016). This is of particular interest since MDD patients showed impaired differentiation of Th subsets, less diverse T-cell-receptor repertoire, and impaired T-cell response compared to matched non-depressed controls (Grosse et al., 2016;Patas et al., 2018). More specifically, MDD patients showed an increased Th1/Th2 ratio (Kubera et al., 2001) and elevated blood levels of Th17 cells with the highest levels of Th17 occurring in patients at high risk of suicide (Schiweck et al., 2020). Recently, intraepithelial lymphocytes were shown to require the presence of Lactobacillus reuteri in combination with a tryptophan-rich diet to reprogram Th cells into regulatory T cells (T reg), emphasizing an interaction between intestinal microbes and the adaptive immune system, which might play a role in the pathogenesis of depression (Cervantes- Barragan et al., 2017). Moreover, Duscha and colleagues found a significant and sustained increase of T reg cells after 2 weeks of propionate intake, as a principal short-chain fatty acid (SCFA) produced by the gut microbiota, whereas Th1 and Th17 cells decreased significantly (Duscha et al., 2020). Interestingly, a similar pattern of microbial changes involving depletion of anti-inflammatory SCFA-producing bacteria and enrichment of pro-inflammatory bacteria has been associated with a variety of mental disorders, suggesting transdiagnostic signals rather than a specific feature of depression (McGuinness et al., 2022;Nikolova et al., 2021). However, clinical subtyping of depression has been largely neglected in microbiome research. As immunometabolic changes have been shown to be much more consistent in the atypical subtype than in other mental disorders, and are closely linked to microbial changes, one might speculate that there are indeed specific microbial perturbations for this subtype.
Early life appears to be a critical window during which the immune system is trained and immune cells learn to tolerate commensal microbiota (Jain, 2020). In turn, the immune system has developed specific mechanisms to ensure that the commensal bacterial load is maintained (Cruz-Pereira et al., 2020;Zheng et al., 2020). Childhood maltreatment (CM), defined as sexual, physical and/or emotional abuse or neglect, has been shown to be an important risk factor for MDD (Humphreys et al., 2020). Furthermore, CM is associated with the course of MDD in terms of earlier onset and increased rates of non-responders, as seen primarily in MDD patients with an atypical profile (Klumparendt et al., 2019;Withers et al., 2013). Consistent with this, evidence suggests that maternal and early life stress not only increases the likelihood of developing obesity and depressive symptoms throughout life, but also increases the diversity of Clostridium genera, which are generally associated with inflammation (Zijlmans et al., 2015). Thus, it could be speculated, that these highly interrelated alterations in the microbiome composition, immune system and metabolic status, observed in both conditions, MDD and obesity, may already arise during early development (see Fig. 1).
Most fundamental events in the formation of host immunity may occur in parallel with critical periods of neurodevelopment during the first years of life (Cenit et al., 2017;Gensollen et al., 2016), in which microbiota composition displays the highest intra-and interindividual variability before reaching a more stable adult-like configuration at the age of ~3 years (Yatsunenko et al., 2012;Zheng et al., 2020). Perturbed crosstalk between the microbiome and the immune system at this time may have long-lasting impacts on multiple immune domains contributing to immune homeostasis, susceptibility to infectious disease and impaired inflammatory response as seen in the atypical depressive subtype and related metabolic disease (Adams et al., 2008;Andersson et al., 2016;Ghilotti et al., 2019;Kaspersen et al., 2015).

Microbiome and systemic low-grade inflammation
Chronic low-grade inflammation, which is a hallmark of numerous chronic conditions such as metabolic disorders (Hotamisligil, 2006), is defined by the persistent presence of elevated levels of circulating pro-inflammatory cytokines (e.g. interleukin (IL)-6, tumor necrosis factor (TNF), IL-1β), and acute phase proteins (e.g. C-reactive protein (CRP)) (Ayyadurai et al., 2022). Common signs and symptoms that develop during chronic inflammation include depressive mood and chronic fatigue. Although there is large heterogeneity in the data, it is now well-established in multiple meta-analyses that pro-inflammatory markers are increased in MDD patients with a fairly unanimous consensus of increased blood levels in IL-6, TNF and CRP compared to healthy controls (Goldsmith et al., 2016;Köhler et al., 2017;Liu et al., 2012). However, these alterations are more specific for MDD patients with an atypical profile, suggesting chronic low-grade inflammation to represent a fundamental mechanism underlying AD and the high reciprocal comorbidity with metabolic diseases (B. W. J. H. . Several mechanisms have been proposed that contribute to inflammation in AD, including poor diet, which is known to activate toll-like receptor 4 (TLR4) (Figueroa-Hall et al., 2020), decreased vagal modulation (C.-H. , mitochondrial dysfunction (Hoffmann et al., 2019), and chronic stress leading to HPA axis dysregulations (Perrin et al., 2019). The human microbiome has been proven to be fundamentally involved in these and other inflammation-inducing pathways (Vetrani et al., 2022). Thus, the gut microbes were revealed as a crucial factor in the regulation of colonic mucus secretion and the homeostasis of the intestinal epithelium, which is vital in maintaining a selectively permeable barrier between the gut lumen and systemic circulation (Pearson and Brownlee, 2010). Increased permeability of the gut barrier results in the translocation of gut bacteria or bacterial components such as lipopolysaccharides (LPS), which are typically safely confined to the gut lumen, triggering a systemic inflammatory response (Butler et al., 2019). Both, dysbiosis, defining an imbalance in the microbial community, and chronic stress promote a "leaky gut" and therefore may present a source of chronic low-grade inflammation (Ilchmann-Diounou and Menard, 2020;Lobionda et al., 2019;Stan et al., 2020). Another way in which microbes induce activation of the host immune system is through Nod-like receptors (NLRs), which respond to microbiota-derived metabolites and induce the inflammasome pathway (Alcocer-Gómez et al., 2017;Levy et al., 2015). Beyond that, principal microbial metabolites like SCFAs and secondary bile acids (BAs) exert other key functions in maintaining intestinal integrity and immune homeostasis. SCFAs mainly result from microbial fermentation of host-indigestible dietary fibers and consist of more than 95% of acetate, propionate and butyrate. In particular butyrate and to a lesser extent acetate are the main substrates for de novo lipid metabolism in colonic epithelial cells, to which it serves as the main energy source, enhancing intestinal integrity (Salvi and Cowles, 2021;Zambell et al., 2003). Accordingly, long-term consumption of a Western-style high-fat diet (HFD), which is low in fiber, leads to systemic inflammation most likely

Fig. 1. Early life stress in the link between depression and related cardiometabolic disorders.
Early life stress is a known risk factor for MDD and related metabolic disorders. Given that MDD patients with atypical features tend to be younger, there may be a strong interrelationship between the development and maturation of the microbiome, immune system, and central nervous system that predispose to the manifestation of AD. The HPA axis, which plays a key role in regulating the body's response to stress, and the immune system are closely linked through a complex network of interactions. Neurodevelopment refers to the processes by which the brain develops and matures, including the formation of neural circuits responsible for various functions such as appetite regulation. Research has shown that early life exposures, such as poor nutrition or exposure to toxins, can affect neurodevelopment and lead to neuroendocrine dysregulation associated with altered hunger and satiety signaling. The human microbiome plays an important role in neurodevelopment, including the production of neurotransmitters and other neuroactive compounds, modulation of immune function, and regulation of inflammation. Disruptions in neurodevelopment can lead to alterations in the function of these circuits, which can affect hunger and satiety signaling and, in turn, promote a chronic inflammatory response and neurodevelopment. due to reduced production of SCFAs (Mathewson et al., 2016), providing another crosslink between immunological and metabolic dysregulations as a common biological correlate of AD and comorbid CVD. Apart from that, evidence suggests SCFAs to suppress cytokine production from myeloid cells and to regulate cell differentiation of Th cells as well as T reg cells, suggesting an overall anti-inflammatory SCFA-mediated influence on the immune system.
BA metabolism, on the other hand, depends on deconjugation and subsequent transformation to secondary BAs by gut microbes. The interplay between BAs and the microbes that metabolize them is complex, as secondary BAs also have potent antimicrobial activity (Begley et al., 2005), suggesting that they are an important regulator of microbial composition. Thus, reduced amounts of secondary BAs may led to intestinal bacteria overgrowth and the disruption of the epithelium, contributing to chronic low-grade inflammation (Butler et al., 2019). Apart from that, BAs have been demonstrated to modulate gut microbial composition by activating genes of the innate immune system in the small intestine (Wahlström et al., 2016).

The oral-microbiome-brain-axis
While the gut microbiome has long been the focus of research, other compartments are now increasingly attracting interest. The oral microbiome, for instance, is one of the most abundant in the human body, second only to the gut (Peng et al., 2022). With regard to immune dysregulations in neuropsychiatric disorders and, in particular, immune-inflammatory processes that are strongly associated with AD, the oral microbiome is of special interest. On the one hand, among patients with severe mental illness, dental caries and periodontal measurement indexes often reach twice the level found in the general population (Gurbuz et al., 2010;Ramon et al., 2003). On the other hand, unlike the vessels from the intestinal tract, the vessels from the oral cavity do not first pass through the liver, where microbes reaching the circulation could already be colonizing or microbial-derived bioproducts may be metabolized to some extent, but instead, continue directly to the brain.
Commensal microbial species are usually site-specific in the host body, performing beneficial functions without causing harm. However, when microbes escape from their niche and travel to other locations, they can cause disease elsewhere in the body by disrupting bioprocesses and secreting bioproducts, that can be toxic in some conditions (Bowland and Weyrich, 2022). A striking example is infectious endocarditis caused by Streptococcus viridans, which is part of the commensal oral microbiome, entering the bloodstream as a result of hemorrhagic oral treatment, such as tooth extraction (Birlutiu et al., 2018).
Although alterations in the oral microbiota are mainly associated with local disease, most commonly caries and periodontal disease (PD) (Gao et al., 2018), the latter may also represent a risk for other conditions (Falcao and Bullón, 2019), including depression with an atypical profile (Maitre et al., 2020;Mohammadi et al., 2019;Wingfield et al., 2021) and related cardiometabolic disease (for review, see (Gasmi Benahmed et al., 2021)). In PD, the inflamed periodontal attachment apparatus shows increased permeability of the gingival epithelium, allowing microbes, endotoxins and inflammatory cytokines to infiltrate into the systemic blood circulation eliciting a systemic immune activation (Kamer et al., 2008). PD has been linked to increased LPS levels in the brain, indicating an accompanying neuro-inflammatory response (Xue et al., 2020). In rodents, the collagen-binding activity of Streptococcus mutans, which is an oral microbe associated with dental carries, can leave the oral cavity in the blood and induce cerebral hemorrhaging by disrupting the blood-brain barrier (BBB) (Hosoki et al., 2020;Watanabe et al., 2016). Similarly, Porphyromonas gingivalis, which exists in many periodontally healthy people (Lassalle et al., 2018) but can be a key pathogen in the development and progression of periodontitis, makes its way to the brain via the bloodstream, where it colonizes and releases neurotoxic proteases called gingipains (Bulgart et al., 2020).
Taken together, these data suggest that commensal microbes in the oral cavity can affect brain function after entering systemic blood circulation that may contribute to the development of atypical depressive symptoms (Fig. 2).

Microbiome-immune communication in the onset of atypical depressive features
Neuroinflammation is increasingly recognized as a key factor interacting with neurobiological correlates of MDD like depletion of brain serotonin, dysregulation of the HPA axis, and alteration of hippocampal neurogenesis and plasticity (for review, see (Troubat et al., 2021)). General symptoms accompanying neuroinflammation are sleepiness, widespread pain and malaise (DiSabato et al., 2016). Intriguingly, neuroimaging studies revealed regional specificity of neuroinflammation, which shows striking overlaps with neural patterns activated in atypical depression but not in the melancholic subtype (for review, see (Woelfer et al., 2019)). These include core regions involved in reward circuitry that may contribute to the preserved ability to experience pleasure from positive events in MDD patients with an atypical profile, but the overall response is still low.
Brain tissue was originally thought to be immune privileged, protected from any invasion by pathogens and somewhat isolated from the peripheral immune system (Cruz-Pereira et al., 2020). With the discovery of the disruption of the BBB, it is only recently that researchers are starting to tease apart the contribution of peripheral and central inflammation in depressive-like behaviors (Engelhardt et al., 2017). Activation of cytokine signaling in BBB endothelial cells mediates recruitment of circulating leukocytes and in some cases induces disruption of tight junctions, resulting in a leaky, permeable BBB . Thus, blocking the peripheral cytokines TNF and IL-6 have been shown to be sufficient to tighten the BBB, which in turn has been sufficient to rescue learned helplessness in mice (Cheng et al., 2018;Menard et al., 2017).
In addition to influencing systemic immune response at the level of the gut, a large body of evidence from mouse models of multiple sclerosis and induced ischemic brain injury indicate a compelling role for the gut microbiome in the mediation of neuro-inflammatory processes (for review, see (Fung et al., 2017)). Evidence suggests that SCFA acts as a natural ligand for the free fatty acid receptor (FFAR), which are widely expressed on peripheral lymphocytes (Nilsson et al., 2003), to regulate the recruitment of trafficking monocytes from the periphery to the brain . Beyond that, microglial maturation and function are influenced by the microbiota. Compared to conventionally colonized controls, GF mice have been demonstrated to exhibit increased numbers of immature microglia, altered microglia morphology (with longer processes and more branching) and compromised immune response to bacterial or viral infection (Erny et al., 2015), which are normalized by postnatal supplementation with SCFA (Borre et al., 2014).
Mechanisms linking inflammatory pathways to synaptic activity also include the modulation of pro-inflammatory cytokines on monoaminergic and glutamatergic transmission as seen in MDD patients with an atypical profile (for reviews see (Sen et al., 2021;Woelfer et al., 2019)). Again, the microbiome is fundamentally involved in these pathways by affecting the host's metabolism at various levels. While some metabolites cross the BBB and directly trigger relevant pathways, others elicit a response in the periphery impacting relevant circuits in the brain. Tryptophan/serotonin and carbohydrate metabolism are remarkable examples (Yano et al., 2015). More than 90% of the body's serotonin is synthesized in the gut. Following absorption from the gut, the precursor molecule tryptophan enters the circulation and travels to the brain, where it is metabolized by two non-protein metabolic pathways: methoxyindoles and kynurenine. Tryptophan availability is the major rate-limiting factor of the methoxyindoles pathway leading to serotonin biosynthesis, as less than 5% of tryptophan is metabolized via this pathway (Gál and Sherman, 1980). The indoleamine 2,3-dioxygenase (IDO), which is responsible for the first step of the kynurenine pathway, is induced in macrophages, dendritic cells and microglia by signaling from pro-inflammatory cytokines, as well as psychological stress or glucocorticoids (Kiank et al., 2010). Activation of IDO leads to reduced tryptophan levels for the synthesis of serotonin in favor of increased kynurenine levels, which has been hypothesized to promote atypical depressive symptoms (Brandacher et al., 2007;Favennec et al., 2015;Savitz, 2017). In line with this, it has been suggested that gut microbes manipulate our eating behavior by generating cravings for specific foods, including carbohydrates, in order to improve their fitness and thus resist selection and evolutionary pressures (Alcock et al., 2014;Anderson et al., 2017). In addition, the microbiome is a critical component of digestion, breaking down complex macronutrients, such as carbohydrates (Oliphant and Allen-Vercoe, 2019). Some studies suggest that the proportion of carbohydrate in total energy intake leads to an increase in tryptophan uptake and synthesis in the brain, reducing the risk of depression (Fernstrom and Wurtman, 1972; "The protective effect of relative carbohydrate intake on depression," 2022). Intriguingly, the severity of depressed mood was associated with an increase in carbohydrate craving in MDD patients with an atypical profile (Brzezinski et al., 1990). This link and the evidence that treatments effective for one of the symptoms also benefit the other, suggests that the two symptoms are different manifestations of the same underlying pathophysiology (Møller, 1992). Unbalanced carbohydrate meals, however, often induce fatigue (Spring et al., 1987). As some patients may turn to food as a way of coping with their emotions, they further exacerbate microbial, inflammatory, metabolic and hormonal changes that underlie the syndrome. Thus, carbohydrate craving in depression appears to be both a compensatory mechanism and part of the underlying and perpetuating factors of the disorder.
Beyond that, there is a complex and multifaceted interaction between inflammation and increased HPA axis activation, which is one of the most enduring and replicated findings in biological psychiatry, at least in a subset of patients with MDD (Keller et al., 2017), and it has rapidly become clear that the microbiome plays a major role in the development and regulation of the HPA axis (for review, see (Sudo, 2014)). In line with this, Zhu et al. recently demonstrated that stress resistance is mediated by Lactobacillus species that are involved in colonic T cell differentiation (Zhu et al., 2023). Lamers and colleagues suggested two different core biological correlates: an overactive HPA axis in melancholic depression and inflammatory/metabolic dysregulation in AD . However, the findings are variable and sometimes contradictory, which may be partly due to differences in methodology. Another important factor is the very heterogeneous clinical presentation of depression, with different and sometimes contradictory symptoms, which is insufficiently compensated for by classification into more homogeneous subtypes. Thus, preserved mood reactivity is the most controversial symptom in the diagnosis of AD, which is also the only obligatory one in DSM IV. In addition, most people have mixed features of melancholic and atypical depression, and most people with long-term depressive illness experience both melancholic and atypical episodes over time (Blanco et al., 2012). HPA hyperactivity, which has been reported in acute melancholic and severe forms of depression with psychotic symptoms (Juruena et al., 2018), may adapt in the long term with altered dynamics (O'Keane et al., 2012). Recently, Iob et al. investigated hair cortisol and plasma CRP with the longitudinal persistence and dimensions of depressive symptoms over a 14-year period. Their results suggest that higher levels of cortisol and CRP are associated with persistent depressive symptoms with a more somatic profile (Iob et al., 2020). These findings are bolstered by gene expression analyses demonstrating glucocorticoid treatment up-regulated several innate immune genes (Horowitz et al., 2020). Thus, stress may prime the immune system for dysfunctional inflammation suggesting synergistic effects in the pathogenesis of depression.
Moreover, stress contributes to a shift of the sympatho-vagal balance in favor of sympathetic modulation, which has been extensively described in depression (Sgoifo et al., 2015). Stimulation of the vagus nerve, which collects a wealth of visceral and inflammatory information from the periphery and is massively influenced by the microbiome as a major signaling pathway of the gut-brain axis, has been shown to have anti-inflammatory effects. (Borovikova et al., 2000).

Dysbiosis of the oral microbiome significantly contributes to the development of periodontal diseases (PD). In PD, the inflamed periodontal attachment apparatus shows increased permeability of the gingival epithelium, which can cause systemic low-grade inflammation. Conversely, successful local periodontal treatment attenuates systemic inflammatory markers. Moreover, hematogenous dissemination of periodontal bacteria or spillover of inflammatory mediators can disrupt the endothelial barrier.
Due to the close proximity, increased vascular endothelial permeability may facilitate the passage of pathogenic microbes and their metabolites across the blood-brain barrier, triggering a neuroinflammatory response.
Besides inflammatory and stress-modulating properties, microbialimmune interactions also play an important role in regulating fetal and adult neurogenesis. Bacterial wall components, crossing the maternal-fetal interface, have been proven to be necessary to activate TLRs, leading to fetal cortical neuroproliferation (Humann et al., 2016), providing further evidence that host-immune interactions during a critical time window in early life may contribute to the pathogenesis of atypical depressive features (for review, see (Suh et al., 2019)) (see also Fig. 1). For instance, impaired hippocampal neurogenesis is a hallmark of clinical depression (Eisch and Petrik, 2012). In that line, antidepressant medication and interventions stimulate adult hippocampal neurogenesis, which in turn dampens stress response (Culig et al., 2017;Santarelli et al., 2003). Brain-derived neurotrophic factor (BDNF), which plays a critical role in the generation of hippocampal neurons in adulthood, were reduced in the hippocampus of GF mice (Bercik et al., 2011;Clarke et al., 2013;Sudo et al., 2004). Such alterations suggest certain gut microbiota to modulate hippocampal neuronal plasticity (Andero et al., 2014). Interestingly, Ogbonnaya and colleagues showed increased adult hippocampal neurogenesis in GF mice, which could not be prevented by microbial colonization during early life, supporting the thesis of a critical window during which the microbiome affects brain function (Ogbonnaya et al., 2015) (Fig. 1). In another study antibiotic treatment decreased hippocampal neurogenesis, which can be restored by probiotic treatment (Möhle et al., 2016). Recently, Kundu et al. have demonstrated increased neurogenesis in the hippocampus of a rodent model, which was associated with an age-sensitive enrichment of butyrate-producing microbes (Kundu et al., 2019). In addition, microbiota transplantation from stressed donors to naive mice has been shown to reduce adult neurogenesis and transmit depressive behavioral symptoms to recipient mice (Chevalier et al., 2020), strengthening the link between the microbiome, maladaptive responses to stress, hippocampal neurogenesis, and the development of depressive symptoms.

Microbiome brain signaling and energy homeostasis in the development of atypical depressive features
Efficient immune responses and effective calorie-storing systems evolved to aid individuals in surviving infections and periods of starvation. Additionally, an organized suite of behavioral changes in response to infection evolved that is called sickness behavior. Sickness behavior includes behavior resembling symptoms of AD, such as social withdrawal, sleepiness and inactivity. Social withdrawal of ill individuals protects the social group from infected individuals by limiting their direct contacts, preventing them from contaminating the environment and broadcasting their health status (Shakhar and Shakhar, 2015). Moreover, sickness behavior allows animals to conserve body resources for the high energetic costs of fever in fighting infection. Thereby, sickness behavior represents another link between an immunoinflammatory response, energy homeostasis, and the onset of depressive symptoms, particularly with an atypical profile.
Beyond the immunoinflammatory links described above, there is a large body of evidence suggesting that the gut microbiome is an important regulator of systemic metabolic homeostasis as a second obvious factor underlying atypical depressive symptoms and related metabolic disorders (see Fig. 3). This is supported by the fact that many conditions wherein food intake behavior is dysregulated (e.g. anorexia nervosa or obesity), are associated with alterations in the gut microbiome (Kleiman et al., 2015;Mack et al., 2016;Turnbaugh et al., 2006). However, the microbiome's definite role in metabolic dysregulations of the host remains elusive, since the key drivers of dysregulated metabolism are also believed to be the key drivers of the microbiome composition: diet and lifestyle (Dabke et al., 2019). Thus, while it is intuitive that the microbiome and host energy metabolism are somehow related, disentangling cause and effect remains a challenge.

Microbiome and endocrine dysregulations associated with atypical depressive symptoms
Dysregulated energy homeostasis is most likely a major contributor to key features of AD like physical and mental exhaustion and hyperphagia (Lamers et al., 2010). There are several neuroendocrine pathways that may be involved in impaired energy metabolism as seen in atypical depression and metabolic disorders (for review, see (Goldstein et al., 2017). Moreover, a plethora of studies has documented decreased diversity and reduced richness in the species to be correlated with dysregulated energy balance (Vallianou et al., 2021). Thus, GF mice have been demonstrated to be largely resistant to HFD-induced obesity, excreting more lipids in feces and have altered cholesterol levels (Rabot et al., 2010). Apart from directly modulating energy intake through the digestion of complex macronutrients, for which the human gut does not have the complete enzymatic capacity (Liu et al., 2018), the microbiome indirectly contributes to energy homeostasis by affecting hunger and satiety signaling via interactions with the gut-brain axis. For instance, microbial-derived metabolites such as SCFAs (Fernandes et al., 2014;Psichas et al., 2015) or caseinolytic protease B (Breton et al., 2016) are involved in the leptin-melanocortin pathway as a key neuroendocrine regulator of hunger and satiety. Dysregulation of this pathway may lead to atypical depressive features such as fatigue and overweight (Baldini and Phelan, 2019). Overeating and lack of physical activity are also linked to leaden paralysis (Ohayon and Roberts, 2015). In rodents, deletion of the leptin receptor results in depressive behavior (Guo et al., 2013) and chronic stress has been found to reduce circulating leptin levels associated with depressive behaviors (Lu et al., 2006). Moreover, peripheral and central administration of leptin produces antidepressant-like effects in animal models (Garza et al., 2012;Yamada et al., 2011). In humans, leptin concentration was strongly associated with clinical symptoms of AD, such as hyperphagia, increased weight and leaden paralysis (Gecici et al., 2005). However, results from observational studies were hampered by small sample sizes and clinical heterogeneity (Milaneschi et al., 2019).
Another potential link in the microbiome-gut-brain communication influencing both, AD and metabolic disorders, lies in the role of gutderived peptides, which act as signaling molecules and are mostly involved in the regulation of digestion and satiety. However, recent findings demonstrate gut hormones and endocrine peptides, including neuropeptide Y, peptide YY, pancreatic polypeptide, cholecystokinin, glucagon-like peptide (GLP), corticotropin-releasing factor, oxytocin and ghrelin, to be also relevant in stress-related psychiatric conditions, such as depression (for review, see (Lach et al., 2018)). Evidence is accumulating suggesting that specific gastrointestinal microbiota and their metabolites are involved in the signaling cascades of these gastrointestinal peptides. For example, the gut microbiota has been reported to influence ghrelin receptor (growth hormone secretagogue receptor) signaling through their metabolites and to regulate circulating ghrelin levels by modulating ghrelin secretion from enteroendocrine cells (for review, see (Leeuwendaal et al., 2021)). Ghrelin, an orexigenic agent synthesized in the enteroendocrine cells (ECC) of the stomach (Dixit et al., 2004), can be considered the counterpart of leptin. Thus, leptin and ghrelin homeostasis may be linked (Brennan and Mantzoros, 2006;Cui et al., 2017;Inui, 2001). By exhibiting neuroprotective properties (Frago et al., 2011) and potentially increasing hippocampal neurogenesis (Abizaid and Anisman, 2014), an increase in ghrelin may protect against depressive outcomes (Lutter et al., 2008).
Finally, evidence points to a stronger association between obesity and MDD in females , which is consistent with the overrepresentation of females in atypical depression (Schuch et al., 2014). Thus, women are at twice the risk for anxiety and MDD as men are, more likely experiencing disturbances of sleep appetite and energy (Kuehner, 2017). While socioeconomic factors and differential trauma exposure are likely contributors (Li and Graham, 2017), a strong body of evidence points to sex hormone fluctuations in women as the primary biological factor driving sex differences not only in depression risk (Altemus et al., 2014;Kundakovic and Rocks, 2022), but also obesity (Cooper et al., 2021;Lovejoy and Sainsbury, 2009). Estrogen, which has been independently linked to changes in the microbiome, especially during menopause (Bridgewater et al., 2017;Graham et al., 2021;Shin et al., 2019), is a potent neuromodulator and has been shown to affect various neurotransmitter systems in the brain, such as serotonergic, noradrenergic, GABAergic, dopaminergic, and glutamatergic systems (Acharya et al., 2023;Barth et al., 2016). Conversely, certain gut bacteria can regulate the expression of estrogen receptors and alter estrogen levels. For example, some bacterial strains can convert estrone to estradiol (for review see (He et al., 2021)).

Bidirectional relationship between mitochondria and the microbiome in depressive symptoms
Recent evidence suggests that mitochondria, which are abundant in neuronal dendrites and synaptic terminals, play a role in many intracellular processes related to synaptic plasticity and cellular resilience, which have been linked to depression (Bansal and Kuhad, 2016). Given that the brain consumes 20% of the body's energy for the average adult in resting state (Watts et al., 2018), it is highly vulnerable to conditions related to impaired energy production. ATP is necessary to attend the energy demands for the activation of downstream signaling following the binding of neurotransmitters to receptors, as well as for vesicle transport and neurotransmitter release (Devine and Kittler, 2018). In addition, mitochondria act as an important mediator of multiple signaling pathways, including those linked to inflammation (Hoffmann et al., 2019). Higher rates of mitochondrial biogenesis are needed for neuronal differentiation, suggesting dysfunctional mitochondria to contribute to impaired neuroplasticity in depression (for review, see (Allen et al., 2018)). Moreover, the brain is particularly vulnerable to the effects of reactive oxygen species (ROS) and free radicals.
On the other hand, it is evident that oxidative stress and mitochondrial dysfunction are also involved in the pathophysiology of overweight, hyperphagia and IR (Meigs et al., 2007;Petersen et al., 2003). Thus, alterations in mitochondrial metabolism in hypothalamic neurons have also been linked to the development of leptin resistance and IR (Kleinridders et al., 2018). Moreover, mitochondrial dysfunction and dysregulation in energy-regulating neuroendocrine metabolic pathways, may contribute to overlapping brain structural alterations in obesity and neuropsychiatric disorders with the most pronounced alterations being lower cortical thickness in the frontal and temporal cortex, which has been consistently shown by previous large-scale neuroimaging research (McWhinney et al., 2022;Opel et al., 2021). It can be speculated that the microbiome might contribute to oxidative stress conditions, since a study in chronic kidney disease demonstrated that microbes release uremic toxins, adding to oxidative stress and inflammation in kidneys (Vaziri et al., 2013). Both, the microbiome and mitochondria, which have descended from the primitive α-Proteobacteria, are maternally inherited. The gut microbiome is responsible for a large proportion of bioactive molecules like SCFAs, vitamins and amino acids that are necessary for the interlinked pathways of glycolysis, tricarboxylic acid cycle (TCA) cycle, and mitochondrial oxidative phosphorylation (OXPHOS). Additionally, microbial-derived metabolites have been shown to regulate key transcriptional co-activators, transcription factors and enzymes involved in mitochondrial biogenesis (Bergeron et al., 2001). By increasing the AMP/ATP ratio, microbial-derived SCFAs are known to directly activate the AMP-activated protein kinase (AMPK), which is a major sensor and regulator of energy homeostasis (den Besten et al., 2015). Alterations in the microbiome composition may lead to decreased SCFA levels, potentially contributing to the pathogenesis of energy-related atypical symptoms. Via inhibition of the mTOR complex 1 pathway AMPK is also involved in regulating inflammatory responses bridging both major underlying substrates of AD and metabolic disorders (Milaneschi et al., 2020;Mizushima et al., 2010). Apart from that, it has been reported that commensal bacteria-induced ROS production can affect regulatory Fig. 3. Vicious circle representing underlying mechanisms that can lead to atypical depressive symptoms. On top: Environmental factors affecting the human microbiome, impaired immune response and imbalanced energy homeostasis contribute to the manifestation of atypical depressive symptoms that further exacerbate disease-related behaviors. In addition, there is a bidirectional relationship between atypical features and disease-related behaviors that contribute to the biological changes in these patients. Central panel: Dysbiosis promotes increased intestinal epithelial permeability, contributing to chronic lowgrade inflammation as a characteristic underlying mechanism linking atypical depression and metabolic disorders. Left panel: Bidirectional relationship between immune system development and the microbiome. Dysregulation of the immune system supports neuroinflammation and chronic low-grade inflammation, contributing to the development and maintenance of atypical depressive symptoms and further reinforcing dysbiosis. Right panel: Chronic low-grade inflammation also contributes to leptin and insulin resistance and oxidative stress, further compromising energy homeostasis. Human microbes modulate energy intake directly by supporting the digestion of complex macronutrients and indirectly by influencing hunger/satiety signaling via the gut-brain axis. Moreover, there is a high reciprocal relationship between the microbiome and mitochondria that may contribute to the development of atypical depressive symptoms. redox sensor proteins, which are essential to maintain gut epithelial barrier function and promote anti-inflammatory IL-10 secretion (Saint-Georges-Chaumet & Edeas, 2016). In turn, genetic variation in the mitochondrial genotype ATP8 synthase gene was demonstrated to result in an increased Firmicutes-to-Bacteriodes ratio in mice (Clark and Mach, 2017;Hirose et al., 2017). Moreover, evidence suggests mitochondrial DNA damage, which accumulates throughout life, to modulate ROS production in the host, leading to decreased diversity of the gut microbiome (Yardeni et al., 2019). Together these findings indicate a fine balance of ROS concentration in the cell to be critical in maintaining microbiome composition and to be not only an important determinant of energy homeostasis but also of the immune response promoting key features of AD.

Dealing with immune-metabolic depression
MDD patients with an atypical symptom profile have been demonstrated to be more likely treatment-resistant to antidepressive medication and to show a chronic course of the disease (Brailean et al., 2020). Treatment strategies targeting the common mechanisms described in this review may be beneficial for both ameliorating AD and preventing cardiovascular mortality in these patients Milaneschi et al., 2019). For instance, lifestyle interventions like modifying dietary habit and physical exercise are effective in reducing body weight, improving related biological dysregulations and depressive symptoms. Thus, a 12-week dietary intervention using a modified Mediterranean diet in MDD patients has shown convincingly that diet can effectively alleviate MDD symptoms (Jacka et al., 2017;Opie et al., 2018). Moreover, the consumption of fiber and fish, as an important part of Mediterranean diet, was associated with an increased abundance of bacteria with anti-inflammatory properties and SCFA producing bacteria as well as shortened episodes of depression (Garcia-Mantrana et al., 2018;Lassale et al., 2019). In addition, several studies have provided evidence for the health benefits of intermittent fasting, usually practiced by restricting eating from 12 to 24 h, to host, that might be mediated by shaping the gut microbiota to some extent Mattson et al., 2017). Moreover, a recent meta-analysis found that fasting groups had lower anxiety, depression levels and BMI (Berthelot et al., 2021). However, it remains unclear whether intermittent fasting is more effective than caloric restriction alone in achieving certain health benefits, such as improving metabolic function and reducing inflammation. More research is needed to fully understand the mechanisms behind these effects. Similarly, evidence suggests moderate levels of exercise have positive effects on stress, immunity and energy homeostasis (Pedersen and Hoffman-Goetz, 2000). Results from a randomized controlled trial suggest that exercise therapy is more effective in reducing depressive symptom scores in patients with higher baseline levels of TNF (Rethorst et al., 2013). More recently, Rethorst et al. demonstrated that patients with AD tended to respond better to aerobic exercise augmentation (Rethorst et al., 2016). Furthermore, free access to exercise was significantly associated with an increase in the relative abundance of the genera Bifidobacterium and Lactobacillus, which have been suggested to alleviate depressive symptoms, as well as with an increase in microbiota diversity (Y. Queipo-Ortuño et al., 2013).
Apart from these interventions, treatments meant to directly alter the composition of the microbiome such as pre-and probiotics, might have a benefit on mental and overall health, too. While prebiotics are substrates that are selectively utilized by host microorganisms conferring a health benefit, probiotics are live microorganisms which when administered in adequate amounts confer a health benefit on the host (Gibson et al., 2017;Hill et al., 2014). The term psychobiotics refers to microbiome interventions specifically designed to promote mental health. The majority of interventional studies exploring the effects of psychobiotics on mental symptoms have been performed in healthy participants and predominantly investigated various Lactobacillus and Bifidobacterium strains with varying results in terms of improving mood, moderating inflammatory or other biomarkers (Akkasheh et al., 2016;Romijn et al., 2017). Another feasible approach when looking at potential therapies by influencing microbiota composition, could become a fecal microbiota transplant from a healthy donor. However, dedicated studies in psychiatric conditions are not available yet.
With respect to the oral microbiome, clinicians often emphasize oral hygiene as a solution to slowing down PD and potentially related neuropsychiatric conditions. Oral hygiene products, such as toothpastes and mouthwashes, are formulated with antimicrobial chemicals or alcohol and alter the oral microbial composition by limiting the growth of certain species. Thus, oral toothpastes containing fluoride have been shown to decrease overall microbial load and diversity (Haraszthy et al., 2019). Ongoing research lines are examining other chemical compositions that may maintain oral hygiene without disrupting the commensal balance in the microbiota (Kong et al., 2021). However, these compounds should also be examined in the context of potentially reducing neuropsychiatric symptoms in future studies.
It is noteworthy, that iproniazid, which is the first antidepressant ever developed, was originally classified as an antibiotic (Butler et al., 2019). Now, it is well established, that the interaction between the microbiota and commonly used non-antibiotic drugs is complex and bidirectional. Several drugs, including psychotropic medication, can influence the composition of the microbiota, but vice versa, gut microbiota can also influence an individual's response to a drug by enzymatically transforming the drug's structure and altering its bioavailability, bioactivity or toxicity Weersma et al., 2020). Thus, on the one hand, the microbiome could contribute to the individual therapeutic response and the occurrence of adverse events, and on the other hand, the antimicrobial effect could be part of their mode of action (Maier et al., 2018).
In addition, there is a comparatively larger body of literature examining the response of depression to treatment that modulates the immune system, although interpretation of these data is limited by small sample sizes, a high risk of bias, high heterogeneity, and the fact that the majority of studies were conducted in patients with comorbid inflammatory diseases (Milaneschi et al., 2020). Thus, it remains to be determined whether improvement is due to their effects on somatic diseases. Furthermore, most studies have used anti-inflammatory medication as add-ons to conventional antidepressants. Trials in this area have focused mainly on non-steroidal anti-inflammatory drugs (NSAIDs) or cytokine inhibitors like etanercept or infliximab (Beurel et al., 2020). Overall, the findings suggest that anti-inflammatory approaches may be beneficial in MDD patients with prominent inflammation (Kappelmann et al., 2018;Köhler-Forsberg et al., 2019;Köhler et al., 2014). However, due to several limitations, this field now needs randomized controlled trials that could pave the way for novel, personalized treatment.

Conclusion
A growing body of data implicates chronic inflammation and metabolic dysregulations as a major player in the pathogenesis of energyrelated atypical depressive symptoms and related metabolic diseases. Recently, the gut and oral microbiome have been found to play an important role in regulating metabolic and immune processes affecting brain function and behavior, which may be linked to the development of atypical depressive symptoms.
This review mainly highlighted biological processes, linking the microbiome and immunometabolic dysregulations in AD, thereby neglecting obvious behavioral and psychosocial factors. However, identifying the biological underpinnings of the atypical subtype of depression would give patients a chance for personalized treatment as this phenotype is believed to be a significant contributor to treatment resistance in depression. For this reason, ongoing research lines should be expanded to include additional markers, such as extended clinical profiling, genetics, or metabolomics, as the immune-metabolic pathways discussed here cover only a portion of potential mechanisms underlying the clinical AD profile. When it comes to microbiota-brain connections in MDD patients, there is a need for better-designed studies with clear clinical phenotyping and multiple timepoints of microbiome collection (from gut and mouth) to examine cause and effect. Beyond that, the potential clinical value of biologically and clinically characterized subtypes needs to be analyzed in appropriate interventional studies, including treatments targeting relevant pathways linking microbiome and immunometabolic dysregulations in clinical presentations of depression, to pave the way for therapeutic alternatives in clinical practice.

Declaration of competing interest
All authors declare to not have any actual or potential conflict of interest including any financial, personal or other relationships, that could inappropriately influence, or be perceived to influence, the work.

Data availability
No data was used for the research described in the article.