GABA-Producing Bacteria as Potential Psychobiotics in Gut–Brain Axis Regulation

GABA is the primary inhibitory neurotransmitter in the mammalian brain, vital for maintaining the excitatory–inhibitory balance and ensuring the stability of neuronal network activity [42,43,44]. In the CNS, its actions are mediated through three receptor subtypes: the ionotropic GABA_A and GABA_C receptors, and the metabotropic GABA_B receptors [45]. GABA_A and GABA_C receptors regulate chloride ion influx, producing rapid synaptic inhibition. Conversely, GABA_B receptors generate slower but longer-lasting effects by modulating potassium and calcium channels via Gi/o protein signalling [46]. In the brain, GABA is synthesised in GABAergic neurons through the decarboxylation of glutamic acid, a reaction catalysed by glutamate decarboxylase (GAD) with pyridoxal phosphate (PLP) as a cofactor [47]. In mammals, two GAD isoforms are present: GAD65 and GAD67, encoded by the gad2 and gad1 genes, respectively [12]. GAD65 is mainly localised at synaptic terminals, supporting rapid GABA synthesis in response to increased neuronal activity [48]. In contrast, GAD67 is distributed within the cytoplasm and maintains basal GABA production, ensuring continuous inhibitory signalling [49,50]. Effective GABAergic signalling is vital for maintaining neuronal stability, synchronising brain oscillations, and regulating cognitive and emotional functions [51]. Dysfunction of GABA signalling contributes to increased excitability in brain regions such as the prefrontal cortex, hippocampus, and amygdala, which are critical for mood regulation, memory processing, and stress responses [52,53]. Clinically, impaired GABAergic activity has been linked to the development of anxiety and depressive disorders, heightened stress reactivity, and, in severe cases, an increased risk of epileptic seizures [54,55,56,57].

The ENS, often called the “second brain,” is a vast network of over 500 million neurons and glial cells distributed throughout the gastrointestinal tract [58]. As a vital part of the gut–brain axis, the ENS regulates sensory, motor, and interneuronal functions, supporting gut homeostasis and overall health [59]. GABA is a key neurotransmitter that controls the activity of the enteric nervous system [60]. In the gut, GABA primarily acts via ionotropic GABA_A and GABA_C receptors, which are located on enteric neurons, epithelial cells, immune cells, and enteroendocrine cells (EECs) [61,62,63]. These receptors enable GABAergic signalling that influences intestinal motility, secretion, barrier integrity, and visceral sensitivity, thereby promoting gastrointestinal stability. GABA in the gut originates from various sources. Endogenous GABA is produced by GABAergic neurons through glutamate decarboxylase (GAD), which converts L-glutamate into GABA [38,63]. Apart from neurons, some EECs, including enterochromaffin (EC) cells, also express GABA_A receptors and respond to GABA. Certain enteroendocrine-like cells produce GAD and may contribute to local GABA synthesis by interacting with enterochromaffin cells and other EECs [64,65]. External sources of GABA include the gut microbiota, particularly LAB and Bifidobacterium, which produce GABA via the GAD system [66,67]. Furthermore, consuming GABA-rich fermented foods can further increase intestinal GABA levels [68].

In the gut, GABA exerts both local and systemic effects. Within the ENS, GABAergic neurons are distributed throughout the myenteric and submucosal plexuses, where they regulate both excitatory cholinergic and inhibitory nitrergic pathways [69]. GABA released from these neurons modulates intestinal motility by inhibiting excitatory neurotransmission, thereby controlling smooth muscle contraction and peristalsis [61,63]. Moreover, GABAergic interneurons in the ENS participate in local reflex circuits that coordinate secretion and blood flow, ensuring optimal digestive function. Additionally, locally produced GABA may act in an autocrine or paracrine manner to modulate hormone release (e.g., serotonin, somatostatin) [70] and influence immune and epithelial cell functions, thereby maintaining gut barrier integrity [71].

Systemically, gut-derived GABA does not readily cross the blood–brain barrier; therefore, its effects on the central nervous system are believed to occur predominantly via indirect pathways involving neural, immune, and metabolic signalling [63,66,72]. Among these mechanisms, immune–brain communication is currently one of the strongest translational links between probiotic exposure and improvements in depressive symptoms, as supported by clinical evidence showing modulation of inflammatory markers following psychobiotic supplementation [73]. Vagal signalling is also considered an important pathway, particularly in preclinical models, in which microbiota-derived neuroactive compounds have been shown to influence vagal afferent activity and stress responsiveness. Through these interconnected pathways, microbial GABA may indirectly modulate neuroendocrine signalling, immune responses, mood regulation, cognitive function, stress reactivity, and hypothalamic–pituitary–adrenal (HPA) axis activity [70,74].

Bacterial biosynthesis of GABA primarily occurs through the glutamate decarboxylase (GAD) system, which consists of two components: the glutamate decarboxylase enzyme (encoded by gadA or gadB) and the glutamate/GABA antiporter (encoded by gadC) [15,78]. In this process, L-glutamate is transported into the bacterial cell by the GadC antiporter and is decarboxylated by GAD in the presence of the cofactor pyridoxal-5′-phosphate (PLP) [79]. This reaction produces GABA and CO₂ [80]. The newly formed intracellular GABA is then exported from the cell via the GadC antiporter, which also imports extracellular glutamate. This substrate–product exchange helps maintain the overall efficiency and balance of the GAD system [81]. GABA biosynthesis requires the conversion of intracellular protons. Therefore, the GABA operon is essential for the acid resistance of LAB [82]. The organization of the gad operon varies among bacterial species.

The presence of gad genes has been reported across multiple bacterial phyla, including Bacteroidetes (e.g., Bacteroides, Parabacteroides, Alistipes, Odoribacter, and Prevotella), Proteobacteria (e.g., Escherichia), Firmicutes (e.g., Enterococcus), and Actinobacteria (e.g., Bifidobacterium) [15]. Among these taxa, LAB, particularly members of the genera Lactobacillus and Lactococcus, have been extensively studied for their ability to synthesise GABA and their potential psychobiotic effects [83,84]. In LAB, the genomic structure of the GAD system varies considerably among species and strains. Typically, gadB, which encodes the glutamate decarboxylase enzyme, is located within the same operon as gadC, which encodes the glutamate/GABA antiporter, allowing coordinated transcription of both genes under the same promoter (gadCB) [15]. This arrangement has been well described in Levilactobacillus brevis [85] and Lactococcus lactis [86]. Additionally, the system is positively regulated by GadR, a transcriptional activator encoded upstream of gadR and expressed from its promoter, thereby enhancing gadB expression and increasing GABA synthesis [87,88]. Within LAB, the amino acid sequence similarity between GadA and GadB is relatively low [78]. However, their PLP-binding domains remain highly conserved, thereby maintaining catalytic function and enabling efficient glutamate decarboxylation [89,90]. In many LAB, gadA or gadB is present, including a high GABA-producing L. brevis strain [91,92,93,94]. Some Lactiplantibacillus plantarum strains harbour gad genes that are not organized into a classic operon because they are located on the plasmid [95]. In Bifidobacterium species, the GAD system, comprising the enzyme GadB and the antiporter GadC, functions synergistically to catalyse the conversion of glutamate to GABA. This GAD system has been well documented in Bifidobacterium dentium, Bifidobacterium angulatum, and Bifidobacterium adolescentis [15].

GABA production is highly strain-dependent and varies significantly even within the same bacterial species [15]. The efficiency of bacterial GABA biosynthesis is influenced by various environmental factors, including pH, temperature, oxygen availability, and substrate concentration (Figure 3). For instance, the growth of L. brevis NCL912 increased with temperature, peaking at 35 °C and declining at higher temperatures [96]. Similarly, L. plantarum DSM19463 exhibited its highest GABA synthesis rate (59 μM/h) within the temperature range of 30–35 °C [97]. Acidic conditions (pH = 5.4–5.5) and the availability of L-glutamate and glutamine as primary and secondary precursors are also key factors in bacterial GABA production [78]. Low pH induces expression of the GAD system, enabling bacteria to adapt to acid stress by increasing GABA production, though the optimal pH for maximal GABA synthesis varies by species. Di Cango et al. [97] showed that L. plantarum DSM19463 produced the maximum GABA (59 μM/h) at pH 6.0. However, Kumar et al. [98] observed that Lacticaseibacillus paracasei NFRI 7415 increased GABA production at 210 mM and pH 5.0. Additionally, GABA production by Streptococcus thermophilus Y2 was notably enhanced when the culture pH was periodically adjusted to 4.5 at 12-h intervals using NaOH or HCl after an initial 24 h of incubation [90]. Furthermore, supplementing culture media with glutamate has been shown to boost GABA biosynthesis. For example, cultivating L. paracasei NFRI 7415 in a medium containing 500 mM glutamate resulted in GABA production reaching 161 mM [99].

In vitro screening of bacterial GABA production is typically performed using de Man, Rogosa, and Sharpe (MRS) medium with varying concentrations of monosodium glutamate (MSG). MSG is used as a substrate because it can be hydrolysed to L-glutamic acid. This approach enables evaluation of strain-specific differences in GABA production efficiency and helps identify optimal substrate levels to maximise neurotransmitter synthesis. Duranti et al. [100] reported that 79% of tested B. adolescentis strains could convert MSG to GABA. Similarly, the highest GABA production by L. plantarum FNCC 260 in culture medium (1226.5 mg/L; ~11.9 mM) was observed at 100 mM MSG, compared to 809.2 mg/L (~7.85 mM) in MRS after 60 h of cultivation [78]. Under conditions with 100 mM of MSG, L. plantarum CGMCC 1.2437T produced 721.35 mM of GABA, which is 7 times higher than when no MSG was added [90]. Transcriptomic analysis showed that MSG increased the expression of key enzymes involved in carbohydrate metabolism, fatty acid synthesis, and amino acid metabolism [90]. Additionally, PLP, a cofactor of glutamate decarboxylase, is often added to enhance enzymatic activity and GABA production. Yang et al. [101] demonstrated that adding 0.2 mM and 0.02 mM PLP to the medium significantly enhanced GABA synthesis by S. thermophilus Y2. Meanwhile, Yogeswara et al. [102] reported that supplementing the medium with 0.6 mM PLP and 0.1 mM pyridoxine in L. plantarum FNCC 260 increased GABA output to 945. 3 mg/L (~9.17 mM) and 969. 5 mg/L (~9.40 mM), respectively. It was also shown that PLP at 10 or 100 µM in the culture medium significantly enhanced GABA biosynthesis by L. paracasei [99]. Furthermore, Cai et al. [103] demonstrated that adding 3% MSG and 2 mmol/L PLP to MRS medium markedly increased GABA production by L. plantarum FRT 7, reaching 1158. 6 mg/L (~11.2 mM). In contrast, cultivation of the same strain in standard MRS medium resulted in a considerably lower GABA concentration of 100. 75 mg/L (~0.98 mM).

An initial screening of the Integrated Microbial Genomes/Human Microbiome Project database by Pokusaeva et al. [104] identified GAD orthologs in 26 bacterial genera. Among Bacteroides, the most abundant and dominant genus in the human gut microbiota, it was reported to be the most prevalent genus harbouring the gadB gene, followed by Escherichia, Fusobacterium, Bifidobacterium, and Lactobacillus. These findings were supported by a meta-analysis of 1159 gut bacterial genomes representing 919 species, which confirmed the widespread presence of gadB orthologs within Bacteroides [105]. In total, 45 Bacteroides strains harbouring gadB were identified, and GABA production was experimentally confirmed in selected strains, including Bacteroides caccae KLE1911, Bacteroides vulgatus KLE1910, Bacteroides ovatus KLE1170, Bacteroides dorei KLE1912, Bacteroides uniformis KLE1913, and Bacteroides fragilis KLE1758 [105]. Further studies demonstrated that nearly 96% of Bacteroides genomes derived from human intestinal isolates contain a complete GAD system [105]. Experimental analysis of 16 human gut-derived Bacteroides strains confirmed GABA production in the range of 0.09–60.84 mM, with the highest levels observed in Bacteroides faecis PB–SESWS, Bacteroides fragilis PB-SZSJC, Bacteroides ovatus DSM 1896, and Bacteroides xylanisolvens DSM 18836. In addition to Bacteroides, GABA-producing capacity has also been reported in Parabacteroides and Eubacterium species [106].

Members of the genus Bifidobacterium are also well-documented GABA producers. Barrett et al. [107] demonstrated that strains such as B. dentium DPC6333, NCFB2243, Bifidobacterium infantisUCC35624↗, and B. adolescentis DPC6044, isolated from the human gastrointestinal tract and faeces, reported concentrations ranging from 2.2 to 12.48 mg/mL (~21.3–121.0 mM). Subsequent studies confirmed GABA production in additional species, including B. adolescentis, B. angulatum, and B. dentium isolated from human faecal samples, with reported concentrations of 4.7–58.2 mM, 25.4–33.6 mM, and 23.9 mM, respectively [15]. Genomic analyses further supported these findings. An in silico study of 1022 Bifidobacterium genomes isolated from the human gastrointestinal tract revealed a high prevalence of gad genes, particularly in B. adolescentis, suggesting that this species may be a key contributor to GABA production in the human gut microbiota [100]. In vitro screening identified high-producing B. adolescentis strains, such as B. adolescentis PRL2019 and HD17T2H, that produced 7.6 mM and 9.43 mM GABA, respectively. Moreover, in vivo studies demonstrated that supplementation with B. adolescentis increased GABA levels in rat faecal samples [100]. Similarly, screening of 16 commensal intestinal isolates obtained from healthy human faecal samples identified B. dentium ATCC 27678 as an efficient GABA-producing strain capable of converting glutamate to GABA via the GadB-mediated decarboxylation pathway [104].

In Lactobacillus, a high GABA-producing potential has been observed in L. plantarum and L. brevis [15,66,77,78,87,99]. For instance, L. brevis DPC6108, isolated from infant faeces, demonstrated a strong ability to convert monosodium glutamate (MSG) into GABA, producing 11.01 and 20.47 mg/mL (~106.8 mM and ~198.5 mM) GABA from 10 and 20 mg/mL MSG, respectively, in MRS medium in vitro. Additional studies showed that supplementation with this strain significantly increased GABA levels during faecal fermentation, reaching 66.25 µg/mL (~0.642 mM) after 4 h and peaking at 70.72 µg/mL (~0.686 mM) after 9 h, suggesting its interaction with gut microbial communities [107]. Other strains, such as L. plantarum 90sk and L. brevis 15f, isolated from adult faeces, have also been identified as effective producers of GABA [108]. The amount of GABA produced by L. plantarum 90sk in culture medium was detected at 200 mg/L (~1.94 mM) [108]. Beyond GABA production, some strains exhibit additional beneficial properties, including antibiotic resistance and antioxidant properties [109]. More recently, Limosilactobacillus fermentum L18, isolated from faecal samples of healthy children, was shown to produce 62.37 mg/mL (~605 mM) of GABA under in vitro conditions after 24 h of cultivation [110]. Lactococcus garvieae MJF010, isolated from healthy human faeces, has also been identified as a strain with strong GABA biosynthesis capacity [111]. GABA-producing bacterial strains isolated from the human gut microbiota are summarized in Table 1.

Numerous studies have shown that LAB can synthesise GABA in food matrices, including fermented vegetables, fermented dairy products, and cheeses [66,113]. Therefore, recent research has concentrated on the isolation, characterisation, and technological evaluation of food-derived GABA-producing strains (Table 2). The aim is to develop functional foods naturally enriched with GABA and to improve the health-promoting properties of fermented products [114].

High GABA-producing strains, including L. paracasei PF6, Lactobacillus delbrueckii subsp. bulgaricus PR1, L. lactis PU1, and L. brevis PM17, have been identified in cheeses such as Pecorino di Filiano and Pecorino del Reatino [115]. The highest GABA concentrations were recorded in Pecorino Marchigiano (289 mg/kg), Pecorino del Reatino (290 mg/kg), Pecorino Leccese (290 mg/kg), Pecorino Umbro (330 mg/kg), and Pecorino di Filiano (391 mg/kg). Among the strains tested, L. paracasei PF6, L. delbrueckii subsp. bulgaricus PR1, L. lactis PU1, L. plantarum C48, and L. brevis PM17 showed the highest GABA production levels (99.9, 63.0, 36.0, 16.0, and 15.0 mg/kg, respectively) [115]. High GABA-producing LAB strains have also been isolated from plant-based fermented foods, including Korean kimchi, traditional Chinese paocai, and Japanese fermented fish. For example, L. plantarum FRT7, isolated from paocai, produced 1158.6 mg/L (~11.2 mM) of GABA under optimised conditions [103].

Recent studies highlight L. brevis as one of the most efficient LAB species for GABA production [116,117]; however, its capacity is highly strain-dependent and affected by the source of isolation and fermentation conditions [86,118,119]. For example, L. brevis NCL912, isolated from paocai, produced approximately 15.4 g/L (~149 mM) of GABA under optimised conditions [103], while L. brevis K203, isolated from kimchi, produced 44.4 g/L (~431 mM) [120]. Similarly, Lim et al. [121] reported that L. brevis HYE1, also isolated from kimchi, produced 18.76 mM GABA under optimal in vitro conditions, while L. brevis CRL1942, isolated from quinoa sourdough, produced up to 255 mM GABA in the presence of 4.5% MSG [122]. Other high-producing strains include L. brevis CRAI, isolated from organic tomatoes, which produced 179.15 mM GABA from an initial concentration of 236.53 mM (4%) MSG [123]. Furthermore, co-cultivation strategies, such as those of L. delbrueckii subsp. bulgaricus with S. thermophilus IFO13957 [124], or L. brevis with L. plantarum from Egyptian dairy products [125], have been shown to enhance GABA production. In addition, strains isolated from Mexican milk kefir grains, including L. lactis (BIOTEC006-008) and Leuconostoc pseudomesenteroides (BIOTEC012), have been identified as GABA producers with potential probiotic properties [126]. Similarly, six L. lactis strains, namely LEY6, LEY7, LEY8, LEY11, LEY12, and LEY13, isolated from raw camel milk, produced GABA at concentrations ranging from 1.74 mM to 1.80 mM [127]. Moreover, all strains produced GABA in a Cabrales-like mini-cheese model, with GABA accumulation ranging from 350 mg/kg (LEY6) to 457 mg/kg (LEY12).

It should also be noted that the high GABA concentrations reported under optimised in vitro culture conditions may not accurately reflect the physiological levels achieved within the gastrointestinal tract in vivo. In the intestinal environment, factors such as substrate availability, microbial competition, intestinal absorption, and host metabolism may substantially influence microbial GABA production, bioavailability, and overall biological activity. Nonetheless, in vitro screening provides valuable preliminary insight into the GABA-producing capacity of psychobiotic strains and remains an important step in identifying candidate microorganisms for further translational and clinical investigation, including their potential incorporation into functional foods and into microbiota-based interventions to support individuals affected by stress-related or neuropsychiatric disorders.

Although this review primarily focuses on the role of GABA in regulating the MGB axis, this pathway is considered within a broader network of microbial-derived signals, including SCFAs and tryptophan metabolites, which collectively orchestrate host–microbiota communication, regulate gut barrier integrity, and influence CNS function. Psychobiotics may affect cognitive function through complex, bidirectional interactions within the MGB axis [27,58,68,72,105,144]. These effects are mediated by dynamic molecular crosstalk at the interface between the intestinal microbiota and the host [3,7,17,18,145]. Current evidence indicates that these interactions are driven by multiple interconnected mechanisms, with the production of bioactive metabolites playing a central role [146,147]. These molecules act as signalling mediators that can influence the CNS both directly and indirectly via peripheral pathways, including immune, endocrine, and neural signalling routes [148,149] (Figure 4).

Preclinical studies using animal models have shown that supplementation with specific Lactobacillus and Bifidobacterium strains capable of producing GABA has beneficial effects on nervous system regulation, including modulation of stress responses and reductions in anxiety- and depression-like behaviours [207,208,209]. The psychobiotic potential of L. rhamnosus JB-1 has been studied in relation to GABAergic signalling and stress-related behaviours [209]. Oral administration of L. rhamnosus JB-1 (1 × 10⁹ CFU) for 28 days resulted in modulation of the GABAergic system, accompanied by decreased circulating corticosterone levels and reduced anxiety- and depressive-like behaviours. In another study, mice subjected to early-life stress showed modulated behavioural responses following administration of L. plantarum PS128, as evidenced by improved psychomotor function and reduced anxiety-like behaviours [210]. The mechanism involves increasing dopamine transporter and β-arrestin expression while reducing phosphorylation of dopamine and cAMP-regulated phosphoprotein (DARPP-32) [211]. In a similar preclinical study, Cowan et al. [212] showed that supplementation with L. helveticus R0052 and L. rhamnosus R0011 mitigated the adverse effects of early-life stress in infant rats. In particular, L. helveticus R0052 was reported to alleviate stress-induced impairments in neuroplasticity and neurogenesis associated with chronic stress exposure. These effects were accompanied by reduced HPA axis activity and autonomic nervous system activation, as reflected by decreased cortisol and catecholamine levels. Similarly, Lonstein et al. [213] demonstrated that L. rhamnosus HN001 reduced stress-associated maternal behaviours in rats, including postpartum anxiety-like behaviours. Decreased levels of norepinephrine, dopamine, and serotonin in the prefrontal cortex supported these effects. However, Luo et al. [214] reported that L. helveticus NS8 alleviated depressive- and anxiety-like behaviours in a rat model of HA-induced neurological dysfunction, likely through modulation of depression-related neuronal pathways. In another study, long-term supplementation with L. paracasei K71 improved cognitive performance in mice, as demonstrated by enhanced performance in the Barnes maze and passive avoidance tests [215]. This effect was accompanied by an upregulation of brain-derived neurotrophic factor (BDNF) expression in the hippocampus. Similarly, L. plantarum GM11 has shown potential in alleviating depressive-like behaviours and other mental health disturbances in rat models. These effects may be associated with the modulation of monoamine neurotransmitter levels, normalisation of HPA axis activity, and regulation of the cAMP response element-binding protein (CREB)–BDNF signalling pathway [216]. On the other hand, evidence from animal studies supports the psychobiotic potential of specific Bifidobacterium strains. For instance, supplementation with Bifidobacterium breve CCFM1025 has been shown to alleviate anxiety- and depressive-like behaviours in a murine model. These effects were accompanied by modulation of the HPA axis and attenuation of stress-induced inflammatory responses [217]. Similar effects have been shown for B. longum Rosell^®-175. The use of this strain in mice, in combination with L. helveticus R0052, influenced the HPA axis under chronic stress, alleviating its effects [218]. In addition, research conducted by Jarosz et al. [219] provided evidence that supplementation of mice’ diets with L. rhamnosus JB-1 and B. longum Rosell^®-175 modifies the expression of brain proteins involved in metabolic and immune processes.

In humans, psychobiotics are increasingly investigated as a therapy for mental health disorders, emphasising the role of the gut microbiota in regulating psychological well-being through the gut–brain axis [220]. A meta-analysis by Zhang et al. [221], covering 13 studies, suggested that psychobiotic interventions targeting the gut microbiota could be a promising approach for managing mild-to-moderate depression. However, other meta-analyses show that the effects of probiotics on mental health outcomes are varied and may depend on the population studied and the parameters evaluated. For example, Ng et al. [222] found that probiotic supplementation provided statistically significant benefits for patients with mild to moderate depression. Conversely, Chao et al. [223] reported that the overall effect of probiotics on depressive symptoms in both depressed individuals and healthy adults was not statistically significant. Similarly, Gao et al. [224] noted variability in probiotic responses, showing different effects on anxiety and depression across studies. Nonetheless, many studies support the positive influence of psychobiotic supplementation on human mental health [225]. For instance, Akkasheh et al. [226] conducted a randomised, double-blind, placebo-controlled clinical trial involving 40 patients with depression, who were assigned to receive either a probiotic formulation (L. acidophilus, L. casei, B. bifidum) or a placebo for 8 weeks. The findings indicated that probiotic supplementation significantly lowered Beck Depression Inventory (BDI) scores compared to the placebo group. Similarly, Kazemi et al. [227] reported that, in an 8-week, randomised, double-blind trial, supplementation with L. helveticus R0052 and B. longum R0175 (1 × 10¹⁰ CFU/mL) substantially reduced BDI scores from 17.39 to 9.1 relative to the placebo. Zhang et al. [221] demonstrated that daily supplementation with L. paracasei Shirota (LcS) for 9 weeks significantly improved depressive symptoms in patients with depression. Another study by Moludi et al. [228] showed that an 8-week course of L. rhamnosus GG alleviated symptoms of chronic inflammation, depression, and anxiety in patients with coronary artery disease (CAD). Steenbergen et al. [229] conducted a study involving 40 healthy students who received a probiotic containing B. bifidum W23, B. lactis W52, L. acidophilus W37, L. brevis W63, L. casei W56, L. salivarius W24, L. lactis W19, and W58, or a placebo, for 28 days. The probiotic reduced negative cognitive responses to low mood, particularly rumination and aggressive thoughts. Another randomised, triple-blind, placebo-controlled trial with 71 participants used a similar probiotic formulation for 8 weeks [230]. The multi-strain probiotic, including B. bifidum W23, B. lactis W51 and W52, L. casei W56, L. salivarius W24, L. lactis W19 and W58, L. brevis W63, and W37, was evaluated for its effects on individuals with varying severity of depression. Results showed a significant reduction in cognitive reactivity and improvements in depressive symptoms, as measured by the BDI and Beck Anxiety Inventory (BAI). A 12-week, randomised controlled trial by Kim et al. [231] found that probiotic supplementation containing B. bifidum BGN4 and B. longum BORI notably decreased pro-inflammatory bacteria in the gut and increased serum BDNF levels in elderly participants. Moschonis et al. [232] reported that a multi-strain probiotic comprising L. fermentum LF16, L. rhamnosus LR06, L. plantarum LP01, and B. longum BL04, taken daily for 12 weeks, yielded beneficial effects in adults with subthreshold depression. Schaub et al. [233] examined the impact of a probiotic blend containing eight strains, including S. thermophilus NCIMB 30438, B. brevis NCIMB 30441, B. longum NCIMB 30436, L. acidophilus NCIMB 30442, L. plantarum NCIMB 30437, L. paracasei NCIMB 30439, and L. delbrueckii subsp. bulgaricus NCIMB 30440, on neuronal processes and microbiome alterations in 60 patients experiencing depressive episodes. Following treatment, those receiving probiotics exhibited a greater reduction in Hamilton Rating Scale for Depression (HAM-D) scores, accompanied by a concomitant increase in Lactobacillus abundance and alleviation of depressive symptoms. In clinical studies on academic stress, an 8-week course of L. casei Shirota (1 × 109 CFU/mL) demonstrated benefits, including reductions in cortisol and anxiety among students undergoing therapy [234]. Another study found that 8 weeks of daily L. casei Shirota supplementation led to notable reductions in anxiety, covering both cognitive and physical symptoms, as well as perceived stress among athletes [235]. Zhu et al. [236] also observed that a twice-daily intake of L. plantarum JYLP-326 for 3 weeks among university students significantly reduced anxiety, depression, and insomnia, and partially restored gut microbiota composition.

A summary of selected trials investigating the use of probiotic strains with psychobiotic potential is presented in Table 3.

In summary, accumulating evidence from in vitro studies, animal models, and clinical research suggests that psychobiotics, particularly strains of Lactobacillus and Bifidobacterium, may be adjunctive strategies for managing mental health disorders [211,215,217,237,245]. In vitro studies have demonstrated that selected probiotic strains can produce neuroactive compounds, modulate inflammatory signalling, and influence neurotransmitter-related pathways involved in MGB axis communication. Animal studies further support these findings, showing that psychobiotic administration may affect stress responsiveness, GABAergic signalling, neuroinflammation, and behaviour associated with anxiety and depression-like phenotypes [209,210]. However, translating these promising preclinical findings into human clinical outcomes remains inconsistent.

Clinical studies have produced mixed results, and the field continues to face substantial methodological and translational challenges [246]. Considerable heterogeneity exists among studies with respect to probiotic strains, dosages, intervention duration, study populations, and assessed endpoints, limiting the comparability and generalizability of findings [145,247]. A meta-analysis conducted by Kadosh et al. [248] reported mixed effects of psychobiotic interventions on stress and anxiety in young people. Although some strains, such as L. casei Shirota and B. bifidum, were associated with reduced salivary cortisol levels and lower self-reported stress during examination periods, most studies did not demonstrate significant reductions in anxiety symptoms. Furthermore, some studies reported potential adverse effects, including increased anxiety scores [249] or elevated heart rate [250] following probiotic supplementation.

The translational gap between animal and human studies was highlighted by Kelly et al. [251], who demonstrated that L. rhamnosus (JB-1), despite producing anxiolytic effects in mice, failed to significantly improve stress-related or cognitive outcomes in healthy human volunteers. Nevertheless, broader evidence remains encouraging. A comprehensive umbrella review by Du et al. [252], including 47 meta-analyses across 12 categories of neuropsychiatric outcomes, concluded that probiotic supplementation may exert beneficial effects on depressive symptoms, stress-related parameters, and cognitive function. However, the authors also emphasized substantial heterogeneity between studies, limited mechanistic evidence, and the need for further well-designed preclinical and clinical investigations to establish causal relationships and clarify the underlying biological mechanisms.

No new data were created or analyzed in this study. Data sharing is not applicable.