Metformin and Sitagliptin Impact the Brain Kynurenine Pathway: Region-Specific Modulation of Neuroactive Metabolites in Non-Diabetic Male Rats

TL;DR

The study shows that metformin and sitagliptin affect brain metabolites involved in the kynurenine pathway differently in various brain regions of non-diabetic rats.

Contribution

This study reveals region-specific modulation of the TRP-KYN pathway by metformin and sitagliptin in the brain.

Findings

Metformin reduced cortical TRP and KYNA, hippocampal KYN, and cerebellar kynurenines.

Sitagliptin did not alter kynurenine levels but both drugs affected TRP/KYN or KYN/KYNA ratios.

Metformin increased KYNA in the striatum while decreasing it in the cortex.

Abstract

An excessive activation of the tryptophan–kynurenine (TRP-KYN) pathway, frequently observed in metabolic and inflammatory disorders, leads to disturbances in the balance between neurotoxic and neuroprotective metabolites. These alterations may contribute to neuronal dysfunction and cognitive impairment, highlighting the importance of modulating this pathway in the context of neuroprotection. Metformin, apart from the AMPK activation and its broad anti-inflammatory actions, has been indicated as a drug capable of influencing the synthesis of TRP metabolites, including the neuroprotective kynurenic acid (KYNA), whereas the effects of sitagliptin in this regard are not known. Here, the effects of sub-chronic metformin or sitagliptin treatment on the brain levels of kynurenines and on functional alterations within the TRP-KYN pathway were evaluated in vivo, in adult non-diabetic Wistar male rats. A 5-day treatment with metformin decreased cortical TRP and KYNA, hippocampal KYN, and cerebellar levels of all studied kynurenines, whereas in the striatum, KYNA level increased. In contrast, sitagliptin did not alter the formation of kynurenines in the examined structures. However, both of the tested drugs had a significant impact on TRP/L-KYN or L-KYN/KYNA ratios in different parts of the brain. These findings indicate a prominent region-specific effect of metformin on brain kynurenines. In conclusion, commonly used antidiabetic agents differ in their impact on central TRP metabolism, which may have significant implications for understanding their potential neuroprotective effects and role in cognitive impairment.

Figures

Full text

1. Introduction

Under physiological conditions, most of ingested tryptophan (TRP) is preferentially metabolized via the kynurenine pathway (KP), while only a limited amount is allocated to serotonin, melatonin, and protein synthesis [1] (Figure 1). Kynurenines—metabolites of TRP generated both peripherally and in the brain—play essential roles in numerous physiological processes, including immune regulation and the maintenance of cellular viability [2]. The KP generates metabolites with a wide range of biological functions. Some of these compounds, such as 3-hydroxykynurenine (3-HK), quinolinate (QUIN), and 3-hydroxyanthranilic acid (3-HANA), have been identified as potentially neurotoxic. QUIN contributes to neuronal damage primarily through excessive stimulation of N-methyl-D-aspartate (NMDA) receptors, whereas the harmful effects of 3-HK and 3-HANA are predominantly attributed to their ability to induce oxidative stress [3].

Kynurenic acid (KYNA) is the key cytoprotective metabolite within the KP. It displays its strongest affinity for the strychnine-insensitive glycine site on the NR1 subunit of the NMDA receptor, while also interacting—though less potently—with AMPA and kainate receptors [4]. KYNA is widely considered a noncompetitive antagonist of nicotinic α7 receptors, although some studies have failed to confirm this effect [5,6]. Beyond receptor antagonism, KYNA can interact with the aryl hydrocarbon receptor (AHR), a transcription factor that regulates xenobiotic metabolism [5], and functions as a ligand for the orphan G protein-coupled receptor GPR35, which is expressed on glial and immune cells. Activation of GPR35 by KYNA may modulate glial metabolism and immune signaling, potentially through a reduction in intracellular cAMP levels [4,7].

Growing evidence indicates that TRP metabolites generated through the TRP–kynurenine (KYN) pathway influence multiple metabolic processes by acting on neurons and glia, as well as immune and muscle cells [8]. The subacute or chronic inflammatory states affect the TRP–KYN pathway, while, conversely, TRP metabolites can modulate immune activity by exerting both pro- and anti-inflammatory effects [9]. Altered concentrations of KP metabolites have been reported in Alzheimer’s disease, Parkinson’s disease, and other dementias in both brain tissue and cerebrospinal fluid [10,11,12]. These metabolic changes correlate with cognitive decline, synaptic dysfunction, and disease severity. Moreover, disease- and region-specific alterations in KP enzyme expression suggest a role of the pathway in selective neuronal vulnerability [10,13]. Recent reviews propose that KP metabolites may serve as biomarkers reflecting neuroinflammatory status and disease progression [10,11].

Metformin, a commonly used antidiabetic drug, exerts diverse neuroprotective effects by modulating key metabolic and inflammatory signaling pathways. Its primary mechanism of action involves the activation of AMP-activated protein kinase (AMPK), a crucial regulator of cellular energy homeostasis. AMPK stimulation concurrently suppresses mTOR signaling and promotes autophagy, mitochondrial renewal, and improved neuronal resilience [14,15]. The AMPK-mediated metformin’s protective effects were demonstrated in various models of neurodegeneration and neurotoxin-induced neuronal injury [16,17,18,19]. Metformin mitigates glutamate-mediated excitotoxicity by maintaining MAPK/ERK and PI3K activity while reducing intracellular Ca^2+^ accumulation and reactive oxygen species generation [20]. Furthermore, it enhances hippocampal neurogenesis following injury or metabolic impairment and promotes neural progenitor cell proliferation through transcriptional regulation involving factors such as TAp73 [21,22,23,24,25,26,27]. Importantly, recent studies indicate that under in vitro conditions, metformin may inhibit the synthesis of KYNA, whose elevated levels were linked with impaired memory and learning [28]. Metformin may also attenuate neuroinflammation by reducing microglial and astroglial activation, suppressing the release of pro-inflammatory cytokines, and modulating NF-κB-dependent signaling pathways, partly through the regulation of miR-141 [29,30,31,32,33,34,35]. In models of Alzheimer’s disease, it decreases Aβ accumulation, tau phosphorylation, and both oxidative and nitrosative stress, thereby mitigating cognitive decline [36,37,38,39]. Through this broad spectrum of metabolic, anti-inflammatory, and neurogenic actions, metformin exhibits a robust neuroprotective profile and is increasingly recognized as a promising therapeutic agent for brain disorders.

Sitagliptin, a selective inhibitor of dipeptidyl-peptidase 4 (DPP-4), prolongs the half-life of endogenous incretin hormones, such as glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP), by preventing their degradation. Elevated GLP-1 and GIP levels in the brain exert neurotrophic and neuroprotective effects, including protection against oxidative stress, enhancement of insulin signaling, and reduction in neuroinflammation [40,41]. Sitagliptin administration has been shown to activate CREB (cAMP response element-binding protein) signaling and increase mitochondrial antioxidant defense, evidenced by elevated manganese superoxide dismutase (MnSOD), resulting in reduced lesion size after experimental traumatic brain injury [42]. In models of neurodegeneration, sitagliptin attenuates pro-inflammatory cytokines (e.g., TNF-α, IL-6, and IL-17) and increases anti-inflammatory mediators (e.g., IL-10, TGF-β), suggesting that it can modulate microglial activation and neuroimmune responses beyond its incretin-based effects [43]. Moreover, sitagliptin prevents brain mitochondrial dysfunction and improves cognition, likely reflecting its combined actions on metabolism, oxidative stress, and neuronal resilience [44]. Finally, studies in non-human primates demonstrate that oral sitagliptin elevates GLP-1 and GIP concentrations in cerebrospinal fluid despite limited central penetration, supporting its translational potential for neurodegenerative conditions [40].

The aim of this study was to answer the question of whether metformin and sitagliptin, antidiabetic drugs with different mechanisms of action and recognized neuroprotective features, would impact the KP and alter the levels of selected kynurenines in four brain regions of male non-diabetic rats in vivo.

2. Results

2.1. Effects of a 5-Day Treatment with Metformin and Sitagliptin on Cortical Levels of Kynurenine Pathway Metabolites

A short-term administration of metformin significantly reduced the amount of TRP (83% of control, p = 0.0005 vs. control) (30.84 ± 2.63 vs. 25.55 ± 1.16 nmol/g tissue) and cortical KYNA levels (81% of the control; p = 0.0006) (61.92 ± 4.79 vs. 49.2912 ± 5.59 pmol/g tissue) in the cerebral cortex. Metformin did not significantly affect cortical L-KYN (1120.9 ± 155.85 vs. 989 ± 135.17 pmol/g tissue) (Figure 2A).

Sitagliptin did not significantly affect levels of TRP, L-KYN, and KYNA in any of the examined structures. In the cortex, the TRP content (nmol/tissue) was 28.25 ± 0.77 vs. 28.37 ± 4.63, L-KYN (pmol/g tissue) content was 848.25 ± 172.67 vs. 997.75 ± 224.16, and KYNA content (pmol/g tissue) reached 61.26 ± 5.5 vs. 57.5 ± 9.14 (Figure 2B).

2.2. Modulation of Kynurenic Acid, Kynurenine, and Tryptophan in the Striatum Following Five-Day Administration of Metformin and Sitagliptin

Metformin application did not affect striatal levels of TRP (29.992 ± 4.6 vs. 31.5 ± 3.6 nmol/g tissue) or L-KYN (953.96 ± 281.8 vs. 863.85 ± 99.97 nmol/g tissue). Treatment with metformin did significantly increase KYNA levels in treated rats (114% of control, p = 0.018) (54.29 ± 5.25 vs. 63.55 ± 7.24) (Figure 3A).

In the striatum, TRP content (nmol/tissue) reached 32.46 ± 3.98 vs. 30.1 ± 5.58, L-KYN (pmol/g tissue) content reached 844.19 ± 158.64 vs. 898.05 ± 209.08, and KYNA content (pmol/g tissue) was 65.31 ± 16.03 vs. 56.91 ± 7.43 (Figure 3B).

2.3. Effects of a 5-Day Treatment with Metformin and Sitagliptin on Hippocampal Levels of Kynurenine Pathway Metabolites

In the hippocampus, metformin did not influence TRP and KYNA levels. The TRP content (nmol/tissue) was 30.94 ± 3.82 vs. 30.33 ± 3.81, and KYNA content (pmol/g tissue) reached 411.23 ± 51.57 vs. 404.52 ± 51.75. The levels of LKYN were significantly decreased (68% of control, p = 0.001) with content reaching 2002.3 ± 293.56 vs. 1359.27 ± 322.1 nmol/g tissue (Figure 4A).

Hippocampal content of TRP reached 29.81 ± 2.78 vs. 30.3 ± 2.38 (nmol/tissue), L-KYN (nmol/g tissue) content reached 2.83 ± 0.65 vs. 2.97 ± 0.58, and KYNA content (pmol/g tissue) was 404.57 ± 41.74 vs. 403.99 ± 31.67 (Figure 4B).

2.4. Effects of a 5-Day Treatment with Metformin and Sitagliptin on Cerebellar Levels of Kynurenine Pathway Metabolites

In the cerebellum, metformin treatment significantly decreased levels of TRP, LKYN, and KYNA by approx. 30% (p = 0.0017, 0.0035, and 0.0019, respectively). The TRP content (nmol/g tissue) was 31.29 ± 5.53 vs. 21.81 ± 3.9, the L-KYN content (pmol/g tissue) was 1535.75 ± 319.66 vs. 1060.5 ± 161.34, and the KYNA content (pmol/g tissue) was 63.99 ± 9.74 vs. 44.3 ± 10.83 (Figure 5A).

After 5-day administration of sitagliptin in the cerebellum, TRP level was 27.07 ± 4.97 vs. 23.71 ± 3.58 (nmol/tissue), L-KYN (nmol/g tissue) content reached 1.08 ± 0.3 vs. 1.23 ± 0.28, and KYNA content (pmol/g tissue) was 53.47 ± 10.23 vs. 52.44 ± 6.21 (Figure 5B).

2.5. TRP/L-KYN, TRP/KYNA, and L-KYN/KYNA Ratios

In addition to the baseline production in individual brain regions, ratios reflecting the relationships between specific components were also calculated.

2.5.1. TRP/L-KYN Ratio

Administration of metformin did not affect the TRP/L-KYN ratio in the cerebral cortex (27.75 ± 2.2 vs. 26.33 ± 4.24), striatum (34.27 ± 12.14 vs. 37.16 ± 7.81), or cerebellum (22 ± 10.24 vs. 20.71 ± 3.46). In the hippocampus, the TRP/L-KYN ratio increased significantly by 47% (15.86 ± 3.54 vs. 23.27 ± 5.66) (p = 0.0073) (Figure 6A).

Five-day administration of sitagliptin did not significantly alter the TRP/L-KYN ratio in the cerebral cortex (34.43 ± 6.44), striatum (39.36 ± 8.48 vs. 34.51 ± 7.65), or hippocampus (10.82 ± 1.53 vs. 10.57 ± 2.24). The TRP/L-KYN ratio decreased in the cerebellum by 23.5% (25.86 ± 4.2 vs. 19.79 ± 3.64) (p = 0.008). This is particularly interesting, as sitagliptin did not exhibit any statistically significant effects on the production or content of these components in any of the examined brain structures (Figure 6B).

2.5.2. TRP/KYNA Ratio

The TRP/KYNA ratio did not change in any of the examined brain structures under the influence of either metformin or sitagliptin. In the cerebral cortex, the ratio was 498 ± 36.73 vs. 511.365 ± 56.38 following metformin administration and 464.19 ± 39.37 vs. 495.21 ± 48.81 following sitagliptin administration. In the striatum, the ratio was 555.48 ± 101.94 vs. 499.26 ± 61.87 for metformin and 505.15 ± 69.91 vs. 529.28 ± 72.51 for sitagliptin. In the hippocampus, after metformin treatment, the ratio was 75.26 ± 0.26 in controls and 75 ± 0.26 in the treated group, while after sitagliptin administration, it was 73.78 ± 2.1 in controls versus 74.97 ± 0.13 in the treated group. In the cerebellum, the ratio was 488.39 ± 33.99 vs. 503.71 for metformin and 509.71 ± 56.3 vs. 454.67 ± 66.89 for sitagliptin (Figure 7).

2.5.3. L-KYN/KYNA Ratio

Both metformin and sitagliptin significantly altered the L-KYN/KYNA ratio, but their effects were region-specific. Metformin significantly decreased the L-KYN/KYNA ratio in the hippocampus by approximately 31.5% (4.98 ± 1.21 vs. 3.41 ± 0.89) (p = 0.0105), while the ratio remained unchanged in the cortex (18.09 ± 2.07 vs. 19.81 ± 3.45), striatum (17.89 ± 6.06 vs. 13.81 ± 2.67), and cerebellum (24.57 ± 6.37 vs. 26.66 ± 4.52) (Figure 8A).

In contrast, sitagliptin did not significantly affect the L-KYN/KYNA ratio in the striatum (13.58 ± 4 vs. 15.94 ± 3.79), hippocampus (6.95 ± 1.06 vs. 7.37 ± 1.44), or cerebellum (20.14 ± 3.67 vs. 23.42 ± 4.19). However, in the cortex, sitagliptin significantly increased the L-KYN/KYNA ratio from 13.87 ± 2.67 in the control group to 17.43 ± 3.3 in the treated group (p = 0.0327) (Figure 8B).

3. Discussion

In our study, a sub-chronic (5-day) treatment with metformin, but not with sitagliptin, caused several region-specific alterations in KP metabolites in the non-diabetic rat brain. Metformin lowered TRP and KYNA in the cortex, increased KYNA in the striatum, reduced L-KYN in the hippocampus, and decreased levels of all of the metabolites in the cerebellum by about 30%. These findings align with our previous in vitro observations, where metformin reduced KYNA formation in cortical slices without direct influence on KAT I and II [28]. Sitagliptin, in contrast, did not affect the levels of studied kynurenines, although a decrease in the TRP/L-KYN ratio in the cerebellum and an increase in the L-KYN/KYNA ratio in the cortex were observed. Overall, the metabolite ratios appeared more sensitive to metformin, indicating its broader influence on pathway balance.

This research focused on short-term (5-day) treatment in order to investigate early, region-specific neurochemical effects of metformin and sitagliptin independently of long-term metabolic adaptations. Such sub-chronic intervention allows mechanistic insights into the modulation of the KP that may precede chronic effects. Further experiments with long-term application of drugs should provide the answer to whether observed changes in the KP evolve in a time-dependent manner.

Regional differences in the effects of metformin may stem from its diverse mechanisms of action. Metformin has been reported to suppress the activity of indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO), the key enzymes responsible for converting tryptophan into kynurenine, as demonstrated in studies on human endometrial tissue and breast cancer cells [45,46]. It also affects systemic KP activity [47,48] and modulates the gut microbiota, which plays an important role in the intestinal synthesis of kynurenines [49,50,51]. The net effects of the KP metabolism and the production of specific kynurenines not only do not rely exclusively on the amount of substrate, but also seem highly region-specific [52,53,54]. Furthermore, it was shown by others that the accumulation of metformin in the brain is heterogeneous. Concerning the structures analyzed in our study, the rank of areas manifesting the highest concentration of metformin to the lowest is cerebellum, cortex, hippocampus, and striatum [55]. In fact, we have observed the most prominent changes in the cerebellum, with relatively less pronounced changes in other regions.

In the cortex, a decrease in TRP was accompanied by a decrease in KYNA. Bearing in mind that the biochemical pathway is rather a continuum than a collection of individual steps, it is conceivable that the decrease in TRP level in the cortex may be associated with (a) an increased build-up into the proteins, or (B) an increased conversion into serotonin and/or melatonin without concomitant alterations within the KP. Alternatively, (C) cortical flow through the KP might be increased as a result of an enhanced activity of IDO/TDO and the enzymes converting KYN to further metabolites such as 3-hydroxy-KYN (3-HK) and, further down, quinolinic acid or NAD+. Indeed, we have observed a decline in cortical KYNA, which suggests that the arm of the KP leading to 3-HK, QUIN, and NAD+ is activated, and there is less available KYN for the KYNA production. These possibilities require further detailed biochemical studies in vivo.

Interestingly, there was a selective increase in KYNA in the striatum, despite decreases in other regions. KYNA is produced in astrocytes via KAT, whereas the alternative conversion of KYN to 3-HK is driven by KMO in microglia. Our previous in vitro work showed that metformin does not influence KAT I or II activity in cortical tissue [28], which suggests that the reductions in cortical KYNA are not due to direct KAT inhibition. However, we cannot exclude that under in vivo conditions, the activity of KATs may be changed in selected areas of the brain. It is of interest that in kat2-/- mice, KYNA decreased in the hippocampus and cortex, while its concentration actually increased in the striatum [53]. Based on a similar pattern of altered KYNA formation as observed in our study, we carefully suggest that metformin treatment may lead to in vivo reduction in KAT II activity in the brain. Striatal increase in KYNA may reflect a compensatory increase in KAT I and III activities or changes in KMO activity. The striatum may also respond differently, possibly because of metformin’s anti-inflammatory effects, such as suppression of NF-κB, mTOR, and NLRP3 signaling pathways [56,57], which may modulate KMO in a region-specific way. These pathways are known regulators of the KP [58,59,60].

In the hippocampus, diminished content of KYN and lack of changes in TRP and KYNA levels were detected after administration of metformin. In line with the above reasoning, the KP seems to be directed towards metabolites downstream of KYN, which may result in its depletion with possible accumulation of NAD+. Diverse environmental and metabolic factors can lead to substantial increases in hippocampal KYNA levels. Prenatal exposure to KYN, sleep deprivation, and long-term ketogenic diet consistently elevate both tissue and extracellular KYNA in adult animals, often accompanied by hippocampus-dependent cognitive impairments [61,62,63]. In contrast, our study showed a reduction in hippocampal KYN without a corresponding decrease in KYNA levels, suggesting that metformin modulates the KP through mechanisms distinct from classical stress- or metabolism-related perturbations. This divergence highlights the complexity and region-specificity of KP regulation in the hippocampus.

Finally, the kynurenine–aryl hydrocarbon receptor (AHR) axis may also contribute to the region-specific regulation of the KP. KYN acts as an endogenous ligand for AHR [5,64,65], and excessive activation of this axis has been linked to metabolic problems, including obesity and glucose intolerance [66]. Metformin has been shown to decrease AHR activity [45,67], and the reductions in KYN in the hippocampus and cerebellum may indicate suppressed AHR signaling, further influencing local KP metabolism.

In the present study, 5-day administration of sitagliptin did not produce significant changes in the concentrations of TRP, L-KYN, or KYNA in any of the examined brain regions. These findings are consistent with clinical observations showing that modulation of the KP by DPP-4 inhibitors does not uniformly translate into measurable alterations in circulating kynurenine metabolites. It was shown that treatment with the DPP-4 inhibitor vildagliptin did not alter TRP, KYN, or the KYN/TRP ratio in diabetic patients, suggesting that short-term or moderate DPP-4 inhibition may be insufficient to modify KYN metabolism at the systemic level [68]. However, experimental data indicate that sitagliptin can influence the KP at the enzymatic level. Sitagliptin upregulates IDO1 expression in human monocyte-derived dendritic cells, a change that would theoretically promote the conversion of TRP to L-KYN [69]. Complementarily, the protease activity of CD26/DPP-4 can induce IDO1 expression and increase KYN production [70]. The absence of changes in metabolite levels in our study may reflect tissue-specific differences, cell-type dependency, or the need for a stronger immunological activation to reveal such enzymatic effects.

An intriguing observation is that sitagliptin significantly decreased the TRP/L-KYN ratio in the cerebellum, despite no measurable changes in the absolute concentrations of TRP or L-KYN. A lowered TRP/L-KYN ratio may indicate subtle shifts in IDO1 or TDO activity that remain below the threshold required to alter bulk metabolite levels but are detectable through alterations in metabolic ratios. Metabolomic evidence suggests that sitagliptin can influence amino-acid profiles, including metabolites derived from TRP [71], raising the possibility that the cerebellum exhibits region-specific sensitivity to DPP-4 inhibition.

Again, it is essential to keep in mind that in the brain, the expression of the KP enzymes is heterogenous and region-specific. Immunological studies indicate that sitagliptin modulates immune cell function, e.g., by reducing Th1/Th17 activation and promoting regulatory phenotypes [72]. Thus, the observed cerebellar effect may reflect early or subtle immune-mediated metabolic modulation. Such changes may require longer exposure, stronger inflammatory stimuli, or additional regulatory inputs to manifest as robust shifts in TRP, L-KYN, or KYNA concentrations.

Taken together, the presented data align with the emerging view that the effects of DPP-4 inhibitors on the KP are highly variable across tissues. Short-term sitagliptin administration may not be sufficient to alter the levels of core TRP metabolites in the brain; however, the cerebellum-specific decrease in the TRP/L-KYN ratio suggests the onset of subtle enzymatic or regulatory changes. These results support the view that sitagliptin’s influence on the KP is gradual and region-dependent, consistent with mechanisms proposed in in vitro and immunological studies.

Our findings should be interpreted with caution in terms of their potential clinical implications. The KP has been postulated as a key metabolic route linking inflammation and metabolic disturbances with neurological outcomes, including cognitive decline. The major impact seems associated with a dysregulation of TRP metabolism and downstream bioactive metabolites [73]. Clinical evidence suggests that metformin therapy may bring potential cognitive benefits in a broad population, including diabetic and non-diabetic individuals, leading to a slower decline in standardized cognitive scores and reduced risk of dementia. These results, however, are not fully clear, and the outcome is likely influenced by the dose and duration of therapy [74]. In agreement with these observations, the region-specific alterations in KP metabolites observed following metformin treatment may reflect early neurochemical changes that precede measurable functional outcomes. DPP-4 inhibitors, such as sitagliptin, have also been linked in observational and clinical studies to improved cognitive performance over time in elderly diabetic populations and to slower memory decline in individuals with comorbid Alzheimer’s disease [75]. Systematic reviews indicate that incretin-based therapies (including DPP-4 inhibitors) are generally associated with better cognitive scores compared with controls [76]. In our experimental paradigm, sitagliptin given for 5 days induced only subtle changes in the TRP/L-KYN ratio, suggesting that modulation of the central KP may require prolonged administration or be more pronounced in the presence of ongoing metabolic or inflammatory stress.

Importantly, our results highlight that commonly used antidiabetic agents differ in their impact on brain KP metabolism, which may have significant implications for understanding their potential neuroprotective effects and role in cognitive impairment. It seems important to follow up on the research presented here and to analyze the impact of studied drugs on behavioral patterns and their potential associations with the KP in both non-diabetic and diabetic animals, as well as in the models of impaired cognition.

Our study has certain limitations. First of all, the administration of metformin and sitagliptin was performed in non-diabetic rats, and we cannot exclude that the response of the KP to these drugs would be different in diabetic individuals. On the other hand, we were prompted to use initially naïve animals to ensure that the potential effects would be observed also under paradigms when the KP intervention is expected among individuals without diabetes. Secondly, the administration of drugs was carried out for 5 days, which is considered a sub-chronic period, but not for a prolonged time. We have studied the effects only in male rats. Separate experiments should reveal the response in females undergoing cyclic hormonal changes; the response to the studied drugs would be different.

4. Materials and Methods

4.1. Experimental Protocol

Adult male Wistar rats, purchased from the Experimental Medicine Center of the Medical University in Lublin, Poland, Breeder nr 077 in the Register of Breeders kept by the Minister of Science and Higher Education, weighing between 270 and 310 g at the start of the study, were used in all experiments. Male rats were chosen as experimental subjects in order to avoid the potential influence of female hormonal changes on the outcome of the experiment. The animals were kept under standard laboratory conditions, according to European Council Directive and the guidelines on animal care, with a 12 h day/night cycle and an ambient temperature of 18 °C, and had unrestricted access to food and water. Rats were randomly allocated to one out of four experimental groups (each group N = 8). Metformin (Sigma-Aldrich, St. Louis, MO, USA) and sitagliptin [(3R)-3-amino-1-[3-(trifluoromethyl)-5,6,7,8-tetrahydro-triazolo[3-a]pyrazin-7-yl]-4-(2,4,5-trifluorophenyl)butan-1-one; Sigma-Aldrich, St. Louis, MO, USA] were administered intraperitoneally (i.p.) at a dose of 250 mg/kg each for 5 days. Injections were performed at the same time each day. The control groups were treated with intraperitoneal (i.p.) injections of physiological saline as a vehicle for metformin control group or saline containing DMSO (Sigma-Aldrich, St. Louis, MO, USA) for sitagliptin control group. Injections were performed once daily for five consecutive days. Two hours following drug or vehicle administration, the animals were sacrificed, and the cortices, striata, cerebella, and hippocampi were carefully dissected. The isolated brain regions were preserved at −72 °C until further biochemical analyses. All procedures and experimental protocols were performed in compliance with the European Council Directive on the protection of animals used for scientific purposes and were approved by the Local Ethical Committee in Lublin, Poland (approval no.: 79/2024, issued 25 November 2024).

4.2. Brain Levels of Kynurenines

Prior to analysis, tissue samples were accurately weighed and homogenized in distilled water at a 1:10 (w/v) ratio with an ultrasonic homogenizer. Throughout the homogenization process, the sample tubes were kept submerged in a 4 °C water bath. The homogenate was then centrifuged at 12,000 rpm for 10 min at 4 °C, and the obtained and the supernatant obtained was collected into Eppendorf tubes for deproteinization with 100 µL 8% HClO_4_ per every 500 µL of supernatant (1:5 ratio). The resulting solution was thoroughly vortexed and subjected to a second centrifugation under the same conditions. The final product was preserved at −72 °C until subsequent evaluations

4.3. Quantification of Tryptophan, L-Kynurenine, and Kynurenic Acid in Brain Samples

Brain levels of TRP, L-KYN, and KYNA were determined quantitatively using an ultra-high-performance liquid chromatography (UHPLC) system coupled with a fluorescence detector (UltiMate 3000, Thermo Fisher Scientific, Waltham, MA, USA) following the method described by Zhao [77]. Separation was achieved using an Agilent HC-C18 analytical column (250 × 4.6 mm, 5 μm). The mobile phase consisted of 3 mM zinc acetate, 20 mM sodium acetate, and 7% acetonitrile, adjusted to pH 6.2, with a flow rate of 1 mL/min. KYNA was quantified using fluorescence detection (excitation 344 nm, emission 398 nm), while L-KYN and TRP were measured using UV detection at 365 nm and 250 nm, respectively. For quantitative evaluation, calibration curves were constructed for each compound, encompassing 0.1–1.0 pmol/100 μL for KYNA, 10–200 pmol/100 μL for L-KYN, and 100–2000 pmol/100 μL for TRP. Chromeleon 7.2 software facilitated both system operation and acquisition of chromatographic data.

4.4. Statistical Analyses

Data are presented as mean ± standard deviation (SD) and expressed as percentages of the values obtained in the corresponding control group, which was set at 100%. Ratios were calculated individually for each animal, and group means ± SD were subsequently determined. Normality of data distribution was assessed using the Shapiro–Wilk test. All datasets showed a Gaussian distribution. Statistical comparisons between control and treatment groups were performed using an unpaired, two-tailed Student’s t-test. Statistical significance was accepted at p < 0.05. Each experimental group consisted of N = 8 animals. All statistical analyses were performed using GraphPad Prism software v9.3.1. (GraphPad Software, San Diego, CA, USA).

5. Conclusions

Sub-chronic metformin administration produced clear, region-specific changes in the brain KP metabolites, whereas sitagliptin induced only subtle alterations limited to the TRP/L-KYN ratio in the cerebellum. These findings show that antidiabetic drugs differ in their ability to modulate central TRP metabolism, suggesting distinct neurochemical profiles relevant to cognitive vulnerability. Metformin appears to influence the KP activity even in the absence of metabolic or inflammatory stress, while the effects of DPP-4 inhibition may require longer exposure or additional regulatory stimuli. Further studies should investigate time-dependent mechanisms, enzyme-level regulation, and the functional impact of these drug-induced changes on brain physiology and cognition in various experimental paradigms in vivo.