Brain region and cell specificity of B4GALT6 in mice with depressive-like behavior: A pilot study

Abstract

BACKGROUND

Major depressive disorder (MDD) is a significant psychiatric condition that poses a serious threat to human life, primarily due to its association with suicidal behavior. The single nucleotide polymorphism rs372369000 is a risk locus for MDD and is located within the B4GALT6 gene. However, the biological and pathological implications of B4GALT6 in the brain concerning MDD remain unclear.

AIM

To reveal the biological and pathological significance of B4GALT6 in the brain and MDD.

METHODS

An inflammation-associated depression mouse model was developed by treating mice with lipopolysaccharides. B4GALT6-like immunoreactivity distribution and expression level was examined throughout the brain.

RESULTS

B4GALT6-like immunoreactivity increased in the hippocampus, PrL, and the visual cortex V1 region of MDD mice. This elevation varied across different brain subregions and cell types. Specifically, pathological alterations in B4GALT6-like immunoreactivity were observed in CA1 microglia, CA2 neurons, CA3 microglia and neurons, as well as V1 microglia and astrocytes. These changes correlated with the pathology of depression.

CONCLUSION

Although the neuropathological role of B4GALT6 in the brain remains to be fully characterized, B4GALT6 may be a potential therapeutic target for MDD.

Key Words: Major depressive disorder; rs372369000; B4GALT6; Potential therapeutic target; Gene expression

Core Tip: Major depressive disorder (MDD) poses a serious threat to human life. The single nucleotide polymorphism rs372369000 located within the B4GALT6 gene is a risk locus for MDD. However, the pathological mechanism of B4GALT6 in the brain in relation to MDD remains unclear. We developed an inflammation-associated depression model and examined B4GALT6-like immunoreactivity distribution and levels throughout the brain. The expression of B4GALT6-like immunoreactivity increased in the hippocampus, PrL, and the visual cortex V1 region of MDD mice. This elevation varied across different brain subregions and cell types. Taken together, our data suggest that B4GALT6 may be a potential therapeutic target for MDD.

INTRODUCTION

Major depressive disorder (MDD) is characterized by significant and persistent depressive symptoms that stem from diverse etiologies. The core features of MDD are disproportionate depressive mood and anhedonia. MDD is a complex psychiatric disorder with approximately 37% heritability[1]. Existing research shows that an intronic variant, rs372369000, located within the B4GALT6 gene, constitutes a risk locus for MDD[2].

B4GALT6 encodes for a type II membrane-bound glycoprotein, which plays a crucial role in lactosylceramide synthesis by transferring galactose from UDP-galactose to glucosylceramide[3-5]. Lactosylceramide serves as a critical precursor for ganglioside biosynthesis, which is essential for neuronal maturation and axon and myelin formation[6]. Studies have shown that abnormal interactions of glycosyltransferases/glycosidases such as B4GALT6 can cause imbalances in key glycosylation modifications, thereby driving the disease process through a triple cascade reaction: (1) Galactosylation defects of neuroadhesion proteins (such as integrins, NCAMs) lead to conformational and functional disorders, blocking integrin-mediated extracellular matrix signal transduction; (2) The impaired integrin signal further weakens downstream pathway activation of f neurotrophic factors (especially BDNF), inactivating pro-survival signals such as PI3K/Akt and ERK/CREB; and (3) The collapse of the above-mentioned signal network ultimately inhibits neural stem cell proliferation and differentiation in the subgranular region of the dentate gyrus in the hippocampus, resulting in a significant reduction in neurogenesis. This phenomenon has been confirmed to be the core pathological basis for hippocampal volume atrophy and cognitive and emotional dysfunction in patients with depression and is directly associated with behavioral deficiencies in chronic stress models. Despite the central role of B4GALT6 in this process, the functional significance of the rs372369000 SNP in the B4GALT6 gene and its potential involvement in MDD remain unclear.

In the current investigation, bioinformatics analysis revealed that rs372369000 was located in a genomic region that governs transcriptional activity. We utilized an lipopolysaccharides (LPS)-induced depressive-like mouse model and found that B4GALT6 expression in the hippocampus, PrL, and V1 brain regions of LPS-treated mice was significantly elevated, and this upregulation was cell subtype-specific. These findings suggest a potential role for B4GALT6 in MDD pathophysiology and further highlight the importance of exploring underlying cell-specific mechanisms in the etiology of this complex disorder.

MATERIALS AND METHODS

Bioinformatics analysis of the regulatory function of rs372369000

RegulomeDB[7,8], a tool for functional variation annotation, was employed to assess whether rs372369000 possesses the capability to regulate gene expression. This tool assigned a probability score for the variant, spanning from 0 to 1, with a score closer to 1 signifying a higher likelihood that the variant has a regulatory function. The Epigenome Roadmap[9] and ENCODE[10-12] databases were utilized to ascertain whether rs372369000 is located in a transcriptionally active region of DNA and unravel potential regulatory mechanisms.

Animals and depressive-like behaviors

Thirty male C57BL/6J mice aged 4-6 weeks were included in the experimental study. These mice were obtained from Changzhou Cavens Experimental Animal Co., Ltd. The experimental animal use was approved by the Experimental Animal Ethics Committee of Southeast University, with approval number 20220104003.

A sucrose preference test was used to assess the core symptom of MDD, known as anhedonia. The tail suspension test and forced swim test were used to evaluate depression mood. Weight measurements were used to evaluate symptoms of weight gain/Loss. The ANY-maze animal behavior analysis system was used to assess animal behavior. The depression model utilized in this study was based on the LPS-induced depressive-like behavior paradigm[13].

After acclimating to the environment for 1 week, baseline behavioral data of the mice were collected. Five mice were excluded as outliers because their baseline data were not within the range of two standard deviations plus or minus the average value. The remaining 25 mice were randomly assigned to the LPS and phosphate-buffered saline (PBS) groups. After no statistically significant difference was observed in two independent samples t-tests, the subsequent experimental process was initiated. LPS (Sigma-Aldrich, No. L2630) was dissolved in PBS and administered intraperitoneally at a dose of 1.0 mg/kg for 7 consecutive days. The PBS group received an equivalent volume of PBS intraperitoneally.

Western blot

Western blots were performed as described previously[2,14]. In brief, RIPA lysis buffer (Servicebio, No. G2002) containing PMSF (Servicebio, No. G2008), NaF (Servicebio, No. G2006), and Na₃VO₄ (Servicebio, No. G2007) was added to mouse brain tissue samples. SDS-PAGE kits (Servicebio, No. G2003) were applied to prepare 10% separating gel and 5% concentrating gel. The FDbio-Dura ECL Kit (Fdbio science, No. FD8030) was used for visualization. The primary antibody was rabbit anti-B4GALT6 (Proteintech, No. 20148-1-AP). The secondary antibody was HRP-conjugated anti-rabbit (Proteintech, No. SA00001-2).

Immunofluorescence

Immunofluorescence was performed as described previously[14]. Briefly, mouse brain tissue was dehydrated and embedded in paraffin. Paraffin sections were dewaxed and subjected to antigen retrieval. After sequentially incubating the primary and secondary antibodies, DAPI staining was performed. The primary antibodies used are as follows: B4GALT6 (Proteintech, No. 20148-1-AP), GFAP (Servicebio, No. GB12096-100), IBA1 (Servicebio, No. GB12105-100), and NeuN (Servicebio, No. GB15138-100). The secondary antibodies used are as follows: Cy3 (Servicebio, No. GB21301) and Alexa Fluor^® 488 (Servicebio, No. GB25303).

Statistical analysis

Experimental results were expressed as mean ± SD and were analyzed using two-sample t-tests.

RESULTS

The rs372369000 SNP is a functional variant

The rs372369000 SNP was located on q12.1 of chromosome 18 in the intronic region of the B4GALT6 gene (Figure 1A). The RegulomeDB probability score for rs372369000 was 0.82805, suggesting a high probability that this SNP regulates gene expression. ENCODE chromatin accessibility data indicated that rs372369000 was located in the open chromatin region of motor neurons (Figure 1B, Supplementary Figure 1). The results of ENCODE chromatin states analysis showed that the main function of rs372369000 was to weaken transcription (Supplementary Figure 2). ENCODE transcription factor chromatin immunoprecipitation sequencing (ChIP-seq) results showed that rs372369000 primarily bound to the transcription factor POLR2A and bound to transcription factors ZEB1 and EZH2 in bipolar neurons (Figure 1B, Supplementary Figure 3). ChIP-seq results of epigenetic modifications in ENCODE showed that the rs372369000 region had H3k27ac and H3k4me3 signal peaks in motor neurons and H3k4me1 signal peaks in BE2C cells (Figure 1B). ChIP-seq results of epigenetic modifications in Roadmap showed that the rs372369000 region had H3k36me3 signal peaks in neural stem cells (Figure 1C). These results suggest that rs372369000 is indeed a functional variant with transcriptional regulation activity.

Figure 1 rs372369000 is a functional variant. A: Rs372369000 is located on q12.1 of chromosome 18 in the intronic region of the B4GALT6 gene; B: ENCODE chromatin accessibility data indicated that rs372369000 was located in the open chromatin region of motor neurons; C: ChIP-seq results of epigenetic modification in Roadmap showed that the rs372369000 region had H3k36me3 signal peaks in neural stem cells. Chr: Chromosome; ATAC-seq: Assay for transposase-accessible chromatin with high throughput sequencing; ChIP-seq: Chromatin immunoprecipitation sequencing.

B4GALT6 expression increases in the hippocampus, PrL and V1 of visual cortex of depressive-like behavior mice

To assess whether B4GALT6 expression is altered in depressive-like behavior, we generated an LPS-induced depressive-like behavior mouse model (Figure 2A-D). We then examined B4GALT6 expression in the hippocampus, PrL, and V1 regions of control and LPS-treated mice by Western blot. B4GALT6 expression significantly increased in the hippocampus, PrL, and V1 regions of the LPS group (Figure 2E and F).

Figure 2 B4GALT6 expression in hippocampus, PrL and V1 of visual cortex increased in depressive-like behavior mice. A: Schematic diagram of the experimental procedure; B: Sucrose preference test outcomes [phosphate-buffered saline (PBS) vs lipopolysaccharides (LPS), 86.20 ± 9.97 vs 64.14 ± 9.1, P = 8.00 × 10^-6]; C: Results of the tail suspension test (PBS vs LPS, 93.42 ± 45.96 vs 133 ± 39.26, P = 0.030); D: Results of the forced swimming test (PBS vs LPS, 50.75 ± 18.14 vs 70.14 ± 23.4, P = 0.037); E: Western blot of B4GALT6 expression; F: Statistical results of Western blot for B4GALT6. FST: Forced swimming test; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase; HIP: Hippocampus; IF: Immunofluorescence; LPS: Lipopolysaccharides; PBS: Phosphate-buffered saline; SPT: Sucrose preference test; TST: Tail suspension test; WB: Western blot.

B4GALT6 expression is altered in specific brain regions and cell subtypes in mice with depressive-like behavior

To assess the cell subtype specificity of B4GALT6 expression in depressive-like behavioral pathology, we costained brain sections with B4GALT6 and IBA1, GFAP, and NeuN using immunofluorescence to identify any differences in cell subtype density. The LPS group had increased cell density of B4GALT6⁺/IBA1⁺ cells within the hippocampal CA1 region (Figure 3) and B4GALT6⁺/NeuN⁺ cells in the hippocampal CA2 region (Figure 4). The densities of B4GALT6⁺/NeuN⁺ neurons and B4GALT6⁺/IBA1⁺ glia were both increased in the hippocampal CA3 region (Figure 5). There was no statistically significant difference in cell subtypes in the dentate gyrus of the hippocampus (Supplementary Figure 4) or the PrL brain region between the PBS and LPS groups (Supplementary Figure 5). Compared to the PBS group, the density of B4GALT6⁺GFAP⁺ and B4GALT6⁺IBA1⁺ cells significantly increased in the LPS group (Figure 6). Bioinformatics analysis revealed that B4GALT6 is evolutionarily conserved across jawed vertebrates, as shown in Supplementary Figure 6. Human brain tissue expression analysis showed widespread B4GALT6 expression, as detailed in Supplementary Figure7. Functional prediction tools GeneMANIA and STRING identified B4GALT6 interactions with other glycosyltransferases and glycosidases (Supplementary Figure 8).

Figure 3 Brains from depressive-like mice have increased B4GALT6⁺IBA1⁺cell density within the hippocampal CA1 region. A: Statistical results of the number of B4GALT6 and GFAP co-labeled cells (B4GALT6⁺GFAP⁺) in the hippocampal CA1 region, with no significant difference between the two groups; B: Statistical results of the number of B4GALT6 and NeuN co-labeled cells (B4GALT6⁺NeuN⁺) in the hippocampal CA1 region, with no significant difference between the two groups; C: Statistical results of the number of B4GALT6 and IBA1 co-labeled cells (B4GALT6⁺IBA1⁺) in the hippocampal CA1 region, with a significant increase in the lipopolysaccharides (LPS) group compared to the phosphate-buffered saline (PBS) group; D: Immunofluorescence co-localization images of the hippocampal CA1 region, showing the co-localization of DAPI (cell nuclei, blue), B4GALT6 (red) with IBA1 (green), GFAP (green), and NeuN (green) in the PBS group and LPS group, respectively. The number of B4GALT6⁺IBA1⁺ cells in the LPS group was increased compared to the PBS group. LPS: Lipopolysaccharides; PBS: Phosphate-buffered saline.

Figure 4 The hippocampal CA2 region of depressive-like mice have an increased density of B4GALT6⁺NeuN⁺ cells. A: Statistical results of the number of B4GALT6 and GFAP co-labeled cells (B4GALT6⁺GFAP⁺) in the hippocampal CA2 region, with no significant difference between the two groups; B: Statistical results of the number of B4GALT6 and NeuN co-labeled cells (B4GALT6⁺NeuN⁺) in the hippocampal CA2 region, indicating that the number of such cells in the lipopolysaccharides (LPS) group was significantly higher than that in the phosphate-buffered saline (PBS) group; C: Statistical results of the number of B4GALT6 and IBA1 co-labeled cells (B4GALT6⁺IBA1⁺) in the hippocampal CA2 region, with no significant difference between the two groups; D: Immunofluorescence co-localization images of the hippocampal CA2 region, displaying the co-localization of DAPI (blue), B4GALT6 (red) with NeuN (green), GFAP (green), and IBA1 (green) in the PBS group and LPS group. The number of B4GALT6⁺NeuN⁺ cells was higher in the LPS group compared to the PBS group. LPS: Lipopolysaccharides; PBS: Phosphate-buffered saline.

Figure 5 The main cellular differences between the phosphate buffer saline group and the lipopolysaccharides group in the hippocampal CA3 region were neurons and microglia. A: Statistical results of the number of B4GALT6 and GFAP co-labeled cells (B4GALT6⁺GFAP⁺) in the hippocampal CA3 region, with no significant difference between the two groups; B: Statistical results of the number of B4GALT6 and NeuN co-labeled cells (B4GALT6⁺NeuN⁺) in the hippocampal CA3 region, revealing that the number of such cells in the lipopolysaccharides (LPS) group was significantly higher than that in the phosphate-buffered saline (PBS) group; C: Statistical results of the number of B4GALT6 and IBA1 co-labeled cells (B4GALT6⁺IBA1⁺) in the hippocampal CA3 region, showing that the number of such cells was significantly higher in the LPS group compared to the PBS group; D: Immunofluorescence co-localization images of the hippocampal CA3 region, presenting the co-localization of DAPI (blue), B4GALT6 (red) with NeuN (green), IBA1 (green), and GFAP (green) in the PBS group and LPS group. The numbers of positive co-labeled cells in neurons and microglia in the LPS group were both higher than those in the PBS group. LPS: Lipopolysaccharides; PBS: Phosphate-buffered saline.

Figure 6 B4GALT6 functions in astrocytes and microglia in the V1 region of the visual cortex. A: Statistical results of the number of B4GALT6 and GFAP co-labeled cells (B4GALT6⁺GFAP⁺) in the V1 region of the visual cortex, indicating that the number of such cells in the lipopolysaccharides (LPS) group was significantly higher than that in the phosphate-buffered saline (PBS) group; B: Statistical results of the number of B4GALT6 and NeuN co-labeled cells (B4GALT6⁺NeuN⁺) in the V1 region of the visual cortex, with no significant difference between the two groups; C: Statistical results of the number of B4GALT6 and IBA1 co-labeled cells (B4GALT6⁺IBA1⁺) in the V1 region of the visual cortex, demonstrating that the number of such cells in the LPS group was significantly higher than that in the PBS group; D: Immunofluorescence co-localization images of the V1 region of the visual cortex, showing the co-localization of DAPI (blue), B4GALT6 (red) with IBA1 (green), GFAP (green), and NeuN (green) in the PBS group and LPS group. The numbers of positive co-labeled cells in microglia and astrocytes in the LPS group were significantly higher than those in the PBS group. LPS: Lipopolysaccharides; PBS: Phosphate-buffered saline.

DISCUSSION

In this study, we observed increased expression of B4GALT6 in the hippocampus, PrL, and V1 brain regions of mice with LPS-induced depressive-like behavior. This elevation was distinct within brain subregions and cell subtypes. Specifically, increased B4GALT6 expression in CA1 microglial cells, CA2 neuronal cells, CA3 microglial and neuronal cells, and V1 microglial and astrocyte cells may be implicated in depressive pathology.

MDD patients typically exhibit smaller hippocampal volumes and compromised hippocampal function[15]. Depressed patients display phenomena including diminished neuron numbers, reduced neuronal branching, and inhibited neuronal generation within the hippocampus[16,17]. Antidepressant medication can facilitate neuronal growth and augment hippocampal volume[18]. The PrL brain region in mice corresponds to the dorsolateral prefrontal cortex (DLPFC) in humans, and the DLPFC in patients with MDD serves as a target for non-invasive brain stimulation therapy[19]. Visual cortex dysfunction plays a role in the pathophysiology and treatment of depression. MDD patients exhibit abnormal visual cortex function, and the effects of many antidepressants were aligned with the enhancement of visual cortex structure and synaptic function[20]. Our research revealed the cell subtype-specific expression of B4GALT6 in the hippocampus and V1 brain region of mice exhibiting depressive-like behaviors. While Western blot analysis indicated heightened B4GALT6 expression in the PrL brain region of these mice, no cell subtype specificity was noted, suggesting that B4GALT6 originating from microglia, astrocytes, and neurons in the PrL brain region collectively contributes to depression.

B4GALT6 is conserved across jawed vertebrates (Supplementary Figure 6) and exhibits widespread expression in human brain tissue (Supplementary Figure 7). B4GALT6 is essential for the synthesis of lactosylceramide, pivotal in neuronal generation and myelin formation[6,21]. B4GALT6 precisely regulates the ratio of gangliosides (such as GM1/GM3) and glycoprotein glycan chain structure by dynamic interaction with glycosphingolipid synthase (UGCG, ST3GAL5) and glycosidase (NEU1). This network imbalance leads to three pathological effects: (1) Synaptic dysfunction: GM1 deficiency destroys the lipid raft microdomain, resulting in abnormal localization of the TrkB receptor, weakening BDNF signal transmission and damaging synaptic plasticity of the prefrontal-hippocampus neural circuit; (2) Neuroinflammatory activation: NEU1-mediated sialidase hydrolysis increased, exposing galactose residues and activating the Galectin-3-TLR4 pathway, triggering the release of glial inflammatory factors (IL-6, TNF-α); and (3) HPA axis dysregulation: Glucolipid metabolism disorder in paraventricular nucleus of the hypothalamus leads to abnormal glycosylation modification of glucocorticoid receptor, which hinders the negative feedback regulation and ultimately leads to sustained stress response hyperactivity. Together, these mechanisms may contribute to the pathological progression of MDD.

Additionally, B4GALT6 is implicated in numerous neurological and psychiatric diseases. Postmortem brain studies reveal reduced B4GALT6 expression in the prefrontal cortex of patients with schizophrenia compared to controls[22]. Loss of B4GALT6 alters developmentally timed Caenorhabditis elegans sleep[23]. In multiple sclerosis, brain tissue analysis revealed elevated levels of B4GALT6 and its product lactosylceramide compared to controls[24,25]. The lactosylceramide produced by B4GALT6 enhances CCL2 expression via the NF-κb and IRF-1 signaling pathways, which in turn activates astrocytes through an autocrine manner, exacerbating conditions in experimental autoimmune encephalomyelitis[24,25]. Additionally, B4GALT6 regulates microglia activation and plays a role in central nervous system inflammation by enhancing astrocyte synthesis of GM-CSF[24,25]. B4GALT6 expression is elevated in the spinal cords of mice with adrenomyeloneuropathy compared to controls[26]. Bioinformatic analysis shows that B4GALT6 is a hub gene for pituitary prolactinoma[27]. To our knowledge, this is the first report that B4GALT6 is involved in depression.

We utilized the gene and protein function prediction tools GeneMANIA[28] and STRING[29] to predict the function of B4GALT6 (Supplementary Figure 8). At the genetic and protein levels, B4GALT6 functions through other glycosyltransferases and glycosidases. Further research is needed to determine whether B4GALT6 participates in the pathophysiological process of MDD through the aforementioned molecules. The elucidation of the B4GALT6 mechanism will shift MDD treatment from 'monoamine coverage' to 'precise regulation of the glycolipid-immune network'. This shift can be achieved through three key advancements: Diagnostic classification, targeted drugs, and delivery technologies. These advancements aim to address the shortcomings of current therapies and reduce relapse rates. This represents not only a technological advancement but also a fundamental paradigm shift in depression treatment.

Our study has several limitations. First, the absence of systematic functional genomics experiments has hindered a full elucidation of how rs372369000 affects B4GALT6 expression. Second, the absence of knockout and conditional knockout mice hinders a more detailed elucidation of the brain region and cell specificity of B4GALT6 in the pathology of MDD. Third, this study did not explore the specific mechanism of B4GALT6 leading to MDD and its effect on cognitive function. Finally, this research serves as a pilot study, and future research should delve into the specific mechanisms of B4GALT6 in models like chronic unpredictable mild stress (CUMS), chronic social defeat stress (CSDS), and chronic restraint stress (CRS). In addition, future studies should comprehensively assess behavioral phenotypes through cognitive function tests such as the Y-maze and the Morris water maze and explore the impact of B4GALT6 Levels on the cognitive functions of brain regions.

CONCLUSION

In summary, we have reported the brain region and cell-specific characteristics of B4GALT6 expression in mice exhibiting depressive-like behaviors. These findings may broaden our understanding and help develop new potential targets for diagnosis and treatment of MDD.