Role of CCM3 in lead-induced neurological damage in diabetes

Abstract

BACKGROUND

China has the largest number of diabetes patients, and nerve injury is an important complication of diabetes. Lead is a residual environmental pollutant with neurotoxicity and endocrine effects.

AIM

To investigate the effects of lead exposure, diabetes, and CCM3 deficiency on neurovascular damage using CCM3^+/- diabetic mice exposed to lead and samples from diabetic patients. The study aims to elucidate the mechanisms by which lead exposure, diabetes, and CCM dysfunction contribute to neurovascular injury and to provide a theoretical foundation for the prevention and treatment of neurological disorders associated with diabetes, lead exposure, and CCM3 gene defects.

METHODS

Using a CCM3^+/- diabetic mouse model exposed to lead, the Morris water maze test was performed to evaluate the effects of lead exposure, blood glucose, and CCM3 deficiency on learning and memory. Neurovascular injury was assessed by immunofluorescence co-localization staining of neural and vascular markers. Differentially expressed proteins and associated signaling pathways were identified through proteomic analysis, and key findings were validated by Western blotting. Glutathione (GSH) levels were measured to assess alterations in these factors and related proteins and metabolite levels in the hippocampus. Lastly, urinary lead and neurotransmitter metabolites were measured using inductively coupled plasma mass spectrometry (ICP-MS) and high-performance liquid chromatography (HPLC), respectively. Single-nucleotide polymorphisms (SNPs) in the CCM3 gene were analyzed using kompetitive allele-specific PCR (KASP) genotyping.

RESULTS

Compared with the control group, CCM3^+/- mice exposed to lead showed the highest blood glucose and blood lead levels. Additionally, diabetic mice exposed to lead exhibited significantly increased escape latency (P = 0.003). The fluorescence intensity of glial fibrillary acidic protein staining was significantly lower in the lead-exposed group compared with that in the control group (P = 0.005). Proteomic analysis identified significant changes in AKT, and Western blotting results showed that PI3K and AKT expression significantly decreased in the lead-exposed group (P = 0.009; P = 0.007). GPX4 expression decreased in both the lead-exposed and diabetic groups (P < 0.001). GSH levels in the hippocampus of the lead-exposed group was lower than the control group (P < 0.001). Finally, in diabetic patients, urine lead levels were weakly but positively correlated with vanillylmandelic acid (VMA) (r_s = 0.426, P < 0.001) and homovanillic acid (HVA) (r_s = 0.410, P < 0.001) in both the low and high urine lead groups. Additionally, there was an interaction between urine lead, blood glucose, and the rs6784267 locus on VMA (F = 9.838, P < 0.001) and HVA (F = 4.788, P = 0.003).

CONCLUSION

In our study, lead exposure, diabetes and the CCM3 gene may jointly increase blood glucose and lead levels, thereby affecting both vascular and neural function, with neural damage occurring earlier than vascular damage. In this process, inhibition of the PI3K-AKT pathway appears to play a crucial regulatory role in ferroptosis. Furthermore, the interaction among urine lead levels, blood glucose, and the rs6784267 locus of the CCM3 SNP was confirmed in diabetic patients.

Key Words: CCM3; Lead; Diabetes; PI3K-AKT; Ferroptosis

Core Tip: Using a diabetic, CCM3-deficient mouse model with lead exposure, this study demonstrates that lead exposure, diabetes, and CCM3 deficiency synergistically increase blood glucose and lead levels, resulting in neural injury that precedes vascular damage. Lead triggers early neural impairment by inhibiting the PI3K-AKT signaling pathway and promoting ferroptosis, a mechanism distinct from glucose-related effects. Epidemiological analyses further confirm that interactions among urinary lead levels, blood glucose, and the CCM3 single-nucleotide polymorphism rs6784267 significantly influence the neurotransmitter metabolites vanillylmandelic acid and homovanillic acid, consistent with animal findings. Collectively, these results provide a scientific basis for the prevention of early neural damage involving environmental, metabolic, and genetic factors.

INTRODUCTION

Lead is an important cumulative toxic heavy-metal environmental pollutant. Currently, the persistent impact of lead exposure on global cardiovascular health is comparable to the cumulative effect of PM2.5 pollution[1]. In recent years, China has implemented stringent measures to control environmental heavy-metal pollution, resulting in continuous improvements in the lead burden across various populations. The prevalence of lead poisoning in children has decreased from 33.80% in 1994 to 12.31%, and the average blood lead level in 220 cities has been maintained at 5.06 ± 1.36 μg/dL. From 2013 to 2017, the average blood lead concentration among 294402 adult patients in China was 28.36 μg/L[2]. However, lead production and consumption in China continue to grow steadily. In 2023 alone, lead production increased by 12.2%. These data indicate that baseline lead exposure remains prevalent in China and that the population continues to face a substantial long-term risk of lead exposure. The nervous system is particularly vulnerable to lead toxicity and therefore remains a major focus of long-term prevention and control efforts.

Diabetes is one of the most prevalent chronic diseases worldwide and is among the top ten causes of death globally. According to the International Diabetes Federation, the number of people with diabetes is projected to reach 783 million by 2045. China has one of the highest prevalences of diabetes, at 10.6%, ranking first in the world. At the same time, China’s healthcare expenditure on diabetes is second only to that of the United States, reaching $165.3 billion. Diabetes is a chronic noncommunicable metabolic disease caused by defects in insulin secretion and/or action that are influenced by both genetic and environmental factors. It can lead to serious complications, including diabetic encephalopathy, kidney damage, and retinopathy[3].

Programmed cell death 10 (PDCD10), also known as CCM3, is the third pathogenic gene associated with cerebral cavernous malformations (CCMs). CCM3 is an important environmental response gene expressed during critical developmental windows of both the vascular and nervous systems during embryogenesis and may represent a key genetic factor underlying neurodegeneration induced by lead exposure[4]. Due to its diverse biological functions, CCM3 deficiency may exacerbate the vascular and neural damage caused by lead exposure. One potential mechanism is its role as a scaffold (anchoring) protein that interacts with a variety of binding partners. CCM3, which has diverse biological functions, is an important member of protein complexes closely related to various serious human diseases, such as CCM, STRIPAK, and GCKIII[5]. The reason why CCM3 deficiency can enhance the effects of lead on blood vessels and nerves may be its ability to bind different types of proteins as an anchoring protein. CCM3 can regulate the expression of key proteins such as PP2A through STRIPAK, which is closely related to diabetes. Therefore, we hypothesize that this gene can respond rapidly to environmental lead exposure and is closely related to diabetes. Thus, CCM3 deficiency may be one of the risk factors for neuropathy under conditions of lead exposure and hyperglycemia.

In summary, based on the above analysis, this study aims to investigate the roles of high blood glucose as an endogenous disease factor, lead as an exogenous environmental factor, and CCM3 as a genetic susceptibility factor in triggering neurological injury in a CCM3^+/- diabetic mouse model exposed to lead. The goal is to provide scientific evidence for the prevention and treatment of neurological disorders caused by lead exposure under diabetic conditions and by CCM3 gene defects.

MATERIALS AND METHODS

Animal and group

SPF-grade CCM3 heterozygous mice (CCM3^+/-) were obtained from Zhongshan University, and SPF-grade C57BL/6J mice were purchased from Hunan Slake Jingda Company (production license number SCXK [Xiang] 2019-0004). Animal protocols were approved by the IACUC of Guilin Medical University (GLMC-IACUC-20241049). Mice were used for subsequent experiments at 8-10 weeks of age and a body weight of 20 ± 2 g. All mice were housed in SPF-grade animal facilities at a temperature of 24 ± 2 °C, and a humidity of 40%-60%, under a 12-hour light/dark cycle, with free access to water and food.

Both CCM3^+/- and wild-type (WT) male mice were randomly allocated to control, lead exposure, diabetes, and diabetes with lead exposure groups (n = 7). Mice in the diabetes and diabetes with lead exposure groups were injected with streptozotocin (STZ), whereas mice in the other groups were inected with 0.01 mmol/L sodium citrate for 5 consecutive days. Fasting blood glucose (FBG) was measured 72 hours after the final injection. An FBG level > 11.1 mmol/L was considered indicative of successful induction of diabetes.

Mice in the diabetes with lead exposure group and the lead exposure group were provided with 0.5% (5 mg/mL) lead acetate solution in their drinking water, whereas the control and diabetes groups received sterilized water. The experiment was terminated after 14 weeks of continuous lead exposure through drinking water.

Study population and group

The study population was recruited from the Department of Endocrinology of a hospital in Guilin City. All participants provided informed consent, and the study was approved by the relevant ethics committee before enrollment (Population Ethics Committee No. GLYY2021074).

Inclusion criteria: Patients diagnosed with diabetes in the endocrinology department of the hospital during the study period who had good independent functional ability and were willing to participate in and cooperate with the study were included.

Exclusion criteria: (1) Patients with a history of type 1 diabetes, pancreatitis, proliferative diabetic retinopathy, or macular disease; (2) Patients with impaired renal function, defined as an estimated glomerular filtration rate < 45 mL/minute/1.73 m²; (3) Patients with gestational diabetes, owing to its distinct diagnostic and management criteria; (4) Patients with long-term use of corticosteroids, antipsychotics, or antidepressants, as well as those with severe comorbidities or vascular complications; (5) Patients who failed to provide qualified biological samples; and (6) Patients with missing information in their electronic medical records. The screening process is detailed in Figure 1.

Figure 1 The blood glucose levels of mice in each group at four time points. WT: Wild-type.

After normality testing, the urinary lead level data were found to have a skewed distribution, with a median of 1.30 μg/L. Therefore, the study population was divided into a high urinary lead group and low urinary lead group using the 75^th percentile (P75; 2.13 μg/L) as the cutoff.

Relevant patient information, including age, sex, height, weight, and FBG level, was obtained from the electronic medical records of patients with diabetes. Body mass index (BMI) was calculated using height and weight according to the formula: BMI = weight (kg)/height² (m²).

Instruments and reagents

STZ (Solable, China), Quick Genotyping Assay Kit for Mouse Tail (Beyotime, China), β-actin antibody (Proteintech, China), PDCD10/CCM3 antibody (Abcam, United Kingdom), GPX4 antibody (Abcam, United Kingdom), PI3K antibody (Abcam, United Kingdom), AKT1 + AKT2 + AKT3 antibody (Abcam, United Kingdom), CD31 antibody (Invitrogen, United States), glial fibrillary acidic protein (GFAP) antibody (Cell Signaling Technology, United States), reduced glutathione colorimetric assay kit (Elabscience, China), inductively coupled plasma mass spectrometer (ICP-MS, PerkinElmer, United States), DigBehv Animal Behavioral Analysis System (Jiliang, China), FluorChem M System (ProteinSimple, United States), CM1950 cryostat microtome (Leica, Germany), inverted fluorescence microscope (Olympus, Japan), liquid chromatography system (Agilent, United States), and real-time PCR system (Thermo Fisher, United States).

Mice and human sample collection and preservation

For mice, blood was collected from the abdominal aorta into EDTA anticoagulant tubes following anesthesia. Whole brains were collected on ice, and a portion was fixed in 4% paraformaldehyde and stored at 4 °C for approximately 12 hours. Subsequently, the brains were transferred to a 20% sucrose solution for dehydration, and the brain tissues were dissected and embedded in optimal cutting temperature compound. Another portion of the brain was dissected to obtain hippocampal tissue. All of the aforementioned samples were stored at -80 °C.

For human subjects, fasting peripheral blood samples (≥ 2 mL) were collected into vacutainer tubes containing EDTA anticoagulant and centrifuged to separate plasma, which was then stored at -80 °C. Urine samples (≥ 5 mL) were collected from the target population into sealed polyethylene centrifuge tubes and stored at -80 °C within 12 hours of collection.

FBG level measurement in mice

FBG levels were measured in mice at four time points during the experiment: 0, 4, 8, and 14 weeks. Prior to each experiment, the mice were fasted overnight for 12 hours with free access to water. The following day, the tips of the tail were punctured after disinfection with 75% ethanol. A drop of blood was collected and used to measure the FBG level using a glucose meter.

Morris water maze experiment

The experiment was conducted in a quiet, dimly lit, and temperature-controlled environment, with the platform consistently positioned at the center of the fourth quadrant. The swimming path and the time required to find and climb onto the platform, known as the escape latency, were recorded. If a mouse failed to locate the platform within 60 seconds, it was manually guided to the platform and allowed to remain there for 10 seconds, and the escape latency was recorded as 60 seconds. Each mouse underwent three training trials per day, with a 20-minute interval between trials, for a total of 5 days.

ICP-MS method for the detection of lead levels in blood and urine

Fifty microliters of collected mouse red blood cells were diluted 30-fold with 1% nitric acid containing 0.05% Triton X-100 and 70 μL of collected human urine was diluted 30-fold with 1% nitric acid. The samples were mixed thoroughly and centrifuged at 12000 rpm for 25 minutes. After centrifugation, 1 mL of the supernatant was collected for analysis. A standard curve was prepared. Lead standard stock solutions were accurately measured and diluted with 1% nitric acid containing 0.05% Triton X-100 (for human urine samples, 1% nitric acid without Triton X-100 was used) to prepare standard solutions with lead concentrations of 0.5, 2, 5, 10, 50, 100, and 200 μg/L. These solutions were then analyzed using an ICP-MS.

Detection conditions: Sweeps, 20; readings, 1; replicates, 3; helium kinetic energy discrimination mode; helium flow rate, 4.

Immunofluorescence in whole brain of mice

Brain tissues were sectioned at a thickness of 10 μm using a cryostat microtome. The sections were washed with phosphate-buffered saline (PBS), treated with 0.3% Triton X-100 at room temperature for 10 minutes, and then blocked with 5% BSA solution at room temperature for 1 hour. Subsequently, the sections were incubated overnight at 4 °C with primary antibodies against CD31 (1:3200 dilution) and GFAP (1:500 dilution). After washing with PBS, the corresponding fluorescent secondary antibodies (1:500 dilution) were applied and incubated at room temperature in the dark for 1 hour. Following another PBS wash, the sections were counterstained with DAPI for 5 minutes to stain the cell nuclei. Finally, the sections were observed and photographed using an inverted fluorescence microscope.

Proteomic analysis of mouse hippocampal tissue

Mouse hippocampal samples from each group were lysed with an appropriate amount of SDT lysis buffer and homogenized using an electric tissue grinder. The homogenates were then incubated in a metal bath at 100 °C for 3 minutes and subsequently centrifuged at 12000 rpm for 20 minutes at 4 °C. The supernatant was collected, and the protein concentration was determined using the BCA method. All mass spectrometry data were processed using Spectronaut 17 software. A library-free data independent acquisition (DIA) method (directDIA) was employed to perform database searching and protein quantification in a single step. The database used was Uniprot Mus musculus (mouse)[10090]-88534-20230417.fasta.

Western blot analysis of mouse hippocampal tissue

Mouse hippocampal tissue samples were placed in centrifuge tubes and lysed with tissue lysis buffer (containing PMSF) at a concentration of 100 mg/mL. After centrifugation, the supernatant was collected, and the protein concentration was determined using the BCA method. The samples were denatured in a metal bath at 100 °C, followed by electrophoresis, membrane transfer, and blocking. Primary antibodies against PDCD10/CCM3 (1:500), GPX4 (1:1000), PI3K (1:1000), and AKT (1:10000) were then added and incubated overnight at 4 °C. The PVDF membrane was washed with PBST and then incubated with the corresponding secondary antibodies (1:5000). A chemiluminescent substrate was applied to the PVDF membrane, followed by imaging using a fully automated multicolor fluorescence, chemiluminescence, and visible light imaging system. Grayscale quantification was performed using ImageJ software (1.8.0).

Detection of glutathione levels in mouse hippocampal tissue

Mouse hippocampal tissue was homogenized in an appropriate amount of PBS using a tissue grinder. After centrifugation, the supernatant was collected, and the protein concentration was determined using the BCA method. Following the manufacturer’s instructions, the supernatant was used to measure the optical density (OD) of reduced glutathione (GSH) at 405 nm using a microplate reader. Finally, the GSH content in the hippocampal tissue was calculated based on the obtained OD values.

Detection of neurotransmitter metabolite levels in human urine by high-performance liquid chromatography

Human urine samples were filtered through a 45 μm filter membrane before analysis. Mixed standard solutions of homovanillic acid (HVA), vanillylmandelic acid (VMA), 3,4-dihydroxyphenylacetic acid (DOPAC), and 5-hydroxyindole acetic acid (5-HIAA) were prepared using deionized water. The concentrations of VMA and HVA were set at 0.1, 0.25, 0.5, 1.0, and 2.0 μg/mL, while the concentrations of 5-HIAA and DOPAC were set at 0.05, 0.125, 0.25, 0.5, and 1.0 μg/mL. These solutions were analyzed using high-performance liquid chromatography. The chromatographic conditions were as follows: Column, 5 μm, 4.6 mm × 250 mm; mobile phase, methanol-0.1 mol/L phosphate buffer (5: 95, v/v); flow rate, 1.0 mL/min; detection wavelength, excitation (Ex) 280 nm and emission (Em) 315 nm; column temperature, 30 °C; injection volume, 10 μL; and total analysis time, 20 minutes.

Detection of single-nucleotide polymorphism mutations in the CCM3 gene in peripheral blood DNA of the human population using the kompetitive allele-specific PCR assay

Peripheral blood genomic DNA was extracted using the red blood cell lysis method. Based on genome-wide association study (GWAS) analysis and previous studies conducted by our research group, three single-nucleotide polymorphism (SNP) loci in the CCM3 gene were selected for detection using the kompetitive allele-specific PCR (KASP) assay. The amplification reaction system had a total volume of 10 μL, including 5 μL of FLU-ARMS 2 × PCR Mix, a mixture of three primers (F1/F2/R primer concentrations: 10 μmol/L), and 2.5 μL of DNA, with water added to bring the final volume to 10 μL. The amplification conditions were as follows: 30 °C for 1 minute and 95 °C for 10 minutes; 95 °C for 15 seconds and 61 °C for 1 minute for 10 cycles (with a decrease of 0.6 °C per cycle); 95 °C for 15 seconds and 55 °C for 1 minute for 35 cycles; and finally, 30 °C for 1 minute to read the fluorescence signal and record the results.

Statistical analysis

Data analysis was performed using SPSS version 28.0 statistical software. For normally distributed data, results were expressed as mean ± SD. For non-normally distributed data, the Mann-Whitney U test was used, and results were presented as the median and interquartile range. Independent sample t-tests or χ² tests were used for comparisons between two groups. For comparisons among multiple groups, one-way analysis of variance (ANOVA) and repeated measures ANOVA were used. If the interaction effect was not significant, the main effects were analyzed; if the interaction effect was significant, post-hoc analysis was conducted using the least significant difference (LSD) method to examine simple effects. The Hardy-Weinberg equilibrium test was used to assess the distribution balance of the three CCM3 gene loci in the study population. Spearman rank correlation analysis was used to examine correlations for data conforming to a normal distribution, and linear regression models were used to analyze interaction effects. The significance level was set at α = 0.05.

RESULTS

FBG levels in mice

Results from repeated-measures ANOVA showed differences in blood glucose levels across different time points (P < 0.001). At weeks 4, 8, and 14, the mean blood glucose levels in the WT diabetic group were significantly higher than those in the WT control group (P < 0.001), indicating successful induction of diabetes in the mice. The results suggested that lead exposure alone had no effect on blood glucose levels but exacerbated the increase in blood glucose levels under diabetic conditions.

Furthermore, comparison between different genotypes within each group revealed that, at week 8, blood glucose levels in CCM3^+/- mice were lower than those in WT mice in the diabetic group (P = 0.016). In the diabetic lead-exposed group, blood glucose levels in CCM3^+/- mice were significantly higher than those in WT mice (P = 0.037), and the mean blood glucose level in the CCM3^+/- diabetic lead-exposed group was higher than that in other groups. Multifactorial ANOVA of blood glucose levels at week 8 showed an interaction among genotype, diabetes, and lead exposure (P = 0.024). These results suggest that blood glucose levels are unaffected by CCM3 gene deficiency alone. However, when diabetes and lead exposure coexist, the CCM3 eficiency exacerbates the increase in blood glucose levels (Table 1 and Figure 1).

Table 1 Multivariate analysis of variance results of blood glucose in mice at week 8.

Factors	F value	P value
Genotype	0.157	0.694
Diabetes	7.591	0.009
Pb exposure	268.756	< 0.001
Genotype × diabetes	5.131	0.030
Genotype × Pb	0.001	0.974
Diabetes × Pb	7.726	0.009
Genotype × diabetes × Pb	5.619	0.024

Pb: Lead.

Mouse blood lead levels

Main effects analysis showed that lead exposure and diabetes significantly affected blood lead levels in mice (P < 0.001), whereas the effect of CCM3 gene was not significant (P > 0.05). There was an interaction between diabetes and lead exposure on blood lead levels in mice (P < 0.001). Post-hoc LSD analysis revealed that blood lead levels in both genotypes were significantly higher in the lead-exposed group compared than in the corresponding control groups (P < 0.001). Additionally, blood lead levels in both genotypes of the diabetic lead-exposed group were significantly higher than those in the corresponding lead-exposed group of the same genotype (P < 0.001). Furthermore, blood lead levels CCM3^+/- mice in the diabetic lead-exposed group were the highest among all groups. These results indicate that diabetes increases blood lead levels and that CCM3 deficiency further increases blood lead levels when diabetes and lead exposure coexist (Table 2 and Figure 2).

Figure 2 The blood lead levels of mice in each group. ^aP < 0.05. WT: Wild-type.

Table 2 Multivariate analysis of variance results of blood lead levels in each group of mice.

Factors	F value	P value
Genotype	0.002	0.964
Diabetes	23.619	< 0.001
Pb exposure	198.072	< 0.001
Genotype × diabetes	0.499	0.486
Genotype × Pb	0.006	0.937
Diabetes × Pb	20.556	< 0.001
Genotype × diabetes × Pb	0.415	0.525

Pb: Lead.

Morris water maze experiment

Repeated measures ANOVA revealed differences in escape latency among mice at different time points (P < 0.001), with escape latency gradually decreasing over successive days. By the fifth day of the experiment, mice in the control group exhibited concise and purposeful swimming trajectories, whereas mice in the diabetic lead-exposed group showed chaotic and longer trajectories without a clear purpose. The escape latency of mice in the diabetic lead-exposed group was significantly higher than that of mice in the control group (P = 0.003). Multifactorial ANOVA of escape latency on the fifth day revealed significant effects of diabetes and lead exposure on mouse escape latency (P = 0.002; P = 0.016, respectively). These results suggest that both diabetes and lead exposure are associated with longer escape latencies, with the longest latency observed in mice exposed to lead under diabetic conditions. However, no significant effect of CCM3 gene deficiency on mouse escape latency was observed (Table 3 and Figure 3).

Figure 3 Results of water maze experiment on mice in each group. A: Representative diagram of swimming trajectory of mice on day 5; B: Escape latency of mice on day 5. ^aP < 0.05 vs wild-type control group. WT: Wild-type.

Table 3 Multivariate analysis of variance results on the escape latency of mice on day 5.

Factors	F value	P value
Genotype	0.056	0.814
Diabetes	9.999	0.002
Pb exposure	6.091	0.016
Genotype × diabetes	0.033	0.856
Genotype × Pb	0.009	0.923
Diabetes × Pb	0.729	0.397
Genotype × diabetes × Pb	0.001	0.998

Pb: Lead.

Immunofluorescent staining of CD31 and GFAP in the mouse hippocampal region

To further confirm neural damage in the brain and its relationship with vascular changes, we performed immunofluorescent co-staining of CD31 and GFAP, markers for endothelial cells and astrocytes in the hippocampal region, respectively. Main effects analysis revealed an interaction between genotype and lead exposure on GFAP expression levels (P = 0.025). Post-hoc LSD analysis showed that, in WT mice, GFAP levels were lower in the lead-exposed group than in the control group (P = 0.005), while in the diabetic lead-exposed group, GFAP levels were significantly lower than those in the diabetic group (P = 0.005). Regarding genotype, GFAP levels in CCM3^+/- mice in the control and diabetic groups were lower than those in WT mice (P = 0.026). These results suggest that lead exposure alone or CCM3 gene deficiency may decrease GFAP expression, whereas diabetes alone has no effect on GFAP expression No interaction among the different factors was observed (Figure 4).

Figure 4 Fluorescence co-localization results of CD31 and glial fibrillary acidic protein in the hippocampus of mice in each group. A: Expression of CD31 and glial fibrillary acidic protein (GFAP) in the hippocampus of mice in each group, B: Average fluorescence intensity of CD31, C: Average fluorescence intensity of GFAP. ^aP < 0.05 vs wild-type control group, ^cP < 0.05 vs wild-type diabetes group. WT: Wild-type.

Proteomic analysis of mouse hippocampal tissue

To elucidate the key regulatory protein changes induced by lead exposure, diabetes, and the CCM3 gene, we conducted proteomic analysis of hippocampal tissue from mice in each group. In the analysis of significantly different proteins, we identified a total of 540 differentially expressed proteins between the WT control group and the CCM3^+/- diabetic group, including 287 upregulated and 253 downregulated proteins. Between the CCM3^+/- control group and the CCM3^+/- diabetic group, 183 differentially expressed proteins were identified, with 121 upregulated and 62 downregulated proteins. Furthermore, between the WT diabetic lead-exposed group and the CCM3^+/- diabetic group, 258 differentially expressed proteins were identified, including 141 upregulated and 117 downregulated proteins. Notably, AKT was identified as a common differentially expressed protein among these comparison groups. AKT is a key regulatory protein in the PI3K-AKT pathway, which is implicated in the regulation of ferroptosis (Figure 5).

Figure 5 Differential protein volcano and AKT protein abundance difference. A: Differential protein volcano between wild-type (WT) control group and CCM3^+/- diabetes group; B: Differential protein volcano between CCM3^+/- control group and CCM3^+/- diabetes group; C: Differential protein volcano between WT diabetes lead exposed group and CCM3^+/- diabetes group; D: AKT protein abundance difference between WT control group and CCM3^+/- diabetes group; E: AKT protein abundance difference between CCM3^+/- control group and CCM3^+/- diabetes group; F: AKT protein abundance difference between WT diabetes lead exposed group and CCM3^+/- diabetes group.

Expression levels of CCM3, GPX4, PI3K, and AKT proteins in mouse hippocampal tissue

Based on the key regulatory proteins identified through proteomic analysis of hippocampal tissue from each group, we validated the expression levels of these proteins using western blot analysis. Main effects analysis revealed significant effects of genotype, diabetes, and lead exposure on CCM3 expression levels (P < 0.001; P < 0.001, P = 0.001, respectively), with a significant interaction among these three factors (P < 0.001). Regarding GPX4 expression levels, diabetes and lead exposure showed significant effects (P = 0.001; P < 0.001, respectively), with significant interactions observed between genotype and diabetes, genotype and lead exposure, and diabetes and lead exposure (P = 0.016; P = 0.013; P < 0.001, respectively). For PI3K expression levels, there were significant effects of genotype and lead exposure (P = 0.014; P = 0.009, respectively). As for AKT expression levels, there was a significant interaction between genotype and lead exposure (P = 0.010).

LSD post-hoc comparisons revealed that CCM3 expression levels in WT mice in the lead-exposed and diabetes groups were significantly lower than those in the control group (P < 0.001). Within the control group, CCM3 expression in CCM3^+/- mice was significantly lower than that in WT mice (P < 0.001). GPX4 expression levels in WT mice in both the lead-exposed and diabetes groups were lower than the control group (P < 0.001), and the diabetes-lead exposed group was lower than those in the lead-exposed group (P < 0.001). In the lead-exposed and diabetes groups, CCM3^+/- mice exhibited higher GPX4 expression levels than WT mice (P < 0.001). For PI3K expression levels, the lead-exposed group showed significantly lower expression levels than the control group in WT mice (P < 0.009), and the diabetes-lead-exposed group showed lower expression levels than the diabetes group (P < 0.009). Within the same treatment group, CCM3^+/- mice exhibited lower expression levels than WT mice (P < 0.014). For AKT expression levels, WT mice in the lead-exposed group exhibited significantly lower expression levels than those in the control group (P < 0.007), and the diabetes-lead-exposed group showed lower expression levels than the diabetes group (P < 0.007). However, in the lead-exposed and diabetes-lead-exposed groups, CCM3^+/- mice exhibited higher expression levels than WT mice (P < 0.015) (Figure 6).

Figure 6 Western blot results of hippocampal tissue of mice in each group. A: Expression of CCM3 and GPX4 in the hippocampus of mice in each group; B: Expression of PI3K and AKT in the hippocampus of mice in each group. ^aP < 0.05. WT: Wild-type.

GSH levels in mouse hippocampal tissue

To further determine the role of ferroptosis under the influence of lead exposure, diabetes, and the CCM3 gene, we measured GSH levels in mouse hippocampal tissue. Main effects analysis showed that genotype, diabetes, and lead exposure had significant effects on GSH levels (P < 0.001), and significant interactions were observed among these three factors.

LSD post-hoc comparisons revealed that, compared with the control group, GSH levels decreased in WT mice in the lead-exposed group but increased in the diabetes group (P < 0.001). The diabetes-lead-exposed group had higher GSH levels than the lead-exposed group but lower GSH levels than the diabetes group (P < 0.001). Within each treatment group, GSH levels in CCM3^+/- mice were lower than those in WT mice (P < 0.001). These results suggest that lead exposure decreases GSH levels, whereas diabetes increases GSH levels. CCM3 gene deficiency also reduces GSH levels (Table 4 and Figure 7).

Figure 7 Glutathione levels in the hippocampus of mice. ^aP < 0.05. WT: Wild-type.

Table 4 Multivariate analysis of variance for glutathione levels in the hippocampus of mice.

Factors	F value	P value
Genotype	80.520	< 0.001
Diabetes	21.955	< 0.001
Pb	228.467	< 0.001
Genotype × diabetes	8.943	0.009
Genotype × Pb	18.901	< 0.001
Diabetes × Pb	6.684	0.020
Genotype × diabetes × Pb	1.760	0.203

Pb: Lead.

General characteristics of patients with type 2 diabetes mellitus and different urinary lead levels

Data were collected from a total of 1017 diabetic patients in the endocrinology department of a hospital in the study area. After screening according to the inclusion and exclusion criteria, information and biological samples from 454 patients were included in the analysis. According to the grouping method described outlined earlier, the study population was divided into a low urinary lead group (341 individuals) and a high urinary lead group (113 individuals). The general characteristics of the two groups were analyzed. The results showed significant differences between the groups in terms of age, urinary lead levels, and the levels of neurotransmitter metabolites VMA, DOPAC, and HVA. In addition, significant differences were observed in urine biomarkers that were not standardized to creatinine (Figure 8 and Table 5).

Figure 8  Research population screening process.

Table 5 General characteristics of different lead groups in urine.

Characteristic	Group		P value
	Low urinary lead group (n = 341)	High urinary lead group (n = 113)
Age (year, mean ± SD)	56.20 ± 12.80	60.91 ± 12.04	0.007
Sex, n (%)			0.202
Male	174 (71.31)	70 (28.69)
Female	121 (77.07)	36 (22.93)
Urinary lead levels (μg/L, mean ± SD)	1.02, 0.86	3.19, 1.81	< 0.001
Fasting blood glucose (mmol/L, M, IQR)	5.99, 3.05	6.41, 3.99	0.837
BMI (kg/m2, M, IQR)	24.10, 4.31	23.67, 4.22	0.059
VMA (μg/mL, M, IQR)	1.17, 1.91	3.17, 4.53	< 0.001
5-HIAA (μg/mL, M, IQR)	1.90, 4.10	1.75, 2.16	0.808
DOPAC (μg/mL, M, IQR)	6.76, 9.43	7.63, 9.43	0.034
HVA (μg/mL, M, IQR)	1.36, 1.89	3.68, 5.00	< 0.001

BMI: Body mass index; VMA: Vanillylmandelic acid; HVA: Homovanillic acid; DOPAC: 3,4-dihydroxyphenylacetic acid; 5-HIAA: 5-hydroxyindole acetic acid; IQR: Interquartile range.

Hardy-Weinberg equilibrium test

After genotyping the three CCM3 gene loci rs9818496, rs3804610, and rs6784267 using the KASP method, the genotype frequency distribution of these loci in the study population were assessed using the Hardy-Weinberg equilibrium test. The results revealed that rs9818496 and rs3804610 did not conform to Hardy-Weinberg equilibrium (P < 0.05), indicating an imbalance in their distribution within the population. However, rs6784267 conformed to Hardy-Weinberg equilibrium (P > 0.05), indicating a balanced distribution within the study population and satisfactory representativeness.

Correlation analysis of urinary lead, blood glucose, and the CCM3 gene locus rs6784267 with four neurotransmitter metabolites

The results indicated that although there was a statistically significant correlation between DOPAC and urinary lead (P = 0.010), the correlation coefficient (0 < |r_s = 0.141| < 0.3) indicated no significant correlation. However, VMA and HVA show statistically significant correlations with urinary lead levels (P < 0.001), with correlation coefficients of 0.3 ≤ |r_s = 0.426; r_s = 0.410| < 0.5, indicating weak positive correlations between VMA, HVA, and urinary lead levels. However, the clinical significance of these correlations remains limited.

Although statistically significant correlations were observed between VMA and blood glucose levels (P = 0.011) and HVA and blood glucose levels (P = 0.003), the correlation coefficients were 0 < |r_s = 0.140; r_s = 0.161| < 0.3, indicating no significant correlation. Similarly, although a statistically significant correlation was observed between VMA and rs6784267 (P = 0.046), the correlation coefficient was 0 < |r_s = -0.113| < 0.3, suggesting no significant correlation (Figure 9).

Figure 9 Heat map of correlation between urinary lead levels, blood glucose levels and rs6784267 locus with four neurotransmitter metabolites. ^aP < 0.05. VMA: Vanillylmandelic acid; HVA: Homovanillic acid; DOPAC: 3,4-dihydroxyphenylacetic acid; 5-HIAA: 5-hydroxyindole acetic acid.

Interaction effects of urinary lead, blood glucose and the rs6784267 locus on VMA, 5-HIAA, DOPAC, and HVA in the study population

A linear regression model was used to investigate the interaction effects of urinary lead, blood glucose, and the rs6784267 locus on VMA, 5-HIAA, DOPAC and HVA in the type 2 diabetes mellitus (T2DM) population. The results indicated that there were interaction effects between urinary lead and blood glucose levels, as well as between urinary lead and the rs6784267 locus, on both VMA and HVA (P < 0.001).

In addition, an interaction effect between blood glucose levels and the rs6784267 locus was observed only for VMA (P = 0.007). Furthermore, a significant three-way interaction effect among urinary lead, blood glucose levels and the rs6784267 locus was observed for both VMA and HVA (P < 0.001 and P = 0.003, respectively) (Table 6).

Table 6 Interaction of urinary lead levels, blood glucose levels and rs6784267 locus with four neurotransmitter metabolites.

Interactive factors	Target	F value	P value
Urinary lead × blood sugar	VMA	14.298	< 0.001
	5-HIAA	0.911	0.404
	DOPAC	2.693	0.070
	HVA	7.856	< 0.001
Urinary lead × rs6784267 locus	VMA	14.939	< 0.001
	5-HIAA	0.116	0.891
	DOPAC	2.274	0.105
	HVA	7.031	< 0.001
Blood sugar × rs6784267 locus	VMA	5.038	0.007
	5-HIAA	2.282	0.104
	DOPAC	0.792	0.454
	HVA	1.955	0.144
Urinary lead × blood sugar × rs6784267 locus	VMA	9.838	< 0.001
	5-HIAA	0.856	0.465
	DOPAC	1.984	0.118
	HVA	4.788	0.003

VMA: Vanillylmandelic acid; HVA: Homovanillic acid; DOPAC: 3,4-dihydroxyphenylacetic acid; 5-HIAA: 5-hydroxyindole acetic acid.

DISCUSSION

Lead is an environmental exposure factor, and diabetes is a pathological factor; both are closely related to the nerves and blood vessels of the brain. Moreover, CCM3 is a crucial genetic factor that is widely expressed in both vascular and neural tissues. However, there is limited research on the combined effects of these three factors. Therefore, this study aimed to investigate the effects and potential mechanisms of these factors in neural and vascular damage using a lead-exposed diabetic mouse model with CCM3 deficiency and patients with T2DM with different urinary lead levels .

As an endocrine-disrupting heavy metal, lead is closely linked to diabetes. First, multiple studies support the notion that environmental lead exposure can increase the incidence of diabetes. Monitoring of a population living in the same community for 20 years revealed a positive correlation between urinary lead levels reaching 2.0 μg/dL and the incidence of diabetes. Additionally, a study of 3053 residents in Wuhan, China, found a close correlation between urinary lead levels, FBG levels, and impaired FBG[6,7]. Second, both lead and diabetes have damaging effects on the nervous system. The central nervous system is an important target of lead toxicity. Lead can accumulate in brain tissue and result in neuronal degeneration and death through mechanisms such as mitochondrial dysfunction, apoptosis, oxidative stress, and epigenetic changes[8]. Long-term lead exposure, acting as a physiological accelerator[9], increases the risk of neurodegenerative diseases and promotes the development of pathological features associated with neurodegeneration[10]. At the same time, diabetic encephalopathy is an important central nervous system complication of diabetes that causes severe impairment of learning and memory functions[11]. Studies in diabetic rat models have shown significant hippocampal damage. These manifestations of diabetes-related nervous system injury may be related to changes in glucose metabolism[12], abnormal cerebral vascular structure[13], impaired insulin pathways[14], and other mechanisms.

The Morris water maze experiment was conducted in this study to assess the neurological damage caused by lead exposure, diabetes and CCM3 gene deficiency in mice. We observed that both the lead-exposed group and the diabetes group had longer escape latencies than the control group. Additionally, mice in the diabetes lead-exposed group had the longest escape latency, suggesting that the combined effect of diabetes and lead exposure on learning and memory impairment are more pronounced than those of either lead exposure or diabetes alone. We speculate that this may be related to the concurrent damage caused by lead exposure and diabetes to the nerves in the brain. Previous studies have indicated that lead exposure or diabetes can damage tissue structure in the hippocampal region of mice by increasing intercellular spaces, causing cell rupture, inducing cytoplasmic condensation, decreasing neuronal cell numbers, and inhibiting excitatory projection neural circuits in the hippocampus (such as the dentate gyrus-CA1 region), thereby impairing learning and memory abilities[15,16]. In this study, no significant effect of CCM3 gene deficiency on the learning and memory abilities of mice was observed. However, because only one method was used to evaluate neurological function in mice, further studies may be required to confirm the impact of the CCM3 gene.

To confirm the effects of lead exposure, diabetes, and CCM3 gene deficiency on the neural and vascular structures of the mouse hippocampus, we performed immunofluorescence co-localization analysis using CD31 and GFAP. CD31 is a marker of endothelial cells that plays an important role in angiogenesis and can serve as a marker of microvessels. Astrocytes are involved in learning and memory functions, and GFAP is a marker of astrocytes[17]. Abnormal activation or damage of astrocytes can cause changes in learning and memory function[18]. In our study, lead exposure, diabetes, and CCM3 gene deficiency did not significantly alter vascular distribution in brain tissue, whereas lead exposure and CCM3 gene deficiency individually reduced GFAP expression levels. Astrocytes are support cells in the nervous system that provide nutritional and metabolic support to neurons, maintain the stability of the brain environment, and regulate communication between neurons[19,20]. Therefore, a decrease in GFAP expression levels suggests damage to astrocytes, which can affect normal nervous system function. These findings indicate that astrocyte damage has already occurred in the mouse model, with lead exposure and CCM3 gene deficiency being the most influential factors. However, no interaction among lead exposure, diabetes, and CCM3 gene deficiency was observed. Based on the immunofluorescence co-localization staining results, we speculate that, under the exposure dose and duration used in this study, lead exposure and CCM3 gene deficiency induced early hippocampal neuronal damage without observable changes in vascular morphology or density at this stage. However, due to the limitations of the immunofluorescence co-localization staining experiment and the lack of further studies on the morphological structure of endothelial cells and astrocytes, it cannot be fully confirmed whether these three factors play a role in the relationship between blood vessels and neurons. Moreover, CCM3 is a regulatory gene with bidirectional effects, and its responses vary across different exposure windows and doses. Further studies are required to confirm these findings.

Although we observed potential effects of lead exposure, glycemic, and CCM3 gene deficiency on the nervous system, the immunofluorescence analysis revealed changes only in the quantified fluorescence intensity of astrocytes. Protein-level analyses may help further elucidate the mechanisms underlying these early changes. Therefore, we conducted proteomic analysis of mouse hippocampal tissue to identify key regulatory proteins responsible for these early alterations. We found that AKT was a common differentially expressed protein in several comparisons, including the WT control group vs CCM3^+/- diabetes group, the CCM3^+/- control group vs the CCM3^+/- diabetes group, and the WT diabetes lead-exposed group vs CCM3^+/- diabetes group. AKT is a crucial regulatory protein in the PI3-AKT signaling pathway, which is closely associated with ferroptosis. Therefore, we further examined the PI3K-AKT signaling pathway and ferroptosis-related factors in the hippocampus of mice of different genotypes in each group.

The PI3K-AKT signaling pathway is closely associated with various human diseases, including diabetes, neurodegenerative diseases, and ischemic brain injury, and plays a crucial role in many biological processes[21]. Additionally, the PI3K-AKT pathway regulates ferroptosis through multiple downstream targets. For instance, activation of the PI3K-AKT-mTORC1 signaling pathway can upregulate lipid synthesis mediated by the downstream sterol regulatory element-binding protein 1, thereby inhibiting ferroptosis in cancer cells[22-25]. In this study, we examined the expression of PI3K and AKT proteins in mouse hippocampal tissue and found that lead exposure inhibited the PI3K-AKT signaling pathway. However, diabetes did not have a significant effect on this pathway. CCM3 gene deficiency inhibited PI3K expression and exacerbated the inhibitory effect of lead exposure on PI3K. Inhibition of the PI3K-AKT pathway has been shown in many studies to induce ferroptosis[26,27]. Ferroptosis is a novel form of programmed cell death distinct from necrosis, apoptosis, pyroptosis, and autophagy, first proposed by Dixon et al[28] in 2012. Its mechanism is complex and involves depletion of reduced GSH, leading to decreased GPX activity, accumulation of lipid peroxides, and ultimately the induction of ferroptosis. Studies have shown that iron levels are negatively correlated with cognitive ability in patients with neurodegenerative diseases and that GPX4 has a protective effect in Alzheimer's disease mouse models[29]. Therefore, the PI3K-AKT pathway may play a key regulatory role in ferroptosis induced by the damaging effects of lead exposure and diabetes on the nervous system and vasculature.

Additionally, decreased GPX4 expression has been observed in diabetes complicated by microvascular complications. In our study, both lead exposure and diabetes decreased GPX4 protein expression levels, and the coexistence of these two factors further enhanced this decrease. CCM3 gene deficiency alone also caused a decrease in GPX4 protein expression; however, no combined effect was observed. Our results also showed that lead exposure decreased GSH levels, whereas diabetes increased GSH levels. The coexistence of these two factors may have offset their effects on GSH levels. CCM3 gene deficiency alone reduced GSH levels and enhanced the reduction in GSH levels caused by lead exposure.

To further validate the findings from the animal experiments, we analyzed the correlations and interaction effects of CCM3 gene SNP polymorphisms with urinary neurotransmitter metabolites based on blood glucose levels, urinary lead levels, and variations in CCM3 SNP loci in the diabetic population. This analysis was performed to determine whether the changes observed in the animal model were also reflected in humans.

HVA, VMA, DOPAC, and 5-HIAA are all metabolites of catecholamines, which play crucial roles in regulating physiological activities such as memory, learning, and mood. The levels of these metabolites can indirectly reflect catecholamine secretion in the body. Lead exposure or diabetes can damage the blood-brain barrier, allowing more peripheral catecholamines to enter the brain. As a compensatory response, more peripheral catecholamines may be secreted, resulting in an overall increase in catecholamine concentrations in the body[30]. It has been reposted that the levels of catecholamines or their metabolites are significantly elevated in the urine of children with increased blood lead levels, as well as in the brain tissue of diabetic rats. Our epidemiological findings revealed that the levels of the neurotransmitter metabolites VMA, DOPAC, and HVA were higher in T2DM patients with high urinary lead levels than in those with low urinary lead levels, and these differences were statistically significant (P < 0.05).

The abnormal elevation of these neurotransmitter metabolites indirectly indicates increased levels of monoamine neurotransmitters such as adrenaline, serotonin, and dopamine in the high-lead group. The abnormal release of monoamine neurotransmitters can spread to the blood, cerebrospinal fluid, and adjacent neurons, thereby damaging neurons and causing neurotoxicity. It also can cause cerebral vasoconstriction, reduce cerebral blood flow, exacerbate cerebral ischemia and hypoxia, and ultimately contribute to local pathological changes. We hypothesize that the elevated levels of neurotransmitter metabolites observed in the high-lead group in our study may exert neurotoxic effects and impair cognitive memory function; however, further investigation is needed.

Spearman’s rank correlation analysis also revealed a weak positive correlation between urinary levels of the neurotransmitter metabolites VMA and HVA and urinary lead levels in the study population. Furthermore, we explored the interaction effects of urinary lead levels, blood glucose levels, and the rs6784267 locus on 5-HIAA, DOPAC, VMA, and HVA using linear regression analysis. We found that these three factors exhibited interaction effects on the neurotransmitter metabolites VMA and HVA. These findings suggest that lead exposure, elevated blood glucose levels caused by diabetes, and CCM3 gene deficiency are all important factors contributing to neural damage in diabetic populations, which is consistent with the results observed in the mouse model.

The advantages of this study include, first, the demonstration of a correlation between lead exposure and diabetes in animal experiments. It was found that when lead exposure and hyperglycemia were present simultaneously, blood lead and blood glucose levels were significantly increased, whereas no significant changes in blood glucose or blood lead levels were observed with lead exposure or diabetes alone. Previous studies have focused on the ability of lead exposure to increase blood glucose levels, whereas our study demonstrated that the presence of diabetes also increased lead levels in vivo. Second, this study found a synergistic effect on neural injury when lead exposure and hyperglycemia coexisted; however, the mechanisms by which lead exposure and diabetes contribute to this injury appear to differ. The effects of CCM3 gene deficiency on neurological damage were more consistent with those of lead exposure, whereas its response to diabetes was less apparent. There are also some limitations to this study. For example, the methods used to analyze the effects and mechanisms of neural injury were limited. Moreover, the role of ferroptosis in lead- and hyperglycemia-induced injury requires further exploration.

CONCLUSION

In summary, our study, conducted using a mouse model with CCM3 gene deficiency, diabetes, and lead exposure, revealed that lead exposure, diabetes, and CCM3 gene deficiency affect the nervous system by increasing blood glucose and lead levels. Moreover, neural damage in the brain appears to precede vascular damage. The findings also indicated that lead exposure may induce early nerve damage by inhibiting the PI3K-AKT pathway, thereby triggering ferroptosis, whereas the mechanism of blood glucose-induced neural damage may differ. Finally, epidemiological studies in diabetic populations further confirmed that urinary lead levels, blood glucose levels, and the rs6784267 SNP of the CCM3 gene interact to affect the levels of the neurotransmitter metabolites VMA and HVA, potentially influencing the nervous system in a manner consistent with the findings from the mouse study. Our findings provide a scientific basis for the precise prevention and effective treatment of early neural damage caused by environmental exposures, internal disease factors, and genetic susceptibility.