The Potential of ANK1 to Predict Parkinson’s Disease

The main cause of Parkinson’s disease (PD) remains unknown and the pathologic changes in the brain limit rapid diagnosis. Herein, differentially expressed genes (DEGs) in the Gene Expression Omnibus (GEO) database (GSE8397 and GSE22491) were assessed using linear models for microarray analysis (limma). Ankyrin 1 (ANK1) was the only common gene differentially down-regulated in lateral substantia nigra (LSN), medial substantia nigra (MSN) and blood. Additionally, DEGs between high ANK1 and low ANK1 in GSE99039 were picked out and then uploaded to the Database for Annotation, Visualization and Integrated Discovery (DAVID) for gene ontology (GO) functional annotation analysis. GO analysis displayed that these DEGs were mainly enriched in oxygen transport, myeloid cell development and gas transport (biological process (BP)); hemoglobin complex, haptoglobin–hemoglobin complex and cortical cytoskeleton (cellular component (CC)); and oxygen transporter activity, haptoglobin binding and oxygen binding (molecular function (MF)). Receiver operating characteristic (ROC) curve analysis showed ANK1 had good diagnostic accuracy and increased the area under the curve (AUC) value when combined with other biomarkers. Consistently, intraperitoneal injection of 1-methyl-4-phenyl-1,2,3,6-tetrahydropy-ridi-ne (MPTP) in C57BL/6J mice reduced ANK1 mRNA expression in both substantia nigra and blood compared to the control group. Thus, ANK1 may serve as a candidate biomarker for PD diagnosis.


Introduction
Parkinson's disease (PD) is a common neurodegenerative disease with complex etiology and pathogenesis. Studies have shown that α-synuclein accumulation, dopaminergic neuron loss, inflammation, oxidative stress and mitochondrial damage are involved in the process of PD [1][2][3][4][5]. However, the exact mechanisms of PD remain obscure and current treatments have limited effectiveness [6,7]. Moreover, PD is a debilitating disorder and characterized by severe motor (e.g., resting tremor, rigidity and hypokinesia tremor) and non-motor symptoms (e.g., cognitive deterioration, hallucination and sensory abnormalities) [8,9]. These symptoms progressively worsen over time and increase family burden [10].
It has been confirmed that PD is a progressive, disabling neurodegenerative disorder, with cases of either sporadic or familial origin. Sporadic PD is a multifactorial disease caused by both environmental and genetic factors. In addition, many genes have been identified in familial PD [11]. Numerous genetic loci associated with PD have been discovered. For example, 23 genes have been correlated with monogenic forms of PD in the Online Mendelian Inheritance in Man resource. Genome-wide association studies have helped researchers identify >90 risk loci associated with PD [12,13]. Studies from the Polish population have shown that four monogenic genes are involved in the PD pathological process, including parkin RBR E3 ubiquitin protein ligase (PRKN), phosphatase and tensin homolog-induced putative kinase 1 (PINK1), Leucine Rich Repeat Kinase 2 (LRRK2) and α synuclein (SNCA). Meanwhile, eight PD risk factor genes-namely, glucocerebrosidase (GBA), mitochondrial transcription factor A (TFAM), Nuclear factor erythroid 2-related factor 2 (NFE2L2), macrophage metalloelastase (MMP12), major histocompatibility complex, (MPTP)-induced mice confirmed ANK1 may offer an efficient diagnosis of PD. Our findings shed light on the crucial role of ANK1 in PD as well as provide a possible biomarker for predicting PD progression.

DEGs Assessment
The downloaded dataset files were standardized by quantiles. DEGs between normal and PD samples were identified via the limma method. The screening criteria were |logFC (fold change)| > 1.5 and adjust p value < 0.05. The common DEGs were identified and visualized using heatmaps and Venn diagrams.

Protein-Protein Interaction Analysis
DEGs in GSE8397 and GSE22491 were transferred to the Search Tool for the Retrieval of Interacting Genes (STRING) database to construct a protein-protein interaction (PPI) network. Subsequently, Cytoscape software (v3.7.2, Cytoscape Consortium, San Diego, CA, USA) was applied to detect the correlations between DEGs. The Molecular Complex Detection (MCODE) plugin in Cytoscape was used to screen hub genes and extract functional modules. The settings were as follows: network scoring degree cutoff, 2; cluster finding, fluff; node density cutoff, 0.1; node score cutoff, 0.2; k-core, 2; max.depth, 100.

Genome Ontology (GO) Enrichment Analysis
GSE99039 datasets were chosen and the median value of ANK1 was calculated. We divided patients into a high ANK1 (h-ANK1) group and a low ANK1 (l-ANK1) group according to the median value and DEGs were excavated. Gene ontology (GO) enrichment analysis is a commonly utilized technique for understanding the functional classification of all DEGs. The three GO categories, including biological process (BP), cellular component (CC) and molecular function (MF), were analyzed by the Database for Annotation, Visualization and Integrated Discovery (DAVID).

Receiver Operating Characteristics (ROC) Curve Analysis
GSE22491 and GSE34287 datasets containing blood samples were downloaded from the GEO database for drawing ROC curve (with sensitivity on ordinate and 1-specificity on abscissa). The ROC curve allows for the detection of the association between sensitivity and specificity. The validity and discriminant ability of genes for diagnosis were assessed using the area under the curve (AUC) value.

Construction of MPTP-Induced PD Model
Mice were divided into two groups: normal group (saline) and MPTP group (MPTP). Normal mice were given intraperitoneal saline injections. The PD construction was performed as previously described. Briefly, the mice received intraperitoneal injection of MPTP-HCl (dissolved in sterile normal saline, 30 mg/kg body weight) per day for 5 successive days. At 1 day after the last MPTP intraperitoneal injection, ANK1 mRNA level was measured. The behavioral tests and protein expression assay were conducted 5 days after the last MPTP intraperitoneal injection.

Behavioral Tests
To test locomotor activity, open-field experiment was applied. Each mouse was placed individually in a box (50 cm length × 50 cm width × 50 cm height) with 16 equal squares. The spontaneous activity (i.e., crossings: the number of squares crossed; rearings: the counts of standing; groomings: cleaning frequency) were recorded for 5 min.
To measure motor coordination, the rotarod experiment was conducted to observe foot movement. Mice were placed on a rotating rod, which rotated along its longitudinal axis. We stopped the apparatus when the mice fell down or when the latency to fall reached 150 s. The length of time that the mice remained on the rotarod was recorded and used for measuring motor behaviors.
The spontaneous forelimb use was evaluated by cylinder test. Mice were placed individually in a glass cylinder for 10 min. We recorded the contact against the cylinder wall (i.e., rearing using the left forepaw, right forepaw or both paws together). The cylinder forelimb asymmetry ratio (CAR) was then calculated using the following formula: [(dominant forelimb touches + 1 2 number of both forelimbs)/(dominant forelimb touches + non-dominant forelimb touches + both) × 100] [37].
The suspension test was designed to measure grasping behavior. A 40 cm wire was fixed tightly in the platform and kept horizontally. Mice were put on the center of the wire and allowed to freely walk for a maximum of 1 min. The time spent by the mice hanging onto the wire was recorded and this index indicated muscle endurance [38].

Western Blot Analysis
Tissues were collected and homogenized in phosphate-buffered saline containing Triton X-100 (PBST). The protein samples were performed on sodium dodecyl sulfatepolyacrylamide gel electrophoresis and subsequently transferred to polyvinylidene fluoride membranes. The membranes were blocked for 2 h at room temperature using 5% skimmed milk. The membranes were then incubated with the primary antibody (anti-ANK1 and anti-β-actin) at 4 • C overnight and washed with PBS containing Tween 20 three times the next day. The blots were treated with secondary antibody for 2 h at room temperature. Finally, enhanced chemiluminescent substrate was used to visualize the bands. β-Actin served as loading control.

Immunofluorescence
Brain slices were blocked with goat serum, followed by incubation with the primary antibody (anti-ANK1). The slices were washed with PBS three times. The following step was incubated with FITC-labelled-goat anti-rabbit antibody for 30 min at 37 • C. After having been washed again three times with PBS, the specimens were sealed and photographed under fluorescence.

Statistical Analysis
DEGs were generated using the limma method. ROC curve analysis was carried out and calculated through SPSS statistical software (version 23, IBM SPSS Statistics, Armonk, NY, USA). Data were expressed as mean ± SEM. Student's t-test was applied to evaluate significant differences between two groups. p vlaue < 0.05 was considered significant.

Identification of ANK1 as a Commonly Down-Regulated Gene
We collected nine PD LSN and seven control LSN samples from GSE 8397. We found 4 up-regulated genes and 41 down-regulated genes ( Figure 1A). The GSE 8397 consisting of MSN (i.e., 15 PD MSN and 8 control MSN) was also analyzed and the generated heatmap showed 7 up-regulated genes and 47 down-regulated genes ( Figure 1B). A microarray study of GSE 22491 on blood was chosen for further analysis. As shown in Figure S1, 114 genes were up-regulated and 56 genes were down-regulated. The exact data are shown in Table S1. To excavate the overlapping genes, the Venn diagram was drawn to display the numbers of intersected gene sets. We observed that ANK1 was the only DEG present ( Figure 1C). For the comparison of LSN samples, ANK1 was lowly expressed in PD, which was verified in MSN ( Figure 1D-E). Consistently, ANK1 in blood samples was also significantly decreased ( Figure 1F). These results indicated that PD tissues may have low ANK1 expression.

Construction of PD Associated Regulatory Network and Module Analysis
To better understand the interplay among the identified DEGs, STRING tool was used to elucidate protein-protein interaction networks. Four up-regulated genes and forty-one down-regulated genes from GSE 8397 LSN samples were submitted to STRING and a network was generated. The network was made up of three modules (13 nodes/30 edges, 7 nodes/8 edges and 5 nodes/5 edges, respectively) ( Figure S2). A crucial module containing ANK1 is shown in Figure 2A. In addition, we used the STRING database to analyze GSE 8397 MSN samples. The network contained 33 nodes. Two modules (11 nodes/24 edges and 15 nodes/21 edges, respectively) were extracted from the network ( Figures 2B and S3). Similarly, there were 97 nodes in the network of DEGs that were identified in our study (GSE 22491). A total of six modules were extracted (26 nodes/133 edges, 17 nodes/78 edges, 9 nodes/18 edges, 19 nodes/30 edges, 9 nodes/13 edges and 4 nodes/ 4 edges, respectively) and ANK1 existed in two significant modules ( Figures 2C and S4). The GO enrichment analysis for these networks is presented in Table S2. Taken together, these findings suggest that ANK1 is associated with PD.  Figure S1, 114 genes were up-regulated and 56 genes were down-regulated. The e data are shown in Table S1. To excavate the overlapping genes, the Venn diagram drawn to display the numbers of intersected gene sets. We observed that ANK1 was only DEG present ( Figure 1C). For the comparison of LSN samples, ANK1 was lo expressed in PD, which was verified in MSN ( Figure 1D-E). Consistently, ANK1 in bl samples was also significantly decreased ( Figure 1F). These results indicated that tissues may have low ANK1 expression.

Construction of PD Associated Regulatory Network and Module Analysis
To better understand the interplay among the identified DEGs, STRING tool used to elucidate protein-protein interaction networks. Four up-regulated genes forty-one down-regulated genes from GSE 8397 LSN samples were submitted to STRI and a network was generated. The network was made up of three modules (13 node edges, 7 nodes/8 edges and 5 nodes/5 edges, respectively) ( Figure S2). A crucial mod containing ANK1 is shown in Figure 2A. In addition, we used the STRING databas analyze GSE 8397 MSN samples. The network contained 33 nodes. Two modules Genes 2023, 14, x FOR PEER REVIEW 6 of 17 erated heatmap showed 7 up-regulated genes and 47 down-regulated genes ( Figure 1B). A microarray study of GSE 22491 on blood was chosen for further analysis. As shown in Figure S1, 114 genes were up-regulated and 56 genes were down-regulated. The exact data are shown in Table S1. To excavate the overlapping genes, the Venn diagram was drawn to display the numbers of intersected gene sets. We observed that ANK1 was the only DEG present ( Figure 1C). For the comparison of LSN samples, ANK1 was lowly expressed in PD, which was verified in MSN ( Figure 1D-E). Consistently, ANK1 in blood samples was also significantly decreased ( Figure 1F). These results indicated that PD tissues may have low ANK1 expression.

Construction of PD Associated Regulatory Network and Module Analysis
To better understand the interplay among the identified DEGs, STRING tool was used to elucidate protein-protein interaction networks. Four up-regulated genes and forty-one down-regulated genes from GSE 8397 LSN samples were submitted to STRING and a network was generated. The network was made up of three modules (13 nodes/30 edges, 7 nodes/8 edges and 5 nodes/5 edges, respectively) ( Figure S2). A crucial module containing ANK1 is shown in Figure 2A. In addition, we used the STRING database to analyze GSE 8397 MSN samples. The network contained 33 nodes. Two modules (11 nodes/24 edges and 15 nodes/21 edges, respectively) were extracted from the network ( Figures 2B and S3). Similarly, there were 97 nodes in the network of DEGs that were identified in our study (GSE 22491). A total of six modules were extracted (26 nodes/133 edges, 17 nodes/78 edges, 9 nodes/18 edges, 19 nodes/30 edges, 9 nodes/13 edges and 4 nodes/4 edges, respectively) and ANK1 existed in two significant modules (Figures 2C

ANK1 Confers Significant Correlation with PD Progression
We selected GSE 99039, a peripheral whole blood microarray dataset with the biggest sample number, to further explore the association between ANK1 and PD. Based on the median value, the GEO cohort was divided into two parts, namely, the high-ANK1 (h-ANK1) group and the low-ANK1 (l-ANK1) group. A volcano plot depicted the log2 fold change on the x-axis and the -log10 of adjusted p-values on the y-axis. The graph presented increased expression of 73 genes and decreased expression of 123 genes ( Figure 3A). Consistently, Parkinson's biomarkers-including membrane palmitoylated protein 1 (MPP1), solute carrier family 4, anion exchanger, member 1 (SLC4A1), erythrocyte membrane protein band 4.2 (EPB42), Selenium-binding protein 1 (SELENBP1), Glycophorin B (GYPB), α-Hemoglobin Stabilizing Protein (AHSP), hemoglobin-delta (HBD), α-synuclein (SNCA) and ferrochelatase (FECH)-were reduced in l-ANK1 compared with that in h-ANK1 ( Figure 3B). To outline the functions of these DEGs, we performed a functional enrichment analysis (Table 1). GO BP analysis revealed that DEGs were mainly enriched in oxygen transport, myeloid cell development, gas transport and erythrocyte differentiation ( Figure 3C). In the CC part, the DEGs were related to hemoglobin complex, haptoglobin-hemoglobin complex and cortical cytoskeleton ( Figure 3D). Additionally, GO MF analysis demonstrated that DEGs participated in oxygen transporter activity, haptoglobin binding, oxygen binding and iron ion binding ( Figure 3E). These mined BP, CC and MF terms have a close relationship with PD, which may explain the involvement between ANK1 and PD. 023, 14, x FOR PEER REVIEW 7 and S4). The GO enrichment analysis for these networks is presented in Table S2. T together, these findings suggest that ANK1 is associated with PD.

ANK1 Confers Significant Correlation with PD Progression
We selected GSE 99039, a peripheral whole blood microarray dataset with the gest sample number, to further explore the association between ANK1 and PD. Base the median value, the GEO cohort was divided into two parts, namely, the high-A

ANK1 Shows a High Diagnosis Capacity in PD
To further evaluate ANK1's potential application in the diagnosis of PD, we performed a ROC analysis to examine the sensitivity and specificity. Data from GSE22491 were analyzed and the results showed the AUC value to be 0.975, suggesting ANK1 may be an effective biomarker for diagnosis ( Figure 4A). In addition, we analyzed GSE34287 datasets and calculated the AUC value. ROC analysis for ANK1 resulted in an AUC of 0.

Validation of ANK1 in MPTP-Induced PD Model
To validate the predictive role for PD, we created a mouse model of PD by intraperitoneal injection of MPTP. A timeline of the experimental design is presented in Figure S5. Mice received a time course experiment and the rotarod test was applied at 1, 2, 3, 4 and 5 days after the last MPTP injection. As shown in Figure S6, at day 4 after the last MPTP injection, the latency to fall was significantly reduced compared with normal mice. Mice were subjected to behavioral tests, and blood and brain samples were collected for biological analysis. As indicated in Figure 5A, mice treated with MPTP exhibited a significant reduction in the number of rearings, crossings and groomings compared to the control mice. In addition, motor coordination was presented in Figure 5B, the latent period of PD mice to fall from the rotarod test was dramatically reduced compared to the control group. The calculated CAR from behavioral observation demonstrated that PD mice displayed higher forelimb asymmetry than normal mice ( Figure 5C). Meanwhile, in the suspension test, the percentage of time spent on the wire was recorded, and MPTP administration considerably diminished this value ( Figure 5D). To explore whether the MPTP-induced PD mice had a differentially expressed ANK1 level, quantitative real-time PCR was op-erated and the result showed decreased ANK1 mRNA expression in PD blood and SN ( Figure 5E,F). The verification in vivo was in agreement with bioinformatic analysis in this study. Furthermore, MPTP treatment reduced ANK1 protein expression in blood and SN ( Figure 5G). Immunofluorescence results showed that the green fluorescent signal intensity was decreased in the SN of MPTP group, indicating ANK1 expression may be down-regulated in PD ( Figure 5H). In addition, in vitro assay indicated that knockdown of ANK1 in PC12 cells can significantly reduce the cell viability compared to normal controls ( Figure 5I). Considering that PD pathogenesis is characterized by neuron loss, the ANK1 deficiency-induced PC12 cell injury further illuminates the potential correlation between ANK1 and PD.

Validation of ANK1 in MPTP-Induced PD Model
To validate the predictive role for PD, we created a mouse model of PD by intra itoneal injection of MPTP. A timeline of the experimental design is presented in Fi S5. Mice received a time course experiment and the rotarod test was applied at 1, 2 and 5 days after the last MPTP injection. As shown in Figure S6, at day 4 after the MPTP injection, the latency to fall was significantly reduced compared with normal Mice were subjected to behavioral tests, and blood and brain samples were collecte biological analysis. As indicated in Figure 5A, mice treated with MPTP exhibited a nificant reduction in the number of rearings, crossings and groomings compared t control mice. In addition, motor coordination was presented in Figure 5B, the laten riod of PD mice to fall from the rotarod test was dramatically reduced compared t control group. The calculated CAR from behavioral observation demonstrated tha mice displayed higher forelimb asymmetry than normal mice ( Figure 5C). Meanwhi the suspension test, the percentage of time spent on the wire was recorded, and M administration considerably diminished this value ( Figure 5D). To explore whethe MPTP-induced PD mice had a differentially expressed ANK1 level, quantitative real-PCR was operated and the result showed decreased ANK1 mRNA expression in blood and SN ( Figure 5E,F). The verification in vivo was in agreement with bioinform

Discussion
The discovery and validation of biomarkers are urgently needed for clinical performance improvement. Bioinformatics analysis on DEGs can provide potential molecules for diseases [39]. The lack of reliable and practical biomarkers is a major obstacle hindering accurate PD detection [40]. Given the potential severity of this condition, there

Discussion
The discovery and validation of biomarkers are urgently needed for clinical performance improvement. Bioinformatics analysis on DEGs can provide potential molecules for diseases [39]. The lack of reliable and practical biomarkers is a major obstacle hindering accurate PD detection [40]. Given the potential severity of this condition, there is a growing requirement to engage research to identify novel assessments. In the present study, we initially obtained GSE datasets from the GEO database, derived from PD patients and healthy controls, and subsequently explored the DEGs and found that ANK1 was commonly differentially expressed in all three tissues (i.e., LSN, MSN and blood). Thus, our research suggested that ANK1 may serve as a biomarker for the diagnosis of PD and further analyses are needed to assess validity. PPI network analysis is an effective approach to identify hub genes with clinical merit [41]. Based on the above observation, a network analysis of DEGs was performed and key modules were extracted. We identified ANK1 as a significant hub gene in the network of GSE8397, a dataset consisting of LSN and MSN. Similarly, in GSE 22491, six modules were gathered and we can see ANK1 occupied two of these modules. Consistent with previous DEGs screening results, PPI network analysis showed that ANK1 was in the key nodes and highly interacted with other genes.
To further illuminate the associations between ANK1 and PD, we divided PD patients into two parts: an h-ANK1 group and an l-ANK1 group. We observed a significant down-regulation of genes, including MPP1, SLC4A1, EPB42, SELENBP1, GYPB, AHSP, HBD, SNCA and FECH. Among these genes, SNCA was identified as the first gene to be involved in PD [42]. It was proven that the relative SNCA transcript level in venous blood was considerably lower in PD patients compared with those in healthy subjects [22]. MPP1 mRNA level has been reported to be decreased in two atypical parkinsonian disorders (multiple system atrophy (MSA) and progressive supranuclear palsy (PSP)) and PD compared with healthy controls. Thus, MPP1 is a useful indicator to diagnose PD [43]. SLC4A1 is responsible for anion exchanger 1 membrane protein production. According to the analysis of expression arrays, research has shown that the SLC4A1 gene is reduced in PD blood versus controls [44,45]. Interestingly, we found a significant down-regulation of genes associated with hemoglobin and ion metabolism, including HBD, AHSP, EPB42, SELENBP1, GYPB and FECH. HBD encodes delta-chains of hemoglobin and a meta-analysis using blood samples was performed and reported that HBD was the second most highly ranked gene after SNCA [24]. Encoded by the FECH gene, ferrochelatase plays an essential role in maintaining iron metabolism [46]. Network-based meta-analysis identified FECH as the down-regulated gene in blood of PD patients [47]. AHSP encodes a molecular chaperone binding specifically to free α-globin and is necessary for hemoglobin assembly [48]. EPB42, a major component of the erythrocyte membrane skeletal network, engages in the maintenance of normal surface area in red blood cells and participates in iron homeostasis [49,50]. Human SELENBP1, a member of the selenium-binding protein family, is a highly conserved protein that covalently binds selenium and mediates the intracellular transport of selenium [51]. GYPB is glycoprotein of the human erythrocyte membrane and may represent changes in the synthesis of erythrocytes [52]. Our study observed that there was a significant decrease in the l-ANK1 group's blood MPP1, SLC4A1, EPB42, SELENBP1, GYPB, AHSP, HBD, SNCA and FECH genes compared with the h-ANK1 group, which was consistent with the previous study that examined the differential blood gene expression between PD patients and healthy controls [24].
Hemoglobin is a protein heavily expressed in red blood cells. In mammals, the main function of hemoglobin is to transport oxygen from lungs to other tissues and interact with small functional molecules (e.g., carbon dioxide, carbon monoxide, and nitric oxide) [53]. As the most significant source of peripheral iron, hemoglobin may modulate iron homeostasis. Additionally, it has been proposed that PD incidence rose significantly as hemoglobin increased [54]. However, there exists evidence of a negative correlation between the hemoglobin level and PD. It has been proposed that PD patients with anemia or low hemoglobin may precede the motor symptoms by 20 or more years [55]. On the other hand, studies have shown iron homeostasis is dysregulated in PD patients [56,57]. The increased iron level has been found in the SN in PD patients in comparison to controls and high iron concentrations may be responsible for PD pathogenesis [58,59]. Intriguingly, increased serum iron may be associated with reduced PD risk, and therefore showing a protective effect of iron against PD [60,61]. Concurrently, most of the iron in the human body is present as heme iron and contained in the hemoglobin of erythrocytes. Its deficiency has been reported to link with anemia [62,63]. In this study, we divided the blood samples from GSE99039 into two parts (i.e., the h-ANK1 group and the l-ANK1 group) according to the median value of ANK1. We found genes associated with hemoglobin and iron homeostasis were down-regulated in the l-ANK1 group. Furthermore, GO analysis has shown that these DEGs were enriched in GO terms related to oxygen transport, myeloid cell development, gas transport and erythrocyte differentiation (BP); hemoglobin complex, haptoglobin-hemoglobin complex and cortical cytoskeleton (CC); and oxygen transporter activity, haptoglobin binding, oxygen binding and iron ion binding (MF). These results demonstrate that ANK1 may serve as a potential indicator for promoting PD diagnosis.
We further investigated the predictive performance of ANK1 by drawing an ROC curve. In GSE22491, the AUC value was 0.975 and this result implies a high predictive ability of ANK1. In addition, GSE34287 was chosen and used to evaluate ANK1's predictive effect. Although the single use of ANK1 may not be effective enough in diagnosing PD, the combination of both genes resulted in improved diagnostic efficiency. For example, some genes such as ZNF134, ARHGAP26, CACNA1D, DBH, EMP1, IFI27, PDE6D, PRG3, SLC14A1, SOD2, TGFB1, THY-1, TRIM24, VAMP4, VAMP8 and ZFAND1 may improve PD's diagnosis [47,[64][65][66][67][68][69][70][71]. We have observed, in combination with ANK1, there was a significant increase in the AUC values of these biomarkers. As an integral membrane and adaptor protein, ANK1 modulates the attachment of membrane proteins (e.g., cell adhesion proteins and ion channels) and is pivotal for cell proliferation, mobility, activation and maintenance of specialized membrane domains [72]. Our results revealed that ANK1 was correlated with PD's potential factors, including the SNCA gene, hemoglobin related genes and ion metabolism. Thus, the application of ANK1 for combined diagnosis may bring benefit in terms of increasing diagnostic accuracy.
Based on the previous results and analysis, we created an MPTP-induced mice PD model. MPTP is an analogue of the narcotic meperidine, which was found to induce parkinsonian-like syndrome [73,74]. The literature has substantiated that MPTP can easily cross the blood-brain barrier and finally be converted to 1-methyl-4-phenylpyridinium (MPP + ). MPP + destroys dopaminergic neurons via inhibiting mitochondrial complex I and inducing oxidative stress [75][76][77]. In recent work, mice received an intraperitoneal injection of MPTP and showed reduced locomotor activity, impaired motor balance, elevated CAR and attenuated grasping strength. We were curious about whether different ANK1 mRNA levels can be found between the control group and the MPTP-induced PD group. Consistent with the aforementioned data, the PCR results revealed that blood and SN ANK1 mRNA expression was down-regulated in the PD group. Meanwhile, MPTP treatment reduced ANK1 protein level when compared to the normal group. Further studies have shown that ANK1 deficiency caused a decrease in PC12 cell viability, indicating the potential association between ANK1 and PD.
In summary, through performing bioinformatics analyses, including DEGs analysis, PPI network analysis and GO enrichment analysis, we identified that ANK1 may serve as a biomarker for PD. Moreover, the AUC values we investigated and our in vivo studies suggest a good discriminative capacity of ANK1 in predicting PD. Our research highlighted the potential clinical value of ANK1 for PD diagnosis. Further studies considering the mechanisms undergirding the links between ANK1 and PD are needed, which may shed light on PD prevention and treatment.
Supplementary Materials: The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/genes14010226/s1, Figure S1: ANK1 was identified as a down-regulated gene in PD blood. DEGs expression heat maps in blood. Figure S2: PPI networks construction and module analysis of GSE8937 LSN DEGs. The nodes represent proteins and edges reflect proteinprotein associations. The texts inside the elliptic nodes represent genes names. Figure S3: PPI networks construction and module analysis of GSE8937 MSN DEGs. The nodes represent proteins and edges reflect protein-protein associations. The texts inside the elliptic nodes represent genes names. Figure S4: PPI networks construction and module analysis of GSE22491 blood DEGs. The nodes represent proteins and edges reflect protein-protein associations. The texts inside the elliptic nodes represent genes names. Figure S5: Timeline of the in vivo study. Representative figure shows experimental design. Mice were injected intraperitoneally with saline or MPTP (30 mg/kg) once daily for 5 successive days. mRNA measurement was performed on day 6. Behavioral tests were carried out on day 10. Mice were sacrificed after behavioral tests and tissues were collected for protein expression assay. Figure S6: Time course experiment. Mice were administered with saline or MPTP and the rotarod test was applied at 1, 2, 3, 4 and 5 days after the last MPTP intraperitoneal injection. Data are expressed as mean ± S.E.M. Two groups were analyzed by Student's t-test. *** p < 0.001. Table S1: DEGs in GSE 8397 and GSE 22491.