Allelic variant in SLC6A3 rs393795 affects cerebral regional homogeneity and gait dysfunction in patients with Parkinson’s disease

Introduction

Parkinson’s disease (PD), characterized by gradually aggravated nigrostriatal dopaminergic neuronal dysfunction, is a progressive neurodegenerative disorder (Fearnley & Lees, 1991). The pathophysiological process of PD is influenced not only by nongenetic factors such as environmental risks and aging but also by genetic polymorphisms (Raza, Anjum & Shakeel, 2019). Recently, the role of associative gene polymorphisms in sporadic PD has increasingly been investigated. The human dopamine transporter (DAT) gene, now known as SLC6A3 (chromosome 5p15.3), is one of the most crucial candidate genes in sporadic PD due to its role in dopamine (DA) neuron function (Zhai et al., 2014). SLC6A3 encodes DAT, which is a membrane protein located in the presynaptic region of dopaminergic neurons, where it can rapidly transport DA from the extracellular space into the cytosol of the presynaptic neuron, contributing to overall DA neurotransmission in brain (Cragg & Rice, 2004; Lohr et al., 2017). Recently, a large prospective trial from the Parkinson Progression Marker Initiative (PPMI) believed that the dopaminergic denervation in presynaptic striatum could be a bad predictor for the later development of freezing of gait (FOG) in almost 400 early PD patients (Kim et al., 2018). Moreover, a pharmacogenetic study suggested that SLC6A3 polymorphisms appeared to be associated with the response to levodopa observed in advanced PD patients with gait disorders (Moreau et al., 2015). Hence, these studies confirmed the important role of SLC6A3 polymorphisms in PD progression. However, in SLC6A3, most previous literature focused on the role of variable number tandem repeat (VNTR) polymorphism in the 3′  untranslated region in PD (Kim et al., 2000; Kang, Palmatier & Kidd, 1999). Recently, growing attention was paid to other single nucleotide polymorphisms (SNPs), such as rs393795. Moreover, several latest studies indicated that SLC6A3 rs393795 played an important role in PD (Purcaro et al., 2019; Kaplan et al., 2014). However, the SLC6A3 rs393795 polymorphisms were not widely investigated in PD, particularly about how the variant affected cerebral spontaneous neuronal activity and clinical manifestations in PD. Therefore, we employed this study to investigate the effects of the SLC6A3 rs393795 allelic variant on the brain spontaneous neuronal activity and relevant clinical features in PD.

Imaging genetics was applied in our research, which could provide us with new insights into the influence of genetic variant on brain function (Bogdan et al., 2017). The method has been widely employed to explore several neurological and psychiatric diseases, such as Alzheimer’s disease (AD) (Apostolova et al., 2018). Actually, resting-state functional magnetic resonance imaging (rs-fMRI) can reveal the brain spontaneous neuronal activity by examining spontaneous fluctuations in the blood oxygen level dependent (BOLD) signal at rest (Prodoehl, Burciu & Vaillancourt, 2014). So, it makes it possible to explore the probable role of gene polymorphisms in brain activity. Calculating the similarity of voxel fluctuations within a given cluster, regional homogeneity (ReHo) can embody the local synchronization of spontaneous BOLD signals (Long et al., 2008). This approach has been widely adopted in PD studies, and even was considered as a potential diagnostic biomarker for PD (Li et al., 2016). However, few researchers focused on the probable role of gene polymorphisms in brain region dysfunction in PD. Thus, we employed this study to explore the role of SLC6A3 rs393795 allelic variant in cerebral spontaneous neuronal activity and relevant clinical presentations in PD, by examining cerebral ReHo signal alterations as well as the correlations between altered brain interactive regional ReHo values and clinical symptoms.

Materials & Methods

Results

Regional homogeneity

The effects of diagnostic groups (PD versus HC) and genotypes (AA versus CA/CC) on ReHo values analyzed by two-way ANCOVA were shown in Table 3. The significant main effect of groups (PD versus HC, regardless of allelic variant status) was observed in left middle temporal gyrus (F = 17.91, p < 0.001, corrected) and pallidum (F = 22.66, p < 0.001, corrected) (Figs. 1A–1J). Furthermore, PD patients exhibited increased ReHo signals in left middle temporal gyrus but reduced ReHo signals in left pallidum in comparison with HC subjects. The significant main effect of genotypes (AA versus CA/CC carriers, regardless of disease status) was discovered in right inferior frontal gyrus (IFG, F = 32.73, p < 0.001, corrected) and supplementary motor area (SMA, F = 22.21, p < 0.001, corrected) (Figs. 1K–1T). Besides, AA genotype participants showed aberrant increased ReHo signals in right IFG and SMA compared with CA/CC carriers. It was worth noting that significant interactions between groups and genotypes (affected by both disease factor and allelic variation) were found in the right inferior temporal gyrus (ITG, F = 25.10, p < 0.001, corrected) (Fig. 2). Hereafter, post hoc tests (Bonferroni) were employed to explore further interactive differences, and p < 0.008 (0.05/6 [all possible pair-wise comparisons]) was considered significant. Specially, abnormal increased ReHo signals in right ITG were observed in PD-AA compared with PD-CA/CC group (p < 0.001), while reduced ReHo signals in right ITG were detected in AA homozygotes in comparison with CA/CC carriers in HC subgroups (p = 0.004).

Table 3:

Groups × genotypes ANCOVA of ReHo.

Brain regions (AAL)	Peak MNI Coordinates x, y, z (mm)			Peak F value	Cluster size (voxels)
(1) Main effect of groups
Temporal_Mid_L	−39	6	−36	17.91	14
Pallidum_L	−18	0	−3	22.66	12
(2) Main effect of genotypes
Frontal_Inf_R	36	33	12	32.73	13
Supp_Motor_Area_R	3	9	57	22.21	12
(3) Groups × genotypes interaction
Temporal_Inf_R	57	−66	−9	25.10	21

DOI: 10.7717/peerj.7957/table-3

Notes:

Two-way factorial analysis of covariance (ANCOVA: groups × genotypes; groups: PD versus HC, genotypes: AA versus CA/CC) was performed, adjusting for age and gender. A corrected threshold by Monte Carlo simulation was set at P < 0.001.

Abbreviations

PD

Parkinson disease

HC

Healthy control

ReHo

Regional homogeneity

AAL

Anatomical automatic labeling

MNI

Montreal Neurological Institute

R

right

L

left

Correlation analysis

Correlations between the clinical scores and ReHo values extracted from significant interactive clusters were investigated in PD patients. ReHo values in right ITG were negatively associated with TMT gait subscale scores (r =  − 0.554, p = 0.026) (Fig. 3A) and positively related to FOG-Q scores (r = 0.581, p = 0.018) (Fig. 3B) in PD patients with AA genotype. These results suggested that the severity of gait disorders in PD was closely associated with altered right ITG functional activity modulated by the AA genotype of SLC6A3 rs393795. Nevertheless, in the two subgroups of PD, no other significant correlations were found between right ITG ReHo values and any other clinical scores (p > 0.05).

Discussion

In this imaging genetic study, we explored the role of SLC6A3 rs393795 allelic variant in brain regional homogeneous activity in sporadic PD patients. Compared with CA/CC carriers, AA genotype individuals showed abnormal increased ReHo signals in right IFG and SMA in all participants. Moreover, our interaction investigation revealed that AA homozygotes exhibited aberrant increased ReHo signals in right ITG compared with CA/CC carriers in PD subgroups. Notably, ReHo values in right ITG were negatively associated with TMT gait subscale scores and positively related to FOG-Q scores in PD-AA group.

SLC6A3 rs393795 was supposed to be the candidate SNP for sporadic PD, supported by numerous imaging and pharmacogenetic studies validating the close relationship between DAT and PD (Djaldetti et al., 2018; Kim et al., 2018). Particularly, the depletion of presynaptic DA was believed to be closely connected with gait disorders in PD, and even could be a bad prognostic sign in terms of FOG development (Kim et al., 2018; Djaldetti et al., 2018). Moreover, the rs393795 SNP (located on intron 4) falls within the same linkage disequilibrium cluster with rs460000 SNP (located within the exon 4-intron 3 boundary) (Cartegni, Chew & Krainer, 2002), which has been associated with frontostriatal response inhibition circuits (Cummins et al., 2012). Consistent with this, our study showed that the SNP AA genotype effected the brain functional activation of right SMA and IFG in all individuals. Furthermore, the right SMA and IFG, involved in frontal cortico-basal ganglia motor network, had been shown to participate in the functional organization of gait initiation or FOG phenomenon (Bartels & Leenders, 2008). Thus, we hypothesized that AA genotype could affect gait in human via altering the spontaneous activity of right SMA and ITG. However, no significant differences were discovered in gait associative scales between PD-AA and PD-CA/CC groups. These findings perhaps implied that the effects of the SNP polymorphisms were compensated and consequently not sufficient to cause significant different clinical manifestations. In good agreement with this, abnormal increased ReHo values in right SMA and IFG were observed in AA carriers compared with CA/CC carriers in our study. Alternatively, the SLC6A3 rs393795 allelic variant might just play a regulatory instead of causative role in PD patients with gait disorders, in other words, not anyone with allelic variation will markedly develop or aggravate PD gait dysfunction. Consistent with this, SLC6A3 rs393795 variant, falling within the same linkage disequilibrium cluster with two of the SNPs (rs458609, rs457702) and with rs460000 SNP, was merely indicated to participate in DAT translations by affecting alternative splicing of the DAT to some extend (Cartegni, Chew & Krainer, 2002; Talkowski et al., 2010). However, these didn’t represent our study worthless. On the contrary, it could deepen our understanding of genetic mechanisms underlying gait disorders in PD, contributing to more effective and precise diagnosis and treatments in the future.

Since both PD and SLC6A3 rs393795 variant had effects on the dysfunction of frontal cortico-basal ganglia motor pathway, their interactions might lead to greater alterations in related brain regions. In good agreement with this, our interaction investigation revealed that AA carriers showed aberrant increased ReHo signals in right ITG in comparison with CA/CC carriers in PD groups. ITG was one of the key hubs in ventral visual pathway (Weller, 1988), which was characterized to support the processing of object quality or identity (Kravitz et al., 2013) and mainly dominated by right-hemisphere (Woolley et al., 2010). It was worth noting that a large amount of evidence indicated the effectiveness of various visual aids in compensating for gait difficulties in PD (Davidsdottir, Cronin-Golomb & Lee, 2005; Lee et al., 2012). Moreover, more and more literature demonstrated that gait was not only an automated motor activity, but also one increasingly dependent on external information, especially visual cues (Uc et al., 2005; Sage & Almeida, 2010). Additionally, dysfunction of right occipitotemporal “visual” networks was also detected in PD patients with gait deficits from different research populations (Tessitore et al., 2012; Wang et al., 2016). Hence, abnormal increased ReHo signals in right ITG observed in PD-AA group could be due to its compensatory activation when PD patients had a worse trend of gait disorders caused by a worse dysfunction of frontal cortico-basal ganglia motor network. Consistently, our further correlation analysis displayed that ReHo values in right ITG were negatively related to TMT gait subscale scores and positively associated with FOG-Q scores in PD-AA group. TMT gait subscale and FOG-Q both showed excellent validity and reliability for the valuation of different gait characteristics and had been widely used in PD researches (Kegelmeyer et al., 2007; Tambasco et al., 2015; Park et al., 2018; Li et al., 2018). Besides, the lower scores of TMT gait subscale represents more serious gait deficits, in contrast, the higher scores of FOG-Q means more serious FOG. Taken together, SLC6A3 rs393795 AA genotype might have a tendency to aggravate the severity of gait dysfunction in PD by altering the SMA and IFG spontaneous activity, and ultimately lead to the compensatory activation of right ITG.

Nevertheless, AA homozygous of SLC6A3 rs393795 was only previously reported to protect the elderly from delirium possibly by reducing the concentrations of DA in the brain (van Munster et al., 2010a; van Munster et al., 2010b). Meanwhile, the combination of 10R/10R ( rs28363170) and A carrier ( rs393795), of the SLC6A3 gene could reduce the risk of levodopa-induced dyskinesias (LIDs) during long-term therapy with levodopa (Purcaro et al., 2019). As we all know, LIDs are characterized by dramatically increased synaptic DA concentrations in brain (Fuente-Fernandez et al., 2004). Thus, SLC6A3 rs393795 AA genotype might be involved in the pathophysiological mechanisms underlying gait disorders in PD by reducing DA concentrations in frontal cortico-basal ganglia motor regions, such as SMA and ITG, and eventually cause compensatory activation of visual regions including right ITG. However, decreased ReHo values in right ITG were found in HC-AA compared with HC-CA/CC. Without the effects of disease, we preferred that decreased ReHo values of right ITG in HC-AA might be mainly caused by genetic polymorphisms. However, the deeper mechanism remains to be studied.

There were other limitations of our research. First, our research sample size is relatively small, i.e., small number genetic subgroups in the PD and HC samples, which may limit our ability to further classify genotypes and explore minor brain functional alterations related to SLC6A3 rs393795. Second, the minor allele frequencies (MAF) of rs393795 (C) was 0.44 in our study, in good agreement with previous Chinese studies (Shang & Gau, 2014; Chu et al., 2019). Because the genetic polymorphisms are frequently different among the races, additional studies further investigating the SNP in PD are needed to clarify its role in different ethnicities. Third, although we had taken some measures to reduce the pharmacological effects on neural activity, we had to admit that the long-term effect of dopaminergic agonists could not be ruled out. Fourth, our study was conducted when PD patients were at rest and without dopaminergic treatments, more realistic studies should be performed, such as during a task or dopaminergic treatments. Fifth, we only investigate SLC6A3 rs393795 in the present study, additional further and deeper studies are needed, such as gene-gene interactions. Finally, although TMT, FOG-Q and TUG tests were used to measure gait in our study, they could not fully assess the subtle gait characteristics that might be affected in PD, therefore more detailed assessments should be applied in future.

Conclusions

With imaging genetics, our findings suggested that SLC6A3 rs393795 allelic variation might have a trend to aggravate the severity of gait disorders in PD patients by altering the SMA and IFG function, and ultimately result in compensatory activation of right ITG. It could provide us with a new perspective for exploring deeply genetic mechanisms of gait disturbances in PD.

Supplemental Information