Background
Parkinson’s disease (PD) is one of the most common neurodegenerative disorders, manifesting mainly as bradykinesia, resting tremor and various non-motor symptoms. Increasing evidence has witnessed that the non-motor symptoms have a significant impact on the quality of life of patients with PD. The non-motor symptoms can occur either after motor onset or before motor symptoms at a prodromal stage [
1]. Sleep disorders are common non-motor symptoms, including excessive daytime sleepiness (EDS), insomnia, restless legs syndrome (RLS), circadian rhythm disorders, sleep attacks, obstructive sleep apnea and rapid-eye-movement behavior disorders (RBD). In addition, some sleep disorders such as RBD are considered as a prodrome of α-synucleinopathies [
2], suggesting that sleep disorders are closely related to the pathophysiology of PD.
Currently, multiple variants of genes, such as alpha-synuclein gene (
SNCA), leucine-rich repeat kinase 2 gene (
LRRK2), glucocerebrosidase gene (
GBA) and parkin gene (
PRKN), have been reported to be causes of PD. Different variants in such causative genes also play a role in the discrepancies in clinical manifestations and prognosis of PD. As one of the most common non-motor symptoms of PD, sleep disorders are strongly associated with some genes, since a growing number of genome-wide association studies have identified genetic risks for sleep disorders [
3]. Therefore, sleep disorders in patients carrying PD-related gene mutations have received much attention for research. However, studies have not reached a consensus regarding the genetic risk factors for sleep disorders in PD or prodromal PD. Some studies have found that PD patients carrying
GBA variants seem to develop RBD more frequently than patients without
GBA variants [
4]. However, another cohort study did not find any difference in the risk of RBD between patients with and without
GBA variants [
5]. Inconsistent results have been found on the prevalence and the severity of other sleep disorders, such as EDS and RLS, from studies on
LRRK2 [
6] or
PRKN variants [
7,
8]. Moreover, controversial conclusions regarding the role of genetic variants on sleep disorders exist when comparing asymptomatic carriers of causative gene variants with non-carrier healthy controls (HCs) [
9,
10]. Differences in the sites of variants of causative genes, disease durations of participants in cohorts, and the study design could contribute to such discrepancies. Therefore, the association of genetic heterogeneity with the risk of sleep disorders may differ, especially at different stages of PD, such as prodromal and clinical stages.
In this context, we systematically reviewed the genetic variants associated with the risk of sleep disorders in PD patients and asymptomatic PD genetic carriers to elucidate this inconsistency.
Discussion
In this systematic review and meta-analysis, we examined the associations between variants of causative genes of PD and sleep disturbance in both PD patients and asymptomatic carriers at the prodromal phase of PD. Our findings suggest that variants of the causative genes play a role in sleep disturbances in PD. We found that GBA variants increased the risk and severity of RBD in patients with PD, while the LRRK2 G2019S variant reduced the risk and severity of RBD. In addition, GBA variants worsened the RBD symptoms at the prodromal stage of PD, while variants of GBA and LRRK2 did not influence the risk of EDS in patients with PD. PRKN variants did not influence the risk of RBD, EDS, or RLS in PD.
Our study revealed relationships between
GBA variants and RBD in patients with PD.
GBA encodes beta-glucocerebrosidase, and biallelic pathogenic variants of
GBA1 cause Gaucher disease (GD), a lysosomal disorder. Both homozygous and heterozygous
GBA variants increase the risk of PD [
52].
GBA variants are classified into severe variants (such as L444P, W291X, H225Q, and IVS2 + 1G > A) and mild variants (such as N370S). In addition, several non-pathogenic variants of
GBA in GD, such as E326K and T369M, have been found to increase the risk of PD [
53]. PD patients with severe variants of
GBA have a younger age of onset, faster progression and more severe cognitive impairment than those with mild variants [
54]. In this meta-analysis, we divided
GBA variants into groups based on the severity of
GBA variants (L444P regarded as the severe variant, N370S as the mild variant, and E326K & T369M as the non-pathogenic variants), and showed that the pathogenic variants of
GBA increase the risk of RBD. Notably, the OR of
GBA L444P variant (2.02) [
19,
22] for RBD in PD patients was higher than that of the
GBA N370S variant (OR 1.68) [
19,
21]. However, there were no differences in the risk of RBD between patients with and without
GBA E326K or T369M variant [
5,
19]. Our current findings indicate that the role of
GBA variants in increasing the risk of RBD in PD differs depending on the severity of
GBA variants. RBD is considered as a prodrome of α-synucleinopathy, and
GBA gene is closely correlated with α-synuclein. Impaired proteolysis caused by GCase deficiency preferentially increases α-synuclein deposition and spread of α-synuclein pathology, while elevated α-synuclein decreases the GCase activity [
55]. Lower levels of α-synuclein in CSF have been detected in PD patients carrying
GBA variants compared to iPD patients, with a downward trend depending on the order of severe, mild and risk variants [
19]. Differences in α-synuclein expression in different
GBA variants may elucidate the discrepancy in phenotypes.
For the prodromal stage of PD, our meta-analysis found that
GBA variant carriers had worse RBD symptoms during the follow-up period [
10,
43]. Interestingly, no difference was found in the severity of RBD between
GBA variant carriers and HCs at baseline [
10,
43]. Statistical analysis of cross-sectional studies failed to find any difference in risk [
9,
10,
41,
42] or severity [
20,
41] between asymptomatic carriers of
GBA variants and HCs. These findings suggest that
GBA variant is a risk factor for the development of RBD at the prodromal stage with disease duration.
LRRK2 is one of the causative genes for autosomal dominant PD. Previous cross-sectional studies have indicated that patients with PD carrying the
LRRK2 G2019S variant, the most common variant, exhibit slower disease progression and milder motor symptoms than iPD patients [
56]. Our meta-analysis found that
LRRK2 G2019S decreased the risk and severity of RBD [
21,
22,
27,
29‐
34], but
LRRK2 G2385R did not affect the risk of RBD in patients with PD [
25,
26,
28,
36]. In prodromal stages, our pooled analysis suggested that
LRRK2 G2019S did not increase or decrease the risk and severity of RBD in asymptomatic carriers, consistent with a review by Tolosa et al. [
57]. In addition, our meta-analysis found no associations between other sleep disturbances, including EDS and RLS, and
LRRK2 G2019S and G2385R variants. Further studies are needed to elucidate the associations between EDS/RLS and
LRRK2 variants.
The pathophysiological mechanisms of
LRRK2 variants in PD are related to a variety of pathways such as kinase activity, autophagy and oxidative stress. The interplay between α-synuclein and
LRRK2 mutations accelerates the progression of α-synuclein-mediated neurodegeneration, with Rab proteins and chaperones serving as mediators [
58]. However, only half of PD patients carryin
g LRRK2 variants were found to contain brainstem synucleinopathy in postmortem studies, suggesting that some
LRRK2 variants may not be involved in central synucleinopathy [
59]. Most PD patients with
LRRK2 variants display loss of dopaminergic neurons in the substantia nigra. Interestingly, patients carrying the
LRRK2 G2019S variant show more Lewy body pathology than patients with other
LRRK2 variants [
60]. Skin biopsy revealed no difference in deposition of phosphorylated α-synuclein between
LRRK2 G2385R carriers and non-carriers [
25], whereas high levels of α-synuclein deposition in sympathetic noradrenergic nerves in skin biopsies were found in patients with
LRRK2 G2019S and R1441G variants [
59]. Reasons for the low risk of RBD in PD patients carrying the
LRRK2 G2019S variant require more pathological research for interpretation.
PRKN is one of the most frequently mutated genes in patients with early-onset PD [
15] and is involved in the mitochondrial function of neural cells. Hatice Kumru found that six out of 10 PD patients carrying
PRKN variants (7 carrying homozygous
PRKN variants) developed mild RBD as assessed by video polysomnography [
61]. However, our pooled analysis of cross-sectional studies showed no correlation between
PRKN variant and the risk and severity of sleep disturbances including RBD [
7,
8,
15,
37‐
40], EDS [
7,
8,
15,
37,
38,
40] and RLS [
7,
8,
40] in patients with PD. At the prodromal stage of PD, only one study showed no difference in the risk of RBD between asymptomatic carriers with and without
PRKN variants [
62]. Large-scale studies are urgently required to dispel the mist of inconsistent results on sleep disorders in patients with
PRKN variants.
SNCA encodes α-synuclein proteins, which abnormally accumulate in PD as a major component of Lewy bodies. Observational studies in cases and families have reported RBD in PD patients carrying various variants of
SNCA, such as A53T variant, duplications [
63] and triplications [
64]. However, two studies on the risk of RBD in PD patients carrying
SNCA A53T [
49,
50] demonstrated contradictory results. As for the prodromal stage of PD, carriers of
SNCA A53T reported RBD before the appearance of motor symptoms in a longitudinal study [
51]. Moreover,
SNCA methylation may play a role in the process of
SNCA expression, which is associated with RBD. Previous studies have found hypomethylation in the promoter and intron 1 regions of
SNCA in idiopathic RBD patients and PD patients compared to HCs [
65].
Patients carrying
GBA,
LRRK2,
PRKN, or
SNCA variants have different manifestations and pathogeneses in essence. Patients with
SNCA,
LRRK2, or
GBA variants have peripheral synucleinopathy, while patients with
PRKN variants do not [
59]. Moreover, PD patients with
PRKN variants mostly lack α-synuclein deposition in the brain and part of PD patients with
LRRK2 variants have α-synuclein deposition in the central nervous system.
GBA,
LRRK2 and
SNCA variants are associated with lysosomal dysfunction, while the
PRKN variant is directly involved in mitochondrial functions [
66]. It has been proposed that patients with
GBA or
SNCA variants typically have initial pathological α-synuclein originating in the enteric and autonomic nervous systems, whereas patients with
LRRK2 or
PRKN variants have initial α-synuclein pathology originating in the brain [
67].