Parkinson’s disease (PD) is usually sporadic, meaning that there is no family history. In approximately 10% of cases, however, there are affected relatives, which may implicate a hereditary-genetic form of the disease. Especially in patients of a young age (disease onset before 40 years of age), an underlying genetic cause is highly possible. Genetic forms may occur even in sporadic cases, again more commonly with a relatively early age of onset. In recent years, further genetic causes of PD have been discovered [1].

Such mutations are not particularly rare in certain populations. The clinical picture and course might vary depending on the exact genetic cause and the specific mutation, but there is also an intrafamilial variability. The genetic diagnosis of the disease is important in these cases not only for family planning and assessment of patient prognosis but also because it can be taken into account for specific therapeutic options, namely device-aided treatments [like Deep Brain Stimulation (DBS) or Levodopa-Carbidopa enteric gel pumps] and even more so for inclusion in emerging clinical studies that are applied to specific genetic forms of the disease, in the context of pharmacogenomics [2]. In the era of individualized therapies, the strategy of treatment in genetic PD forms may differ substantially compared to idiopathic PD [3].

The discovery of Parkinson’s disease (PD) related genes has revolutionized PD research during the last few years. Following the discovery of the SNCA gene in 1997 [4], an ever-increasing number of causative genes for PD have been described. Pathogenic mutations in autosomal-dominant genes like the α-synuclein (SNCA), leucine-rich repeat kinase 2 (LRRK2), and gene encoding for the vacuolar protein-sorting 35 ortholog (VPS35) and autosomal recessive genes like PRKN, PINK1 and DJ1 have been increasingly recognized as important genetic factors in young-onset PD or in cases of inheritance within families. The importance of the risk factor Glucocerebrosidase gene (GBA1), accounting for as many as 5–10% of PD cases, is outstanding. However, there has been a bias in existing literature data as most previous studies have assessed patient groups of European ancestry, mainly in Europe and North America. The majority of other races are underrepresented, including even different racial minorities of the aforementioned countries [especially in the United States or in European countries [5].

Missense mutations in the LRRK2 (leucine-rich repeat kinase 2) gene segregated with the disease in cases with autosomal-dominant inheritance. Such mutations were first described by two different groups of PD patients [6]. LRRK2 mutations are thought to relate to a gain-of-function mechanism. LRRK2 is considered to regulate the phosphorylation of Rab GTPases, thus regulating vesicular trafficking ^[1][7][1,7]. The p.G2019S mutation is notably the most common mutation (up to 4% of familial cases and 1% of sporadic cases in a large worldwide study) [8]. The phenotype of affected patients resembles idiopathic PD. Additional pathogenic mutations include p.R1441G/C/H [6], p.Y1699C [6] and p.I2020T [9].

In a multicenter study, patients harboring a LRRK2 mutation exhibited a rather favorable motor course and preservation of cognition, as assessed through the use of the MiniMental State Examination (MMSE), compared to idiopathic PD ^[8][10][8,10]. Moreover, a number of LRRK2 carriers might exhibit hyposmia [8]. Interestingly, REM Sleep Behavior Disorder (RBD) appears to be rather rare [11].

A notable number of cases with familial and/or sporadic forms of the disease with the p.G2019S mutation has been reported in Ashkenazi Jews (up to 15%) and North African Arab-Berbers (up to 30–40%) ^[12][13][12,13]. In contrast, this variant is relatively rare in familial PD in different populations, such as in Greece [14]. From a global perspective, the G2019S LRRK2 mutation [prevalent in patients of European ancestry (notably the Iberian Peninsula) and Ashkenazi Jewish/Northern African ancestry] is probably due to a common founder effect that seems to be scarce in other populations, except for those of Central Asia. The G2019S LRRK2 mutation was identified in Uzbek patients with idiopathic and familial PD (5.7% and 17.6%, respectively) [15]. A Whole-Exome Sequencing (WES) study including 50 index individuals with young-onset PD from Kazakhstan yielded a low prevalence of PD-related genes with two non-related familial PD patients with the LRRK2 p.(Arg1441Cys) mutation, while a few additional cases had extremely rare variants of uncertain significance in LRRK2 and other genes [16]. Notably, East Asian patients carried different variants, mainly R1441G, R1441C, G2385R and p.I2020T in Japanese populations. G2385R and R1628P were reported mainly in East Asian countries, where they could be detected in ~5–10% of cases. These have been considered to act as genetic risk factors, almost doubling PD risk [17]. When evaluating East Asian populations, A419V, R1628P, and G2385R LRRK2 variants were prominent [18]. In a large Japanese study, 1402 PD patients were assessed (sporadic PD n = 653 and familial PD n = 749). In total, 23 patients harbored pathogenic LRRK2 variants (four with p.R1441G, five with p.R1441H, seven with p.G2019S, and seven with p.I2020T), while additional patients carried other rare variants. Additional risk variants, p.P1446L and p.G2385R, were also detected in two groups of 10 and 146 patients. [17].  Finally, a large study from mainland China assessed 1676 unrelated patients with Parkinson’s disease (192 patients from families with autosomal-recessive Parkinson’s disease, 242 patients from families with autosomal-dominant Parkinson’s disease, and 1242 sporadic early-onset PD patients). In total, eight patients harbored pathogenic/likely pathogenic LRRK2 variants, including the p.N1437D, p.R1441C and p.R1441H variants, but not the variant common in Europe, p.G2019S ^[19][22]. A second Chinese cohort showed similar outcomes regarding LRRK2 variants, with only 2 out of 832 cases of PD carrying a pathogenic LRRK2 mutation, while p.G2019S was again not detected ^[20][23]. LRRK2 mutations were relatively rare in Indians. In the study of Vishwanathan and co-authors, after evaluating 140 Indian PD patients, only one harbored the G2019S LRRK2 mutation, while other studies failed to identify any carriers ^[21][24]. LRRK2 mutations appear to be ultra rare or absent in Black African population studies. In two Nigerian population studies (a modest cohort of Sub-Saharan African PD patients (n = 126) and controls HC (n = 54) and a second cohort of 92 Nigerians with PD and 210 HC), all patients and controls were negative for the p.G2019S mutation and other rare LRRK2 pathogenic mutations described in Caucasians, Asians, and persons of mixed ancestry ^[22][23][27,28]. 

23.1.2. SNCA

The identification of the p.A53T mutation in the α-synuclein (SNCA) gene was a milestone of PD genetics research [4]. Since 1997, additional missense SNCA mutations have been reported [1]. Moreover, multiplication mutations like SNCA duplication or triplication have been described. These mutations result in a gain of function that enhances the aggregation of the insoluble form of the protein and exerts a deleterious action likely at the level of the synapse and lysosomal function ^[24][32]. The p.A53T mutation was found in members of Italian or Greek families, possibly as a consequence of a common founder effect ^[4][25][4,33] and subsequently in a limited number of patients from other countries. The p.E46K mutation has been described in members of a family of Spanish Basque background ^[26][34]. Other variants, including p.A30P and p.G51D, have been reported in patients from different countries like Germany ^[27][35] and can have a variable atypical presentation. The phenotype of SNCA point mutations is largely heterogeneous. Non-motor features are common, including olfactory and autonomic dysfunction. RBD, depression or cognitive decline may also be often present ^[1][28][1,36]. SNCA triplications cases exhibit more severe symptoms, including concomitant cognitive decline and an early disease onset ^[29][37]. SNCA locus duplications have a variable clinical phenotype ^[30][38]. Non-motor symptoms, like dysautonomia, depression, or psychosis, were reported in approximately half of the reported cases. Pathogenic SNCA variants were either rare or absent in most populations of non-European ancestry. A study on the early-onset PD of the Kazakhstan population failed to identify any pathogenic SNCA mutation [16]. The p.A53T SNCA mutation has been identified in a single case of a Han Chinese PD patient ^[31][39]. In the large 2020 Chinese screening study, four probands carried SNCA pathogenic/likely pathogenic variants, including three patients with SNCA duplications and one patient with an unreported missense variant (p.P117S) ^[19][22]. A common SNCA pathogenic variant, p.A53V, identified in a Chinese cohort, occurs exclusively in Asians ^[32][40]. 

23.1.3. VPS35

Variants in the gene regulating the vacuolar protein sorting 35 ortholog (VPS35) protein are considered to be another cause of monogenic PD [1]. Affected carriers have typical PD symptoms, much like LRRK2 carriers, apart from an early PD initiation in some cases. In terms of pathophysiology, mutations in VPS35 impair vesicular recycling ^[33][50]. When first described, the mutation p.D620N was found in Austrian and Swiss families ^[34][51]. Clinically, PD patients harboring this mutation exhibited symptoms typical of idiopathic PD. No cognitive problems were noted except for a few patients with mild cognitive impairment. Moreover, the olfaction was decreased in about half of the subjects, and rarely, neuropsychiatric features have been described ^[34][51]. In a large Chinese study no VPS35 gene c.1858G > A (p.Asp620Asn) mutation could be detected in a sample of 1011 sporadic PD patients and 1016 controls. The authors suggested that the VPS35 variants are not associated with PD in the mainland Chinese population ^[35][54]. Accordingly, in a recent large study from mainland China, assessing 1676 unrelated patients with Parkinson’s disease, only one VPS35 mutation carrier was found ^[19][22].

2.2. Autosomal-Recessive Genes

3.2. Autosomal-Recessive Genes

23.2.1. PRKN (Parkin)

Pathogenic variants in the PRKN gene, which encodes for Parkin, represent a major genetic cause of early-onset PD [1]. Such variants were initially reported in patients from Japan with an autosomal recessive inheritance pattern ^[36][58]. Genetic changes are either missense or nonsense mutations, but they can often be Copy Number Variants (CNVs). It is apparent that these mutations cause a loss of function. Parkin is an E3 ligase whose role in the Ubiquitin–Proteasome System is crucial and results in the degradation of certain substrates. It is highly possible that PRKN mutations are linked to PD via mitochondrial dysfunction (in mitophagy which involves the inactivation of damaged mitochondria by means of autophagy) ^[37][59]. Pathology assessment in Parkin-related PD showed neuronal loss in substantia nigra and locus coeruleus, but no Lewy body pathology was evidenced. The phenotypic spectrum involves slow motor progression, frequent concomitant dystonia (mainly prominent in the lower limbs) with gait abnormalities and early-onset dyskinesias and motor fluctuations ^[38][60]. Cognitive function is not affected even in later disease stages ^[38][60], but psychiatric manifestations, including depression, anxiety, obsessive–compulsive symptoms, panic attacks or, occasionally, psychosis, may occur. PRKN pathogenic mutations appear to have a widespread distribution across the globe (Sub-Saharan Africa, East and Central Asia, India and Latin America). PRKN pathogenic mutations were present in Central Asian populations in Kazakhstan (1 patient out of 50 young-onset PD carried a homozygous pathogenic PRKN p.(Arg84Trp) variant) [16]. In East Asia, PRKN was particularly common in young-onset PD in mainland Chinese studies, in Taiwan, Japan and Korea. In the large Chinese 2020 cohort, PRKN was the most prevalent pathogenic variant, as 50 out of 192 patients with an autosomal-recessive inheritance pattern (26.04%) were carriers (27 PD patients harbored homozygous and 23 compound heterozygous variants). The prevailing form of mutations appeared to be exon deletions, mostly in exon 3 and exon 4. Additionally, 21 probands carried PRKN point mutations (7 p.G284R and 3 p.M1T). PRKN mutations were also the most common genetic finding in the sporadic early-onset cohort, accounting for 33 out of 1242 patients (2.66%) ^[19][22]. Regarding other Asian populations, a study assessing the genetic landscape of PD in Malaysia showed comparable outcomes, as 11 out of 161 young onset PD patients (6.8%) harbored PRKN mutations. PRKN exon 7 deletion appeared to be common among Malay patients ^[39][64]. In a Vietnamese early-onset PD population (N = 83), among 24 pathogenic/likely pathogenic mutations found in this study group, the second most common altered alleles were detected in PRKN (second to LRRK2), corresponding to 29%. ^[40][20].  In the Indian subcontinent, PRKN mutations are often encountered in the genetic analysis of PD cohorts. An assessment of 97 patients, including 52 sporadic early-onset PD and 45 familial PD, revealed a total of 25 (25.77%) patients carrying PRKN variants (5 homozygous/compound heterozygous and 20 heterozygous mutations only) ^[41][66]. In another Indian PD population study assessing 250 patients, 21 carried known pathogenic variants in recessive genes, and a novel PRKN variant in a homozygous state was identified ^[42][55]. Chaudhary and co-authors reported a prevalence of PRKN gene mutations in 8.5% of the PD population in India. They failed to find any homozygous deletions, but 9.2% of patients had heterozygous exon rearrangements ^[43][67].  PRKN mutations are occasionally found in Black African PD patients. Four PD cases harboring PRKN mutations have been described: two from South Africa carrying a heterozygous PRKN duplication in exon 2 and an exon 9 deletion; a heterozygous PRKN deletion in exon 4 and a point mutation, respectively ^[44][45][70,71]. Moreover, a patient from Zambia carried a heterozygous PRKN exon 2 and a PRKN exon 4 deletion [46], and a patient from Nigeria harbored a homozygous PRKN exon 4 deletion ^[47][72]. 

23.2.2. PINK1

After Parkin, PTEN-induced putative kinase 1 (PINK1) mutations are considered to be an important causative gene of young-onset PD worldwide and account for 1–5% of cases. As in the case of PRKN, there is variation as related to ethnicity [1]. Point mutations or dosage mutations (small deletions/insertions) have been described in a compound heterozygote or homozygous state ^[48][76]. There is evidence that the mechanism of action involves mitochondria. Its exact role is still obscure, but it possibly has a crucial role in mitophagy ^[49][77]. Regarding a patient’s phenotype, Ishihara-Paul (2008) evaluated sporadic and familial Tunisian PD cases ^[50][78]. PINK1 carriers often have a rather early age of PD onset. From a clinical perspective, these patients have a rather benign clinical course but present more often with difficulties in gait and balance. Cognition is less affected compared to other genetic PD causes. Again, as the case is with PRKN variants, the importance of monoallelic PINK1 mutations as a risk factor for PD remains elusive ^[51][79]. PTEN-induced kinase 1 (PINK1) has been reported to be an important recessive PD gene after PRKN in sporadic and familial PD in East Asia (China, Korea) and India. A group of Malaysian PD patients from multiple ethnic backgrounds (185/499 were Malays) revealed five cases of homozygous p. Leu347Pro variants, resulting in a frequency of 6.9% among Malays with EOPD ^[52][80]. PINK1 is a rare cause of early-onset PD in Japan, as a recent study assessing 1700 patients (842 with positive family history and 858 sporadic PD patients from Japan) reported thirty patients carrying heterozygous, three homozygous, and three digenic variants of PINK1-PRKN ^[51][79]. In a large 2020 Chinese cohort, PINK1 was rarer than PRKN since five patients harboring homozygous PINK1 pathogenic/possibly pathogenic variants and one patient harboring a compound heterozygous PINK1 mutation were detected in cases with autosomal-recessive inheritance, while only one case was found in sporadic PD ^[19][22]. In a Korean study, 70 patients with EOPD were screened for PD-related genes; three rare variants in PINK1 were detected, two of which were likely pathogenic ^[53][81].

23.2.3. DJ-1

DJ-1 represents a scarce recessive type of early-onset PD. DJ-1 protein is important in the anti-oxidant response and possibly shares a biological pathway with PINK-1 and Parkin ^[1][54][1,86]. A homozygous deletion involving DJ-1 was initially reported in the Netherlands, while a homozygous point mutation, p. L166P, was described in a family from Italy ^[54][86]. The clinical features resemble Parkin-related PD, with early age of onset, occasional dystonic symptoms, marked L-Dopa-induced motor fluctuations, and anxiety or depression ^[55][87]. Parkinson disease protein 7 (DJ-1) has been described rather rarely in East Asia (China, Taiwan, Malaysia and Korea). In a Chinese study, 3 out of 127 (2.4%) unrelated early-onset PD patients carried possibly pathogenic DJ-1 variants ^[56][88]. Zhao and co-authors reported only one familial DJ-1 case (possibly a pathogenic variant) with autosomal recessive inheritance in a large Chinese cohort ^[19][22].

2.3. GBA1

3.3. GBA1

GBA1 gene mutations represent a common genetic predisposing factor for PD. Glucocerebrosidase (GCase) (the protein product of the GBA1 gene) is a lysosomal enzyme that degrades glucosylceramide into glucose and ceramide. A multitude of GBA1 variants have been identified, with the majority causing severe attenuation of glucocerebrosidase activity [1]. Except for its role in Gaucher’s disease, GBA1 mutation heterozygotes lead to an increased incidence of PD ^[57][92]. The proposed causative mechanisms of PD comprise the aggregation of α-synuclein and the dysregulation of the homeostasis of lysosomes. GBA1 mutations result in an earlier PD initiation by about 1.7–6.0 years ^[57][58][92,93]. Glucocerebrosidase gene variants (GBA1) were found in over 5% of Asian cases [Mainland China, Taiwan, Korea and Thailand (mostly the p.L444P, RecNcil and R120W while the N370S was rare)] (Table 1). GBA1-related PD was rather prevalent in the Chinese population. Yu and co-authors evaluated a Chinese cohort (184 PD and 130 HC) by means of complete sequencing of the GBA1 gene. The prevalence of GBA1 variants in this group was 8.7%. Identified variants included three novel (5-bp deletion (c.334_338delCAGAA), L264I and L314V) and nine already reported GBA1 variants (R163Q, F213I, E326K, S364S, F347L, V375L, L444P, RecNciI and Q497R) ^[59][96]. GBA1 mutations are also present in patients of Malaysian origin (with the L444P mutation being the commonest). A large study of a multiethnic PD population in Malaysia evaluated the spectrum of GBA1 variants. In this study, 496 patients were screened (n = 496) belonging to Chinese, Malay, and Indian ethnical backgrounds. The researchers managed to identify 14 heterozygous GBA alleles in 25 patients (5.0%). The most common variant was p.L483P [p.L444P] (including RecNciI, (2.2%)), found in all groups along with newly described variants. However, the usual mutations found in Europe, p.E365K, p.T408M, and p.N409S (N370S), were not reported ^[60][100]. GBA1 mutations also represent an important risk of PD in the Thai population. A group of 108 EOPD patients and 100 PD with disease initiation over 50 years were evaluated for GBA1 mutations. Heterozygous GBA1 mutations were detected in 24 patients (5%) The seven identified GBA1 point mutations included p.L444P, p.N386K, p.P428S, IVS2+1G > A, IVS9+3G > C, IVS10-9_10GT > AG and c.1309delG, of which five variants had not been previously described ^[61][101]. GBA1 variants have also been found in a considerable number of cases in India, mainly the L444P, and other variants rare in Europeans. Kukkle and co-authors analyzed whole-genome sequencing data from a cohort of 90 young-onset Parkinson’s disease and found GBA1 heterozygous variants in 13 individuals (14.4%) [including E365K, L483P (L444P), R159Q, D448H, P454L, R502P and R502H variants] ^[62][102]. GBA1 mutations have been rarely identified in Black African populations (with different variants as opposed to common risk GBA1 variants in Europeans like N370S). In a study on thirty South African PD patients of black ancestry, all 11 exons of GBA1 were evaluated, and the research group found two carriers of pathogenic mutations (p.R120W in and p.R131L) ^[63][104]. Furthermore, in a Nigerian population study, the screening of GBA1 mutations revealed six putative pathogenic mutations only in patients (p.W184R, p.L383PfsX3, and 3 p.L444P) (6.5% of PD cohort) ^[64][105]. At this point, we should highlight the fact that a very recent extensive genome-wide association study (GWAS) in people of African and African mixed ancestry with and without PD managed to identify a novel risk factor in GBA1 at the rs3115534 g locus in people of African ancestry. This non-coding variant (found in 39% of Sub-Saharan PD patients) seemed to be associated with reduced Glucocerebrosidase activity, which was absent in European populations and could, consequently, constitute a novel disease-causing mechanism for PD in African populations. This discovery, in particular, highlights the value of performing genetic studies in PD cohorts that are diverse ethnically, beyond Caucasian populations ^[65][106].

2.4. Other Rare Genes

.4. Other Rare Genes

Mutations in additional rare genes have been described in Chinese and East Asian studies (CHCHD2, ATP13A2, PLA2G6, EIF4G1 and FBX07). Phospholipase-associated Neurodegeneration (PLAN) with PLA2G6 mutation involves a spectrum of three distinct clinical manifestations, including PLA2G6-associated dystonia-parkinsonism. In a Chinese study, variants in PLA2G6 were not particularly rare and accounted for 1.89% of early-onset PD ^[20][23]. An Indian group described juvenile PLA2G6-Parkinsonism due to a p.R741Q mutation ^[66][109]. A Chinese population study showed that the Pro2Leu variant in CHCHD2 may be related to the development of PD among Asians ^[67][110].