Introduction

Parkinson’s disease (PD) is the second most common neurodegenerative disease after Alzheimer’s but ranks first in the world as neurodegenerative movement disorder [1]. The various monogenic forms of PD account for a minority of PD cases. PD most commonly occurs in an idiopathic form, where genetics, environmental exposure, and aging are related risk factors [2, 3].

PD’s pathological characteristics include dopaminergic neuronal loss in the substantia nigra at the central nervous system (CNS), the subsequent dopamine (DA) level reduction affecting motor function, and Lewy bodies’ presence [4, 5]. Also, mitochondrial dysfunction, oxidative stress (OS), and abnormal protein accumulation are involved in PD’s pathogenesis [6].

PD patients may suffer from motor and non-motor symptoms (cognitive and behavioral) and those related to the autonomic nervous system (ANS) [7]. PD cardinal motor symptoms are tremor, bradykinesia, rigidity, and postural instability [8]. Some pathologic changes that may appear up to 20 years before the motor symptoms onset [7] are related to ANS failures, including gastrointestinal, urinary, and cardiovascular symptoms, which increase with age, the severity of the disease, and higher dopaminergic medication doses [9]. Constipation is an early gastrointestinal symptom that precedes motor symptoms. Men with less than one bowel movement per day have 4.1 times the risk of developing PD compared to men who have two bowel movements per day, and the risk increases 4.5 times compared to men who have more than two bowel movements per day [10]. There, the importance of non-motor symptoms as a potential warning for PD progression since they can precede motor symptoms by decades.

Paradoxically, although the current technological advances are rapidly growing, the generation of new effective disease-modifying treatment is not yet available in the clinic. The current dopaminergic replacement therapy for PD prevails 60 years after the dopaminergic deficit discovery [11]. Although there are several FDA-approved medications for PD (anticholinergics, carbidopa/levodopa, catechol-o-methyl transferase (COMT) inhibitors, dopamine agonists, MAO-B inhibitors, and NMDA receptor inhibitor), all of them target symptomatology, primarily motor symptoms [12].

Several biomolecules, including non-enzymatic antioxidants, and synthetic drugs, have yielded positive outcomes in PD models. However, they have failed to reproduce such results in clinical trials [13]. Besides, numerous drugs are still under evaluation in clinical trials [14]. Nonetheless, there are currently no PD-modifying treatments [15].

Current Diagnosis of Parkinson’s Disease

Despite advances in neurodegenerative disease research and improved neuroimaging and genetic studies, PD diagnosis remains mainly dependent on observational clinical criteria, mostly clinical motor symptoms [16] manifesting when patients have lost about 60–80% of dopaminergic neurons [17].

Since the development of the Unified Parkinson’s Disease Rating Scale (UPDRS) in the 1980s, the Movement Disorder Society (MDS) has evaluated [18] and updated the diagnostic criteria to include also the non-motor symptoms [19]. An update for the research criteria for prodromal PD has also been published to achieve an early diagnosis [20]. Prodromal PD encompasses patients in a stage where they do not fulfill diagnostic criteria for PD but exhibit nonmotor signs and symptoms, increasing the risk of developing motor skills symptoms and PD in the future. Among prodromal PD symptoms are hyposmia, constipation, mood disorders, and REM sleep behavior disorder (RBD), which significantly impact the life quality early and once the disease has progressed to motor PD [21]. However, these criteria only allow a phenotypic clinical classification of PD that includes the subtypes of tremor dominant (TD), postural instability-gait disorder (PIGD), nonmotor mild cognitive impairment (PD-MCI), and dementia (PDD).

Notwithstanding the effort to improve the diagnosis accuracy, it is not yet satisfying. The diagnosis accuracy is 75% for general neurologists [22] and 79% for movement disorders experts. The overall validity of clinical diagnosis has not improved in the last 25 years [23].

Currently, clinical diagnosis is supported by imaging studies, including (1) computed tomography (CT) scan, consisting of passing X-rays from different angles to obtain brain cross-sectional images, helps to rule out other diseases with similar symptoms to those of PD [24]. (2) Magnetic resonance imaging (MRI) scan obtains weighted images by measuring the tissues’ water diffusion speed. Because of the dopaminergic neuronal death and the reduction of the region’s volume in PD, a greater water molecule diffusivity occurs, helping to distinguish PD from similar diseases [25]. (3) Dopamine transporter single-photon emission computed tomography (DaTSCAN™ SPECT) scan requires the intravenous administration of ioflupane 123I as 123I-FP-CIT, which binds to the DA transporter (DAT) in the dopaminergic neurons presynaptic membrane, making evident the loss of dopaminergic neurons in PD [26]. (4) Positron emission tomography (PET) scan uses radiotracers that emit positrons to evaluate the presynaptic DAT presence. The radiotracers employed are 18F or 11C radiolabeled DA analogs. Besides, the vesicular monoamine transporter 2, which also transports DA, can be assessed using radiolabeled dihydrotetrabenazine (DTBZ) [25]. Since the specific patterns of regional glucose metabolism are known, the 18F-FDG radiotracer is used to evaluate the brain’s glucose metabolism as a neuronal function marker to detect PD’s specific alterations due to synaptic dysfunction, allowing differential diagnosis between parkinsonism and PD [27]. (5) Scintigraphy of cardiac 123I-metaiodobenzylguanidina (123I-MIBG), which is a synthetic analog of norepinephrine, allows differential diagnosis of PD and MSA (multiple system atrophy) and is based on postganglionic sympathetic neurons integrity. In PD, pre and postganglionic autonomic neurons are affected by α-synuclein; therefore, cardiac 123I-MIBG uptake is disrupted. In contrast, in MSA, where autonomic deterioration is mainly preganglionic, uptake of 123I-MIBG is assumed to be preserved [28].

However, all these imaging techniques have some disadvantages when used in PD diagnosis. For instance, CT and MRI scans are used for differential diagnosis but not to confirm PD. DaTSCAN™ SPECT and PET scan evaluate DATs. Therefore, they can detect the first signs of dopaminergic damage but cannot differentiate PD from atypical parkinsonism with dopaminergic dysfunction [25]. Furthermore, DaTSCAN™ SPECT and PET scans are costly, and of limited access, so the latter is mainly used in research. In the case of cardiac MIBG scintigraphy, its specificity is limited in the presence of cardiac damage such as cardiomyopathy and myocardial infarction, in addition to peripheral nerve diseases such as diabetes and other polyneuropathies [29].

The main obstacles to finding a cure for PD are inaccurate diagnosis and deficient availability of high-quality human tissues [22]. Therefore, pursuing the identification of biomarkers for early PD diagnosis is critical; with effective, not invasive biomarkers, a chemical test could be created, so those at risk could be identified and treated before any overt signs or symptoms.

Biomarkers

Biomarkers are defined as a characteristic that can be objectively measured and evaluated as an indicator of normal or pathogenic processes or pharmacological responses to therapies [30]. Biomarkers can be cells, lipids, proteins, genes, or metabolites; their presence; variation in concentration; or even their molecular modification [31]. Biomarkers can be detected and measured in biological samples such as tissues, cells, or ideally in body fluids (blood, plasma, saliva, urine) due to their easy access [32].

Pursuing an ideal PD biomarker is challenging, mainly when signs and symptoms take decades to appear, and even more, to chase early stages or predictive biomarkers when there are no signs of disease and its origin is idiopathic. Therefore, it is likely that a panel of biomarkers will provide greater diagnostic accuracy than any single marker.

Major concerns related to biomarkers search are individual or group variability, sensitivity, specificity, and effect modifiers. Therefore, reliability is essential, which is usually affected by differences in storage, transport, methods, and instruments [33].

Herein, we discuss integrating multiple candidate biomarkers, proposing those less invasive and more reliable, to form a potential panel of early PD diagnosis and its progression. Therefore, CSF biomarkers are not addressed since CSF sampling is an invasive outpatient procedure that has to be performed by an experienced and specialized physician and may not be performed in developing countries as easily as in developed countries. Besides, CSF biomarkers were recently reviewed by Parnetti et al. [34]. Establishing a validated biomarker panel for early PD diagnosis ideally may allow treatment of patients before or at the earliest stages of a diagnosable form of the disease and interventions targeting specific biological processes associated with disease progression [35].

Potential Biomarkers for the Diagnosis of PD

Is α-Synuclein the Leading PD Candidate Biomarker?

α-Synuclein (α-Syn) is the main component of Lewy bodies, whose presence is one of the primary PD’s pathological characteristics [5]. α-Syn is encoded by the SNCA gene, which has three main domains: (1) the N-terminal domain, which has a multiple repeat consensus sequence (KTKEGV); (2) the central domain, called non-amyloid β component (NAC), which is highly hydrophobic and participates in α-Syn aggregation when it acquires the structure of the β-sheet; and (3) the C-terminal domain, enriched in proline residues and negatively charged, which provides flexibility [36]. α-Syn has dynamic conformations stabilized by long-range interactions between the C-terminal end and the NAC region, and between the N and C-terminal ends, probably due to hydrophobic and electrostatic contacts that may prevent aggregation [37]. SNCA gene multiplication and point mutations, environmental changes, and post-translational modifications can alter the α-Syn native structure, inducing misfolding and aggregation [38, 39]. Among the post-translational modifications, phosphorylation affects the α-Syn function [38], and OS increases phosphorylated α-Syn at S129 (pS129-α-Syn) and inclusion formation [40]. Also, pS129-α-Syn is found abundantly in various synucleopathies, including PD [41]. It has been reported that α-Syn is not a sensitive biomarker for PD diagnosis in the early stages when analyzed using the immunohistochemical staining method in the skin and submandibular gland (invasive examination), obtaining a sensitivity of 24.1% and 56.1%, respectively [42]. Interestingly, an increase of α-Syn has been found in peripheral tissues and fluids using other methods, including real-time quaking-induced conversion (RT-QuIC) and protein misfolding cyclic amplification (PMCA), which are described below, along with other promising PD biomarkers (Fig. 1).

Fig. 1
figure 1

Potential systemic biomarkers for the diagnosis of Parkinson’s disease. *The fibroblasts of PD patients reflect the changes observed in dopaminergic neurons during the development of PD, such as OS and mitochondrial and autophagy dysfunction. Created with Biorender.com

Full size image

Skin Biomarkers

The skin is a promising organ for the biomarkers search for diagnosing PD because it shares the embryological origin (ectoderm) with neural tissue, is highly innervated, easily accessible, and its fibroblasts reflect environmental and aging changes [43, 44].

The association between the brain and the skin was evidenced by detecting α-Syn inclusions in the epidermis and skin appendages of PD patients while they were absent in control subjects. Besides, relatively high α-Syn levels were observed in PD patients’ skin, while atypical parkinsonism patients had few inclusions, providing a guideline to diagnose PD and differentiate between these two movement disorders [45].

Recently, α-Syn has also been detected in the skin through the RT-QuIC and PMCA techniques. In the RT-QuIC, a soluble recombinant prion protein (rPrP-sen) is used as a substrate to amplify minute amounts of the abnormal prion protein (PrPSc) with vigorous intermittent shaking that induces rPrP-sen aggregation, forming fibrils. Then, protease-resistant rPrP fibrils (rPrP-res) are detected with thioflavin T fluorescent dye [46].

Similarly, the PMCA takes advantage of the nucleation-dependent prion replication process to accelerate the conversion of cellular prion protein (PrPC) to PrPSc but using ultrasound waves to fragment the PrPSc polymers and increasing the number of seeds present in the infected sample without affecting its ability to act as a conversion nucleus. The final result is detected by western blot [47].

Through RT-QuIC, α-Syn differences between postmortem skin samples of PD patients and control subjects were detected with high sensitivity and specificity, varying depending on the sample fixation method. Formalin-fixed and paraffin-embedded skin sections showed 75% sensitivity and 83% specificity, whereas frozen skin tissues had 96% sensitivity and specificity [48]. In another study, skin samples from PD patients were analyzed by the RT-QuIC and PMCA techniques, obtaining 95% sensitivity and 100% specificity for RT-QuIC, and 80% sensitivity and 90% specificity for PMCA [49]. In addition, PMCA has a lower sensitivity and specificity; as a disadvantage, it allows the detection of approximately 1 attogram (10−21 kg) of PrPSc in approximately 3 weeks, while RTQuIC can detect approximately one lethal prion dose in one day [50].

Additionally, increased α-Syn deposition was found in cutaneous sympathetic adrenergic and cholinergic peripheral nerve fibers of PD patients related to higher autonomic dysfunction in advanced PD [51]. Also, skin fibroblasts from PD patients could be used to monitor alterations in morphology, growth patterns, mitochondrial function, autophagy, and OS [44].

Recently, a hyperosmic or “super smeller” individual contributed to identifying a distinctive signature of volatile metabolites from the skin’s sebum in PD patients. This PD distinctive signature consists of the perillic aldehyde decrease and the eicosane increase compared to the controls, resulting in a slight modification in people’s aroma [52]. Therefore, skin biopsies may be a non-invasive alternative for PD diagnosis.

Digestive System Biomarkers

The search for biomarkers in the digestive system is crucial because PD patients can suffer early intestinal inflammation and dysfunction [53]. Up to 30% of PD patients report gastrointestinal symptoms, commonly observed at all stages of the disease but frequently precede motor symptoms, indicating an early digestive tract involvement in the pathological process [54]. Indeed, there is a direct relationship between the gut and the CNS through the “gut-brain axis,” which is bidirectional [55].

Exosomal α-Syn

Exosomes are cell-derived vesicles (30–100 nm) generated by the endosomal pathway and released through exocytosis to the extracellular space and circulation [56]. Exosomes enclose specific macromolecules mirroring the cytosol content of their cellular origin [57]. They are secreted by many cells, including neurons, astrocytes, endothelial [58, 59], and abundant in distant cells and different body fluids, such as saliva [60].

Most proteins involved in neurodegenerative diseases are transported in exosomes [61]. Cerebrospinal fluid (CSF) α-Syn was transported towards blood, with a small portion being contained in exosomes relatively specific to the CNS [62]. α-Syn oligomers were found outside and inside exosomes, probably as a clearance mechanism, particularly when autophagy is altered. These exosomes are more likely to be internalized by recipient cells and are more toxic than free α-Syn oligomers [63]. Exosomes found in saliva samples from PD patients contained a higher proportion of oligomeric α-Syn than controls [64]. Another study showed that total α-Syn levels in saliva from early, moderate, and advanced PD patients were not significantly different between PD and healthy groups [42].

Oral Microbiota

The oral cavity has a moist and warm, suitable environment for microbiota adhesion and colonization [65]. Salivary antimicrobials are critical to maintaining a symbiotic relationship between the host and its resident microbiota [66]. Some oral bacteria routinely survive transit to the gut, and their functional activity is significantly reduced [67]. From oral to the gut, the increased microbial transmission was observed in diseased patients [68]. Early-stage PD patients have shown changes in the oral cavity’s microbiome, most frequently in the taxa abundance. Among the significantly increased bacteria were Acidaminococcus, Bifidobacterium, Brucella, Cellulosimicrobium, Clavibacter, Gardnerella, Lactobacillus, Methylobacterium, Rhodococcus, and Scardovia, while Buchnera, Chryseobacterium, and Wenyingzhuangia were significantly decreased compared to the controls. The bacteriophage Streptococcus phage PhiSpn 200 was significantly decreased, while the yeasts Candida albicans, C. dubliniensis, and Saccharomyces cerevisiae were increased in PD subjects. The microbial metabolic pathways related to serine metabolism, glycolysis, and pentose phosphate showed an increased expression, while the tryptophan’s metabolism and tricarboxylic acid cycle pathways decreased in PD subjects [69]. A strong separation of groups was achieved using a microbiota data set transformed in 5 proportions of 11 taxa in total, and only 13 subjects were misclassified from a total of 84 subjects, with an overall precision of 84.5%. A partial least squares discriminant analysis (PLSDA) showed almost complete separation of PD and control subjects when examined in a multidimensional setting [69]. Saliva might then be a feasible alternative for studying oral microbiota in health and disease [70]. However, there are currently few studies on the analysis of the oral microbiota in PD, so there is a long way to go before they can be used for the early diagnosis of PD.

Intestinal Permeability and Inflammation

Intestinal inflammation has been detected in PD patients in ascending colon biopsies [71]. The inflammatory process and intestinal barrier dysfunction are interrelated. The intestinal mucosal barrier’s integrity depends mainly on the epithelium tight junctions (TJ), which regulate the biomolecules’ paracellular transport. When the intestinal epithelium is damaged, serum proteins diffuse into the gut lumen. Intestinal permeability is reversibly regulated by zonulin (pre-haptoglobin 2) modulating TJ [72]. Upon intestinal exposure to enteric bacteria, zonulin is secreted to the lumen, followed by an increased intestinal permeability because of the zonula occludens 1 protein dissociation from the TJ complex [73]. The protease α-1 antitrypsin is also produced in the liver; it plays an immunomodulatory function by inhibiting neutrophil elastase and may reach the intestinal lumen, mainly when there is damage [74]. Therefore, elevated concentrations of fecal zonulin and α-1 antitrypsin reflect an increased intestinal permeability.

The loss of the intestinal barrier’s integrity leads to inflammation. Neutrophils are the first to arrive at the inflammation sites; monocytes/macrophages follow them and together orchestrate immune responses against infections [75]. Upon stimulation, neutrophils and monocytes/macrophages release calprotectin, a calcium binding-protein that belongs to the S100 protein family [76, 77]. Calprotectin regulates myeloid cell adhesion to endothelium and extracellular matrix and has antimicrobial properties by sequestering zinc [78, 79]. Likewise, lactoferrin is an iron-binding glycoprotein that belongs to the transferrin family; it is also secreted by activated neutrophils and has a critical role in the innate immunity as bactericidal [80]. Since calprotectin and lactoferrin stability and presence in feces are proportional to neutrophil migration to the gastrointestinal tract and the inflammation degree, both are considered inflammation biomarkers [81, 82].

The four biomarkers described above were evaluated in PD patients’ stool samples. Zonulin, α-1-antitrypsin, and calprotectin were significantly elevated in PD patients than age-matched controls, while lactoferrin showed a non-significant trend towards elevated concentrations. However, none of the four fecal biomarkers correlated with disease severity, subtype, dopaminergic therapy, or constipation [83]. No significant differences were observed in zonulin in another study, but calprotectin levels were significantly higher in PD patients than in healthy subjects. However, there was no correlation between the calprotectin level and the disease duration [84].

α-Syn in the Gastrointestinal-Nervous System

Lewy bodies containing α-Syn have been found in the mucosa, submucosal, and myenteric plexus in the enteric nervous system (ENS) of PD patients [85, 86]. α-Syn presence in the ENS was higher in the stomach, intestine, and appendix biopsies from PD patients than controls [87]. PD patients’ intestinal hyperpermeability significantly correlated with bacterial invasion, OS, and α-Syn accumulation in the intestinal mucosa [88]. Besides, a large-scale study found that α-Syn accumulation in the gastrointestinal tract occurs before the motor symptoms onset and principally consists of pS129-α-Syn [86].

Intestinal Microbiota

The brain and gut are connected through the ENS, the vagus nerve, the immune system, and the microbiota’s metabolites. Intestinal microbiota affects the development and function of the CNS through the microbiota-gut-brain axis [55]. Upon vagotomy, the therapeutic effects on the brain caused by probiotic bacteria such as Lactobacillus rhamnosus and Bifidobacterium longum are absent [89, 90]. Using transgenic mice overexpressing α-Syn, it was demonstrated that gut microbiota is essential for motor deficits, microglia activation, and α-Syn accumulation. Antibiotic treatment mitigated, while microbial recolonization promoted the pathology. Besides, oral administration of specific microbial metabolites to germ-free mice promoted neuroinflammation and motor symptoms. Also, colonization of α-Syn-overexpressing mice with microbiota from PD patients intensified physical impairments compared to microbiota transplants from healthy donors [91]. The intestinal microbiota also plays an essential role in the microglia maturation and proper function in the CNS. A study showed that germ-free mice had microglia with an immature phenotype and altered cell ratios, resulting in an impaired innate immune response. Microglia also were defective in mice in which the microbiota was temporarily depleted or had limited complexity. On the contrary, after recolonization with complex intestinal microbiota, the microglia were partially restored. Furthermore, these effects are regulated by the short-chain fatty acids (SCFA) that are produced from the dietary fibers during bacterial fermentation in the large intestine [92].

Interestingly, the elderly has unusual microbiota proportions and extreme variability [93]. Modifications in the intestinal microbiota profile in PD patients have also been reported. So far, due to the intestinal microbiota’s complexity, it has not been possible to establish a specific composition as a biomarker for this disease. However, few studies in different populations have shown modifications in the intestinal microbiota composition in PD patients (Table 1). Among the more consistent modifications of the intestinal microbiota, the families Lactobacillaceae, Enterococcaceae, Christensenellaceae, Verrucomicrobiaceae, and Enterobacteriaceae were increased.

Table 1 Modifications to the PD-associated microbiota
Full size table

In contrast, Bacteroidetes, the family Prevotellaceae, and Lachnospiraceae were decreased in PD patients compared to control subjects (Table 1), and the fecal SCFA were also reduced [94]. Fecal SCFA includes acetate, propionate, and butyrate [95]. Butyrate plays a vital role in intestinal mucosa homeostasis as it reduces OS by increasing glutathione and exerts an effect on the ENS [96, 97].

Interestingly, modifications in the microbiota are correlated with mitochondrial metabolism and mitochondrial DNA (mtDNA) haplogroup variants [98, 99], which may be explained because both mtDNA and microbiota are inherited from the mother [100, 101]. Also, mitochondrial redox status and reactive oxygen species (ROS) production changes are associated with intestinal microbiota modifications, indicating that mitochondrial function controls the microbiota composition [102].

Blood Biomarkers

Seeking blood biomarkers has excellent advantages like easy sample obtention, minimal invasion, and low cost [103].

Erythrocyte α-Syn

α-Syn can be detected in the blood, which is higher in individuals with SNCA gene multiplication [104]. More than 99% of α-Syn in the blood is localized in erythrocytes [105]. Total and aggregated α-Syn is higher in the erythrocyte membrane, while pS129-α-Syn is higher in the cytosolic compartment of PD patients than in control subjects [106]. However, a recent study found that total α-Syn levels in blood from early, moderate, and advanced PD patients were not significantly different between PD and healthy groups [42].

Nonmercaptalbumin

Albumin is one of the primary antioxidants in human serum (HAS) capable of scavenging hydroxyl radicals through its reduced cysteine residue (Cys34). Human albumin with the reduced cysteine thiol group is known as human mercaptalbumin (HMA), while the oxidized form is called human nonmercaptalbumin (HNA). Because albumin is widely distributed intravascularly and extravascularly, its oxidation reflects a systemic oxidative state [107].

Therefore, the redox ratio of HNA to HSA, defined as % HNA, was investigated in patients with idiopathic PD and autosomal recessive familial PD due to parkin mutations (PARK2). Patients with idiopathic PD had a higher % HNA than healthy controls, while patients with PARK2 mutations did not show a significant difference compared to controls [108].

Neurofilament Light Chain

Neurofilament light chain (NfL) is a neuronal cytoplasmic protein highly expressed in large-caliber myelinated axons [109]. Under normal conditions, axons constantly release NfL in an age-dependent manner to release the highest levels at older ages [110, 111]. However, NfL release increases in response to axonal damage due to inflammatory or neurodegenerative injury in the CNS [109]. Therefore, NfL levels have been evaluated in several studies in the blood of PD and atypical parkinsonism patients. Unfortunately, there is no consensus because of the inconsistency of the results. In some studies, differences were found in blood NfL levels between PD patients and healthy subjects [112, 113], while others did not find differences [111, 114, 115]. However, an increase in NfL was reported in patients with atypical parkinsonism (such as progressive supranuclear palsy, multiple system atrophy, and corticobasal syndrome) compared to healthy subjects and even to PD patients [111, 114, 115]. Differences between the results may be due to variations between populations, as demonstrated in a study where NfL levels were analyzed in PD patients from Sweden and London. Significant differences were found between PD patients and control subjects in the London population, while no significant differences were detected in the Switzerland population [116]. Furthermore, NfL levels are directly related to the prognosis of PD patients [111, 117].

Mesencephalic Astrocyte-Derived Neurotrophic Factor

Mesencephalic astrocyte-derived neurotrophic factor (MANF) is an 18-kDa arginine-rich soluble protein localized in the endoplasmic reticulum (ER), but it is also secreted upon ER stress-induction [118]. MANF exerts a selective protective role on dopaminergic neurons [119]. Its neuroprotective effects may occur through the ER stress pathway [120] induced by unfolded protein accumulation and leading to unfolded protein response (UPR). The UPR increases the protein folding capacity and the misfolded proteins’ degradation to counterbalance ER stress [121]. Importantly, MANF has immune-modulatory properties. Its levels decline with age in serum [122], which correlates with some PD pathogenesis features, like misfolded protein aggregation and inflammation. However, MANF serum concentration was elevated in PD patients [123], probably to counteract misfolded protein aggregation and inflammation.

Resolvin D1

Pro-resolving mediators, including resolvins, stimulate the inflammation resolution in PD models and modulate the disease progression of PD patients [124]. The resolvin D1 (RvD1) is an anti-inflammatory lipid mediator [125]. The evidence suggests a strong link between RvD1 and mitochondrial homeostasis. In a recent study, where mice with traumatic brain injury were treated with RvD1, damaged mitochondria and mitochondrial ROS were eliminated by mitophagy in astrocytes [126]. Also, plasma RvD1 levels were drastically reduced in early PD patients (duration of symptoms 13 ± 5 months) compared to control subjects [127]. However, it needs to be validated in a larger group.

Urate and Homocysteine

Since OS plays a crucial role in PD dopaminergic neuronal loss [128], it is not surprising that antioxidant levels are altered in PD patients. In humans, urate is the end product of purine metabolism and is present intracellularly and in body fluids as the anionic form of uric acid [129]. Uric acid circulates in the blood at high concentrations near its solubility limits with a physiological range between 240 and 350 µM [130] and represents 60% of the antioxidant capacity of plasma [131]. Its decrease has been related to a greater susceptibility to OS and is a high-risk predictor for developing PD [132]. Also, postmortem substantia nigra of PD patients showed 54% less uric acid levels than control subjects [133]. However, the association of low urate levels with a higher risk of PD and the faster progression of the disease have been observed only in men [134, 135] but not in women [136, 137]. Therefore, this metabolite may not be useful in the diagnosis of PD in women.

On the other hand, homocysteine (Hcy) is an amino acid that contains a thiol group generated through methionine demethylation. Methionine reacts with ATP to produce S-adenosyl-L-methionine, which is then demethylated to form Hcy [138]. It is present in plasma in three forms: bound to proteins through disulfide linkage (~ 70–80%), as free Hcy in a reduced form (~ 2%), and the remaining portion correspond to Hcy-cysteine disulfide [139]. Total Hcy ranges between 5 and 15 µmol/L in healthy individuals [140]. Elevated Hcy in plasma has been associated with cognitive impairment and dementia [141]. Notably, the total mean Hcy was significantly elevated in idiopathic PD patients’ plasma than controls [142]. Recently, low urate and high Hcy serum levels were associated with decreased motor function, while only high Hcy serum levels were associated with cognitive decline in early PD progression [143].

Mitochondrial Creatine Kinase

Creatine kinase (CK) is a central controller of cellular energy homeostasis by catalyzing a phosphoryl group’s reversible transfer from MgATP to creatine, producing phosphocreatine and MgADP [144]. There are four major CK isoenzymes: two cytosolic forms, the muscle (MMCK) and brain (BBCK) forms [145], and two mitochondrial forms (MtCK), the sarcomeric (sMtCK) and the ubiquitous (uMtCK) forms [146]. MtCK is highly susceptible to reactive oxygen and nitrogen species damage since most of them originate from the mitochondrial respiratory chain [147]. Since the pathogenesis of PD is linked to mitochondrial dysfunction, MtCK is a potential PD biomarker [148]. A recent study found no differences in PD patient’s serum sMtCK activity. In contrast, a significant decrease in uMtCK activity was observed compared to the control group, associated with the disease progression rate [149].

Metabolomic Profiling

The metabolomic profile comprises a large-scale analysis of metabolites and is widely applied as a comprehensive diagnostic tool because it is a chemical reflection of a phenotype of a particular biological system. Therefore, it is implemented to understand the pathophysiological processes involved in the progression of the disease and search for new diagnostic or prognostic biomarkers of various diseases [150].

Several studies have evaluated the metabolomic profile in serum of PD patients. Among the metabolites that have found increased in PD patients compared to healthy subjects are glutathione, which is a non-enzymatic antioxidant [151]; pyruvate, which is involved in the energy metabolism [152]; amino acids, such as methionine, threonine, alanine, and serine [153]; bile acids like cholic acid, deoxycholic acid, and lithocholic acid [154, 155]; and pyroglutamate, which is a metabolite of glutathione [153]. The metabolites that have been found decreased in PD patients are uric acid, which is a non-enzymatic antioxidant [151, 156]; hypoxanthine, a metabolite of purine [156]; galactitol, glycerol, methylamine, trimethylamine, ethanolamine, suberate, glutarate, malate, methylmalonate, succinate, acetate, gluconate, threonate, glucolate, ascorbate, isocitrate, and citrate, which is involved in the energy metabolism [152]; caffeine metabolites [155]; and C16-C18 saturated and unsaturated fatty acids [153].

Also, fatty acid 14:1 was associated with disease severity, while its levels and those of phosphatidylcholine 34:2, and indolelactic acid were associated with the disease duration [155].

MicroRNAs

MicroRNAs (miRNAs) are endogenous non-coding RNA molecules, approximately 21 to 24 nucleotides in length, that play a role in the post-transcriptional regulation of gene expression during development [157]. The first miRNA (lin-4) was discovered in Caenorhabditis elegans [158]. Several miRNAs are essential neuroinflammation regulators. Inflammatory and neurodegenerative disorders have display variations in microRNA [159, 160]. Therefore, a study evaluated miRNAs’ expression profiles related to aging and cellular senescence in peripheral blood mononuclear cells from PD patients. The expression of miR-885 and miR-17 increased along with PD’s severity.

In contrast, the expression of miR-361 decreased in PD patients compared to controls. Furthermore, the lowest miR-361 levels were observed in the initial stages, while the highest levels were detected in the disease’s advanced stages. Combining detection of these three miRNAs provides a set of biomarkers that allows discrimination of PD patients from healthy subjects [161].

Urine Biomarkers

The urine is another non-invasive sample with diagnostic value for searching PD biomarkers. Importantly, urine is not subject to homeostatic mechanisms as blood and accumulates many changes that may reflect the body’s status [162].

As mentioned before, OS is involved in the loss of dopaminergic neurons. The increase in ROS causes oxidative modifications in lipids, proteins, and DNA. Heterocyclic bases in DNA are prone to oxidative damage, particularly guanine, which is more susceptible to form 8-hydroxydeoxyguanosine (8-OHdG) [163]. Then, 8-OHdG was evaluated in PD patients’ urine, whose concentration was higher and dependent on disease progression than control subjects [164]. Also, there was a positive correlation between urine 8-OHdG levels and the hallucinations presented in PD patients [165].

Another potential urine biomarker associated with OS is kynurenine, a tryptophan metabolite. In mammalian cells, tryptophan is degraded mainly by the kynurenine pathway, where the following metabolites are included: kynurenic acid (KYNA), quinolinic acid (QUIN), 3-hydroxyquinurenine (3-HK), and picolinic acid (PIC). KYNA and QUIN are considered neuroactive, while 3-HK and PIC have pro-oxidant and antioxidant properties. Therefore, the kynurenine pathway may play an essential role in PD pathogenesis [166]. Kynurenine concentration in PD patients’ urine was significantly higher than in control samples, and it was associated with the severity of the disease [167].

Also, the metabolic phenotype analysis in urine samples using high-performance liquid chromatography coupled with high-resolution mass spectrometry has yielded significant differences between PD patients and control subjects. Differences in related metabolic pathways were observed, including steroidogenesis, β-oxidation of fatty acids, metabolism of histidine, phenylalanine, tryptophan, tyrosine, and nucleotide, generating a panel of altered metabolites associated with PD [168]. On the other hand, proteomic analysis of urine exosomes from patients has identified synaptosomal-associated protein 23 (SNAP23), a component of the SNARE complex, and calbindin, a calcium-binding protein, as novel PD biomarkers [169]. Perturbations of some metabolites and protein levels in PD patients’ urine indicate that these can be used as a potential biomarker signature and may help in PD diagnosis and searching for a potential therapeutic target.

Concluding Remarks

The early PD diagnosis biomarker pursuing is an arduous task that still has a long way to go. Several biomarkers can be used to evaluate the prognosis and development of the disease. Some of them are related to the disease’s predominant pathophysiological characteristics, including mitochondrial damage, OS, and protein accumulation (Fig. 2), which have been detected in both the CNS and systemically in a wide variety of samples in PD patients, indicating that the study and diagnosis of PD should not only focus on the CNS but should be approached systemically. Many potential biomarkers in non-invasive samples have been reported and are summarized in this review.

Fig. 2
figure 2

Potential biomarkers and their relationship with predominant pathophysiological characteristics of Parkinson’s disease. Mitochondrial dysfunction, OS, and abnormal protein accumulation are involved in dopaminergic neuronal death in PD pathogenesis. Biomarkers related to these events are shown, including the sample type for their detection. Created with Biorender.com

Full size image

α-Syn seems to be the leading candidate biomarker for PD detection, as it can be found in diverse tissue and fluid samples, including skin, saliva, blood, and urine. However, there is still controversy regarding its sensitivity and specificity. Increasing evidence supports that microbiota influences α-Syn accumulation in the gut, from where it travels through the vagus nerve to the CNS; there, it starts aggregating and forming Lewy bodies until they cross the blood–brain barrier and disseminate systemically.

Significant progress has been achieved on the human metabolome in the last decade. Without a doubt, metabolomic profiles are a promising diagnostic tool of a diseased state as they reflect the metabolic state. Changes in biofluid metabolites seem to be a potential biomarker profile given the ease of obtaining the sample. However, these profiles have only been analyzed in patients who already present one or more motor symptoms. Still, there is a long way to find a biomarker profile for early idiopathic PD diagnosis.

PD patients suffer from early intestinal inflammation and dysfunction. In agreement with the gut-brain axis bidirectionality and its implication in PD pathogenesis, it is reasonable to hypothesize that the biomarkers with the best potential for early idiopathic PD diagnosis are those found in feces. Feces represent a non-invasive sample, where zonulin, α-1 antitrypsin, and calprotectin are stable and reflect increased intestinal permeability, damage, and inflammation. However, these biomolecules are not specific for PD by themselves, but they are connected to the gut modifications that PD patient suffers years before clinical symptoms are present.

The intestinal microbiota plays a crucial role in PD development. Despite intestinal microbiota’s complexity, some bacteria families have been found to increase or decrease in two or three PD patients’ population groups compared to control subjects (Table 1). Microbiota variations in the reported populations (German, Chinese, Italian, American, and Finnish) need to be validated in more diverse populations; other and additional differences could be present and be population-specific. To deploy a more precise medicine, potential confounders such as medications, diet, gastrointestinal symptoms, demographics, and others, can be controlled through a systematic approach [170]. Further studies are crucial to assure this biomarker panel sensitivity, specificity, repeatability, and reliability.

Can We Exploit These Biomarkers as a Diagnostic Approach?

Probably, a panel consisting of zonulin, α-1 antitrypsin, and calprotectin, which are related to early intestinal inflammation and dysfunction, along with Lactobacillaceae, Enterococcaceae, Christensenellaceae, Verrucomicrobiaceae, Enterobacteriaceae, Prevotellaceae, Lachnospiraceae, and Bacteroidetes quantification, all of which can be detected from feces, have an excellent potential for early idiopathic PD diagnosis. Protein biomarker modification can be detected with specific antibodies, while microbiota alterations can be simultaneously detected by real-time multiplex PCRs of the 16S rRNA gene with family-specific oligonucleotide sets. Further studies are crucial to assure this biomarker panel sensitivity, specificity, repeatability, and reliability. Additionally, particular caution must be exercised in PD patients since substantial variability was observed in PD subtype classification within 1 year of analysis in a Parkinson’s Progression Markers Initiative (PPMI) [171]. Hopefully, this biomarker panel may help identify subjects in the early stages of the illness and reduce the heterogeneity of disease patients in clinical trials or epidemiologic studies, leading to a better understanding of PD pathogenesis.