With several fluid biomarkers for Alzheimer’s in hand, the search is on for counterparts in Parkinson’s disease and dementia with Lewy bodies (DLB). Four new papers introduce two candidates: DOPA decarboxylase (DDC), the enzyme that converts levodopa into dopamine, and mitochondrial DNA damage.

  • People with Lewy body disease (LBD) make ample DOPA decarboxylase (DDC).
  • High DDC in CSF or blood distinguishes LBD from controls with up to 91 percent accuracy.
  • DDC pegs preclinical LBD, foretells imminent symptom onset.
  • Another marker, mitochondrial DNA damage, detected Parkinson’s with 85 percent accuracy.

In the September 13 Nature Communications, researchers led by Charlotte Teunissen at Amsterdam University Medical Centers, The Netherlands, reported that high cerebrospinal fluid levels of DDC distinguished people with DLB from controls. Per Svenningsson, Karolinska Institutet, Stockholm, Sweden, found the same in PD, as reported in the September 4 Translational Neurodegeneration. So, too, did Oskar Hansson, Lund University, Sweden, in both DLB and PD. In the September 18 Nature Aging, his group also reported that DDC was up in blood from people with those diseases and that CSF DDC predicted progression within three years in people with preclinical DLB or PD.

Scientists led by Laurie Sanders, Duke University School of Medicine, Durham, North Carolina, measured mitochondrial DNA damage in blood cells. In the August 30 Science Translational Medicine, they reported that blood cells of people with PD and non-symptomatic carriers of a PD-causing mutation had more mitochondrial DNA damage than controls.

“A blood-based biomarker is critically needed for synucleinopathies as current modalities involve either procedures, e.g., lumbar puncture for CSF or skin biopsy, or expensive brain imaging,” wrote Lawren VandeVrede, University of California, San Francisco (comment below).

Current diagnostic options for PD include DaTscan SPECT to assess dopamine transporter activity, the Syn-One test to detect α-synuclein aggregates in a skin biopsy, or the SYNTap α-synuclein seed amplification assay, which detects aggregate-prone α-synuclein in CSF (Dec 2015 conference news; Scott et al., 2021). Hansson used a similar seed amplification assay to detect misfolded α-synuclein lurking in CSF from controls, foretelling PD or DLB in some (Aug 2023 conference news).

To find other LBD biomarkers, first author Marta del Campo and colleagues measured proteins in CSF from 109 people with DLB, 235 with Alzheimer’s disease, and 190 controls from three cohorts: the Amsterdam Dementia Cohort, Dementia with Lewy bodies Project (DEvELOP), and the University of Pennsylvania Center for Neurodegenerative Disease Research. Most participants were in their 50s and 60s; 65 percent were men.

The most upregulated protein in DLB was DDC, distinguishing patients from controls with an area under the curve of 0.91 (see image below). AUC is a measure of sensitivity and specificity, with 1 being perfect. CSF DDC also separated DLB from Alzheimer’s with an AUC of 0.81, suggesting it is a disease-specific marker.

Distinctive DDC. DOPA decarboxylase (dot in upper right corner) was the most dysregulated protein in CSF from people with DLB compared to controls (left panel). DDC levels distinguished people with DLB from controls and people with AD with 91 and 81 percent accuracy, respectively (right). [Courtesy of del Campo et al., Nature Communications, 2023.]

DDC also topped Svenningsson’s CSF proteomics list in Parkinson's. Co-first authors Wojciech Paslawski and Shervin Khosousi measured proteins in CSF from 117 controls, 132 people with PD, and 67 with atypical Parkinsonian disorders, including 24 with multiple system atrophy, 21 with progressive supranuclear palsy, and 22 with corticobasal syndrome. The samples came from the Swedish BioPark cohort, the American BioFIND cohort, and the University of California, San Francisco.

Among the top six upregulated CSF proteins in PD? You guessed it: DDC. It had the most diagnostic potential, distinguishing people with PD from controls with an AUC of 0.80. DDC was also high in CSF from people with atypical Parkinsonian disorders, hinting that it might be a broader biomarker of neurodegenerative diseases with dopamine deficiencies.

Hansson’s findings agree with this idea. In their larger study, co-first authors Joana Pereira and Atul Kumar analyzed proteins in CSF from 682 Swedish BioFINDER participants: 347 controls, 33 people with DLB, 48 with PD, 40 with atypical Parkinsonian disorders, 172 people with AD, 23 with frontotemporal dementia, and 19 with vascular dementia.

Again, DDC was the most upregulated CSF protein in DLB and other Parkinsonian disorders (see image below). It identified DLB and PD, lumped together as Lewy body disease (LBD), from controls with an AUC of 0.89, and atypical Parkinson’s disorders from controls with 79 percent accuracy. CSF DDC also distinguished LBD from the three non-Parkinsonian neurodegenerative diseases with an AUC of 0.83. The authors suspect neurons might crank out more DDC to compensate for low dopamine levels, hence the marker’s upregulation only in Parkinsonian disorders.

Ditto DDC. In a different cohort, DDC was again the most dysregulated CSF protein in people with LBD versus controls (left image, upper right corner). It pegged DLB with 89 percent accuracy (right). [Courtesy of Pereira et al., Nature Aging, 2023.]

Hansson was initially skeptical that DDC was so high in LBD. He thought it might be related to treatment. People with PD typically take levodopa plus a DDC inhibitor that stays in the periphery, allowing most of the dopamine precursor drug to enter the brain. Could this influx of levodopa drive up DDC production? While the protein was indeed higher in the LBD participants taking medication than in those who weren't yet, DDC was still the most upregulated CSF protein in untreated patients, meaning the disease drove the biomarker change. Likewise, del Campo and Svenningsson saw high CSF DDC in treatment-naïve PD participants and even higher levels in those taking levodopa and a DDC inhibitor.

Importantly, DDC was able to detect preclinical LBD and predict progression when paired with an α-synuclein seed amplification assay. Among the 347 controls, 35 tested positive for α-synuclein aggregates, which Hansson and colleagues took to mean preclinical LBD. DDC was already high in their CSF, distinguishing them from controls with an AUC of 0.81. Over three years, 12 of them developed symptoms, and those with the highest CSF DDC levels were 3.7 times more likely to progress than their counterparts with low DDC levels. “If a person has both Lewy body pathology and a disturbance in the brain's dopaminergic system, they are close to developing symptoms,” Hansson concluded.

High levels of CSF DDC also correlated with worse cognition and memory in LBD, as measured by the mPACC and Alzheimer’s Disease Assessment Scale–Cognitive Subscale.

Hansson envisions combining the CSF DDC and α-synuclein seed assays in the clinic. A person testing positive for both would likely have LBD, whereas someone with only high DDC would have an atypical Parkinsonian disorder, as Lewy bodies do not develop during those diseases.

A blood biomarker would be better for clinical use. How about plasma DDC? This research is only starting, but so far, among 33 people with LBD, 56 with atypical Parkinson’s disorders, and 54 controls, having lots of plasma DDC identified LBD with 92 percent accuracy and atypical PD with 85 percent. “We need blood-based biomarkers to not rely on CSF for large population-based screens in the future,” wrote Brit Mollenhauer, University Medical Center Göttingen, Germany (comment below).

Are there other factors in the blood that could flag PD? Sanders and colleagues took a different route, analyzing damage to mitochondrial DNA (mtDNA), rather than proteins, in blood cells. Mitochondrial damage looms large in PD, with defective disposal of these old or broken organelles and subsequent buildup of their spoiled bits driving the disease (Sep 2015 news; reviewed by Malpartida et al., 2021).

To measure mitochondrial DNA damage, co-first authors Rui Qi of Sanders’ lab and Esther Sammler at the U.K.’s University of Dundee developed a PCR-based assay that quantifies how much mtDNA is in cells. The assumption is that the more intact mtDNA the assay detects, the less has been damaged or destroyed. Qi and Sammler analyzed peripheral blood mononuclear cells in blood samples from 22 controls, 58 people with PD—including 28 who carried the PD-causing G2019S LRRK2 mutation—and 17 non-symptomatic carriers from the Michael J. Fox Foundation's Bionet cohort. Sanders had previously tied this LRRK2 variant to mtDNA damage in cultured human neurons (Sanders et al., 2013).

Compared to blood cells from controls, cells from PD patients had 50 percent more mtDNA damage. So, too, did asymptomatic LRRK2 carriers, placing mitochondrial damage early in pathogenesis (see image below). “It remains to be investigated whether these non-manifesting LRRK2 mutation carriers develop clinical PD,” the authors wrote.

Mitochondrial genome damage load distinguished PD, with or without the LRRK2 mutation, from controls, posting AUCs of 0.85 or 0.84, respectively. The marker was less accurate in asymptomatic carriers, identifying them from noncarriers with an AUC of 0.74 (see image below).

Marred MitochondriaCompared to blood cells from controls (upper left panel, black dots), cells from idiopathic PD (blue), PD with LRRK2 mutation (red), or asymptomatic mutation carriers (purple) had about 50 percent more mitochondrial DNA damage. The level of damage identified PD or LRRK2 carriers vis-à-vis controls with 74 to 85 percent accuracy. [Courtesy of Qi et al., Science Translational Medicine, 2023.]

Mitochondrial damage was specific to PD, as 10 people with AD had no more of it than seven controls from a Duke Memory Disorders Clinic cohort. “I wondered how specific this mitochondrial lesion assay is for PD, so I appreciated the inclusion of AD patients,” VandeVrede noted.

All told, DDC concentration in CSF or plasma and mtDNA damage in blood cells offer two potential markers of PD and other dopaminergic neurodegenerative disorders. “I think this is an exciting time, not only with the robust seed amplification assays showing high sensitivities and specificities in CSF of neuronal synuclein disorders but also with these emerging blood biomarkers,” said Mollenhauer. Douglas Galasko, University of California, San Diego, agreed. “The ability to detect increased DDC in plasma raises the possibility of early screening and detection, and potentially the differential diagnosis of PD from atypical Parkinsonism if CSF is also obtained for seed amplification,” he wrote (comment below).—Chelsea Weidman Burke

Comments

  1. This is an interesting set of papers. Qi et al. extends these authors’ prior work shedding light on the pathobiology of PD, particularly carriers of pathogenic LRKK2 variants. A blood-based biomarker is critically needed for synucleinopathies, as current modalities involve either procedures, e.g. LP for CSF (SynTap), skin biopsy (SynOne), or expensive imaging (DaT scan). The basic science in this paper appears unimpeachable, and the validation efforts are intriguing.

    I would like to see the results independently replicated in larger cohorts that also include other synucleinopathies and atypical presentations such as DLB, MSA, PSP, and CBS. I particularly appreciated the inclusion of AD patients because I wondered how specific this mitochondrial lesion assay is for PD. Regardless, it looks like it provides justification for a trial of mitochondrial agents in prodromal LRRK2 mutation carriers, and the assay itself might be useful to show on-target pathway effects in trials of LRRK2 inhibitors.

    Pereira et al. offer a nice validation effort of a biomarker of dopaminergic dysfunction, essentially a fluid biomarker version of a DaT scan (though the plasma results were not as encouraging as CSF). It would be interesting to see how it compares to DaT. Pereira et al. make innovative use of CSF SAA assay to define prodromal LBD. 

    Campo et al. represents a good proteomic discovery and biomarker validation study. DLB has poor diagnostic accuracy, but this is a known limitation. I appreciate the autopsy validation and also the subgroup analyses with LBD-specific clinical features (e.g., REM).

    DLB versus CON is a great comparison, as is inclusion of three PD cohorts. Alas, I’m not sure of the need for a DLB versus AD comparison of diagnostic performance—it makes sense in the discovery to increase specificity of findings but great CSF biomarkers are already available to detect AD so the clinical use is less clear. I would be interested to see the ability to detect LBD co-pathology in AD.

    Though the authors focus on a biomarker panel, based on my read of Figure 2c, it appears DDC is the primary driver of difference between DLB and controls. In Fig 5, the AUC is nominally higher for the panel compared to DDC in a few cohorts, but is it significant? Just like in the other study, I’d want to see this replicated, especially compared to DaT including in atypical PD (PSP/CBS). Somewhat less exciting given it’s another CSF assay for LBD and doesn’t have a head-to-head with SAA, but overall very impressive.

  2. This is an exciting time, not only with the robust seed amplification assays showing high sensitivities and specificities in cerebrospinal fluid of neuronal synuclein disorders, but also with emerging blood biomarkers.

    DDC indeed has been identified before as a marker, e.g. in the PPMI cohort study. It was thought to be due to a medication effect, but Pereira et al. now show increased values in preclinical LBD, as well, which is interesting. It could well be a compensatory effect before the motor/cognitive disease develops. This is encouraging, as we urgently need biomarkers for identifying candidates for clinical trials in these very early stages of the disease. And we need blood-based biomarkers so we no longer have to rely on cerebrospinal fluid for large population-based screens in the future.

    Considering that PD may start in the periphery, there must be a biomarker pattern in peripheral blood that is detectable with newer methods, including mass spectrometry or more sensitive immunoassays.

    The mtDNA paper is also a path forward toward a blood panel to identify subjects at risk if developing motor/cognitive disease. It could be used in upcoming clinical trials, with the ultimate goal of preventing the disease.

  3. Protein biomarkers for Parkinson’s disease and DLB have been sought extensively, with far fewer hits than in AD. Recent efforts using the O-link multiplex immune-PCR assay system for discovery have yielded interesting results. In a large, multicenter CSF study, the most significantly increased protein in DLB and PD was aromatic amino acid decarboxylase, aka dopa decarboxylase or DDC, which distinguished DLB from AD (Del Campo et al., 2023). Another study noted a similar increase in DDC and a correlation with treatment of PD, which often includes a DDC inhibitor (Paslawski et al., 2023).

    The exciting study by Pereira et al. expands on these efforts by using the Swedish BioFINDER cohort, in which CSF and plasma are available and CSF samples were analyzed for α-synuclein seeding via RT-QuIC. They found that DDC was increased in SAA+ cases, including DLB, PD, and cases of preclinical PD, i.e., the changes occurred early in the course of disease. Changes were present whether patients were taking L-dopa/carbidopa or not. CSF levels also were increased in patients with atypical Parkinsonism, including MSA, PSP and CBS (unlike Paslawski et al.), but not in AD. In plasma, levels of DDC were also increased in SAA+ DLB in a separate cohort. The power of having concomitant SAA+ data allowed Pereira et al. to identify CSF DDC as a strong marker of PD and DLB, but the findings of an increase in atypical PSP suggests that DDC is a "damage" marker related to dopaminergic and possibly other aminergic nerves.

    The ability to detect increase DDC in plasma raises the possibility of early screening and detection, and potentially of differential diagnosis of PD versus atypical Parkinsonism if CSF is also obtained for SAA.

    Replicating these findings and developing a different assay for DDC will be important steps in translation. Staging systems for PD are being proposed, e.g. Neuronal Synuclein Disorder, which includes DaTscan as a marker of dopaminergic neurodegeneration. It will be interesting to see how CSF or plasma DDC might fit into those proposed schemes. 

    A different approach to PD biomarkers was taken by Qi et al., who adapted PCR to amplify mitochondrial DNA in peripheral leukocyte pellets. The elegant concept is that if mitoDNA damage is present, amplification will be less efficient. This strategy gets around many of the difficulties of analyzing mitochondrial function, and allowed frozen cell pellets to be studied. Qi et al. found that mitochondrial damage was present in cells from patients with idiopathic PD and LRKK2 mutation carriers, with very good sensitivity and specificity in discovery and replication cohorts. The test was semiquantitative, and further iterations to develop a quantitative test raise the possibility of identifying patients with greater or lesser degrees of mitochondrial damage. This could allow patient selection for personalized approaches to treatment.

    Combining mitochondrial biomarker analyses with biomarkers such as SAA and DDC may help to further characterize heterogeneity of disease mechanisms in PD.

    References:

    . CSF proteome profiling reveals biomarkers to discriminate dementia with Lewy bodies from Alzheimer´s disease. Nat Commun. 2023 Sep 13;14(1):5635. PubMed.

    . Large-scale proximity extension assay reveals CSF midkine and DOPA decarboxylase as supportive diagnostic biomarkers for Parkinson's disease. Transl Neurodegener. 2023 Sep 4;12(1):42. PubMed.

  4. Two recent publications have identified novel biomarkers for PD based on pathways that have long been known to be associated with its pathogenesis (Qi et al., 2023). Interestingly, these biomarkers also reflect changes in blood that are concurrently reported in the central nervous system (CNS).

    Pereira et al. utilize an unbiased proteomics approach using Olink technology to identify elevation of dopa decarboxylase in individuals with PD and atypical parkinsonism in the Swedish BIOFINDER study (Cui et al., 2022; The Swedish BioFINDER study. Accessed September 19, 2023). DDC is highly relevant because it is the enzyme that converts levodopa, the diagnosis-defining PD medication, to dopamine, the neurotransmitter whose levels are reduced in the CNS due to neuronal cell death in the midbrain of individuals with PD. With high reliability, CSF levels of DDC differentiate healthy controls from individuals with PD or atypical parkinsonism.

    DDC CSF levels may also be a potential prodromal biomarker, as it reliably differentiated individuals without PD who were positive for a novel α-synuclein biomarker, the seed aggregation assay (SAA). Moreover, plasma elevation in DDC also demonstrated similar ability to discriminate individuals with parkinsonism from healthy controls, which again reflects the systemic nature of PD.

    The directionality of this effect, whether CNS to periphery or vice versa, requires further investigation. The result that DDC levels do not differentiate PD from atypical parkinsonism suggest that this biomarker may reflect dopaminergic dysfunction rather than PD-specific pathology.

    Limitations of this work include lack of validation with protein detection methods besides Olink, lack of use of a non-parkinsonism neurologic disease group, and use of a relatively genetically and ethnically homogenous population. Since this biomarker does not differentiate PD from other forms of parkinsonism, DDC could be developed for use in a similar manner to how DAT scans are used or for use in individuals who have barriers to imaging.

    Qi et al. developed a novel method to measure mitochondrial DNA damage in a high-throughput method, termed mito DNADX (Qi et al., 2023). Mitochondrial dysfunction is a well-known phenomenon in PD; it has been demonstrated in selective brain regions as well as peripheral blood cells (Legati et al., 2023). Outside of assessing its role as a biomarker, the development and validation of a quantitative, sensitive, and reproducible method to assess mitochondrial DNA quality is impressive in itself.

    This study validates that mitochondrial DNA damage in peripheral blood mononuclear cells can be used as a biomarker to reliably differentiate individuals with PD from healthy controls (AUC = 0.8-1.0) in multiple cohorts. Notably, this difference was seen in individuals with PD regardless of whether they carry the known G2019S risk mutation in LRRK2 and was also seen in LRRK2 mutation carriers without PD. Despite use of a LRRK2 inhibitor improving mitochondrial DNA quality in immortalized lymphocytes from individuals with PD, there were no identified differences in phosphorylation of known LRRK2 substrates.

    Even if identified mitochondrial DNA changes are not specific to LRRK2, this manuscript certainly provides a new peripheral biomarker for in vivo mitochondrial damage in PD (and likely other disorders). Whether this biomarker can be used in prodromal populations, and how it relates to SAA, was not definitively examined in this study. Furthermore, while mitochondrial DNA changes were not seen in AD, the ability to differentiate PD from atypical parkinsonism was not tested. The sensitivity of measured mitochondrial DNA quality to sample processing and storage, as well as the relatively highly technical nature of the mito DNADX assay, are significant roadblocks for development into a readily accessible clinical assay.

    Both these biomarkers will require further validation in other cohorts on a larger scale to determine feasibility of use in a clinical setting. Regardless, these findings certainly raise mechanistic questions in PD pathogenesis, particularly with respect to the interplay of central and peripheral mechanisms in dopamine metabolism and mitochondrial function. These studies increase our hope that reproducible blood-based biomarkers for PD and other neurodegenerative disorders are on the horizon.

    References:

    . A blood-based marker of mitochondrial DNA damage in Parkinson's disease. Sci Transl Med. 2023 Aug 30;15(711):eabo1557. PubMed.

    . High-throughput proteomics: a methodological mini-review. Lab Invest. 2022 Nov;102(11):1170-1181. Epub 2022 Aug 3 PubMed.

    . Parkinson's Disease, Parkinsonisms, and Mitochondria: the Role of Nuclear and Mitochondrial DNA. Curr Neurol Neurosci Rep. 2023 Apr;23(4):131-147. Epub 2023 Mar 7 PubMed.

  5. It is amazing to see three papers with complementary information indicating that CSF Dopa decarboxylase (DDC) is a novel biomarker that might be useful to identify disorders characterized by dopamine deficiency in their very early stages. More studies are probably needed to validate plasma findings and its potential implementation, taking medication status into consideration, since treatment can increase the activity of DDC in the serum (van Rumund et al., 2021). We expect that the new immunoassay for DDC that has been developed and validated by the Teunissen lab will help to answer remaining questions and advance the implementation of this biofluid marker.

    References:

    . Peripheral decarboxylase inhibitors paradoxically induce aromatic L-amino acid decarboxylase. NPJ Parkinsons Dis. 2021 Mar 19;7(1):29. PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Brain Imaging: What Does it See in DLB?
  2. Finally, a Diagnostic Marker for Lewy Body Disease?
  3. PINK1 Can Act Alone to Destroy Mitochondria, But Parkin Helps

Paper Citations

  1. . Fluid and Tissue Biomarkers of Lewy Body Dementia: Report of an LBDA Symposium. Front Neurol. 2021;12:805135. Epub 2022 Jan 31 PubMed.
  2. . Mitochondrial Dysfunction and Mitophagy in Parkinson's Disease: From Mechanism to Therapy. Trends Biochem Sci. 2021 Apr;46(4):329-343. Epub 2020 Dec 13 PubMed.
  3. . LRRK2 mutations cause mitochondrial DNA damage in iPSC-derived neural cells from Parkinson's disease patients: Reversal by gene correction. Neurobiol Dis. 2013 Oct 19;62C:381-386. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . DOPA decarboxylase is an emerging biomarker for Parkinsonian disorders including preclinical Lewy body disease. Nat Aging. 2023 Oct;3(10):1201-1209. Epub 2023 Sep 18 PubMed.
  2. . Large-scale proximity extension assay reveals CSF midkine and DOPA decarboxylase as supportive diagnostic biomarkers for Parkinson's disease. Transl Neurodegener. 2023 Sep 4;12(1):42. PubMed.
  3. . CSF proteome profiling reveals biomarkers to discriminate dementia with Lewy bodies from Alzheimer´s disease. Nat Commun. 2023 Sep 13;14(1):5635. PubMed.
  4. . A blood-based marker of mitochondrial DNA damage in Parkinson's disease. Sci Transl Med. 2023 Aug 30;15(711):eabo1557. PubMed.