In the present study, we investigated CSF PENK- and PDYN- derived peptides as candidate biomarkers in PD and HD.

First, we added strength to our previous results [8] by showing reduced CSF levels of PDYN-derived peptides in an independent and larger cohort of HD patients in comparison to controls and other disease groups. Moreover, we found a similar profile for CSF PENK-derived peptides, in accordance with other studies [9, 10], thus supporting the notion that both CSF PENK and PDYN might reflect the striatal neurodegeneration along with the loss of MSNs occurring in HD [9, 10]. Consistently, a decrease in PDYN mRNA expression as well as in PDYN- and PENK-derived peptides has been reported in HD brains suggesting a combined effect of transcriptional dysregulation and loss of MSNs expressing PENK and PDYN genes [8, 9, 28,29,30]. Accordingly, CSF PENK and PDYN were within the normal range in ALS and AD, two neurodegenerative diseases with lack or low degree of striatal dysfunction/atrophy [11,12,13].

Interestingly, we found correlations between biomarker levels and both motor and cognitive parameters in HD patients, thus suggesting a potential role of both analytes as markers of disease severity in HD patients. Similarly, Niemela et al. [10] reported a decrease in CSF PENK levels along with increased disease severity, together with reduced biomarker concentrations in symptomatic compared to pre-symptomatic patients, and a trend toward lower levels in the latter group compared to controls. Given the urgent need for surrogate endpoints in ongoing clinical trials for HD, PDYN- and PENK-derived peptides may be used together with neurofilament light chain protein (NfL) as biomarkers of disease severity that could be potentially influenced by future disease-modifying therapies [31, 32]. In this regard, larger studies including longitudinal samples of HD symptomatic and pre-symptomatic subjects are needed to fully elucidate the dynamics of CSF PENK and PDYN levels during disease course and to evaluate their potential predictive role in the pre-symptomatic phase. Here, results in a small cohort documented an inverse correlation between CSF PENK levels and the 5-year risk of onset among pre-symptomatic HD cases [10]. We also acknowledge that the lack of association between biomarker levels and disease duration in HD patients might possibly depend on the relative homogeneity of our cohort (i.e. most HD patients in TFC stage 1).

On another issue, we did not find any biomarker changes in a large group of PD and PD-MCI subjects, compared to other diagnostic groups. Given that dopaminergic signaling modulates opioids synthesis by inducing PDYN and inhibiting PENK production, respectively [33], alterations in CSF biomarker levels were expected to be found in PD patients. However, animal model data showed that a subtotal striatal dopamine depletion should be a pre-requisite to produce a significant alteration in brain PDYN and PENK levels [34]. Thus, a possible explanation of our findings may rely on the inclusion of a relatively high proportion of PD patients in the early-middle disease stage (median disease duration 3 years).

Interestingly, the use of dopaminergic therapy influences considerably the production of endogenous opioids in PD animal models with an upregulation of PDYN and a downregulation of PENK, respectively [33, 35]. Accordingly, by comparing treated and untreated PD patients, we found higher CSF PDYN levels in the former compared to the latter group. However, this difference was not maintained in the PD-MCI group probably due to the advanced disease stage and the lower response to dopaminergic therapy. Similarly, CSF PENK levels were not altered in both PD and PD-MCI subjects after stratification according to the treatment state, suggesting possibly a less powerful effect of the therapy on CSF PENK levels.

The major strength of our study relies on the analysis of two new potential biomarkers in the largest cohort to date of HD patients. Regarding potential limitations, we would mention the cross-sectional nature of the study, which did not help in tracking the longitudinal evolution of biomarker values according to disease stage. Moreover, further clinical and therapeutical data (e.g. treatment duration and side effects, motor fluctuations) were not investigated in our cohort and deserve to be explored in future studies. The finding of positive associations between PENK peptide levels and age in controls, PD and sALS patients is challenging and deserves further explorations in bigger cohorts. Nevertheless, in all the above-mentioned groups both PENK-derived peptides were within the normal range and in PD, there was no correlation between biomarkers levels and disease severity scales, suggesting that age-related associations in PD and sALS may be driven by other pathophysiological phenomena compared to those of HD (i.e. striatal neurodegeneration). Furthermore, despite the very promising results, the main limit for the implementation of CSF PENK and PDYN analyses in the clinical diagnostic setting is the lower distribution of LC−MS/MS compared to classic ELISA techniques.

In conclusion, we provided further evidence on the performance of CSF PENK- and PDYN-derived peptides as promising candidate biomarkers reflecting ongoing striatal neurodegeneration and disease severity in HD. In PD patients, CSF PDYN showed a limitative role as a possible pharmacodynamic marker during dopaminergic therapy.