Cerebrospinal fluid ceramides and cognition as a function of striatal asymmetry in early Parkinson's disease

doi:10.21203/rs.3.rs-2404396/v1

A growing interest in studying the impact of brain asymmetry on physiological markers and cognition in people with Parkinson's disease (PwPD) recently emerged. The aim of this study was to test the impact of striatal denervation asymmetry on brain markers such as ceramides accumulation (ng/ml), as well as its impact on cognitive performances in early PwPD. We analysed data from 329 PwPD patients at the beginning of the disease (mean 6.9 months after diagnosis) (72 left-asymmetric, 83 right-asymmetric and 174 symmetric PwPD) and 167 healthy controls. Asymmetry was based on the difference in putamen denervation greater than 20%. Patients with genetic mutations were excluded. We performed generalized linear mixed models introducing the amount of cerebrospinal fluid ceramides (ng/ml) and cognitive functions as discriminating factors. Finally, Spearman correlations were used to highlight the relationship between cognition and cerebrospinal fluid ceramides on the whole pathological group and within each sub-group, as a function of striatal denervation asymmetry. First, a reduced concentration of ceramide (C18:0 CER) in the symmetrical group compared to the asymmetrical group and the control group was observed. More specifically left denervation predominant PwPD had significantly more ceramides (C22:0 CER) than the right and controls. Second, poorer cognitive performance for the symmetric group compared to the asymmetric and control groups was reported. Overall, positive correlations between cognition and ceramides (C18:0, C22:0 CER) was observed. Conversely, glucosylceramide correlation (C22:0 GlcCer) showed a negative link with cognition. This study shows that asymmetry of striatal degeneration in early Parkinson's disease is accompanied by metabolic variations related to cognitive processes.

Health sciences/Neurology/Neurological disorders/Parkinson's disease

Health sciences/Biomarkers/Diagnostic markers

Brain asymmetry

ceramides

glucosylceramides

cognition

Parkinson’s disease

A recent and interesting way of characterising aetiologies leading to different clinical phenotypes in people with Parkinson’s disease (PwPD) is the so-called alpha-Synuclein Origin site and Connectome (SOC) model ¹. Accordingly, two PwPD phenotypes can be distinguished based on the development of the disease, either by a “body first” or a “brain first” start. This distinction in disease onset would have an impact on the development of motor and non-motor symptoms, and in fine, on the potential appearance of an associated dementia ². Thus, body first patients would start the disease with symmetrical damage to the nigrostriatal regions, demonstrating rather vegetative symptoms. This would be accompanied by symmetrical motor symptoms and would cause the disease to progress more rapidly towards neurocognitive dementia. In contrast, brain first patients would begin with asymmetric brain damage and would be accompanied by asymmetric motor symptoms with few non-motor symptoms. Therefore, there would be a slower progression of the disease to dementia in this second phenotype. While there are currently no physiological markers able to capture PwPD clinical phenotypes ³, the potential involvement of glucosylceramides metabolism seems to be an emerging lead to follow ^3–7. As a reminder, glucosylceramides are more broadly glucocerebrosides involved in the ceramide and sphingomyelin metabolism pathways active at the cerebral region in the constitution of neuronal tissue ⁸. Thus, the relationships between these metabolic pathways and genetic factors show an involvement of ceramide accumulation ⁹ in mitochondrial cellular stress taking place in Parkinson’s disease (PD), without distinction of striatal denervation asymmetry ^3,10. Interestingly, when comparing PwPD to a group of healthy individuals, results revealed not only an increase in some ceramides (C18:0 CER), but also a decrease in the concentration of other CER (C24:0 CER) ⁴. Another recent study showed links between glucosylceramides in cerebrospinal fluid and global cognitive performance in PD ¹¹. In this study, the ratio of glucosylceramides to sphingomyelin (GLCCER/ SM) seems to be related to an accelerated cognitive decline in the longitudinal follow-up of patients. Another study postulated a strong implication of ceramides in the neurodegenerative phenomena of PwPD by reducing the concentration of certain ceramides in brain regions (such as the grey matter of the anterior cingulate cortex) affected in PD ⁴. Finally, genetic mutations in glucocerebrosidase were linked to a significant neurodegenerative asymmetry phenomenon and a more rapid progression of motor problems ⁵. It would therefore seem that variations in the concentration of glucosylceramides in the cerebrospinal fluid are associated with the possibility to differentiate idiopathic PwPD phenotypes by taking genetic factors into account ¹¹. In view of recent studies showing differences in motor (MDS-UPDRS III)¹² and cognitive (MoCA)¹³ symptoms according to striatal degeneration asymmetry, as well as studies highlighting relationships between ceramides and neurodegenerative processes, a better understanding of the relationship between metabolism and cognition as a function of striatal dopaminergic denervation asymmetry seems of crucial interest^2,11,14.

In this context, the aim of the present study was to explore the impact of asymmetric (vs. symmetric) striatal denervation by comparing cerebrospinal fluid ceramide accumulation rates and cognition, as well as the relationship between ceramide accumulation and global cognitive efficiency in each phenotype of patients. From the literature presented above ^1,11, we expected that PwPD in the early stages of the disease, without genetic mutations, and with asymmetric degeneration in the putamen (striatal specific binding ratios [SBR] + 20%), would have lower levels of ceramide accumulation in cerebrospinal fluid (ng/ml) and less cognitive disorders than patients with a symmetric degeneration. Secondly, based on the results obtained in Huh et al,¹¹, in regard to the relationship between sphingolipid ratio and cognition, we expected positive correlations between ceramide/glucosylceramide accumulation and cognitive performance in the whole PwPD group and in the PwPD subgroups.

2.1 Intergroup differences in cerebrospinal fluid ceramide concentration (Table 1)

Asymmetric vs. symmetric PwPD. Asymmetric PwPD had a higher concentration of ceramides C18:0 CER than symmetric PwPD (t = -2.26; p = .024), false discovery rate (FDR) corrected (see Table 1).

Left vs. right PwPD. We found a significant difference for C22:0 CER (t = -2.60; p = .010) with a higher concentration for left asymmetry than for right PwPD (see Fig. 1–2), FDR corrected.

Left vs. healthy controls (HC). Intergroup differences in C22:0 CER (t = -2.65; p = .008) showed a higher rate for the left asymmetric group (see Fig. 2), FDR corrected.

Right vs. HC and asymmetric vs. HC. There was no intergroup difference for ceramides.

Symmetric vs. HC. We observed more C18:0 CER in the symmetric group (t = 2.42; p = .016), FDR corrected (see Fig. 1).

All other comparisons were not significant after FDR correction.

2.2 Intergroup differences in cognitive performance (Table 1)

Asymmetric vs. symmetric PwPD. Asymmetrical PwPD had better cognitive performances on the total score of symbol digit modalities test (SDMT) (t = -3.62; p < .001) and the Montreal cognitive assessment (MoCA) scores (t = -2.60; p = .010), FDR corrected.

Left vs. symmetric PwPD. Left PwPD had better performances in the SDMT total score (t = -3.45; p = .001) (see Fig. 1–2), after FDR correction.

Right vs. symmetric PwPD. The groups differed on the global cognitive scale (MoCA), showing better performances for right asymmetric PwPD than for symmetric PwPD (t = -2.62; p = .009) (see Fig. 2).

Left vs. HC. We observed differences for MoCA scores, with better performance for the HC group (t = 3.50; p < .001), after FDR correction.

Right vs. HC. We observed here better performances for the HC group in the SDMT (t = 2.73; p = .006) and the MoCA (t = 2.64; p = .009), after FDR correction.

Asymmetric vs. HC. We observed better performances for the HC group in the SDMT (t = 2.50; p = .013) and the MoCA (t = 3.74; p < .001), after FDR correction.

Symmetric vs. HC. Finally, we observed differences in cognitive performance for the SDMT (t = 6.23; p < .001), MoCA (t = 6.48; p < .001) and HVLT immediate recall (t = 4.03, p < .001) after FDR correction, marked by improved performance of the control group.

All other comparisons were not significant after FDR correction.

Table 1

Ceramides levels and cognitive performances according to striatal denervation asymmetry groups
	HC ^ç n = 123 ^# n = 167 (Mean ± SD)	LAPD ^ç n = 57 ^# n = 72 (Mean ± SD)	RAPD ^ç n = 72 ^# n = 83 (Mean ± SD)	SPD ^ç n = 148 ^# n = 174 (Mean ± SD)	ASPD ^ç n = 129 ^# n = 155 (Mean ± SD)
C16:0 CER ^ç	0.35 ± 0.13	0.38 ± 0.17	0.35 ± 0.14	0.34 ± 0.16	0.36 ± 0.15
C16:0 GlcCer ^ç	0.24 ± 0.11	0.29 ± 0.30	0.23 ± 0.10	0.24 ± 0.14	0.26 ± 0.22
C18:0 CER ^ç	2.21 ± 0.72	2.27 ± 0.74	2.17 ± 0.76	2.03 ± 0.77 ^*	2.21 ± 0.75 ^!
C18:0 GlcCer ^ç	0.61 ± 0.28	0.66 ± 0.33	0.59 ± 0.23	0.58 ± 0.23	0.62 ± 0.28
C20:0 CER ^ç	0.29 ± 0.10	0.32 ± 0.14	0.29 ± 0.12	0.28 ± 0.12	0.31 ± 0.13
C20:0 GlcCer ^ç	0.04 ± 0.10	0.05 ± 0.13	0.04 ± 0.10	0.04 ± 0.11	0.05 ± 0.11
C22:0 CER ^ç	0.36 ± 0.14 ^£	0.42 ± 0.28	0.37 ± 0.14 ^£	0.36 ± 0.16	0.39 ± 0.21
C22:0 GlcCer ^ç	0.48 ± 0.39	0.70 ± 0.70	0.49 ± 0.42	0.54 ± 0.43	0.58 ± 0.56
C23:0 CER ^ç	0.05 ± 0.12	0.13 ± 0.25	0.07 ± 0.13	0.09 ± 0.15	0.10 ± 0.20
C23:0 GlcCer ^ç	0.36 ± 0.28	0.48 ± 0.37	0.38 ± 0.30	0.41 ± 0.32	0.42 ± 0.33
C24:0 CER ^ç	0.54 ± 0.24	0.65 ± 0.61	0.54 ± 0.28	0.60 ± 0.42	0.59 ± 0.46
C24:0 GlcCer ^ç	1.81 ± 1.08	2.25 ± 1.62	1.87 ± 1.13	2.00 ± 1.35	2.04 ± 1.38
C24:1 CER ^ç	0.72 ± 0.25	0.83 ± 0.48	0.74 ± 0.35	0.74 ± 0.30	0.78 ± 0.41
C24:1 GlcCer ^ç	1.82 ± 1.17	2.17 ± 1.87	1.78 ± 1.06	1.99 ± 1.48	1.95 ± 1.48
HVLT imm. ^#	26.23 ± 4.65	24.64 ± 4.60	25.34 ± 4.57	24.11 ± 5.16^*	25.01 ± 4.58
HVLT ret. ^#	0.90 ± 0.18	0.84 ± 0.17	0.86 ± 0.18	0.84 ± 0.20	0.85 ± 0.18
SDMT ^#	47.24 ± 10.76	44.86 ± 9.40 ^!	42.93 ± 8^*	40.09 ± 10.50^*	43.83 ± 8.70 ^!*
SFT ^#	52.26 ± 11.32	48.40 ± 10.84	51.82 ± 12.93	48.41 ± 11.70	50.23 ± 12.08
MoCA ^#	28.22 ± 1.15 ^£	27.3 ± 2.50	27.53 ± 2.03 ^!*	26.80 ± 2.31^*	27.42 ± 2.25 ^!*

Note: The number of people is lower for the ceramide data (n annotated ç) compared to the neuropsychological data (n annotated #) due to the absence of ceramide data for some patients. FDR-corrected P-value threshold for ceramide comparisons: HC vs. LAPD (p < .05 ); HC vs. SPD (p < .05); LAPD vs. SPD (p < .016); LAPD vs. RAPD (p < .05); ASPD vs. SPD (p < .05). FDR-corrected P-value threshold for cognitive comparisons: HC vs. LAPD (p < .05); HC vs. RAPD (p < .025); HC vs. SPD (p < .016); HC vs. ASPD (p < .025); ASPD vs. SPD (p < .025); RAPD vs. SPD (p < .05); LAPD vs. SPD (p < .05). ASPD: Asymmetric putamen denervation (SBR) grouping the right and left predominance of PwPD; CER: ceramides; HC: Healthy controls; HVLT imm.: Hopkins verbal learning test immediate recall; HVLT ret: Hopkins verbal learning test retention; LAPD: Predominant denervation of the left putamen (SBR) for PwPD; MoCA: Montreal cognitive assessment; PwPD: people with Parkinson disease; RAPD: Predominant denervation of the right putamen (SBR) for PwPD; SBR: striatal specific binding ratios; SDMT: Symbol digit modalities test; SFT: Semantic fluency total score; SPD: Symmetric putamen denervation (SBR) for PwPD.

^* Significance with the control group FDR corrected; ^£ Significance with the LAPD group FDR corrected; ^! Significance with the symmetric group FDR corrected

2.3 Relation between cerebrospinal fluid ceramides and cognition as a function of striatal denervation asymmetry (Table 2 and SI 1)

Interestingly, we observed almost uniquely positive correlations with cognition when it comes to ceramides, and negative correlations with cognition when it comes to glucosylceramides. We observed the same patterns of results when analyses were performed on the whole sample (see Table 2), or within subgroups (see SI 1), except for the left asymmetry group where no correlation was found.

Table 2

**Relationship between ceramides, glucosylceramides, and cognition in the whole cohort and by sub-groups (with symmetrical and asymmetrical striatal denervation (SBR)).**
Ceramides/ Glucosylceramides	Cognitive tests	Whole group n = 277	(p value)	SPD n = 148	(p value)	ASPD n = 129	(p value)
C22:0 GlcCer	SDMT	Ns.	Ns.	r = -0.20	p = .015*	Ns.	Ns.
C22:0 GlcCer	SFT	r = -0.18	p < .001*	Ns.	Ns.	r = -0.25	p = .004*
C18:0 CER	SDMT	r = 0.16	p = .002*	Ns.	Ns.	r = 0.21	p = .017*
C22:0 CER	SDMT	Ns.	Ns.	Ns.	Ns.	r = 0.20	p = .021
C18:0 CER	HVLT ret.	r = 0.13	p = .011*	Ns.	Ns.	Ns.	Ns.

Note. The whole group includes people with predominantly right, left and symmetrical striatal denervation (SBR). ASPD: Asymmetric putamen denervation (SBR) grouping the right and left predominance of PwPD; CER: ceramides; GlcCer: glucosylceramides; HVLT ret: Hopkins verbal learning test retention; Ns.: not significant; SDMT: Symbol digit modalities test; SFT: Semantic fluency test; SPD: Symmetric putamen denervation (SBR). * FDR corrected by subgroup.

In this study, we observed results that qualify for the SOC model, by specifying metabolic and cognitive differences according to cerebral asymmetry. Here, we confirmed that symmetrical PwPD are characterised by variation in certain cerebrospinal fluid ceramides (C18:0 CER) and more cognitive deficits than asymmetrical PwPD and HC. Moreover, our metabolic and cognitive results go beyond the predictions that can be derived by the SOC model in the sense that we observed a specificity according to the side of striatal denervation. Left PwPD are distinguished by significantly more ceramides (C22:0 CER) than both right PwPD, and HC. Finally, these results were reinforced by the observation of positive correlations between cognition and ceramides (C22:0 CER and C18:0 CER).

Considering ceramide concentrations, we therefore highlighted a difference in PD clinical phenotypes, with altered metabolic homeostasis ¹⁰ as evidenced by the variations in C22:0 CER and C18:0 CER between our groups. Indeed, the symmetrical group would be depleted of the C18:0 CER. This pattern of disrupted homeostasis coincides with the apoptosis and degeneration phenomena described in ^6,7,Mencarelli and Martinez–Martinez ^8,9. In PD trajectories, a susceptibility to left hemispheric asymmetry would be associated with an increased risk of developing dementia compared to right asymmetry ^13,15.

As pointed out in recent works on the relationship between sphingolipid mechanisms and PD ^3,10, ceramides metabolism could be a very interesting marker in discriminating the intensity of PD. Ever since, a few studies have focused on the relationship between cognition and cerebral ceramides ¹⁶. One of them showed that the accumulation of C16:0, C18:0, C20:0, C22:0 and C24: 1 ceramides in PwPD was associated with more global cognitive impairment ¹⁷. Another study in PD showed that levels of ceramides C14:0 and C24:1 were negatively correlated with verbal memory performance, and that ceramides C22:0, C20:0 and C18:0 were associated positively with higher anxiety levels and the presence of hallucinations ¹⁸. Lastly, it is only recently that longitudinal relationships have been observed between cognition and ceramides levels. Indeed, Huh, et al. ¹¹, showed in PwPD that high sphingolipid ratios were associated with a more rapid cognitive decline over 3 years.

In response to the crucial interest in characterising ceramides variations in PD in order to understand its aggressiveness and adjust our therapeutics ¹⁹, we showed here that ceramides are associated to certain cognitive performances. Indeed, negative relationships between verbal (SFT) and non-verbal (SDMT) executive tasks were found only with accumulation rates of glucosylceramides (C22:0 GlcCer) meaning that the less accumulation of this glucosylceramide we observed, the better the cognitive performance for these processes. However, it is interesting to note that in the opposite direction, only ceramides are positively correlated with cognition (total SDMT and HVLT) indicating that the accumulation of specific ceramides (C18:0 and C22:0 CER) were more associated with better cognitive performance for specific processes.

Finally, we also showed the importance of distinguishing the putamen denervation asymmetry groups according to cognition. In general, we observed lower cognitive performance for the symmetric group compared to the asymmetric groups. This is line with the SOC model according to which a symmetric dopaminergic denervation reflects higher neuropathological burden and is thus associated with accelerated cognitive decline ¹. More precisely, we observed specificities depending on the side of the asymmetry. The symmetrical PwPD group performed worse on a non-verbal executive task (SDMT) compared to the left asymmetrical PwPD group; and worse on a verbal executive task compared to the right asymmetrical PwPD group. On a global cognitive efficiency task (MoCA), the symmetrical group differed only from the right asymmetrical PwPD group. Our observations suggest that symmetric denervation shares similar vulnerabilities with left asymmetric denervation for cognitive decline ¹⁵. It therefore seems crucial, based on our results and the literature, to take into account variations in ceramides and glucosylceramides in the investigation of cognitive impairment according to striatal denervation asymmetry in early PD.

This work has certain limitations. First, we controlled for genetic factors that are known to impact on the development of PwPD, however this does not exclude other genetic elements that may have an impact, such as synuclein alpha (SNCA) gene polymorphisms ²⁰. Second, we observed that the absolute values of putamen denervation in symmetrical patients had greater baseline denervation than the right putamen of left asymmetrical patients and the left putamen of right asymmetrical patients. These results highlight that baseline denervation is more pronounced in the symmetrical group suggesting a neurodegenerative distinction between our groups. We controlled for these differences in denervation in our analyses, but these differences remain to be fully understood. Finally, sociodemographic aspects related to lifestyle, geography and diet could have an impact on metabolism and the observed neurodegenerative trajectories ^21,22.

In view of the existing literature and our results on the relationship between PD, the SOC model, cognition and ceramides metabolism to characterise the development of different PD trajectories, it seems of crucial interest to investigate the relationship between physiology and cognition in more detail. It also seems necessary to study these relationships longitudinally during the development of the disease using cerebrospinal fluid and serum markers. The results on metabolism open the door to the creation of more personalized treatments ¹⁹.

Our results suggest different levels of precision regarding the metabolism of ceramides, glucosylceramides and their links to cognition in brain asymmetry in PwPD. In the absence of a known genetic mutation, cerebrospinal fluid derived ceramides and cognitive performance appear to distinguish subgroups of PwPD with left from right striatal denervation asymmetry or symmetrical denervation. We thus highlight the discriminatory potential of metabolism and the links with cognition in the context of PwPD.

5.1 Participants

The dataset used in this article was obtained from the Parkinson's Progression Markers Initiative (PPMI) database (www.ppmi-info.org/data)²³. The global PPMI cohort is composed of men or women over 30 years of age, not taking any medication within 6 months of the first baseline visit. Participants had at least two of the following symptoms: resting tremor, bradykinesia and rigidity (must have either resting tremor or bradykinesia). The presence of a stage I or II on the Hoehn and Yahr assessment at the first baseline visit was observed. For updated information on this study, see www.ppmi-info.org.

Participants with data from ceramides concentrations, and DAT-SPECT analysis were retained. Patients with LRRK2 G2019S, G105R/G183E, I479L, R78C, R83C genotypes ^11,24 and/or no DAT-SPECT data of putamen denervation or other variables of interest (ceramides, cognition) were excluded from the study. We thus selected a total of 496 individuals composed of n = 167 healthy individuals and n = 329 PwPD (see Table 3). We analysed the data at baseline of the disease, no dopaminergic treatment had been introduced yet in PwPD, and participants were comparable in age, gender, education, and time from diagnosis to the start of the study measurements (see Table 3).

Table 3

Sociodemographic features of the cohort
HC n = 167	LAPD n = 72	RAPD n = 83	SPD n = 174	p value
Mean Age in years (± SD)	60.3 ± 11.6	59.4 ± 9.8	59.3 ± 9.96	62.4 ± 8.45	.057
Men %	64%	51%	65%	64%	.871
Average years of education (± SD)	16.2 ± 2.8	15.6 ± 2.4	16.1 ± 2.6	15.7 ± 3.0	.849
Average time between diagnosis and baseline measurements (± SD)	Na.	6.8 ± 7.11	7.5 ± 7.83	6.62 ± 6.26	.574
Mean putamen denervation (SBR): left/right (± SD)	Na.	1.15 ± 0.40/ 0.62 ± 0.22	0.60 ± 0.22/ 1.13 ± 0.37	0.80 ± 0.32 / 0.80 ± 0.32	< .001
Handedness (L/ambidextrous/R)	22/9/135	9/2/61	5/6/72	15/1/157	.035
Mean UPDRS total score (± SD)	4.72 ± 4.24	33.44 ± 11.09	29.46 ± 12.39	33.68 ± 13.39	< .001

Note: HC: healthy controls; LAPD: Predominant denervation of the left putamen (SBR); PwPD asymmetry (left/right) of the putamen greater than 20%. Bil.: bilateral; L: left; PwPD: People with Parkinson Disease; NA: not available; RAPD: Predominant denervation of the right putamen (SBR); R: right; SBR: striatal specific binding ratios; SPD: Symmetric putamen denervation (SBR); UPDRS: Unified Parkinson’s Disease Rating Scale; SD: standard deviation. Left asymmetric PwPD have greater denervation on the left putamen than symmetric or right asymmetric PwPD. Similarly, right asymmetric PwPD have greater denervation of the right putamen than other groups. However, we observed that symmetrical PwPD have a higher baseline denervation: the right denervation of left asymmetric PwPD is significantly lower than symmetric PwPD, as is the left denervation of right asymmetric PwPD compared to symmetric PwPD (all comparisons involved p < .001).

Striatal asymmetry calculation. Analysis of striatal asymmetry was based on Kaasinen ²⁵ preliminary calculations of SBR ([striatal region count density/occipital count density]-1). As in the Kaasinen ²⁵ and Fiorenzato, et al. ²⁶ studies, we used the difference in SBR [(right-left putamen SBR)/ (right + left putamen SBR)] to calculate cerebral asymmetry. We determined an asymmetry of the putamen when the ratio was superior to 20% ²⁶. These analyses enabled us to split our sample of patients in n = 83 PwPD patients with predominantly right striatal denervation, n = 72 PwPD patients with predominantly left striatal denervation, and n = 174 PwPD patients with relatively symmetric denervation. However, we observed a greater striatal denervation baseline for symmetrical PwPD compared to asymmetrical left and right PwPD (see Table 3 and Fig. 3).

5.2 Glucosylceramides (GlcCer), ceramides (Cer)

Ceramides in human CSF were extracted by protein precipitation from the PPMI database. We extracted data for C16:0, C18:0, C20:0, C22:0, C24:1 and C24:0 CER and glucosylceramides (GlcCer) (for more details see Huh, et al. ¹¹). Ceramides or glucosylceramides determinations were reported in ng/mL.

5.3 Cognition

We used measures of cognition related to verbal episodic memory with the HVLT immediate recall ²⁷ and executive function, particularly verbal executive function, with the SFT ²⁸, and non-verbal executive function with the SDMT ²⁹. Within verbal episodic memory, we considered the scores for retention of information and the ratio of immediate to total recall. We also retained a global measure of cognition with the MoCA ³⁰.

5.4 Statistical analysis

Kolmogorov Smirnov test revealed the non-parametric nature of the data. Therefore, non-parametric Kruskall-Wallis analyses were performed for sociodemographic and clinical data. A power analysis to estimate an ideal sample was performed. The analysis was set at 1-β = 0.80 and α = 0.05 and based on the sample of Huh et al,¹¹ split between 114 HC with a median fraction of 0.79% glucosylceramides and 189 idiopathic PwPD with a median fraction of 0.93% glucosylceramides, indicated that we needed to split 85 people into groups to best detect an effect. Our groups were split as (HC = 167; ASPD = 155; SPD = 174) and the ASPD was split into (LAPD = 72; RAPD = 83) and are therefore close to or beyond the estimated sample. Second, for the data of interest, due to a non-parametric nature of the data and a potential effect of confounding variables in generalized linear mixed models (GLMM), pairwise group comparisons were performed by GLMM with the left and right putamen denervation rates as random factors. These models control for random effects such as inter-individual variability (handedness, denervation), in addition to fixed effects. We used a GLMM model following gamma regression for each ceramide and each cognitive test. For example, for a GLMM model we considered 3 within-participant factors: the level of a ceramides or glucosylceramides (12 levels: C16:0 CER, C18:0 CER, C20:0 CER, C22:0 CER, C24:1 CER, C24:0 CER and C16:0 GlcCer, C18:0 GlcCer, C20:0 GlcCer, C22:0 GlcCer, C24:1 GlcCer, C24:0 GlcCer), cognitive performance (3 levels: MoCA, SFT, SDMT) and putamen denervation asymmetry (SBR) (4 levels: predominantly left striatal denervation, right striatal denervation, symmetric striatal denervation and HC). The false discovery rate (FDR) correction was applied according to each asymmetry subgroup comparison for all ceramides/glucosylceramides and cognitive processes. We chose to perform a Benjamini-Hochberg FDR correction. Studies have shown that Bonferroni is not an adequate correction for positive variables and for a large number of analyses ³¹. Our decision to perform Benjamini-Hochberg FDR corrections allowed us to avoid type II errors ³². We added the significance thresholds for each group comparison in the Table 1. Finally, we performed two-way Spearman correlations between significantly discriminating ceramides and cognition on the whole sample and by subgroups. Analyses were performed with SPSS version 27.0.0.0.

ASPD: Asymmetric putamen denervation (SBR) grouping the right and left predominance; CER: Ceramides; GlcCer: Glucosylceramides; HC: healthy controls; HVLT: Hopkins verbal learning test; LAPD: Predominant denervation of the left putamen (SBR); MoCA: Montreal cognitive assessment; PwPD: people with Parkinson’s disease; RAPD : Predominant denervation of the right putamen (SBR); SBR: Striatal specific binding ratios; SDMT: Symbol digit modalities test; SFT: Semantic fluency test; SOC: alpha-Synuclein Origin site and Connectome; SPD: Symmetric putamen denervation (SBR); UPDRS: Unified Parkinson disease rating scale.

Data availability

All data analysed are available at www.ppmi-info.org/access-data-specimens/download-data.

Ethics

The study was conducted in accordance with the Declaration of Helsinki. All participants in the PPMI clinical database had the ability to provide informed consent in accordance with Good Clinical Practice (GCP), the International Conference on Harmonisation (ICH) and local regulations.

Data availability

The data used in this study were obtained from the PPMI database (www.ppmi-info.org/data). For more information, visit http://www.ppmi-info.org.

Acknowledgements

This work was supported by funding from the PPMI – a public-private partnership –funded by the Michael J. Fox Foundation for Parkinson's Research and funding partners 4D Pharma, Abbvie, Acurex Therapeutics, Allergan, Amathus Therapeutics, ASAP, Avid Radiopharmaceuticals, Bial Biotech, Biogen, BioLegend, Bristol-Myers Squibb, Calico, Celgene, Dacapo Brain Science, Denali, The Edmond J. Safra Foundaiton, GE Healthcare, Genentech, GlaxoSmithKline, Golub Capital, Handl Therapeutics, Insitro, Janssen Neuroscience, Lilly, Lundbeck, Merck, Meso Scale Discovery, Neurocrine Biosciences, Pfizer, Piramal, Prevail, Roche, Sanofi Genzyme, Servier, Takeda, Teva, UCB, Verily, and Voyager Therapeutics.” The first author was funded by Swiss National Foundation grant no. 105314_182221 (PI: Pr Julie Péron). The funders had no role in data collection, discussion of content, preparation of the manuscript, or decision to publish.

Author contributions

Anthony Nuber-Champier: Carried out the analysis and drafting of the manuscript

Philippe Voruz: Carried out the proofreading and construction of the manuscript

Ioana-Medeleine Constantin: Participated in the writing and reviewing

Alexandre Cionca: Performed data cleaning and manuscript construction

Julie Péron: Participated in the writing, conceptualisation of the study, Supervision, Reviewing and Editing.

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Borghammer, P. The α-synuclein origin and connectome model (SOC Model) of Parkinson’s disease: explaining motor asymmetry, non-motor phenotypes, and cognitive decline. Journal of Parkinson's Disease 11, 455–474 (2021).
Knudsen, K. et al. Asymmetric dopaminergic dysfunction in brain-first versus body-first Parkinson’s disease subtypes. Journal of Parkinson's Disease, 1–11 (2021).
Lin, G., Wang, L., Marcogliese, P. C. & Bellen, H. J. Sphingolipids in the Pathogenesis of Parkinson’s Disease and Parkinsonism. Trends in Endocrinology & Metabolism 30, 106–117 (2019).
Abbott, S. K. et al. Altered ceramide acyl chain length and ceramide synthase gene expression in Parkinson's disease. Movement Disorders 29, 518–526 (2014).
David, F. J. et al. Subthalamic Peak Beta Ratio Is Asymmetric in Glucocerebrosidase Mutation Carriers With Parkinson's Disease: A Pilot Study. Frontiers in neurology, 1661 (2021).
Zunke, F. et al. Reversible conformational conversion of α-synuclein into toxic assemblies by glucosylceramide. Neuron 97, 92–107. e110 (2018).
Scherzer, C. R. & Feany, M. B. Yeast genetics targets lipids in Parkinson's disease. Trends in genetics 20, 273–277 (2004).
Mencarelli, C. & Martinez–Martinez, P. Ceramide function in the brain: when a slight tilt is enough. Cellular and Molecular Life Sciences 70, 181–203 (2013).
Mazzulli, J. R. et al. Gaucher disease glucocerebrosidase and α-synuclein form a bidirectional pathogenic loop in synucleinopathies. Cell 146, 37–52 (2011).
Plotegher, N., Bubacco, L., Greggio, E. & Civiero, L. Ceramides in Parkinson’s disease: from recent evidence to new hypotheses. Frontiers in Neuroscience 13, 330 (2019).
Huh, Y. E. et al. Glucosylceramide in cerebrospinal fluid of patients with GBA-associated and idiopathic Parkinson’s disease enrolled in PPMI. npj Parkinson's Disease 7, 1–7 (2021).
Ricciardi, L. et al. Working on asymmetry in Parkinson’s disease: randomized, controlled pilot study. Neurological Sciences 36, 1337–1343 (2015).
Voruz, P., Constantin, I. M. & Péron, J. A. Biomarkers and non-motor symptoms as a function of motor symptom asymmetry in early Parkinson's disease. Neuropsychologia, 108419 (2022).
Riederer, P. et al. Lateralisation in Parkinson disease. Cell and tissue research 373, 297–312 (2018).
Voruz, P. et al. Motor symptom asymmetry predicts non-motor outcome and quality of life following STN DBS in Parkinson's disease. Scientific reports 12, 1–9 (2022).
Kalinichenko, L. S., Gulbins, E., Kornhuber, J. & Müller, C. P. Sphingolipid control of cognitive functions in health and disease. Progress in Lipid Research, 101162 (2022).
Mielke, M. M. et al. Plasma ceramide and glucosylceramide metabolism is altered in sporadic Parkinson's disease and associated with cognitive impairment: a pilot study. PloS one 8, e73094 (2013).
Xing, Y. et al. Associations between plasma ceramides and cognitive and neuropsychiatric manifestations in Parkinson's disease dementia. Journal of the neurological sciences 370, 82–87 (2016).
Custodia, A. et al. Ceramide metabolism and Parkinson’s disease—Therapeutic targets. Biomolecules 11, 945 (2021).
Pedersen, C. C. et al. A systematic review of associations between common SNCA variants and clinical heterogeneity in Parkinson’s disease. npj Parkinson's Disease 7, 1–10 (2021).
Ruiz-Núñez, B., Pruimboom, L., Dijck-Brouwer, D. J. & Muskiet, F. A. Lifestyle and nutritional imbalances associated with Western diseases: causes and consequences of chronic systemic low-grade inflammation in an evolutionary context. The Journal of nutritional biochemistry 24, 1183–1201 (2013).
Mattson, M. P. & Arumugam, T. V. Hallmarks of brain aging: adaptive and pathological modification by metabolic states. Cell metabolism 27, 1176–1199 (2018).
Marek, K. et al. The Parkinson progression marker initiative (PPMI). Progress in neurobiology 95, 629–635 (2011).
Liu, G. et al. Specifically neuropathic Gaucher's mutations accelerate cognitive decline in Parkinson's. Annals of neurology 80, 674–685 (2016).
Kaasinen, V. Ipsilateral deficits of dopaminergic neurotransmission in Parkinson's disease. Annals of clinical and translational neurology 3, 21–26 (2016).
Fiorenzato, E., Antonini, A., Bisiacchi, P., Weis, L. & Biundo, R. Asymmetric dopamine transporter loss affects cognitive and motor progression in Parkinson's disease. Movement Disorders 36, 2303–2313 (2021).
Benedict, R., Schretlen, D., Brandt, J. & Groninger, L. Revision of the Hopkins verbal learning test: reliability and normative data. Clin Neuropsychol 12, 43–55 (1998).
Spreen, O. Neurosensory center comprehensive examination for aphasia. Neuropsychological Laboratory (1977).
Smith, A. Symbol digit modalities test. (Western Psychological Services Los Angeles, 1973).
Nasreddine, Z. S. et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. Journal of the American Geriatrics Society 53, 695–699 (2005).
Thissen, D., Steinberg, L. & Kuang, D. Quick and easy implementation of the Benjamini-Hochberg procedure for controlling the false positive rate in multiple comparisons. Journal of educational and behavioral statistics 27, 77–83 (2002).
Lee, S. & Lee, D. K. What is the proper way to apply the multiple comparison test? Korean journal of anesthesiology 71, 353–360 (2018).

(Not answered)

NuberChampieretalSI.docx

Cerebrospinal fluid ceramides and cognition as a function of striatal asymmetry in early Parkinson's disease

Status:

Version 1

Abstract

Figures

1. Introduction

2. Results

2.1 Intergroup differences in cerebrospinal fluid ceramide concentration (Table 1)

2.2 Intergroup differences in cognitive performance (Table 1)

3. Discussion

4. Conclusion

5. Method

5.1 Participants

5.3 Cognition

5.4 Statistical analysis

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1