Phosphatidylethanolamines are the main lipid class altered in red blood cells from patients with VPS13A disease/chorea-acanthocytosis

Introduction

VPS13A disease (formerly known as chorea-acanthocytosis) is a neurodegenerative disorder of the young adulthood and an important differential of Huntington’s disease¹. Together with XK disease (McLeod syndrome) it has been classified as neuroacanthocytosis syndrome as it is characterized by striatal degeneration and the presence of deformed red blood cells (RBCs), referred to as acanthocytes². Typical clinical manifestations include a variety of movement disorders (chorea and dystonia with orofacial predominance, in later stages parkinsonism), epilepsy, behavioral and cognitive impairment as well as peripheral neuro- and myopathy¹.

The autosomal-recessive condition is caused by biallelic pathogenic variants in the VPS13A gene leading in most cases to a complete loss of the respective protein, VPS13A/chorein³. VPS13A belongs to a family of 4 proteins, VPS13A-D, that are all related to neurodegenerative or neurodevelopmental disorders, such as hereditary parkinsonism (VPS13C) or ataxia (VPS13D)^4-6. Only recently, VPS13A has been assigned to the new protein superfamily of bridge-like lipid transfer proteins (BLTPs)^7,8. Forming hydrophobic grooves that span between two organellar membranes at membrane contact sites, these proteins mediate direct bulk lipid transfer, most likely selective for phospholipids⁹. VPS13A localizes between the endoplasmic reticulum and mitochondria, lipid droplets or the plasma membrane⁹. At the plasma membrane, it has been shown to form a complex with the putative scramblase XK^10,11. In support of a pathomechanistic role of altered membrane lipid distribution and supply in VPS13A disease, elevated levels of several sphingo- and phospholipids have been recently found in the striatum of VPS13 patients¹². Also, in Huntington’s disease, a distinct shift in the sphingolipid profile of the caudate has been reported¹³.

In this exploratory study, we aimed to study the lipid composition of RBCs from VPS13A patients for various reasons: 1) RBCs are – besides neurons – prominently affected by the disease as a high proportion are acanthocytic. 2) RBCs are “products” of a complex rearrangement of membranes and organelles during erythropoiesis¹⁴ potentially requiring membrane lipid transfer. 3) RBCs are – in contrast to brain tissue – easily obtainable. RBC lipid composition may therefore be an ideal biomarker candidate.

Methods

Five genetically confirmed VPS13 patients (4 males, 1 female, mean age 45.6, min 32, max 52 years) and 12 healthy controls (9 males, 3 females, mean age 40.7, min 23, max 56 years) were included in this study. Demographic and clinical data of patients are shown in Table S1. The study was approved by the ethics committee at the Technische Universität Dresden (EK45022009, EK78022015). All participants gave written informed consent in accordance with the Declaration of Helsinki.

EDTA blood samples were processed in accordance to the sample preparation guidelines from Lipotype GmbH (Dresden, Germany; see supplementary material). Mass spectrometry-based lipid analysis was performed by Lipotype GmbH as previously described using a QExactive Orbitrap mass spectrometer (Thermo Fisher Scientific, Darmstadt, Germany)¹⁵. Table S2 shows the list of analyzed lipid classes and the respective structural detail level of the analysis.

Lipid data was analyzed in mol% for better comparability. An occupational threshold was applied to filter lipid (sub)species that were not adequately measured to solidify findings (see supplementary material).

Multiple t-tests on all remaining (sub)species were performed (without assuming a consistent standard deviation). In order to account for multiple testing, the Benjamini Hochberg procedure with a false discovery rate of 5% was used. Analysis was performed at different levels: lipid class level, lipid (sub)species level, structural or functional category level as well as grouped by chain length/double bounds number/OH-groups number. For visualization of the results on the lipid (sub)species level, a Vulcano plot was generated.

Results

There was no statistically significant difference in age and sex distribution between the two groups analyzed in this study (Table S3).

Untargeted lipidomic analysis was achieved by mass spectrometry by screening for 23 lipid classes. In total, 575 lipid species and subspecies were measured. After application of the occupational threshold, 313 lipid species and subspecies remained for further analysis. Out of the 23 classes, 13 were easily detectable including ceramide (Cer), cholesterol (Chol), hexosylceramide (HexCer), lyso-phosphatidylcholine (LPC), phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylcholine-ether (PE O-), phosphatidylinositol (PI), phosphatidylserine (PS) and sphingomyelin (SM), in agreement with previous reports on RBCs ¹⁶. Moreover, we were able to detect additional species such as lyso-phosphatidylethanolamine (LPE), phosphatidate (PA) and phosphatidylcholine-ether (PC O-).

On the lipid class level, no relevant differences between healthy controls and VPS13A patients could be detected (Fig 1A). The most distinct difference was seen in the PI class with less mol% in the disease group, however, this did not reach statistical significance (p=0.07, q=0.51). Also, after grouping the lipids into structural (glycerophospholipids, sphingolipids, and sterols) or functional categories (lyso vs. membrane lipids), analysis did not reveal striking differences (Fig 1 B, C), as was the case for the number of double bounds and hydroxyl groups of the lipids (Fig 1 D, E). Likewise, while there was a tendency of decrease of lipids with medium length fatty acid chains (30-36 C-atoms) and of increase of lipids with very long chains (39-44 C-atoms) in the patient group compared to control, none of these differences reached statistical significance (Fig 1 F).

• Download figure
• Open in new tab

Fig 1.

Lipidomics analysis on lipid class (A), structural category (B), functional category (C) level. Structural analysis of lipids in respect to number of double bounds (D), fatty acid chain length (E), and number of hydroxyl groups (F).

Ctrl controls, VPS13A VPS13A patients, Cer Ceramide, Chol Cholesterol, HexCer Hexosylceramide, LPC lyso-Phosphatidylcholine, LPE lyso-Phosphatidylethanolamine, PA Phosphatidate, PC (O-) Phosphatidylcholine (-ether), PE (O-) Phosphatidylethanolamine (-ether), PI Phosphatidylinositol, PS Phosphatidylserine, SM Sphingomyelin, GPL glycerophospholipids, SL sphingolipids, ST sterols, LYS lyso lipids, MEM membrane lipids.

On the lipid species and subspecies level, however, distinct differences with small effect size (fold change) were observed. The increase of Cer34:1;2 was the most significant finding (Fig 2 A, B). Most of the (sub)species with significant change after Benjamini Hochberg procedure belonged to the PE or PE O-classes. Within the PEs, a shift in the fatty chain lengths became obvious: PE subspecies with longer fatty acid chains tended to be increased, species with shorter chains to be decreased (Fig. 2 C). This is in line with the (nonsignificant) shift that has been observed in the overall chain length analysis (Fig. 1 F). In addition, decrease of single PC and SM species were also detected.

• Download figure
• Open in new tab

Fig 2.

Lipidomics analysis on species and subspecies level. Volcano blot showing all analysed (sub)species (A). The horizonal dotted line represents the threshold of statistical significance after Benjamini Hochberg procedure with a false discovery rate of 5%. (B) shows all significantly different (sub)species between healthy control and VPS13A disease groups. (C) Relative change of Phosphatidylethanolamine (PE) subspecies in VPS13A disease sorted by fatty acid chain length and double bond number; results are expressed as percentage of controls (healthy control values were set 100%).

Ctrl controls, VPS13A VPS13A patients, Cer Ceramide, Chol Cholesterol, HexCer Hexosylceramide, LPC lyso-Phosphatidylcholine, LPE lyso-Phosphatidylethanolamine, PA Phosphatidate, PC (O-) Phosphatidylcholine (-ether), PE (O-) Phosphatidylethanolamine (-ether), PI Phosphatidylinositol, PS Phosphatidylserine, SM Sphingomyelin

Discussion

VPS13A disease has recently become paradigmatic for a new pathophysiological concept in neurodegeneration: disturbed bulk lipid transfer at membrane contact sites^2,9,17. VPS13A is a bridge-like protein enabling direct bulk lipid transfer between intracellular membranes. It seems that disturbances of this process are central for VPS13A and related diseases although the exact role of bulk lipid transfer in neuronal and other mainly affected cells such as RBCs is subject of further research.

Based on these recent molecular developments in the field, we studied for the first time the lipid composition of RBCs from VPS13A patients using state-of-the-art lipidomics analysis. RBCs are clearly affected by the disease as acanthocytosis is a core feature and VPS13A deficient RBCs reflect at least partly the pathophysiology of the disease^1,18-20. Therefore, RBCs may be an easily accessible surrogate and biomarker for pathology of the nervous system. Of note, there is significant evidence that lipid dysmetabolism is causal for the development of acanthocytosis. First, hypo-/abetalipoproteinemia leads to the appearance of acanthocytes ^21,22. Second, in liver failure where acanthocytosis is observed, irregularities in lipid metabolism, particularly an excess of chol, have been associated with the deformation of RBCs²³. On the other side, lipid analysis in the “pre-lipidomics era” has not revealed consistent differences in RBCs from neuroacanthocytosis patients¹⁸.

In line with that, our study did not show a generalized major disturbance of lipid classes, but revealed interesting distinct changes at the lipid species and subspecies level, including PE (O-) and Cer, but also single PC, and SM (sub)species.

PE play a central role in autophagosome formation and is a regulator of autophagy²⁴, a process which has been shown to be impaired in VPS13A disease, as reflected by the presence of membrane remnants^19,20, and which is essential during erythropoiesis. Moreover, the relative abundance of PE species highly evolves during reticulocyte maturation into RBCs: longer and more unsaturated PE species decrease whereas smaller and more saturated PE species increase (Minetti et al, BioRxiv, https://doi.org/10.1101/2023.06.02.543386). Interestingly, we showed here a reverse “acyl chain remodeling”: The former group increased in VPS13A patients while the latter decreased, suggesting impairment of RBC maturation. Therefore, these findings may be related to acanthocyte formation during erythropoiesis and the pathophysiology of VPS13A disease. Supporting this possibility, Cer, PC and PE species were also shown to be affected in hypobetalipoproteinemia, another acanthocyte-related disease²². Interestingly, Csf1, a Vps13-like BLTP, has been shown to transfer PEs to the ER for GPI anchor synthesis²⁵.

Also, as PE is a non-bilayer forming lipid, especially with longer chain length²⁶, an increase of longer PE species might at least partially explain the morphological alterations of acanthocytes. Moreover, non-bilayer lipids may affect integration of proteins into membranes, their lateral movement and their function²⁷.

Cer act as the foundational element for complex sphingolipids and is important in cellular signaling. Accordingly, abnormal Cer levels have been associated with several neurodegenerative conditions²⁸.

The rather distinct lipid alterations may be viewed as unexpected considering the function of VPS13A as lipid transfer protein. However, as most of the altered lipids are phospholipids, this finding is consistent with the suspected role of VPS13A as (specifically) phospholipid transferring protein⁹. Also, quantitative analyses might not have captured localized changes in membrane composition without major disturbances of overall lipid content. Furthermore, variations within the RBC population (e.g., acanthocytic vs. non-acanthocytic) could have masked specific alterations in a subgroup of cells. As VPS13A plays a role in lipid transfer between organellar membranes and as mature RBCs lack such organelles, the pathogenic process occurs potentially mainly during erythropoiesis. Therefore, quantitative lipid analysis might reveal more pronounced alterations in RBC precursor cells. Another limitation of this study relates to the low number of patient samples due to the ultra-rarity of the disease which might have resulted in an insufficient statistical power to detect even more sub2tle changes.

The RBC lipidomics data presented are not conclusive with the data derived from post mortem brain tissue¹². This is partially due to the differences in covered lipid classed by the measurements. However, this may also point to different effects of VPS13A deficiency in RBCs and the brain.

In summary, the alterations of particular lipid species in RBCs in VPS13A disease contributes to the pathophysiological understanding. Further studies need to focus on lipid composition of RBC precursor cells and on potential localized changes in distinct RBC membrane domains.

Also, our results may result in the development of PE species as biologically relevant biomarker for VPS13A disease which is a prerequisite for future clinical studies²⁹.

Authors’ Roles

KP, HG, AH – study design. KP and AH: patient recruitment, KP, AS, HG – execution. KP, AS, GMM, BF, CK, DT – data analysis. KP, AH: funding acquisition. AH: supervision, project administration; KP, AS, GMM, DT – writing. All authors - editing of final version of the manuscript.

Financial Disclosures of all authors (for the preceding 12 months)

KP: Received funding from the Rostock Academy of Science (RAS) and the Deutschen Gesellschaft für Parkinson und Bewegungsstörungen e.V. Stock Ownership in medically-related fields.

AH has received personal fees and non-financial support from ITF Pharma, Biogen and Desitin during the conduct of the study outside of the submitted work. He received funding from the Schilling Stiftung für medizinische Forschung im Stifterverband, VDI/BMBF, ESF, target ALS foundation outside the submitted work.

CK is CTO and shareholder of Lipotype GmbH.

AS, GMM, HG, BF, DT have no financial disclosure to declare.