Fourth International Symposium on Neuroacanthocytosis

Abstracts from the Fourth International Symposium on Neuroacanthocytosis

July 1-2, 2008
London and Oxford

Chairs: Prof. Kailash Bhatia, MD, FRCP, Institute of Neurology, University College London; Prof. Anthony P. Monaco, MD, PhD, Wellcome Trust Centre for Human Genetics, University of Oxford

Organizers: Antonio Velayos-Baeza, PhD; Susanne Schneider, MD; Glenn Irvine

8-2 Neuroacanthocytosis - A structural point of view
R. Prohaska
Max F. Perutz Laboratories Medical University of Vienna, Austria

Neuroacanthocytosis (NA) symptoms can result from several gene defects that cause the loss or dysfunction of the respective proteins, namely chorein/VPS13A, Kx, junctophilin-3, PANK2, and additional new candidates like the glucose transporter GLUT1. These proteins have little in common. VPS13A is a homologue of yeast vacuolar protein sorting 13 protein, a large protein reminiscent of cytoskeletal or tethering proteins, Kx is a putative transporter of neutral amino acids, with 10 transmembrane domains, junctophilin-3 is a membrane protein of the endoplasmic/sarcoplasmic reticulum (ER/SR) and a component of junctional complexes with the plasma membrane mediating cross-talk between the cell surface and intracellular ion channels, and PANK2 is a pantothenate kinase, an enzyme involved in lipid metabolism.

Because of the lack of structural similarities, the question is whether these proteins share a common pathway, e.g. in metabolism, or if the mutated proteins are misfolded and prone to aggregation and thus create a proteinopathic phenotype, or both. Misfolded and aggregated proteins are recognized by chaperones and by the unfolded protein response (UPR)-proteins at the ER membrane and are subsequently ubiquitinylated and degraded in the proteasomes. Larger aggregates cannot fit in the proteasome and are deposited in aggresomes, which are microtubule-associated inclusion bodies at the microtubule organizing centre (MTOC) near the centrosomes. The formation of the aggresome is largely believed to be a protective response, sequestering potentially cytotoxic aggregates and also acting as a staging centre for eventual autophagic clearance from the cell. Certain cellular inclusions seen in human disease are thought to represent an aggresomal response, for example the Lewy body seen in neurons in the brain in Parkinson's disease.

Apparently, many neurological diseases are caused by deposited material, mainly protein or lipid, which cannot be degraded in the normal way. Often, this is due to an enzymatic defect like in the sphingolipidoses, Niemann-Pick disease, Tay-Sachs disease, leukodystrophy, and other lysosomal storage diseases. The accumulation of undegraded material eventually leads to (neuronal) cell death by apoptosis and autophagy, processes that are highly regulated and connected. Autophagy is the major route for the clearance of mutated huntingtin in Huntington's disease and α-synuclein in Parkinson's disease. In Alzheimer's disease, a key protein of autophagy, beclin1, was found to be reduced in affected brain regions and in beclin1 knock-out mice the lack of beclin1 led to intraneuronal amyloid β (Aβ) accumulation and neurodegeneration. There is also experimental evidence that the loss of autophagy in the central nervous system causes neurodegeneration and therefore restoring normal autophagy is thought to have therapeutic potential in neurodegenerative diseases.

Autophagy and the apoptotic pathway also play an essential role in erythrocyte maturation. When human erythroid precursor cells expel the nucleus, the resulting reticulocytes clear all remaining organelles by autophagy. Concomitantly, the large amounts of transferrin receptor (TfR) on the cell surface are internalised by endocytosis, Fe3+ is transported through the endosomal membrane and TfR is targeted to multivesicular bodies (MVBs) that bud off luminal vesicles from the limiting membrane, the so-called exosomes, which are expelled into the extracellular medium after fusion of MVBs with the plasma membrane (PM). MVBs and autophagosomes share signalling pathways, because in late stage erythropoiesis MVBs are fused to autophagosomes. The autophagosomes also fuse with lysosomes to generate autophagolysosomes, which digest all enclosed material and eventually fuse with the PM. This process is also essential for the restructuring of the PM from the bizarre shape of early reticulocytes to the smooth, discoid shape of mature erythrocytes and is tightly regulated. It is conceivable that a defect in this process may lead to a deviation of the discocyte shape, as in acanthocytosis.

Regarding the formation of the acanthocyte shape, there are two possible mechanisms according to the bilayer couple hypothesis (reviewed by Gordon Stewart in Neuroacanthocytosis Syndromes II): either the outer leaflet of the bilayer membrane expands or the inner leaflet contracts. Expansion of the outer leaflet could be caused by insertion of lipids that are not equally distributed between the outer and inner leaflet via the flip-flop mechanism like the sphingolipids or by insertion of (lipo)proteins like glycosylphosphatidylinositol (GPI)-linked proteins, which are known to be transferred from one cell to another. Contraction of the inner leaflet could be caused by loss of "inner" phospholipids like phosphatidylserine (PS), phosphatidylethanolamine (PE), and phosphatidylinositol phosphates (PIPs). Particularly the turnover of PIPs is tightly regulated and correlated with the attachment of cortical actin filaments. Aggregation of PI(4,5)P2-containing lipid rafts could lead to unequal distribution of the actin cytoskeleton and thereby cause an acanthocytic shape. Moreover, attachment of cytosolic proteins to the inner leaflet like the annexin family- and BAR domain-proteins could cause an acanthocytic shape. Last, not least, a conformational change of a large transmembrane protein like band 3 could also create acanthocytes. It is not clear, which of these mechanism(s) might play a role in NA. If autophagy is defect in NA, certain proteins might survive, which normally would be destroyed. Since autophagy is essential for discocyte formation, that is creating the right balance between surface area and cytoskeleton attachment, it is feasible that a defect would create misshaped red cells.

The biogenesis of autophagosomes has been studied in yeast and mammalian cells. In yeast, 17 autophagy genes (Atg) have been identified. A key component that is regulating autophagy is the class III phosphatidylinositol 3-kinase, PI(3)KC3. In mammalian cells, the PI(3)KC3 complex contains the proteins UVRAG, Vps15, Vps34, and Beclin1. Recently, the protein Bif-1 was also found to be associated with the PI(3)KC3 complex. Bif-1 contains an N-terminal BAR domain that is associated with lipids and due to its wedge-shaped structure induces curvature on membranes. Bif-1 also contains a coiled-coil domain for oligomerization and an SH3 domain for association with a variety of proteins like UVRAG, dynamin, amphiphysin, Bax, and huntingtin. One can envisage that Bif-1 forms oligomeric complexes with various proteins bound to the SH3 domains, possibly also VPS13A, which could then interact with each other.

Although many Vps proteins were found to play a role in the biogenesis of autophagosomes, VPS13A was not (yet) identified in the relevant complexes. However, it may play a role in endosome-vacuole/phagosome tethering rather than biogenesis. Several large tethering complexes are known that connect various endosomes with intracellular membranes, like the TRAPP complexes between ER and Golgi and within the Golgi apparatus, the CORVET complex tethering trans-Golgi network (TGN) vesicles with the TGN, the exocyst complex tethering TGN vesicles to the PM, and the HOPS complex tethering TGN vesicles to vacuoles. Vps proteins are tethered by the HOPS complex, however, Vps13 was not (yet) identified as part of this complex. However, Vps13A has been identified on phagosomes of Tetrahymena (L. Klobutcher) and human VPS13A was found to associate with large vesicles when overexpressed in mammalian cells (A. Velayos-Baeza). Moreover, VPS13A was identified in human erythrocytes suggesting that it plays a role not only in neurons but also in red cells and probably in the other cell types that express it. In summary, VPS13A may act as a tethering component targeting endosomes/MVBs to vacuoles/autophagosomes for the clearing of cell debris. Non-functional VPS13A may lead to accumulation of debris in the cell and eventually lead to cell death.

The Kell/Kx complex has no similarity to any tethering complex. The Kx protein that is mutated in McLeod Syndrome (MLS) has a similarity to transporters of neutral amino acids and oligopeptides. It is also similar to the C. elegans protein CED-8, which is involved in the apoptosis pathway. Mutations in CED-8 inhibit apoptosis and enhance cell survival but may lead to inefficient cell differentiation/maturation (of red cells and neurons), a process that relies on apoptotic steps and autophagy. There is a balance of anabolism and catabolism (apoptosis and autophagy) in the cell and this balance is dependent on the metabolic state. Neutral amino acids in the cytosol, particularly leucine, activate the mammalian target of rapamycin (mTOR), a Ser/Thr-kinase, which is a key regulator of cell metabolism stimulating protein synthesis and inhibiting autophagy. It is not known, if Kx actually transports amino acids, nor if the lack of Kx has an effect on mTOR signalling, however, the inhibition of apoptosis and autophagy may have long-term consequences on the autophagic removal of accumulated debris in the cell. The mTOR C2 complex is also an important regulator of the cytoskeleton and therefore could be involved in the formation of the red cell acanthocytic shape.

It is certainly desirable to get more insight into the function of the players that have been identified so far, i.e. VPS13A, Kx, JPH3, PANK2, and the newly discovered proteins. Insights might come from cell biological studies and biochemical analyses. Cell biological studies may start with the localization in the cell, overexpression of wild type and mutated proteins, and knock-down of the proteins, respectively. Live microscopy of the GFP-tagged proteins (wild type and mutants) will give insight into the dynamics of the membrane-bound proteins. The biochemical analyses will characterize the proteins and identify binding partners. When soluble fragments can be produced, X-ray crystallography might be the ultimate goal for understanding the structural basis. Insights will also come from comparative proteomic analyses of red blood cells by identifying gained or lost proteins and changes in post-translational modifications in the acanthocytes. Because lipids play an essential role in membrane structure, it is also advisable to perform lipid analyses, preferably lipidomics in collaboration with a specialized laboratory.