A disease causing ATLASTIN 3 mutation affects multiple endoplasmic reticulum-related pathways
- 497 Downloads
Atlastins (ATLs) are membrane-bound GTPases involved in shaping of the endoplasmic reticulum (ER). Mutations in ATL1 and ATL3 cause spastic paraplegia and hereditary sensory neuropathy. We here show that the sensory neuropathy causing ATL3 Y192C mutation reduces the complexity of the tubular ER-network. ATL3 Y192C delays ER-export by reducing the number of ER exit sites, reduces autophagy, fragments the Golgi and causes malformation of the nucleus. In cultured primary neurons, ATL3 Y192C does not localize to the growing axon, resulting in axon growth deficits. Patient-derived fibroblasts possess a tubular ER with reduced complexity and have a reduced number of autophagosomes. The data suggest that the disease-causing ATL3 Y192C mutation affects multiple ER-related pathways, possibly as a consequence of the distorted ER morphology.
KeywordsAtlastin Hereditary spastic paraplegia Endoplasmic reticulum Secretory transport
Length-dependent axonopathies comprise a large group of hereditary disorders affecting both afferent and efferent neurons with exceptional long axons [1, 2]. Among this large group, peripheral motor neurons are predominantly affected in Charcot–Marie–Tooth (CMT) neuropathies and distal pure hereditary motor neuropathies (HMN), whereas sensory neurons are affected in hereditary sensory and autonomic neuropathies (HSAN). Upper motor neurons, in contrast, are affected in hereditary spastic paraplegias (HSP) . Common to these heterogeneous groups of neurological disorders is the length-dependent axonopathy of projecting neurons.
Causative mutations for axonopathies have been identified in genes functioning in several neuronal processes, among them intracellular transport, endosomal function and others (for review see Refs. [1, 3, 4]). Of interest are a number of axonopathy-causing genes involved in the architecture of the endoplasmic reticulum (ER). Many of them code for proteins with a reticulon-domain or reticulon-like domain, namely ARL6IP1, ATL1, ATL3, FAM134B, REEP1, RTN2 and SPAST (reviewed in Ref. ). The Atlastins (ATLs), gene products from ATL1-3, are large dynamin-related GTPases that are involved in ER network formation [5, 6] and maintenance . ATL1 was identified as the gene product of the ATL1/SPG3A gene locus  and is mutated in app. 10% of autosomal-dominant pure/uncomplicated forms of HSP  and in rare cases of HSAN1 . Two heterozygous mutations causing HSAN were described in ATL3, Y192C  and P338R . Both mutations are defective in dimerization and fusion, resulting in aberrant bundling of ER tubules .
The role of ATLs in ER-membrane fusion is well established. ATLs can dimerize in cis and trans, and are believed to tether adjacent membranes and bring them closely together to allow fusion. ATLs can mediate fusion in vitro, but for efficient fusion additional factors were proposed . Mammalian ATLs differ in tissue expression, with ATL1 being mainly expressed in the CNS [8, 15] whereas ATL2 and ATL3 are more ubiquitously expressed . All ATLs localize to the ER, ATL1 throughout ER tubules, ATL2 and 3 more concentrated to three-way-junctions (3WJs) [7, 11, 17].
A role for ATLs in membrane traffic is less clear. Two publications found no role for ATL1-3 in ER–Golgi transport, assayed by knock-down, KO or overexpression of dominant negative ATLs [16, 18]. Another study concluded that ATL1 plays a role in vesicle trafficking at the ER/Golgi interface, but transport was not directly assessed . Recently, it was shown that knockdown of ATL2 but less so of ATL3 in HeLa cells reduced ER–Golgi transport .
The pathomechanism of ATL mutations thus remains largely unclear. ATLs are directly or indirectly involved in many ER-related cellular processes like trafficking , ER stress  and lipid droplet biogenesis . Many findings support the notion that mutated ATLs are defective in forming a fully functional axonal ER (reviewed in Ref. [1, 3]). Interestingly, not all HSP or HSAN disease variants of ATLs are defective in ER network formation, GTP-hydrolysis, dimer formation and fusion .
We here report that the ATL3 disease variant Y192C slows down ER–Golgi trafficking, induces Golgi disruption, causes ER morphology defects, reduces autophagosome formation and affects nuclear shape.
Materials and methods
Antibodies, plasmids and chemical compounds
See supplementary experimental procedures.
For construction of untagged human ATL3, ATL3 was amplified from hATL3-myc pCI neo  using forward (fwd) primer 5′-ggatccATGTTGTCCCCTCAGCGAGTGG-3′, introducing a BamHI site and reverse (rev) primer 5′-gcggccgcCTATTGAGCTTTTTTATCCATGGATGGTCTTCC-3′, introducing a stop codon and a NotI restriction site. For construction of N-terminal myc-tagged human ATL3, ATL3 was amplified from hATL3-myc pCI neo using primers fwd 5′-ggatccATGGAACAAAAACTTATTTCTGAAGAAGATCTGTTGTCCCCTCAGCGAGTGG-3′, introducing a N-terminal myc-tag and a BamHI site and rev: 5′-gcggccgcCTATTGAGCTTTTTTATCCATGGATGGTCTTCC-3′, introducing a stop codon and a NotI restriction site. In both cases, the restriction sites were used for cloning the respective fragment into pcDNA3.1 (+) Hygro. N-terminally GFP-tagged human ATL3 was constructed by amplifying ATL3 from hATL3-myc pCI neo using fwd primer 5′-CTCGAGCTATGTTGTCCCCTCAGCGAGTGG-3′, introducing a XhoI site, and rev primer 5′-ggatccCTATTGAGCTTTTTTATCCATGGATGGTCTTCC-3′, introducing a stop codon and a BamHI restriction site. The restriction sites were used for cloning the fragment into pEGFP C1. The Y192C mutation was introduced by standard site-directed mutagenesis. All sequences were validated.
Cell lines, cell culture, transfection
HeLa Kyoto (in the following called HeLa) and HuH7 cells were maintained in Dulbecco’s modified Eagle Medium + GlutaMax (Invitrogen) supplemented with 10% FBS and incubated at 37 °C, 95% relative humidity and 5% CO2. Cells were transfected using Lipofectamine 2000 according to manufacturer´s instruction. Primary fibroblasts from a patient with the heterozygous Y192C mutation in ATL3 gene were derived from a 3-mm diameter punch biopsy. The procedures were approved by the local ethics committee (reference number A145/11) of the Christian-Albrechts University Medical School (Kiel, Germany). Control fibroblasts 1.1 (AG13334) and 1.2 (AG04151) were obtained from the National Institute of Ageing collection of the Coriell Cell Repository (Camden) and are from the same donor at different age. Patient fibroblasts and controls were maintained in Dulbecco’s modified Eagle Medium + GlutaMax (Invitrogen) supplemented with 20% FBS and incubated at 37 °C, 95% relative humidity and 5% CO2. Fibroblasts were immortalized by lentiviral infection with pCDH hTERT.
Neuronal cell culture
Cortical neurons were isolated from murine embryonic brains (E15.5) and maintained in glia-conditioned neurobasal medium. For transfection, the calcium-phosphate method was used as described  with the following changes: DNA-calcium phosphate-precipitates were prepared by mixing 80 ng/µl plasmid DNA and 250 mM CaCl2 with equal volumes of 2 × BES buffered saline. Coverslips with attached neurons were transferred to a dish containing transfection medium (800 µM sodium pyruvate, 8 mM HEPES, 0.16% glucose, 5.5% ddH2O in MEM, pH 7.65). Subsequently, neurons were incubated at 37 °C for 1 h 30 min. Afterwards, neurons were incubated for 10 min in washing medium (800 µM sodium pyruvate, 8 mM HEPES, 0.16% glucose, 5.5% ddH2O in MEM, pH 7.35) and moved back to their original dish containing glia-conditioned medium. After 24 h, neurons were processed for immunocytochemistry.
Immunocytochemistry and microscopy of cells
For immunofluorescence cells were grown on coverslips. After incubation, cells were fixed with 4% paraformaldehyde and processed for immunofluorescence as described  with antibodies as indicated. For staining of RTN4, cells were fixed with 4% paraformaldehyde containing 0.1% glutaraldehyde as described . For Fig. S1a cells were fixed with 3% glyoxal at pH 5 as described . For Fig. S2a cells were fixed with ice-cold MeOH for 20 min at − 20 °C. Nuclei were stained with Hoechst 33342 (Invitrogen H1399). Images were acquired on a Zeiss Axiovert200 using 20 ×, 40 × or 63 × objective and Zen2012 software. For some settings a confocal-like Apotome slider was used. Images were assembled and processed in Adobe Photoshop. Care was taken that identical settings were applied where images were to be compared. Weak signals like small vesicles or thin ER tubules were enhanced using gamma settings < 1 to match the image perception by eye. For live cell imaging, cells were plated on 3.5-cm glass bottom microwell dishes and imaged in DMEM without phenol red.
Quantification of immunocytochemistry
The number of fluorescently stained structures within cells was quantified using ImageJ software. For this purpose, images were converted to 8-bit grayscale pictures before converting to binary pictures to distinguish objects of interest from background. Overlapping objects were separated using the ‘Watershed’ function. Finally, distinct particles were counted using the ‘Analyze particle’ function. Y192C-expressing HuH7 cells were smaller in size, therefore, α-tubulin staining and CellProfiler were used to quantify and normalize cell size. For quantification of the Golgi fragmentation index, the relative frequency distribution of the number of fragments per cell for each condition was tabulated and the index was calculated using a published formula .
Cell lysis, western blotting, deglycosylation assay
Cells were lysed in STEN buffer, separated by SDS-PAGE, transferred to PVDF membranes and probed with antibodies as described . Deglycosylation of VSVG-EYFP was performed basically as described  but using direct cell lysates instead of immunoprecipitates.
Cells seeded on cover slips were transfected with respective plasmids for 6 h followed by medium change. After 24 h, cells were treated with control medium or pretreated with medium containing 50 μM chloroquine (CQ) for 1 h followed by starvation medium  with 50 μM CQ for 3 h. Finally, cells were fixed and processed for immunocytochemistry.
ER stress assay
HeLa cells were seeded in 6 well plates and transfected the next day with respective plasmids for 6 h, followed by medium change. 24 h after transfection, cells were treated with 2.5 µg/ml tunicamycin or DMSO for 24 h, lysed and processed for western blotting.
For measuring secretion activity HeLa cells stably expressing SEAP were used as described .
Stain-free gel imaging and protein quantification
Stain-free gel imaging was used to assess total protein levels (loading control). After separation, gels were soaked in 10% TCA solution for 5 min followed by three washing steps with dH2O. Subsequently, gels were imaged using the Gel Doc System (Bio-Rad). The intensity of whole lanes was quantified after background subtraction using the Rolling Ball algorithm in ImageJ. The obtained intensity values were used to normalize the western blot signals of proteins of interest. As described by others [32, 33] we found this method superior to normalization to actin or tubulin or other house-keeping genes.
Reported values represent mean ± SEM, unless stated otherwise. Statistical differences between the means of two groups were determined using Welch’s t test, unless stated otherwise. Numbers of independent replicates and p values are indicated in figures or figure legends, a p value < 0.05 was considered significant.
ATL3 Y192C disrupts the ER network and deforms the nucleus
ATL3 Y192C delays ER export by reducing the number of ERES
Taken together, the data suggest that by reducing the number of ERES, ATL3 Y192C causes a reduced transport of cargo from the ER, resulting in disruption of the ERGIC and fragmentation of the Golgi.
ATL3 Y192C reduces the formation of autophagosomes
ATL3 Y192C is excluded from axonal ER
ATL3 Y192C patient cells display a reduced ER network complexity and are compromised in autophagy
In sum, the data suggest that ATL3 Y192C causes numerous defects in ER and Golgi morphology, ER-exit, autophagy and neurite outgrowth.
Causative heterozygous mutations for HSP (SPG3A) and HSAN were identified in ATL1 and ATL3, respectively [8, 11, 12]. Both ATL1 and ATL3 are ER-shaping membrane proteins, but the precise pathomechanism has not been elucidated. We here show that the HSAN-causing ATL3 mutation Y192C but not ATL3 affects multiple ER-related pathways when overexpressed in cell lines or cultured neurons. The ATL3 Y192C is a dominant negative mutation , justifying the use of an overexpression paradigm to study the role of the mutation, although this does not fully match the patient situation. However, since in all our assays only the overexpression of ATL3 Y192C but not ATL3 caused a phenotype we are confident that the observed effects are indeed due to the mutation and not the overexpression per se. In addition, using fibroblasts from a HSAN patient carrying the Y192C mutation heterozygously, we could confirm that features we detected by transfection (ER morphology, autophagy defects) are similarly observed in the heterozygous patient cells. The overexpression approach most likely allows detection of cellular dysfunctions that may go unnoticed in patient fibroblasts in vitro. Subtle changes in, for example, ER–Golgi transport might be difficult to detect, but still be relevant for disease progression. The age of onset of HSAN1 caused by ATL3 Y192C is 14–30 years and only sensory neurons are affected , despite the wide-spread expression of ATL3. This indicates that cellular dysfunctions are subtle in patients and may need to be exaggerated in in vitro studies to detect them, but it should be kept in mind that the overexpression does not exactly mimic the patient situation.
Why are neurons with their extremely long axons preferentially affected? We speculate that their proteostasis systems might be simply already at their limits without further reserves, just coping with the load at young age. Any further burden imposed by aging of the neuron in conjunction with the mutation then may lead to neurodegeneration. Less polarized cells may have more reserves (and in fact in most cases a much shorter lifespan) and therefore, remain asymptomatic.
Using sensitive assays for ER–Golgi transport (with VSVG-EYPF) and secretion (with SEAP) we detected effects of the disease-causing ATL3 Y192C variant on ER export. The reduced secretion is caused by a reduction in the number of ERES, and this also affects the structure of the ERGIC and the Golgi, probably as a consequence of the reduced ER-export. Indeed, ERGIC53, a marker for the ERGIC, seems to accumulate in the ER due to its reduced export. In contrast, using overexpression of dominant-negative ATL1-3, Rismanchi et al. found no effect on transport of VSVG-GFP to the plasma membrane . However, the transport was not quantified and small changes might have gone unnoticed. Namekawa et al. indirectly concluded from changes in ER/Golgi morphology that ATL1 HSP-mutations affected trafficking, but no transport was analyzed . In addition to the transport defects we observed a strongly compromised ER-network, reduced number of autophagosomes and increased p62-levels, defects in nuclear shape and neurite outgrowth deficits in primary neurons.
Are all the observed cellular changes directly caused by ATL3 Y192C, or are they indirectly the consequence of a single primary deficit? This is difficult to answer at this point. ATLs have mainly been associated with ER-network formation [5, 6] and a convincing model how ATLs mediates fusion was published . The ATL3 Y192C mutation also causes severe deficits in establishing an ER network, up to total loss of tubules in highly overexpressing cells (this work, ). While this manuscript was in preparation, it was reported that the ATL3 Y192C mutation is defective in ER fusion and results in aberrant tethering of ER tubules . It is conceivable that for proper assembly and function ERES need an exactly defined ER structure with the right degree of curvature and sheet/tubule ratio . A condensed, non-branched ER induced by ATL3 Y192C would simply not provide enough suitable assembly sites for ERES components, resulting in reduced ER-export. The reduced numbers of ERES could also explain the deficits in autophagy, since ERES are tightly linked to the formation of autophagosomes [52, 53], and the axonal outgrowth deficits. Our findings on autophagy are in contrast to a recent report in which increased autophagy flux was observed in HeLa cells expressing ATL3 Y192C, while patient cells were not analyzed . This discrepancy remains unclear and may be attributed to different cellular systems or levels of overexpression. Finally, the nuclear shape malformations we detected might be indirectly induced by the condensed juxtanuclear ER. Interestingly, ATLs via their function in maintaining a proper ER topology, sustain the efficient targeting of proteins to the inner nuclear membrane . Maybe the nuclear malformations are a consequence of compromised transport to the inner nuclear membrane.
On the other hand, it is conceivable that ATLs mediate not only the fusion of ER tubules but have additional direct functions in the pathways affected here. This is supported by the study of Ulengin et al. who analyzed many ATL1 HSP-mutations and categorized them in two classes: one class comprised mutations with a reduced GTPase activity that had deficits in dimerization and membrane fusion, the other one mutations that had no obvious effect in the respective assays . Additional functions of ATLs could encompass a direct role in ERES formation and/or a direct role in autophagosome formation.
ATL3 Y192C caused neurite growth defects in young neurons and interestingly was not present in distal axons in DIV2 and DIV8 cultured neurons. This cannot be explained by non-establishment or retraction of distal axonal ER or a non-continuous ER. Soluble KDEL-tagged ER-marker is distributed throughout the axon in ATL3 Y192C expressing neurons, indicating a continuous ER ranging into distal axons. It remains to be shown to what extend the ultrastructure of the ATL3-free axonal ER in ATL3 Y192C expressing neurons is changed and to what extend this contributes to pathology. How the mutant ATL3 is prevented from localizing to distal axons remains an open question, maybe involving unknown immobilization/segregation mechanisms.
Why does a mutation in a homologous position (Y196C in ATL1, Y192C in ATL3) of two very conserved proteins cause different diseases? One explanation could be different expression patterns, with ATL1 being more expressed in motor neurons and ATL3 in sensory neurons. Another explanation could be that ATLs have overlapping, but not identical functions. ATL3 is more concentrated in 3WJs, whereas ATL1 is more evenly distributed along the ER [7, 11, 17]. This was suggested to be the consequence of a slower GTPase activity of ATL3 , pointing to differences between the molecules. Endogenous ATLs do not seem to form heterodimers , but if this is true for disease-causing mutated proteins in patient cells has to be investigated. Single or double knock-down/knock-out of one or two of ATL1-3 has little effect on ER morphology [18, 20]. Does this suggest that ATL1 Y196C and ATL3 Y192C are forming heterodimers with other ATL family members, explaining why they act dominant-negatively? During revision of this manuscript Liang et al. demonstrated that exogenously expressed ATL2 interacted with endogenous ATL3 , but further work is needed to fully address this point.
Future studies will also tell if one or more of the observed cellular deficits are causative for the axonal degeneration. A careful assessment and comparison of mutations in ATL1 and 3 but also of mutations in other ER-shaping HSP- or HSAN-genes with as many assays as possible might reveal commonalities or differences that point to the cellular pathway(s) whose malfunctions are actually responsible for the disease.
We are grateful to Eric Snapp, Bertrand Kleizen, Cagatay Günes and Rainer Pepperkok for providing cDNA or antibodies. We thank the FLI facilities Imaging, Functional Genomics and Animal Facility. Special thanks to Daniela Reichenbach for neuronal cultures and Christina Valkova for pplss-tdTomato-KDEL. We thank Franz-Josef Müller (Kiel, Germany) for providing patient fibroblasts.
CK and IK conceived the project. LB performed experiments and analyzed the data. CK designed experiments, analyzed data and wrote the paper with help of all authors. All authors discussed the results and implications at all stages.
Compliance with ethical standards
Conflict of interest
The authors declare no conflict of interest.
- 19.Namekawa M, Muriel MP, Janer A, Latouche M, Dauphin A, Debeir T, Martin E, Duyckaerts C, Prigent A, Depienne C et al (2007) Mutations in the SPG3A gene encoding the GTPase atlastin interfere with vesicle trafficking in the ER/Golgi interface and Golgi morphogenesis. Mol Cell Neurosci 35:1–13CrossRefGoogle Scholar
- 25.Wacker I, Kaether C, Kromer A, Migala A, Almers W, Gerdes HH (1997) Microtubule-dependent transport of secretory vesicles visualized in real time with a GFP-tagged secretory protein. J Cell Sci 110:1453–1463Google Scholar
- 34.Gallione CJ, Rose JK (1985) A single amino acid substitution in a hydrophobic domain causes temperature-sensitive cell-surface transport of a mutant viral glycoprotein. J Virol 54:374–382Google Scholar
- 35.Toomre D, Keller P, White J, Olivo JC, Simons K (1999) Dual-color visualization of trans-Golgi network to plasma membrane traffic along microtubules in living cells. J Cell Sci 112:21–33Google Scholar
- 40.Misumi Y, Misumi Y, Miki K, Takatsuki A, Tamura G, Ikehara Y (1986) Novel blockade by brefeldin A of intracellular transport of secretory proteins in cultured rat hepatocytes. J Biol Chem 261:11398–11403Google Scholar
- 43.Klumperman J, Schweizer A, Clausen H, Tang BL, Hong W, Oorschot V, Hauri HP (1998) The recycling pathway of protein ERGIC-53 and dynamics of the ER–Golgi intermediate compartment. J Cell Sci 111(Pt 22):3411–3425Google Scholar
- 44.Hauri HP, Kappeler F, Andersson H, Appenzeller C (2000) ERGIC-53 and traffic in the secretory pathway. J Cell Sci 113(Pt 4):587–596Google Scholar
- 54.Krols M, Asselbergh B, De Rycke R, De Winter V, Seyer A, Muller FJ, Kurth I, Bultynck G, Timmerman V, Janssens S (2018) Sensory neuropathy-causing mutations in ATL3 affect ER-mitochondria contact sites and impair axonal mitochondrial distribution. Hum Mol Genet. https://doi.org/10.1093/hmg/ddy352 Google Scholar
OpenAccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.