ALS/FTD-associated mutation in cyclin F inhibits ER-Golgi trafficking, inducing ER stress, ERAD and Golgi fragmentation

Amyotrophic lateral sclerosis (ALS) is a severely debilitating neurodegenerative condition that is part of the same disease spectrum as frontotemporal dementia (FTD). Mutations in the CCNF gene, encoding cyclin F, are present in both sporadic and familial ALS and FTD. However, the pathophysiological mechanisms underlying neurodegeneration remain unclear. Proper functioning of the endoplasmic reticulum (ER) and Golgi apparatus compartments is essential for normal physiological activities and to maintain cellular viability. Here, we demonstrate that ALS/FTD-associated variant cyclin FS621G inhibits secretory protein transport from the ER to Golgi apparatus, by a mechanism involving dysregulation of COPII vesicles at ER exit sites. Consistent with this finding, cyclin FS621G also induces fragmentation of the Golgi apparatus and activates ER stress, ER-associated degradation, and apoptosis. Induction of Golgi fragmentation and ER stress were confirmed with a second ALS/FTD variant cyclin FS195R, and in cortical primary neurons. Hence, this study provides novel insights into pathogenic mechanisms associated with ALS/FTD-variant cyclin F, involving perturbations to both secretory protein trafficking and ER-Golgi homeostasis.

Cell lysis iPSC-derived motor neurons were harvested after 12 days of maturation for western blot analysis.Motor neurons were washed with PBS and lysed with RIPA buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 0.1% SDS and 1% protease inhibitors).Cell lysate was subsequently centrifuged at 16,000 rcf for 10 min at 4 °C and the supernatant (the RIPA soluble protein fraction) was collected.

VSVG ts045 transport assay
SH-SY5Y cells co-expressing cyclin F and VSVG ts045 were incubated overnight at 40 °C to accumulate VSVG ts045 in the ER.Cycloheximide (Sigma) diluted 20 µg/ml in DMEM was then added and the cells were incubated at 32 °C for 30 min to allow VSVG ts045 to traffic to the Golgi.After one wash in PBS, samples were fixed in 4% PFA in 0.1 M PBS (pH 7.2) for 10 min and processed for immunocytochemistry as described above.At least 20 cells expressing both cyclin F and VSVG ts045 were examined in each group.www.nature.com/scientificreports/

In vitro budding assay
A modified in vitro assay was used to analyse ER vesicle budding 14,29 .Briefly, HEK293T cells co-transfected with constructs encoding cyclin F and VSVG ts045 were incubated overnight at 40 °C to accumulate VSVG ts045 in the ER.Cells were then washed in PBS, resuspended in 5/90 buffer (50 mM HEPES and 90 mM potassium acetate in demineralized water) and incubated with rat liver cytosol (Thermofisher) and an energy regenerating system (50 mM creatine phosphate, 0.2 mg/ml creatine phosphokinase and 1 mM ATP) at 32 °C for 30 min.Identical samples were incubated at 4 °C to monitor non-specific ER fragmentation.The cells were removed by low-speed centrifugation at 4000g for 1 min at 4 °C, followed by 15000g for 1 min, and budded vesicles were recovered by centrifugation at 100,000g for 1 h at 4 °C from the resulting supernatant.The levels of VSVG ts045 cargo in the budded vesicle fractions were quantified by Western blotting using anti-VSVG (1:1000; Sigma V4888) and anti-COPII (Sec23; 1:500; Pierce, Rockford, IL, PA1-069A) antibodies.The relative intensity of VSVG and COPII/ Sec23 to β-actin was normalised to untreated cells.

ERAD assay
SH-SY5Y cells co-transfected with constructs encoding cyclin F and either NHK-Venus, NHK-ddVenus or SS-ddVenus for 48 h, were fixed in 4% PFA in 0.1 M PBS, pH 7.2, and mounted as above.Images were acquired using an Axio Imager Z2 fluorescent microscope at 20x/na = 0.8 magnification.At least 100 cells expressing cyclin F and Venus/ddVenus were scored as the percentage of NHK-Venus, NHK-ddVenus or SS-Venus fluorescent cells from three different experiments.

Subcellular fractionation to produce ER-rich fractions
HEK293T cells expressing cyclin F for 48 h were resuspended in 500 μl fractionation buffer, containing 20 mM HEPES (pH 7.5), 10 mM KCl, 1.5 mM MgCl 2 , 1 mM EDTA, 1 mM DTT and 0.1% protease and phosphatase inhibitor cocktail (Roche).Cells were passed 10 times through a 27 G needle and incubated at 4 °C for 20 min.The cell lysate was then centrifuged at 720g for 5 min and the supernatants, containing the cytoplasmic, membrane and mitochondrial fractions, were centrifuged at 10,000g for 5 min to pellet the mitochondrial fraction.The supernatant was then centrifuged at 100,000g for 1 h to pellet the membrane fraction and washed again by adding 200 μl of fractionation buffer, following resuspension by pipetting and passing through a 25-gauge needle.Following centrifugation again for 1 h at 100,000g, the pellet was resuspended in 100 μl RIPA buffer with 0.1% SDS and retained as the ER fraction.

Co-localization analysis between calnexin and cyclin F
Mander's coefficient was calculated for each cell to determine the degree of colocalisation (where 0 indicated no colocalisation and 1 indicated total colocalisation) of VSVG ts045 (50 + cells per group) or cyclin F (20 + cells per group) with either calnexin or GM130, using the JaCoP plugin 30  www.nature.com/scientificreports/

Quantitative analysis of cells displaying ER stress
The percentage of cells displaying nuclear immunoreactivity to CHOP or XBP1 as specific markers was quantified from 30 + primary neurons per group and at least 100 + HEK293T cells per group expressing cyclin F from n ≥ 3 independent experiments.Apoptotic cells identified by their condensed or fragmented nucleus 31 were excluded from the analysis.

Quantitative analysis of cells with Golgi fragmentation
The percentage of cells expressing cyclin F with Golgi fragmentation was quantified from 10 to 30 primary neurons per group and from at least 50 + SH-SY5Y or 100 + HEK293T cells per group from n = 3 independent experiments.Only cells where the Golgi structure was clearly visible were analysed.Normal Golgi morphology was evident by the presence of continuous stacked membranous vesicles.In contrast, fragmented Golgi were detected by the presence of disconnected elements and tubular-vesicular clusters.The Golgi was considered fragmented when at least 5 fragments were clearly visible.The area covered by the Golgi fragments was also calculated using ImageJ.Apoptotic cells identified by their condensed or fragmented nucleus 31 were excluded from the analysis.All analyses were performed blind.

Quantitative analysis of apoptotic nuclei
Apoptotic nuclei were defined as condensed when they were under 5 µm in diameter or fragmented (multiple condensed Hoechst-positive structures in one cell), following previous methods 31 .The percentage of apoptotic cells was quantified from 10 to 30 primary neurons per group or from at least 100 + HEK293T cells per group expressing cyclin F from n ≥ 3 independent experiments.Cells undergoing division were excluded from analysis.

Sytox Blue staining
HEK293T transfected with cyclin F constructs for 72 h were harvested by adding trypsin for 1 min at room temperature.The cells were then collected in PBS, centrifuged at 1200 rpm for 5 min and resuspended in 200 µl of buffer containing 10 mM HEPES, 140 mM NaCl, 2.5 mM CaCl 2 , pH 7.4.The cell suspension was treated with 1 µM Sytox Blue nucleic acid stain (Invitrogen) for 10 min at room temperature in the dark.The cell suspension was then analysed for SYTOX blue positive cells after gating for cells positive for mCherry fluorescence, using a BD FACS Canto™ II flow cytometer (BD Biosciences).

Statistics
Data are presented as mean value ± standard error of the mean (SEM).Statistical comparisons between group means were performed using GraphPad Prism 6 software (Graph Pad software, Inc.).One-way or two-way ANOVA followed by post hoc Tukey test for multiple comparisons was used when justified.The significance threshold was set at p = 0.05.Size distribution curves for ERES clusters were plotted for 10 nm bins in Airyscan resolved images and 100 nm bins for whole cell images.Binning was performed using the pivot table function in Microsoft Excel.

Ethical approval
All husbandry and experimental procedures were performed in compliance with the Animal Ethics Committee, Macquarie University, NSW, Australia (ARA 2017/030) and the Internal Biosafety Committee, Macquarie University (NLRD 5,201,401,007 and 5974-52,019,597,412,350).

ALS/FTD cyclin F S621G inhibits ER-Golgi transport
We expressed p.S621G ALS/FTD-associated cyclin F (Fig. 1a) 3 , tagged with mCherry, in two cell lines, SH-SY5Y cells and HEK293T.We previously confirmed similar expression of cyclin F S621G to cyclin F WT in two cell lines, including HEK293T 11 .SH-SY5Y cells were used in immunocytochemistry studies only, due to the low transfection efficiency of cyclin F in this cell line, where individual cells were analysed.Only transfected cells were included so that outcomes were not dependent on transfection efficiency.We first examined if cyclin F S621G inhibits ER-Golgi trafficking following widely used methods 14,15 , using a temperature sensitive mutant of vesicular stomatitis viral glycoprotein (VSVG ts045 ) 14,[32][33][34] as a reporter 35,36 (Fig. 1b).Cells co-expressing GFP-tagged VSVG ts045 and mCherry-tagged cyclin F were incubated overnight at 40 °C to accumulate misfolded VSVG ts045 in the ER.They were then incubated for 30 min at 32 °C to refold VSVG ts045 with cycloheximide (CHX) to inhibit protein synthesis, so that VSVG ts045 is released synchronously from the ER, before transport to the Golgi apparatus.VSVG ts045 transport was examined by immunocytochemistry using anti-calnexin (Fig. 1c) or anti-GM130 (Fig. 1e) antibodies to label the ER and Golgi compartments, respectively.Mander's coefficient was used to quantify the degree of colocalization between fluorophores, representing the extent of overlap between images.This was quantified in a range from 0, indicating no co-localisation to 1, representing high co-localisation.
VSVG ts045 co-localised significantly more with calnexin in cyclin F S621G expressing cells (0.6 ± 0.03), after incubation at the permissive temperature (32 °C), compared to control cells expressing either cyclin F WT (0.3 ± 0.03) or mCherry only (0.06 ± 0.009) (Fig. 1d).Hence more VSVG ts045 was retained in the ER in cyclin F S621G expressing cells compared to controls.Consistent with these findings, in cells expressing cyclin F S621G significantly less VSVG ts045 colocalised with GM130 (0.2 ± 0.02) compared to controls expressing mCherry only (0.8 ± 0.02) or cyclin F WT (0.6 ± 0.03) (Fig. 1f).Hence, VSVG ts045 extensively colocalised with GM130 in control cells, indicating efficient ER-Golgi transport, but not in cyclin F S621G cells.Additional controls from each group were also examined whereby cells remained at 40 °C incubation for 30 min after CHX treatment, instead of switching to the permissive temperature (32 °C) (Supplementary Fig. 1).As expected, VSVG ts045 was retained within the ER for each group at 40 °C and there were no differences detected in either ER or Golgi localisation (Fig. 1d, f).Inhibition of ER-Golgi transport was not caused by non-specific over-expression of protein, because efficient  (d, f) Graphs represent mean ± SEM.Symbols represent 3 independent experiments, two-way ANOVA (factors: "temperature" and "genotype") followed by post-hoc Tukey test for multiple comparisons; ***p < 0.001 vs cells expressing mCherry only, ### p < 0.001 vs cells expressing cyclin F WT .After switching to the permissive temperature (32 °C), VSVG ts045 is retained in the ER and less is transported to the Golgi apparatus, in cells expressing cyclin F S621G compared to both WT and mCherry only cells.However at 40 °C, VSVG ts045 is retained in the ER in all groups, demonstrating the specificity of the assay.
VSVG ts045 transport was detected in cells expressing mCherry (27 kDa) only, where little retention in the ER was observed (Fig. 1d, f).Moreover, we previously established that transport of VSVG ts045 was not inhibited in cells expressing wildtype versions of SOD1, TDP-43, FUS, or UBQLN2 or GFP only 14,15 .Similarly, we also previously confirmed that a mutant protein unrelated to ALS and neurodegeneration did not inhibit ER-Golgi transport of VSVG ts045 in this assay 14 : R311K Nck adaptor protein 2 37 .However, we detected a small, but significant, inhibition of transport in cells expressing cyclin F WT compared to mCherry only, but this was significantly less than cyclin F S621G .Therefore, in summary, more VSVG ts045 is retained in the ER and less is transported to the Golgi in cells expressing ALS/FTD-variant cyclin F compared to controls expressing cyclin F WT cyclin F or mCherry only, indicating perturbation of ER-Golgi transport.

ALS/FTD-associated cyclin F S621G inhibits COPII vesicular ER budding
ER-Golgi trafficking is initiated when COPII vesicles (containing protein cargo) bud from the ER membranes 18,19 .We next examined whether this initial stage of transport was inhibited in cells co-expressing cyclin F and VSVG ts045 as a cargo 14,[32][33][34] (Fig. 2).Sec31 and Sec23 were employed as specific markers of newly-budded, ERderived COPII vesicles at ERES.
First, to confirm that VSVG ts045 is transported from ER-to-Golgi in COPII vesicles, cells expressing GFPtagged VSVG ts045 were incubated overnight at 40 °C to accumulate misfolded VSVG ts045 in the ER, followed by 30 min incubation at the permissive temperature (32 °C) to allow VSVG ts045 to transport to the Golgi.Immunocytochemistry for Sec31 (as outer coat marker) revealed that VSVG ts045 was associated with Sec31-positive vesicles (Fig. 2a), consistent with its packaging into COPII vesicles exported from the ER.
We next used a modified in vitro assay 29 to examine vesicle budding from the ER (Fig. 2b).This assay examines whether VSVG ts045 -GFP is associated with membranes that form following ER retention, including those from COPII vesicles, the ER-Golgi intermediate compartment (ERGIC) and intra/post Golgi carriers.It therefore aims to reconstitute ER vesicle budding using semi-intact cells (as a source of ER), rat liver cytosol (as a source of soluble COPII) and an ATP regeneration system (Fig. 2b).ER-derived COPII vesicles are released into the buffer during incubation 29,38 , given that the integrity of the ER is preserved under these conditions 29,39,40 .Semiintact HEK293T cells co-expressing cyclin F and VSVG ts045 for 24 h were incubated at 40 °C overnight to retain VSVG ts045 in the ER, then incubated at 32 °C to allow VSVG ts045 incorporation into COPII vesicles.The light membranes, including ER-derived vesicles, were then recovered by cellular fractionation (Fig. 2b).
Control experiments revealed that if cells were incubated at 40 °C rather than the permissive temperature, VSVG ts045 was not incorporated into COPII vesicles after the budding reaction, confirming that it is retained in the ER at 40 °C (Supplementary Fig. 2a).VSVG ts045 was also examined in control lysates, sampled before the budding reaction, to examine its expression in the different cell populations.However, no significant differences in VSVG ts045 were detected between all groups (Fig. 2 c, d).Similarly, no significant differences in expression of Sec23, an inner COPII coat marker, were present amongst all cell lysates before the budding reaction (Supplementary Fig. 2b,c).Hence there are similar levels of protein expression in all the different populations, ruling this out as a possible confounding factor.
VSVG ts045 was then quantitated in budded vesicular preparations by Western blotting for VSVG (Fig. 2d,e).Significantly less VSVG ts045 was present in cells expressing cyclin F S621G compared to untransfected cells (3.4-fold) and cells expressing mCherry only (3.2-fold) (Fig. 2f).COPII was also assessed in fractions after budding using Sec23 as a marker of the vesicular interior (Supplementary Fig. 2d).Significantly less Sec23 was present in vesicles obtained from cells expressing cyclin F S621G compared to untransfected cells and mCherry only expressing cells (3-4 fold) (Supplementary Fig. 2e), consistent with less COPII vesicular transport.
Hence, less COPII and protein cargo (VSVG ts045 ) are present in ER-derived vesicles prepared from cells expressing cyclin F S621G compared to controls.This suggests that the first step of ER-Golgi trafficking, vesicle budding from the ER, is perturbed by cyclinF S621G in ALS/FTD.

ERES are perturbed in cells expressing ALS/FTD-associated cyclin F S621G
COPII-coated vesicles form at ERES, where several hundred vesicles often cluster together 41 .Hence, ERES can be distinguished from the surrounding ER membranes by the presence of COPII proteins, including Sec31.We examined the morphology of Sec31-positive clusters in HEK 293 T cells using whole cell images and observed that most clusters (84.54 ± 0.03%) displayed larger diameters (> 118 nm) (Supplementary Fig. 3a-c).However, COPII vesicles are ~ 60-90 nm in diameter 42 , implying that individual vesicles and smaller clusters were not resolved by examining the whole cell.Hence, we examined the morphology of Sec31-positive clusters in HEK 293 T cells expressing mutant cyclin F by airyscan microscopy to examine these clusters with higher resolution (Fig. 3a-c).These clusters were identified as hollow, spherical structures staining positive for Sec31, with a diameter 60-350 nm.The mean diameter of ERES clusters was significantly decreased (1.2-fold, p < 0.05) in cells expressing cyclin F S621G , compared to untransfected cells or those expressing either mCherry or cyclin F WT (Fig. 3b).Furthermore, significantly more ERES clusters with smaller diameters (90 nm) and conversely, significantly less ERES clusters with larger diameters (< 120 nm) were present in cells expressing cyclin F S621G compared to untransfected cells, and mCherry or cyclin F WT cells (Fig. 3c).This was reflected by a shift to the left when the size distribution of the diameters of Sec31-positive clusters was examined in cells expressing cyclin F S621G , compared to the other groups (Fig. 3d).These results suggest that COPII vesicular clusters at ERES are narrower and thus irregular in cells expressing variant cyclin F 621G , consistent with inhibition of ER-Golgi trafficking.

ALS/FTD-associated cyclin F S621G induces ERAD
Inhibition of the early stage of ER-Golgi transport by cyclin F S621G implies that partially-folded proteins within the ER cannot efficiently exit this compartment, leading to their accumulation and hence ER stress.This would for each group were co-transfected with VSVG ts045 and either mCherry only, cyclin F-mCherry (WT or variant), or untransfected cells as a control.VSVG ts045 was trapped in the ER by incubating cells overnight at 40 °C and ER-vesicle budding was reconstituted in vitro in perforated cells expressing mCherry only or mCherry-tagged cyclin F proteins by incubating cells at 32 °C with rat liver cytosol as a source of ER membranes and an ATP regenerating system.Budded vesicles were collected by centrifugation.(c) Western blotting of cyclin F and VSVG ts045 in lysates from untransfected cells (UT) or cells expressing mCherry only, cyclin F WT -mCherry (WT), or variant cyclin F S621G -mCherry (S621G), before budding was performed and ER-derived vesicles were isolated.β-actin was used as a loading control for protein expression.(d) Relative intensity of VSVG ts045 to β-actin (normalized to control, untransfected cells: a line parallel to the x axis set at 1) in (c).Mean ± SEM.Symbols represent independent experiments, one-way ANOVA followed by post-hoc Tukey test for multiple comparisons, ns, non-significant.No differences in expression of VSVG ts045 between groups were detected.(e) Western blotting for cyclin F and VSVG ts045 of ER-derived vesicle preparations isolated from untransfected cells (UT) and cells expressing cyclin F WT -mCherry (WT), or cyclin F S621G -mCherry (S621G), after vesicle budding was performed.β-actin was used as a loading control for protein expression.(f) Relative intensity of VSVG ts045 in blots in (e) normalised to β-actin and control, untransfected cells (a line parallel to the x axis set at 1).Mean ± SEM.Symbols represent independent experiments, one-way ANOVA followed by a post-hoc Tukey test for multiple comparisons, **p < 0.01 vs cells expressing mCherry or untransfected cells (UT).www.nature.com/scientificreports/activate ERAD, an important quality control function to degrade misfolded ER proteins 43,44 .Hence, we next examined ERAD in cyclin F cells.ERAD first involves the recognition of substrate, followed by ubiquitination, retro-translocation and then proteasomal degradation.We used a specific ERAD substrate with an ER-targeted signal sequence (K b -SS) fused to a Venus variant, ddVenus (deglycosylation-dependant Venus), where Asp is substituted to Asn at position 82 25 (Fig. 4a).This mutation results in glycosylation and reduced fluorescence, which is restored to wildtype levels when Asn is converted back to Asp.Removal of oligosaccharides by endogenous peptide:N′glycanase (PNGase) in the cytosol results in deamidation of glycosylated Asn, converting it to Asp which restores fluorescence 25 (Fig. 4b).As ERAD involves two stages-entry of substrate into the ER, followed by retro-translocation to the cytoplasm-both glycosylation of ddVenus in the ER and deglycosylation in the cytosol are required for fluorescence.Hence accumulation of ddVenus indicates ERAD activation.In addition, a second ERAD substrate was used, fluorescent ddVenus fused to the null Hong Kong genetic variant of α1-antitrypsin (NHK-ddVenus), which misfolds terminally in the ER and is also degraded specifically by ERAD 24,25,45 .In cells expressing mCherry only, cyclin F WT or cyclin F S621G , ERAD was probed using substrates NHK-Venus (control for transfection efficiency), NHK-ddVenus or SS-ddVenus (Fig. 4c).Quantification demonstrated that significantly more fluorescent cells were present in cyclin F S621G expressing populations compared to cyclin F WT (NHK-ddVenus: 1.7-fold; SS-ddVenus: 1.5-fold,) or mCherry only (NHK-ddVenus: 33.1-fold; SS-ddVenus: 33.8-fold) (Fig. 4d-f).Hence, expression of cyclin F S621G activates ERAD compared to cyclin F WT and controls.

ALS/FTD cyclin F S621G induces ER stress in neuronal cells, primary neurons and iPSC-derived motor neurons
These results imply that cyclin F S621G induces ER stress.To investigate this possibility further, we examined whether cyclin F S621G activates UPR following established methods [46][47][48] (Fig. 5).During ER stress, misfolded proteins accumulate in the ER, which activates and phosphorylates UPR sensors, including PERK and IRE1.The latter activates transcription factor XBP1, which becomes spliced in the nucleus (forming spliced XBP1, sXBP1) (Fig. 5a).Following chronic ER stress, transcription factor CHOP is activated only when the UPR becomes proapoptotic 49 .Both XBP-1 and CHOP translocate to the nucleus when activated during ER stress.Hence, splicing of XBP-1 and nuclear immunoreactivity to both XBP-1 and CHOP therefore indicates activation of pro-apoptotic phases of the UPR (Fig. 5a).Expression of cyclin F S621G significantly increased the proportion of cells with nuclear XBP1 immunoreactivity, compared to cells expressing cyclin F WT (twofold), mCherry only (5.9-fold) or untransfected cells (173-fold) (Fig. 5b,c), implying that cyclin F S621G induces ER stress.This was confirmed following immunocytochemistry for CHOP.Significantly more cells with nuclear CHOP immunoreactivity were detected in populations expressing variant cyclin F 621G compared to cells expressing cyclin F WT (1.7-fold), mCherry only (5.9-fold) or untransfected cells (31.4-fold; Fig. 5d,e).The proportion of untransfected cells and mCherry-expressing cells with XBP1 or CHOP immunoreactivity was not significantly different (Fig. 5d,e), confirming that ER stress was not induced by non-specific protein overexpression.Compared to mCherry, expression of cyclin F WT resulted in significantly more cells with nuclear immunoreactivity to CHOP, but not to XBP1, implying that wildtype cyclin F also induces low levels of ER stress, albeit significantly less than the mutant.These results indicate that ALS/FTD-associated cyclin F S621G induces ER stress in neuronal cell lines.This finding raises the question of whether ER stress is induced directly, by the presence of cyclin F S621G in the ER, or indirectly, by cyclin F S621G localised in the cytoplasm.Dysregulation of ER-Golgi trafficking can induce ER stress by cytoplasm-located proteins 50 , including those associated with ALS 14 .Hence, we next examined whether cyclin F S621G is present in the ER, which would imply it directly triggers ER stress.Subcellular fractionation of lysates to produce ER-rich fractions was performed.Western blotting for calnexin confirmed the presence of ER membranes (Supplementary Fig. 4a) and blotting for cyclin F revealed similar levels of both cyclin F WT and cyclin F S621G in the ER fractions (Supplementary Fig. 4b).To further confirm these findings, immunocytochemistry for calnexin was performed.Analysis using Mander's coefficient (where 0 = no co-localisation and 1 = high co-localisation) demonstrated little co-localisation between wildtype (0.3) or cyclin F S621G and calnexin (0.2). Furthermore, no significant differences were detected between localisation of calnexin with either cyclin F WT or cyclin F S621G (Supplementary Fig. 4c,d), despite differences in induction of ER stress by these proteins.Hence, these results imply that ALS/FTD cyclin F S621G does not induce ER stress by excess accumulation of cyclin F in the ER.
To confirm the findings obtained in cell lines, ER stress was next examined in mouse primary cortical neurons expressing cyclin F S621G by immunocytochemistry (Fig. 5f).Quantification demonstrated that significantly more primary neurons expressing cyclin F S621G displayed nuclear immunoreactivity to CHOP (1.7-fold), compared to cyclin F WT or mCherry only cells (Fig. 5f,g).Similarly, nuclear localisation of XBP-1 was detected in primary neurons expressing cyclin F S621G unlike controls (Supplementary Fig. S5), although quantification could not be performed due to low transfection efficiency.Hence, cyclin F S621G activates ER stress in primary cortical neurons, confirming the results obtained in cell lines.
Finally, we also examined iPSC-derived motor neurons from an ALS patient carrying the cyclin F S621G mutation 27 and the equivalent CRISPR-corrected isogenic control cells (cyclin F wild-type), after 12 days of maturation.Cell lysates from these cells were subjected to western blotting for XBP1, revealing that expression of spliced XBP-1 was significantly increased in iPSC-derived motor neurons expressing cyclin F S621G compared to the isogenic control (Fig. 5h,i, 1.8-fold).Hence these findings provide further evidence that cyclin F S621G activates ER stress.

ALS/FTD cyclin F S621G induces fragmentation of the Golgi
Golgi fragmentation follows ER stress 51,52 , and inhibition of ER-Golgi trafficking 14,[53][54][55] , raising the possibility that cyclin F S621G also induces this phenotype.Hence, next the morphology of the Golgi apparatus was examined by immunocytochemistry for GM130 as a marker.In SH-SY5Y (Fig. 6a) and HEK293T (Supplementary Fig. 6a, b) cells expressing mCherry only or cyclin F WT , the Golgi displayed its typical morphology of continuous, stacked membranous cisternae.However, in cyclin F S621G expressing HEK293T cells and SH-SY5Y cells, the Golgi apparatus was fragmented, displaying multiple disconnected elements or tubular-vesicular clusters 56 .Quantification revealed that significantly more cells with fragmented Golgi were present in populations expressing cyclin F S621G (HEK293T: 1.5-fold; SH-SY5Y: twofold) compared to cyclin F WT , mCherry or untransfected cells (Fig. 6b, Supplementary Fig. 6).
To provide evidence that Golgi fragmentation correlates with impairment of ER-Golgi transport, we next examined Golgi morphology in the GM130-immunostained SH-SY5Y cells used in the VSVG ts045 assay (Fig. 1).Quantification revealed that significantly more cyclin F S621G cells displayed fragmented Golgi compared to cyclin F WT (2.6-fold) or mCherry-cells (13.2-fold change, Supplementary Fig. 7a).The proportion of cells co-expressing VSVG ts045 and cyclin F with Golgi fragmentation (57.3 ± 8.6%) was also similar to results obtained in Fig. 6b (64.6 ± 1%).Hence Golgi fragmentation is present in cells in which VSVG ts045 secretion is impaired, confirming that inhibition of ER-Golgi transport correlates with fragmentation of the Golgi.www.nature.com/scientificreports/To further confirm these results, a more unbiased quantification method was used.Golgi stacks can be dispersed (mini-stacks) or completely disassembled, hence the surface area covered by the Golgi also indicates fragmentation 51 .Quantification revealed a significant increase in Golgi fragment surface area in cells expressing cyclin F S621G (1.9-fold) compared to cyclin F WT and control untransfected or mCherry cells (Fig. 6c).Similarly, we also examined the GM130-immunostained cells prepared for the VSVG ts045 assay (Fig. 1).Quantification revealed that a significant increase in the area covered by fragmented Golgi in cells expressing cyclin F S621G compared to cyclin F WT (2.6-fold) or mCherry cells (13.5-fold, Supplementary Fig. 7b) (57.4%), similar to Fig. 6 (64.7%).Together these results demonstrate that cyclin F S621G induces Golgi fragmentation in neuronal cells, which correlates with inhibition of ER-Golgi transport.
To further confirm the above results, we next examined mouse cortical primary neurons expressing cyclin F. Immunocytochemistry was performed for GM130 to examine the morphology of the Golgi as above (Fig. 6d).Significantly more neurons expressing cyclin F S621G displayed Golgi fragmentation compared to those expressing WT (2.2-fold) and mCherry (6.3-fold) (Fig. 6e).Hence, ALS/FTD-associated cyclin F S621G induces Golgi fragmentation in primary cortical neurons, confirming the results obtained in cell lines.

ALS/FTD mutant cyclin F induces cell death
ER stress induces apoptosis when prolonged/severe, and Golgi fragmentation also triggers apoptosis 57,58 .Hence cell death was next analysed by flow cytometry following Sytox Blue staining, in SH-SY5Y cells expressing cyclin F (Fig. 7a,b).Quantitative analysis demonstrated significantly more dead cells in populations expressing cyclin F S621G compared to untransfected cells (3.7-fold) and those expressing cyclin F WT (1.4-fold) or mCherry (3.8fold), indicating that cyclin F S621G expression induces cell death.
This was next examined in primary neurons, where the presence of condensed nuclear morphology indicated induction of apoptosis, as previous 46,59 .Quantitative analysis of primary neurons expressing cyclin F demonstrated that significantly more neurons were undergoing apoptosis in populations expressing cyclin F S621G compared to wildtype (1.9-fold) or mCherry (cyclin F S621G : 7.9-fold) (Fig. 7c,d).Hence, these data confirm that ALS/FTD-cyclin F S621G induces apoptosis.
Finally, we also examined a second ALS/FTD-associated variant, p.S195R, to probe whether ER/Golgi perturbations are specific to cyclin F S621G only.Golgi morphology was examined as a marker of trafficking dysfunction in SHSY5Y cells using GM130.Significantly more Golgi fragmentation (1.6-fold) was detected when cyclin F S195R was expressed, compared to cyclin F WT , mCherry and untransfected cells (Supplementary Fig. 8a-c).Similarly, significantly more primary neurons expressing cyclin F S195R displayed Golgi fragmentation compared to those expressing WT (2.9-fold) and mCherry (8.1-fold) (Supplementary Fig. 8d-e).Hence the Golgi is fragmented by ALS-FTD-associated mutant F S195R .Similarly, significantly more primary neurons expressing cyclin F S195R displayed nuclear immunoreactivity to CHOP (2.2-fold), compared to cyclin F WT or mCherry only expressing cells (Supplementary Fig. 8 f-g), indicating activation of ER stress.Consistent with this result, significantly more neurons bearing apoptotic nuclei were present in cyclin F S195R populations compared to cyclin F WT (1.8-fold) or mCherry (7.3-fold) (Supplementary Fig. 8h-i).There were also similar levels of cyclin F S195R in the ER-rich fractions compared to both cyclin F WT or cyclin F S621G (Supplementary Fig. 4a,b).Hence these findings imply that the second ALS/FTD variant also perturbs ER-Golgi homeostasis.were excluded.Mean ± SEM; n = 100 + cells/group.One-way ANOVA, post-hoc Tukey test; ***p < 0.001 vs untransfected (UT) and cells expressing mCherry; ### p < 0.001 vs cells expressing cyclin F WT .(f) Fluorescence microscopy of mouse primary cortical neurons expressing mCherry only, or mCherry-tagged cyclin F, following immunocytochemistry for CHOP and Hoechst.Nuclei: white outline.Arrows: nuclear CHOP immunoreactivity.Scale bar = 10 μm.(g) Proportion of neurons with nuclear CHOP in (f).Neurons undergoing apoptosis (with condensed nuclei) were excluded.Mean ± SEM; n = 10-30 neurons/group; one-way ANOVA, post-hoc Tukey test.ns, non-significant (WT vs mCherry), ***p < 0.001 vs mCherry only; # p < 0.05 vs cells expressing cyclin F WT .(h) Western blotting of sXBP1 in lysates from iPSC-derived motor neurons from a cyclin F S621G patient and CRISPR-corrected isogenic control, GAPDH as loading control.uXBP1, unspliced XBP1 (i) sXBP1 expression from (h) relative to GAPDH.Three iPSC differentiation replicates for both cyclin F S621G and isogenic control were analysed on the same blot, with two technical replicate blots.The mean sXBP1 intensity in cyclin F S621G relative to GAPDH per blot is shown, all samples were normalised to the first differentiation isogenic control replicate.Mean ± SD; paired t-test, *p < 0.05.

Discussion
This study demonstrates that ALS/FTD-associated cyclin F S621G perturbs ER homeostasis by inhibiting an important ER function: transport of secretory/transmembrane proteins to the Golgi.This induces Golgi fragmentation, ER stress, ERAD, and apoptosis, in both cell lines and mouse cortical primary neurons.Together these data imply that ER dysfunction is important in neurodegeneration induced by cyclin F S621G , thus providing novel insights into CCNF-associated ALS/FTD.ER-Golgi transport is a vital gateway to the endomembrane system, and one-third of all proteins transit via this pathway before reaching their final locations 18,60 .Our results imply that ALS/FTD-associated cyclin F S621G perturbs the first stage of ER-Golgi transport: budding of COPII vesicles from the ER.The COPII coat is essential for the formation of transport vesicles on the cytosolic face of the ER membrane.Curvature of the ER membrane, concentration of cargo and vesicular release then results.Defective COPII vesicles are known to inhibit secretion 20 .Here we detected several abnormalities in COPII in cyclin F S621G cells, less vesicular cargo (VSVG ts045 ) and narrower Sec31-positive clusters, demonstrating that the ERES are abnormal.Typical COPII vesicles are 60-90 nm in diameter, and it was not possible to resolve individual COPII vesicles or ERES using the methods used here.The vesicle clusters were found to contain smaller diameters in mutant cyclin F expressing cells, implying that they are irregular and possibly misshapen.Furthermore, whilst Sec23 levels correlate with the load of protein cargo 20 , here the levels of Sec23 were comparable to cargo load, suggesting that the defect is related to COPII vesicles themselves, rather than the mechanisms involving incorporation of protein cargo.expression of cyclin F S621G and cyclin WT in the ER by both immunocytochemistry and subcellular fractionation, raising the question of how ER stress is induced in ALS/FTD.As ALS-cyclin F mislocalizes to the cytoplasm, where it promotes the cytoplasmic aggregation of TDP-43 87 , it is possible that cyclin F S621G induces ER stress from the cytoplasm by inhibiting ER-Golgi transport.However, we cannot rule out the possibility that cyclin F S621G directly induces ER stress from the ER itself.Given that the cytosol and ER are closely associated, it is hard to conclusively determine whether a fraction of a cytosolic protein is bound to the ER membrane using immunocytochemistry.Furthermore, the ER-rich fractions isolated displayed weak reactivity to GAPDH (Supplementary Fig. 3a), indicating possible low levels of contamination from the cytoplasm.Hence further studies are required to confirm the contribution of both cytosolic and ER-localised to induction of ER stress by mutant cyclin F.
Golgi fragmentation is also well-described in ALS 23,56,88 , resulting in morphological changes and disruption of its characteristic ribbon-like structure 88,89 .Fragmentation of the Golgi apparatus is present in sporadic ALS patients 90 and before disease onset in SOD1 G93A mice 91 , prior to neuromuscular denervation and axon retraction 92 .It is also a feature of other neurodegenerative diseases 88 , but it has not been previously described for cyclin F in ALS.Here, we demonstrate that ALS/FTD-associated cyclin F variants trigger Golgi fragmentation.Proper organization of the Golgi depends on efficient bidirectional ER-Golgi vesicle transport 89,93 and both membrane flow and cargo load influence its structure and function.Hence, blocking export of cargo-containing ER carriers 94,95 or depleting cargo receptors 96 results in Golgi fragmentation.Also the tubulo-vesicular Golgi clusters can further fuse with the ER, increasing ER stress 97 .In addition, inhibition of intra-Golgi trafficking or vesicle transport from the Golgi to the plasma membrane induces Golgi fragmentation if prolonged [53][54][55] .Hence, in this study, it is possible that the inhibition of ER-Golgi transport by cyclin F S621G triggers Golgi fragmentation and ER stress.However, other stressors can induce Golgi fragmentation, so the directionality of these links cannot be conclusively established.
We also examined a second ALS/FTD variant in this study.In neuronal cells or primary neurons expressing cyclin F S195R , Golgi fragmentation and ER stress, leading to apoptosis, was detected.This implies that the trafficking perturbations induced by cyclin F S621G are shared by other variants in ALS.However further studies are required to confirm this possibility.
In summary, this study identifies novel cellular mechanisms triggered by ALS/FTD-associated variant cyclin F S621G .In Fig. 8, we provide one hypothetical model to illustrate how these events are triggered.However, it is also possible that ER stress or Golgi fragmentation is the upstream trigger, which would subsequently impair ER-Golgi transport, given these events are all closely related.Alternatively, it is possible that the cellular events detected in this study result from a combination of these defects.Further studies are therefore required to probe the directionality of these links.

Figure 1 .
Figure 1.ALS/FTD cyclin F S621G inhibits ER to Golgi trafficking.(a) Schematic diagram illustrating domain structure of cyclin F showing the location of the p.S621G mutation in the PEST domain.NLS, nuclear localisation signal.(b) Experimental paradigm: cells expressing VSVG ts045 for 24 h are incubated at 40°C overnight (ON) to misfold VSVG ts045 , leading to its accumulation in the ER.Cells are then incubated for 30 min in cycloheximide (CHX) to inhibit further VSVG ts045 protein synthesis to synchronously release it from the ER before transport to the Golgi.When cells are placed at the permissive temperature (32 °C), VSVG ts045 refolds and transits to the Golgi apparatus within 30 min.(c) Confocal microscopy images following immunocytochemistry for calnexin (as an ER marker) in SH-SY5Y cells co-expressing GFP-tagged VSVG ts045 and either mCherry only (mCherry), cyclin F WT-mCherry (WT) or variant cyclin F S621G mCherry (S621G), after incubation at 40 °C overnight and 32 °C for 30 min as in (b).Scale bar = 5 µm.(d) The degree of co-localisation of VSVG ts045 with calnexin from the images in (c) was quantified using Mander's coefficient.(e) Confocal microscopy images following immunocytochemistry for GM130 (as a Golgi marker) in SH-SY5Y cells co-expressing GFPtagged VSVG ts045 and either mCherry only (mCherry), cyclin F WT-mCherry (WT) or cyclin F S621G mCherry (S621G), after incubation at 40 °C overnight and 32 °C for 30 min as in (b).(f) The degree of colocalisation of VSVG ts045 with GM130 from the images in (c) was quantified using Mander's coefficient.(d,f) Graphs represent mean ± SEM.Symbols represent 3 independent experiments, two-way ANOVA (factors: "temperature" and "genotype") followed by post-hoc Tukey test for multiple comparisons; ***p < 0.001 vs cells expressing mCherry only, ### p < 0.001 vs cells expressing cyclin F WT .After switching to the permissive temperature (32 °C), VSVG ts045 is retained in the ER and less is transported to the Golgi apparatus, in cells expressing cyclin F S621G compared to both WT and mCherry only cells.However at 40 °C, VSVG ts045 is retained in the ER in all groups, demonstrating the specificity of the assay.

Figure 2 .
Figure 2. ALS/FTD cyclin F S621G inhibits the load of protein cargo (VSVG ts045 ) into COPII vesicles.(a) Immunocytochemistry for Sec31 in SH-SY5Y cells expressing GFP-tagged VSVG ts045 , after incubation at 40 °C overnight and 32 °C for 30 min.Upper panel: Low magnification images of cells immunostained for Sec31; region enlarged in the lower panel is indicated by box shown in the inset.Lower panel: higher magnification image of inset, illustrating the presence of VSVG ts045 in Sec31-positive vesicles.Scale bar = 1 µm [upper panel] or 0.5 µm [lower panel].VSVG ts045 co-localises with Sec31 (arrows), and hence COPII vesicles.(b) Schematic diagram depicting the experimental procedure used for cell-free ER vesicle budding reactions.The same number and volume of HEK293T cellsfor each group were co-transfected with VSVG ts045 and either mCherry only, cyclin F-mCherry (WT or variant), or untransfected cells as a control.VSVG ts045 was trapped in the ER by incubating cells overnight at 40 °C and ER-vesicle budding was reconstituted in vitro in perforated cells expressing mCherry only or mCherry-tagged cyclin F proteins by incubating cells at 32 °C with rat liver cytosol as a source of ER membranes and an ATP regenerating system.Budded vesicles were collected by centrifugation.(c) Western blotting of cyclin F and VSVG ts045 in lysates from untransfected cells (UT) or cells expressing mCherry only, cyclin F WT -mCherry (WT), or variant cyclin F S621G -mCherry (S621G), before budding was performed and ER-derived vesicles were isolated.β-actin was used as a loading control for protein expression.(d) Relative intensity of VSVG ts045 to β-actin (normalized to control, untransfected cells: a line parallel to the x axis set at 1) in (c).Mean ± SEM.Symbols represent independent experiments, one-way ANOVA followed by post-hoc Tukey test for multiple comparisons, ns, non-significant.No differences in expression of VSVG ts045 between groups were detected.(e) Western blotting for cyclin F and VSVG ts045 of ER-derived vesicle preparations isolated from untransfected cells (UT) and cells expressing cyclin F WT -mCherry (WT), or cyclin F S621G -mCherry (S621G), after vesicle budding was performed.β-actin was used as a loading control for protein expression.(f) Relative intensity of VSVG ts045 in blots in (e) normalised to β-actin and control, untransfected cells (a line parallel to the x axis set at 1).Mean ± SEM.Symbols represent independent experiments, one-way ANOVA followed by a post-hoc Tukey test for multiple comparisons, **p < 0.01 vs cells expressing mCherry or untransfected cells (UT).

Figure 3 .
Figure 3. ALS/FTD cyclin F S621G perturbs ER-exit sites.(a) Fluorescent confocal airyscan microscopy images of HEK293T cells following immunocytochemistry for Sec31 and Hoechst staining in untransfected cells (UT) or cells expressing mCherry only, mCherry-tagged cyclin F WT or mutant cyclin F S621G for 48h.Left and middle columns: scale bar = 5 µm.Right column: Higher magnification images acquired with Airyscan of the area delimited by white squares on images on middle column: scale bar = 0.5 µm 98 .As ERES generate COPII-coated vesicles with an average diameter of 60-90 nm 42 , each green dot represents clusters of Sec31positive ERES, rather than a single Sec31-positive COPII vesicle.(b) Diameter of ERES clusters calculated from images obtained in (a).For mCherry, cyclin F WT and cyclin F S621G expressing cells, symbols represent individual experiments.ERES cluster diameter was calculated from n = 30 cells.Mean ± SEM.Symbols represent independent experiments; one-way ANOVA followed by a post-hoc Tukey test for multiple comparisons, *p < 0.05 vs untransfected cells and cells expressing either mCherry only or cyclinF WT .(c) Proportion of HEK293T cells with diameters of ERES clusters below 90 nm; between 90 and 120 nm; and over 120 nm.Mean ± SEM; symbols represent n = 3 independent experiments, **p < 0.01 vs untransfected cells (UT), mCherry only and cyclin F WT cells, one-way ANOVA, followed by post-hoc Tukey test.The total number of ERES clusters quantified in each group across the three replicates was UT: n = 1827; mCherry: n = 2114; cyclin F WT : n = 1722; cyclin F S621G : n = 1746.(d) Size distribution curve of Sec31-positive ERES clusters in untransfected HEK293 cells and cells expressing mCherry or cyclin F. n = 3 independent experiments, 380 + ERES clusters per group.

Figure 4 .
Figure 4. ALS/FTD variant cyclin F S621G activates ERAD.(a) Fluorescent ERAD substrates used to monitor ERAD.Wild-type Venus and mutant ddVenus, in which Asn is substituted for Asp at position 82 (SS-ddVenus), were fused to a well-known ERAD substrate, the unstable null Hong Kong variant of alpha-1-antitrypsin (NHK-Venus; NHK-ddVenus) 25 .SS-ddVenus contains ER-targeted signal sequence (K b -SS) fused to dd-Venus.(b) The system is based on glycosylation and deglycosylation of Venus.Removal of a glycan by the cytosolic enzyme Peptide:N′glycanase (PNGase) results in deamidation of the glycosylated asparagine (N), converting it to an aspartic acid (D) residue, restoring fluorescence.The fluorescence results from protein that both reaches the ER and also retrotranslocates to the cytosol.Both glycosylation and deglycosylation (ER entry and exit) are required for fluorescence 25 .(c) Representative fluorescent microscopy merge images of SH-SY5Y cells co-expressing ERAD substrates (green) with either mCherry, cyclin F WT or cyclin F S621G (red).Arrows represent ddVenus fluorescent cells expressing cyclin F S621G .NHK fused to Venus was used as a control for cell transfection.Scale bar = 10 µm.(d-f) Graphs illustrate the proportion of cells co-expressing cyclin F with NHK-Venus (d), NHK-ddVenus (e) or SS-ddVenus (f) relative to the total number of mCherry-fluorescent cells.Graphs represent mean ± SEM.Symbols represent three independent experiments, ***p < 0.001 vs mCherry only; # p < 0.05, ### p < 0.001 vs cyclin F WT , One-way ANOVA followed by post-hoc Tukey test for multiple comparisons. https://doi.org/10.1038/s41598-023-46802-9

◂Figure 6 .
Figure 6.ALS/FTD cyclin F S621G induces Golgi fragmentation.(a) Fluorescent confocal microscopy images following immunocytochemistry for GM130 and Hoechst staining, of SH-SY5Y cells expressing mCherry only or mCherry-tagged cyclin F WT or variant S621G.Arrows: fragmented Golgi.Apoptotic cells displaying a condensed nucleus were excluded from the analysis.Scale bar = 10 µm.(b) The proportion of cells with fragmented Golgi in (a) was quantified.Apoptotic cells displaying a condensed nucleus were excluded from analysis.Mean ± SEM.Symbols represent 3 independent experiments, n = 50 + cells per group from, one-way ANOVA, post-hoc Tukey test for multiple comparisons; **p < 0.01 and ***p < 0.001 vs untransfected cells (UT) and mCherry only; ## p < 0.01 vs cells expressing cyclin F WT .(c) The area covered by fragmented Golgi was quantified per cell using the images in (a).Mean ± SEM.Symbols represent 3 independent experiments, n = 50 + cells per group one-way ANOVA followed by a post-hoc Tukey test for multiple comparisons; ns, nonsignificant (WT vs UT and mCherry), ***p < 0.001 vs UT and mCherry; ### p < 0.01 vs cells expressing cyclin F WT .(d) Fluorescent confocal microscopy images of following immunocytochemistry for GM130 and Hoechst staining in mouse primary cortical neurons expressing mCherry only or cyclin F (WT or S621G).Arrows: fragmented Golgi.The dashed white line delimits the outline of the neuron.Scale bar = 5 µm.(e) The proportion of primary neurons with fragmented Golgi was quantified.Apoptotic neurons displaying a condensed nucleus were excluded from the analysis and thus are not shown here.Mean ± SEM.Symbols represent 3 independent experiments, n = 10-30 neurons per group; one-way ANOVA followed by a post-hoc Tukey test for multiple comparisons; ns, non-significant (WT vs UT), **p < 0.01 vs mCherry; # p < 0.05 vs cells expressing cyclin F WT .
Images acquired by confocal microscopy under × 100 objective were used for analysis of Sec31-positive clusters in whole cells.Individual cells were outlined by the freehand selection tool in ImageJ, converted to RGB images and colour threshold level set to 40-255.The ' Analyze particles' feature of ImageJ was then used to obtain diameter of selected clusters.Eight to ten individual cells per group were analyzed.A size distribution curve was plotted for 100 nm bins using the pivot table function in Microsoft Excel.