Spermidine Recovers the Autophagy Defects Underlying the Pathophysiology of Cell Trafficking Disorders

ABSTRACT

Cell trafficking alterations are a growing group of disorders and one of the largest categories of Inherited Metabolic Diseases. They have complex and variable clinical presentation. Regarding neurological manifestations they can present a wide repertoire of symptoms ranging from neurodevelopmental to neurodegnerative disorders. The study of monogenic cell trafficking diseases draws an scenario to understanding this complex group of disorders and to find new therapeutic avenues. Within their pathophysiology, alterations in autophagy outstand as a targetable mechanism of disease, ammended to be modulated through different strategies. In this work we have studied the pathophysiology of two cell trafficking disorders due to mutations in SYNJ1 and NBAS genes. Specifically, we have assesed the autophagic flux in primary fibroblast cultures of the patients and gender/age‐matched controls and whether it could be address with a therapeutic purpose. The results have shaped autophagy as one of the hallmarks of the disease. Moreover, we tested in vitro the effect of spermidine, a natural polyamine that acts as an autopagy inductor. Due to the positive results, its efficacy was evaluated later on the patients as well, in a series of n‐of‐1 clinical trials, achieving improvement in some clinical aspects related to motricity and cognition. Defining autophagy alterations as a common feature in the pathophysiology of cell trafficking disorders is a great step towards their treatment, as it represents a potential actionable target for the personalized treatement of these disorders.

1. Introduction

Cell trafficking is the biological process for the exchange of signals and metabolites between cell compartments and the outside of the cell, including four primary mechanisms: membrane trafficking (trafficking of molecules in vesicle‐bound compartments), autophagy, cytoskeletal transport and membrane contact sites. The number of monogenic diseases of cell trafficking has increased over 7 folds in the last 10 years [1], forming the largest group in the recent International Classification of Inherited Metabolic Disorders (ICIMD) [2]. Despite the importance of trafficking on cellular homeostasis and its impact on health and disease, cell trafficking disorders pose a significant challenge in both diagnostic and treatment, partially due to their heterogeneous clinical presentations and complex underlying pathophysiological mechanisms.

Among cellular trafficking mechanisms, autophagy outstands as a central process. Shortly, it starts with the signaling of a molecule or cellular component to be degraded in the cytoplasm, followed by the nucleation of a phagophore around the cargo. The membrane of the phagophore then elongates, forming a double‐membraned vesicle named autophagosome. Upon completion, autophagosomes undergo maturation through fusion with lysosomes, forming autolysosomes, where the cargo is degraded by lysosomal hydrolases. This sequence of events results in the efficient removal of cytoplasmic components and damaged proteins, outstanding as a crucial cellular mechanism for maintaining homeostasis and responding to stress conditions [3].

Mechanisms of cell trafficking, as most cellular activities, are not independent buy highly interconnected. As a result of this, alterations in autophagy can underlie in the pathophysiology of other cell traffic disorders, as those related to ER‐Golgi anterograde and retrograde transport processes, as recently reviewed [4]. The involvement of autophagy as part of the pathophysiological landscape is observed in diseases across all four groups of trafficking‐related disorders. Moreover, defects in autophagy have been described within several non‐metabolic diseases as neurodegenerative disorders or cancer. The implication of autophagy in disease progression, together with the knowledge on this process modulation, underscores the importance of its study. Autophagy's involvement in cellular trafficking disorders represents an opportunity to unveil novel strategies for their treatment.

In this work, we have evaluated both in vitro and in vivo the potential therapeutic impact of spermidine supplementation in two specific cellular trafficking disorders, SYNJ1 and NBAS deficiencies. While the first protein is directly involved in autophagy along with functions in endocytosis, NBAS function has been related to retrograde transport, showcasing a different cell trafficking scenario. Our results confirm the implication of autophagy in the pathophysiology of these cell trafficking diseases and underscore spermidine's potential in the management of these disorders. Spermidine, a natural polyamine present in cells, represents a safe‐to‐use alternative for autophagy modulation. Defining autophagy alterations as a common feature in the pathophysiology of unrelated cell trafficking disorders is a great step towards their treatment, as it represents a potential actionable target for the personalized treatement of these disorders.

2. Patients, Materials and Methods

2.1. Patients

2.1.1. Patient 1: Patient^SYNJ1

Patient presenting an early‐onset encephalopathy associated with epilepsy, microcephaly and a hyperkinetic movement disorder, corresponding to choreo‐dystonia. Genetic analysis confirmed the homozygous mutation SYNJ1:c.2537G > A (p.Arg846His). This boy presented an uneventful pregnancy and perinatal period, born full term with a normal birth weight (3360 Kg). Delayed psychomotor development with axial hypotonia were observed during the first months. By 5 months of age, he had his first seizure, that was a focal seizure with oculocephalic deviation. By 6 months, he was admitted to our center for study due to neurodevelopmental regression. Several studies were performed, as metabolic work‐up and brain MRI that were normal. The EEG showed bilateral parieto‐occipital epileptiform abnormalities, and treatment with valproic acid was started. By 8 months, he presented with epileptic spasms. His EEG showed slowed and poorly organized brain activity, with multifocal paroxysms. He started treatment with vigabatrin and later ACTH and levetiracetam, with transient response; however, neurodevelopment was stagnated. In later months, due to new crises, the therapeutic scheme was changed, adding zonisamide and topiramate. In the last assessments he maintained treatment with the combination of lamotrigine and levetiracetam. The patient is currently 7 years old, he has no verbal communication and the connection is poor. He presents hand stereotypes and severe drooling, for which glycopyrrolate has been started. He also presents a severe hypotonia with choreodystonic movements of trunk and extremities, and myoclonus, with not prominent pyramidal signs. From age 4 to 7, multiple admissions due to epileptic seizures were required, and seizures persisted with poor response to antiepileptic drugs, although seizures were less frequent than in the first years. The predominant clinical features, besides epilepsy, were the hyperkinetic movement disorder and the severe intellectual disability.

2.1.2. Patient 2: Patient^NBAS

The patient was first seen at our center at the age of 6 years with a diagnosis of leukoencephalopathy and SOPH syndrome (short stature; optic atrophy; Pelger‐Huet anomaly) due to NBAS mutations: NBAS: c.451G > A (p.Glu151Lys) and c.6840G > A (p.Thr=; described as splicing mutation [5]), diagnosed in another hospital. She presented with intrauterine growth retardation, delivery at 37 weeks gestation, birth weight 1956 g (2nd percentile) and a head circumference of 32 cm. The patient's clinical presentation aligns with the phenotype described for mutations in both the β‐propeller domain and the C‐terminal region of the protein [6], illustrating a mixed genotype–phenotype correlation. Other history of interest included the following: ‐ hypoglycaemia at birth which resolved spontaneously after oral intake. At 10 days of life, she was readmitted for fever and failure to thrive and sustained hypoglycaemia that required feeding by NGUS. ‐Neonatal seizures treated with phenobarbital for 3 months. At this time, cerebral white matter alterations were detected. Reappearance of focal seizures were treated with Levetiracetam. She is free of seizures since the age of 4. ‐ Vision loss began at the age of 2 years with progression until practically blindness at the age of 4 years up to the present day. ‐ Psychomotor development: appropriate for her age. No cognitive alterations. Adequate language. ‐ Low weight gain and short stature. ‐Intermittent C1‐C2 rotational luxation since the age of 3 years. ‐ Persistent hypertransaminemia and hypogammaglobulinemia. ‐ Leukoencephalopathy, encephalomyelitis with contrast‐enhancing inflammatory component, recurrent triggered by fever (and also with no recognizable trigger) leading to paraparesis. Treatment with monthly immunoglobulins, mycophenolate and methylprednisolone during leukoencephalopathy outbreaks.

2.2. Cell Culturing

Skin biopsy was taken from each patient to obtain primary cultures of fibroblasts, along with gender/age‐matched controls. Fibroblasts were cultured following standard culture conditions in Dulbecco's Modified Eagles' Medium high‐glucose (DMEM) supplemented with 10% fetal bovine serum (FBS), 1% L‐Glutamine and 1% penicillin–streptomycin at 37°C in a 5% CO₂ incubator. Unless stated otherwise, experiments were accomplished when fibroblasts were at 80% confluence and between the 7th and 14th passages.

For autophagy inhibition Bafilomycin A1 (BafA1; Sigma‐Aldrich, 10900010) was used at 200 nM in complete DMEM for 4 h at 37°C. For autophagy induction, fibroblasts were cultured in Earle's Balanced Salt Solution (EBSS; Sigma‐Aldrich, E2888) for 4 h at 37°C. Before incubating these modulators, the fibroblasts were washed twice with phosphate buffer saline (PBS).

Spermidine (Sigma‐Aldrich, S0266) is a natural polyamine and was used as a stimulator cytoprotective autophagy. This nutracient was resuspended in 100% dimethyl sulfoxide (DMSO) to obtain a stock solution at 1 M. We prepared an intermediate solution at 1 mM diluted in mili‐Q water and stored at −20°C for less than 1 month. The working solution is 1 μM in supplemented DMEM for 24 h.

2.3. Immunofluorescence

Fibroblasts were seeded onto glass coverslip at 60% confluence for 24 h, washed with PBS and fixed with 4% paraformaldehyde (PFA) for 20 min at room temperature. Cells were permeabilized with PBS 0.1% Triton X‐100 for 30 min and they were blocked with PBS 10% FBS for 1 h at room temperature. The specific primary antibody α‐p62/SQSTM1 rabbit polyclonal (1:1000; Sigma‐Aldrich, P0067) was incubated in a solution of PBS 1% FBS for 1 h in a humidity camber. Secondary antibody anti‐Rabbit (Alexa Fluor 488, Goat; 1:1000; Thermo Fisher Scientific, A11034) and α‐Phalloidin CruzFluor 647 Conjugate (1:1000; Santa Cruz Biotechnologies, sc‐363 797) were incubated for 1 h at room temperature in the same solution. After, fibroblasts were washed with PBS three times and mounted on a coverslip using 4′,6‐diamidino‐2‐phenylindole (DAPI; Thermo Fisher Scientific, P36966).

Images were acquired with a Leica TCS SP8 Confocal Microscope (Leica, Wetzlar, Germany) using 40× immersion oil objective. The original data was stored as 8‐bit greyscale images with a spatial resolution of 1024 × 1024 pixels. In order to compare the data, identical settings were used for image acquisition of different experiments and the image analysis were performed using Image J/Fiji software (NIH, US National Institutes of Health). All Z‐stacks were projected in a Sum Slices projection and an automated Yen‐thresholding was applied to 8 bit‐images before the particle analysis. The number of particles of each cell was normalized by the cell area to obtain the number of p62 aggregates (particles)/μm². To facilitate visualization, images of p62 aggregates have been displayed in an inverted grey scale and Phalloidin in magenta LUT. Merge images are displayed in the main figures and separate individual channels in the Figures S2 and S5.

2.4. Traffic‐Light LC3B Assay

To visualize and quantify autophagosomes (APG) and autophagolysosomes (APGL) we transfected control and patients fibroblasts with an mCherry‐GFP‐LC3 plasmid (Addgene, 110060), which codes for a LC3B protein linked to GFP and mCherry. As explained in [7, 8] when internalized in an autophagosome, both fluorophores can be detected, while the acidic environment of the autolysosome prevents GFP fluorescence. Quantification of the particles in each channel reflects the total amount of autophagic vesicles, along with the distribution between autophagosomes (green and red) and autolysosomes (red without green). Cells were observed by confocal microscopy following the previously mentioned conditions; Z‐stacks were projected into a Maximum Intensity projection and an automated Otsu‐thresholding was applied to 8‐bit images before the particle analysis.

2.4.1. Plasmid Transfection

For plasmid transfection, fibroblasts were seeded at 80% confluence 24 h prior to the transfection. 500 ng of plasmid were transfected per M24 well using Lipofectamine 2000 (Thermo Fisher Scientific, 11668030) and following the manufacturer instructions.

2.5. Transferrin Internalization Assay

To monitor Transferrin (Tf) internalization and recycling, fibroblasts were initially cooled for 10 min and washed with cold HEPES buffer (Sigma‐Aldrich, H0887) supplemented with 1% BSA and 20 mM Glucose. After, cells were incubated with pHrodo Red‐conjugated Tf (25 μg/mL) (Thermo Fisher Scientific, Invitrogen, P35376) in supplemented HEPES for a 10 min‐pulse at 37°C. Then, cells were washed twice with PBS to remove the excess of unbound Tf and incubated with supplemented DMEM at 37°C in a 5% CO₂ incubator. The cells were chased at different times after a 10 min pulse. The following timepoints after the addition of conjugated Tf were measured: 0′ 5′, 10′, (pulse), 15′, 30′ and 60′ (chase). At each time point, cells were fixed with 4% PFA and stored in darkness and cold. For each condition, we performed a time 0 incubation by add and removing the probe, and a negative control which was only incubated with supplemented HEPES.

2.5.1. Transferrin Assay Analysis

For each quantified cell, the fluorescence of 3 equally ROIs in the perinuclear zone, equivalent to endosome distribution, was measured. Background fluorescence was measured for each image and subtracted from the measures of the cells of interest to eliminate image signal background. All values presented are net and comparable between conditions, as each condition has been adjusted by subtracting the average basal autofluorescence signal from each line acquired in the negative control images. The average of triplicate values of each cell was used to calculate the mean intensity of fluorescence for each condition at each time measurement, as well as their respective standard error of the mean (SEM). At least 15 cells per experiment were analyzed and between two and four independent experiments were performed.

2.6. Western Blot

Fibroblasts were homogenized in RIPA Buffer (Sigma‐Aldrich, R0278) supplemented with a protease and phosphatase inhibitor cocktail. Homogenates were in cold during 30 min and then centrifuged at 13500 rpm for 10 min. The protein concentration of the supernatant was quantified by Bradford method and prepared at a homogeneous concentration in 3× Laemmli Buffer. Lysates were resolved in sodium dodecyl sulfate‐polyacrylamide gels (SDS‐Page) and transferred onto Amersham Hybond PVDF membranes 0.22 mm pore (GE Healthcare GVWP04700) at 100 V 1 h. Membranes were blocked with 5% defatted‐milk in TBS—0.05% Tween20. Primary antibodies were incubated overnight at 4°C in 5% milk, at the following concentrations: GAPDH (1:20000, Ms.; Proteintech, 60 004‐1‐Ig), LC3B (1:5000, Rb; Cell Signaling, D11), NBAS (1:1000, Rb; Thermo Fisher Scientific, PA5‐49534), p31/USE1 (1:250, Rb; Sigma‐Aldrich Prestige Antibodies, HPA026851), p62/SQSTM1 (1:1000, Rb; Sigma‐Aldrich, P0067), Rab4 (1:1000, Ms.; Santa Cruz Technologies, sc‐376 243), Rab5 (1:1000, Ms.; Cell Signaling, 46 449), Rab7 (1:1000, Ms.; Cell Signaling, 95 746), SYNJ1 (1:500, Rb; Thermo Fisher Scientific, PA5‐82442), Tubulin (1:50000, Ms.; Abcam, ab7291). GAPDH and Tubulin were used as a loading control. The primary antibodies were detected using secondary antibodies conjugated to horseradish peroxidase (HRP) at 1:5000 dilution: goat anti‐Rabbit (Thermo Fisher Scientific, A31458) and goat anti‐Mouse (Thermo Fisher Scientific, A16017) IgG antibodies. Proteins were detected by chemiluminescence using the Pierce ECL Western Blotting substrate (Thermo Fisher Scientific, 32 106). For the LC3B protein detection, we needed 10 ug protein in each sample, 1:100000 secondary antibody concentration and the incubation with SuperSignal West Atto Ultimate Sensitivity Substrate (Thermo Fisher Scientific, A38555) for 5 min. Quantification of protein levels was performed using Fiji software by calculating protein densitometry relative to controls and normalized to each loading control.

2.7. Autophagy Markers

Autophagy flux has been quantified based on (1) the formation of lipidated form LC3B‐II, by expressing the ratio of lipidated to non‐lipidated form (LC3B‐II/I ratio) by Western Blot, and (2) the aggregation of the autophagy receptor p62/SQSTM1 by immunofluorescence [9]. Both parameters were studied under the effect of two autophagy modulators: Bafilomycin A1 (i) and EBSS medium (ii); the expected effect of these modulators in control situation is represented in the schematic view in Figure S1. (i) Bafilomycin A1 is an inhibitor of autophagosome‐lysosome fusion [10]. In a control situation an increase LC3B‐II/I ratio and p62 aggregates is expected upon inhibition with bafilomycin, representing the accumulation of the autophagosomes at the last part of the route without proper degradation. (ii) Opposite to that, the induction of autophagy by nutrient deprivation through EBSS incubation leads to a rapid increase in the formation of autophagosomes. In a control situation, an increase LC3B‐II/I ratio and a decreased number of p62 aggregates due to accelerated degradation is expected.

2.8. Patients' Neurodevelopment Assessment and Treatment With Spermidine

Both patients were treated with wheat germ extract 0.1% spermidine, receiving a total spermidine amount of 6 mg/day, split in three 2 mg doses. The effect of the treatment was assessed by evaluating the patient's neurodevelopment through the Vineland Adaptive Behavior Scale—third edition (VABS‐III), specifically the Comprehensive Parent Caregiver Form. VABS‐III semi‐structured interview was conducted by a trained neuropsychologist before the treatment and after 3 months of spermidine treatment. Growth Scale Value (GSV) have been calculated, derived from the standardized scores obtained in each domain, and based on normative benchmarks. With a mean of 100 and a typical range of 85–115 in normotypic subjects, it is used to measure progress.

2.9. Statistical Analyses

All results have been obtained by a minimum of three independent experiments, including at least sample duplicates in each experiment. For statistical analysis, GraphPad Prism (version 10.0; La Jolla, CA, United States) software was used. First, outliers were identified by ROUT test and discarded, and then data normality was determined by Shapiro test. For compare two groups, t Test was used for both parametric and nonparametric test. For more than two groups, ANOVA test for parametric data or Kruskal‐Wallis for non‐parametric data were performed, with posterior multiple comparisons analysis correction when was appropriate. Results are expressed as a mean ± standard deviation (SD). RStudio was used for plot representation of transferrin assay.

Differences were considered as statistically significant at p‐values less than 0.05 and are indicated by asterisks *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

3. Results

3.1. Spermidine Partially Recovers the Autophagy Alterations in Synj1 Deficiency

Synaptojanin‐1 (Synj1) is a phosphoinositide phosphatase involved in various cell trafficking processes. Through its two different enzymatic domains it plays a role in membrane trafficking, both in endocytosis and recycling and in autophagy. The mutation described in the SYNJ1‐patient was located towards the end of the 5′ phosphatase domain, that dephosphorylates phosphatidylinositol 4, 5‐ bisphosphate (PI(4, 5)P2) to PI(4)P and is involved in endocytosis and recycling (Figure S2a). Assessment of the protein expression in primary fibroblasts of SYNJ1‐patient (referred to as patient^SYNJ1) and gender and age‐matched controls supported the pathogenicity of the mutation; while we did not observe an altered expression in the 170 kDa canonical isoform, we detected an 90% decrease in the expression of the 145 kDa isoform, which is the mainly expressed isoform in brain (Figure S2b). Moreover, a Transferrin receptor recycling assay confirmed an alteration in the endocytosis‐recycling activity, as the patient^SYNJ1 fibroblasts showed a different Transferrin internalization profile compared to controls (Figure S2c). Given the alterations in this process, we explored the three main populations of endosomes involved in recycling through specific markers: RAB5 for early endosomes (EE), RAB4 for recycling endosomes (RE, which mark the recycling route) and RAB7 for late endosomes (LE), marking the degradation route (Figure S2d). As shown in Figure S2e, patient^SYNJ1 fibroblasts showed a significant increase in the expression of markers of EE and a decrease in those of LE compared to controls, suggesting a potential impairment in the degradative pathway.

Since disruptions in autophagy have been described in SYNJ1 deficiency, and considering the mentioned alterations in the degradative pathway, we explored the autophagy flux in the patient^SYNJ1 and control fibroblasts through two markers: p62/SQSTM1 total expression and aggregations (indicative of the initial phases of autophagy) and LC3B II/I ratio, as a surrogate for the autophagosome formation and autophagy progression. We observed, under normal culture conditions, an increase on total p62 expression and the number of aggregates per cell in the patient^SYNJ1 fibroblasts, together with a significant increase in the LC3B II/I ratio (Figure 1A; Figure S3a). Since this suggested an alteration in the autophagy flux, either due to an increased autophagy activity or a decreased degradation of the autophagy markers, we delved into its analysis by studying both parameters in two opposite scenarios: in the presence of the lysosome inhibitor bafilomycin at 200 nM for 4 h, which prevents the degradation of the autophagy markers, and the opposite, stimulating autophagy by incubating the fibroblasts for 4 h in Earle's Balanced Salt Solution (EBSS). Treating the cells with bafilomycin resulted in a higher increase in p62 expression and aggregation in patient^SYNJ1 than in control (Figure 1B, upper panel; Figure S3b). This difference was not mirrored in the LC3B II/I ratio, as its increase over basal conditions was equal in both patient and control cells (Figure 1B, lower panel). Analysis of the opposite scenario, this is, induction of autophagy with 4 h EBSS incubation, resulted in a decrease in p62 aggregates in the control fibroblasts but not in the patient^SYNJ1 ones (Figure 1C; Figure S3c), although not reflected in total protein detection. Accordingly, the response in LC3B II/I ratio to EBSS stimulation was lower in the patient^SYNJ1 cells (Figure 1C, lower panel). Altogether, these results pointed towards an impairment in the progression of the autophagy flux.

FIGURE 1.

Characterization of Synj1 patient's pathophysiology by the study of the mutation and its role in autophagic (A–F) pathway. (A–F) Quantification of autophagy markers p62/SQSTM1 by immunofluorescence and Western blot (upper panels) and LC3B by Western Blot (lower panels) under basal conditions (A), autophagy inhibition with Bafilomycin A1 (Baf) (B) and autophagy stimulation with EBSS (C) for 4 h. Response to spermidine (SPM) supplementations was analyzed by the same markers in basal (D) and EBSS‐stimulated conditions (E). (F) Graphical representation of the total number of autophagic vesicles and proportion of autophagosomes (APG) and autophagolysosomes (APGL), evaluated through a traffic‐light LC3B assay (mCherry‐GFP‐LC3 plasmid transfection) in control and patient fibroblasts with or without spermidine. A representative image of the merge is shown for each condition. On top of each subsection a schematic view of the autophagy pathway under the different conditions is provided; the diagrams highlight in yellow the section of the autophagic pathway that each treatment allows to focus on. Bafilomycin treatment, which inhibits the final steps of autophagy, enables a focus on the early stages of the pathway, while EBSS treatment, which stimulates autophagy, shifts the focus to the mid‐stages of the process. Western blot and confocal images are representative figures of at least three independent experiments. Confocal microscopy scale bars represent 8 μm for p62 immunofluorescence images (6× zoomed‐in section) and 30 μm for mCherry‐GFP‐LC3 images. Statistical analyses were performed as described in Materials and Methods. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Absence of asterisk means no statistically significant difference.

Considering the impairment in the autophagy progression in SYNJ1 patient fibroblasts, we wondered whether the nutraceutical spermidine could modulate such alterations. Spermidine is a polyamine naturally produced by the cells and has been described as an autophagy regulator. To test its in vitro effect in the context of the disease, after proving the experimental setup in control fibroblasts (Figure S4), we treated the patient cells with spermidine 1 μM for 24 h previous to the analysis of the mentioned markers. We did not observe any significant difference between the spermidine‐treated and untreated patient^SYNJ1 fibroblasts in basal conditions (Figure 1D). However, when cells were grown in starvation (EBSS), as shown in Figure 1E, we observed a reduction in p62 aggregates in spermidine‐treated patient^SYNJ1 cells, together with a significant increase in LC3B II/I ratio. These results point towards an effect of spermidine in recovering the autophagy progression in SYNJ1 patient fibroblasts. This was confirmed using a traffic‐light LC3B assay, as spermidine partially restored the impaired formation of autolysosomes observed in patient^SYNJ1 cells (Figure 1F). Noteworthy, it was not accompanied by a change in the endosome populations markers expression and only a modest correction in the endocytosis/recycling process (Figure S5).

3.2. Autophagy Impairment Underlies in NBAS Deficiency and Can Be Amended With the Nutraceutical Spermidine

Based on the positive results regarding autophagy modulation with spermidine in SYNJ1 deficiency, we wondered whether autophagy alterations that underlie in the pathophysiology of other cell trafficking disorders where targetable with spermidine as well. For that, we studied a second and unrelated patient (patient^NBAS, Figure S6a), who carried two variants affecting NBAS function. This protein is a subunit of the NRZ tethering complex (together with ZW10 and RINT1), and by interacting with RE‐resident proteins as p31, controls the docking of COPI vesicles in Golgi‐ER retrograde transport (Figure S6b). The first mutation was located in the region of the protein that interacts with p31, while the second affected ZW10 interacting region (Figure S6c). While the expression of NBAS itself was not reduced in the patient fibroblasts, the expression of its interacting protein p31 was significantly decreased, supporting the pathogenicity (Figure S6d).

Analysis of the autophagy flux revealed an augmented expression of p62 and p62 aggregates per cell in the NBAS‐patient (patient^NBAS) fibroblasts compared to controls, together with a significant increased LC3B II/I ratio in basal conditions (Figure 2A; Figure S7a). Following the previously described methodology, we deepened into the autophagy alterations. Upon treatment with bafilomycin 200 nM for 4 h we observed that, although not reflected in total protein expression, the increase in p62 aggregates compared to the basal quantification was significantly lower in the NBAS patient fibroblasts (Figure 2B, upper panel; Figure S7b). Similarly, while LC3B II/I ratio increased in both patient and controls upon treatment with bafilomycin, this increase was significantly lower in the patient^NBAS (Figure 2B, lower panel). Induction of autophagy by starvation led to a reduction in the p62 aggregates which was higher in the patient^NBAS than in the control cells, and an increase in LC3B II/I ratio, similar in both cell lines (Figure 2C; Figure S7c). Integration of these results pointed towards an impairment in the degradative phases of the autophagy flux.

FIGURE 2.

Characterization of Nbas patient's pathophysiology by the study of its role in autophagic pathway (A–F). (A–F) Quantification of autophagy markers p62/SQSTM1 by immunofluorescence and p62 and LC3B by Western Blot under basal conditions (A), or modulation of the pathway with Bafilomycin A1 (Baf) (B) and EBSS (C) for 4 h in absence (A–C) or presence of spermidine as a nutraceutical autophagy modulator (D–F). On top of each subsection a schematic view of the autophagy pathway under the different conditions is provided; the diagrams highlight in yellow the section of the autophagic pathway that each treatment allows to focus on. Bafilomycin treatment, which inhibits the final steps of autophagy, enables a focus on the early stages of the pathway, while EBSS treatment, which stimulates autophagy, shifts the focus to the mid‐stages of the process. Western blot and confocal images are representative figures of at least three independent experiments. GAPDH images in panels B, D and E are the same for p62 and LC3B blots in each panel, as both are crops of the same membrane and share the same loading control. Confocal microscopy scale bars represent 8 μm (6× zoomed‐in section). Statistical analyses were performed as described in Materials and Methods. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Absence of asterisk means no statistically significant difference.

We next explored the effect of spermidine supplementation as previously done, by treating the patient fibroblasts with spermidine 1 μM for 24 h. We did not observe any change neither in p62 expression or aggregates formation nor in LC3B II/I ratio in basal conditions (Figure 2D; Figure S7d). However, when we analyzed the bafilomycin‐treated scenario, we observed a greater increase in p62 aggregates formation, even though the total protein expression remained unaltered (Figure 2E, upper panel; Figure S7e). This was followed by a 3‐folds increase in LC3B II/I in spermidine‐treated patient^NBAS fibroblasts, suggesting a recovery of the autophagy flux progression (Figure 2E, lower panel). Similarly, autophagy stimulation through EBSS incubation also resulted in a normalization of the p62 aggregates consumption in spermidine‐treated patient^NBAS fibroblasts (Figure 2F; Figure S7f).

3.3. Spermidine Can Constitute a Potential Treatment for Cell Trafficking Disorders

Due to the positive in vitro response to spermidine treated, we evaluated its potential therapeutic effect in the patients. We evaluated four different domains of neurodevelopment through a Vineland test before and after a 3 months treatment with spermidine at a dose of 6 mg/day. Spermidine was well tolerated in both patients and none of them experienced any adverse effect.

As shown in Figure 3, we observed an improvement across the different evaluated neuropsychological domains, with disparities in both patients. Both patients experienced a similar increase in the socialization domain (4 standard points). They both experienced an improvement in daily skills (1 standard point for patient^SYNJ1 and 6 standard points for patient^NBAS) and psychomotor development (2 standard points for patient^SYNJ1 and 9 standard points for patient^NBAS). Only SYNJ1‐patient experienced an improvement in the communication domain (2 standard points), while no value changes were observed in the NBAS‐patient. Finally, a comparable increase in the adaptive behavior composite was observed in both patients (2 standard points for patient^SYNJ1 and 3 standard points for patient^NBAS).

FIGURE 3.

Visualization of clinical aspects assessed by Vineland study for both patients in a n‐of‐1 clinical trial pre‐ and post‐spermidine treatment. (A) Schematic representation of patient with workflow of Vineland parameters acquisition pre‐ and post‐ 3 months of spermidine treatment in a 6 mg/day doses. (B) Representation of score improvements post‐spermidine treatment during 2 months in Synj1 patient. (C) Representation of score improvements post‐spermidine treatment during 3 months in Nbas patient. GSV (growth scale values) are represented in the horizontal axis. Gray shadows represent the average values for controls in each patients age.

Besides the neurodevelopmental improvements, treatment with spermidine also resulted in a stabilization of patientNBAS symptomatology, especially regarding the succession of outbreaks of frequent symptoms, such as distal paresthesias, often the prodrome of a relapse of the leukoencephalopathy outbreak.

4. Discussion

Cell trafficking is the biological processes for the exchange of signals and metabolites between cell compartments and with their environment. This includes membrane trafficking, autophagy, transport along the cytoskeleton and membrane contact sites. The proper functioning of these pathways is essential for cell homeostasis. As a matter of fact, monogenic diseases of cell trafficking form the largest group in the recent International Classification of Inherited Metabolic Disorders (ICIMD), and the number of monogenic diseases of cell trafficking has increased over 7 folds in the last 10 years [1]. These disorders have a heterogeneous clinical presentation and diverse pathophysiological features [11], which hinders their understanding and treatment. Nonetheless, within the interconnected web of cellular trafficking defects, autophagy emerges as a potential shared factor in their pathophysiology. Although comprehensive and systematic studies investigating the role of autophagy in cell trafficking disorders are scarce, disruptions in autophagy have been observed in various pathways of cellular trafficking, ranging from exocytosis (such as defects in TRAPPC11 [12, 13, 14]) to membrane contact sites (like mutations in Mfn2 [15]). Given our knowledge on autophagy modulation [16], defining its implication in the pathophysiological landscape of cellular trafficking disorders will unveil potential therapeutic strategies for their treatment.

Besides specific disorders, the close relationship between membrane trafficking and autophagy is particularly evident in the synaptic vesicle cycle [17, 18]. In fact, several agents bridging both processes have been described, such as Atg9 [19], EndophilinA [20] or Synaptojanin‐1 [21]. Particularly, the different domains of Synj1 protein are directly involved in the regulation of both activities. On one hand, through the 5′ phosphatase domain, Synj1 dephosphorylates phosphatidylinositol 4,5‐biphosphate (PI(4,5)P₂) to PI(4)P, a phosphoinositide concentrated at the plasma membrane, involved in endocytosis [22, 23]. On the other side, Synj1‐ SAC1 domain hydrolyzes PI(3)P and PI(3,5)P₂, which is key for autophagosome maturation [24]. Mutations affecting the SAC1 domain result in a failure in autophagosome maturation in Synj1^+/− astrocytes [25], resulting in an increase in the autophagy markers p62 and LC3B [26] similar to the variations suggested by our results in the Synj1‐patient fibroblasts in basal conditions. Expanding on the study of the autophagy impairment, we have studied the expression of both markers upon autophagy inhibition or stimulation. Our results revealed that upon autophagy inhibition with bafilomycin, p62 increase over basal condition was higher in the patient^SYNJ1 fibroblasts than the observed in controls. On the other hand, autophagy EBSS‐stimulation did not result in a consumption of p62 aggregates as observed in control cells, while the rise in LC3BII/I ratio was similar in both patient^SYNJ1 and controls. Narrowing down the defects in autophagy require a profound and multifaceted analysis; while might appear as counterintuitive, an increase in autophagy markers in one isolated condition can reflect various different alterations, requiring the study under different scenarios. Altogether, our results pointed towards an impairment in autophagosome maturation, that was confirmed by direct tracking of autophagolysosome formation in a traffic‐light experiment. These results confirm an alteration in the progression of the autophagy flux in Synj1‐patient fibroblasts.

While Synj1 is a bi‐functional enzyme involved in both autophagy and membrane trafficking through its different domains, this is not the most frequent case of proteins involved in cell trafficking. Yet, autophagy alterations have been described underlying in the pathophysiology of cell trafficking diseases as mutations in TRAPPC11 [14], Atg5 [27], VPS13D [28] or VPS13B [29]. Based on this premise, we explored whether alterations in autophagy contribute to the pathophysiology of an unrelated cell trafficking disease such as NBAS deficiency. The protein NBAS, together with ZW10 and RINT1 form the trimeric NRZ tethering complex, which is involved in the Golgi‐ER retrograde transport [30]. Specifically, the complex interacts with ER SNARE proteins and with the vesicle coatomer, tethering the COPI vesicle to the ER. As recently reviewed [4], NRZ complex is also involved in the coordination between retrograde transport and autophagy through RINT1. Agreeing with the described defects in RINT1 mutations, we demonstrate that the NBAS‐patient fibroblasts also show an impairment in autophagy flux. On top of the increase of both p62 aggregation and LC3BII/I ratio in basal conditions, inhibition with bafilomycin resulted in a lower rise of the detection of both markers compared to the observed in controls, suggesting an impairment in the degradative phases of autophagy. Moreover, upon autophagy induction through starvation, p62 aggregates were consumed at a higher rate in the patient^NBAS cells, which was not paired to a higher increase in LC3BII, supporting the cargoes degradation impairment. Altogether, these results pointed towards the involvement of autophagy in the pathophysiology of NBAS deficiency, not previously described and outstanding as a potential therapeutic target.

Delimiting autophagy involvement in the pathophysiology of cell trafficking disorders opens a promising avenue for their treatment given the availability of different autophagy‐regulating compounds, both activators and inhibitors [16]. Among the autophagy modulators we can find the nutraceutical spermidine, a naturally occurring polyamine. Due to its role as an autophagy inductor, and given its age‐induce natural decline, several studies exploring its supplementation for autophagy modulation in healthy population have been carried out (mostly, but not only, in the context of geroprotection) [31, 32, 33, 34]. Treatment with spermidine in the context of both Synaptojanin‐1 and NBAS deficiency improved the impairment in the autophagy flux progression. In SYNJ1 deficient cells this was evidenced in the EBSS‐grown cells, which upon treatment with spermidine recovered the p62 aggregates consumption and increased the formation of LC3BII. Likewise, in NBAS patient fibroblasts spermidine treatment resulted in an increase in p62 and LC3B in bafilomycin‐treated cells and a recovery on the p62‐LC3BII relation in EBSS‐grown cells. These in vitro results support spermidine as a potential positive modulator for the different autophagy defects potentially associated to cell trafficking disorders.

Encouraged by the positive in vitro results, and given that spermidine is a nutrient that is safe to use and has been tested for supplementation in humans, we investigated its effect as a possible treatment for both patients. To this end, we supplemented with 6 mg/day of spermidine in three 2 mg doses for 3 months, evaluating different areas of neurodevelopment before and after treatment through a Vineland test. While certain differences between patients' response was recorded, both patients improved in the socialization and daily kills domain, along with psychomotor development. The changes were not only measurable by neurodevelopment tests but also perceived by the patients' caregivers. NBAS patient family reported that the patient was in a “better shape” overall and improved the sleeping disturbances. These results are aligned with the previous studies reporting spermidine as a cognitive enhancer [31]. However, and although positive, the effects were modest, perhaps also conditioned by the short follow‐up time. To our knowledge, this is the first study exploring spermidine as a potential treatment for cell trafficking disorders in pediatric population. Further studies should explore the effect of sustained supplementation with spermidine and the long‐term effect, along with the efficiency in other cell trafficking disorders.

In summary, our results highlight that impairments in autophagy underlie in the pathophysiology of cell trafficking diseases, and that spermidine is a potential treatment for them, with positive effects both in vitro and in vivo. We acknowledge that the role of autophagy in trafficking‐related pathologies may vary significantly between diseases; while in some conditions autophagy may play a major role, in others it may be just one of several contributing factors to their pathophysiological landscape. Further studies should analyze systematically whether such defects are common to all cell trafficking disorders and the long‐term effect of spermidine supplementation, to unveil its full potential. Study of autophagy and its modulation in cell trafficking disorders can reveal new therapeutic strategies for this growing group of inborn errors of metabolism.

Ethics Statement

All parents or legal representatives of the patients gave written informed consent. This study was approved by the local institutional ethics committee (Sant Joan de Déu Hospital ID number: PIC‐131‐18). All research work has been carried away following the principles under the Helsinki declaration.

Conflicts of Interest

A.O. is a co‐founder of the Hospital Sant Joan de Déu start‐up ‘Neuroprotect Life Sciences’ A.G.C. has received honoraria for research support and lectures from PTC Therapeutics, Immedica and Nutricia, honoraria for lectures from Biomarin and Recordati Rare Diseases Foundation, and is a co‐founder of the Hospital Sant Joan de Déu start‐up ‘Neuroprotect Life Sciences’. The content of the article has not been influenced by the sponsors.

Supporting information

Figure S1. Schematic representation of autophagic flux in control cells through the study of two main autophagic markers: p62/SQSTM1 and LC3B‐II.

Figure S2. Characterization of Synj1 patient’s pathophysiology by the study of the mutation and its role in endosomal pathway.

Figure S3. Representative images of p62 in control and Synj1 patient fibroblasts, with or without spermidine treatment.

Figure S4. Modulation of spermidine treatment effect in control fibroblasts: characterization of p62 and LC3B protein basal levels.

Figure S5. Characterization of Synj1 patient mutation and endosomal pathway modulation with spermidine treatment.

Figure S6. Characterization of Nbas patient’s pathophysiology by the study of the mutations.

Data Availability Statement

All data has been included in the manuscript.