Regulation of lysosomal trafficking of progranulin by sortilin and prosaposin

Abstract Haploinsufficiency of the progranulin protein is a leading cause of frontotemporal lobar degeneration. Accumulating evidence support a crucial role of progranulin in the lysosome. Progranulin comprises 7.5 granulin repeats and is known to traffic to lysosomes via direct interactions with prosaposin or sortilin. Within the lysosome, progranulin gets processed into granulin peptides. Here, we report that sortilin and prosaposin independently regulate lysosomal trafficking of progranulin in vivo. The deletion of either prosaposin or sortilin alone results in a significant decrease in the ratio of granulin peptides versus full-length progranulin in mouse brain lysates. This decrease is further augmented by the deficiency of both prosaposin and sortilin. A concomitant increase in the levels of secreted progranulin in the serum was observed. Interestingly, while the deletion of both prosaposin and sortilin totally abolishes lysosomal localization of progranulin in neurons, it has a limited effect on lysosomal trafficking of progranulin in microglia, suggesting the existence of a novel sortilin and prosaposin independent pathway mediating progranulin lysosomal trafficking. In summary, our studies shed light on the regulation of lysosomal trafficking and processing of progranulin in vivo.


Graphical Abstract Introduction
Frontotemporal lobar degeneration (FTLD) is a devastating neurodegenerative disease which affects 250 000 people in the USA. 1,2 Heterozygous mutations in the granulin (GRN) gene, resulting in progranulin (PGRN) haploinsufficiency, are one of the major causes of FTLD with TDP-43 and ubiquitin-positive inclusions (FTLD-TDP). [3][4][5] PGRN is an evolutionarily conserved, secreted glycoprotein of 88 kDa, comprising 7.5 granulin modules. [6][7][8][9][10] The neurotrophic and anti-inflammatory functions of PGRN and granulin peptides are thought to prevent neurodegeneration in the ageing brain. 9,[11][12][13][14][15][16][17][18][19][20][21] However, multiple recent studies have suggested a critical role of PGRN in the lysosome. 22 First, homozygous PGRN mutant human patients exhibit neuronal ceroid lipofuscinosis (NCL), a lysosomal storage disorder. 23,24 PGRN knockout mice also accumulate the lysosomal byproduct, lipofuscin. 17 Most importantly, NCL-related phenotypes are reported in FTLD patients with GRN mutation, [25][26][27] supporting the theory that lysosomal dysfunction might serve as a common pathogenetic mechanism of these two diseases. Secondly, PGRN is transcriptionally co-regulated with a number of essential lysosomal genes by the transcriptional factor TFEB, a master regulator of lysosomal biogenesis. 28,29 Finally, PGRN has been shown to be a lysosome resident protein and there are two independent pathways for PGRN lysosome trafficking. Sortilin, a trafficking receptor of the vacuolar protein sorting 10 (VPS10) family, was shown to interact with PGRN and traffic PGRN to the lysosome. 30 In addition, PGRN interacts with another lysosomal protein prosaposin (PSAP) to get a 'piggy-back' ride to the lysosome through the PSAP receptors mannose 6-phosphate receptor (M6PR) and a low-density lipoprotein receptor-related protein 1 (LRP1) in a sortilin independent manner. 31 In the lysosome, PSAP is processed to saposin peptides, which regulate enzymes involved in glycosphingolipid degradation. 32 Interestingly, recently several studies have shown that PGRN is processed in a similar manner to individual granulin peptides in the lysosome and several proteases, including cathepsin B, L and D have been implicated in the process. [33][34][35] The granulin peptides have been shown to regulate the activities of several lysosome enzymes, including cathepsin D 27,36-38 and glucocerebrosidase. [39][40][41] Furthermore, heterozygous mutations in the GRN gene not only leads to the haploinsufficiency of full-length PGRN but also decreased levels of granulin peptides. 33 Anti-sortilin antibodies which block the PGRN-sortilin interaction are currently in clinical trials attempting to boost circulating PGRN levels to treat FTLD patients with GRN mutations (NCT04111666, NCT04374136, NCT03987295). Thus, it is important to examine the effect of sortilin ablation on PGRN trafficking and processing in vivo. In this study, we generated mice deficient in both PSAP and sortilin and analysed PGRN lysosome localization, the levels of granulin peptides as well as serum PGRN levels. Our results further confirm that PSAP and sortilin are two independent lysosome trafficking pathways for PGRN and support the existence of an additional lysosomal trafficking pathway of PGRN in microglia.
Cell culture and biochemical assays BV2 cells were maintained in Dulbecco's Modified Eagle's Medium (Cellgro) supplemented with 10% foetal bovine serum (Sigma) in a humidified incubator at 37°C with 5% CO 2 . To generate CRISPR constructs against mouse Psap, two oligonucleotides with the sequences 5 ′ -CACCGAAGAGGGCGAGGGCGTACA-3 ′ and 5 ′ -AAA CTGTACGCCCTCGCCCTCTTC-3 ′ were annealed and ligated to pLenti-CRISPRv2 (Addgene). Lentiviruses were generated by transfecting HEK293T cells with pLenti-CRISPRv2 constructs together with pMD2.G and psPAX2 plasmids. For CRISPR-mediated genome editing, BV2 cells were infected with lentiviruses containing pLenti-CRISPRv2 harbouring guide RNA sequences targeted to mouse Psap. The cells were selected with puromycin (2 µg/ml) 7 days after infection and the knockout is confirmed by western blot and immunostaining.

Western blot analysis
Mice were perfused with phosphate-buffered saline (PBS) and the tissues were dissected and snap-frozen with liquid nitrogen and kept at −80°C. On the day of the experiment, the frozen tissues were thawed and homogenized on the ice with bead homogenizer (Moni International) in ice-cold RIPA buffer [150 mM NaCl, 50 mM Tris-HCl (pH 8.0), 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate] with 1 mM phenylmethylsulfonyl fluoride, and 1× protease inhibitors (Roche). After centrifugation at 14 000 × g for 15 min at 4°C, supernatants were collected. Protein concentrations were determined via BCA assay, and then standardized. Samples were separated by 4-12% Bis-Tris PAGE (Invitrogen) and transferred to 0.2 µm nitrocellulose. Membranes were blocked with LiCor Odyssey blocking buffer for 2 h at room temperature followed by incubation with primary antibody overnight at 4°C with gentle rocking. Membranes were then washed with Tris-buffered saline with 0.1% Tween-20 (TBST) three times, 10 min each, and incubated with fluorescently tagged secondary antibodies (LI-COR Biosciences) for 1 h at room temperature, followed by three washes. Membranes were scanned using an Odyssey Infrared Imaging System (LI-COR Biosciences). Densitometry was performed using Image Studio (LI-COR Biosciences) and Image J.

Immunofluorescence staining, image acquisition and analysis
For brain section staining, mice were perfused with cold PBS and tissues were post-fixed with 4% paraformaldehyde. After dehydration in 30% sucrose buffer, tissues were embedded in O.C.T compound (Electron Microscopy Sciences). Twenty-micrometre-thick brain sections were cut with cryotome. Tissue sections were blocked and permeabilized with 0.1% saponin in Odyssey blocking buffer before incubating with primary antibodies overnight at 4°C. The next day, sections were washed 3× with cold PBS followed by incubation with secondary fluorescent antibodies and Hoechst at room temperature for 1 h. The slides were then mounted using mounting medium (Vector Laboratories). To block the autofluorescence in aged mice, brain sections from 12-month-old mice were incubated with 1× TrueBlack Lipofuscin Autofluorescence Quencher (Biotium) in 70% ethanol for 30 s at room temperature after the staining process. Images were acquired on a CSU-X spinning disc confocal microscope (Intelligent Imaging Innovations) with an HQ2 CCD camera (Photometrics) using 100× objectives, 10-12 different random images were captured. Lower magnification images were captured by 20× objectives on a Leica DMi8 inverted microscope, three to five images were captured from each sample. Data from ≥3 brains in each genotype were used for quantitative analysis.
For the quantitative analysis of neuronal PGRN levels in the brain sections, first neurons were selected based on the size of nuclei and PGRN expression, then the fluorescence intensity and area were measured directly using ImageJ after a threshold application. To quantify microglia PGRN levels in the brain sections, microglia was selected based on microglia maker IBA1 staining and the images were analysed using the same procedure in ImageJ. The total fluorescence signals were quantified. To quantify the degree of colocalization between PGRN and the lysosomal marker LAMP1, the JACoP plugin was used to generate Manders' overlap coefficients. 45 Three brain sections per mouse, separated by 100 µm, were used for quantification. The mean from the three sections was used to be representative of each mouse. Data were normalized to age-matched controls.

Enzyme-linked immunosorbent assay
Mouse serum was collected and analysed using mouse PGRN ELISA kit (R&D Systems, Catalogue no. DY2557) according to the manufacturer's instructions as previously described. 31

Statistical analysis
All statistical analyses were performed using GraphPad Prism 8. All data are presented as mean + SEM. Statistical significance was assessed by unpaired Student's t-test (for two groups comparison) or one-way ANOVA tests with Bonferroni's multiple comparisons (for multiple comparisons). P-values of ≤0.05 were considered statistically significant. *P , 0.05; **P , 0.01; *** P , 0.001.

Availability of data and material
All data have been included in the manuscript. Additional data are available upon request.

PGRN trafficking and processing in sortilin-deficient mice
Previously, we have shown that the ablation of sortilin in mice leads to an increase of PGRN levels in the brain lysates and in the serum. 30 To examine the role of sortilin in PGRN trafficking more thoroughly, we stained the brain sections from age-matched WT and Sort −/− mice with antibodies against PGRN, the main lysosomal membrane protein LAMP1 and microglia maker IBA1. Consistent with our previous findings, we found that neuronal signals of PGRN and more specifically, lysosomal pool of PGRN in neurons are significantly reduced by sortilin ablation in both 3-and 12-month-old mice (Fig. 1A and B and Supplementary  Fig. S1). However, microglial PGRN remains localized in the lysosome compartment in Sort −/− brain sections and microglial PGRN levels are not altered ( Fig. 1A and C), which is consistent with low expression levels of sortilin in microglia compared to neurons. 30 Next, we examined the levels of granulin peptides and PGRN in the cortical lysates from 3-month-old WT and Sort −/− mice by western blot analyses, using polyclonal antibodies that recognize both full-length PGRN and granulin peptides. 46 Concomitant with an increase in the levels of fulllength PGRN in Sort −/− mice, 30 a significant reduction in the levels of granulin peptides were observed in the cortical lysates from Sort −/− mice, which results in a significant decrease in the ratio between granulin peptides and full-length PGRN (Fig. 2). Thus, lysosomal trafficking defects caused by sortilin deletion are associated with a decrease in PGRN processing.

PGRN trafficking and processing in PSAP-deficient mice
Previously, we have shown that the PSAP interacts with PGRN and mediates sortilin independent lysosomal trafficking of PGRN. 31 To further determine the role of PSAP in PGRN processing in vivo, we examined the levels of both granulin peptides and PGRN in the cortical lysates from post natal day 21 (P21) WT and Psap −/− mice by western blot analyses, since Psap −/− mice die around weaning age. 44 The ablation of PSAP leads to severe lysosomal abnormalities, 31,44 and thus, a significant increase in PGRN mRNA levels ( Supplementary Fig. S2) and levels of both full-length PGRN and granulin peptides in the brain (Fig. 3). However, a significant decrease in the ratio between granulin peptides and full-length PGRN was observed, which supports that PSAP plays a critical role in regulating PGRN lysosomal trafficking in vivo.

Ablation of PSAP and sortilin leads to defects in lysosomal trafficking and processing of PGRN
To further analyse the role of PSAP and sortilin in PGRN trafficking in vivo, we generated mice with both sortilin and PSAP ablated. Then we analysed the PGRN lysosomal localization in the brain sections from Psap −/− Sort −/− mice. The PGRN signals retained in the lysosome compartment in Sort −/− neurons are completely lost in Psap −/− Sort −/− neurons ( Fig. 4A and B), indicating that PSAP and sortilin are the only two pathways involved in the PGRN lysosome trafficking in neurons. Significant colocalization between PGRN and the lysosomal marker LAMP1 was observed in both Psap −/− and Psap −/− Sort −/− microglia (Fig. 4A), indicating that there exists additional PSAP and sortilin independent PGRN lysosome trafficking mechanism in microglia. The increased intensity of PGRN was observed in both Psap −/− and Psap −/− Sort −/− microglia due to the transcriptional up-regulation of PGRN ( Supplementary  Fig. S2). To further confirm this finding, we ablated PSAP in the microglial cell line BV2, which has very low levels of sortilin expression ( Supplementary Fig. S3), and examined PGRN localization. PGRN remains localized in the lysosomal compartment in PSAP-deficient BV2 cells (Fig. 5A and B), further supporting that PGRN lysosomal trafficking in microglia is sortilin and PSAP independent.
Next we examined the levels of granulin peptides in the cortical lysates from 3-week-old WT, Sort −/− , Psap −/− Sort −/− mice as another readout of PGRN lysosomal trafficking. In the 3-week-old mice, the ablation of sortilin results in a slight increase in the levels of both full-length PGRN and granulin peptides. A slight decrease in the ratio between the granulin peptides and PGRN was also observed (Fig. 6). Deficiency of both PSAP and sortilin results in an obvious increase in the levels of both PGRN and granulin peptides in the cortical lysates, with a significant decrease in the ratio between granulin peptides and full-length PGRN (Fig. 6). However, there are still significant amount of granulin peptides present in Psap −/− Sort −/− cortical lysates (Fig. 6A), consistent with our immunostaining results that PGRN lysosomal trafficking is affected in neurons but not microglia in Psap −/− Sort −/− mice (Fig. 4). Taken together, these data support that PSAP and sortilin serve as two independent lysosome trafficking pathways for PGRN in vivo and are the main two pathways to deliver PGRN to lysosomes in neurons, but there exist other sortilin and PSAP independent mechanisms to mediate PGRN lysosome trafficking in microglia, and possibly other cell types as well.
Disruption of PGRN lysosome trafficking also results in increased sorting of PGRN to the secretory pathway and thus an increase in the serum levels of PGRN. 30,31 To further analyse the role of PSAP and sortilin in PGRN trafficking in vivo, we measured the levels of secreted PGRN in the serum. In the WT mice, the concentration of PGRN in the serum is 500 ng/ml, which is increased to 3200 ng/ml in Sort −/− mice. Deletion of one copy of Psap in Sort −/− background (Psap +/− Sort −/− ) leads to a further increase in PGRN levels to 4500 ng/ml, indicating that PSAP haploinsufficiency exacerbates PGRN lysosome trafficking defects (Fig. 7A). Similarly, haploinsuffiency of both PSAP and Sortilin in Psap +/− Sort +/− mice leads to a further increase in serum PGRN levels compared to Psap +/− or Sort +/− mice (Fig. 7B). In addition, ablation of both PSAP and sortilin leads to an increase of PGRN levels in the serum to 17 000 ng/ml, which is much more than in Psap −/− mice ( 2500 ng/ml) or Sort −/− mice ( 3200 ng/ml) (Fig. 7C), further confirming that PSAP and sortilin are two independent lysosome trafficking pathways for PGRN in vivo.

Discussion
Previously, we have shown that PSAP and sortilin are two independent pathways mediating PGRN lysosomal trafficking. 31,47       sortilin ( Supplementary Fig. S3), does not affect lysosomal trafficking of PGRN. All these data support that while PSAP and sortilin are the two main pathways mediating PGRN lysosomal trafficking in neurons, there exist additional PSAP and sortilin independent mechanisms in microglia and possibly the other cell types to facilitate PGRN lysosomal delivery. It remains to be determined whether PGRN binds to a distinct lysosomal trafficking receptor or gets a 'piggy-back' ride from another lysosomal protein to reach lysosomes in microglia. Nevertheless, our data support that PGRN takes multiple routes to the lysosome and each cell type might utilize different pathways for PGRN lysosomal delivery.
In agreement with the finding that PGRN is processed to granulin peptides in the lysosome, defects in PGRN lysosomal trafficking are associated with a decrease in the ratios between granulin peptides and full-length PGRN. Although the PSAP traffics with PGRN to the lysosome, PSAP does not seem to be required for PGRN, processing per se since PGRN can still be processed to the granulin peptides in PSAP-deficient cells and the defects are correlated with the lysosomal trafficking defects. It is possible that PSAP might affect the dynamics of PGRN processing and future work is required to fully examine this possibility.
Our analyses of PGRN processing and PGRN serum levels are consistent with the notion that PGRN lysosomal trafficking defects cause concomitant decrease in the levels of granulin peptides and increase in serum PGRN levels. The ablation of sortilin in mice has been shown to result in a significant increase in serum PGRN levels 30 and the increases in the sortilin levels due to the single nucleotide polymorphism (SNP) have been associated with decreased plasma levels of PGRN in humans. 48 Thus, sortilin has become a hot target for FTLD-GRN therapeutics to boost circulating PGRN levels. Small molecules and anti-sortilin antibodies have been developed to inhibit PGRN-sortilin interaction and anti-sortilin antibodies are currently in clinical trials for FTLD-GRN [49][50][51] (NCT04111666, NCT04374136, NCT03987295). Recent studies have shown that anti-sortilin antibodies not only block PGRNsortilin interaction but also downregulate sortilin levels. 49 However, given the critical role of PGRN and granulin peptides in the lysosome 10,22 and the fact that GRN mutations in FTLD also cause haploinsufficiency of granulin peptides, 33 restoration of both extracellular and lysosomal pools of PGRN in the brain of FTLD-GRN patients will be needed to have better therapeutic outcomes.

Supplementary material
Supplementary material is available at Brain Communications online.
Ethical approval and consent to participate All applicable international, national and/or institutional guidelines for the care and use of animals were followed. The work under animal protocol 2017-0056 is approved by the Institutional Animal Care and Use Committee at Cornell University.