Chaperone-Mediated Autophagy Ablation in Pericytes Reveals New Glioblastoma Prognostic Markers and Efficient Treatment Against Tumor Progression

Background: The lack of knowledge of the progression mechanisms of glioblastoma (GB), the most aggressive brain tumor, contributes to the absence of successful therapeutic strategies. Our team has recently demonstrated a crucial new role for chaperone-mediated autophagy (CMA) in pericytes (PC)-acquired immunosuppressive function, which prevents anti-tumor immune responses and facilitates GB progression. The possible impact that GB-induced CMA in PC has on other functions that might be useful for future GB prognosis/treatment, has not been explored yet. Thus, we proposed to analyze the contribution of CMA to other GB-induced changes in PC biology and determine if CMA ablation in PC is a key target mechanism for GB treatment. Methods: Studies of RNA-seq and secretome analysis were done in GB-conditioned PC with and without CMA (from knockout mice for LAMP-2A) and compared to control PC. Different therapeutic strategies in a GB mouse model were compared. Results: We found several gene expression pathways enriched in LAMP2A-KO PC and affected by GB-induced CMA in PC that correlate with our previous findings. Phagosome formation, cellular senescence, focal adhesion and the effector function to promote anti-tumor immune responses were the most affected pathways, revealing a transcriptomic profiling of specific target functions useful for future therapies. In addition, several molecules associated with tumor mechanisms and related to tumor immune responses such as gelsolin, periostin, osteopontin, lumican and vitamin D, were identified in the PC secretome dependent on GB-induced CMA. The CMA ablation in PC with GB cells showed an expected immunogenic phenotype able to phagocyte GB cells and a key strategy to develop future therapeutic strategies against GB tumor progression. A novel intravenous therapy using exofucosylated CMA-deficient PC was efficient to make PC reach the tumor niche and facilitate tumor elimination. Conclusion: Our results corroborate previous findings on the impaired immunogenic function of PC with GB-induced CMA, driving to other altered PC functions and the identifications of new target markers related to the tumor immune responses and useful for GB prognosis/therapy. Our work demonstrates CMA ablation in PC as a key target mechanism to develop a successful therapy against GB progression.

detection of those from low-abundance proteins (Pascovici et al., 2016). Thus, for sample preparation, we use BluePrep Major Serum Protein Removal Kit (SERVA) following manufacturer´s recommendations. Control GB media, control cell culture media, cell culture media from WT PC or KO PC co-cultured with and without GB, were depleted from the six major serum proteins through chromatography spin-columns (SERVA). Subsequently, the proteins of the cell culture media were quantified and digested with trypsin, identified by means of HPLC-MS/MS analysis and validated using auto thresholds by the Proteomics facilities of the University of Murcia-IMIB Arrixaca.
In-solution trypsin digestion. Samples were digested with the following standard procedure. Samples were dissolved in 100 l of 50 mM ammonium bicarbonate buffer pH 8.5 containing 0.01% ProteaseMax (Promega). This surfactant enhances the trypsin digestion. Protein samples were reduced by adding 20 mM DTT at 56C for 20 min. Then, samples were alkylated by adding 100 mM IAA during 30 min at room temperature in the dark. Finally, digestion was performed by adding 1 g of Trypsin Gold Proteomics Grade (Promega) (1:100 w/w) for 3h at 37 C. Reaction was stopped with 0.1% formic acid and filtered through 0.22 m. Finally, samples were dried using an Eppendorf Vacuum Concentrator model 5301.

HPLC-MS/MS analysis.
The separation and analysis of the tryptic digests of the samples were performed with a HPLC/MS system consisting of an Agilent 1290 Infinity II Series HPLC (Agilent Technologies) equipped with an Automated Multisampler module and a High Speed Binary Pump, and connected to an Agilent 6550 Q-TOF Mass Spectrometer (Agilent Technologies) using an Agilent Jet Stream Dual electrospray (AJS-Dual ESI) interface.
Dry samples from trypsin digestion were resuspended in 20 l of buffer A, consisting in water/acetonitrile/formic acid (94.9:5:0.1). Samples were injected onto an Agilent AdvanceBio Peptide Mapping HPLC column (2.7 m, 150  1.0 mm, Agilent technologies), thermostatted at 55 C, at a flow rate of 0.05 ml/min. This column is suitable for peptide separation and analysis. After the injection, the columns were washed with buffer A for 5 min and the digested peptides were eluted using a linear gradient 0-40% B (buffer B: water/acetonitrile/formic acid, 10:89.9:0.1) for 90 min followed by a linear gradient 40-95% B for 20 min. The column was equilibrated in the initial conditions for 10 min before every injection.
The mass spectrometer was operated in the positive mode. The nebulizer gas pressure was set to 35 psi, whereas the drying gas flow was set to 14 l/min at a temperature of 300 C, and the seath gas flow was set to 11 l/min at a temperature of 250 ºC. The capillary spray, nozzle, fragmentor and octopole RF Vpp voltages were 3500V, 100V, 360V and 750V respectively. Profile data were acquired for both MS and MS/MS scans in extended dynamic range mode. MS and MS/MS mass range was 50-1700 m/z and scan rates were 8 spectra/sec for MS and 3 spectra/sec for MS/MS. Auto MS/MS mode was used with precursor selection by abundance and a maximum of 20 precursors selected per cycle. A ramped collision energy was used with a slope of 3.6 and an offset of -4.8. The same ion was rejected after two consecutive scans.
The MS/MS search against the appropriate and updated protein database was performed with the following criteria: variable modifications search mode (carbamidomethylated cysteines, STY phosphorylation, oxidized methionine, and N-terminal glutamine conversion to pyroglutamic acid); tryptic digestion with 5 maximum missed cleavages; ESI-Q-TOF instrument; minimum matched peak intensity 50%; maximum ambiguous precursor charge +5; monoisotopic masses; peptide precursor mass tolerance 20 ppm; product ion mass tolerance 50 ppm; and calculation of reversed database scores. Validation of peptide and protein data was performed using auto thresholds.
Experimental parameters for HPLC and Q-TOF were set in MassHunter Workstation Data Acquisition software (Agilent Technologies, Rev. B.08.00) for free label quantification and identification (Kronstrand et al., 2018) in three pooled of concentrated supernatants from three different experiments and for each experimental line (WT PC; KO PC; WT PC + GB; KO PC + GB; control GB; control cell culture media). The proteins in the culture medium from negative controls (WT PC, KO PC, GB, control cell culture media) were subtracted from the KO PC + GB and WT PC + GB averages and the ratio of the averages was determined.
Immunofluorescence and microscopy: For detection of protein associated to the phagosome function ( Supplementary Figure 2A), WT and KO PC, alone or cocultured with U87 and U373 GB lines, were 100% methanol-fixed, permeabilized with Triton X-100 0.1%, blocked with BSA and incubated with antibodies against mouse ATP6V1A (Abcam) and alpha-Smooth Muscle Actin (α-SMA; Abcam). Anti-Rabbit AlexaFluor 488 (Invitrogen) was used as secondary antibody against ATPase and Anti-Mouse Cyanine5 (Invitrogen) against α-SMA.
For in vitro phagocytosis assay (Supplementary Figure 3), a protocol similar to that described in Diaz-Aparicio et al. was followed (Diaz-Aparicio et al., 2016). Briefly, WT PC and KO PC were allowed to rest and settle for at least 48 h before phagocytosis experiments in 24-well plates. GB cell lines were previously labeled with the cell tracker DiI and treated with 60 M of staurosporine (Cayman Chemical) for 48 h to induce apoptosis. Only the floating dead-cell fraction was collected from the supernatant and added to the PC cultures in a proportion of 1:1. Apoptotic cells were visualized and quantified by trypan blue exclusion. Because cell membrane integrity is still maintained in early induced apoptotic cells, cells not labeled with trypan blue were considered apoptotic. After 2 h, cells were fixed with 4% paraformaldehyde in PBS after washing away with media to discard all apoptotic cells non-trapped by PC. Due to its non-adherent nature, after cell washing, only apoptotic cells that have been trapped by PCs (in different stages of phagocytosis) remain. Remanent apoptotic cells trapped by PC and PC were stained with AlexaFluor 488-labeled Phalloidin (Invitrogen) to detect F-actin cytoskeleton.
Images were acquired with a Delta Vision RT (Applied Precision) restoration microscope coupled to a Coolsnap HQ camera (Photometrics), with a 60x/1.42 Plan Apo or 100x/1.40 Uplan Apo objectives. Morphometric measurements and quantification of cells, including histochemical quantification of phagocytic populations shown in Supplementary Figure 5, were performed using ImageJ (NIH, USA) software. Pictures for illustrations and quantitative analysis were uploaded from direct microscopic images and were not manipulated in subsequent steps of figures preparation, except for framing and scaling.
Flow cytometry analysis. Expression of MHC class I (H-2Kb, clone: AF6-88.5; BD Biosciences) and MHC class II (I-A/I-E, clone: M5/114.15.2; eBioscience,) were analyzed using DiI labeling solution (Invitrogen) for tracking and fluorescence separation of cells and specific anti-mouse antibodies. Background fluorescence was analyzed using labeled isotype monoclonal antibodies, and GB cells were measured as negative controls to discard nonspecific labeling of human cells using anti-mouse antibodies. Stained cells were analyzed by flow cytometry using a FACSCanto flow cytometer (BD Biosciences) and data were analyzed with Flowjo analysis software (FlowJo, LLC).
For HCELL expression, wild type pericytes or LAMP-2A knock out pericytes were exofucosylated and analyzed for expression of sLeX with HECA452 antibody or control isotype by flow cytometry. All results are representative of at least four independent experiments using both wild type and LAMP-2A knock out pericytes. Table   (S1,A Nonspecific fluorescence was measured using specific isotype monoclonal antibodies, and GB cells were used as negative control (not shown). Data represents mean ± SD obtained from at least 5 independent experiments using U373 and U87 GB lines; *P < 0.05; **P < 0.01.