Intracellular and extracellular domains of protein tyrosine phosphatase PTPRZ-B differentially regulate glioma cell growth and motility.

Gliomas are primary brain tumors for which surgical resection and radiotherapy is difficult because of the diffuse infiltrative growth of the tumor into the brain parenchyma. For development of alternative, drug-based, therapies more insight in the molecular processes that steer this typical growth and morphodynamic behavior of glioma cells is needed. Protein tyrosine phosphatase PTPRZ-B is a transmembrane signaling molecule that is found to be strongly up-regulated in glioma specimens. We assessed the contribution of PTPRZ-B protein domains to tumor cell growth and migration, via lentiviral knock-down and over-expression using clinically relevant glioma xenografts and their derived cell models. PTPRZ-B knock-down resulted in reduced migration and proliferation of glioma cells in vitro and also inhibited tumor growth in vivo. Interestingly, expression of only the PTPRZ-B extracellular segment was sufficient to rescue the in vitro migratory phenotype that resulted from PTPRZ-B knock-down. In contrast, PTPRZ-B knock-down effects on proliferation could be reverted only after re-expression of PTPRZ-B variants that contained its C-terminal PDZ binding domain. Thus, distinct domains of PTPRZ-B are differentially required for migration and proliferation of glioma cells, respectively. PTPRZ-B signaling pathways therefore represent attractive therapeutic entry points to combat these tumors.

A pENTR/U6 derivative was customized by inserting oligonucleotide heteroduplex set 6 between the unique SalI-XbaI sites, rendering pENTR/NotI-XhoII. To clone PTPRZ-B cDNA (Refseq NM_001206838.1), we first tailored pENTR/NotI-XhoI by inserting an SstII site-containing linker (set 7) in the unique HindIII site. The PTPRZ-B open reading frame was produced by reverse-transcriptase PCR using oligonucleotide set 8 and E98 total RNA as template. The SstII and NotI digested cDNA was subsequently inserted into the SstII sitecontaining pENTR/NotI-XhoI variant, resulting in plasmid pENTR-wtPTPRZ-B. A shPTPRZ1-resistant PTPRZ-B version (further indicated as pENTR-PTPRZ-B) was created via introduction of a silent C-T mutation at nucleotide position 1861 (numbering according to NM_001206838) in the shPTPRZ1 recognition site, using oligonucleotide set 9 and the Quickchange site-directed mutagenesis kit (Stratagene) according to manufacturer"s instructions.
PTPRZ-B cDNA was subsequently adjusted via site-directed mutagenesis to encode an enzymatically inactive PTPRZ-B C/S mutant (GC to CG, at positions 3613-3614) using oligonucleotide set 10 and the afore-mentioned protocol. Furthermore, a C-terminally VSVtagged full-length PTPRZ-B variant was generated by first generating a KpnI site (AGTTTAA to GGTA, pos. 4757-4763) using oligonucleotide set 11, and subsequently introducing oligonucleotide heteroduplex set 12, encoding an in-frame C-terminal VSV-G epitope tag (flanked by KpnI sites). To enable expression of the PTPRZ-B ecto-domain only, first a KpnI site was created at the codon preceding the PTPRZ-B transmembrane-encoding region (AGTTATA to GGTA, pos. 2726-2732) using oligonucleotide set 13 in the mutagenesis protocol. Subsequent ClaI digestion and re-ligation resulted in removal of residues 776-1448 comprising the PTPRZ-B intracellular domains. Finally, the new KpnI site was used to insert the VSV-G epitope tag-encoding heteroduplex 12. All resulting pENTR-3 PTPRZ-B plasmid variants were sequence-verified before being used in Gateway LR cloning reactions with pLenti6/PGK-DEST-TagRFP as destination vector. Also using Gateway cloning, the empty pENTR/NotI-XhoI vector served to generate pLenti6/PGK-EV-TagRFP as empty vector control.

Lentiviral transduction of glioblastoma cells and spheroids
Lentiviruses were produced using HEK-293FT cells according to the manufacturer"s instructions (Invitrogen). Briefly, 95% confluent 10cm culture dishes with HEK-293FT cells were transfected overnight, using JetPRIME reagent (Westburg) and the appropriate plasmid cocktail. The next day, medium was refreshed and 48-72 hrs later virus-containing medium was harvested, passed through a 0.45 µm pore size filter and stored at -80 °C. E98 Glioma cells or E434 spheroids were transduced by adding virus-containing medium to the cultures, at a 1:2 to 1:5 virus to medium ratio. After an overnight incubation, cells were superinfected with virus to increase the percentage of transduced cells. Routinely, this led to 80-100% transduction efficiency for E98 cells and 40-80% for the spheroid E434 cultures. Stably transduced E98 cells were selected by adding Blasticidin (2 µg/mL; Invitrogen). For rescue experiments, cells were first transduced twice with shPTPRZ1-expressing lentiviruses.
Several days later, two or three rounds of transduction with PGK promoter-driven rescue constructs were performed and cells were subjected to the proliferation and migration assays 72 hrs later.

Real-time quantitative RT-PCR
Total RNA was isolated using RNA-Bee (Tel-Test Inc. cs104B) using standard trizolchloroform extraction methods, and concentrations were measured spectrophotometrically.
Reverse transcriptase reactions were performed using Iscript tm cDNA synthesis kit (Bio-Rad) according to the supplier's specifications. Specificity and efficacy of real-time quantitative PCR primer pairs for PTPRZ1 and β-actin (Qiagen) have been verified previously by Schmidt et al. [7]. Reactions, containing 3 μL of cDNA, 1 µL of the pre-mixed primer pair, 5 μL of SYBR Green PCR master mix (Bio-Rad) and 1 µL MQ, were run on a CFX96 tm Real Time system using the C1000 tm Thermal Cycler (Bio-rad). Reactions were initialized at 95 °C for 15 minutes and then cycled 40 times at 95 °C for 15 s and 60 °C for 40 s. After the last cycle, a dissociation curve was recorded between 60 °C and 95 °C with and increment of 0.5 °C.

4
The amount of PTPRZ1 RNA was determined relative to ACTB levels using the delta Ct method [8].
To visualize PTPRZ-B ecto-VSV in E98 conditioned medium, immunoprecipitation was performed. 30 μL ProtG sepharose beads (GE Healthcare, #17-0618-01) were incubated with mouse-VSV antibody overnight at 4 °C. Beads were washed 5 times with TBS and incubated with E98 conditioned medium overnight at 4 °C. Beads were washed 5 times with TBS and suspended in SDS sample buffer. After centrifugation, the supernatant was used for gel loading and blotted as described above. For purification purposes PTPRZ-B ecto-VSV was produced in HEK-293FT cells transfected with pLenti6/PGK-PTPRZ-B ecto-VSV-TagRFP, using JetPRIME according to the manufacturer instructions. The secreted PTPRZ-B ectodomain was purified from conditioned medium via immunoprecipitation using mouse anti-VSV bound to ProtG sepharose beads, as described above. VSV-tagged proteins were eluted from the beads using excess VSV peptide [9].
For co-immunoprecipitation, mouse anti-VSV was coupled to ProtG sepharose beads by overnight rotation at 4 °C in TBS. After 5 subsequent washes with TBS, conditioned medium from HEK293FT cells transfected with either pLenti6/PGK-EV-TagRFP or pLenti6/PGK-PTPRZ-B ecto-VSV-TagRFP was added to the beads allowing coupling to 5 VSV overnight at 4 °C. After 5 washes, E98 cell lysates (prepared as described above) were added to the beads and were incubated overnight at 4 °C. The next day, beads were washed 5 times with TBS before being taken up in 2x SDS sample buffer. Samples were processed for immunoblotting as described above.

Immunohistochemistry
FFPE sections of 4 μm were subjected to immunohistochemical stainings according to standard procedures [10]. In brief, sections were de-paraffinized in xylene and rehydrated in PBS. Endogenous peroxidases were blocked in 3% H 2 O 2 in PBS, followed by epitope retrieval (10 min boiling in 10 mM sodium citrate, pH 6.0). Slides were then washed twice in PBS, blocked in 20% normal serum (from the species in which the secondary antibody was Quantification of immunodetected EGFP and TagRFP signals was done using KS400 software (Carl Zeiss AG) and a custom-written macro. Sections of FFPE brains with orthotopic glioma xenografts were included in the analysis (n=3 and 2 for E98 and E434, respectively) and at least five non-overlapping microscopic fields (magnification x200) were measured per immunostaining for each animal. The TagRFP-or GFP-positive area per tumor field was divided by the total tumor area as determined via nuclear haematoxylin staining, and average values per animal were determined and used to calculate TagRFP/GFP ratios. Ratios were compared to those prior to injection using the one-sample Student"s t-test.

Peptide microarray analysis
E98 cells stably expressing shSCR/GFP and shPTPRZ1/TagRFP constructs were grown to 80% confluency in 6-well plates (4 wells per sample). Cells were washed twice with ice-cold PBS prior to lysis with M-PER Mammalian Extraction Reagent supplemented with protease and phosphatase inhibitor cocktails (Thermo Scientific) for 30 min at 4 °C. Lysates 6 were centrifuged (15 min, 14,000 rpm, 4 °C), and supernatants were snap frozen in liquid nitrogen and stored at -80 °C. Protein concentrations were determined using the BCA protein assay (Thermo Scientific). Kinase activity measurements were performed in quadruplicate on Tyrosine kinase PamChip arrays on a PamStation 12 instrument (PamGene International BV) essentially as described [11]. Sample input was 5 µg per array. A Student"s t-Test was used to identify the peptides that are significantly (p<0.05) different between the treatments.

Statictical analysis:
Statistical analysis was performed using GraphPad Prism 5 or PamGene"s proprietary