Article Figures & Data
Figures
- Figure 1.
GPC4 is expressed on the astrocyte surface via a GPI-anchorage. A, Diagram of GPC4 sequence, protein, and posttranslational modifications. GPC4 contains an N-terminal signal peptide and GPI-anchor signal peptide for GPI-anchor attachment. Mature GPC4 is subject to furin cleavage, intrachain di-sulfide bonds, heparan sulfate attachment, and GPI-anchorage to the cell surface membrane. B, Western blotting of HA tagged GPC4 construct expressed in HEK293T cells. The ∼37-kDa band shows the HA tagged N-terminal GPC4 in the reducing condition. PI-PLC treatment in HEK293T cells drives robust release of GPC4; however, GPC4 is detected in media in the absence of PI-PLC. C, Western blotting of concentrated ACMs for endogenous GPC4, IGFBP2, and TSP1, with and without PI-PLC treatment. PI-PLC treatment strongly facilitates release of endogenous GPC4 from astrocytes. D, Biotinylation of surface GPC4 shows that pretreatment of PI-PLC removes endogenous GPC4 from the cell surface of astrocytes. Extended Data Figure 1-1 supports Figure 1.
- Figure 2.
Luciferase assay for quantifying the release of GPC4 from astrocytes. A, Nluc is inserted at the N terminus, after the endogenous signal peptide, to preserve GPC4 trafficking. B, Primary astrocytes were nucleofected with Nluc-GPC4 and treated with and without PI-PLC. Western botting with α-Nluc antibody showed the expected 20 kDa size shift in the N-terminal fragment. PI-PLC treatment facilitates the release of Nluc-GPC4, confirming the GPI-anchorage of the construct. C, Representative trace of one experiment showing the linear kinetics of GPC4 release from astrocyte culture (R2 = 0.998). Error bars indicate the standard error of the mean. D, Astrocytes expressing Nluc-GPC4 were incubated in fresh media with and without PI-PLC for 3 h, and Nluc signal was measured in the cell lysate and media. Nluc signal was normalized to untreated lysate conditions for each biological replicate. PI-PLC treatment resulted in the decrease in the luciferase activity of the cell lysate and the corresponding increase in the activity in the media. These data show Nluc-GPC4 is quantitative in measuring released versus surface pools of GPC4. Error bars indicate 95% CI of the mean here and in following graphs. The requirement of the GPI-anchorage for PI-PLC-dependent release of GPC4 is shown in Extended Data Figure 2-1. E, The release rate (media over lysate activity) of Nluc-GPC4 and Nluc-Prion was normalized to Nluc-GPC4 release. Nluc-GPC4 is released ∼2-fold more than Nluc-Prion (t test p < 0.0001, Cohen’s d = 3.59); ***p < 0.001.
- Figure 3.
Characterizing the release mechanism of GPC4 using Nluc-GPC4. A, After biotinylating the surface proteins of astrocytes expressing Nluc-GPC4, cells were further incubated in the fresh medium for 5 h. The released Nluc-GPC4 is collected from the medium, and the ratio of biotinylation of the released Nluc-GPC4 was measured (luciferase activity of avidin pulldown over medium input). As a control for the release from the cell surface, PI-PLC were treated during the incubation time. Both basal and PI-PLC induced GPC4 release occur from the cell surface. GPC4sec, which does not contain the GPI-anchor signal peptide and is thus not GPI-anchored, is directly secreted without membrane attachment and lacks biotinylation (basal vs PI-PLC ANOVA p = 0.0079, Cohen’s d = 1.012). B, Ultracentrifugation of ACMs containing Nluc-GPC4 was used to test for GPC4 association with extracellular vesicles as a potential release mechanism. Supernatant (soluble) fraction and pellet (vesicular) fraction Nluc were normalized to input Nluc luminescence. Nluc-GPC4 did not associate with the pellet (vesicular) fraction. An exosome marker, CD9, was enriched in the pellet (vesicular) fraction (t test p < 0.0001, Cohen’s d = 7.829). C, alpha toxin (AT), which binds to the glycan core of GPI-anchors, was used to pull down GPI-anchored proteins released through lipase activity. PI-PLC, lipase released Nluc-GPC4 is used as a positive control, while GPC4sec, lacking a GPI anchor, is used as a negative control. Nluc-GPC4 pulled down by AT is normalized to input Nluc-GPC4 measurements. Basal release shows partial binding to AT, indicating that ∼30% of Nluc-GPC4 is released through a lipase mechanism (PI-PLC vs basal release ANOVA p < 0.0001, Cohen’s d = 7.659); **p < 0.01, ***p < 0.001.
- Figure 4.
ADAM9 mediates the release of GPC4 from astrocytes. A, GM6001 (25 μm), an inhibitor of Zn2+-based proteases including the MMP and ADAM family proteases, reduced the release of GPC4 (t test p = 0.0003, Cohen’s d = 1.118). B, shRNA knock-down of Adam9 and not Adam10 reduces the release of Nluc-GPC4 in astrocytes (control vs Adam9 shRNA1 Tukey’s p = 0.0019, Cohen’s d = 1.832; control vs Adam9 shRNA2 Tukey’s p = 0.0010, Cohen’s d = 2.63; control vs Adam10 shRNA1 Tukey’s p = 0.8999, Cohen’s d = 0.291; control vs Adam9 shRNA2 Tukey’s p = 0.8999, Cohen’s d = 0.315). C, qPCR data of astrocyte Adam9 and Adam10 knock-down replicates. Δ-CT between Adam and Gapdh and ΔΔ-CT between control and target shRNA conditions were used to determine the fold change in expression. One biological replicate for Adam9 shRNA2 did not show reductions in Adam9 gene expression. Adam9 shRNA1 one-sample t test p = 0.0190, Cohen’s d = 1.393; Adam9 shRNA2 p = 0.1076, Cohen’s d = 0.799; Adam10 shRNA1 p < 0.0001, Cohen’s d = 10.338; Adam10 shRNA2 p = 0.0444, Cohen’s d = 1.089. D, ADAM9 overexpression, and not ADAM10 overexpression (O/E), is capable of inducing the release of Nluc-GPC4 from astrocytes (control vs ADAM9 O/E ANOVA p < 0.0001, Cohen’s d = 3.482); *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant.
- Figure 5.
The released Nluc-GPC4 includes the heparan sulfate side chain. A, Heparinase II and III treatment of media Nluc-GPC4 reduced the binding of GPC4 to DEAE anion exchange column, by removing the negatively charged heparan sulfate side chains (one-sample t test p = 0.0365, Cohen’s d = 3.482). B, The surface Nluc-GPC4 released by PI-PLC showed similar binding to DEAE as basal release GPC4, indicating that both lipase and protease release mechanisms contain the heparan sulfate modification and are capable of known synaptogenic signaling functions (t test p = 0.9467, Cohen’s d = 0.390); *p < 0.05, ns: not significant.
Tables
Set FP/RP Primer sequence Gene name Vector backbone Set 1 FP CGATGGCCGGCCACCATGGCACGCTTAGGCTTGCTC Gpc4 pCAG GPC4 RP GCACTTCCGAGCAACTTTTC Set 2 FP CCACCATGGCACGCTTAGGCTTGCTCGCGCTCCTCTGCACCCTGGCCGCGCTCAGCGCCTCGCTGCTGGCTGATATCGCGGAGCT Gpc4 (endogenous signal peptide) pCAG GPC4 RP CCGCGATATCAGCCAGCAGCGAGGCGCTGAGCGCGGCCAGGGTGCAGAGGAGCGCGAGCAAGCCTAAGCGTGCCATGGTGGCCGG Set 3 FP TCGCTGCTGGCTGATATGGTCTTCACACTCGAAGATTTCG Nano luciferase pCAG GPC4 RP CTTGAGCTCCGCGATCTTATCGTCGTCATCCTTGTAATCCGCCAGAATGCGTTCGCACAG Set 4 FP ATAGGCTAGCCTCGAGACCATGGCACGCTTAGGCTTG Nluc-GPC4 pZac2.1 Nluc-GPC4 RP CCGGGTCGACTCTAGATTATCTCCACTCTCCCTGCAC Set 4A RP ATCGTCTAGATCATTACTTGTACAGCTCGTCCAT Nluc-GPC4TM Set 5 FP ATTCTGGCGTCCGGAAAAAAGCGGCCAAAGCCTGGA Prion pZac2.1 Nluc-Prion RP TCCAGGCTTTGGCCGCTTTTTTCCGGACGCCAGAAT Set 6 FP AATTCGCGCCTAGGTCCGAACCCCGTTCGGACCTAGGCGCGAATTATTTTT shRNA control pSilencer 1.0 U6 RP AATTAAATTCGCGCCTAGGTCCGAACGGGGTTCGGACCTAGGCGCGAATTGGCC Set 7 FP ACAAGAACCACAATTACTAGATCAAGAGTCTAGTAATTGTGGTTCTTGTATTTTT Adam9 shRNA1 pSilencer 1.0 U6 RP AATTAAAAATACAAGAACCACAATTACTAGACTCTTGATCTAGTAATTGTGGTTCTTGTGGCC Set 7A FP GCCAATAACGTCTGCATTTTCAAGAGAAATGCAGACGTTATTGGCATTTTT Adam9 shRNA2 pSilencer 1.0 U6 RP AATTAAAAATGCCAATAACGTCTGCATTTTCAAGAGAAATGCAGACGTTATTGGCGGCC Set 8 FP GCAAATGATACCCTTACAGTTTCAAGAGAACTGTAAGGGTATCATTTGCATTTTT Adam10 shRNA1 pSilencer 1.0 U6 RP AATTAAAAATGCAAATGATACCCTTACAGTTCTCTTGAAACTGTAAGGGTATCATTTGCGGCC Set 8A FP GGAAGACAGTTCAACCTACGATCAAGAGTCGTAGGTTGAACTGTCTTCCATTTTT Adam10 shRNA2 pSilencer 1.0 U6 RP AATTAAAAATGGAAGACAGTTCAACCTACGACTCTTGATCGTAGGTTGAACTGTCTCCGGCC Set 9 FP GGGGTGTGTAACAGCAATAAG ADAM9 Sequencing primers RP CAAGGGGGACGATTAGGAAG Set 10 FP AACAGGCTTATCGCTATG ADAM10 Sequencing primers RP CTGCTGCTGACTTCTAAT Set 11 FP CTGGAGAAACCTGCCAAGTA GAPDH Sequencing primers RP AGTGGGAGTTGCTGTTGAAG Data Data structure
(normality test)Type of test Power Figure 2D, lysate measurement of Nluc-GPC4 with and without PI-PLC treatment Yes Unpaired t test Cohen’s d = 7.29, p < 0.0001 Figure 2D, media measurement of Nluc-GPC4 with and without PI-PLC treatment Yes Unpaired t test Cohen’s d = 3.00, p < 0.0001 Figure 2E, overnight release of GPC4 compared with Prion, normalized to lysate Yes Unpaired t test Cohen’s d = 3.59, p < 0.0001 Figure 3A, live surface Biotinylation of GPC4 release over time Yes ANOVA; Tukey’s multiple
comparisonsCohen’s d = 1.012, p < 0.0001 Figure 3B, ultracentrifugation of Nluc-GPC4 in ACMs Yes Unpaired t test Cohen’s d = 7.829, p < 0.0001 Figure 3C, pulldown of Nluc-GPC4 by α Toxin Yes ANOVA; Tukey’s multiple
comparisonsCohen’s d = 7.659, p < 0.0001 Figure 4A, reduction of Nluc-GPC4 by protease inhibitor GM6001 Yes Unpaired t test Cohen’s d = 1.118, p = 0.0003 Figure 4B, reduction of Nluc-GPC4 release by Adam9 shRNA1 KD Yes ANOVA; Tukey’s multiple
comparisonsCohen’s d = 1.832, p < 0.0001 Figure 4B, reduction of Nluc-GPC4 release by Adam9 shRNA2 KD Yes ANOVA; Tukey’s multiple
comparisonsCohen’s d = 2.63, p < 0.0001 Figure 4C, fold gene expression by qPCR for shRNA KD experiments Adam9 shRNA1 Yes One-sample t test Cohen’s d = 1.393, p = 0.0190 Figure 4C, fold gene expression by qPCR for shRNA KD experiments Adam9 shRNA2 Yes One-sample t test Cohen’s d = 0.799, p = 0.1076 Figure 4C, fold gene expression by qPCR for shRNA KD experiments Adam10 shRNA1 Yes One-sample t test Cohen’s d = 10.33, p < 0.0001 Figure 4C, fold gene expression by qPCR for shRNA KD experiments Adam10 shRNA2 Yes One-sample t test Cohen’s d = 1.089, p = 0.0444 Figure 4D, induction of Nluc-GPC4 release by Adam9 Overexpression Yes ANOVA; Tukey’s multiple
comparisonsCohen’s d = 3.482, p < 0.0001 Figure 5A, heparinase treatment of Nluc-GPC4 vs Control Yes One-sample t test Cohen’s d = 2.93, p = 0.0365 Figure 5B, pulldown of Nluc-GPC4 by α Toxin in PI-PLC control vs basal release conditions Yes Unpaired t test Cohen’s d = 0.390, p = 0.9467 Extended Data Figure 1-1, cytotoxicity assay of PI-PLC treatment Yes ANOVA; Tukey’s multiple
comparisonsCohen’s d = 0.907, p = 0.964 Extended Data Figure 1-1, cytotoxicity assay of Edelfosine treatment Yes ANOVA; Tukey’s multiple
comparisonsCohen’s d = 2.870, p < 0.0001 Extended Data Figure 2-1, lysate measurement of Nluc-GPC4-TM with and without PI-PLC treatment Yes Unpaired t test Cohen’s d = 0.49, p = 0.19 Extended Data Figure 2-1, media measurement of Nluc-GPC4 = TM with and without PI-PLC treatment Yes Unpaired t test Cohen’s d = 0.50, p = 0.27
Extended Data Figure 1-1
Extended data supporting Figure 1C. Cell toxicity assay of WT astrocytes with and without overnight PI-PLC treatment. Higher levels of bioluminescence indicate a loss of cell integrity and leakage of intracellular adenylate kinase into the media. Edelfosine treatment (50 μm, 1 h) is included as a positive control for cell death. Tukey’s multiple comparison Ctrl versus PI-PLC p = 0.964, Cohen’s d = 0.907. Tukey’s multiple comparison Ctrl versus Edelfosine p < 0.0001, Cohen’s d = 2.870. Download Figure 1-1, TIF file.
Extended Data Figure 2-1
Extended data supporting Figure 2D. Astrocytes expressing Nluc-GPC4-TM was incubated in fresh media with and without PI-PLC for 3 h, and Nluc signal was measured in the cell lysate and media. Nluc signal was normalized to untreated lysate conditions for each biological replicate. PI-PLC treatment did not result in the release of GPC4-TM from astrocytes. Lysate +/– PI-PLC unpaired t test p = 0.19, Cohen’s d = 0.49. Media +/– PI-PLC unpaired t test p = 0.27, Cohen’s d = 0.50. Download Figure 2-1, TIF file.
In this issue
