MARCH Proteins Mediate Responses to Antitumor Antibodies - PubMed (original) (raw)
MARCH Proteins Mediate Responses to Antitumor Antibodies
Jailal N Ablack et al. J Immunol. 2020.
Abstract
CD98, which is required for the rapid proliferation of both normal and cancer cells, and MET, the hepatocyte growth factor receptor, are potential targets for therapeutic antitumor Abs. In this study, we report that the antiproliferative activity of a prototype anti-CD98 Ab, UM7F8, is due to Ab-induced membrane-associated ring CH (MARCH) E3 ubiquitin ligase-mediated ubiquitination and downregulation of cell surface CD98. MARCH1-mediated ubiquitination of CD98 is required for UM7F8's capacity to reduce CD98 surface expression and its capacity to inhibit the proliferation of murine T cells. Similarly, CD98 ubiquitination is required for UM7F8's capacity to block the colony-forming ability of murine leukemia-initiating cells. To test the potential generality of the paradigm that MARCH E3 ligases can mediate the antiproliferative response to antitumor Abs, we examined the potential effects of MARCH proteins on responses to emibetuzumab, an anti-MET Ab currently in clinical trials for various cancers. We report that MET surface expression is reduced by MARCH1, 4, or 8-mediated ubiquitination and that emibetuzumab-induced MET ubiquitination contributes to its capacity to downregulate MET and inhibit human tumor cell proliferation. Thus, MARCH E3 ligases can act as cofactors for antitumor Abs that target cell surface proteins, suggesting that the MARCH protein repertoire of cells is a determinant of their response to such Abs.
Copyright © 2020 by The American Association of Immunologists, Inc.
Figures
Figure 1:. Anti-CD98 limits proliferation by MARCH1 mediated ubiquitination and downregulation of CD98.
(A) Inset: Schematic of antibody treatment regimen. Cells are treated 1) with either anti-CD98 (Clone UM7F8) or isotype control (IgG) and 2) with crosslinking anti-mouse IgG F(ab’)2 or non-crosslinking anti-mouse IgG F(ab’). The mean change in cell number relative to IgG control is depicted ± SEM from two experiments performed in triplicate of RAMOS cells treated as outlined above. (B) Histograms depicting surface expression of CD98 on RAMOS cells treated as in (A). (C) Dose response of UM7F8 downregulation of surface CD98. RAMOS cells were treated with either clone UM7F8 or Control IgG in combination with F(ab’)2 or F(ab’). Surface expression of CD98 is depicted ± SEM from three independent experiments. (D) Endogenous CD98 immunoprecipitated from RAMOS and Jurkat cells treated with either UM7F8 or IgG control and in the presence of F(ab’)2 was assayed for recovered CD98 or covalent modification with ubiquitin by immunoblotting. (E) Jurkat cells stably expressing HA-tagged 8KR-CD98, a mutant of CD98 resistant to ubiquitination on cytoplasmic K residues (8KR), or its wild type counterpart (wt) were treated as in (D) and HA-tagged CD98 8KR or wt was immunoprecipitated and probed for recovered HA-tagged CD98 or covalent modification with ubiquitin by immunoblotting. (F) OT1 T cells were stimulated with cognate peptide (SIINFEKL) antigen and were then transduced with retroviruses encoding human CD98 and a Thy1.1 marker. Transduced cells were treated with UM7F8 or IgG as in (D) and surface expression of human CD98 was detected by flow cytometry. The mean percent change in surface expression of ectopically expressed human CD98 relative to IgG control from three independent experiments (n=3 donor mice) is depicted ± SEM. (G) Primary T cells from March1 −/− mice or March1 +/+ littermate controls were stimulated, transduced with human CD98 and treated with anti-CD98 or control as in (F). The mean percent change in human CD98 expression relative to untreated cells ± SEM from three independent experiments (N=3 littermate pairs) is depicted. *** denotes P < 0.001 by one way ANOVA with Bonferroni’s posttest, NS denotes not significant. * denotes P < 0.05 by Student’s T test, NS denotes not significant. Raw data appear in Supplemental Figure S2.
Figure 2:. Anti-CD98 blocks CD98-dependent proliferation of T cells by promoting CD98 ubiquitination.
(A) Schematic of retrogenic derivation of OT1 T cells with customized human CD98 genotypes. Briefly, donor OT1 mice with conditional knockout of endogenous mouse CD98 in all T cells (Slc3a2 f/f, CD4 CRE+, OT1+) were treated with 5-Flurouracil to deplete cycling cells enriching hematopoietic stem cells. Five days after 5FU treatment bone marrow was isolated and cultured in the presence of cytokines to support stem cell cycling (SCF, IL-3, IL6). After 24 hours in culture, cells were transduced with retroviruses expressing either 8KR or wt human CD98 with a Thy1.1 surrogate marker. Transduced cells were transplanted into lethally irradiated recipient mice. Approximately 8 weeks after transplant CD8+ T cells derived from donor bone marrow were detectable in circulation. (B) CD98 phenotype of purified naïve OT1 T cells from the spleens of retrogenic mice used for experiments. (C) Purified OT1 T cells of the indicated phenotypes confirmed in (B) were labeled with CFSE and stimulated with SIINFEKL peptide. Representative histograms depicting CFSE dilution of the indicated stimulated human CD98 positive, mouse CD98 negative, CD8+ OT1 T cells stimulated with SIINFEKL peptide, in the presence of UM7F8 (black trace) or IgG (grey trace) and anti-mouse F(ab’)2. (D) The mean proliferation index ± SEM from three independent experiments (n=3 donor mice) is depicted. * denotes P < 0.01, NS denotes not significant by one way ANOVA and Bonferroni post-test.
Figure 3:. Anti-CD98 induced ubiquitination blocks CD98 dependent colony formation of cancer initiating AML cells.
(A) Overview of CD98 customized MLL-AF9/NRAS AML. Briefly, hematopoietic stem cells (KLS) from donor mice with inducible global knockout of endogenous mouse CD98 were transduced with retroviruses encoding oncogenes (MLL-AF9 fusion and NRAS) sufficient to initiate AML. Retroviruses included a surrogate marker to select transduced cells (human NGFR and YFP for MLL-AF9 and NRAS, respectively). Cancer initiating cells were transplanted into recipient mice to initiate AML. After approximately 21 days established AML cells were recovered and transduced with retroviruses encoding either the 8KR mutant of human CD98 (8KR), wt human CD98 (hCD98) or empty EGFP vector. Viable, EGFP+ AML initiating cells were plated in methyl cellulose media containing 4 hydroxy tamoxifen (4OHT) or ethanol (Vehicle), and treated with UM7F8 or IgG and Anti-mouse F(ab’)2. (B) Colony formation assays. After 5 days in culture colonies were enumerated. Colony formation in the presence of UM7F8 expressed as a percent of IgG treatment from two independent experiments performed in triplicate is shown ± SEM. ** denotes P < 0.01, NS denotes not significant by one-way ANOVA and Bonferroni post test. (C) Verification of human and mouse CD98 phenotypes in established AML (EGFP+) after treatment with anti-CD98 and subsequent clonal expansion. Raw colony counts appear in Supplemental Figure S2 panel F.
Figure 4:. MET is regulated by MARCH1/4/8 through ubiquitination of membrane proximal cytoplasmic lysine residues.
(A) Surface expression of endogenous MET in HeLa cells transfected with either EGFP, MARCH4-EGFP, MARCH1-EGFP, MARCH8-EGFP or catalytically inactive mutants of MARCH1 or 8-EGFP. Dot plots are representative of three independent experiments. (B) Surface expression of endogenous MET in HeLa cells transfected with MARCH4-EGFP, MARCH1-EGFP, MARCH8-EGFP or a catalytically inactive mutant of MARCH8-EGFP (C.I.) relative to EGFP control as a function of MARCH expression (EGFP fluorescence) mean±SEM from three independent experiments is shown. (C) Differential MARCH sensitivity of CD98 and MET in HeLa cells. Cells transfected with cDNA encoding either MARCH4-EGFP or MARCH8-EGFP as in A were assayed for cell surface CD98 or MET Mean% of control±SEM from three independent experiments for MET and two independent experiments is shown for CD98. (D) as in (B) depicting CD98 or MET expression as a function of MARCH4 or MARCH8 expression. Note that the CD98 is resistant to MARCH4. (E) Amino acid alignment of the membrane proximal region of mammalian MET proteins. Conserved lysines are shaded. All four conserved membrane proximal lysines of human MET were mutated to arginine to generate 4KR-MET (marked by asterisks). (F) Surface expression of MET as a percentage of EGFP control from MET-knock out HeLa-FlpIn TREX cells, rescued with a single inducible integration of either HA-tagged 4KR MET or HA-tagged MET ± SEM from three independent experiments is depicted. (G) Surface expression of endogenous MET from MKN45 cells 72 hours post-transfection with either MARCH8 siRNA or control (non-targeting) siRNA. (H) quantification of MET surface expression from 3 independent experiments performed as in (G). (I) Validation of MARCH8 knockdown in MKN45 cells treated with MARCH8 or control siRNAs. Mean relative MARCH8 expression 72 hours post transfection from three independent experiments is depicted. *** denotes significantly different from EGFP control, P < 0.01, by one way ANOVA and Dunnet’s post test. Raw data appear in Supplemental Figure S3.
Figure 5:. Emibetuzumab induces MARCH8-dependent ubiquitination and downregulation.
(A) Downregulation of endogenous MET with Emibetuzumab in HeLa cells. Histograms depicting MET surface expression 24 hours post-treatment with Emibetuzumab (Emi) or human IgG4 isotype control (IgG), with and without anti-human F(ab’)2. (B) Quantification of Emibetuzumab induced downregulation of MET in HeLa cells. Cells were treated and surface MET expression determined as in (A). 24 hours post-treatment the percent change in MET relative to hIgG4 isotype control with or without anti-human F(ab’)2 was determined. Mean percent change in MET surface expression ± SEM from three independent experiments is depicted. (C) 4KR-MET is partially resistant to Emibetuzumab induced downregulation. 24 hours post-treatment surface expression of MET or 4KR-MET was determined as in (B). Mean percent change in surface expression of MET wt or 4KR-MET± SEM from three independent experiments is depicted. (D) A Kinase Dead+ 4KR mutant of MET is resistant to Emibetuzumab induced downregulation. as in (C) with the indicated MET mutant or wt. Mean percent change in MET surface expression± SEM from three independent experiments is depicted. (E) Immunoblot of Immunoprecipitated HA-tagged MET for covalent modification with ubiquitin from HeLa cells stably expressing HA-tagged 4KR-MET (4KR) or its wt counterpart 24 hours post-treatment with 100nM Emibetuzumab or hIgG4 isotype control. (F) Emibetuzumab induced downregulation is impaired by MARCH8 knockdown in MKN45 cells. As in (B) comparing the mean percent change in endogenous MET surface expression of MKN45 cells 72 hours post-transfection with the indicated siRNA from 3 independent experiments is depicted. (G) 4KR-MET renders MKN45 cells resistant to the anti-proliferative effects of Emibetuzumab. The Emibetuzumab sensitive cell line, MKN45, was transfected with 4KR-MET-EGFP or MET-EGFP. 24 hours post-transfection viable, EGFP positive cells were sorted, plated then treated with either Emibetuzumab (Emi), hIgG4 Isotype control (IgG). 96 hours post treatment viable cell number was determined by MTS assay. Mean relative cell number compared to untreated cells ± SEM from four independent cultures is depicted. * denotes P < 0.01 by one way ANNOVA and Bonferroni post-test. Raw data appear in Supplemental Figure S4.
References
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Medical
Molecular Biology Databases
Miscellaneous