An amino acid transporter subunit as an antibody-drug conjugate target in colorectal cancer

Abstract

Background: Advanced colorectal cancer (CRC) is difficult to treat. For that reason, the development of novel therapeutics is necessary. Here we describe a potentially actionable plasma membrane target, the amino acid transporter protein subunit CD98hc.

Methods: Western blot and immunohistochemical analyses of CD98hc protein expression were carried out on paired normal and tumoral tissues from patients with CRC. Immunofluorescence and western studies were used to characterize the action of a DM1-based CD98hc-directed antibody-drug conjugate (ADC). MTT and Annexin V studies were performed to evaluate the effect of the anti-CD98hc-ADC on cell proliferation and apoptosis. CRISPR/Cas9 and shRNA were used to explore the specificity of the ADC. In vitro analyses of the antitumoral activity of the anti-CD98hc-ADC on 3D patient-derived normal as well as tumoral organoids were also carried out. Xenografted CRC cells and a PDX were used to analyze the antitumoral properties of the anti-CD98hc-ADC.

Results: Genomic as well proteomic analyses of paired normal and tumoral samples showed that CD98hc expression was significantly higher in tumoral tissues as compared to levels of CD98hc present in the normal colonic tissue. In human CRC cell lines, an ADC that recognized the CD98hc ectodomain, reached the lysosomes and exerted potent antitumoral activity. The specificity of the CD98hc-directed ADC was demonstrated using CRC cells in which CD98hc was decreased by shRNA or deleted using CRISPR/Cas9. Studies in patient-derived organoids verified the antitumoral action of the anti-CD98hc-ADC, which largely spared normal tissue-derived colon organoids. In vivo studies using xenografted CRC cells or patient-derived xenografts confirmed the antitumoral activity of the anti-CD98hc-ADC.

Conclusions: The studies herewith reported indicate that CD98hc may represent a novel ADC target that, upon well-designed clinical trials, could be used to increase the therapeutic armamentarium against CRC.

Keywords: Antibody–drug conjugates; CD98hc; Colorectal cancer; Targeted therapy.

Fig. 1

Expression of CD98hc in normal and tumoral tissues from patients with CRC. A Expression of CD98hc in paired samples (normal and tumoral colon tissue) of patients with CRC. CD98hc was analyzed by Western using the anti-CD98hc^V509 antibody. ERK1/2 was used as a loading control. N = normal tissue, T = tumoral tissue. The position of the molecular weight markers is shown at the right. B and C represent the relative protein levels in arbitrary units of CD98hc of the data shown in A. P value in B was calculated using Mann Whitney U test. Quantitation of CD98hc was made as described in the methods section. D Immunohistochemical staining of CD98hc in paired colon samples (normal and tumoral tissue) stained with the anti-CD98hc^V509 antibody. Magnification: 40X. E Immunohistochemical staining of CD98hc in a sample of a patient which includes tumoral as well as normal tissue. Magnification: 20 X

Fig. 2

Internalization of an antibody against CD98hc in colon cancer cells. A Interaction of the anti-CD98hc^ECTO with native cell surface CD98hc. One mg of tissue extract from three paired patient samples (normal and tumoral) was immunoprecipitated with the anti-CD98hc^ECTO antibody and the immunocomplexes and extracts of these samples were analyzed by Western blot with the anti-CD98hc^V509 antibody. Cell extracts (50 µg) from the same samples were also loaded. Calnexin was used as a loading control for the cell extract samples. B Expression of CD98hc in a panel of colon cancer cell lines. Colon cancer cell lines were lysed, and cell extracts (20 µg) used to identify CD98hc by Western blot with the anti-CD98hc^V509 antibody. Calnexin was used as a loading control. C Cell surface immunoprecipitation of CD98hc. Four CRC cell lines were in vivo treated or not with 10 nM of anti-CD98hc^ECTO for 30 min at 37°C. Cells were lysed and cell extracts precipitated with protein A-sepharose. CD98hc in those immunoprecipitates was analyzed by Western blot with the anti-CD98hc^V509 antibody. D Internalization of the anti-CD98hc^ECTO in HT29 cells, analyzed by immunofluorescence. The cells were seeded on coverslips and treated with 10 nM of anti-CD98hc^ECTO for 30 min, 6, 12 and 24 h. The images at the right correspond to magnifications of a cell present in the images obtained at 24 h. The colocalization of CD98hc and LAPM1 is show in the merged images

Fig. 3

Anti-proliferative activity of the anti-CD98hc ADC in colon cancer cell lines. A Dose–response analyses of anti-CD98hc-DM1 in a panel of CRC cell lines. Cells were treated with the ADC for four days at the indicated doses. The data are plotted as the percentage of MTT metabolization with respect to control. Results are shown as the mean ± SD of quadruplicates of an experiment repeated two times. The table on the right indicates the IC₅₀ (nM) of anti-CD98hc-DM1 for each cell line. B Effect of nude anti-CD98hc and anti-CD98hc-DM1 on the proliferation of CRC cell lines. Cells were treated with anti-CD98hc or anti-CD98hc-DM1 10 nM for four days. The data are plotted as the percentage of MTT metabolization with respect to control. Results are shown as the mean ± SD of triplicates of an experiment repeated twice. C and D HT29 (C) and HCT116 (D) cells were treated with the doses indicated of anti-CD98hc, anti-CD98hc-DM1 or DM1 for four days. The data are plotted as the percentage of MTT metabolization with respect to control. Results are shown as the mean ± SD of triplicates of an experiment repeated twice. E HT29 cells were infected with lentivirus containing the shRNA control (sh-Control) or the shRNA sequences targeting CD98hc (sh-CD98hc #3 and sh-CD98hc #7). Knockdown efficiency was verified by Western with the anti-CD98hc^V509 antibody. Calnexin was used as a loading control. F Knockout of CD98hc in HT29 cells by CRISPR/Cas9. Parental HT29 cells and ten different clones knocked out for CD98hc were lysed and the levels of expression of CD98hc analyzed by Western blot with the anti-CD98hc^V509 antibody. Calnexin and tubulin were used as a loading controls. G Impact of CD98hc knockdown on the antiproliferative effect of anti-CD98hc-DM1. HT29 cells were treated with anti-CD98hc-DM1 (1 nM and 5 nM) for four days. The data are plotted as the percentage of MTT metabolization with respect to control. Results are shown as the mean ± SD of triplicates of an experiment repeated twice. H Dose–response analyses of the effect of anti-CD98hc-DM1 in the proliferation of parental and CD98hc CRISPR #5 and #11 HT29 cells. Results are shown as the mean ± SD of quadruplicates of an experiment repeated three times

Fig. 4

The anti-CD98hc-DM1 antibody induces cell cycle arrest in mitosis and mitotic catastrophe. A Effect of anti-CD98hc-DM1 (10 nM, 24 h) on the morphology of HT29 and HCT116 cells grown as monolayers. B Effect of anti-CD98hc-DM1 (10 nM, 24 h) on the distribution of the different cell cycle phases (G0/G1, S and G2/M) in HT29 and HCT116 cell lines. C Effect of anti-CD98hc-DM1 on the levels of expression and phosphorylation of proteins implicated in cell cycle progression. HT29 and HCT116 cells were treated with anti-CD98hc-DM1 (10 nM) and lysed at the indicated times. The levels of expression or phosphorylation of the different proteins studied were analyzed by Western. GAPDH was used as loading control. D Action of anti-CD98hc-DM1 on spindle assembly and organization. HT29 cells seeded on coverslips were treated with CD98hc-DM1 (10 nM) for 48 h, fixed and stained. β-Tubulin in green and DAPI in blue. Scale bars = 7.5 µm. E Percentage of viable (Annexin V-negative/PI-negative) and non-viable HT29 and HCT116 cells after 72 h of treatment with 10 nM anti-CD98hc-DM1. F Effect of anti-CD98hc-DM1 on the levels of different apoptosis-related proteins

Fig. 5

The anti-CD98hc-DM1 has anti-proliferative effect on tumoral patient-derived organoids. A Levels of expression of CD98hc in normal and tumoral PDOs. Levels of expression of CD98hc and CA2 (normal colon marker) were analyzed by Western blot with the anti-CD98hc^V509 antibody. GAPDH was used as a loading control. B Immunofluorescence detection of CD98hc in tumoral PDOs. Tumoral PDOs were incubated with 10 nM of the anti-CD98hc^ECTO antibody for 1.5 h and the subcellular distribution of CD98hc was analyzed by immunofluorescence. Scale bar = 50 µm. The drawing on the left represents an organoid and the red dashed lines indicate the organoid regions where images were taken. C Evaluation of anti-CD98hc ADC on the cell viability of different tumoral PDOs. The tumoral PDOs were treated with 10 nM of anti-CD98hc-DM1 for 4 days and the cell viability was analyzed as indicated in the methods sections. Results are shown as the fold-change respect to untreated organoids ± SD of triplicates of an experiment repeated twice. D Dose–response analyses of the effect of anti-CD98hc-DM1 on cell viability of paired normal and tumor PDO (#175), analyzed at four days. Results are shown as the fold-change respect to the untreated organoids ± SD of triplicates of an experiment repeated twice. P values were calculated using Student t test (two-sided). E Immunofluorescence analyses of the effect of the anti-CD98hc or anti-CD98hc-DM1 on a tumoral PDO. Organoids were treated with anti-CD98hc or anti-CD98hc-DM1 at 10 nM for 4 days and the effect on cell morphology was evaluated by inmunofluorecescence. CD98hc is stained in green, nuclei are stained in blue. Scale bars = 25 µm. Arrows indicate nuclear fragmentation

Fig. 6

The anti-CD98hc-DM1 ADC has antitumoral activity on in vivo CRC models. A Evaluation of the antitumoral effect of anti-CD98hc-DM1 on tumor growth in nude mice implanted with HT29 cells. Arrows indicate days of administration of anti-CD98hc-DM1 (15 mg/Kg). Data are plotted as mean tumor volumes ± SEM. P values were calculated using Student’s t test (two-sided). B Expression of anti-CD98hc-DM1 and CD98hc in the tumors from mice. Tumor samples were obtained on day 21 after initiation of treatments (seven days after the last treatment). Tissue extracts of the tumors were used to analyze the levels of expression of anti-CD98hc-DM1 and CD98hc by Western blot with anti-DM1 and anti-CD98hc^V509 antibodies, respectively. GAPDH was used as a loading control. C Immunohistochemical staining of CD98hc with the anti-CD98hc^V509 antibody in normal and tumoral tissue of patient BT6224 used to generate the PDX. Magnification: 40 X. D Analysis of expression of CD98hc by Western blot with the anti-CD98hc^V509 antibody in a PDX derived from BT6224, and in HT29 cells. Calnexin was used as a loading control. E Effect of anti CD98hc-DM1 on tumor growth in nude mice carrying BT6224-derived PDXs. Arrows indicate days of administration of anti-CD98hc-DM1 (15 mg/Kg). Data are plotted as mean tumor volumes ± SEM. P values were calculated using Student t test (two-sided). F Evaluation of anti-CD98hc-DM1 and CD98hc in tumoral samples by western blot with anti-DM1 and anti-CD98hc^V509 antibodies, respectively. GAPDH was used as a loading control