Exploitation of Sulfated Glycosaminoglycan Status for Precision Medicine of Triplatin in Triple-Negative Breast Cancer

Abstract

Triple-negative breast cancer (TNBC) is a subtype of breast cancer lacking targetable biomarkers. TNBC is known to be most aggressive and when metastatic is often drug-resistant and uncurable. Biomarkers predicting response to therapy improve treatment decisions and allow personalized approaches for patients with TNBC. This study explores sulfated glycosaminoglycan (sGAG) levels as a predictor of TNBC response to platinum therapy. sGAG levels were quantified in three distinct TNBC tumor models, including cell line-derived, patient-derived xenograft (PDX) tumors, and isogenic models deficient in sGAG biosynthesis. The in vivo antitumor efficacy of Triplatin, a sGAG-directed platinum agent, was compared in these models with the clinical platinum agent, carboplatin. We determined that >40% of TNBC PDX tissue microarray samples have high levels of sGAGs. The in vivo accumulation of Triplatin in tumors as well as antitumor efficacy of Triplatin positively correlated with sGAG levels on tumor cells, whereas carboplatin followed the opposite trend. In carboplatin-resistant tumor models expressing high levels of sGAGs, Triplatin decreased primary tumor growth, reduced lung metastases, and inhibited metastatic growth in lungs, liver, and ovaries. sGAG levels served as a predictor of Triplatin sensitivity in TNBC. Triplatin may be particularly beneficial in treating patients with chemotherapy-resistant tumors who have evidence of residual disease after standard neoadjuvant chemotherapy. More effective neoadjuvant and adjuvant treatment will likely improve clinical outcome of TNBC.

Figure 1.. Sulfated glycosaminoglycans are a potential target for Triplatin in TNBC.

A. IHC H-scores of TNBC PDX samples on TMA (n= 42) stained with Ab 10E4. Each data point represents the mean H-score for duplicate samples. B. Representative images of Ab 10E4 staining of TNBC PDX TMA samples.

Figure 2.. sGAG expression levels affect cellular accumulation, cytotoxicity, and tumor uptake of Triplatin.

A. Structures of Triplatin and carboplatin B. DFT optimized structure for the interaction of [Pt(NH₃)₄]²⁺, the central platinum of Triplatin, with sulfated monosaccharide, GlcNS(6S). C. Mutation and phenotype of GAG- and sulfation-deficient CHO carcinoma cell lines D. Quantification of the total cell surface sGAG levels in CHO cell lines using the Blyscan sGAG quantitation assay (n=5 biological replicates). E. Relative levels of N-sulfated HS (Ab 10E4) expression by flow cytometry. Histogram representative of 3 independent experiments. F. ICP-MS quantification of Triplatin and carboplatin cellular uptake. G-H. Analysis of cell proliferation after 15 min treatments with the indicated concentrations of Triplatin or carboplatin. % Confluence was quantified every 2 h for 210 h using live cell analysis (Incucyte). Each data point represents the mean ± SD (n=2 biological replicates with 4 technical replicates each). IC₅₀ calculations are shown in G. I. Right panel, Image of a mouse injected with wt CHO and A475. Middle panel, IHC image of N-sulfated HS (Ab 10E4) expression in CHO wt and 745 tumors and staining were quantified (left panel). Each bar represents the mean ± SD (n=3). J, K. ICP-MS analysis of Pt tumor accumulation in CHO wt and A-745 tumors 24h after treatment with Triplatin (0.3 mg/kg; n=7) and carboplatin (40 mg/kg; n=10). Statistical analysis: D-F, One-way ANOVA, Tukey’s posttest; G unpaired student’s t-test; J and K, paired student’s t-test. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 3.. Triplatin inhibits tumor and metastatic growth in a high sGAG-expressing MDA-MB-231-luc model of TNBC.

A. Representative images of IHC analysis of N-sulfated HS (Ab 10E4) reactivity in MDA-MB-231-luc tumors with or without heparinase III digestion. B. Quantitation of Ab 10E4 IHC (n=3). C. Analysis of primary tumor growth. Arrows, drug treatment [Triplatin (0.3mg/kg), carboplatin (40 mg/kg), or vehicle (saline)] given i.p. on days 1, 5, and 9 after randomization (day 16). D. Tumor weight at endpoint. E. Survival analysis: MDA-MB-231BR cells were injected into left ventricle of female NSG mice (day 0). Animals were randomized on day 10. Drug treatment [Triplatin (0.3mg/kg), cisplatin (3 mg/kg), or vehicle (saline)] given i.p. on days 1, 5 and 9 after randomization. F. Quantification of overall metastases in vivo on day 21 (total flux). G. Quantification of bone metastases ex vivo on day 21 (total flux). H. Quantification of liver metastases ex vivo on day 21 using HLA staining. Statistical analysis: B, F and G, student’s t-test; C, D and E, One-way ANOVA; Tukey’s posttest *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 4.. sGAG levels affect the survival benefit of Triplatin treatment.

A. Flow cytometry analysis shows levels of N-sulfated HS (Ab 10E4) in MDA-MB-231 XYLT1/2 KO cell lines is significantly decreased compared to wt and CRIPSR control MDA-MB-231 cells. B. ICP-MS quantification of Triplatin and carboplatin cellular uptake in vitro. C. ICP-MS analysis of Pt tumor accumulation in parental MDA-MB-231 tumors and MDA-MB-231 XYLT1/2 KO-F and KO-G tumors, 24 h after treatment with Triplatin (0.3 mg/kg). D, E. Comparison of % tumor growth inhibition (TGI) of MDA-MB-231 wt and MDA-MB-231 XYLT1/2 KO-F and KO-G tumors (D) and survival benefit (E) of female NSG mice treated with 0.3 mg/kg Triplatin given i.p. on days 1, 5 and 9 after randomization. Statistical analysis; A-B, One-way ANOVA, Tukey’s posttest. C, paired student’s t-test. D, Two-way ANOVA, Dunnett’s multiple comparison test E, Survival curve analysis, Log-rank (Mantel-Cox) test. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Effect size, Glass’s ∆ = mean difference between Triplatin and vehicle divided by the standard of deviation of the vehicle group.

Figure 5.. Triplatin efficacy against PDX TNBC primary tumor growth correlates with sGAG levels.

A IHC H-scores of Ab 10E4 reactivity in TNBC PDX tumors. H-scores expressed as mean ± SD (bars). Statistical significance evaluated by one-way ANOVA (n=3). B Representative images of Ab 10E4 IHC staining. Scale bar represents 50 µm. (C-E) Analysis of primary tumor growth; 1×10⁵ cells injected into the T4 mammary gland of female NSG mice. Animals randomized and injected i.p. Q4Dx3 with Triplatin (0.3mg/kg), carboplatin (40 mg/kg), or vehicle (saline). Tumor growth measured by caliper and IVIS imaging. Statistical analysis: two-way ANOVA (n=5). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001

Figure 6.. Efficacy of Triplatin against PDX TNBC metastatic growth correlates with sGAG levels.

WHIM30, WHIM2, or UCD52 PDX cells (5×10⁵) were injected into left ventricle of female NSG mice. Animals were randomized and injected i.p. Q4Dx3 with Triplatin (0.3mg/kg), carboplatin (40 mg/kg), or vehicle (saline). A Total metastases were imaged in vivo. B Total metastases were quantified for each model. C-D Tumors and organs were harvested for ex vivo imaging and quantified. E Light photographs of WHIM30 liver metastases. Visible metastases highlighted using false coloration. F Light photographs of WHIM2 ovary metastases. Statistical significance was determined by one-way ANOVA, Tukey’s posttest. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001