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.

This article is featured in Highlights of This Issue, p. 243

Triple-negative breast cancer (TNBC) represents 15% to 20% of breast cancers and is defined by the lack of estrogen- and progesterone-receptor expression and absence of overexpression of HER2. Chemotherapy remains the primary treatment, and the standard of care is to use taxanes and anthracyclines as first-line therapy for both early-stage and metastatic TNBCs (1, 2). TNBC is known to be the most aggressive subtype, recurs in half of patients after curative treatment, and rapidly develops drug resistance and metastases. The average life expectancy of approximately 1 year after recurrence underscores the need for novel therapeutic regimens (3).

Advancements in patient survival continue through new approaches to precision medicine and biomarker research, such as the recently FDA-approved pembrolizumab (PD-L1 antibody), for use in early-stage TNBC in combination with standard neoadjuvant chemotherapy (NACT); the Trop-2 antibody–drug conjugate, sacituzumab govitecan, for use in metastatic TNBC; and a combination of atezolizumab (PD-L1 antibody) and nab-paclitaxel, also for treatment of metastatic TNBC (4–6). Other strategies aim to optimize patient responses and survival through new combinational regimens (7). Recent clinical trials for early-stage TNBC show that addition of carboplatin or carboplatin/pembrolizumab to standard NACT regimens significantly increases the number of patients who achieve a pathological complete response (pCR), which is a known surrogate for better survival (7, 8). For patients with metastatic TNBC harboring BRCA mutations, the phase III TNT trial showed patients who received carboplatin had higher overall response rates (68% vs. 33%; P = 0.01) than those who received docetaxel (9). Biomarkers predicting response to therapy will continue to improve treatment decisions and allow for personalized approaches for patients with TNBC.

Studies aimed at the identification of biomarkers and treatment targets of TNBC have mainly focused on mRNAs, miRNAs, and proteins (10). However, carbohydrates, such as glycosaminoglycans (GAGs), are relatively unexplored as biomarkers or treatment targets for TNBC. The non-template–driven nature of GAG biosynthesis involves the concerted action of many enzymes, which complicates the analysis of expression at the genomic/proteomic level and their role in clinical oncology. GAGs, such as heparan sulfate (HS) and chondroitin sulfate (CS) are linear polymers of repeated disaccharide units that are modified with sulfate groups at various positions. The resulting sulfated GAGs (sGAGs) are highly negatively charged and are known to mediate interactions of angiogenic factors, such as cytokines and growth factors with their respective cell surface receptors (11). Changes in chain length, sulfation, and the expression of proteoglycans (containing GAG attachment sites) at the cell surface and within the extracellular matrix are associated with the progression of TNBC and other cancers (11–13).

Recently, we discovered that the positively charged platinum agent, Triplatin (BBR3464), has high affinity for sGAGs, which mediates its cellular accumulation and cytotoxicity, whereas these properties are not shared by either cisplatin or carboplatin (14). Triplatin–sGAG interactions inhibit cancer cell growth and invasion by blocking growth factor signaling and heparanase activity (15). In this contribution, we explore sGAGs as a predictive marker for Triplatin efficacy against models of TNBC primary tumor growth and metastasis.

Reagents

Triplatin was synthesized by published procedures (16). Carboplatin and cisplatin were purchased from Sigma.

Cell lines and culture conditions

CHO-K1, CHO pgsE-606, and CHO pgsA-745 were purchased from the ATCC. CHO pgsF-17 was provided by Dr. Jeffrey Esko at the University of California (San Diego, CA). All CHO cell lines were cultured in DMEM-F12 media with 10% FBS and 1% penicillin/streptomycin. MDA-MB-231-BrM2-luc cells (provided by Dr. Joan Massauge at Sloane Kettering Cancer Center) and MDA-MB-231-luc cells (provided by Dr. Danny Welch at University of Kansas, Lawrence, KS) were cultured in DMEM media (Invitrogen) with 10% FBS, 1% l-glutamine, and 1% penicillin/streptomycin (Gibco). The MDA-MB-231-luc cells were generated by infection with lentiviruses containing the pFULT vector–expressing firefly luciferase. The lentiviruses were generated by DNA/RNA Delivery Core, Skin Disease Research Center, Department of Dermatology, Northwestern University Feinberg Medical School. 4T1-luc cells (provided by Dr. Sarah Spiegel at Virginia Commonwealth University) were cultured in RPMI media with 10% FBS and 1% penicillin/streptomycin. All cells were grown in a humidified atmosphere at 37°C with 5% CO2. Cells were routinely checked for Mycoplasma contamination using the Mycoplasma detection kit (Lorenzo).

Knockout of xylt1/2

sGAG-deficient MDA-MB-231 cell lines were generated by transfecting the parental/wild-type (WT) MDA-MB-231-luc cell line according to the protocol found in ref. (17). Cells received either Cas9 plasmid and empty DNA backbone (control cells) or Cas9, Xylt1, and Xylt2 gRNA plasmids [Knockout (KO) cells]. Following transfection, cells were single-cell sorted based upon GFP expression and allowed to grow in 96-well plates for 4 weeks, to allow for the clonal expansion of individual cell lines. WT and generated control, and KO cells were then screened via flow cytometry using sGAG-specific antibodies, to detect sGAG-deficient cell lines with XYLT1/2 KO and to confirm that control cell lines matched WT sGAG levels. Following screening, cells were used for in vitro and in vivo studies. Genomic alterations in KO cells were confirmed via cloning of Xylt1/2 PCR products and ligation into TOPO TA-sequencing plasmid. Sequencing was carried out using Euro Fins Sanger Sequencing, M13F primer. The following are PCR primers—Xylt1fwd (gtttctctcccctcttctccag), Xylt1rev (atgtcacacttagggggctg), Xylt2fwd (cccagtgatgtgtctgcatc), Xylt2rev (gtatctccgtgtggtgggaag). Each cell line was sequenced six times for each gene to ensure accuracy.

sGAG quantitation

Blyscan assay

The Blyscan assay was used per the manufacturer's instructions (Biocolor). Cells were seeded onto 100-mm dishes and grown to 90% confluence. Cells and cellular matrix were dissociated from dishes in 1 mL trypsin-EDTA and placed directly on ice. After the number of cells in each sample were determined, 100 μL of the suspension was added to 1 mL of 1,9-dimethylmethylene blue (DMMB) dye reagent and mixed via inversion every 5 minutes for 30 minutes. The samples were centrifuged at 12,000 rpm for 10 minutes. to pellet precipitated GAG-dye complexes. Complexes were then resolved in SDS-containing dissociation reagent and the absorbance of bound dye was measured at 656 nm. Values were compared with chondroitin-4-sulfate standards and normalized to cell number.

Flow cytometry

To study the relative amounts of cell surface sGAGs expressed, the binding of the 10E4 antibody (clone F58–10E4, Amsbio) to sulfated HS can be quantitated using flow cytometry. Briefly, cells are harvested using versene and washed twice in cold PBS. 1.5×106 cells are resuspended in PBS containing 0.5% BSA and kept on ice throughout the entire procedure. The cells are incubated with 1:100 10E4 primary antibody for 30 minutes. After washing twice with PBS, the cells are incubated with 1:400 goat anti-mouse IgM Alexa-647 conjugated secondary antibody (Life Technologies). Cells incubated with only secondary antibodies serve as negative controls. Fluorescence is measured using a Fortessa flow cytometer (BD Biosciences).

IHC

All tissues were fixed in 10% neutral-buffered formalin for at least 5 days before paraffin embedding and sectioning. A tissue microarray (TMA) was obtained from Champions Oncology containing breast cancer patient-derived xenograft (PDX) samples. IHC staining for human leukocyte antigen (HLA) primary antibody (Abcam, clone EMR8–5, 1:50 dilution) and N-sulfated HS (Amsbio, clone F58–10E4, 1:100 dilution) was performed in the VCU Cancer Mouse Models Core with the Leica Bond RX autostainer. All antibodies were incubated for 30 minutes at 25°C. For HLA staining heat-induced epitope retrieval buffer 2 (Leica, EDTA pH 8.0) was used. For negative control of N-sulfated HS staining (10E4), slides were incubated with 50 mU Heparinase III (Sigma-Aldrich) in Tris buffer (20 mmol/L Tris-HCL, pH 7.5 containing 0.1 mg/mL BSA and 4 mmol/L CaCl2) for 45 minutes at 37°C, three times, before primary antibody incubation. Digestion with heparinase III cleaves the HS chain and results in a loss of the 10E4 epitope. For the positive N-sulfated HS staining (10E4), slides were incubated with the Tris buffer only for 45 minutes at 37°C, three times, before primary antibody incubation. Bond polymer refine detection kit was used for DAB chromogen and hematoxylin counterstain (Leica). The tissue slides were visualized with a Vectra Polaris Automated Quantitative Pathology Imaging System (Akoya Biosciences) using whole-slide scanning at ×20. Phenochart Whole Slide Contextual Viewer software was used to visualize the scans and inFrom software (Akoya Biosciences) was used to quantify the staining (18–20). The CHOK1 and psA-745 were used to titrate the 10E4 antibody and validate staining. Liver metastases were quantified by metastases cell number identified by the HLA staining using Cell Counter on ImageJ and then normalized to total liver area measured in ImageJ. n = 5, each group.

Cell growth inhibition

One thousand cells/well were seeded in 96-well plates. After 24 hours, cells were treated with either carboplatin or Triplatin for 15 minutes in 200-μL media. Following treatment, drug media were removed, wells were washed with PBS, and replaced with drug-free media. Images and confluence data were acquired every 2 hours for approximately 200 hours using the IncuCyte Live Cell Analysis Imaging System (Sartorius). The AUC was calculated for each concentration using GraphPad Prism 8.0, normalized against the control, and used to calculate the IC50 values for Triplatin and carboplatin.

Cellular platinum uptake

Intracellular platinum content was determined as previously described (14). Cells were seeded onto 100-mm dishes, allowed to attach for 24 hours, and treated with 10 μmol/L carboplatin or Triplatin. Following 1- or 3-hour incubation, cells were harvested via trypsinization and counted. Cell pellets were washed repeatedly with PBS and digested in nitric acid for 72 hours. Samples were filtered and analyzed by inductively coupled plasma mass spectrometry (ICP-MS). Values were compared with platinum standards and normalized to cell number.

Tumor platinum uptake

CHO-K1 and pgsA-745 (3.0×106) cells were suspended 1:2 with 100 μL of PBS: 14 mg/mL of Cultrex BME Type 3 (Bio-Techne) and subcutaneously injected into the left and right flanks of male NSG mice, respectively, as described previously in ref. (21). MDA-MB-231-luc Xylt1/2 KO cells (1.0×104) were injected intraductally into female NSG mice. Once the tumors grew to a size of approximately 400 mm3, mice were given a single dose of Triplatin (0.3 mg/kg) or carboplatin (40 mg/kg) and euthanized 24 hours after administration. Tumors were harvested, weighed, dissolved in nitric acid for 72 hours, and filtered. Platinum content was determined using ICP-MS as previously described (14). Values were normalized to wet tumor weight.

Drug efficacy

In vivo mouse models

Virginia Commonwealth University Institutional Animal Care and Use Committee approval was obtained for all experiments. NOD.Cg.-PrkdcscidIL2rgtm1Wjl/SzJ (NSG) mice were purchased from the Cancer Mouse Models Core Laboratory at VCU Massey Cancer Center. Balb/c mice were purchased from Envigo.

Orthotopic primary tumor studies

MDA-MB-231-luc, MDA-MB-231-luc Xylt1/2 KO-F and KO-G (1×104) cells, WHIM2-luc, UCD52-luc, and WHIM30-luc (1×105) PDX cells were suspended in 100 μL of a 1:9 mix of PBS:Cultrex BME Type 3 (14 mg/mL, Bio-Techne) and injected into the fourth lactiferous duct of 6- to 8-week-old female NSG mice as described previously in ref. (22, 23). For 4T1 studies (1×104 cells suspended in 20-μL 1:9 PBS: Cultrex BME Type 3) were injected in the second lactiferous duct of 6- to 8-week-old Balb/c mice. Mice were weighed 2x weekly. Because the mice experienced weight loss during drug treatments, a supplementary diet (Diet Gel 76A; ClearH20) was given ad lib to the mice starting 1 day after injection of the cells until the end of the experiment. Drug treatments [vehicle (saline), cisplatin (3 mg/kg), carboplatin (40 mg/kg), or Triplatin (0.3 mg/kg)] were given interperitoneally on days 1, 5, and 9 (three times at intervals of 4 days) after randomization (day 1 was the day of randomization). Tumor growth was measured by using either digital caliper or bioluminescent imaging with the IVIS Spectrum (PerkinElmer). Using Studylog laboratory management software, the mice were randomized with the 2 factor (tumor volume and body weight) randomization Multi-Task program into groups once tumors reached an average of 150 mm3. At the endpoint, all mice were imaged live, euthanized, and tumors were harvested and weighed. The bioluminescence of harvested tumors and organs were quantified by ex vivo imaging and analysis. Each organ was immersed for 5 minutes in luciferin/PBS (0.3 mg/mL).

Experimental metastasis studies

MDA-MB-231-BrM2-luc cells (1×105 for endpoint study or 1×104 for survival study) or WHIM2-luc, UCD52-luc, and WHIM30-luc (5×105 for endpoint studies) PDX cells in 100-μL sterile PBS were injected into the left cardiac ventricle of female 5-week-old NSG mice as described previously in ref. (23). BALB/c female 6- to 8-week-old mice were injected into the left cardiac ventricle with 4 × 104 murine mammary cancer 4T1-luc cells (in 100-μL PBS) on day 0. The cells were mixed with luciferin (1.5 mg/mL) to permit immediate in vivo imaging (IVIS Spectrum, PerkinElmer) to verify widespread seeding of tumor cells. Mice were randomized using Studylog laboratory management software with the 2 factor (total flux values and body weight) randomization Multi-Task program on day 1 for 4T1 injected mice and day 10 for MDA-MB-231-BrM2-luc, WHIM2-luc, UCD52-luc, and WHIM30-luc injected mice. The groups were as follows: two groups (control and Triplatin) for MDA-MB-231-BrM2-luc cells endpoint studies or three groups (vehicle, carboplatin, Triplatin) for 4T1, WHIM2-luc, UCD52-luc, and WHIM30-luc endpoint studies, or three groups (vehicle, cisplatin, and Triplatin) for MDA-MB-231-BrM2-luc cells survival studies. After randomization, mice were treated interperitoneally with saline vehicle, cisplatin (3 mg/kg), carboplatin (40 mg/kg) or Triplatin (0.3 mg/kg) on days 1, 5, and 9 after randomization with day 1 being the day of randomization. Tumor burden was quantified by bioluminescence (radiance/sec) emitted from the cells after a subcutaneous injection of luciferin (150 mg/kg diluted in PBS) allowing for live in vivo imaging using the IVIS Spectrum and Living Image software (PerkinElmer). For the endpoint study, mice were euthanized, and organs (kidneys, heart, lung, ovaries, liver, brain, and skeleton) were harvested for ex vivo imaging as described above.

Statistical analysis

One-way or two-way ANOVA was performed to determine significant differences among the control and treatment groups. Tukey's multiple comparisons tests (for one-way ANOVA) or Dunnett's multiple comparison test (for two-way ANOVA) were performed to for comparison across groups. Unpaired t tests were performed to compare two groups. Kaplan–Meier survival plots were generated and changes in survival were analyzed by the log-rank test. All analyses were calculated in GraphPad Prism 8 software and P values <0.05 were considered significant.

sGAGs may be exploited for therapeutic benefit in TNBC

Dysregulation of sGAG expression affects all stages of tumorigenesis and correlates with poor prognosis in multiple cancers (12, 13). These findings have raised considerable interest in the development of sGAG-based qualitative and quantitative methods that include alcian blue, 1,9-dimethylene blue (Blyscan), capillary electrophoresis, high-performance liquid chromatography, mass spectrometry, and IHC. These methods have all been used to determine the relevance and significance that sGAGs play in biological processes, as identification of changes at the RNA/protein level are not sufficient to accurately predict changes in cell sGAG profile. Several reports observed increases in sGAG levels in TNBC and other BC cell lines as well as clinical samples compared with normal tissue, as shown in Table 1 (24–31).

Table 1.

Expression of sGAGs in breast cancer.

ObservationMethodSample typeRef
↑ In total sulfated GAGs Carbazole, PAGE Clinical (BC) 24 
↑ In total sulfated GAGs Alcian Blue (IHC), DMMB Clinical (BC), MDA-MB-231 25 
↑ N-sulfated HS Ab 10E4 (IHC) Clinical (BC, TNBC) 26–28 
↑ N-sulfated HS Ab 10E4 (flow cytometry) MDA-MB-231, HCC38 29 
↑ Sulfated CS Ab CS-56 (IHC, flow cytometry) Clinical (BC), MDA-MB-231 30 
↑ Sulfated CS VAR2CSA (IHC) Clinical (BC) 31 
ObservationMethodSample typeRef
↑ In total sulfated GAGs Carbazole, PAGE Clinical (BC) 24 
↑ In total sulfated GAGs Alcian Blue (IHC), DMMB Clinical (BC), MDA-MB-231 25 
↑ N-sulfated HS Ab 10E4 (IHC) Clinical (BC, TNBC) 26–28 
↑ N-sulfated HS Ab 10E4 (flow cytometry) MDA-MB-231, HCC38 29 
↑ Sulfated CS Ab CS-56 (IHC, flow cytometry) Clinical (BC), MDA-MB-231 30 
↑ Sulfated CS VAR2CSA (IHC) Clinical (BC) 31 

On the basis of these observations, we analyzed a TMA of TNBC PDX samples for sGAG levels using two GAG-specific antibodies; Ab 3G10, which reacts with an HS neoepitope after heparinase III digestion and Ab 10E4, which reacts with N-sulfated HS. The Ab 10E4 staining showed a wide range of expression, with two distinct populations staining high (40%) and low (60%; Fig. 1A and B), whereas Ab 3G10 staining was high overall and less dynamic (Supplementary Fig. S1). These results confirmed that sGAGs are upregulated in TNBC and thus, sGAG levels may be exploited for precision medicine with a compatible therapeutic.

Figure 1.

sGAGs 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 1.

sGAGs 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.

Close modal

Triplatin tumor uptake is sGAG dependent

The positively charged polynuclear platinum agent, Triplatin (Fig. 2A) interacts with negatively charged sGAGs with high affinity. The optimized structure for the interaction of [Pt(NH3)4]2+, the central platinum of Triplatin, with a sulfated monosaccharide, GlcNS(6S) is shown in Fig. 2B (32). WT CHO-K1 cells and GAG-deficient CHO mutant cell lines pgsF-17, pgsE-606, and pgsA-745 (Fig. 2C) were used to determine how sGAG levels affect platinum uptake. The pgsF-17 (F17) cell line, harboring a mutation in HS 2–0-sulfotransferase (HS2ST1), does not show significant changes in total sGAG levels as measured by DMMB-based Blyscan assay (Fig. 2D), and instead shows a slight compensatory increase in N-sulfated HS (Fig. 2E). The pgsE-606 (E606) cell line, with a mutation in HS N-sulfotransferase (NDST1), and the pgsA-745 (A745) cell line, with a mutation in one of the rate-limiting controllers of sGAG biosynthesis, xylosyltransferase-2 (XYLT2), show a significant reduction in both total sGAG and N-sulfated HS (Fig. 2C–E and I). To determine the effect of sGAG levels on platinum uptake, we compared the cellular accumulation of Triplatin and carboplatin in these cell lines (Fig. 2F). The overall cellular accumulation of Triplatin in the control cell line was significantly higher than that of carboplatin, as expected from previous results (14). The cellular accumulation (Fig. 2F) and cytotoxicity (Fig. 2G and H) of Triplatin increase with increasing sGAG levels (IC50 decreases from 120 ± 4.2 nmol/L to 69.5 ± 0.3 nmol/L). Surprisingly, the cellular accumulation (Fig. 2F) and cytotoxicity (Fig. 2G and H) of carboplatin negatively correlate with sGAG levels (IC50 increasing from 434 ± 71 nm to ≥ 1,000 nmol/L). Furthermore, significantly more Triplatin accumulated in vivo in the WT compared with the GAG-deficient A745 tumor, whereas carboplatin showed a slightly increased accumulation in the A745 tumor compared with the WT (Fig. 2I and K). These results raise the intriguing possibility that sGAGs not only mediate Triplatin uptake but also serve as a predictive biomarker for patient treatment.

Figure 2.

sGAG expression levels affect cellular accumulation, cytotoxicity, and tumor uptake of Triplatin. A, Structures of Triplatin and carboplatin. Pt, platinum. 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 three independent experiments. F, ICP-MS quantification of Triplatin and carboplatin cellular uptake. G–H, Analysis of cell proliferation after 15-minute treatments with the indicated concentrations of Triplatin or carboplatin. The percentage of confluence was quantified every 2 hours for 210 hours using live cell analysis (Incucyte). Each data point represents the mean ± SD (n = 2 biological replicates with four technical replicates each). IC50 calculations are shown in G. I, Right, Image of a mouse injected with WT CHO and A475. Middle, IHC image of N-sulfated HS (Ab 10E4) expression in CHO WT and 745 tumors and staining were quantified (left). Each bar represents the mean ± SD (n = 3). J and K, ICP-MS analysis of Pt tumor accumulation in CHO WT and A-745 tumors 24 hours 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 post-test; G, unpaired Student t test; J and K, Paired Student t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 2.

sGAG expression levels affect cellular accumulation, cytotoxicity, and tumor uptake of Triplatin. A, Structures of Triplatin and carboplatin. Pt, platinum. 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 three independent experiments. F, ICP-MS quantification of Triplatin and carboplatin cellular uptake. G–H, Analysis of cell proliferation after 15-minute treatments with the indicated concentrations of Triplatin or carboplatin. The percentage of confluence was quantified every 2 hours for 210 hours using live cell analysis (Incucyte). Each data point represents the mean ± SD (n = 2 biological replicates with four technical replicates each). IC50 calculations are shown in G. I, Right, Image of a mouse injected with WT CHO and A475. Middle, IHC image of N-sulfated HS (Ab 10E4) expression in CHO WT and 745 tumors and staining were quantified (left). Each bar represents the mean ± SD (n = 3). J and K, ICP-MS analysis of Pt tumor accumulation in CHO WT and A-745 tumors 24 hours 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 post-test; G, unpaired Student t test; J and K, Paired Student t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

Triplatin shows higher efficacy against sGAG-expressing tumors due to increased platinum accumulation

The effectiveness of carboplatin and Triplatin was compared against primary tumor growth using the parental MDA-MB-231-luc mouse TNBC model. MDA-MB-231 tumors show high reactivity with the 10E4 antibody (H-score >200), comparable with that of the WT CHO tumors (Fig. 3A and B). To confirm the specificity of Ab 10E4 staining, the MDA-MB-231-luc tumor samples were incubated with heparinase III, which resulted in cleavage of the HS polymer and loss of the epitope against N-sulfated HS (Fig. 3A). To validate our hypothesis that Triplatin will be efficacious in tumor models with high levels of sGAGs, MDA-MB-231-luc cells were injected into the mammary gland of female NSG mice. Once tumors reached approximately 150 mm3, mice were randomized and treated with either saline (vehicle), Triplatin or carboplatin on days 1, 5, and 9 with day 1 being the day of randomization. As predicted, Triplatin was more effective than carboplatin at reducing tumor volume, tumor weight, and lung metastases in this model (Fig. 3C and D; Supplementary Fig. S2A and S2B). The mice had weight loss during treatment but remained healthy overall and gained back weight after the end of treatment (Supplementary Fig. S2C).

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 (0.3 mg/kg), carboplatin (40 mg/kg), or vehicle (saline)] given interperitoneally 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 the left ventricle of female NSG mice (day 0). Animals were randomized on day 10. Drug treatment [Triplatin (0.3 mg/kg), cisplatin (3 mg/kg), or vehicle (saline)] given interperitoneally 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 t test; C, D, and E, One-way ANOVA; Tukey's post-test *, P < 0.05; **, P < 0.01; ****, 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 (0.3 mg/kg), carboplatin (40 mg/kg), or vehicle (saline)] given interperitoneally 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 the left ventricle of female NSG mice (day 0). Animals were randomized on day 10. Drug treatment [Triplatin (0.3 mg/kg), cisplatin (3 mg/kg), or vehicle (saline)] given interperitoneally 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 t test; C, D, and E, One-way ANOVA; Tukey's post-test *, P < 0.05; **, P < 0.01; ****, P < 0.0001.

Close modal

Because lower tumor burden can also result in decreased metastasis, we next determined the effect of Triplatin on metastatic growth using an intracardiac MDA-MB-231BR model of TNBC metastasis that allows analysis of Triplatin on inhibition of growth at metastatic sites seen in patients with TNBC, for example, the brain, bone, lung, and liver. Triplatin-treated mice gave marked increases in survival compared with control and cisplatin-treated mice (Fig. 3E). In an endpoint study, when all mice were sacrificed 21 days after cancer cell injection, total metastases and, especially, bone and liver metastases were dramatically reduced in mice treated with Triplatin compared with control mice (Fig. 3F–H; Supplementary Fig. S3). Furthermore, the ability of Triplatin to directly inhibit lung metastases was evaluated using the syngeneic murine 4T1 mastectomy model (33). The 4T1-luc mouse–derived mammary cancer cell line injected into immune-intact mice mimics human cancer progression and metastasizes more efficiently than conventional human xenografts. The IHC analysis of 4T1 tumors grown orthotopically showed high Ab 10E4 staining (Supplementary Fig. S4A), suggesting that this model will be sensitive to Triplatin. We observed that Triplatin inhibited lung metastases that occurred after mastectomy (Supplementary Fig. S4B and S4C) and significantly increased survival (Supplementary Fig. S4D). Furthermore, 4T1 cells were intracardially injected into mice to mimic widespread metastasis. Mice treated with Triplatin had reduced metastases in the lungs and bone compared with carboplatin-treated mice and saline control (Supplementary Fig. S5A and S5B). Together, these results show that Triplatin inhibits primary tumor growth and metastasis in carboplatin resistant TNBC models expressing high levels of sGAGs.

To confirm that sGAG levels play an important role in the efficacy of Triplatin in TNBC tumors, we generated XYLT1/2 knockout MDA-MB-231 cell lines using well-characterized CRISPR plasmids as described previously in ref. (16). Created CRISPR clones KO-F and KO-G both possess indels in exon 3 of XYLT1 and exon 2 of XYLT2, as determined by sanger sequencing (Supplementary Fig. S6A and S6B). These generated cell lines, MDA-MB-231 XYLT1/2 KO-F and KO-G, had >90% reduction in sGAG epitopes in comparison with both parental and control cells (Fig. 4A). Consistent with our model that high sGAG levels facilitate Triplatin accumulation, the MDA-MB-231 XYLT1/2 KO-F and KO-G cell lines had reduced uptake of Triplatin and showed higher uptake of carboplatin in vitro (Fig. 4B) compared with WT MDA-MB-231. Differences in Triplatin accumulation were also seen in vivo (Fig. 4C), similar to Triplatin accumulation in WT CHO K1 compared with CHO-pgsA-745 cells. These differences in accumulation affected the efficacy of Triplatin. WT MDA-MB-231 tumors had a higher percentage of tumor growth inhibition (%TGI) when treated with Triplatin compared with vehicle than did the sGAG-deficient MDA-MB-231 XYLT1/2 KO-F and KO-G tumors (Fig. 4D). This difference translated to an increased survival benefit and effect size for the mice with WT MDA-MB-231 tumors (survival benefit 16.7 days, effect size 4.17) compared with those with the sGAG-deficient MDA-MB-231 XYLT1/2 KO-F and KO-G tumors (survival benefit 11 and 7 days and effect size 2.36 and 0.99, respectively; Fig. 4E).

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 with 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 hours after treatment with Triplatin (0.3 mg/kg). D and E, Comparison of the percentage 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 interperitoneally on days 1, 5, and 9 after randomization. Statistical analysis; A and B, One-way ANOVA, Tukey's post-test. C, Paired Student 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 Δ = mean difference between Triplatin and vehicle divided by the standard of deviation of the vehicle group.

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 with 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 hours after treatment with Triplatin (0.3 mg/kg). D and E, Comparison of the percentage 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 interperitoneally on days 1, 5, and 9 after randomization. Statistical analysis; A and B, One-way ANOVA, Tukey's post-test. C, Paired Student 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 Δ = mean difference between Triplatin and vehicle divided by the standard of deviation of the vehicle group.

Close modal

The efficacy of Triplatin against TNBC PDX models correlates with sGAG levels

Use of PDX models in preclinical studies correlates well with clinical response (34–36). Therefore, the efficacy of Triplatin was compared in TNBC PDX with high and low levels of sGAGs to test the hypothesis that PDX models with high levels of sGAG staining will be most responsive to Triplatin. We examined three PDX TNBC models, previously investigated for sensitivity to carboplatin—UCD52 (carboplatin-sensitive), WHIM30 (carboplatin-sensitive), and WHIM2 (carboplatin-resistant; refs. 37–40). WHIM30 and WHIM2 had significantly higher sGAG levels compared with UCD52 (Fig. 5A–B). WHIM30 primary tumor growth was significantly inhibited by Triplatin, and carboplatin compared with the control (Fig. 5C). Triplatin was more effective than carboplatin at inhibiting WHIM2 tumor growth (Fig. 5D), whereas carboplatin was more effective than Triplatin at inhibiting UCD52 primary tumor growth (Fig. 5E). These results indicate that the activity of Triplatin against primary tumor growth is correlated with sGAG levels in TNBC PDX tumors.

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, 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 interperitoneally three times at intervals of 4 days with Triplatin (0.3 mg/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 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, 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 interperitoneally three times at intervals of 4 days with Triplatin (0.3 mg/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.

Close modal

The effects of Triplatin and carboplatin on metastatic growth of WHIM30, WHIM2, and UCD52 PDXs were further evaluated using an intracardiac injection model. As previously described, WHIM30 cells formed metastases in the lung and liver. Similar to the primary tumor study, Triplatin and carboplatin decreased metastatic burden overall (Fig. 6A and B) and specifically, in the liver (Fig. 6C and E) and the lung (Fig. 6D). WHIM2 cells formed metastases in the brain, lung, liver, and ovaries with the highest burden in the lung and ovaries. Triplatin, but not carboplatin, decreased metastatic burden overall (Fig. 6A and B) specifically in the ovaries (Figs. 6C and 5F) and the lungs (Fig. 6D). UCD52 formed metastases in the lung, liver, and brain with the highest burden in the liver and brain, which were overall more sensitive to carboplatin than Triplatin (Fig. 6A and B). In specific organs, carboplatin inhibited metastases to the liver more efficiently than Triplatin (Fig. 6C and D), and only carboplatin inhibited brain metastatic growth (Fig. 6D). Taken together, these results demonstrate that Triplatin may be more effective than carboplatin in treatment of TNBC with high levels of sGAGs.

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 interperitoneally three times at intervals of 4 days with Triplatin (0.3 mg/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 post-test. *, 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 interperitoneally three times at intervals of 4 days with Triplatin (0.3 mg/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 post-test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

Although numerous reports describe molecular characterizations of TNBC (including genetic mutations, copy-number alterations, and protein expression), quantitation of carbohydrate levels is generally omitted. However, a few pivotal studies have identified upregulation of HS and CS in BC compared with normal tissue (Table 1). Gill and colleagues (26) showed high expression of the HS 10E4 antibody epitope in clinical samples of BC and, in particular, TNBC. We applied this antibody to a TMA of TNBC PDX tumors and found distinct groups of low- and high-staining tumors. Tumors with H-scores >100 were predictive of Triplatin efficacy, even in models that were carboplatin-resistant.

The desirability of incorporating a platinum into treatment regimens stems from data suggesting that a high frequency of DNA-damage repair defects renders TNBC particularly susceptible to DNA-modifying agents (9, 41). Platinums exert their effect by creating DNA–drug lesions that inhibit DNA replication and induce cell death. The synergy of two different mechanisms that potentially induce DNA damage (platinums causing DNA damage and “BRCAness” with ineffective DNA damage repair), known as synthetic lethality, is the rationale for the potential benefit of platinums in this setting. Triplatin was originally developed to generate DNA lesions that are distinct from other platinums and less recognizable by DNA repair proteins. The 1,4 DNA interstrand crosslinks produced by Triplatin, unlike the 1,2 intrastrand crosslinks produced by cisplatin/carboplatin, evade binding of high-mobility group proteins and repair by the nucleotide excision repair machinery (42–44). The results shown here are consistent, in that Triplatin showed similar activity in high sGAG-expressing PDX models, WHIM30 (mtBRCA), and WHIM2 (wtBRCA), regardless of BRCA status, whereas carboplatin was markedly less effective in the WHIM2 model with wtBRCA (37, 38). Together, these results suggest that the efficacy of Triplatin may be independent of BRCA status, although future studies with more BRCA mutational models will be needed.

The majority of patients with TNBC in the United States present with local or locally advanced (stage I–III) disease at the time of diagnosis. Patients who achieve pCR, defined as elimination of invasive cancer from the breast and lymph nodes, have excellent chances of recurrence-free survival. However, those with residual disease (RD) have a higher risk of relapse and poorer prognosis than other BC subtypes. Paradoxically, a subset of TNBC appears to be particularly chemosensitive, with 30% to 40% achieving pCR after NACT using standard anthracycline- and taxane-based combinations. Four recent randomized trials, GeparSixto, CALGB 40603, Brightness and I-SPY 2, have investigated the addition of carboplatin to standard NACT regimens. On average, the pCR rate increased to 50% to 55% when carboplatin was included (7, 31, 45–47). These results were confirmed in the Keynote-522 trial and with the addition of pembrolizumab, the number of patients achieving pCR further increased to approximately 65% (7). Together, the results support adding platinum to NACT regimens in treatment of locally advanced TNBC, which may again be optimized in the context of BRCA status and/or homologous repair deficiencies.

The major advantage of NACT is the ability to immediately identify patients with chemotherapy-resistant disease who may benefit from second adjuvant therapy following surgery (48). Recent approval of capecitabine, an oral antimetabolite of 5-flurouracil, resulted from the CREATE-X trial showing a significant improvement in the rate of recurrence-free survival at 5 years (69.8% in the capecitabine group vs. 56.1% in the control group; ref. 49). The ongoing EA1131 trial will challenge this success by comparing capecitabine directly with six cycles of carboplatin (7). Future studies profiling RD for markers of selective targeting agents, such as Triplatin, may further benefit patients with chemotherapy-resistant disease.

Patients who developed distant metastases during the course of their disease survive less than a year, underscoring the importance of developing more effective treatment of patients living with metastatic TNBC. Even with the addition of recent treatment options for metastatic TNBC, such as immunotherapy (nab-paclitaxel and atezolizumab) for PD-L1–positive tumors and carboplatin/PARP-inhibitors for BRCAness-positive tumors, the survival benefit is measured in months (7, 9, 50). For this reason, the inclusion of patients in ongoing trials evaluating new targeted agents and predictive biomarkers is typically encouraged. The results presented here find that Triplatin is effective at prolonging survival of mice with metastatic TNBC and inhibiting growth of TNBC cells at metastatic sites in TNBC with high levels of sGAGs in multiple models.

In light of our discovery that Triplatin cellular and tumor uptake is sGAG dependent and increase efficacy is seen in tumors with higher levels of sGAGs, a re-evaluation of the clinical potential of Triplatin in this context is warranted. The primary dose-limiting toxicities of Triplatin were neutropenia and diarrhea. Using the recommended phase II dosing schedule, 1.1 mg/m2 on day 1 of a 28-day cycle, Triplatin had a manageable toxicity profile. Phase I trials showed objective responses in advanced metastatic BC; in combination with 5-FU (one partial response, three stable disease/14; ref. 51). Phase II clinical trials in cisplatin-relapsed ovarian cancer gave partial responses (5/28 patients; ref. 44), and in several cases durable responses >3 years (personal communication). In summary, these results show activity of Triplatin in TNBC tumors with high sGAG levels, regardless of BRCA status and carboplatin resistance. Although the added toxicities of incorporating platinum into any regimens should be considered, the benefits may be substantial, especially in early-stage cases showing resistance, RD after neoadjuvant therapy, and in the metastatic setting. Elevation of pCR rate is expected to significantly improve patient survival.

J.B. Turner reports grants from Veterans Affairs during the conduct of the study. N.P. Farrell reports grants from Commonwealth Research Commercialization Fund and Commonwealth Heath Research Board and Massey Cancer Center during the conduct of the study; and started a small start-up company—Bioplatinum Technologies LLC—to facilitate technology transfer of polynuclear platinum complexes. This company is nascent; has no revenue or employees but nevertheless added here for completeness. J.E. Koblinski reports grants from Virginia Commonwealth University/VCUHealth, Massey Cancer Center, Center for Innovative Technology, and Commonwealth Health Research Board during the conduct of the study. No disclosures were reported by the other authors.

J.D. Hampton: Conceptualization, formal analysis, investigation, visualization, methodology, writing–original draft, writing–review and editing. E.J. Peterson: Conceptualization, formal analysis, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. S.J. Katner: Conceptualization, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. T.H. Turner: Resources, investigation, writing–review and editing. M.A. Alzubi: Resources, investigation, writing–review and editing. J.C. Harrell: Resources, investigation, writing–review and editing. M.G. Dozmorov: Formal analysis, writing–review and editing. J.B.M. Turner: Investigation, writing–review and editing. P.J. Gigliotti: Investigation, writing–review and editing. V. Kraskauskiene: Investigation, writing–review and editing. M. Shende: Investigation, writing–review and editing. M.O. Idowu: Validation, investigation, writing–review and editing. M. Puchallapalli: Validation, investigation, methodology, writing–review and editing. B. Hu: Investigation, writing–review and editing. L. Litovchick: Conceptualization, writing–review and editing. E. Katsuta: Conceptualization, resources, validation, investigation, visualization, writing–review and editing. K. Takabe: Conceptualization, resources, investigation, writing–review and editing. N.P. Farrell: Conceptualization, resources, supervision, funding acquisition, writing–original draft, project administration, writing–review and editing. J.E. Koblinski: Conceptualization, resources, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing.

We thank Dr. J. Esko for the CHO-F17 mutant cell line and Dr. H. Bear for the use of 4T1 tumor samples for IHC. The lentivirus stocks have been generated by Gene Editing, Transduction and Nanotechnology Core, Skin Biology and Disease Resource-Based Center, Northwestern University, supported in part by the NIH-NCI Cancer Center support grant (P30 CA060553). Services in support of the research project were provided by the Virginia Commonwealth University Massey Cancer Center Flow Cytometry Core, Cancer Mouse Models Core Laboratory, and Microscopy Facility supported, in part, with funding to the Massey Cancer Center from NIH-NCI Cancer Center Support Grant P30 CA016059. This work was supported by the Massey Cancer Center Molecules to Medicine Initiative, Commonwealth Health Research Board (CHRB 236–05–17), Commonwealth Research Commercialization Fund (MF17–028-LS) and Value and Efficiency Teaching and Research Initiative (VETAR, School of Medicine, VCU). National Institutes of Health grant R01CA160688 (to K. Takabe). National Cancer Institute (NCI) grants P30CA016056 and U24CA232979, involving the use of Roswell Park Comprehensive Cancer Center's Shared Resources were awarded to K. Takabe and E. Katsuta.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Cardoso
F
,
Senkus
E
,
Costa
A
,
Papadopoulos
E
,
Aapro
M
,
André
F
, et al
4th ESO–ESMO international consensus guidelines for advanced breast cancer (ABC 4) These guidelines were developed by the european school of oncology (ESO) and the european society for medical oncology (ESMO)
.
Ann Oncol
2018
;
29
:
1634
57
.
2.
Bevers
TB
,
Anderson
BO
,
Bonaccio
E
,
Buys
S
,
Daly
MB
,
Dempsey
PJ
, et al
Breast cancer screening and diagnosis
.
J Natl Compr Canc Ne
2009
;
7
:
1060
96
.
3.
Waks
AG
,
Winer
EP
. 
Breast cancer treatment: a review
.
JAMA
2019
;
321
:
288
300
.
4.
Schmid
P
,
Cortes
J
,
Pusztai
L
,
McArthur
H
,
Kümmel
S
,
Bergh
J
, et al
Pembrolizumab for early triple-negative breast cancer
.
N Engl J Med
2020
;
382
:
810
21
.
5.
Bardia
A
,
Hurvitz
SA
,
Tolaney
SM
,
Loirat
D
,
Punie
K
,
Oliveira
M
, et al
Sacituzumab govitecan in metastatic triple-negative breast cancer
.
N Engl J Med
2021
;
384
:
1529
41
.
6.
Schmid
P
,
Adams
S
,
Rugo
HS
,
Schneeweiss
A
,
Barrios
CH
,
Iwata
H
, et al
Atezolizumab and nab-paclitaxel in advanced triple-negative breast cancer
.
N Engl J Med
2018
;
379
:
2108
21
.
7.
Garufi
G
,
Palazzo
A
,
Paris
I
,
Orlandi
A
,
Cassano
A
,
Tortora
G
, et al
Neoadjuvant therapy for triple-negative breast cancer: potential predictive biomarkers of activity and efficacy of platinum chemotherapy, PARP- and immune-checkpoint-inhibitors
.
Expert Opin Pharmaco
2020
;
21
:
687
99
.
8.
Tokumaru
Y
,
Joyce
D
,
Takabe
K
. 
Current status and limitations of immunotherapy for breast cancer
.
J Surg
2020
;
167
:
628
30
.
9.
Tutt
A
,
Tovey
H
,
Cheang
MCU
,
Kernaghan
S
,
Kilburn
L
,
Gazinska
P
, et al
Carboplatin in BRCA1/2-mutated and triple-negative breast cancer BRCAness subgroups: the TNT trial
.
Nat Med
2018
;
24
:
628
.
10.
Tokumaru
Y
,
Oshi
M
,
Katsuta
E
,
Yan
L
,
Satyananda
V
,
Matsuhasi
N
, et al
KRAS signaling enriched triple negative breast cancer is associated with favorable tumor immune microenvironment and better survival
.
Am J Cancer Res
2020
;
10
:
897
907
.
11.
Zhang
F
,
Zhang
Z
,
Linhardt
RJ
. 
Chapter 3—glycosaminoglycans
.
In
:
Cummings
RD
,
Pierce
JM
,
editors
.
Handbook of Glycomics
.
San Diego
:
Academic Press
; 
2010
; p.
59
80
.
12.
Fuster
MM
,
Esko
JD
. 
The sweet and sour of cancer: glycans as novel therapeutic targets
.
Nat Rev Cancer
2005
;
5
:
526
42
.
13.
Afratis
N
,
Gialeli
C
,
Nikitovic
D
,
Tsegenidis
T
,
Karousou
E
,
Theocharis
AD
, et al
Glycosaminoglycans: key players in cancer cell biology and treatment
.
FEBS J
2012
;
279
:
1177
97
.
14.
Silva
H
,
Frézard
F
,
Peterson
EJ
,
Kabolizadeh
P
,
Ryan
JJ
,
Farrell
NP
. 
Heparan sulfate proteoglycan-mediated entry pathway for charged tri-platinum compounds: differential cellular accumulation mechanisms for platinum
.
Mol Pharm
2012
;
9
:
1795
802
.
15.
Peterson
EJ
,
Daniel
AG
,
Katner
SJ
,
Bohlmann
L
,
Chang
C-W
,
Bezos
A
, et al
Antiangiogenic platinum through glycan targeting
.
Chem Sci
2017
;
8
:
241
52
.
16.
Qu
Y
,
Kipping
RG
,
Farrell
NP
. 
Solution studies on DNA interactions of substitution-inert platinum complexes mediated via the phosphate clamp
.
J Chem Soc Dalton Trans
2015
;
44
:
3563
72
.
17.
Narimatsu
Y
,
Joshi
HJ
,
Yang
Z
,
Gomes
C
,
Chen
Y-H
,
Lorenzetti
FC
, et al
A validated gRNA library for CRISPR/Cas9 targeting of the human glycosyltransferase genome
.
Glycobiology
2018
;
28
:
295
305
.
18.
Sunshine
JC
,
Nguyen
PL
,
Kaunitz
GJ
,
Cottrell
TR
,
Berry
S
,
Esandrio
J
, et al
PD-L1 expression in melanoma: a quantitative immunohistochemical antibody comparison
.
Clin Cancer Res
2017
;
23
:
4938
44
.
19.
Rehman
JA
,
Han
G
,
Carvajal-Hausdorf
DE
,
Wasserman
BE
,
Pelekanou
V
,
Mani
NL
, et al
Quantitative and pathologist-read comparison of the heterogeneity of programmed death-ligand 1 (PD-L1) expression in non–small cell lung cancer
.
Mod Pathol
2017
;
30
:
340
9
.
20.
Parra
ER
,
Villalobos
P
,
Zhang
J
,
Behrens
C
,
Mino
B
,
Swisher
S
, et al
Immunohistochemical and image analysis-based study shows that several immune checkpoints are co-expressed in non–small cell lung carcinoma tumors
.
J Thorac Oncol
2018
;
13
:
779
91
.
21.
Nakase
I
,
Konishi
Y
,
Ueda
M
,
Saji
H
,
Futaki
S
. 
Accumulation of arginine-rich cell-penetrating peptides in tumors and the potential for anticancer drug delivery in vivo
.
J Control Release
2012
;
159
:
181
8
.
22.
Harrell
JC
,
Dye
WW
,
Allred
DC
,
Jedlicka
P
,
Spoelstra
NS
,
Sartorius
CA
, et al
Estrogen receptor positive breast cancer metastasis: altered hormonal sensitivity and tumor aggressiveness in lymphatic vessels and lymph nodes
.
Cancer Res
2006
;
66
:
9308
15
.
23.
Puchalapalli
M
,
Zeng
X
,
Mu
L
,
Anderson
A
,
Hix Glickman
L
,
Zhang
M
, et al
NSG mice provide a better spontaneous model of breast cancer metastasis than athymic (Nude) mice
.
PLoS ONE
2016
;
11
:
e0163521
.
24.
Weyers
A
,
Yang
B
,
Yoon
DS
,
Park
JH
,
Zhang
F
,
Lee
KB
, et al
A structural analysis of glycosaminoglycans from lethal and nonlethal breast cancer tissues: toward a novel class of theragnostics for personalized medicine in oncology?
OMICS
2012
;
16
:
79
89
.
25.
Singh
V
,
Jha
KK
,
JK
M
,
Kumar
RV
,
Raghunathan
V
,
Bhat
R
. 
Iduronate-2-sulfatase-regulated dermatan sulfate levels potentiate the invasion of breast cancer epithelia through collagen matrix
.
J Clin Med
2019
;
8
:
1562
.
26.
Gill
RM
,
Mehra
V
,
Milford
E
,
Dhoot
GK
. 
Short SULF1/SULF2 splice variants predominate in mammary tumours with a potential to facilitate receptor tyrosine kinase-mediated cell signalling
.
J Histochem Cell Biol
2016
;
146
:
431
44
.
27.
Guo
C-H
,
Koo
C-Y
,
Bay
B-H
,
Tan
P-H
,
Yip
GW
. 
Comparison of the effects of differentially sulphated bovine kidney- and porcine intestine-derived heparan sulphate on breast carcinoma cellular behaviour
.
Int J Oncol
2007
;
31
:
1415
23
.
28.
Koo
C-Y
,
Bay
B-H
,
Lui
PC-W
,
Tse
GM-K
,
Tan
P-H
,
Yip
GW-C
. 
Immunohistochemical expression of heparan sulfate correlates with stromal cell proliferation in breast phyllodes tumors
.
Mod Pathol
2006
;
19
:
1344
50
.
29.
Sembajwe
LF
,
Katta
K
,
Grønning
M
,
Kusche-Gullberg
M
. 
The exostosin family of glycosyltransferases: mRNA expression profiles and heparan sulphate structure in human breast carcinoma cell lines
.
Biosci Rep
2018
;
38
:
BSR20180770
.
30.
Svensson
KJ
,
Christianson
HC
,
Kucharzewska
P
,
Fagerström
V
,
Lundstedt
L
,
Borgquist
S
, et al
Chondroitin sulfate expression predicts poor outcome in breast cancer
.
Int JOncol
2011
;
39
:
1421
8
.
31.
Salanti
A
,
Clausen
TM
,
Agerbæk
,
Al Nakouzi
N
,
Dahlbäck
M
,
Oo
HZ
, et al
Targeting human cancer by a glycosaminoglycan binding malaria protein
.
Cancer Cell
2015
;
28
:
500
14
.
32.
Gorle
AK
,
Rajaratnam
P
,
Chang
C-W
,
von Itzstein
M
,
Berners-Price
SJ
,
Farrell
NP
. 
Glycans as ligands in bioinorganic chemistry. probing the interaction of a trinuclear platinum anticancer complex with defined monosaccharide fragments of heparan sulfate
.
Inorg Chem
2019
;
58
:
7146
55
.
33.
Katsuta
E
,
Rashid
OM
,
Takabe
K
. 
Murine breast cancer mastectomy model that predicts patient outcomes for drug development
.
J Surg Res
2017
;
219
:
310
8
.
34.
Zhang
X
,
Lewis
MT
. 
Establishment of patient-derived xenograft (PDX) models of human breast cancer
.
Curr Protoc Mouse Biol
2013
;
3
:
21
9
.
35.
Fiebig
HH
,
Neumann
HA
,
Henss
H
,
Koch
H
,
Kaiser
D
,
Arnold
H
. 
Development of three human small-cell lung cancer models in nude mice
.
Recent Results Cancer Res
1985
;
97
:
77
86
.
36.
Gao
H
,
Korn
JM
,
Ferretti
S
,
Monahan
JE
,
Wang
Y
,
Singh
M
, et al
High-throughput screening using patient-derived tumor xenografts to predict clinical trial drug response
.
Nat Med
2015
;
21
:
1318
25
.
37.
Turner
TH
,
Alzubi
MA
,
Sohal
SS
,
Olex
AL
,
Dozmorov
MG
,
Harrell
JC
. 
Characterizing the efficacy of cancer therapeutics in patient-derived xenograft models of metastatic breast cancer
.
Breast cancer Res and TR
2018
;
170
:
221
34
.
38.
Turner
TH
,
Alzubi
MA
,
Harrell
JC
. 
Identification of synergistic drug combinations using breast cancer patient-derived xenografts
.
Sci Rep
2020
;
10
:
1493
.
39.
Alzubi
MA
,
Sohal
SS
,
Sriram
M
,
Turner
TH
,
Zot
P
,
Idowu
M
, et al
Quantitative assessment of breast cancer liver metastasis expansion with patient-derived xenografts
.
Clin Exp Metastas
2019
;
36
:
257
69
.
40.
Alzubi
MA
,
Turner
TH
,
Olex
AL
,
Sohal
SS
,
Tobin
NP
,
Recio
SG
, et al
Separation of breast cancer and organ microenvironment transcriptomes in metastases
.
Breast Cancer Res
2019
;
21
:
36
.
41.
Loibl
S
,
Weber
KE
,
Timms
KM
,
Elkin
EP
,
Hahnen
E
,
Fasching
PA
, et al
Survival analysis of carboplatin added to an anthracycline/taxane-based neoadjuvant chemotherapy and HRD score as predictor of response—final results from GeparSixto
.
Ann Oncol
2018
;
29
:
2341
7
.
42.
Kasparkova
J
,
Zehnulova
J
,
Farrell
N
,
Brabec
V
. 
DNA interstrand cross-links of the novel antitumor trinuclear platinum complex BBR3464: conformation, recognition by high mobility group domain proteins, and nucleotide excision repair
.
J Biol Chem
2002
;
277
:
48076
86
.
43.
Zehnulova
J
,
Kasparkova
J
,
Farrell
N
,
Brabec
V
. 
Conformation, recognition by high mobility group domain proteins, and nucleotide excision repair of DNA intrastrand cross-links of novel antitumor trinuclear platinum complex BBR3464
.
J Biol Chem
2001
;
276
:
22191
9
.
44.
Farrell
NP
. 
Multi-platinum anti-cancer agents. Substitution-inert compounds for tumor selectivity and new targets
.
Chem Soc Rev
2015
;
44
:
8773
85
.
45.
Pandy
JGP
,
Balolong-Garcia
JC
,
Cruz-Ordinario
MVB
,
Que
FVF
. 
Triple negative breast cancer and platinum-based systemic treatment: a meta-analysis and systematic review
.
BMC Cancer
2019
;
19
:
1065
.
46.
Park
JH
,
Ahn
J-H
,
Kim
S-B
. 
How shall we treat early triple-negative breast cancer (TNBC): from the current standard to upcoming immuno-molecular strategies
.
ESMO Open
2018
;
3
:
e000357
.
47.
Sikov
WM
,
Berry
DA
,
Perou
CM
,
Singh
B
,
Cirrincione
CT
,
Tolaney
SM
, et al
Impact of the addition of carboplatin and/or bevacizumab to neoadjuvant once-per-week paclitaxel followed by dose-dense doxorubicin and cyclophosphamide on pathologic complete response rates in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance)
.
J Clin Oncol
2015
;
33
:
13
21
.
48.
Kennedy
WR
,
Tricarico
C
,
Gabani
P
,
Weiner
AA
,
Altman
MB
,
Ochoa
LL
, et al
Predictors of distant metastases in triple-negative breast cancer without pathologic complete response after neoadjuvant chemotherapy
.
J Natl Compr Canc NE
2020
;
18
:
288
96
.
49.
Masuda
N
,
Lee
S-J
,
Ohtani
S
,
Im
Y-H
,
Lee
E-S
,
Yokota
I
, et al
Adjuvant capecitabine for breast cancer after preoperative chemotherapy
.
N Eng J Med
2017
;
376
:
2147
59
.
50.
Caparica
R
,
Lambertini
M
,
de
AE
. 
How I treat metastatic triple-negative breast cancer
.
ESMO Open
2019
;
4
:
e000504
.
51.
Gourley
C
,
Cassidy
J
,
Edwards
C
,
Samuel
L
,
Bisset
D
,
Camboni
G
, et al
A phase I study of the trinuclear platinum compound, BBR 3464, in combination with protracted venous infusional 5-fluorouracil in patients with advanced cancer
.
Cancer Chemother Pharmacol
2004
;
53
:
95
101
.

Supplementary data