TH1902, a new docetaxel-peptide conjugate for the treatment of sortilin-positive triple-negative breast cancer

Abstract

Triple-negative breast cancer (TNBC) is a heterogeneous subgroup of cancers which lacks the expression and/or amplification of targetable biomarkers (ie, estrogen receptor, progestrogen receptor, and human epidermal growth factor receptor 2), and is often associated with the worse disease-specific outcomes than other breast cancer subtypes. Here, we report that high expression of the sortilin (SORT1) receptor correlates with the decreased survival in TNBC patients, and more importantly in those bearing lymph node metastases. By exploiting SORT1 function in ligand internalization, a new anticancer treatment strategy was designed to target SORT1-positive TNBC-derived cells both in vitro and in two in vivo tumor xenografts models. A peptide (TH19P01), which requires SORT1 for internalization and to which many anticancer drugs could be conjugated, was developed. In vitro, while the TH19P01 peptide itself did not exert any antiproliferative or apoptotic effects, the docetaxel-TH19P01 conjugate (TH1902) exerted potent antiproliferative and antimigratory activities when tested on TNBC-derived MDA-MB-231 cells. TH1902 triggered faster and more potent apoptotic cell death than did unconjugated docetaxel. The apoptotic and antimigratory effects of TH1902 were both reversed by two SORT1 ligands, neurotensin and progranulin, and on siRNA-mediated silencing of SORT1. TH1902 also altered microtubule polymerization and triggered the downregulation of the anti-apoptotic Bcl-xL biomarker. In vivo, both i.p. and i.v. administrations of TH1902 led to greater tumor regression in two MDA-MB-231 and HCC-70 murine xenograft models than did docetaxel, without inducing neutropenia. Altogether, the data demonstrates the high in vivo efficacy and safety of TH1902 against TNBC through a SORT1 receptor-mediated mechanism. This property allows for selective treatment of SORT1-positive TNBC and makes TH1902 a promising avenue for personalized therapy with the potential of improving the therapeutic window of cytotoxic anticancer drugs such as docetaxel.

Keywords: docetaxel; peptide-drug conjugate; receptor-mediated chemotherapy; sortilin; triple-negative breast cancer.

Conflict of interest statement

MD, AL, RB and BA were scientific founders of Katana Biopharma. CM is senior vice president and chief medical officer at Theratechnologies.

© 2021 The Authors. Cancer Science published by John Wiley & Sons Australia, Ltd on behalf of Japanese Cancer Association.

Figures

FIGURE 1

SORT1 expression in breast cancers. A, Immunohistochemistry stainings were performed to assess SORT1 expression in normal adjacent tissue and in stage IIIC infiltrating duct carcinoma (IDC) and involved lymph node carcinoma (ILNC). B, SORT1 staining intensity was evaluated using the immunohistostaining (IHS) method on nine normal adjacent tissues, 35 IDC, 10 ILNC, and 19 TNBC tumors. Data are represented as mean ± SEM

FIGURE 2

In silico analysis of SORT1 gene expression on the survival of triple‐negative breast cancer (TNBC) patients. (A) TNBC patient survival curves were obtained using the Kaplan‐Meier plotter software (http://kmplot.com/analysis/) for patients expressing low (black curve) or high (red curve) levels of SORT1 (n = 161 cases) or (B) in TNBC patients with lymph node metastases (n = 72 cases)

FIGURE 3

SORT1‐mediated TH19P01 peptide uptake in MDA‐MB‐231 cells. Transient SORT1 gene silencing was performed as described in the Methods section and confirmed by (A) Western blotting and (B) densitometric analysis of SORT1 protein expression. Experiment was performed in triplicate and the densitometric data analyzed using a t‐test. (C) The binding (left panel) or uptake (right panel) of 200 nM Alexa488‐labeled TH19P01 was performed in control (siScrambled) or SORT1‐deficient (siSORT1) MDA‐MB‐231 cancer cells. The levels of fluorescence in the two categories of cells were measured in triplicate (binding and uptake n = 3). Affinity constants were extrapolated for (D) Alexa488‐labeled TH19P01 cell surface binding (KD), and inhibitory constants (Ki) for (E) uptake of 200 nM Alexa488‐labeled TH19P01 performed in the presence of increasing TH1902, TH19P01, or Neurotensin as indicated. Data are represented as mean ± SD and statistical analysis was performed using Student's unpaired t test (*P < .05, **P < .01)

FIGURE 4

Assessment of SORT1 expression requirement for TH19P01 peptide uptake in different cancer cell models. The uptake of 200 nM Alexa488‐labeled TH19P01 was evaluated in (A) human TNBC cells (MDA‐MB‐231), human colon cancer cells (MC38), and (B) murine ovarian cancer cells (ID8), and murine TNBC cells (4T1). This was performed in the absence (white bar) or presence (black bars) of excess unlabeled TH19P01 (50 µM), Neurotensin (NT, 10 µM) or Progranulin (PGRN, 1.5 nM). Data are represented as mean ± SD (ID8, n = 3; 4T1, n = 5; MDA‐MB‐231 and MC38, n = 2) and statistical analysis was performed using Dunnet's test (*P < .05, **P < .01). Inserts are representative immunodetections of Sortilin in whole cell lysates from each of the cell models tested

FIGURE 5

Schematic representation of docetaxel conjugation to the TH19P01 peptide. (A) Synthesis of TH1902 was performed as described in the Materials section. Briefly, DIEA was added dropwise to a solution of DoceSuOH and TBTU in DMSO in order to preactivate the DoceSuOH. The completion of preactivation was monitored by UPLC/MS, then a solution of TH19P01‐peptide in DMSO was added. The mixture was stirred at room temperature and the reaction was again monitored by UPLC/MS until completion. The reaction mixture was purified using a 30RPC resin column and an AKTA purifier system (10%‐80% ACN) to give TH19P01‐(SuDoce)2 or TH1902 as a white powder after lyophilization. (B) The purity (>95%) of the conjugate (peak 1) was evaluated by UPLC/MS, and (C) had an estimated m/z mass of 1852.91 (2+ charge)

FIGURE 6

In vitro anticancer properties of TH1902. A, A representative plot of the in vitro antiproliferative activities of TH19P01, docetaxel and TH1902 was determined from MTT assays in MDA‐MB‐231 as described in the Methods section. B, IC50 values were calculated using a log[inhibitor] vs response‐constant slope equation with GraphPad Prism software. IC50 values are presented as mean ± SD from human TNBC‐derived MDA‐MB‐231 and HCC‐70 cells. All assays were performed in quadruplicate (n = 3 experiments, except for TH1902 n = 4). C, A representative cell‐cycle phase distribution of MDA‐MB‐231 cells was determined based on cellular DNA content (propidium iodide staining) after a 2 h treatment with 4 µM docetaxel or 2 µM TH1902 then followed by a 22 h incubation in complete media alone. D, Quantification of MDA‐MB‐231 G2/M cell cycle arrest following docetaxel or TH1902 treatment. Data represent mean ± SD (n = 4)

FIGURE 7

Cell death induction by docetaxel and TH1902 in MDA‐MB‐231 cells. A, MDA‐MB‐231 cells were treated with vehicle, 10 µM docetaxel or 5 µM TH1902 for 24 h and cellular morphology was examined under a light microscope. B, MDA‐MB‐231 cells were treated for 5 or 24 h with 10 µM docetaxel (white bars), or with 5‐10 µM TH1902 (black bars). Cells were then harvested and the extent of apoptotic cell death determined by flow cytometry following staining with AnnexinV‐FITC and propidium iodide. Data are represented as mean ± SEM (n = 3, except for the 5 µM data where n = 7). C, TH1902‐mediated apoptosis in MDA‐MB‐231 cells was competed by an excess of free TH19P01 peptide (50 µM) or one of the SORT1 ligands neurotensin (NT, 10 µM) or progranulin (PGRN, 1 nM). Following 5 h of incubation, cells were stained with AnnexinV‐FITC/PI. All data are represented as mean ± SD (n = 3) and statistical analysis was performed using one‐way ANOVA with Dunnet's multiple comparison test (B and C) (*P < .05, **P < .01)

FIGURE 8

Effect of TH1902 on MDA‐MB‐231 microtubules and cell migration. A, MDA‐MB‐231 cells were treated with vehicle (DMSO), 2 µM docetaxel or 1 μM TH1902 for 24 h, fixed and immunostained with anti‐α‐tubulin antibody, DNA stained with DAPI, and then imaged using confocal fluorescence microscopy. Representative pictures from each condition are displayed. (B) Representative immunoblot showing the protein expression level of Bcl‐xL in MDA‐MB‐231 cells after treatment with docetaxel (100 nM) or TH1902 (50 nM) for 48 h. GAPDH was used as a loading control. Data are representative of three independent experiments. (C) Quantification of Bcl‐xL was performed by scanning densitometry. Data represent mean ± SD of arbitrary units and statistical analysis was performed using one‐way ANOVA with Dunnet's test compared to control, (*P < .05, n = 3). The impact of (D) docetaxel and (E) TH1902 on MDA‐MB‐231 cell migration was evaluated with and without siRNA‐mediated silencing of SORT1. Transfected cells were pretreated for 2 h with TH1902 (1 µM) or docetaxel (2 µM). Cells were harvested and their migratory potential assessed in real time as described in the Methods section. Data of normalized arbitrary units are represented as mean ± SD of two experiments performed in duplicate. Statistical analysis of data at 10 h was performed using one‐way ANOVA with Dunnet's multiple comparisons test (**P < .01)

FIGURE 9

Comparative efficacy of docetaxel and TH1902 on MDA‐MB‐231 xenograft tumor growth. A, Nude mice were subcutaneously implanted with MDA‐MB‐231‐Luc cells and then treated with vehicle, docetaxel or TH1902 as described in the Methods section. Intraperitoneal (i.p.) treatments are indicated by red arrows. B, Representative bioluminescence images were taken at 14 and 74 days post‐implantation of MDA‐MB‐231‐Luc cells into the mice. In the color bar, violet and red indicate the highest and lowest intensity of exposure, respectively. Results are expressed in terms of tumor luminescence intensity at day 0 post‐treatment for each mouse. C, Tumor burden was expressed as net luminescence intensity at days 14 and 74 post‐treatment. Data are represented as mean ± SEM (six mice/group). At day 74, all control mice were already sacrificed. Therefore, NA in this panel indicates that bioluminescence is not available (NA) for this group, whereas the bioluminescence in TH1902‐treated mice was below detectable levels (BDL)

FIGURE 10

Intravenous administration of TH1902 inhibits tumor growth in a MDA‐MB‐231 TNBC xenograft model. A, Nude mice were subcutaneously implanted with TNBC‐derived MDA‐MB‐231 cells and then treated with vehicle or with either docetaxel (15 mg/kg, three cycles) or TH1902 (35 mg/kg, five cycles) as described in the Methods section. Intravenous (i.v.) treatments are indicated by red arrows. B, Tumor volume progression measured as above for vehicle, docetaxel (3.75 mg/kg), or TH1902 (8.75 mg/kg) treatments. C, Representative bioluminescence images were taken at day 27 post‐implantation of MDA‐MB‐231‐Luc cells into the mice. In the color bar, violet and red indicate the highest and lowest intensity of exposure, respectively. D, Tumor burden was expressed as net luminescence intensity at day 27. Data are represented as mean ± SEM and statistical analysis was performed using one‐way ANOVA with Dunnet's test compared to control (A, n = 5 mice/group; B‐D, n = 6 mice/group, ***P < .001)

FIGURE 11

Intravenous administration of TH1902 inhibits tumor growth in a HCC‐70 TNBC xenograft model. A, Nude mice were subcutaneously implanted with TNBC‐derived HCC‐70 cells and then treated with vehicle or with either docetaxel (3.75 mg/kg, ¼ MTD) or TH1902 (8.75 mg/kg) as described in the Methods section. Intravenous treatments are indicated by red arrows. B, Tumor volume progression measured as above for vehicle, docetaxel (3.75 mg/kg), or TH1902 (8.75 mg/kg) treatments. Data are represented as mean ± SEM and statistical analysis was performed using one‐way ANOVA with Dunnet's test compared to control (vehicle, n = 7; docetaxel and TH1902, n = 6 mice/group; **P < .01)

FIGURE 12

Absence of hematotoxicity of TH1902 in mice. A, Athymic nude mice were treated with docetaxel at MTD (15 mg/kg/wk, three cycles) or with an equivalent dose of TH1902 (up to six cycles). Blood was collected 4 days before the initiation of treatments and 4 days after the first, third and sixth treatments to assess the drug effects on neutrophil blood cell counts. Neutrophils count was not available (NA) for mice treated with docetaxel since treatment was stopped after three cycles. Data are represented as mean ± SEM and statistical analysis was performed using a t‐test against initial values observed at treatment 0 (n = 5 mice for treatments 0, 1 and 6; n = 2 mice for treatment 3; *P < .05, **P < .01). B, Body weight of mice was monitored with relation to the treatments. The number of mice with variations in body weight (>5% and 10%) after three or six treatments is presented (n = 5 mice/group)

FIGURE 13

Preliminary pharmacokinetics of TH1902 in mice. Normal CD‐1 mice were injected i.v. with TH1902 (50 mg/kg). Plasma samples for pharmacokinetic characterization were collected at the indicated times after i.v. bolus injection of TH1902. TH1902 and docetaxel concentrations were determined by UPLC/MS as described in the Methods section. Both TH1902 and free docetaxel plasmatic concentrations were plotted as a function of time. Cmax, termination half‐life and AUC were estimated using the PK solution software. Data are represented as mean ± SEM (n = 3 to 6 mice per time point)