A novel CD44-targeting aptamer recognizes chemoresistant mesenchymal stem-like TNBC cells and inhibits tumor growth

Abstract

Triple-negative breast cancer (TNBC) represents a significant therapeutic challenge owing to the scarcity of targeted medicines and elevated recurrence rates. We previously reported the development of the nuclease-resistant RNA sTN58 aptamer, which selectively targets TNBC cells. Here, sTN58 aptamer was employed to capture and purify its binding target from the membrane protein fraction of cisplatin-resistant mesenchymal stem-like TNBC cells. Mass spectrometry in conjunction with aptamer binding assays across various cancer cell lines identified CD44 as the cellular target of sTN58. By binding to CD44, sTN58 inhibits the invasive growth and hyaluronic acid-dependent tube formation in chemoresistant TNBC cells, where CD44 serves as a key driver of tumor cell aggressiveness and stem-like plasticity. Moreover, in vivo studies demonstrated the aptamer's high tumor targeting efficacy and its capacity to significantly inhibit tumor growth and lung metastases following intravenous administration in mice with orthotopic TNBC. Overall, our findings reveal the striking potential of sTN58 as a targeting reagent for the recognition and therapy of cancers overexpressing CD44.

Keywords: Aptamer; Biomarker identification; CD44; Chemoresistant triple-negative breast cancer; Targeted cancer therapy.

Graphical abstract

Fig. 1

Identification of the target of sTN58 aptamer on TNBC cell surface. (A) Schematic representation of biotin-sTN58-mediated affinity purification. Membrane-protein fraction from Cis-Pt-R cells were subjected to a preclearing step to remove non-specific components prior to the sTN58-mediated precipitation. The colloidal Blue-stained SDS-PAGE (10 %) displayed is utilized for the analysis of target purification mediated by the sTN58 aptamer. The molecular weights of protein markers are reported. Lane 1, molecular markers; lane 2, membrane extracts; lane 3, 15 μg aliquot of unbound proteins from SCR-mediated purification; lane 4, proteins captured with sTN58. Red boxes indicate the regions excised for MS analyses. (B) Comparison of transcript expression values of best candidates in different BC cell lines. The normalized transcript expression values (nTPM), according to HPA, are reported relative to MDA-MB-231 target cells, whose expression level is arbitrarily set to 1. Box indicates the 5 candidates chosen for experimental validation. (C) Immunoblot analysis of EphA2, CD44, integrin β1, myoferlin, liprin β1 and ZO-1, and of the housekeeping proteins α-tubulin and vinculin. The molecular weights of protein markers are reported. Black dashed lines delineate the boundary between non-contiguous lanes of the same gel. (D) The histogram shows the relative fold-change in expression levels of the indicated proteins compared to the housekeeping protein α-tubulin or vinculin, normalized to MDA-MB-231 target cells, whose expression level is arbitrarily set to 1. (E) Binding affinity (1/Kd) of TN58 aptamer to the indicated cell lines expressed relative to MDA-MB-231 target cells. Dose response curves and binding affinity calculations for MDA-MB-231 and Cis-Pt-R and Dox-R chemoresistant cells, as well as non-TNBC BT-474, MCF-7 and A431 cells were previously reported [6,7]. The dose response curve used for Kd calculation in relation to BT-549 is shown in Fig. S2. "NB", no binding.

Fig. 2

sTN58 binds to CD44-positive TNBC cell lines. Following 5 min incubation at RT with 2 μM Alexa 647-sTN58, Cis-Pt-R (A) or BT-474 (B) cells were stained with CD44 Ab, visualized by confocal microscopy, and photographed. Alexa 647-SCR was used as a negative control. Alexa 647-sTN58, CD44 Ab and nuclei are visualized in red, green, and blue, respectively. Magnification 63×, 1.0× digital zoom, scale bar = 10 μm. Co-localization results appear yellow in the merged images (Overlap Coefficient 0.74). All digital images were captured under identical settings to allow direct comparison of staining patterns. (C) Flow cytometry analyses of Cis-Pt-R, MDA-MB-231, BT-549 and BT-474 cells treated with CD44-PE Ab or Alexa 647-sTN58. (D) Quantification of the geometric mean fluorescence intensity (gMFI) of Alexa 647-sTN58- or CD44-PE Ab-treated cells normalized to the gMFI of the untreated cells. Bars depict mean ± SD of at least two independent experiments. ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001 relative to untreated cells; ns, no significant.

Fig. 3

CD44 silencing results in reduced sTN58 binding. (A, E) Cis-Pt-R (A) and BT-549 (E) cells were left untreated or transfected with si-CD44 or siRNA ctrl. At 48 h post-transfection, cells were harvested, and cell lysates prepared and immunoblotted with CD44 Ab. Equal loading was confirmed by immunoblot with anti-α-tubulin antibody. Molecular weights of protein markers are reported. (B, F) The histogram depicts the densitometric ratio of CD44 expression to α-tubulin. Values are shown relative to the untreated control, arbitrarily set to 1. ∗P < 0.05, ∗∗P < 0.01 relative to siRNA ctrl. (C, G) Binding of CD44-PE Ab (left) and Alexa 647-sTN58 (right) to Cis-Pt-R (C) and BT-549 (G) cells following 48 h transfection with si-CD44 (green) and siRNA ctrl (gray). (D, H) The histogram shows gMFI of si-CD44-transfected cells treated with sTN58 aptamer or CD44 Ab, normalized to the gMFI of untreated cells, and expressed as percentage with respect to siRNA ctrl-transfected cells. ∗∗∗∗P < 0.0001 relative to siRNA ctrl. (I) Immunoblot analysis of CD44 and the housekeeping protein α-tubulin. The molecular weights of protein markers are reported. (J) The histogram shows the relative fold-change in CD44 expression levels compared to α-tubulin, normalized to MDA-MB-231 target cells, whose expression level is arbitrarily set to 1. ∗P < 0.05, ∗∗∗P < 0.001 relative to MDA-MB-231. (K) Flow cytometry analyses of Cis-Pt-R, MCF 10A, THP-1 and HS-5 cells treated with Alexa 647-sTN58. (L) Quantification of the gMFI of Alexa 647-sTN58-treated cells normalized to the gMFI of the untreated cells. ∗∗P < 0.01, ∗∗∗∗P < 0.0001 relative to untreated cells; ns, no significant. In B, D, F, H, J, L, bars depict mean ± SD of at least two independent experiments.

Fig. 4

sTN58 aptamer and CD44 Ab colocalize with integrin β1 Ab on Cis-Pt-R cells. (A) Following 5 min incubation at RT with 2 μM Alexa 647-sTN58, Cis-Pt-R cells were stained with integrin β1 Ab, visualized by confocal microscopy, and photographed. (B) Cell lysates from Cis-Pt-R cells left untreated or treated for 48 h with 100 nM siRNA ctrl or si-ITGB1 were analyzed by immunoblotting with integrin β1, CD44 and anti-α-tubulin antibodies. Molecular weights of protein markers are reported. (C) The histogram shows the protein expression/α-tubulin ratio based on the densitometric signals. Values are shown relative to untreated samples, arbitrarily set to 1. (D) Binding of integrin β1-APC-Cy7 Ab (left) and Alexa 647-sTN58 (right) to Cis-Pt-R cells following 48 h transfection with siRNA ctrl (gray) and si-ITGB1 (pink). (E) The histogram shows gMFI of si-ITGB1-transfected cells treated with Alexa 647-sTN58 or integrin β1-APC-Cy7 Ab, normalized to the gMFI of untreated cells, and expressed as percentage with respect to siRNA ctrl-transfected cells. Bars depict mean ± SD of two independent experiments. ∗∗∗P < 0.001; ns, no significant. (F) Confocal microscopy analyses of A431 cells treated with sTN58 and stained with integrin β1 Ab, as in A, or stained with CD44-PE and integrin β1 antibodies. Alexa 647-SCR was used as a negative control. In A, F, aptamers and CD44-PE Ab are visualized in red, integrin β1 Ab in green and nuclei in blue. All digital images were captured at the same setting to allow direct comparison of staining patterns. Magnification 63×, 1.0× digital zoom, scale bar = 10 μm. Co-localization results appear yellow in the merged images. Arrowheads indicate some co-localization points between sTN58 and integrin β1 Ab (Overlap Coefficient, 0.80). (G) Flow cytometry analyses of A431 cells treated with CD44-PE Ab, Alexa 647-sTN58 and integrin β1-APC-Cy7 Ab. (H) Quantification of the gMFI of sTN58-, CD44 PE- and integrin β1-APC-Cy7-treated cells normalized to the gMFI of the untreated cells. Bars depict mean ± SD of three independent experiments. ∗∗∗∗P < 0.0001 relative to untreated cells; ns, no significant.

Fig. 5

sTN58 inhibits invasive growth and vessel like-structures formation in 3D cell cultures. (A) Representative phase-contrast images of Cis-Pt-R cells grown in 2D or in Matrigel (3D) in the presence of 500 nM sTN58 or SCR for the indicated time points. (B) The invasive ability of Cis-Pt-R cells at 5 days is expressed as the number of colonies and branches per field. (C, D) Invasion of Cis-Pt-R (C) and Dox-R (D) cells toward 10 % FBS was analyzed by transwell invasion assay in the presence of 500 nM sTN58 or SCR for 72 h. Photographs of a representative experiment are shown. (E, F) Data are presented as percentage of invaded cells in the presence of sTN58 compared with SCR control. (G, I) Representative phase-contrast images of Cis-Pt-R cells (G) and Dox-R (I) grown on Matrigel in the presence of HA and treated with 500 nM sTN58 or SCR for 24 h. (H, J) Tube formation ability is expressed as the number of junctions and loops per field. (A, C, D, G, I) Magnification 10×, scale bar = 200 μm. (B, E, F, H, J) Bars depict means ± SEM of at least two independent experiments. ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001 relative to SCR-treated cells. (K) AF3 molecular docking model of sTN58-CD44 binding complex; sTN58, HA-binding site and CD44 folded domain are shown in blue, white and green, respectively.

Fig. 6

sTN58 selectively targets CD44-positive 4T1 xenografts. (A) Representative confocal images of 4T1 (8.0 × 10⁴ cells) incubated with 2 μM Alexa 647-sTN58 or Alexa 647-SCR and then fixed and stained with CD44 Ab. Aptamers, CD44 Ab and nuclei are visualized in red, green and blue, respectively. Co-localization results appear yellow in the merged images (Overlap Coefficient, 0.76). All digital images were captured at the same setting to allow direct comparison of staining patterns. Magnification 63× , 1.0× digital zoom, scale bar = 10 μm. (B) Binding of CD44-PE Ab (left) and Alexa 647-sTN58 (right) to 4T1 cells using flow cytometry. (C) Quantification of the gMFI of sTN58 aptamer- or CD44 Ab-treated cells normalized to the gMFI of the untreated cells. Bars depict mean ± SD of two independent experiments. ∗∗P < 0.01, ∗∗∗P < 0.001 relative to untreated cells. (D) Mice bearing mammary fat pad orthotopic 4T1 tumors were i.v. injected with 0.75 nmol of either NIR-sTN58 or NIR-SCR and analyzed with FMT at the indicated time points; Pre, before injection. Representative volume renderings taken at the same color gating for NIR-sTN58 and NIR-SCR injected mice are shown. (E) The amount of fluorescence (pmol) was quantified in specific VOIs encompassing the tumor in the animal. (F) Representative ex vivo FRI imaging of tumor and major organs (liver, kidneys, spleen, lung, heart and muscle) harvested from mice at 24 h post-injection of NIR-sTN58 and NIR-SCR. (G) The histogram indicates the mean FRI Signal Intensity of tumors and organs in the two groups. (E, G) Bars depict mean ± SD. ∗∗P < 0.01; ∗∗∗P < 0.001 relative to NIR-SCR; ns, no significant. (H) Plasma pharmacokinetic profile of sTN58. Concentration of aptamer is shown as a function of time following a single i.v. injection in Balb/c mice. Data are presented as the mean ± SEM.

Fig. 7

Effect of sTN58 treatment on tumor growth and lung metastases formation. (A) Mice bearing mammary fat pad orthotopic 4T1 tumors were i.v. injected with sTN58 or SCR aptamer (at day 0, 3, 5, 10 and 13, indicated by arrowheads). Tumor growth was monitored by calipers over time and experimental raw data (expressed as fold increase) were interpolated with no curve fitting or regression analysis. Day 0 marks the start of treatments. (B) Mice body weight was measured at the indicated days and the mean weight of each group is shown. (A, B) The mean ± SD (n = 5) was calculated for all the groups. ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001 relative to SCR. (C) Shown are images from one representative tumor sample for each treatment group stained for H&E (upper panels) or with Ki-67 antibody (lower panels). Arrows identify features of cancer cells, as described in the text. Magnification 40×, scale bar = 50 μm. (D) Ki-67 proliferation index was calculated as percentage of Ki-67 positive cells/total cell count for randomly selected 40× microscopic fields considering the SCR-group as 100 %. Bars depict mean ± SD. (E) Lysates from recovered tumors were immunoblotted with the indicated antibodies. Equal loading was confirmed by immunoblot with anti-vinculin or anti-α-tubulin antibody. Molecular weights of protein markers are reported. (F) The histogram shows the relative fold of expression of the indicated proteins against the housekeeping protein α-tubulin or vinculin. Each data point represents the sample from an individual mouse (n = 5). (G) Shown are images from one representative lung sample for each treatment group stained for H&E. Magnification 2×; scale bar = 1000 μm. Arrows point to metastasis of breast cancer in the lung. (H) The histogram shows the ratio between the metastasis area and the entire lung tissue section, expressed in percentage. Bars depict mean ± SEM (n = 5). (A, D, F, H) ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001 relative to SCR.