The human TNBC cell lines MDA-MB-231, MDA-MB-468, and MDA-MB-231-FLuc (ATCC, Manassas, VA, USA) were cultivated in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS, v/v) and 1% penicillin/streptomycin (Pen/Strep, v/v). The mouse TNBC cell lines 4T1 and 4T1-FLuc (ATCC) were maintained in RPMI-1640 medium with 10% FBS and 1% Pen/Strep. The seed train of Expi293F cells for LSR mAb production was maintained in Expi293 Expression Medium containing 4 mM GlutaMAX and 4 g/L glucose. All cell cultures were incubated at 37°C in a humidified incubator with 5% CO₂ (Caron, Marietta, OH, USA). Media and supplements were obtained from Fisher Scientific (Waltham, MA, USA) or Millipore Sigma (Burlington, MA, USA) unless otherwise specified.

First, the anti-human LSR mAb, targeting the Ig-like V-type extracellular domain (Met1-Asp259) of the surface receptor, was developed using the well-established hybridoma technology as previously described (32–34). Once mAb production was detected in the mice injected with immunogen, the splenocytes were collected and fused with myeloma cells Sp2/0-Ag14 (ATCC) to generate hybridoma cells. Two rounds of single cell cloning were performed to identify the top hybridoma clone with high affinity and mAb titer, quantified via enzyme-linked immunosorbent assay (ELISA) using the extracellular peptide as the coating antigen. Second, the top LSR mAb clone was engineered by fusing its variable region with the Fc region of a human IgG1 for future clinical application.

The engineered LSR mAb was produced from Expi293F cells using transient expression system (Gibco) fed with glucose (6 g/L) and L-GlutaMAX (6 mM). The antibody productions were performed in 2-L stirred-tank bioreactor with process parameters of Temp 37°C, agitation 140 rpm, DO 40% and pH 7.2, or in a 3-L shaker flask at Temp 37°C, agitation 130 rpm, and 8% CO₂ (33–36). Purification of LSR mAb was performed utilizing a Next Generation Liquid Chromatography system (Bio-Rad, Hercules, CA, USA) following our previous procedures (32–35, 37–40). Specifically, the spent medium containing the mAb product was loaded onto 1-mL or 5-mL Bio-Scale Mini UNOsphere SUPrA affinity chromatography cartridges (Bio-Rad) with buffer A comprised of 0.02 M Na₃PO₄ and 0.02 M Na₃C₆H₅O₇ (pH 7.5). The mAb was eluted with buffer B comprised of 0.1 M NaCl and 0.02 M Na₃C₆H₅O₇ (pH 3.0).

The LSR mAb-DM1 ADC was constructed using our previously established conjugation platform (32, 34, 40) with minor modifications. Briefly, the purified LSR mAb was neutralized, desalted via buffer exchange, and concentrated to 5 mg/mL with a 10 kDa MWCO PES concentrator. During the conjugation reaction, 5 mg/mL mAb, 10 mg/mL sulfo-SMCC linker, and 10 mM of DM1 were mixed at a molar ratio of 1:14:18.2. After incubation at 37 °C for 2 hours, the constructed ADC was purified using a Protein A column, desalted using a 10 kDa Slide-A-Lyzer dialysis cassette, and further concentrated with a 10 kDa PES concentrator.

The purity and drug-antibody ratio (DAR) of the ADC were analyzed using a High-Performance Liquid Chromatography (HPLC) system (Shimadzu, Columbia, MD, USA) equipped with a MAbPac hydrophobic interaction chromatography (HIC)-butyl column (5 µm, 4.6x100 mm). Two aqueous mobile phases with flow rates of 1.0 mL/min were utilized: Buffer A, 1.5 M (NH₄)₂SO₄ and 50 mM Na₃PO₄ at pH 7.0 and Buffer B, 50 mM Na₃PO₄ at pH 7.0. UV/Vis spectroscopy was used to confirm ADC conjugation and mass spectrometry was used to quantitate DAR at University Mass Spectrometry and Proteomics facility.

The human and mouse TNBC cells were seeded in 96-well plates at densities of 10,000 cells (MDA-MB-468), 1,000 cells (MDA-MB-231), and 500 cells (4T1) per well. The cells in each well were treated with LSR mAb-DM1 or free DM1 at final concentrations of 0, 2, 4, 10, 20, 50, 100, or 200 nM. After a 5-day incubation at 37°C and 5% CO₂, relative viability was tested using the MTT Cell Proliferation Assay Kit. The absorbance at 570 nm was measured using a microplate reader (BioTek, Winooski, VT, USA).

The surface binding of engineered LSR mAb to TNBC MDA-MB-231, MDA-MB-468 and 4T1 cells was tested using flow cytometry as described before (41–46). Specifically, the anti-LSR mAb was labeled with fluorescent Alexa Fluor™ 647 kit. Approximately 1×10⁶ (6) TNBC cells were stained with 5 µg of LSR mAb-AF647 or saline (control) at 37 °C for 60 minutes. After washing with PBS three times, the stained cells were analyzed on a BD LSRII flow cytometer (BD Biosciences, San Jose, CA, USA) and data were processed using FlowJo software. The gating was set so that the negative control (i.e. TNBC cells stained with saline) had <0.5% fluorescent population.

The internalization of LSR mAb into TNBC cells was assessed using live-cell confocal microscopy. Approximately 10,000 MDA-MB-468 cells were seeded in 35-mm glass-bottom dishes (Cellvis, Mountain View, CA, USA) in 1.5-mL of medium. As described before (47), the cytoplasm of cells was stained with BacMam GFP Transduction Control, and the nucleus was visualized with NucBlue™ Live ReadyProbes Reagent following the manufacturer’s instructions. The LSR mAb labeled with cyanine-5.5 fluorescent dye (Lumiprobe, Hunt Valley, MD, USA) was added to TNBC cultures at a final concentration of 5 µg/mL and incubated for 12 hours. Live-cell images were acquired using an A1R-HD25 confocal microscope (Nikon, Melville, NY, USA).

The 6-week-old NOD SCID gamma (NSG) female mice (Jackson Lab, Bar Harbor, ME, USA) were subcutaneously (s.c.) injected with 10×10⁶ MDA-MB-231-FLuc cells to establish TNBC xenograft models. When tumor volumes reached ~100 mm³, 80 µg of LSR mAb-Cy5.5 was administered intravenously (i.v.) via tail vein. After 24 hours, live-animal imaging was collected at an excitation/emission wavelength of 660/710 nm following intraperitoneal (i.p.) injection of the FLuc substrate luciferin. Mice were then sacrificed to collect major organs, such as brain, heart, lung, spleen, liver and kidneys, for ex vivo imaging.

A total of 10x10⁶ MDA-MB-231-Flux cells were s.c. injected into 6-week-old female NSG mice (Jackson Lab). Once tumor volumes reached ~50 mm³, the mice were randomized into three groups (n = 6) and received i.v. administrations of saline (control), 12 mg/kg ADC, or 24 mg/kg ADC every three or four days (i.e. twice each week). Tumor dimensions were measured with a caliper and volumes were calculated using the formula (length x width ²)/2. IVIS imaging was also performed to monitor tumor growth. Mouse body weight was measured twice per week. At the end of the study, all tumor tissues and major organs, such as brain, heart, lung, kidneys, liver, and spleen, were collected for further post-treatment analysis.

The 6-week-old BALB/cJ female mice were s.c. implanted with 2x10⁶ 4T1-Flux cells per mouse. Tumor growth and body weight were monitored twice each week. When the average tumor volume reached ~50 mm³, mice were randomized into two groups (n = 4) and i.v. injected with either saline (control) or 24 mg/kg ADC every three or four days (i.e. twice each week). The study concluded when ulcerations reached 2 mm, after which tumors and organs were collected for further analysis.

A TNBC patient tissue microarray (TMA, Cat# BR1102, TissueArray, Derwood, MD, USA) and normal organ tissue microarray (Cat# FDA662d, TissueArray) were stained with our engineered anti-LSR mAb following a published IHC staining protocol (47). The stained TMA slides were scanned with a Lionheart FX automated microscope (BioTek, Winooski, VT, USA). LSR expression was quantified with ImageJ and the score was calculated using the formula ((red/blue)_TNBC/(red/blue)_{positive control})-1)x100. High, medium, and low or no expression levels were defined as scores of >0.5, 0.5 to -0.5 and <-0.5 respectively, using positive control as baseline. Proliferation and apoptosis in the tumor microenvironment after treatment were evaluated using a similar IHC procedure to stain the tissue slides.

The LSR RNA data for breast cancers (TNBC, ER⁺/PR⁺, HER2⁺) and normal breast tissue were obtained from The Cancer Genome Atlas (TCGA) via the UCSC Xena platform (https://xenabrowser.net/datapages/). The raw expression data, originally in fragments per kilobase of transcript per million mapped reads (FPKM), were converted to transcripts per million (TPM) for normalization. Student’s t-test was conducted to assess the differential expression of LSR between various breast cancer subtypes and normal tissue.

Lysates of TNBC cells were prepared using RIPA buffer and quantified using the Pierce protein assay kit. After separation on a NuPAGE 4-12% gradient SDS-PAGE gel, host cell proteins were transferred onto a PVDF membrane using a power supply (Bio-Rad Laboratories, CA, USA). The PVDF membrane was blocked in 5% non-fat milk in TBST buffer for 1 hour, stained with primary antibodies, i.e. our engineered LSR mAb (2 µg/mL) or anti-β-tubulin antibody (1:1000, Abcam), overnight at 4 °C, and washed three times with TBST. Following the washes, PVDF membrane was incubated with anti-human IgG-HRP (1:5000) or anti-rabbit IgG-HRP (1:3000) in 5% non-fat milk for 1 hour at room temperature. After color development, the washed PVDF membrane was imaged with Azure 300 imaging system (Azure Biosystems, Dublin, CA, USA).

Paraffin-embedded tumor and organ slides were dewaxed in xylene, rehydrated using a graded ethanol series (100% → 50%) and distilled water, and stained with hematoxylin. After dipping in 1% HCl and 70% ethanol, slides were immersed in 1% NH₄OH for blue color development, followed by a 30-second eosin stain. Finally, the slides were dehydrated in 95% and 100% ethanol and cleared with xylene. Images of the H&E-stained slides were captured using a Lionheart FX Automated Microscope (BioTek, Winooski, VT, USA).

Experimental data were analyzed and presented as mean ± standard error of the mean (SEM). Within-animal changes were evaluated using paired t-tests. Tumor growth and body weight were compared among groups using a two-way ANOVA followed by post-hoc (Dunnett’s) analysis. GraphPad Prism was used for statistical analysis, a significance threshold of P-value< 0.05 was used for all tests. All adjusted P-values were calculated using the Benjamini-Hochberg method to control the FDR.

All data generated during this study are included in the main text of this article. Most materials used in the study, except the developed and engineered LSR mAb, are commercially available. The commercial NSG (NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ) mice and BALB/cJ mice were purchased from Jackson Laboratory.

To assess LSR protein expression in TNBCs, we performed IHC staining on a patient-derived tumor microarray (TMA, n = 110) using our developed LSR mAb ( Figure 1A ). Representative IHC images of tissue cores displaying varying LSR expression levels are shown in Figure 1B . High LSR expression was observed in 35 patient samples (32.7%), and moderate expression in 32 samples (29.9%), collectively accounting for 62.6% of TNBC cases ( Figure 1B ). To complement these findings, LSR transcript expression was analyzed using The Cancer Genome Atlas (TCGA) dataset. As depicted in Figure 1C , LSR mRNA levels were significantly elevated in TNBC compared to normal breast tissue. These transcriptomic data align with the TMA protein expression results, suggesting LSR as a promising therapeutic target. In addition to TNBC, the ER⁺/PR⁺ patients and HER2⁺ patients also showed high counts of LSR transcripts. Therefore, targeting LSR could cover and benefit multiple subtypes of breast cancer. LSR expression was further evaluated in TNBC cell lines, including human MDA-MB-231, human MDA-MB-468 and mouse 4T1. Western blot analysis revealed consistently high LSR expression across these three cell lines ( Figure 1D ), corroborating the TMA protein and TCGA transcript analyses. This study used MDA-MB-231 and 4T1 cell lines for in vitro anti-cancer cytotoxicity assays and in vivo anti-TNBC efficacy evaluations. Collectively, these results underscore LSR as a promising therapeutic target for TNBC and potentially other breast cancer subtypes.

We developed a new mAb to target the extracellular domain (Met1-Asp259) of human LSR (UniProtKB A0A8Z5ABK9). The top hybridoma clone 3C4A7, which had high titer and binding affinity for the extracellular domain of CTYQMTSTPTQPIV, was identified ( Figure 2A ). After sequencing, the top clone was further engineered by fusing its variable region with the Fc region of a human IgG1 ( Figure 2B ). The engineered LSR mAb was produced in fed-batch suspension culture using transient expression system. As shown in Figure 2C , 41.6 mg/L of mAb was produced from a 9-day culture with maximum viable cell density (VCD) of 4.9x10⁶ cells/mL. Glucose (4 g/L) and GlutaMAX (4 mM) were added to the production culture on Days 3 and 5. Purified mAb was generated from affinity purification using liquid chromatography system ( Figure 2D ).

To evaluate the surface binding of our engineered anti-LSR mAb to TNBC cells, we performed flow cytometry on human TNBC cell lines (MDA-MB-231 and MDA-MB-468) and a murine TNBC cell line (4T1) ( Figure 3A ). The mAb exhibited high surface binding, with rates of 100% in MDA-MB-468, 95.6% in MDA-MB-231, and 73.3% in 4T1. The extracellular domain (residues 89-259) of human LSR (UniProt: Q86X29), used as the immunogen, shares 91% amino acid sequence similarity with the corresponding extracellular region (residues 1-171) of mouse LSR (UniProt: Q99KG5). This high homology accounts for the cross-species reactivity of our anti-human LSR mAb, though binding to murine TNBC cells was comparatively lower. The ability to target both human and murine LSR supports the use of mouse models for future preclinical studies. Additionally, IHC staining of TNBC patient tissues ( Figure 1A ) confirmed the specificity of our chimeric anti-LSR mAb for human TNBC.

Next, mAb internalization in TNBC cells was investigated using live-cell confocal microscopy. As presented in Figure 3B , the Cy5.5 labelled LSR mAb (displayed as red color) was detected in the cytoplasm of MDA-MB-468 cells expressing GFP (green color) 12 hours after incubation. As previously reported (33, 40, 47), the mAb is internalized via receptor-mediated endocytosis and trafficked to late endosomes for lysosomal degradation, which releases the conjugated cytotoxic drugs into the cytoplasm, causing cell death (48).

Finally, the MDA-MB-231-FLuc xenografted mouse models were established to assess the in vivo TNBC-specific targeting and biodistribution of the LSR mAb. Live-animal IVIS imaging displayed colocalization of the Cy5.5 fluorescence signal (mAb) with the FLuc bioluminescence signal (TNBC tumor) 24 hours after tail vein injection ( Figure 3C ). The ex vivo IVIS analysis confirmed strong accumulation of LSR mAb (Cy5.5) in the tumor, with no detectable signal in the brain, heart, lung, spleen, liver, and kidneys ( Figure 3D ). These results demonstrate that our engineered LSR mAb selectively targets TNBC in vivo with minimal off-target effects, highlighting its potential to effectively deliver therapeutic payloads for cancer treatment.

The LSR-targeting ADC was constructed by conjugating the LSR mAb with DM1 via sulfo-SMCC linker ( Figure 4A ). HPLC characterization of ADC confirmed the homogenous conjugation of mAb and chemotherapy with a conversion rate of ~100% ( Figure 4B ). SDS-PAGE analysis verified the structural integrity of the LSR ADC ( Figure 4C ). The anti-TNBC cytotoxicity of free DM1 and LSR mAb-DM1 ADC was assessed in human and mouse TNBC cell lines. The mass spectrometry analysis showed that the drug to antibody ratio (DAR) of LSR ADC was 3.91 ( Figure 4D ). As presented in Figure 4E , free DM1 exhibited IC₅₀ values of 6.6 nM (4T1), 10.7 nM (MDA-MB-231) and 25.5 nM (MDA-MB-468). The IC₅₀ values for LSR mAb-DM1 ADC were 46.0 nM (4T1), 43.1 nM (MDA-MB-231) and 35.5 nM (MDA-MB-468) ( Figure 4F ). These cytotoxicity results illustrate that LSR-DM1 ADC can effectively kill TNBC cells in vitro. The ADC has lower cytotoxicity to TNBC cells than free drug at low concentrations of 10–40 nM, which can be explained by the multi-step (receptor binding, internalization, and drug release) targeted drug delivery process.

The anti-tumor efficacy study was first conducted in human MDA-MB-231-FLuc xenografted NSG mice. As described in Figure 5A , the TNBC tumor volumes reached 612.2 mm³ (saline), 435.8 mm³ (12 mg/kg-BW LSR mAb), 212.4 mm³ (12 mg/kg LSR mAb-DM1 ADC), and 95.7 mm³ (24 mg/kg ADC) on Day 32. Compared to the saline group, LSR mAb reduced tumor burden by 29%, which could be attributed to the inhibition of lipid uptake or synthesis by the mAb. The ADC at 12 and 24 mg/kg reduced tumor burden by 65% and 85%, respectively, demonstrating dose-dependent tumor growth inhibition. TNBC tumor volume, assessed via FLuc flux at the end of treatment, corroborated that the 24 mg/kg ADC dose effectively suppressed TNBC growth ( Figure 5B ). The IVIS images, collected before treatment and on Day 31, also confirmed the ADC’s inhibitory effect on tumor growth ( Figure 5C ). H&E staining of paraffin-embedded tumor tissue sections revealed reduction of tumor intensity in ADC treatment groups, with no evident cancer cell death in the saline group ( Figure 5D ). IHC staining with anti-Ki67 antibody confirmed that ADC treatment significantly inhibited TNBC cell proliferation ( Figure 5E ).

The anti-TNBC efficacy of the LSR mAb-DM1 ADC was validated in immunocompetent BALB/cJ mice bearing 4T1 xenografts. Treatment with 24 mg/kg ADC significantly reduced tumor volume and caused tumor shrinkage ( Figure 6A ), without affecting mouse body weight, suggesting minimal systemic toxicity ( Figure 6B ). Live-animal IVIS imaging confirmed the effectiveness of ADC treatment ( Figure 6C ), as indicated by the undetectable or weak FLuc signal in the treatment group. H&E staining of tumor tissues harvested at the end of the study revealed severe tumor cell death in the treatment group ( Figure 6D ). The tumor volume slightly increased on Days 18–20 in Figure 6A , but the dead cells in H&E staining indicated that this increase could be caused by non-cellular factors (e.g., inflammation, vascular formation, necrosis) instead of tumor growth recurrence. IHC staining of tumor sections showed that the ADC induced apoptosis, as indicated by the upregulation of cleaved caspase 3, and inhibited cancer cell proliferation, as displayed by the downregulation of Ki67 ( Figure 6E ). Live-cell confocal microscopy demonstrated that anti-LSR mAb reduced lipid uptake (green fluorescence) in MDA-MB-231 cells ( Figure 6F ), consistent with literature reports that LSR promotes lipid uptake and fatty acid oxidation in gastric cancer (29, 49).

In animal studies, ADC treatment at 12 and 24 mg/kg did not cause body weight loss ( Figure 7A ), indicating low systemic toxicity. H&E staining of major organs (heart, brain, lung, liver, kidney, spleen) showed no signs of inflammation, apoptosis, necrosis, or damage in the mAb or ADC groups ( Figure 7B ). IHC staining of 33 normal human tissues (n = 2) with the engineered LSR mAb revealed no significant off-target binding, as confirmed by whole-slide scanning ( Figure 7C ). High-resolution images of heart, brain, lung, liver, kidney, and spleen showed no significant mAb binding ( Figure 7D ). These findings suggest that the LSR mAb selectively targets LSR-overexpressing TNBC cells with minimal off-target effects on normal tissues.