Published online Jan 15, 2024.
https://doi.org/10.4048/jbc.2023.0177
Potential Perturbations of Critical Cancer-regulatory Genes in Triple-Negative Breast Cancer Cells Within the Humanized Microenvironment of Patient-derived Xenograft Models
,1
Jihui Yun,2,3
Hye-Youn Son,4
Woohang Heo,4
Jong-Il Kim,2,3,5,6
and Hyeong-Gon Moon2,5,7,8
Abstract
Purpose
In this study, we aimed to establish humanized patient-derived xenograft (PDX) models for triple-negative breast cancer (TNBC) using cord blood (CB) hematopoietic stem cells (HSCs). Additionally, we attempted to characterize the immune microenvironment of the humanized PDX model to understand the potential implications of altered tumor-immune interactions in the humanized PDX model on the behavior of TNBC cells.
Methods
To establish a humanized mouse model, high-purity CD34+ HSCs from CB were transplanted into immunodeficient NOD scid γ mice. Peripheral and intratumoral immune cell compositions of humanized and non-humanized mice were compared. Additionally, RNA sequencing of the tumor tissues was performed to characterize the gene expression features associated with humanization.
Results
After transplanting the CD34+ HSCs, CD45+ human immune cells appeared within five weeks. A humanized mouse model showed viable human immune cells in the peripheral blood, lymphoid organs, and in the tumor microenvironment. Humanized TNBC PDX models showed varying rates of tumor growth compared to that of non-humanized mice. RNA sequencing of the tumor tissue showed significant alterations in tumor tissues from the humanized models. tumor necrosis factor receptor superfamily member 11B (TNFRSF11B) is a shared downregulated gene in tumor tissues from humanized models. Silencing of TNFRSF11B in TNBC cell lines significantly reduced cell proliferation, migration, and invasion in vitro. Additionally, TNFRSF11B silenced cells showed decreased tumorigenicity and metastatic capacity in vivo.
Conclusion
Humanized PDX models successfully recreated tumor-immune interactions in TNBC. TNFRSF11B, a commonly downregulated gene in humanized PDX models, may play a key role in tumor growth and metastasis. Differential tumor growth rates and gene expression patterns highlighted the complexities of the immune response in the tumor microenvironment of humanized PDX models.
INTRODUCTION
Among the different categories of breast cancer in humans, triple-negative breast cancer (TNBC), characterized by the absence of estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2, represents the most aggressive form, with no specific targeted treatment currently available [1, 2]. Efforts have been made to propose novel targeted approaches, such as mammalian target of rapamycin or androgen receptor inhibition, using preclinical models of TNBC based on the genomic profiling of heterogeneous diseases [3].
Patient-derived xenograft (PDX) models are attractive alternatives to traditional cell-line models because they can recapitulate the inter- and intratumor heterogeneity of solid tumors [4]. By characterizing the tumor heterogeneity, one can identify potential therapeutic targets or understand the molecular mechanisms of drug resistance [5, 6]. PDX models are also useful preclinical models for evaluating the in vivo efficacy of novel targeted therapies [7]. For instance, the potential value of FGFR or CDK targeting in TNBC was demonstrated by using a PDX model-based approach [8, 9].
A major disadvantage of PDX models is the absence of an intact immune microenvironment. The immune microenvironment is a major factor that determines tumor progression and metastasis in TNBC [10]. To address this issue, researchers have established humanized mouse models harboring intact human immune cells to study various human diseases, including breast cancer [11, 12, 13]. While these efforts have demonstrated that humanized PDX models are suitable for studying the efficacy of novel immunotherapeutics [11], they have also shown that the humanization process results in changes in the molecular profiles of the tumor as well as in tumor-microenvironment interactions [12, 14].
In this study, we investigated changes in the molecular characteristics of xenografted human breast cancers in humanized TNBC PDX models. Our data demonstrated distinct transcriptomic changes in humanized TNBC PDX models, including key genes that regulate tumor cell phenotypes, such as tumor necrosis factor receptor superfamily member 11B (TNFRSF11B).
METHODS
Participants and cord blood (CB) collection
Between September 2020 and November 2023, 40 CB units were evaluated at Seoul Metropolitan Public Cord Blood Bank (Allcord, Seoul, Korea). The CB units were generously donated by pregnant Korean women who provided informed consent. Participants completed self-administered medical questionnaires, while obstetricians provided perinatal details, including gestational age, birth weight, sex, and delivery method. Written informed consent and donation-related medical questionnaires were approved by the Institutional Review Board of Seoul National University Boramae Hospital. The collection and processing of CB units adhered to the guidelines outlined in the “Act on Cord Blood Management and Research” and followed established standard operating procedures. Furthermore, exemption determinations were obtained from the Institutional Review Board of Seoul National University Boramae Hospital (E-2111-005-1267).
Human CD34+ hematopoietic stem cells (HSCs) isolation and purification from CB
Whole mononuclear cells (MNCs) were isolated from blood samples by Ficoll-Plaque (Cytiva, Danaher Corporation, Washington, D.C., USA) density gradient centrifugation, which is a straightforward and efficient technique. Varying densities of different cellular components were exploited, resulting in the separation of the cells into distinct layers within the solution based on their density. A suitable CD34+ secondary antibody was conjugated to the beads using a magnetic bead separation system following the guidelines provided by the manufacturer (MACS, North Rhine-Westphalia, Germany). Subsequently, the cells conjugated with beads were introduced into a magnetic separation column and cell separation was performed. The separated cells were collected and used for > 90% purification of CD34+ cells by flow cytometric (fluorescence activated cell sorting [FACS]) analysis (Supplementary Figure 1). The hCD34− MNCs were stored at −80°C until further use.
Humanized PDX model
NOD.Cg-Prkdcscid Il2rgtm1Wjl mice, commonly known as NOD scid γ (NSG) mice (Jackson Laboratory, Bar Harbor, USA), were utilized as the recipient strain for intravenous engraftment of human CD34+HSCs. Two days before introducing HSC into the mice, bone marrow depletion was achieved by administering a single dose of busulfan at 25 mg/kg (Sigma-Aldrich, St. Louis, USA). Following busulfan administration, the mice were inoculated with 3 × 105 CD34+ HSCs through the lateral tail vein. Three weeks post-transplantation, 3 × 105 CD34-MNCs, stored to overcome insufficient human T-cell development, were administered via the tail vein. HSCs were engrafted, and peripheral blood samples were collected from the retro-orbital sinus of recipient mice for analysis using FACS. Following the protocols of Pearson et al. [15] and Meraz et al. [16], we confirmed the detection of hCD45+ cells as early as four weeks after the engraftment of HSCs. Achieving 25% engraftment of mature human white blood cells serves as a benchmark for successful humanization. PDX (passage No. 4 of PDXs, Supplementary Table 1 were transplanted into the mammary fat pads of humanized mice following the protocol described in a previous study [5].
Isolation of human immune cell lineages from the blood, spleen, and bone marrow of mice
Blood was collected in EDTA tubes (BD Biosciences, Franklin Lakes, USA) and centrifuged at 300 × g to remove the plasma. The bone marrow was harvested from the femur. The spleen tissue was chopped into 1–2 mm2 sections and filtered through a 70 μm cell strainer (Falcon, Cary, USA). Red blood cells in the blood, bone marrow, and spleen were lysed using Red Blood Cell Lysis Buffer (Roche, Basel, Switzerland).
Flow cytometry (antibodies)
For staining of single cells, the cells were incubated with the appropriate fluorescent dye-conjugated antibodies in phosphate-buffered saline (PBS) for 30 minutes at 4°C. Data were acquired using a flow cytometer (BD Biosciences) and analyzed using BD FACSDiva Software, Version 6.1.3. Eight color antibodies were used in this study (BD Biosciences): mCD45-BV421 (#103106), hCD45-APC/Fire750 (#368517), hCD4-FITC (#357405), hCD19-BV711 (#363021), hCD3-perCP/Cyanine5.5 (#317335), hCD8α-BV605 (#301039), hCD11b-PE/Cyanine 7 (#301321), and hCD56-BV421 (#318327).
RNA sequencing and analysis
Total RNA was isolated from the tumor tissues of the control (non-humanized) or humanized mouse PDX (X1, X2, and X3). The RNA samples were then subjected to library preparation using the Illumina TruSeq Stranded mRNA Library Preparation Kit (Illumina, San Diego, USA), and sequencing was performed on an Illumina NovaSeq 6000 platform. Sequence reads were aligned to the combined human and mouse reference genomes using a STAR aligner [17]. Alignment of reads and processing steps were performed using a previously described pipeline [6]. HTSeq was used to count the sequence reads mapped to each annotated gene [18]. Raw read counts were converted to fragments per kilobase of transcript per million mapped reads (FPKM) using the RPKM function in the edgeR package [19]. DESeq2 was used to differentially expressed genes (DEGs) between the two groups [20]. Significant DEGs were identified based on the following criteria: adjusted p-value < 0.05, log2 fold-change ≥ 1, and average expression (FPKM) ≥ 1 across all samples. To determine the significantly enriched pathways using the Kyoto Encyclopedia of Genes and Genomes [21], we used the Database for Annotation, Visualization, and Integrated Discovery (https://david.ncifcrf.gov/). Statistical significance was defined as a p-value < 0.05. Sequencing data were analyzed using a computing server at the Genomic Medicine Institute Research Service Center.
Immunohistochemistry (IHC)
Fresh tumor, liver, and lung tissues were collected and fixed in 4% formalin. After fixation, the tissues were embedded in paraffin and cut into 4-μm thick sections. The tissue sections were treated with xylene to remove paraffin and then rehydrated using a series of alcohol solutions. Antigen retrieval was performed using a solution (Vector Laboratories, Inc., Burlingame, USA) to expose the antigens. To block endogenous peroxidase activity, sections were treated with 3% hydrogen peroxide and subsequently blocked with normal goat serum (ImmunoBioScience Corp., Mukilteo, USA). The sections were then incubated overnight at 4°C with primary antibodies against mCD45 (R&D Systems, Minneapolis, USA) and hCD45 (LSbio, Washington, D.C., USA), diluted at 1:2,000. Following primary antibody incubation, sections were exposed to a secondary antirabbit/mouse antibody, followed by incubation with a peroxidase solution. Finally, the sections were developed using an HRP/DAB IHC polymer detection kit (Agilent, Santa Clara, USA) and counterstained with hematoxylin. Images were captured using ZEN Blue 3.3 microscopy software under an optical microscope (NICON, Tokyo, Japan). Quantification of positive signals in IHC images was performed using the ImageJ IHC Image Analysis Toolbox (National Institutes of Health, Bethesda, USA).
shRNA knockdown and stable cell lines
To generate human TNFRSF11B knockdown cell lines, the specific short hairpin TNFRSF11B sequence 5′-GCTCAGTTTGTGGCGAATAAA-3′ was inserted into a shRNA piggyBac Transposase expression-GFP vector (System Biosciences, California, USA). The transposase and cumate switch transposon PiggyBac vectors were co-transfected into Hs578T and MDAMB231 cells using Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, USA). Puromycin was used to select and establish TNFRSF11B knockdown stable cell lines.
Cell culture, migration and invasion assay
Breast cancer cell lines were obtained from American Type Culture Collection (Manassas, USA). The cells were maintained in Dulbecco’s modified Eagle’s medium (Welgene, Seoul, Korea) or Roswell Park Memorial Institute 1640, supplemented with 10% fetal bovine serum (Welgene) and 1% penicillin/streptomycin (Gibco, Waltham, USA). The cells were incubated at 37°C with 5% CO2. For migration assay, a total of 2 × 105 cells were seeded in a trans-well insert with an 8-μm pore size (Corning Incorporated, Corning, USA) using a serum-free medium. For the invasion assays, the inserts were precoated with 1 mg·mL−1 Matrigel before seeding the cells. The lower chambers were filled with medium supplemented with 10% fetal bovine serum. The cells were then incubated for 24 hours, fixed using 4% paraformaldehyde (Biosesang, Seoul, Korea), and stained with 0.1% crystal violet (Sigma-Aldrich). Quantitative analysis of the migrated and invaded cells was conducted using the ImageJ (Java 1.8.0_172) software (National Institutes of Health).
Immunocytochemistry
Cells were cultured on two-well chamber slides at a density of 0.5 × 105 cells per well. The cells were rinsed with PBS and fixed in 4% paraformaldehyde for 20 minutes at room temperature. After another PBS rinse, cells were permeabilized with 0.1% Triton X-100 in PBS for 5 minutes and rinsed again with PBS. The cells were then stained with TMEM176B-Alexa Fluor® 488 (Invitrogen, Waltham, USA) and TNFRSF11B-Alexa Fluor® 647 antibodies for 1 hour at room temperature, followed by staining with 4′,6-diamidino-2-phenylindole. After extensive washing with PBS and a brief rinse with distilled water, the cells were mounted on a glass slide using a mounting reagent (Vector Laboratories, Inc.). Images were captured using a fluorescence microscope (Leica, Wetzlar, Germany).
3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay
Cells were seeded onto flat-bottom 96-well culture plates at a density of 5 × 104 cells (MDAMB231) or 0.25 × 104 cells (Hs578T) per well. After incubation, the cells were treated with 0.5 mg·mL−1 thiazolyl blue tetrazolium bromide (Sigma-Aldrich) for 3 hours. The resulting formazan crystals were dissolved using dimethyl sulfoxide (Duchefa Biochemistry, Haarlem, The Netherlands). The absorbance of each sample was measured at 540 nm using a microplate reader (BioTek Instruments, Winooski, USA).
RNA isolation and quantitative real-time polymerase chain reaction (PCR)
The cells of interest were harvested and transferred to a TRIzol reagent (Favorgen, Pingtung, Taiwan). For reverse transcription of RNA, the Prime Script 1st strand cDNA Synthesis Kit (Takara Bio, Kusatsu, Japan) was used. Subsequently, the amplified cDNA was obtained using the Power SYBR® Green PCR Master Mix (Applied Biosystems, Waltham, USA). The primer sequences used were as follows: For TNFRSF11B: forward 5’-GGT CTC CTG CTA ACT CAG AAA GG-3’ and reverse 5’-CAG CAA ACC TGA AGA ATG CCT CC-3’. For GAPDH: forward 5’-GAGTCCACTGGCGTCTTC-3’ and reverse 5’-GGAGGCATTGCTGATGATC-3.’
Protein extraction and immunoblotting
The cells were lysed using RIPA buffer supplemented with a protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific). Proteins from cell lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes (MERCK, St. Louis, USA). The membranes were incubated with specific antibodies and the signal was amplified using chemiluminescence reagents (Thermo Fisher Scientific) and detected using an Amersham Imager 680 (GE Healthcare Life Sciences, Pittsburgh, USA). The primary antibodies used were as follows: anti-hTNFRSF11B (#ab73400; Abcam, Cambridge, UK), anti-β-actin (#sc-47778; Santa Cruz Biotechnology, Dallas, USA), and anti-α tubulin (#ab52866; Abcam).
Xenograft mouse model
MDA-MB-231-shCTL or shTNFRSF11B cells (1 × 106 cells) were inoculated into the mammary fat pads of immune-deficient BALB/c nude mice (Orient Bio, Seongnam, Korea). Tumor dimensions were measured using a caliper and tumor volume was calculated using the following formula: 0.523 × length × width2 (mm3). Mice were purchased from KOATECH (Pyeoungtaek, Korea). All experiments were approved by the Institutional Animal Care and Use Committee of the Seoul National University Hospital, and the animals were maintained in a facility accredited by AAALAC International (#001169) in accordance with the Guide for the Care and Use of Laboratory Animals 8th edition, National Research Council (2010) (IACUC No.20-0199-C1A0).
Statistical analysis
Statistical analysis was conducted using GraphPad Prism 8.2 software (GraphPad Software, Inc., San Diego, USA). The results are presented as means ± standard error. To determine statistical significance, a t-test was used to compare two populations, whereas analysis of variance was used to compare more than two populations, assuming a normal distribution. For non-normal distributions, the Mann-Whitney U test was used to compare two populations, and the Kruskal-Wallis H-test was used to compare more than two populations. A p-value of less than 0.05 was considered statistically significant, denoted as *p < 0.05, †p < 0.01, ‡p < 0.001, §p < 0.001.
RESULTS
Establishment of humanized PDX models of TNBC using CB HSCs
Our protocol for establishing humanized TNBC PDX models is shown in Figure 1A. After transplanting high-purity CD34+ HSCs (purity > 90%, Supplementary Figure 1) obtained from CB into NSG mice, CD45+ human immune cells appeared in most experimental mice within five weeks (Figure 1B). CD45+ human immune cells constitute more than one-third of the cells in the peripheral blood. After the 5th week, the proportion of CD45+ human immune cells was sustained or increased until the 10th week, when the number of human immune cells began to decline (Figure 1B). In some mice, after the transplantation of CD34+ cells, we observed severe hair loss (Figure 1C) or significant weight loss (Figure 1D), which are representative signs of graft-versus-host disease (GVHD) [22]. However, humanized mice without signs of GVHD did not show significant differences in weight (Figure 1D) and were used for subsequent tumor tissue transplantation.
Figure 1
HSC = hematopoietic stem cell; MNC = mononuclear cell; GVHD = graft-versus-host disease; NK = natural killer; FACS = fluorescence activated cell sorting; SD = standard deviation. *p < 0.05, §p < 0.001.
Establishment of humanized patient-derived xenograft models of triple-negative breast cancer by using cord blood hematopoietic stem cells.
(A) The schematic illustration shows the process and schedule of generating a humanized patient-derived xenograft model by recapitulating the human immune system in NOD/Shi-scid/IL-2Rγnull murine hosts through hematopoietic stem cell engraftment from cord blood. (B) After the transplantation of CD34+ HSCs, the hCD45+ is assessed using mice peripheral blood at the 4-week mark, followed by regular FACS analysis until the end of the study. Following the transplantation of CD34+ HSCs, some cases exhibit symptoms indicating GVHD, such as excessive hair loss, a hunched posture (C), and rapid weight loss, ultimately resulting in mortality (D). Data are presented as mean and error ± SD; two-way analysis of variance. (E) The percentage of differentiated human immune lineages from hCD34+ cells is analyzed via FACS in the blood, bone marrow, and spleen at the time of harvest in the non-humanized, humanized CD34+, and humanized CD34+MNC+ groups. Data show median values ± SD; Kruskal-Wallis test. CD34, a cell surface marker for HSCs and progenitor cells; CD45 a cell surface marker for white blood cells or leukocytes.
In mice that received CD34+ HSCs exclusively for human immune reconstitution, B cells constituted the major cell type in the peripheral blood, bone marrow, and spleen (Figure 1E, Supplementary Figure 2). However, in mice transplanted with CD34- MNC alongside CD34+ HSCs, we observed an increased appearance of T cells, especially CD4+ T cells, in the peripheral blood and tissues (Figure 1E). Additionally, our data demonstrated that the transplanted human immune cells could differentiate into CD4+ or CD8+ T cells, natural killer cells, and myeloid cells.
In vivo tumor growth in the humanized PDX models of TNBC
We orthotopically implanted three TNBC PDX tumors (X1–3) in humanized and non-humanized NSG mice and compared their in vivo tumor growth. After transplantation of HSCs, as shown in Figure 2A, peripheral blood was collected from the humanized mouse model, and hCD45+ cells were analyzed by FACS. When the percentage exceeded 25%, PDXs were transplanted (Figure 2B, Supplementary Figure 3). The impact of humanization on tumor growth rate was different for each TNBC tumor (Figure 2C). X1 PDX tumors showed significantly faster tumor growth in humanized models than in non-humanized models, whereas the other two TNBC tumors showed similar growth rates. We determined the presence of human immune cells in humanized TNBC mouse models. As shown in Figure 2D, hCD45+ immune cells were present in tumor tissues as well as in other organs, such as the lungs and liver. Murine immune cells (mCD45+) were also detected in the humanized mouse models (Supplementary Figure 4).
Figure 2
HSC = hematopoietic stem cell; PDX = patient-derived xenograft; TNBC = triple-negative breast cancer; SD = standard deviation; IHC = immunohistochemistry; ns = not significant. ‡p < 0.001, §p < 0.001.
The impact of human leukocyte infiltration on humanized tumors and organs.
Overview of diverse growth patterns of TNBC PDX in non-humanized and humanized in vivo models. (A) Diagram representing the process by which PDX was successfully transplanted into a humanized model. (B) Percentage of human CD45+ cells at 4 weeks post-HSC transplantation in peripheral blood. Data represents means ± SD; Mann-Whitney tests. (C) Visual representation of TNBC PDX after surgical removal (X1; top, X2158; middle and X3; bottom, respectively). Mean tumor volume is depicted by tumor growth curves with error bars indicating SD; two-way analysis of variance for X1, X2, and X3 tumors in both groups. (D) Microscopic IHC images of mouse and human CD45+-stained tumor, liver, and lung sections demonstrating the infiltration of mouse and human leukocytes. Scale bar = 50 µm. Quantification of human CD45+ IHC in tumors, livers, and lungs. Data represent means ± SD; Mann-Whitney test within each group.
Gene expression characteristics of TNBC tumors of the humanized PDX models
We performed RNA sequencing of fat pad TNBC tissues from humanized and non-humanized mouse models to investigate gene expression differences, as shown in Supplementary Table 2. The analysis of DEGs revealed notable and distinct differences in the transcriptomic profiles between the groups in two of the three TNBC models (Figure 3A). Importantly, in X1 and X2 humanized TNBC tumors, we observed the upregulation of multiple immune response-related signaling pathways, including antigen processing and presentation, as well as autoimmune diseases, such as allograft rejection, GVHD, systemic lupus erythematosus, and autoimmune thyroid disease (Figure 3B). Furthermore, we observed the upregulation of antigen processing and positive regulation of T-cell activation in humanized tumors during the analysis of biological processes (Supplementary Table 3).
Figure 3
KEGG = Kyoto Encyclopedia of Genes and Genomes; PDX = patient-derived xenograft; FC = fold change; TNFRSF11B = tumor necrosis factor receptor superfamily member 11B or osteoprotegerin.
Gene expression characteristics of the triple-negative breast cancer patient-derived xenografts and cells under human immune system reconstituted environment.
(A) Principal component analysis plot for X1, X2, and X3 PDX tumors of both non-humanized and humanized mice. (B) Top 15 KEGG pathways for upregulated genes in the humanized samples X1 and X2 with log2 fold change > 1, and adjusted p-value < 0.05 were considered significant (differentially expressed genes; n = 1,498; X1 and n = 204; X2, respectively). The pathways represented by the 10 overlapping terms indicate shared enrichment in both X1 and X2 humanized PDXs that are upregulated. (C) In the volcano plot, the FC represents the average expression level of each gene (fragments per kilobase of transcript per million mapped reads > 1, log2 fold change > 1 and adjusted p-value < 0.05). Each dot on the graph corresponds to a single gene, with green dots indicating downregulated genes and purple dots representing upregulated genes in humanized PDXs. Red spot shows that TNFRSF11B is identified as commonly downregulated genes between X1 and X2 humanized models. (D) TNFRSF11B expression levels analysis in protein expression and mRNA level in 10 breast cancer cell lines (n = 3).
Among the DEGs in these humanized TNBC PDX models, TNFRSF11B was significantly downregulated in both the X1 and X2 humanized models (Figure 3C). When tested in multiple breast cancer cell lines, the TNFRSF11B protein and mRNA were highly expressed in several TNBC cell lines, including MDA-MB-231 and Hs578T cells (Figure 3D).
TNFRSF11B silencing inhibits breast cancer growth and metastasis
To address the functional importance of TNFRSF11B, a gene associated with the humanized microenvironment, we generated TNFRSF11B knockdown breast cancer cells by shRNA transfection of MDA-MB-231 and Hs578T cells. TNFRSF11B was successfully downregulated in shTNFRSF11B transfected breast cancer cells (Figure 4A-C). TNFRSF11B silencing reduced the proliferation, migration, and invasion of both MDA-MB-231 and Hs578T cells in vitro (Figure 4D and E).
Figure 4
shCTL = control; shTNFRSF11B = TNFRSF11B-knockdown; TNFRSF11B = tumor necrosis factor receptor superfamily member 11 B, or osteoprotegerin; DAPI = 4′,6-diamidino-2-phenylindole; SD = standard deviation. *p < 0.05, †p < 0.01, ‡p < 0.001, §p < 0.001.
Tumor necrosis factor receptor superfamily member 11B associated with human immune system-reconstitution affected diverse phenotypes of breast cancer cells.
(A) Fluorescent microscopy images show control and knockdown of TNFRSF11B in MDAMB231 cells, with the nucleus stained in blue (DAPI) and TNFRSF11B labeled in red. Scale bar = 50 μm. Level of TNFRSF11B mRNA (B) and protein (C) in Hs578T and MDAMb231 (mean ± SD, n = 5 by Mann-Whitney test). (D) 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay of shCTL and shTNFRSF11B breast cancer cells, Hs578T and MDAMB231 (mean ± SD, n = 3; analysis of variance). (E) Migration and invasion assay of shCTL and shTNFRSF11B transfected Hs578T and MDAMB231 using a trans-well insert. Scale bar = 100 μm. Data are presented as mean ± SD, n = 4; Mann-Whitney test.
Subsequently, we investigated the in vivo tumorigenicity and metastatic capacity associated with TNFRSF11B by orthotopically injecting MDA-MB-231 cells. TNFRSF11B-silenced MDA-MB-231 cells showed a modest but significant reduction in primary tumor growth regarding both tumor weight and volume (Figure 5A). Interestingly, TNFRSF11B silencing significantly reduced lung metastasis in a spontaneous metastasis model (Figure 5B). The metastasis-suppressing effect of TNFRSF11B was also observed in an experimental lung metastasis model of tail vein injection of tumor cells (Figure 5C). These data indicated that TNFRSF11B, which is regulated by the presence of a humanized microenvironment, can regulate tumor growth and metastasis in breast cancer.
Figure 5
shCTL = control; shTNFRSF11B = TNFRSF11B-knockdown; TNFRSF11B = tumor necrosis factor receptor superfamily member 11B or osteoprotegerin; SD = standard deviation. *p < 0.05, †p < 0.01.
Tumor necrosis factor receptor superfamily member 11B associated with human immune system-reconstitution affected diverse phenotypes of breast cancer cells.
(A) Image, weight and growth curve of tumor volume derived from MDAMB231 cells with or without shTNFRSF11B xenograft in athymic nude mice (mean ± SD, n = 5; Mann-Whitney and two-way analysis of variance, respectively). Microscopic lung metastasis images and quantifications of cytokeratin19 immunohistochemistry staining of lung tissue sections (B) spontaneous and (C) experimental lung metastasis from MDAMB231-shCTL or shTNFRSF11B (mean ± SD, scale bar = 100 μm, n = 5 by Mann-Whitney).
DISCUSSION
In the present study, the successful establishment of humanized PDX models of TNBC using CB HSCs represents the possibility of wide use of this preclinical research model for studying tumor-immune interactions in a more physiologically relevant context. The engraftment of high-purity CD34+ HSCs and the subsequent appearance of CD45+ human immune cells in most experimental mice underscores the capability of humanized models to recapitulate the human immune system in an immunocompromised murine host, at least partially. However, despite the successful engraftment of human immune cells, humanized PDX models have significant limitations. Partial reconstitution of the human immune system in murine hosts failed to replicate the complexity and diversity observed in the human body. The immunodeficient nature of the NSG mice restricts the full range of immune cell interactions and functional responses that would be present in an intact human immune system [23, 24, 25, 26, 27, 28]. Consequently, while humanized models provide valuable insights, they should be interpreted with caution, and additional complementary approaches, such as humanized mouse models with more comprehensive immune reconstitution or patient-derived immune cells, should be considered to complement the findings. To increase the duration of humanization more effectively, other research teams considered two approaches: the simultaneous engraftment of HSCs and human tissues in neonatal mice for the humanized model and the use of radiation therapy instead of busulfan treatment for bone marrow depletion [21]. Additionally, the use of a limited number of TNBC PDX tumors (X1–3) in our study may not have fully captured the extensive heterogeneity that characterizes TNBC. The inherent variability in tumor characteristics and immune responses across different patient tumors may influence the outcomes observed in humanized PDX models. Expanding the repertoire of PDX tumors and including more representative samples of diverse TNBC subtypes may enhance the generalizability and robustness of our findings, thereby enabling a more comprehensive understanding of the interplay between the human immune system and TNBC.
The observed variability among TNBC PDX tumors in humanized models underscores the complexity of tumor-immune interactions and highlights the tumor-specific nature of the immune response [16, 29]. The microenvironmental context, including tumor antigenicity, immunogenicity, and immune cell infiltration, could play a crucial role in determining the extent of humanization’s impact on tumor growth [30, 31]. Further investigation of the factors governing the variable responses to humanization could provide valuable insights into the determinants of immunogenicity in TNBC and inform the development of personalized immunotherapeutic strategies. Previous studies have reported that the human immune system facilitates the enhancement of hepatocellular carcinoma proliferation and angiogenesis [32]. Additionally, there was no significant difference in ER+ breast PDX growth between humanized and non-humanized control mice [14]. Consequently, variations in tumor growth between humanized and non-humanized environments in PDX models are shaped by inherent traits and genetic diversity.
RNA sequencing analysis of TNBC tumor tissues from humanized and non-humanized mouse models revealed several immune response-related signaling pathways and autoimmune disease-associated pathways that were upregulated in X1 and X2 TNBC tumors. Downregulation of TNFRSF11B in both the X1 and X2 humanized models suggests a potential role of this gene in shaping the tumor-immune microenvironment. TNFRSF11B, also known as osteoprotegerin, is involved in various biological processes including cell migration, immune regulation, and bone homeostasis [33, 34, 35]. Its differential expression in humanized models indicates a plausible link between the presence of human immune cells and immunomodulatory characteristics of the tumor microenvironment. Elucidating the mechanisms by which TNFRSF11B influences tumor growth and immune responses could provide novel targets for immunotherapeutic interventions. TNFRSF11B is correlated with distant organ metastasis in breast cancer. Furthermore, a decrease in TNFRSF11B expression was linked to shorter median overall survival and signified an unfavorable prognosis [36]. Another study presented a groundbreaking discovery, highlighting the role of osteoprotegerin as a critical paracrine factor in the transition of normal cells to tumor cells, while also providing novel insights into the mechanisms through which osteoprotegerin influences proliferation, cell cycle, and aneuploidy in normal mammary epithelial cells [37]. Phenotypic changes involving TNFRSF11B silencing in MDA-MB-231 and Hs578T cells provided valuable evidence of its involvement in regulating key cellular processes, such as proliferation, migration, and invasion. In vivo experiments further demonstrated its significant effect on primary tumor growth and lung metastasis, supporting its potential as a promising therapeutic target for the treatment of TNBC. Targeting TNFRSF11B or its associated immune response pathways could potentially modulate the immune landscape of the tumor microenvironment, thereby enhancing the antitumor immune response and improving therapeutic outcomes.
In conclusion, the establishment of humanized PDX models represents a crucial step towards elucidating the intricate interplay between the human immune system and TNBC tumors. Despite the inherent limitations of achieving a fully representative human immune system in murine hosts, these models offer valuable insights into the complexities of the humanized microenvironment in TNBC. By identifying TNFRSF11B as a critical gene associated with the humanized microenvironment and its functional implications in tumor growth and metastasis, our findings provide a foundation for developing innovative immunotherapeutic strategies for TNBC.
SUPPLEMENTARY MATERIALS
Information about the patient-derived xenografts and its donorSupplementary Table 1
The list of differential expression genes for X1 and X2 patient-derived xenograftsSupplementary Table 2
Biological process (top 15-humanized mice upregulated)Supplementary Table 3
The purification of CD34+ cells from whole mononuclear cells for fluorescence activated cell sorting analysis. Representative flow cytometry plots of CD34 staining mononuclear cells isolated from cord blood.Supplementary Figure 1
Exploring the development and function of human immune lineage in humanized mice.Supplementary Figure 2
Fluorescence activated cell sorting analysis for hCD45+ post-injection of CD34+ hematopoietic stem cells at 4 weeks for patient derived xenografts engraftment.Supplementary Figure 3
Mouse leukocyte infiltrates in humanized mice for immune system.Supplementary Figure 4
Funding:This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (HI22C0497), by a grant supported by the National Research Foundation of Korea (NRF-2019R1A2C2005277, 2020R1A6A1A03047972), and a research grant from Seoul National University Hospital (grant No. 0320210280). Also, this research was partly supported by a research donation to Seoul National University Hospital by Mr. Sung Koo Lee.
Conflict of Interest:The authors declare that they have no competing interests.
Data Availability:In accordance with the ICMJE data sharing policy, the authors have agreed to make the data available upon request.
Author Contributions:
Conceptualization: Kim JI, Moon HG.
Data curation: Her Y, Yun J.
Formal analysis: Her Y, Yun J.
Funding acquisition: Moon HG.
Investigation: Her Y.
Methodology: Her Y, Son HY.
Project administration: Moon HG.
Resources: Heo W.
Supervision: Moon HG.
Validation: Yun J.
Visualization: Her Y.
Writing - original draft: Her Y.
Writing - review & editing: Moon HG.
References
-
Pearson T, Greiner DL, Shultz LD. Creation of "humanized" mice to study human immunity. Curr Protoc Immunol 2008;Chapter 15:15.21.1–21.
-
-
Brehm M, Bortell R, Verma M, Shultz L, Greiner D. Humanized mice in translational immunology. In: Tan SL, editor. Translational Immunology: Mechanisms and Pharmacological Approaches. Amsterdam: Elsevier; 2016. pp. 285-326.
-
MeSH Terms
Figures
1 / 5
Funding Information
-
Korea Health Industry Development Institute
HI22C0497
-
National Research Foundation of Korea
NRF-2019R1A2C20052772020R1A6A1A03047972
-
Seoul National University Hospital
0320210280
