MiR-210 Is Overexpressed in Tumor-infiltrating Plasma Cells in Triple-negative Breast Cancer

Abstract

Triple-negative breast cancer (TNBC) is a heterogeneous group of breast cancer and is characterized by aggressiveness and poor prognosis. MicroRNA represents a new class of biomarkers, and accumulating evidence indicates that microRNAs contribute to tumorigenesis and cancer metastasis. It has been described that miR-210 is highly expressed in TNBC, and its overexpression had been linked to poor prognosis. In a previous work, we showed that in TNBC miR-210 is expressed in tumor cells and also in the tumor microenvironment (TME), particularly in inflammatory CD45-LCA positive cells. However, the exact identity of these cells remained unknown. In this study, we performed in situ hybridization and immunohistochemistry using validated antibodies for the different specific immune cell markers on adjacent sections of 23 TNBC infiltrated with immune cells. We found that miR-210 expressing cells in the TME were stained positive with CD79a, a B-cell lineage marker. These tumor-infiltrating cells were negative for CD20 and Ki-67 but positive for MUM1 and CD38 and also expressed immunoglobulins, indicating that they are immunoglobulin-producing plasma cells (PCs). To the best of our knowledge, this is the first study demonstrating miR-210 expression in tumor-infiltrating PCs.

Introduction

Triple-negative breast cancer (TNBC) represents about 15% to 20% of breast cancer (BC) cases and is characterized by the lack of estrogen receptor, progesterone receptor, and HER2 gene amplification.¹ TNBC is considered to be an aggressive type of BC with high-grade, frequent distant metastasis and the poorest outcome.² TNBC represents a major clinical challenge due to heterogeneity and a lack of specific markers, well-defined molecular targets, and efficient therapy.³

Cancer immunity has emerged as a clinical hallmark of cancer, and the immune system plays a multifaceted role in cancer biology by either promoting or inhibiting tumor growth and thus influences the oncological outcome.⁴ In TNBC, the presence of high level of tumor-infiltrating T-cells has been associated with good prognosis, improved disease-free survival and overall survival, and improved responses to neoadjuvant chemotherapy.⁵ Similarly, dense infiltration of CD20⁺ B-cells has been associated with improved outcome and prognosis in some cancer.^6,7

MicroRNAs (miRNAs) are short single-stranded, non-coding RNAs that function at the post-transcriptional level to control and fine-tune gene expression.⁸ Increasing evidence indicates that the miRNA profile varies according to cancer type and can be a useful tool in diagnosis and prognosis prediction.⁹ The role of miRNA is not limited to cancer cells but also implicated in the activation of immune cells and the modulation of immune response.¹⁰

MiR-210 has been largely studied in the past years and has been identified as a major miRNA induced under hypoxia. MiR-210 expression is frequently upregulated in several types of cancer. In TNBC, miR-210 was shown to be upregulated and its overexpression is correlated with aggressiveness and poor prognosis.¹¹

In a previous work, we showed that in TNBC miR-210 is expressed in tumor cells and in the tumor microenvironment (TME), particularly in inflammatory CD45-LCA positive cells.¹² However, the exact nature and the identity of these cells remained unknown. In this study, we have performed in situ hybridization (ISH) analysis for miR-210 and dual immunohistochemistry (IHC) staining for the different immune cell markers on adjacent sections of TNBC. We show that miR-210 expressing cells in the TME express CD79a, a B-cell lineage marker. These cells also express MUM1, CD38, and immunoglobulins (Igs), indicating that they are Ig-producing plasma cells (PCs). The high expression of miR-210 is not observed in PCs present in non-tumoral tissues. Taken together, these results highlight for the first time the overexpression of the miR-210 in tumor-infiltrating PCs.

Materials and Methods

Study Population

Formalin-fixed paraffin-embedded tissues were selected from the IPG Biobank. We analyzed 23 TNBC, two tonsils, and one nasal polyp. All analyzed TNBC lacked estrogen receptor or progesterone receptor expression (Allred score of 0/8) and were negative for HER2 amplification (confirmed by fluorescent ISH). A basal phenotype was defined by being positive for any of the basal markers (CK5/6, P63, and EGFR). CK5/6, P63, and epidermal growth factor receptor (EGFR) were considered positive if 1% or more of tumor cells exhibited immunoreactivity (weak or strong) as described previously.¹³ All the experiments involving human tissues were conducted with the permission of the ethics committee (G2-2019-E001) of Grand Hôpital de Charleroi. Patients’ characteristics are summarized in Table 1.

Table 1.

Characteristics of TNBC Patients.

Age at diagnosis	Mean	49
	Range	28–69
Tumor size (cm)	Mean	2.8
	Range	1–9
Lymph node invasion n (%)	Positive	13 (56.5)
	Negative	6 (26.0)
	Unknown	4 (17.5)
KI67 (%)	Mean	42
	Range	2–90
Basal markers n (%)	CK5/6 positive	12 (52.2)
	P63 positive	11 (47.8)
	EGFR positive	13 (56.2)
	Basal	17 (73.9)
	Non-basal	6 (27.1)
Histological grade n (%)	3	19 (82.6)
	2	4 (17.4)
	1	0

Abbreviations: TNBC, triple-negative breast cancer; EGFR, epidermal growth factor receptor.

Immunohistochemistry

IHC was performed on 4-µm paraffin sections. Heat-induced antigen retrieval was performed using PT-link (DAKO, Agilent Technologies; Diegem, Belgium). Detection of CD45-LCA (M0701), IgG (IR512), and IgM (IR513) was performed on DAKO autostainer automate. The antibodies were from DAKO (Agilent Technologies).

Dual staining was performed using Enzo Multiview IHC kit (ENZ 181-0150; ENZO Life Sciences, Brussels, Belgium) according to the manufacturer’s instructions. The IHC signal was developed using the HIGHDEF AP Red chromogen/substrate and then DAB chromogen/substrate.

The polyclonal rabbit anti-human CD3 (IR503), monoclonal mouse anti-human CD20 (IR604), anti-human CD68 (IR613), monoclonal mouse anti-human CD56 (IR628), monoclonal mouse anti-human Ki-67 (IR626), monoclonal mouse anti-human CK5/6 (IR780), and monoclonal mouse anti-human p63 (GA662) were from DAKO (Agilent Technologies). Polyclonal rabbit anti-human CD38 (HPA022132) was from Sigma-Aldrich (Overijse, Belgium) and used at 1:1000 dilution. Monoclonal rabbit anti-human MUM1 (MRQ-43) and monoclonal rabbit anti-EGFR (3C6) (790-2988) were from Roche Applied Science (Mannheim, Germany).

MiRNA Locked Nucleic Acid (LNA) In Situ Hybridization

The hsa-miR-210-3p (YD00612239), scramble-miR (YD00699004), hsa-miR-21-5p (YD00 619870), and hsa-U6 probe (YD00699002) as well as the miRNA ISH buffer with proteinase K (339450) were purchased from Qiagen (Benelux). ISH was performed according to the recommended “One-day miRNA ISH protocol.” Probes were diluted with 1× ISH buffer (50 nM for miR-210, miR-21, and scramble probes and 1 nM for U6 probe) and incubated at 55C for 60 min. The miRNA LNA-ISH probe was detected with anti-DIG-alkaline phosphatase antibody (1:800, cat. no 11 093 274 910; Roche Applied Science). The miRNA expression was visualized by 4-nitro-blue-tetrazolium and 5-bromo-4-chloro-3-indolynitrolphosphate substrate (cat. no 11 697 471 001; Roche Applied Science). The slides were counterstained with nuclear fast red (H-3403; Vector laboratories, Burlingame, CA).

Results

MiR-210 Is Expressed in Tumor-infiltrating B-lymphocyte Lineage Cells

We have previously shown that miR-210 is detected in the TME of TNBC. These miR-210-positive cells colocalized with regions positive for the immune cell marker, CD45-LCA.¹² To identify the nature of these miR-210 expressing immune cells, we performed ISH analysis for miR-210 (ISH validation illustrated in Supplemental Fig. S1) and IHC analysis using validated antibodies for the different specific immune cell markers (CD3/CD68 and CD79a/CD11c double staining, CD56 single staining) on adjacent sections of 23 TNBC (basal and non-basal TNBC) samples presenting an miR-210 signal in the TME (Fig. 1A and Supplemental Fig. S2). Of note, the proportion of miR-210-positive cells in the TME of the 23 samples varies from few positive cells to a majority of tumor-infiltrating cells. Keeping with previous observations,¹² we found that for all the 23 samples, the miR-210 signal was detected in cancer cells as well as in CD45-LCA positive cells of TME (Fig. 1B). We found that the miR-210-positive cells in the TME did not express CD3 or CD68 (Fig. 1B and C), suggesting that they are not T-cells or macrophages. These cells did not express CD11c or CD56, but they expressed the CD79a marker (Fig. 1B and C), suggesting that the miR-210-positive cells in TME are B-cell lineage.

Figure 1.

MiR-210-positive cells in the tumor microenvironment colocalize with B-lymphocyte lineage cell marker: (A) Schematic representation of serial sections used for in situ hybridization (ISH) and immunohistochemistry. (B) and (C) Representative images of immunostaining (single or double) in triple-negative breast cancer. MiR-210 was detected by ISH. The miR-210-positive cells exhibited a strong expression of CD45-LCA (hematopoietic lineage marker) and CD79a (B-cell marker), whereas they were negative for CD3 (T-cell marker), CD56 (natural killer cell marker), CD11c (dendritic cell marker), and CD68 (macrophage marker). Original magnification: (B) 5× (scale bar: 500 µm); (C) 40× (scale bar: 50 µm).

MiR-210 Is Expressed in Tumor-infiltrating MUM1⁺CD38⁺Ig⁺CD20⁻Ki-67⁻ Plasma Cells

Activation of B-cells leads to their differentiation into memory B-cells or plasmablasts (PBs) that can be distinguished via the expression of several markers.¹⁴ To identify the B-cell subsets that express miR-210, tumor sections from eight samples highly infiltrated with miR-210-positive immune cells were stained with antibodies to MUM1 and CD20 to distinguish PBs/PCs (which are CD20 negative but express MUM1) from naive and memory B-cells (which are MUM1 negative but express CD20) (Fig. 2A). We found that miR-210 signal in non-tumor cells did not colocalize with CD20 (Fig. 2B and C), suggesting that miR-210 expressing B-cells are not naive or memory cells. However, these regions colocalized with the PB/PC marker, MUM1 (Fig. 2B and C). To further confirm these results, we investigated the expression of another PB/PC marker, CD38. Similarly to MUM1, we found that the miR-210 regions were CD38 positive (Fig. 2B and C).

Figure 2.

MiR-210-positive cells are MUM1⁺CD38⁺CD20⁻Ki67⁻ plasma cells (PCs): (A) Schematic representation of serial sections used for in situ hybridization (ISH) and immunohistochemistry. (B) and (C) Representative images of immunostaining in triple-negative breast cancer tumors. MiR-210 was detected by ISH. The miR-210-positive cells (8/8 samples) expressed MUM1 and CD38, whereas they were negative for CD20 and Ki67, indicating that they are PCs (8/8 samples; scale bar 5×: 500 µm, scale bar 40×: 50 µm). (D) Representative images of H&E staining. MiR-210 expressing cells have a dense nucleus with heterochromatin in a “cartwheel” pattern, abundant cytoplasm, and enlarged Golgi zone (scale bar 10×: 250 µm, scale bar 40×: 25 µm).

Mature PCs are quiescent and non-cycling and differentiated from transitional highly proliferating PBs.¹⁴ To determine whether the miR-210 expressing cells are PCs or PBs, we assessed the expression of Ki-67, a nuclear protein associated with cell proliferation.¹⁵ Interestingly, we found that the miR-210-positive cells were Ki-67 negative and thus stopped in cell cycle, indicating that they correspond to PCs and not PBs (Fig. 2B and C). PCs can also be distinguished by their morphological appearance as they have a dense nucleus with heterochromatin in a “cartwheel” pattern, abundant cytoplasm, and enlarged Golgi zone.¹⁶ H&E staining and morphological analysis of miR-210 expressing cells confirmed that miR-210 expressing cells in the TME are PCs (Fig. 2D).

As mature PCs are committed to the production and secretion of antibodies,¹⁴ we analyzed the expression of Ig isotypes. Eight TNBC tumors with the miR-210-positive PCs were stained with IgG and IgM antibodies. The majority of miR-210 expressing PCs were IgG positive, whereas a small fraction was IgM positive (Fig. 3), indicating that miR-210 expressing cells correspond to antibody-producing PCs.

Figure 3.

MiR-210-positive cells express immunoglobulins: (A) Schematic representation of serial sections used for in situ hybridization (ISH) and immunohistochemistry. (B) Serial sections of eight tumors were stained with antibodies against IgG or IgM. MiR-210 was detected by ISH. The miR-210 expressing plasma cells exhibited a strong expression of IgG compared with IgM (8/8 samples). Original magnification: 5× (scale bar: 500 µm); 40× (scale bar: 50 µm).

To explore whether miR-210 expression is a marker of B-cell differentiation into PCs or is specific to tumor-infiltrating PCs, we investigated whether miR-210 is expressed in PCs in non-tumoral tissues. We performed ISH analysis on different human samples known to contain many PCs. Interestingly, we found that PCs in tonsils and nasal polyps (MUM1 and IgG positive and CD20 negative) expressed very weak level of miR-210 compared with tumor-infiltrating PCs (Fig. 4).

Figure 4.

Plasma cells in non-tumoral tissues express a very low level of miR-210: (A) Schematic representation of serial sections used for in situ hybridization (ISH) and immunohistochemistry. (B) Representative images of miR-210 detection by ISH in human tonsil and nasal polyp. Serial sections of human tonsils (two samples) and nasal polyps were stained with antibodies against MUM1/CD20 (double staining) and IgG. The MUM1-positive cells expressed a very low level of miR-210 but expressed the positive control probe U6 in the nucleus. These cells exhibited a strong expression of IgG (scale bar: 50 µm).

Collectively, these results indicate that in addition to tumor cells, miR-210 is highly expressed in tumor-infiltrating antibody-producing PCs.

Discussion

This study identifies the nature of the miR-210 expressing cells in the TME in TNBC infiltrated with immune cells. We show that miR-210 expressing cells in the TME express CD79a, a B-cell lineage marker. These cells also express MUM1, CD38, and Igs, indicating that they are antibody-producing PCs. These results highlight for the first time the expression of miR-210 in tumor-infiltrating PCs.

Many studies have demonstrated the role of miR-210 in cancer progression, and it was shown that miR-210 is overexpressed in numerous cancers and miR-210 overexpression has been linked to poor prognosis.^11,17 MiR-210 has numerous functions in tumor epithelial cells and also in the TME, which plays an important role in tumor progression.¹⁷ Considering that miR-210 has numerous functions in the TME, a high expression of miR-210 in the tumor-infiltrating PCs might also have an impact on the function of PCs. Previous reports suggested that miR-210 affects the immune system. MiR-210 is involved in Th17 differentiation.¹⁸ In regulatory T-cells, miR-210 negatively regulates FOXP3.¹⁹ Most importantly, it was reported that miR-210 limits autoantibody production in B-cells.²⁰ In their study, the authors used models of knockout and transgenic mice and found that mice deficient in miR-210 lead to the development of autoantibodies, whereas mice overexpressing miR-210 result in impaired IgG antibody secretion.²⁰ This is in contrast with our results where the miR-210-positive tumor-infiltrating PCs express a high level of IgG. It should be kept in mind that the conclusions of Mok et al.²⁰ are based on antibody assessment in mice serum. Of note, the secretion of high amount of antibodies requires a high secretory capacity of PCs. However, PC differentiation can proceed normally in the absence of high Ig secretion.²¹ It is thus possible that the phenotypic alterations observed in miR-210-deficient mice are the result of inhibition of antibody secretion. Thus, in vivo, miR-210 overexpressing PCs might be able to differentiate and mature, but still failing in antibody secretion. Such view is consistent with other reports on PC differentiation and maturation regulation.²¹

Tumors have developed strategies to evade the immune surveillance.²² During tumor immune escape, cancer cells can alter the expression of miRNAs to suppress the immune system. For instance, the expression pattern of miRNAs in T-cells can be modulated by tumor cells, thereby counteracting tumor-directed T-cell responses.¹⁰ So far, nothing is known about the regulation of miR-210 expression in PCs. Our study indicates that miR-210 expression is very low in PCs present in non-tumoral tissues compared with tumor-infiltrating PCs. It is thus possible that tumor cells modulate miR-210 expression in PCs to regulate their function. MiR-210 is considered as the major miRNA induced under hypoxic microenvironment.¹¹ As hypoxia is an important feature of solid tumors, miR-210 expression in PCs might be due to hypoxic conditions where it is likely to be regulated by a mechanism independent of HIF-1-alpha.¹² Another possibility would be that increased miR-210 expression might be due to high activation of PCs in tumor stroma that would allow high production of antibodies. Further investigations are needed to elucidate the mechanism(s) regulating miR-210 expression in PCs in the TME.

Studies developed over the past decade showed that the presence of high levels of lymphocytic infiltration has been consistently associated with good prognosis and overall survival.⁴ However, few studies evaluated the prognostic value of PCs, and the predictive efficacy of PC infiltration remains largely controversial.^6,7,23 Further studies taking into account the miR-210 expression in PCs might allow a better correlation between PC infiltration and prognosis.

Although TNBCs exhibit more abundant lymphoid cell infiltrates compared with other BC cell types, it is worth noting that not all TNBCs are highly infiltrated with immune cells.²⁴ The tumors used in our study correspond to a subgroup of TNBCs that are highly infiltrated with immune cells. We believe that miRNA expression in PCs is not limited to TNBC but may also be expressed not only in other subtypes of BC but also in other types of tumors with high PC infiltration.^6,25 The expression of miR-210 in tumor-infiltrating PCs warrants further investigation.

In conclusion, we describe for the first time the overexpression of miR-210 in tumor-infiltrating PCs. These results will pave the way for a new research on PC regulation and function in the TME. Elucidating the precise mechanism of PC modulation by miR-210 is critical for an improved understanding of the potential effect of miR-210 on PC function, thus providing an opportunity for their use as prognostic biomarkers or as therapeutic targets.