Transcriptional Factor KLF11 is a new prognostic biomarker and potential therapeutic target in human sporadic invasive breast cancer

doi:10.21203/rs.3.rs-1605154/v1

Download PDF

Research Article

Transcriptional Factor KLF11 is a new prognostic biomarker and potential therapeutic target in human sporadic invasive breast cancer

https://doi.org/10.21203/rs.3.rs-1605154/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Breast cancer (BC) is a heterogeneous disease characterized by distinct molecular subtypes. Krüppel like Factor (KLF) family members in BC have been widely studied. However, the roles of KLFs as tumor-promoters or tumor-suppressors are context-dependent. KLF11, as a member of this family, the expression pattern and prognostic relevance of which in BC remains unclear.

Methods

To clarify whether KLF11 acts as a tumor-promoter or a tumor-suppressor in BC, we explored the prognostic role of the KLF11 in BC patients by performing immunohistochemistry (IHC) staining of KLF11 in 298 patients samples. Then we correlated the expression to clinicopathological characteristics and survival outcomes. To investigate the function of KLF11 in tumor cells, we performed different assays of cell viability, cell proliferation, and cell apoptosis assays in three distinct molecular subtypes of BC cell lines with siRNA-mediated loss-of-function of KLF11.

Results

In the BC patient cohort, We demonstrated that KLF11 was higher in highly proliferative BC and lowest in luminal A-like BC. Furthermore, "KLF11-high" BC patients had shorter disease-free survival (DFS), distant-metastasis-free survival (DMFS), and local recurrence-free survival (LRFS). High KLF11 expression remained an independent indicator for DFS and DMFS of BC. The multivariate cox regression-dependent, KLF11-related nomograms of DFS and DMFS showed high accuracy in predicting BC patients' 3-,5- and 10 -year survival probability. Our results indicate that KLF11 might play a pro-tumor role in BC.

From the KLF11-BC tumor cells perspective, the cellular functional assays confirmed that KLF11 acts as a pro-tumor transcription factor (TF) in BC cells via promoting proliferation and/or decreasing cell apoptosis. However, due to neoplastic intratumor heterogeneity, different cancer molecular subtypes could have different underlying mechanisms.

Conclusions

Prior research has demonstrated that KLF11 expression is lower in BC compared to healthy breast tissue. However, we now could show that high KLF11 expression in BC is probably pro-tumorigenic and is associated with an impaired prognosis. Our study indicates that KLF11, as a transcriptional factor, acts as a new prognostic biomarker and might also be potentially a synergistic or an alternative therapeutic target for conventional targeting treatments of human BC.

KLF11

transcription factor

breast cancer

prognostic factor

proliferation

apoptosis

Breast cancer (BC) is a malignancy with high mortality among women (1–4). It is a complicated and heterogeneous disease characterized by diverse molecular subtypes. For clinical purposes, BC subtypes are classified according to the prognostic biomarkers: estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), and the proliferation marker Ki-67 (5). The cutoff values for high versus low Ki-67 expression vary from 10–29% and remain debatable, thus limiting its clinical utility (6). In recent decades, specific therapies based on these molecular subtypes-associated parameters, like endocrine therapies for ER-positive or anti-HER2 therapies for HER2 amplification subtypes, have dramatically improved the prognosis of specific patient groups (7–10). However, in BC patients with ER-positive status, endocrine therapies show unsatisfactory effects in about 20% of all patients due to primary or acquired resistance (11). In addition, despite the substantial efficacy of the HER2-targeted therapies, some patients with HER2-amplified BC show inadequate responses to the treatment (12). Furthermore, triple-negative BC (TNBC), which is defined by the absence of ER, PR, and HER2 expression, lacks similar specific targeted therapies (e.g., against ER and HER2) (13). Non-specific chemotherapy remains the primary treatment option for this subtype, indicating the urgent requirement of alternative therapeutic targets for TNBC. In summary, molecular subtype classification has achieved therapy improvement. However, the heterogeneity of tumor cell molecular subtypes enables different types of BC to have distinct therapy strategies and exhibit different therapeutic effects. In search of therapeutic improvement, exploring alternative and/or synergistic therapeutic targets, such as transcription factors (TFs), etc., for conventional targets is a promising concept for improved treatment of BC.

Krüppel like factors (KLFs) are TFs that belong to the triple-fingered proteins of the C2H2-type zinc finger family. The three homologous zinc finger motifs in the C-terminus of KLFs bind to CACCC-, GC-, or GT- box elements found in the promoter or enhancer DNA (14). The N-terminus of KLFs bind to their cofactors(15). The different bindings of the C-terminus and N-terminus contribute to the distinct cellular functions in which KLFs are involved (15). Remarkably, the aberrant expression of several KLFs is observed in BC (16). It has been described that they regulate cell proliferation, apoptosis, differentiation, invasion, migration, and cell metabolism of different molecular subtypes of BC cell lines (16). In addition, KLFs are highly correlated to the clinicopathological characteristics and survival outcomes of BC patients (16). However, the roles of KLFs as tumor-promoters or tumor suppressors are context-dependent and have not yet reached scientific consensus and need to be further explored in BC. Since that KLFs regulate multiple biological processes and are essential for the function of BC tumor cells, the therapy concepts that target several KLFs have been achieved in BC (17–28).

KLF11 as a member of the KLF family, has been initially characterized as a transforming growth factor-beta (TGFβ) inducible early gene (29). However, several studies have demonstrated that it is also inducible by growth factors and hormones, and several cytokines (30–32). In addition to the C-terminal DNA binding and N-terminal transcription regulatory domains, within KLF11, the nuclear localization signaling (NLS) domain is closely next to or among the three zinc finger motifs (15). The NLS exerts the function of KLF nuclear localization(15). KLF11 can regulate several gene network transcriptions as both an activator and a repressor (33–35).

Furthermore, KLF11 suppresses cell proliferation and induces cell apoptosis of untransformed cells such as pancreatic epithelial cells, normal ovary cells, etc., in vitro and in vivo (29, 31, 36–39). However, the role of KLF11 in carcinogenesis is context-dependent. The expression analysis of KLF11 mRNA in small samples has demonstrated its downregulation in pancreatic, breast, renal, ovarian, colon, and gastric cancers (36). Interestingly, KLF11 reveres its role as a cell growth suppressor in normal cells to a cell growth promoter in tumor cells of several cancer types, e.g., pancreatic cancer, ovarian cancer, etc. (39), suggesting that KLF11 might also be a cell growth-promoting factor in breast tumor cells. The incidental results of two studies reported that the genomic region of KLF11 is hypermethylated in lower-grade BC (40), and the mRNA expression of KLF11 is reduced in BC samples (36). Furthermore, MiR-30d decreases cell apoptosis and promotes the epithelial-mesenchymal transition (EMT) process of BC cells via downregulating the expression of KLF11 (41). However, the direct role of KLF11 in the specific function in BC remains largely unknown. However, there is no comprehensive study on the role of KLF11 in BC. Moreover, the prognostic relevance and expression pattern of KLF11 in BC patients remain unclear, not to mention the establishment of KLF11-related prognostic models. Therefore, it is worth exploring whether KLF11 actually acts as an oncogene or a tumor suppressor in BC, implying its potential ability to predict the prognosis and to be a targeting therapy for BC patients.

Here, we explored the prognostic role of the KLF11 in our BC cohort by performing immunohistochemistry (IHC) staining of KLF11 in 298 patients samples, followed by clinicopathological correlation and survival analysis to assess the pro- or anti-tumor role of KLF11 in the BC progression. Afterward, we explored the mechanism underlying KLF11-mediated BC progression from the tumor cell's perspective. It has been demonstrated that the role of KLF11 in other types of cancer cells is mainly related to cell growth, like cell proliferation or cell apoptosis (39). Furthermore, the functional significance of KLFs in BC tumor cells is distinct due to the different molecular subtypes of BC cell lines (16). Therefore, we evaluated the siRNA-mediated loss-of-function of KLF11 by performing cell viability assays, cell proliferation assays, and cell apoptosis assays in three different molecular subtypes of BC cell lines to reveal the direct role of KLF11 in tumor cells of BC.

BC Patients Cohort

In this study, 320 consecutive patients who underwent surgery for BC from 2000 to 2002 at the Department of Obstetrics and Gynecology, Ludwig-Maximillian's-University of Munich, Germany, and of whom tumor tissue was still available were primarily included. To diagnose BC, all patients had undergone tumor biopsy prior to surgery for BC, but no patient has undergone any other prior treatment. In the further analyses, only cases with a diagnosis of sporadic BC and without family history for BC were included (n = 306). Patients with primary distant metastases (n = 6) and patients with only ductal carcinoma in situ (DCIS) but without invasive BC (n = 2) were excluded from further analyses. Then, in total, 298 patients were included. The Institute of Pathology, Ludwig-Maximillian's-University of Munich, assigned the tumor grading (according to the Elston-Ellis system). Patient data regarding patient age, ER status, PR status, HER2-amplification, metastasis, local recurrence, progression, and survival were retrieved from the Munich Cancer Registry. The surrogate molecular subtypes of these patients were defined: Luminal A-like (ER/PR+, HER2 unamplified, Ki-67 ≤ 14%), Luminal B-like (ER/PR positive, HER2 unamplified, Ki-67 > 14%), TNBC (ER-, PR-, HER2 unamplified), HER2 amplified Luminal-like (HER2 amplified, ER/PR+) and HER2 amplified non-Luminal-like (HER2 amplified, ER-, PR-) (42). Survival outcomes were statistically analyzed after an observation period of up to 12 years.

Immunohistochemistry

Formalin-fixed and paraffin-embedded (FFPE) slides of BC patients were used for immunohistochemistry (IHC) analysis. The samples were dewaxed in xylene substitute Roticlear® (Art.-Nr. A538.5, Carl Roth), followed by rinsing with 100% ethanol. Next, the samples were soaked with 0.3% or 0.6% H₂O₂ diluted in methanol for 20 minutes to inhibit the endogenous peroxidase reaction. Subsequent gradually rehydrated in 100%, 70%, and 50% ethanol, followed by rinsing with distilled water. Afterward, a pressure cooker performed antigen retrieval with sodium citrate buffer (pH 6.0). Then, after twice washing with phosphate-buffered saline (PBS) and blocking with Reagent 1 of ZytoChem Plus HRP Polymer Bulk kit (Nr.POLHRP-100, Zytomed System), the slides were incubated with the primary anti-KLF11-antibody (Mouse IgG2a kappa, 1:200, Nr.H00008462-M03, Novus Biologicals) for 16 hours at 4°C. The next day, after washing twice with PBS, the slides were post blocked with Reagent 2 of ZytoChem Plus HRP Polymer Bulk kit (Nr.POLHRP-100, Zytomed System). Afterward, the samples were incubated with HRP-Polymer (Reagent 3 of ZytoChem Plus HRP Polymer Bulk kit, Nr.POLHRP-100, Zytomed system). The color was then developed by Liquid DAB + Substrate Chromogen System kit (Nr. K3468, Dako). Finally, the samples were counterstained with Mayer's hemalum. Subsequent gradually dehydrated in 50%, 70%, 96%, 100% ethanol to Roticlear®, followed by mounting with Roti®Mount medium (Art.-Nr. HP68.1, Carl Roth). Placenta tissue was used as negative and positive control for the staining. The staining samples were finally assessed by Leitz microscope (Type 307-148.001 514686, Wetzlar).

Evaluation of KLF11 Expression

KLF11 was positively stained in the nuclei of cells of BC tissues. Immunoreactive score (IRS) was calculated to evaluate the protein level of KLF11. This semi-quantitative score is calculated as follows: the optical staining intensity (grades: 0 = none, 1 = weak, 2 = moderate, 3 = strong staining) is multiplied by the total percentage of positively stained cells (0 = none, 1 < 10%, 2 = 10–50%, 3 = 51–80% and 4 > 80% of the cells) (43). The range of IRS is 0–12 (43). The IRS of KLF11 expression was assessed by two experienced investigators independently. Receiver operating characteristic (ROC) analysis is commonly used to define optimal cutoff points of biomarkers with a curve based on the measurement of the accuracy evaluated by the area under curve (AUC) (44). This analysis was performed to determine the cutoffs of KLF11 categorization. The protein level of KLF11 was further categorized as "KLF11-high" and "KLF11-low". The optimized cutoff IRS of KLF11 was 2. The IRS of KLF11 ≤ 2 was categorized as "KLF11-low", and IRS > 2 was categorized as "KLF11-high". The ordinal variables of the IRS of KLF11 and the categorized variables of "KLF11-high" and "KLF11-low" correlated to clinical and pathological characteristics. The categorized variables of "KLF11-high" and "KLF11-low" were then correlated to the overall survival (OS), disease-free survival (DFS), distant-metastasis-free survival (DMFS), and local recurrence-free survival (LRFS) of BC.

Forest plots for Subgroup analyses

Forest plots were used to display the prognostic role of KLF11 for OS, DFS, DMFS, and LRFS in BC subgroups grouped by clinicopathological parameters. The R package "ggplot2 (version 3.3.3)" was used to plot the analyzed univariate cox results of subgroups. The R scripts were performed in R programmer version 3.6.3.

Nomogram Construction and Calibration

Multivariate Cox regression-identified independent prognostic variables were integrated to create nomograms for survival probability predictions. Nomograms were modeled and developed using R packages "survival" and "rms". The R scripts were performed in R programmer version 3.6.3. The nomogram capability of discrimination can be assessed by the concordance index (c-index) (45). The value of the c-index can be 0.5-1.0 (45). C-index with a value closer to 1.0 indicates a perfect discriminative performance (45). The prediction accuracy of the established model was evaluated with a calibration curve. In the calibration curve, the predicted prediction that falls on the diagonal indicates a very accurate prediction model.

BC Cell Lines Culture

MCF7 (Nr.86012803, ECACC) as Luminal type, MDA-MB-231 (Nr.92020424, ECACC) as TNBC type, and SK-BR-3 (Nr.ACC736 DSMZ) as HER2 type were used to validate the functional findings of KLF11 involved in BC cell lines. The cell lines were cultured in RPMI medium 1640 (1X) + GlutaMAX™-1 Nr. 61870-010, Gibco) containing 10% fetal bovine serum (FBS Qualified, Nr.10270-106, Gibco). No antibiotics or antimycotics were added to the culture medium. The culture condition of the cells was an incubator with a stable temperature (37°C), stable CO2 level (5%), and high relative humidity (95%). In subsequent experiments, only cells detected to be free of mycoplasma infection were used. Neubauer cell chambers were used to count cells when performing cell experiments.

siRNA-Mediated Knockdown

For siRNA knockdown experiments, BC cell lines were transfected with AllStars Negative Control (NC) siRNA (Nr.1027280, Qiagen), and two independent siRNAs target KLF11: KLF11-S1: CACGTAGATAACCGAGAGAAT (Nr.SI04139751, Qiagen) and KLF11-S2: AGGAAGCGGCATGACAGCGAA (Nr.SI4291175, Qiagen). The entire transfection procedure was performed according to the manufacturer's protocol of the Lipofectamine RNAiMAX reagent (Nr.13778-075, Invitrogen). BC cell lines were pre-seeded in six-well plates. siRNA-transfection was performed when the cell density reached approximately 60%. Opti-MEM Reduced Serum Medium (Nr.31985-047, Gibco) containing siRNAs (25pmol) and Lipofectamine RNAiMAX Reagent (7.5µl) was used for the transfection of the cells. After 48 hours, BC cell lines were harvested for protein or RNA extraction and detection or further cell functional experiments.

RNA Extraction and Real-Time PCR

RNA was extracted following the procedure documented in the manual of RNeasy Mini Kit (Nr.74104, Qiagen). Subsequent reverse transcription was carried out with Biozym cDNA Synthesis Kit (Nr.331470L, Biozym) and implemented as the manual document. Real-time PCR (rtPCR) was carried out to validate the knockdown of KLF11 expression from mRNA level in different BC cell lines with FastStart Essential DNA Probes Master kit (Nr.06402682001, Roche) gene-specific primers using the LightCycler Nano (Roche). ACTB (β-actin) was used as the reference gene. The 20 µL reaction system was used, consisting of cDNA (1 µl), H₂O (8 µl), and Mastermix (10 µL) and the mixture of forward primer (0.5 µl) and reverse primer (0.5 µl). The process of rt-PCR cycling was set as 10 minutes for pre-incubation and two step-amplification: 40 cycles of 20 seconds at 95 ℃ and 40 seconds at 60 ℃. The primers were KLF11 (Forward: 5’-CTTCCATTCTTTATCGACTCTGTG-3' and Reverse: 5'- GATGGCTCCACGAGATCAG-3', Nr.100154265, Roche) and ACTB (Forward: 5'- TCCTCCCTGGAGAAGAGCTA-3' and Reverse: 5'- CGTGGATGCCACAGGACT-3', Nr.100143492, Roche)

Colorimetric Cell-Based KLF11 ELISA

The KLF11 protein level in BC cell lines was detected using the colorimetric cell-based KLF11 ELISA Kit (Dr.DEIA-XYA1113, CD Creative Diagnostic). The kit was used to validate the siRNA-mediated knockdown of KLF11 from protein levels in different cultured BC cell lines. The siRNA transfection for KLF11 knockdown of BC cell lines was conducted as described above. Afterward, the entire measurement procedure was performed as the colorimetric cell-based KLF11 ELISA Kit manual documented. Cells were then seeded (20000 cells/well) into 96-well plates with triplicates for each cell line and grown in the 10% FBS-containing RPMI 1640 medium for 24 hours. The absorbance of optical density (OD)₄₅₀ was then measured for the target protein. The absorbance of OD₅₉₅ was measured for crystal violet cell staining using Elx800 universal Microplate Reader and analyzed using Gene 5 software. For data normalization, the absorbance of OD₄₅₀ of the target protein and internal reference protein were normalized using the absorbance of OD₅₉₅ via the portion, then normalized the target protein values with the internal reference protein values. Three replicates were performed with each cell line. To guarantee the reliability, we repeated the experiment at least three times.

Cell Viability Assay

Cell viability assay of BC cell lines was performed using the Methylthiazolyldiphenyl-tetrazolium bromide (MTT). The siRNA transfection for KLF11 knockdown of BC cell lines was conducted as described above. Cells were then seeded (3500 cells/well) into three sets of 96-well plates with five replicates for each cell line and cultured in 10% FBS-containing RPMI 1640 medium for 24 hours, 48hours, and 72hours, respectively. Then incubation ended at three different time points, and then at each time point to each well, the MTT solution (20 µl/well, 5 mg/ml, Nr.M5655, Sigma-Aldrich) was pipetted. Afterward, the plates were put back in the cell incubator for another 1.5 hours. After removing the MTT-containing incubation medium, dimethyl sulfoxide (DMSO, 200 µl/well) was added. A shaker was used to mix DMSO and formazan crystal thoroughly for 5 min at room temperature (RT). The dissolved formazan crystals were used to represent the number of viable cells. The absorbance of OD₅₉₅ was measured using an Elx800 universal Microplate Reader and analyzed using Gene 5 software. To guarantee the reliability, we repeated the experiment at least three times.

Cell Proliferation Assay

Cell proliferation assay was performed with the Cell Proliferation ELISA kit (Nr.11647229001, Roche) based on 5-bromo-2’-deoxyuridine (BrdU) labeling and detection. The entire procedure has been performed according to the kit's manual. The siRNA transfection for KLF11 knockdown of BC cell lines was conducted as described above. Cells were then seeded (5000 cells/well) into two sets of 96-well plates with five replicates for each cell line and cultured in10% FBS-containing RPMI 1640 medium for 24 hours and 48 hours, respectively. Then incubation ended at two different time points, and then each time point to each well was supplemented with BrdU (20 µl/well), and the plates were put back in the cell incubator for another 24 hours. After removing the BrdU-contained incubation medium, the fixing solution (200 µl/well) was pipetted in the wells to fix the cells. Afterward, cells were incubated with anti-BrdU-POD (100 µl/well) to label the BrdU that was incorporated into the cellular DNA. After thoroughly washing, a substrate solution (100 µl/well) was pipetted in and incubated until the development of blue color. Sulfuric acid (1M, 25 µl/well) was then pipetted in to stop the reaction. The absorbance of OD₄₅₀ was measured using Elx800 universal Microplate Reader and analyzed using Gene 5 software. To guarantee the reliability, we repeated the experiment at least three times.

Cell Apoptosis Assay

Cell apoptosis assay was performed using the Cell Death Detection ELISA kit (Nr.11544675001, Roche). The apoptosis was determined by DNA fragment analysis. The whole detection procedure followed the manual instructions of the kit. The siRNA transfection for KLF11 knockdown of BC cell lines was conducted as described above. BC cell lines were then harvested and prepared as sample lyses for further detection. After forming the "sandwich", which consists of anti-histone antibody, sample lyses containing nucleosomes, and anti-POD-DNA-peroxidase determined the amount of peroxidase retained in the immune complex with substrate solution. The absorbance of OD₄₀₅ was measured using Elx800 universal Microplate Reader and analyzed using Gene 5 software. To guarantee the reliability, we repeated the experiment at least three times.

Statistical Analysis

IBM SPSS (version 26) was used for statistical analysis of the data. Adobe Illustrator (AI) and Graphpad Prism 8.1 were used for illustrations. Data analyzed statistically was first subjected to the Shapiro-Wilk normality test and Levene's equal variances test. Proper statistical test methods were selected according to normality and variances and the number of comparison groups. Student's t-test, Welch's test, and Mann-Whitney U test were used to perform two-group comparisons. One-way ANOVA test, Welch's test, and Kruskal-Wallis test were used to perform multiple-group comparisons. Subsequent Dunn's test was implemented for pairwise comparisons within the multiple groups. In addition, Chi-Square Test was performed for the comparison of categorical variables. For survival analysis of BC patients, Kaplan-Meier (KM) analysis was performed for the generations of survival curves with the Log-rank test. Univariate and multivariate Cox regression analyses were also performed for the survival analysis with a hazard ratio (HR) and a 95% confidence interval (CI). The proportional hazards assumption test was performed for each variable in all cox models using the Schoenfeld statistical test. All reported p values are two-sided. P-values < 0.05 were considered statistically significant.

The Quantification and Distribution of KLF11 Protein Expression in BC

To explore the prognostic role of KLF11 in BC, we performed IHC staining of KLF11 in 298 specimens of BC patients was performed and quantified by the IRS. KLF11-IRS were then correlated to clinicopathological characteristics and survival outcomes. KLF11 was successfully stained in 292/298 samples (6 samples without sufficient staining due to technical issues). We demonstrated that KLF11 protein levels (measured by IRS) were positively associated with Ki-67 expression (P = 2.4e-03, Fig. 1a-1c). KLF11 also correlated with the BC molecular subtypes (P = 0.027, Fig. 1d-1f) defined by surrogate biomarkers in our BC cohort. The lowest KLF11 expression was shown in luminal A-like BC (Fig. 1f). The further pairwise comparison demonstrated that KLF11 expression was higher in Luminal B-like BC than in Luminal A-like BC (P = 0.016, Fig. 1f).

We then classified the BC patients into the two subgroups "KLF11-high/-low" determined by KLF11 protein levels and compared these subgroups to clinicopathological parameters. Of all 292 stained samples, 202 (69.2%) cases showed high, while 90 cases (30.8%) showed low KLF11 protein levels. We found that there were significantly less “KLF11-high” patients with luminal-A subtype than with a non-luminal A subtype of BC (64.4% vs. 76.4%, p = 0.028, Table 1) and vice versa with “KLF11-low” (35.6% vs. 23.6%, p = 0.028, Table 1). Additionally, there were significantly less “KLF11-high” patients in BC with Ki-67 expression ≤ 14% than in BC with Ki-67 > 14% patients (64.6% vs. 79.7%, p = 0.033, Table 1), and vice versa with “KLF11-low” (35.4% vs. 20.3%, p = 0.028, Table 1). No further associations of KLF11 protein levels with other clinic pathological parameters in BC patients were found.

Taken together, these results with both the absolute KLF11-IRS and the categorized "KLF11-high/-low" subgroups confirmed that KLF11 was higher in highly proliferative BC and lowest in luminal A-like BC.

Table 1

Distribution of KLF11 protein expression
	Total	KLF11-low protein level	KLF11-high protein level	P value
Characteristics	Number of cases (%)	Number of cases (%)	Number of cases (%)	P value
N	292 (98.3%)	90 (30.8%)	202 (69.2%)
Age(yo)				0.127
༜50 yo	72 (24.7%)	17 (23.6%)	55 (76.4%)
≥ 50 yo	220 (75.3%)	73 (33.2%)	147 (66.8%)
Tumor histology				0.995
Invasive Ductal	118 (40.5%)	36 (30.5%)	82 (69.5%)
Invasive lobular	83 (28.5%)	26 (31.3%)	57 (68.7%)
Mixed type	37 (12.7%)	11 (29.7%)	26 (70.3%)
other types	53 (18.2%)	17 (32.1%)	36 (67.9%)
Molecular subtype				0.254
Luminal A-like	163 (56.2%)	58 (35.6%)	105 (64.4%)
Luminal B-like	58 (20%)	12 (20.7%)	46 (79.3%)
Triple negative	37 (12.8%)	10 (27.0%)	27 (73.0%)
Her2 amplified luminal-like	23 (7.9%)	6 (26.1%)	17 (73.9%)
Her2 amplified non luminal-like	9 (3.1%)	2 (22.2%)	7 (77.8%)
Luminal A like subtype				0.028*
No	127 (43.8%)	30 (23.6%)	97 (76.4%)
Yes	163 (56.2%)	58 (35.6%)	105 (64.4%)
Tumor grade				0.328
G1	15 (9.3%)	6 (40.0%)	9 (60.0%)
G2	102 (63.4%)	27 (26.5%)	75 (73.5%)
G3	44 (27.3%)	16 (36.4%)	28 (63.6%)
Tumor focus				0.429
Unifocal	157 (53.8%)	52 (33.1%)	105 (66.9%)
Multifocal&Multicentric	135 (46.2%)	38 (28.1%)	97 (71.9%)
Axillary lymph node metastasis				0.706
No	161 (56.3%)	51 (31.7%)	110 (68.3%)
Yes	125 (43.7%)	37 (29.6%)	88 (70.4%)
pT classification				0.907
pT1	190 (65.1%)	59 (31.1%)	131 (68.9%)
pT2-4	102 (34.9%)	26 (30.4%)	60 (69.6%)
ER status				0.467
Negative	56 (19.2%)	15 (26.8%)	41 (73.2%)
Positive	236 (80.8%)	75 (31.8%)	161 (68.2%)
PR status				0.294
Negative	123 (42.1%)	42 (34.1%)	81 (65.9%)
Positive	169 (57.9%)	48 (28.4%)	121 (71.6%)
HER2 status				0.467
Negative	259 (89%)	81 (31.3%)	178 (68.7%)
Positive	32 (11%)	8 (30.6%)	24 (69.4%)
Expression of Ki-67				0.033*
≤ 14%	164 (73.5%)	58 (35.4%)	106 (64.6%)
> 14%	59 (26.5%)	12 (20.3%)	47 (79.7%)
yo, year old; pT, pathological Tumor size; ER, Estrogen receptor; PR, Progesterone receptor; HER2, Human epidermal growth factor receptor 2; , p < 0.05; , p < 0.01; **, p < 0.001.

Prognostic Significance of KLF11 for OS, DFS, DMFS, LRFS in BC

In the overall patient cohort, patients with a high KLF11 expression showed an impaired DFS (HR = 2.41, 95%CI = 1.54–3.77, p = 0.001, Fig. 2b), an impaired DMFS (HR = 2.11, 95%CI = 1.23–3.63, p = 0.018, Fig. 2c) and an impaired LRFS (HR = 2.62, 95%CI = 1.42–4.81, p = 0.01, 2d). However, no association of KLF11 with OS of BC patients was found (HR = 1.13, 95%CI = 0.72–1.78, p = 0.601, Fig. 2a).

The subgroup analysis of DFS in our BC cohort showed that higher KLF11 led to an inferior DFS both in patients aged of < 50 yo (HR = 5.021, 95%CI = 1.174–21.477, n = 72, p = 0.03, Additional file 2) and patients aged ≥ 50 yo (HR = 2.016, 95%CI = 1.103–3.686, n = 220, p = 0.023, Additional file 2). KLF11 was also negatively associated with DFS both of patients with unifocal tumors (HR = 2.487, 95%CI = 1.090–5.672, n = 157, p = 0.03, Additional file 2) and with the appearance of multifocal/multicentric tumors (HR = 2.505, 95%CI = 1.201–5.221, n = 135, p = 0.014, Additional file 2). In addition, a higher KLF11 level lead to an inferior DFS of the patients with histological type of invasive ductal/invasive lobular/mixed type (HR = 2.173, 95%CI = 1.247–3.787, n = 238, p = 0.006, Additional file 2), of Luminal A-like BC (HR = 2.831, 95%CI = 1.341–5.978, n = 163, p = 0.006, Additional file 2), of patients with G1 or G2 BC ( HR = 3.108, 95%CI = 1.075–8.976, n = 117, p = 0.036, Additional file 2), of patients without lymphatic metastasis (HR = 4.255, 95%CI = 1.650-10.976, n = 161, p = 0.003, Additional file 2), of patients with tumor size smaller than 2cm (HR = 3.268, 95%CI = 1.517–7.028, n = 190 p = 0.002, Additional file 2), of patients with ER positive (HR = 2.740, 95%CI = 1.488–5.045, n = 236 p = 0.001, Additional file 2), of patients with PR positive ( HR = 2.591, 95%CI = 1.259–5.334, n = 169 p = 0.01, Additional file 2) and of patients with HER2 unamplified BC (HR = 2.414, 95%CI = 1.343–4.339, n = 259, p = 0.003, Additional file 2) and in the BC subgroup with a low proliferation rate (Ki-67 ≤ 14%) (HR = 2.831, 95%CI = 1.341–5.978, n = 164, p = 0.006, Additional file 2).

Similarly, a DMFS analysis across various patient subgroups was performed. A high KLF11 protein level significantly correlated to worse DMFS of the BC patients with histological type with invasive ductal/Invasive lobular/mixed type (HR = 1.926, 95%CI = 1.008–3.683, n = 238, p = 0.047, Additional file 3), of patients without lymphatic metastasis ( HR = 6.256, 95%CI = 1.451–26.962, n = 161, p = 0.014, Additional file 3), of patients with tumor size ≤ 2cm ( HR = 2.690, 95%CI = 1.004–7.205, n = 190, p = 0.049, Additional file 3), of patients with ER positive (HR = 2.308, 95%CI = 1.097–4.853, n = 236, p = 0.028, Additional file 3), of patients with PR positive (HR = 2.751, 95%CI = 1.047–7.228, n = 169, p = 0.04, Additional file 3) and of patients with HER2 unamplified BC (HR = 2.169, 95%CI = 1.076–4.374, n = 259, p = 0.03, Additional file 3).

Additionally, the subgroup analyses of LRFS showed that a high KLF11 level correlated with an impaired LRFS of patients aged older than 50y (HR = 2.382, 95%CI = 1.032–5.493, n = 220, p = 0.042, Additional file 4), of patients with histological type of invasive ductal/Invasive lobular/mixed type (HR = 2.297, 95%CI = 1.054–5.005, n = 238, p = 0.036, Additional file 4), of Luminal A-like BC (HR = 3.261, 95%CI = 1.220–8.722, n = 163, p = 0.019, Additional file 4), of patients with unifocal tumors (HR = 3.001, 95%CI = 1.028–8.761, n = 157, p = 0.044, Additional file 4), of patients without lymphatic metastasis (HR = 3.946, 95%CI = 1.157–13.463, n = 161, p = 0.028, Additional file 4), of patients with tumor size smaller than 2cm (HR = 4.329, 95%CI = 1.504–12.461, n = 190, p = 0.007, Additional file 4), of patients with ER positive (HR = 3.651, 95%CI = 1.522–8.756, n = 236, p = 0.004, Additional file 4), of patients with PR positive (HR = 2.662, 95%CI = 1.017–6.968, n = 169, p = 0.046, Additional file 4) and of patients with HER2 unamplified BC ( HR = 2.627, 95%CI = 1.156–5.970, n = 259, p = 0.021, Additional file 4) and also of the BC subgroup with a low proliferation rate (Ki-67 ≤ 14%) (HR = 3.261, 95%CI = 1.220–8.722, n = 164, p = 0.019, Additional file 4).

Concerning OS, no correlation with KLF11 was found in the subgroup survival analysis (Additional file 1), which was consistent with the result of the overall-BC cohort.

KLF11 was an independent Indicator for DFS and DMFS of BC

Univariate cox regression analysis showed that KLF11 (HR = 2.433, 95%CI = 1.407–4.208 p = 0.001), grading (HR = 1.940, 95%CI = 1.106–3.403, p = 0.021), tumor size (HR = 1.991, 95%CI = 1.301–3.047, p = 0.002) and lymph node status (HR = 1.832, 95%CI = 1.187–2.829, p = 0.006) were significantly associated with DFS in the BC cohort (Table 2). Multivariate regression was further performed with the univariate significant factors: grading, tumor size, lymph node status and KLF11. The multivariate Cox regression revealed that KLF11 (HR = 2.610, 95%CI = 1.241–5.488, p = 0.011), grading (HR = 2.260, 95%CI = 1.262–4.047, p = 0.006) and tumor size (HR = 2.624, 95%CI = 1.384–4.975, p = 0.003) were independent factors for DFS of BC patients (Table 2).

Table 2

Univariate and multivariate Cox regression analyses of KLF11 and clinicopathological characteristics for DFS in BC patients
Characteristics	Univariate analysis			Multivariate analysis
	p	HR	95% CI	p	HR	95% CI
Age(༜50 yo vs. ≥50 yo)	0.223	1.337	0.893–2.130	n.i.	n.i.	n.i.
Molecular subtype (non-LuA like vs.LuA-like)	0.137	1.379	0.903–2.106	n.i.	n.i.	n.i.
Tumor histology (Invasive Ductal & Invasive lobular & Mixed type vs. other types)	0.075	1.938	0.935–4.015	n.i.	n.i.	n.i.
Grading (G3 vs. G1-G2)	0.021*	1.940	1.106–3.403	0.006**	2.260	1.262–4.047
Tumor focis (multifocal & muticentric vs.unifocal )	0.238	1.292	0.844–1.976	n.i.	n.i.	n.i.
Tumer size (pT2-pT4 vs. pT1)	0.002**	1.991	1.301–3.047	0.003**	2.624	1.384–4.975
Axillary lymph node status (yes vs. no)	0.006**	1.832	1.187–2.829	0.901	1.042	0.547–1.985
ER status (ER + vs. ER-)	0.608	0.875	0.525–1.457	n.i.	n.i.	n.i.
PR status (PR + vs. PR-)	0.404	1.202	0.780–1.852	n.i.	n.i.	n.i.
HER2 status (HER2 + vs. HER2-)	0.544	1.208	0.565–2.226	n.i.	n.i.	n.i.
Expression of Ki-67(Ki-67༞14% vs. Ki-67 ≤ 14%)	0.091	1.576	0.931–2.671	n.i.	n.i.	n.i.
KLF11 (High vs. Low)	0.001**	2.433	1.407–4.208	0.011*	2.610	1.241–5.488
KLF11, Krüppel like Factor 11; BC, Breast cancer; yo, years old; DFS, Disease-free survival; pT, pathological Tumor size; ER estrogen receptor; PR progesterone receptor; HER2 human epidermal growth factor receptor 2; LuA-like luminal A-like; non-LuA like, not luminal A-like; HR, Hazard ratio; CI Confidence interval; n.i., not included in multivariate model, as p > 0.05 in univariate analysi; , p < 0.05; , p < 0.01; **, p < 0.001.

Furthermore, univariate cox regression revealed that KLF11 (HR = 2 .132, 95%CI = 1.125–4.043, p = 0.02), molecular subtype (HR = 1.772, 95%CI = 1.054–2.981, p = 0.031), grading (HR = 2.689, 95%CI = 1.281–5.644, p = 0.009), tumor size (HR = 3.044, 95%CI = 1.818–5.099, p = 0.000023) and lymph node status (HR = 2.328, 95%CI = 1.355–3.998, p = 0.002) were significantly associated with DMFS in BC (Table 3). Multivariate analysis was then performed with the univariate significant factors: molecular subtype, grading, tumor size, lymph node status and KLF11. The result showed that KLF11 (HR = 2.744, 95%CI = 1.017–7.403, p = 0.046), grading (HR = 3.276, 95%CI = 1.424–7.536, p = 0.005), and tumor size (HR = 5.729, 95%CI = 2.266–14.484, p = 0.000225) were independently related to DMFS of BC patients (Table 3). Concerning OS and LRFS, no independent prognostic significance of KLF11 was found in the multivariate cox regression.

Table 3

Univariate and multivariate Cox regression analyses of KLF11 and clinicopathological characteristics for DMFS in BC patients
Characteristics	Univariate analysis			Multivariate analysis
	p	HR	95% CI	p	HR	95% CI
Age(༜50yo vs. ≥50yo)	0.514	1.211	0.682–2.150	n.i.	n.i.	n.i.
Molecular subtype (non-LuA like vs.LuA-like)	0.031*	1.772	1.054–2.981	0.367	1.488	0.628–3.529
Tumor histology (Invasive Ductal & Invasive lobular & Mixed type vs. other types)	0.100	2.158	0.862-5.400	n.i.	n.i.	n.i.
Grading (G3 vs. G1-G2)	0.009**	2.689	1.281–5.644	0.005**	3.276	1.424–7.536
Tumor focis (multifocal & muticentric vs.unifocal )	0.074	1.607	0.955–2.704	n.i.	n.i.	n.i.
Tumer size (pT2-pT4 vs. pT1)	0.000023***	3.044	1.818–5.099	0.000225***	5.729	2.266–14.484
Axillary lymph node status (yes vs. no)	0.002**	2.328	1.355–3.998	0.942	1 .035	0.415–2.579
ER status (ER + vs. ER-)	0.101	0.624	0.355–1.096	n.i.	n.i.	n.i.
PR status (PR + vs. PR-)	0.479	0.831	0.497–1.388	n.i.	n.i.	n.i.
HER2 status (HER2 + vs. HER2-)	0.182	1.565	0.811–3.108	n.i.	n.i.	n.i.
Expression of Ki-67(Ki-67༞14% vs. Ki-67 ≤ 14%)	0.083	1.796	0.927–3.480	n.i.	n.i.	n.i.
KLF11 (High vs. Low)	0.02*	2.132	1.125–4.043	0.046*	2.744	1.017–7.403
KLF11, Krüppel like Factor 11; DMFS, Distant metastasis-free survival; pT, pathological Tumor size; non-LuA-like, not luminal A-like; ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; LuA-like, luminal A-like; HR, hazard ratio; CI, confidence interval; n.i., not included in multivariate model, as p > 0.05 in univariate analysis; , p < 0.05; , p < 0.01; **, p < 0.001.

Nomograms Construction and Calibration for the Established KLF11 Related DFS and DMFS Cox Model

Based on the independent predictors found in the Cox regression analyses, we developed two nomograms for predicting BC patients' DFS (Fig. 3.3.10 A) and DMFS (Fig. 3.3.10 C). Nomograms estimate the survival probability after 3, 5, and 10 years based on a total score calculated by the addition of zero to 100 points for every individual prognostic factor. Then the 3-, 5- and 10-year survival probabilities can be estimated by adding up all points. In the nomogram for predicting DFS (Fig. 3a), the KLF11 protein level showed a high impact due to the high level of KLF11 almost added up to 100 points to the final score. However, regarding DMFS, KLF11 protein level only showed a minor impact on outcome prediction (Fig. 3c).

The developed nomograms were internally validated using bootstrapping. The internal validation of the underlying regression models showed optimism adjusted c-index values of 0.694 for DFS and 0.8 for DMFS, respectively calculated by using the R programmer. The calibration of the model was assessed with calibration curves, which analyze the fit between the model established by the Cox regression method and the actual situation. The calibration curve of the nomogram-predicted probability of DFS (Fig. 3b) indicated that the model's predicted survival probabilities for 3-, 5- 10- years were close to observed probabilities. Calibration for the 10-year probability of DFS prediction for the nomogram showed perfect model calibration with a high correlation of nomogram predicted probability of DFS and observed probability of DFS estimated by the KM method. The calibration plots for the probabilities of 3-, 5- and 10-year DMFS (Fig. 3d) also showed that the predicted-DMFS by nomogram fit well with the actual DMFS of BC patients. However, this prediction accuracy was not as good as the nomogram of the DFS prediction.

Taken together, we demonstrated in the BC patient cohort that KLF11 was higher in highly proliferative BC and lowest in luminal A-like BC. Furthermore, BC patients in the subgroup "KLF11-high" had a shorter DFS, DMFS, and LRFS. In addition, high KLF11 expression remained to be an independent indicator for DFS and DMFS of BC. KLF11-related nomograms of DFS and DMFS have a high accuracy in predicting the 3-,5- and 10 -year survival probability of BC patients. Our results indicate that KLF11 might act as an oncogene in BC.

siRNA-mediated loss-of-function of KLF11 in BC cell lines

Afterward, we explored the possible mechanism underlying KLF11-mediated BC progression from the perspective of tumor cells. To investigate the pro-tumor significance of KLF11 in BC cells, we performed siRNA-mediated loss-of-function of KLF11 with cell viability assay, cell proliferation assay, and cell apoptosis assay in ER-positive luminal type, HER2 amplified type, and triple-negative type BC cell lines.

Then, MCF7 as luminal type, SK-BR-3 as HER2 type, and MDA-MB-231 as TNBC type were transfected with non-targeting control siRNA as negative control (NC), and two independent siRNAs target KLF11 (KLF11-S1 and KLF11-S2). The knockdown efficiency was confirmed at the mRNA level by RT-PCR (Additional file 5a) and at the protein level by colorimetric cell-based KLF11 ELISA (Additional file 5b).

Furthermore, we performed relevant functional experiments in MCF7, SK-BR-3, and MDA-MB-231 cell lines. All cell lines showed a significant reduction of viable cells after transfection with KLF11-siRNAs compared to KLF11-NC (Fig. 4a-c). In addition, the highest inhibitory effect on cell viability was observed in the MDA-MB-231 cell line (Fig. 4c). These results indicated that KLF11-knockdown has an inhibitory effect on tumor cell survival. Moreover, we aimed to assess whether the cell number increase was based on the influence of cell proliferation. After the knockdown of KLF11, a high inhibition of proliferation was obtained in MDA-MB-231 cells (Fig. 4f). Only modest inhibition of proliferation was observed in MCF7 cells (Fig. 4d). However, there was no proliferation inhibitory effect in the SK-BR-3 cells (Fig. 4e). We could demonstrate that KLF11-knockdown only inhibited cell proliferation of HER2 negative BC cell lines, the most effective impact was obtained on the TNBC cell line. Next, cell apoptosis assays were performed in MCF7, SK-BR-3, and MDA-MB-231 cells transfected with KLF11-siRNAs. This might be another reason that induces the inhibition of cell survival by downregulating KLF11 expression. We found that KLF11-knockdown induced MCF7, SK-BR-3, and MDA-MB-231 cells apoptosis. In addition, a higher induction of apoptosis was obtained in MCF7 (Fig. 4g) and SK-BR-3 cell lines (Fig. 4h). Only a modest induction of apoptosis was observed in the MDA-MB-231 cell line (Fig. 4i).

In summary, there was a high degree of consistency in cellular functional assays suggesting that KLF11 acts as a pro-tumor TF in BC cells via promoting cell proliferation and/or decreasing cell apoptosis. However, due to neoplastic intratumor heterogeneity, different cancer molecular subtypes are likely to have different underlying mechanisms as the local signaling microenvironment could cause context-specific changes.

In our BC patients cohort, we demonstrated in the BC patient cohort that KLF11 was higher in highly proliferative BC and lowest in luminal A-like BC. Furthermore, patients with high KLF11 expression had a shorter DFS, DMFS, and LRFS. In addition, high KLF11 expression remained to be an independent indicator for DFS and DMFS of BC. Our patient-related results indicate that KLF11 might act as an oncogene in BC. Besides, different types of parameters mediated nomograms for predicting OS or DFS in BC patients have been developed (46–50). KLF11-related nomograms of DFS and DMFS developed by the current study showed high accuracy in predicting the 3-,5- and 10 -year survival probability of BC patients. We are glad to establish well-fitting DFS and DMFS prediction models from our BC cohort. In cell culture assays suppressing KLF11-function via siRNA knockdown, we demonstrated that KLF11 acts as a pro-tumor TF in BC cells via promoting cell proliferation and/or decreasing cell apoptosis. However, due to neoplastic intratumor heterogeneity, different cancer molecular subtypes are likely to have different underlying mechanisms.

Consistent with our results, a previous bioinformatics study also suggests that KLF11 relates to the impaired prognosis of BC patients (51). Higher KLF11 expression is observed in a higher grade than in the lower grade tumors, in lymphatic metastasis cases than in cases without lymphatic metastasis, and in TNBC patients than in other molecular subtypes (51). Furthermore, the study that explored the clinical role of KLF11 in endometrial cancer, a highly hormone-dependent female cancer like BC, demonstrated that KLF11 is positively associated with aggressive features (52). In addition, endometrial cancer patients with higher KLF11 levels exhibit poorer survival outcomes (52). This previous study can be supportive evidence of our results. KLF11 might also have a similar impact on progression and prognosis in BC. Furthermore, in prognostic analyses of KLF11 in BC patients grouped by clinicopathological features, we found that KLF11 was related to an impaired prognosis in BC patients of specific subgroups but not in their corresponding contrary subgroups (e. g., ER-positive vs. ER-negative BC), indicating that the prognostic role of KLF11 might be BC microenvironment-dependent.

It has been demonstrated previously that miR-30d promotes BC cell growth, depending on the low level of KLF11 (41). This also indicates that KLF11 might act as a tumor-suppressor in BC. However, the current study experimentally confirmed that KLF11 played a pro-tumor role in BC cell lines. We speculate that when restricted to a tumor context, the transcription-regulated genes by KLF11 in BC tumor cells were reversed compared to in the normal breast tissue. However, we do not know the exact mechanism of how KLF11 influences the growth of BC tumor cells. Physiologically, KLF11 can strengthen TGFβ-induced cell growth inhibition and TGFβ-induced cell apoptosis (38, 54, 55). The mechanism of KLF11 enhancing the TGFβ signaling pathway is documented as follows: TGFβ-induced growth inhibition works via activating the intracellular Smad signaling pathway (56, 57). This signaling is mediated by the heteromeric complexes of specific TGFβ-type II and -type I kinase receptors, which phosphorylate the Smad2/Smad3 complex (57). Phosphorylated Smad2/Smad3 forms a complex with Smad4 (58). Afterward, the complex locates in the nucleus and regulates gene transcription, including KLF11 (58). KLF11 then represses the expression of Smad7 by binding its promoter, which is mediated through the recruitment of the histone deacetylase co-repressor mSin3a (54, 59, 60). There is a negative feedback loop within the TGFβ/Smad signaling pathway, Smad7, which binds to activated TGFβ type I receptors and prevents entry and phosphorylation of the Smad2/Smad3 receptors (61, 62). Due to the induction of upstream KLF11, inhibitory Smad7 expression is decreased, which leads to the disruption of the negative feedback loop of TGFβ/Smad signaling. This further strengthens the TGFβ/Smad signaling (54). Disruption of TGFβ/Smad signaling is common in human cancers and facilitates tumor cells to evade TGFβ-induced cell inhibition (63, 64). Interestingly, KLF11 does suppress growth in untransformed cells and even transforms normal fibroblasts (39). However, it appears to be a pro-tumor regulator of several cancer types, e.g., pancreatic cancer, ovarian cancer, esophageal cancer, etc. (39). From a mechanistic perspective, KLF11 reverses its role as a cell growth suppressor to a cell growth promoter in pancreatic cancer. This reversion depends on a RAS mutation that inhibits the TGFβ-induced inhibition of proliferation is inhibited in pancreatic cancer cells (55). Due to the oncogene-RAS mutation, the KLF11-mSin3a interaction is disrupted by the phosphorylated KLF11, which diminishes the termination of the inhibitory Smad7 loop in the TGFβ-inhibitory effect (54). Consequently, KLF11 turns out to promote tumor growth (55). From another mechanistic perspective, KLF11 inhibits the proliferation of esophageal adenocarcinoma cell lines in vitro via its suppression of the biosynthesis of prostaglandin E2 (PGE2). However, in esophageal cell lines with upregulated EGFR/AKT signaling, the tumor-suppressive role of KLF11 appears to be reversed (65). In esophageal adenocarcinoma cell lines, KLF11 binds to the promoter of cytosolic phospholipase A2α (cPLA2α), a rate-limiting enzyme of oncogenic PGE2 biosynthesis, which in turn recruits chromatin co-suppressor Sin3a/HDAC, then suppresses the cPLA2α expression as well as decreases the PGE2 level (65). When turning to an EGFR/AKT signaling activated context, due to the phosphorylation of the KLF11-Sin3a interaction domain, the whole signal pathway trend turned to be reversed (65). These situations might also happen in BC, and in-depth studies are necessary to explain and explore the context-dependent regulatory mechanism of KLF11 in BC.

However, there are limitations. Our BC cohort is 20 years old (primary diagnosis 2000–2002). The advantage of an old patient cohort in terms of very long follow-up data is, in this case, also a disadvantage, as data on treatment cannot be obtained anymore. In addition, at that time, neoadjuvant chemotherapy therapies were not applied regularly in our clinic. However, the advantage of our data is that no patient in our cohort had undergone any prior treatment before surgery was performed. Regarding adjuvant therapy, even if we do not display it specifically, it is assumed to be the same in the two groups now sub-divided by KLF11-high/-low. Additionally, KLF11 was an independent prognostic factor in our cohort, which suggests that treatment (which depends on the other factors in this analysis, of which KLF11 were independent) should not cause any bias in our data. Since neoadjuvant chemotherapy has become common recently, this factor must be regarded in any research using more recent patient samples.

We showed a higher KLF11 expression in the BC patients accompanied by a highly proliferative molecular subtype. Moreover, a high KLF11 protein level was associated with an impaired prognosis. Furthermore, KLF11 remained an independent and unfavorable prognostic indicator for DFS and DMFS of BC. Although it has been demonstrated to be lower expressed in BC tissues than the normal breast tissues, KLF11 appeared to be a pro-tumor KLF member and a prognostic risk factor in BC. Our study indicates that KLF11, as a transcriptional factor, acts as a new prognostic biomarker and might also be a potential synergistic or an alternative therapeutic target for conventional targeting treatments in human BC.

AUC	Area under curve
BC	Breast cancer
BrdU	5-bromo-2’-deoxy-uridine
CI	Confidence interval
C-index	Concordance-index
DCIS	Ductal carcinoma in situ
DFS	Disease-free survival
DMFS	Distant-metastasis-free survival
DMSO	Dimethyl sulfoxide
EMT	Epithelial-mesenchymal transition
ER	Estrogen receptor
FFPE	Formalin-fixed and paraffin-embedded
HER2	Human epidermal growth factor receptor 2
HER2 Lu-like	HER2 amplified luminal-like
HER2 non-Lu-like	HER2 amplified non-luminal-like
HR	Hazard ratio
IHC	Immunohistochemistry
IRS	Immunoreactive score
KLF	Krüppel like Factor
KM	Kaplan-Meier
LRFS	Local recurrence-free survival
LuA	Luminal A
LuB	Luminal B
MTT	Dimethyl-thiazolyl-diphenyl-tetrazolium bromide
NR	Nuclear receptor
NLS	Nuclear localization signal
OD	Optical density
OS	Overall survival
PBS	Phosphate buffered saline
PGE2	Prostaglandin E2
PR	Progesterone receptor
pT	pathological Tumor size
ROC	Receiver operator characteristic
RT	Room temperature
TF	Transcription factor
TGFβ	Transforming growth factor-beta
TNBC	Triple-negative BC
TNFα	Tumor necrosis factors alpha
yo	years old

Acknowledgments

This research article is part of the doctoral dissertation of the first author, Lili Lin.

Ethics approval and consent to participate

This study has been approved by the Ethics Committee of the Ludwig-Maximilian-University Munich (approval number 048–08). The BC specimens were obtained in clinically indicated surgeries. All diagnostic procedures were completed when the current study was performed, and the patients' data were anonymized. The ethical principles adopted in the Declaration of Helsinki 1975 have been respected. As per declaration of our ethics committee, no written informed consent of the participants or permission to publish is needed given the circumstances described above. Researchers were blinded from patient data during experimental and statistical analysis.

Consent for publication

All authors read the manuscript and agreed to the publication of the manuscript.

Availability of data and materials

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests

Anna Hester has received a research grant from the "Walter Schulz" foundation and advisory board, speech honoraria, and travel expenses from Roche and Pfizer. Research support, advisory board, honoraria, and travel expenses from AstraZeneca, Clovis, Medac, MSD, Novartis, PharmaMar, Roche, Sensor Kinesis, Tesaro, Teva have been received by Sven Mahner. Fabian Trillsch declares research support, advisory board, honoraria, and travel expenses from AstraZeneca, Medac, PharmaMar, Roche, and Tesaro. Doris Mayr declares advisory board and investigator meeting honoraria from AstraZeneca, Novartis, and Merck. All other authors declare no conflict of interest.

Funding

This research received funding from the "Walter-Schulz-Stiftung".

Authors' contributions

LL: Participated in the design and coordination of the study, performed the statistical analysis, and performed the cell functional assays; LL wrote the manuscript; LL, KP: performed the immunohistochemistry assays and analysis; CK performed technical assistance of the immunohistochemistry assays; MR: performed technical assistance of the cell functional assays; ND: participated in the design of the study and carefully read the manuscript for important intellectual content; ES, DM: supervised immunohistochemistry as a gynecologic pathologist and participated in immunohistochemistry analysis as well as in the design and coordination of the study; LP, FT, MK, and SM revised the manuscript for important intellectual content; UJ and AH: conceived of the study and participated in its design and coordination, approved the final version of the manuscript. All authors analyzed and interpreted the data and read and approved the final manuscript.

Woolston C. Breast cancer. Nature. 2015;527(7578):S101.
Harbeck N, Gnant M. Breast cancer. Lancet. 2017;389(10074):1134–50.
Waks AG, Winer EP. Breast Cancer Treatment: A Review. JAMA. 2019;321(3):288–300.
Torre LA, Islami F, Siegel RL, Ward EM, Jemal A. Global Cancer in Women: Burden and Trends. Cancer Epidemiol Biomarkers Prev. 2017;26(4):444–57.
Coates AS, Winer EP, Goldhirsch A, Gelber RD, Gnant M, Piccart-Gebhart M, et al. Tailoring therapies–improving the management of early breast cancer: St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2015. Ann Oncol. 2015;26(8):1533–46.
Muftah AA, Aleskandarany MA, Al-Kaabi MM, Sonbul SN, Diez-Rodriguez M, Nolan CC, et al. Ki67 expression in invasive breast cancer: the use of tissue microarrays compared with whole tissue sections. Breast Cancer Res Treat. 2017;164(2):341–8.
Turner NC, Neven P, Loibl S, Andre F. Advances in the treatment of advanced oestrogen-receptor-positive breast cancer. Lancet. 2017;389(10087):2403–14.
The L. Breast cancer targeted therapy: successes and challenges. Lancet. 2017;389(10087):2350.
Harvey JM, Clark GM, Osborne CK, Allred DC. Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer. J Clin Oncol. 1999;17(5):1474–81.
Higgins MJ, Baselga J. Targeted therapies for breast cancer. J Clin Invest. 2011;121(10):3797–803.
Osborne CK, Schiff R. Mechanisms of endocrine resistance in breast cancer. Annu Rev Med. 2011;62:233–47.
Loibl S, Gianni L. HER2-positive breast cancer. Lancet. 2017;389(10087):2415–29.
Bianchini G, Balko JM, Mayer IA, Sanders ME, Gianni L. Triple-negative breast cancer: challenges and opportunities of a heterogeneous disease. Nat Rev Clin Oncol. 2016;13(11):674–90.
Kaczynski J, Cook T, Urrutia R. Sp1- and Kruppel-like transcription factors. Genome Biol. 2003;4(2):206.
Swamynathan SK. Kruppel-like factors: three fingers in control. Hum Genomics. 2010;4(4):263–70.
Zhang J, Li G, Feng L, Lu H, Wang X. Kruppel-like factors in breast cancer: Function, regulation and clinical relevance. Biomed Pharmacother. 2019;123:109778.
Jia Y, Zhou J, Luo X, Chen M, Chen Y, Wang J, et al. KLF4 overcomes tamoxifen resistance by suppressing MAPK signaling pathway and predicts good prognosis in breast cancer. Cell Signal. 2018;42:165–75.
Dong MJ, Wang LB, Jiang ZN, Jin M, Hu WX, Shen JG. The transcription factor KLF4 as an independent predictive marker for pathologic complete remission in breast cancer neoadjuvant chemotherapy: a case-control study. Onco Targets Ther. 2014;7:1963–9.
Li Z, Dong J, Zou T, Du C, Li S, Chen C, et al. Dexamethasone induces docetaxel and cisplatin resistance partially through up-regulating Kruppel-like factor 5 in triple-negative breast cancer. Oncotarget. 2017;8(7):11555–65.
Liu R, Shi P, Nie Z, Liang H, Zhou Z, Chen W, et al. Mifepristone Suppresses Basal Triple-Negative Breast Cancer Stem Cells by Down-regulating KLF5 Expression. Theranostics. 2016;6(4):533–44.
Lin Y, Liu R, Zhao P, Ye J, Zheng Z, Huang J, et al. Discovery of novel mifepristone derivatives via suppressing KLF5 expression for the treatment of triple-negative breast cancer. Eur J Med Chem. 2018;146:354–67.
Liu R, Zhi X, Zhou Z, Zhang H, Yang R, Zou T, et al. Mithramycin A suppresses basal triple-negative breast cancer cell survival partially via down-regulating Kruppel-like factor 5 transcription by Sp1. Sci Rep. 2018;8(1):1138.
Farrugia MK, Sharma SB, Lin CC, McLaughlin SL, Vanderbilt DB, Ammer AG, et al. Regulation of anti-apoptotic signaling by Kruppel-like factors 4 and 5 mediates lapatinib resistance in breast cancer. Cell Death Dis. 2015;6:e1699.
Ebert R, Meissner-Weigl J, Zeck S, Maatta J, Auriola S, Coimbra de Sousa S, et al. Probenecid as a sensitizer of bisphosphonate-mediated effects in breast cancer cells. Mol Cancer. 2014;13:265.
Reinholz MM, An MW, Johnsen SA, Subramaniam M, Suman VJ, Ingle JN, et al. Differential gene expression of TGF beta inducible early gene (TIEG), Smad7, Smad2 and Bard1 in normal and malignant breast tissue. Breast Cancer Res Treat. 2004;86(1):75–88.
Ali A, Ielciu I, Alkreathy HM, Khan AA. KLF17 attenuates estrogen receptor alpha-mediated signaling by impeding ERalpha function on chromatin and determines response to endocrine therapy. Biochim Biophys Acta. 2016;1859(7):883–95.
Ismail IA, El-Sokkary GH, Saber SH. Low doses of Paclitaxel repress breast cancer invasion through DJ-1/KLF17 signalling pathway. Clin Exp Pharmacol Physiol. 2018.
El-Sokkary GH, Ismail IA, Saber SH. Melatonin inhibits breast cancer cell invasion through modulating DJ-1/KLF17/ID-1 signaling pathway. J Cell Biochem. 2019;120(3):3945–57.
Cook T, Gebelein B, Mesa K, Mladek A, Urrutia R. Molecular cloning and characterization of TIEG2 reveals a new subfamily of transforming growth factor-beta-inducible Sp1-like zinc finger-encoding genes involved in the regulation of cell growth. J Biol Chem. 1998;273(40):25929–36.
Rakhshandehroo M, Hooiveld G, Muller M, Kersten S. Comparative analysis of gene regulation by the transcription factor PPARalpha between mouse and human. PLoS One. 2009;4(8):e6796.
Yin P, Lin Z, Reierstad S, Wu J, Ishikawa H, Marsh EE, et al. Transcription factor KLF11 integrates progesterone receptor signaling and proliferation in uterine leiomyoma cells. Cancer Res. 2010;70(4):1722–30.
Fan Y, Guo Y, Zhang J, Subramaniam M, Song CZ, Urrutia R, et al. Kruppel-like factor-11, a transcription factor involved in diabetes mellitus, suppresses endothelial cell activation via the nuclear factor-kappaB signaling pathway. Arterioscler Thromb Vasc Biol. 2012;32(12):2981–8.
Fernandez-Zapico ME, van Velkinburgh JC, Gutierrez-Aguilar R, Neve B, Froguel P, Urrutia R, et al. MODY7 gene, KLF11, is a novel p300-dependent regulator of Pdx-1 (MODY4) transcription in pancreatic islet beta cells. J Biol Chem. 2009;284(52):36482–90.
Seo S, Lomberk G, Mathison A, Buttar N, Podratz J, Calvo E, et al. Kruppel-like factor 11 differentially couples to histone acetyltransferase and histone methyltransferase chromatin remodeling pathways to transcriptionally regulate dopamine D2 receptor in neuronal cells. J Biol Chem. 2012;287(16):12723–35.
Lomberk G, Grzenda A, Mathison A, Escande C, Zhang JS, Calvo E, et al. Kruppel-like factor 11 regulates the expression of metabolic genes via an evolutionarily conserved protein interaction domain functionally disrupted in maturity onset diabetes of the young. J Biol Chem. 2013;288(24):17745–58.
Fernandez-Zapico ME, Mladek A, Ellenrieder V, Folch-Puy E, Miller L, Urrutia R. An mSin3A interaction domain links the transcriptional activity of KLF11 with its role in growth regulation. EMBO J. 2003;22(18):4748–58.
Wang Z, Spittau B, Behrendt M, Peters B, Krieglstein K. Human TIEG2/KLF11 induces oligodendroglial cell death by downregulation of Bcl-XL expression. J Neural Transm (Vienna). 2007;114(7):867–75.
Gohla G, Krieglstein K, Spittau B. Tieg3/Klf11 induces apoptosis in OLI-neu cells and enhances the TGF-beta signaling pathway by transcriptional repression of Smad7. J Cell Biochem. 2008;104(3):850–61.
Lin L, Mahner S, Jeschke U, Hester A. The Distinct Roles of Transcriptional Factor KLF11 in Normal Cell Growth Regulation and Cancer as a Mediator of TGF-beta Signaling Pathway. Int J Mol Sci. 2020;21(8).
Faryna M, Konermann C, Aulmann S, Bermejo JL, Brugger M, Diederichs S, et al. Genome-wide methylation screen in low-grade breast cancer identifies novel epigenetically altered genes as potential biomarkers for tumor diagnosis. FASEB J. 2012;26(12):4937–50.
Han M, Wang Y, Guo G, Li L, Dou D, Ge X, et al. microRNA-30d mediated breast cancer invasion, migration, and EMT by targeting KLF11 and activating STAT3 pathway. J Cell Biochem. 2018;119(10):8138–45.
Goldhirsch A, Wood WC, Coates AS, Gelber RD, Thurlimann B, Senn HJ, et al. Strategies for subtypes–dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011. Ann Oncol. 2011;22(8):1736–47.
Remmele W, Stegner HE. [Recommendation for uniform definition of an immunoreactive score (IRS) for immunohistochemical estrogen receptor detection (ER-ICA) in breast cancer tissue]. Pathologe. 1987;8(3):138–40.
Unal I. Defining an Optimal Cut-Point Value in ROC Analysis: An Alternative Approach. Comput Math Methods Med. 2017;2017:3762651.
Wolbers M, Koller MT, Witteman JC, Steyerberg EW. Prognostic models with competing risks: methods and application to coronary risk prediction. Epidemiology. 2009;20(4):555–61.
Su J, Miao LF, Ye XH, Cui MS, He XF. Development of prognostic signature and nomogram for patients with breast cancer. Medicine (Baltimore). 2019;98(11):e14617.
Li S, Zhao J, Zhu L, Su F, Chen K. Development and validation of a nomogram predicting the overall survival of stage IV breast cancer patients. Cancer Med. 2017;6(11):2586–94.
Gong Y, Ji P, Sun W, Jiang YZ, Hu X, Shao ZM. Development and Validation of Nomograms for Predicting Overall and Breast Cancer-Specific Survival in Young Women with Breast Cancer: A Population-Based Study. Transl Oncol. 2018;11(6):1334–42.
Rudloff U, Jacks LM, Goldberg JI, Wynveen CA, Brogi E, Patil S, et al. Nomogram for predicting the risk of local recurrence after breast-conserving surgery for ductal carcinoma in situ. J Clin Oncol. 2010;28(23):3762–9.
Lischka A, Doberstein N, Freitag-Wolf S, Kocak A, Gemoll T, Heselmeyer-Haddad K, et al. Genome Instability Profiles Predict Disease Outcome in a Cohort of 4,003 Patients with Breast Cancer. Clin Cancer Res. 2020;26(17):4606–15.
Cheng L, Shi L, Dai H. Bioinformatics analysis of potential prognostic biomarkers among Kruppel-like transcription Factors (KLFs) in breast cancer. Cancer Biomark. 2019;26(4):411–20.
Viola L, Londero AP, Bertozzi S, Orsaria M, Marzinotto S, Antoniazzi F, et al. Prognostic Role of Kruppel-Like Factors 5, 9, and 11 in Endometrial Endometrioid Cancer. Pathol Oncol Res. 2020;26(4):2265–72.
Song CZ, Gavriilidis G, Asano H, Stamatoyannopoulos G. Functional study of transcription factor KLF11 by targeted gene inactivation. Blood Cells Mol Dis. 2005;34(1):53–9.
Ellenrieder V, Buck A, Harth A, Jungert K, Buchholz M, Adler G, et al. KLF11 mediates a critical mechanism in TGF-beta signaling that is inactivated by Erk-MAPK in pancreatic cancer cells. Gastroenterology. 2004;127(2):607–20.
Buck A, Buchholz M, Wagner M, Adler G, Gress T, Ellenrieder V. The tumor suppressor KLF11 mediates a novel mechanism in transforming growth factor beta-induced growth inhibition that is inactivated in pancreatic cancer. Mol Cancer Res. 2006;4(11):861–72.
Massague J, Blain SW, Lo RS. TGFbeta signaling in growth control, cancer, and heritable disorders. Cell. 2000;103(2):295–309.
Derynck R, Zhang YE. Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature. 2003;425(6958):577–84.
Spittau B, Krieglstein K. Klf10 and Klf11 as mediators of TGF-beta superfamily signaling. Cell Tissue Res. 2012;347(1):65–72.
Cook T, Gebelein B, Belal M, Mesa K, Urrutia R. Three conserved transcriptional repressor domains are a defining feature of the TIEG subfamily of Sp1-like zinc finger proteins. J Biol Chem. 1999;274(41):29500–4.
Zhang JS, Moncrieffe MC, Kaczynski J, Ellenrieder V, Prendergast FG, Urrutia R. A conserved alpha-helical motif mediates the interaction of Sp1-like transcriptional repressors with the co-repressor mSin3A. Mol Cell Biol. 2001;21(15):5041–9.
Hayashi H, Abdollah S, Qiu Y, Cai J, Xu YY, Grinnell BW, et al. The MAD-related protein Smad7 associates with the TGFbeta receptor and functions as an antagonist of TGFbeta signaling. Cell. 1997;89(7):1165–73.
Nakao A, Afrakhte M, Moren A, Nakayama T, Christian JL, Heuchel R, et al. Identification of Smad7, a TGFbeta-inducible antagonist of TGF-beta signalling. Nature. 1997;389(6651):631–5.
de Caestecker MP, Piek E, Roberts AB. Role of transforming growth factor-beta signaling in cancer. J Natl Cancer Inst. 2000;92(17):1388–402.
Dumont N, Arteaga CL. Targeting the TGF beta signaling network in human neoplasia. Cancer Cell. 2003;3(6):531–6.
Buttar NS, DeMars CJ, Lomberk G, Rizvi S, Bonilla-Velez J, Achra S, et al. Distinct role of Kruppel-like factor 11 in the regulation of prostaglandin E2 biosynthesis. J Biol Chem. 2010;285(15):11433–44.

Competing interest reported. Anna Hester has received a research grant from the "Walter Schulz" foundation and advisory board, speech honoraria, and travel expenses from Roche and Pfizer. Research support, advi-sory board, honoraria, and travel expenses from AstraZeneca, Clovis, Medac, MSD, Novartis, PharmaMar, Roche, Sensor Kinesis, Tesaro, Teva have been received by Sven Mahner. Fabi-an Trillsch declares research support, advisory board, honoraria, and travel expenses from AstraZeneca, Medac, PharmaMar, Roche, and Tesaro. Doris Mayr declares advisory board and investigator meeting honoraria from AstraZeneca, Novartis, and Merck. All other authors declare no conflict of interest.

KLF11forBreastcancerresearchAdditionalfiles.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Transcriptional Factor KLF11 is a new prognostic biomarker and potential therapeutic target in human sporadic invasive breast cancer

Status:

Version 1

Abstract

Figures

Background

Methods

BC Patients Cohort

Immunohistochemistry

Evaluation of KLF11 Expression

Forest plots for Subgroup analyses

Nomogram Construction and Calibration

BC Cell Lines Culture

siRNA-Mediated Knockdown

RNA Extraction and Real-Time PCR

Colorimetric Cell-Based KLF11 ELISA

Cell Viability Assay

Cell Proliferation Assay

Cell Apoptosis Assay

Statistical Analysis

Results

The Quantification and Distribution of KLF11 Protein Expression in BC

Prognostic Significance of KLF11 for OS, DFS, DMFS, LRFS in BC

KLF11 was an independent Indicator for DFS and DMFS of BC

Nomograms Construction and Calibration for the Established KLF11 Related DFS and DMFS Cox Model

siRNA-mediated loss-of-function of KLF11 in BC cell lines

Discussion

Conclusions

List of abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1