ANKRD44 Gene Silencing: A Putative Role in Trastuzumab Resistance in Her2-Like Breast Cancer

Abstract

Trastuzumab is an effective therapeutic treatment for Her2-like breast cancer; despite this most of these tumors develop resistance to therapy due to specific gene mutations or alterations in gene expression. Understanding the mechanisms of resistance to Trastuzumab could be a useful tool in order to identify combinations of drugs that elude resistance and allow a better response for the treated patients. Twelve primary biopsies of Her2+/hormone receptor negative (ER-/PgR-) breast cancer patients were selected based on the specific response to neoadjuvant therapy with Trastuzumab and their whole exome was sequenced leading to the identification of 18 informative gene mutations that discriminate patients selectively based on response to treatment. Among these genes, we focused on the study of the ANKRD44 gene to understand its role in the mechanism of resistance to Trastuzumab. The ANKRD44 gene was silenced in Her2-like breast cancer cell line (BT474), obtaining a partially Trastuzumab-resistant breast cancer cell line that constitutively activates the NF-kb protein via the TAK1/AKT pathway. Following this activation an increase in the level of glycolysis in resistant cells is promoted, also confirmed by the up-regulation of the LDHB protein and by an increased TROP2 protein expression, found generally associated with aggressive tumors. These results allow us to consider the ANKRD44 gene as a potential gene involved in Trastuzumab resistance.

Keywords: ANKRD44; Her2+ breast cancer; LC-MS/MS; Trastuzumab resistance; gene silencing; next generation sequencing.

Figure 1

Whole exome analysis. (A) The dendrogram derived from cluster analysis classified 12 Her2 breast cancer patients based on the mutational profile of 18 informative genes. (B) Histogram representing the ANKRD44 gene relative mutational frequency between full (FR) and partial (PR) responder patients. There is a 2-fold higher mutation frequency in the PR than in the FR samples. Values are normalized on the total amount of mutations found per group (FR and PR).

Figure 2

ANKRD44 silencing confers partial trastuzumab resistance and a more aggressive phenotype: (A) ANKRD44 mRNA expression detected by quantitative real-time RT-PCR analysis in BT474 cells with control empty vector (shCTRL) and with ANKRD44 short interfering RNA (shANK). (A) Gene expression values were normalized on housekeeping GAPDH mRNA. (B) Western Blot of proteins extracted from shANK and shCTRL cells. β-Tubulin was used as reference. The band intensity of ANKRD44 was normalized against the respective β-tubulin, and the ratios shown are against the shCTRL. Data were obtained with ImageLab™ 4.1 Software (Bio-Rad). (C) Viable cell count assessed by WST1 assay. Values were normalized on untreated cells. (D) Colony formation assay histogram. shANK cells present a 10% increase in the colony formation number respect to the shCTRL cells. Colonies larger than 100 μm in diameter were counted from five randomly selected fields. (E) Cell cycle distribution of shCTRL and shANK cells performed with propidium iodide staining in both untreated (–) and treated (+). Bars present the percent of cells in G1/G0-phase (black), in S phase (dark gray) and in G2/M phase (gray). (F) Two examples of plurinucleated cells with anisonucleosis obtained by cytochemistry analysis in shANK cells (Papanicolaou stain × 200). Scale bars: 25 μm. (G) Plurinucleate cells distribution of shANK (right) and shCTRL (left) cells fixed and labeled with propidium iodide in a FL3/SSC dot plot. Quadrant gate regions identify cells with higher scatter intensity (high SSC) and higher concentration of PI, index of cells with big dimensions, more intracellular complexity and nuclei dysmetria. Percentage of gated cells has been reported in a histogram and shANK cells present a significant (p < 0.01) increase of multinucleated cells respect to shCTRL cells. Each experiment was repeated in triplicate (D,E) and octuplicate (C) and each bar represents the mean ± SD of 3 independent experiments. Statistical significance was examined using Student t-test. (^*p < 0.05; ^**p < 0.01; ^***p < 0.001).

Figure 3

ANKRD44 silencing increases NF-kβ (p65) activity through TAK1/Akt-pathway. BT474 cells with empty vector (shCTRL) and with ANKRD44 shRNA (shANK). Cells were treated (+) and untreated (–) with trastuzumab 21 μg/ml for 48 h. (A) ANKRD44, Akt and p-Akt-Ser473 western blots. (B) Western blot of p-TAK1 (Thr 184/187). Proteins band intensity was normalized against the respective β-tubulin using ImageLab™ 4.1 Software (Bio-Rad) and normalized on untreated shCTRL cells. (C) NF-kβ (p65) activity of shANK cells and shCTRL cells. Values represent absorbance measurement (450 nm) with an Infinite 200 PRO NanoQuant plate reader (Tecan). Values were then normalized on the respective amount of loaded protein. Columns, mean of 3 determinations. Statistical significance was examined using Student t-test. (^*p < 0.05).

Figure 4

ANKRD44 silencing confers a different proteomic pattern in shANK cells respect to the control. Volcano plot showing the protein expression comparison between shCTRL and shANK cells before (A) and after (B) treatment with Trastuzumab. Gray dots represent protein not differently expressed (double sided t-test with FDR < 0.05). Green and red dots represent proteins significantly up-regulated and down-regulated in the shANK cells, respectively. (C) Log2 of the ERBB2 expression levels for the shCTRL and shANK before (–) and after (+) treatment with Trastuzumab. (D,E) LDHB and TACSTD2 mRNA expression detected by droplet digital PCR (ddPCR) analysis in BT474 cells with control empty vector (shCTRL) and with ANKRD44 short interfering RNA (shANK) both treated and untreated for 48 h with Trastuzumab 21 μg/ml. Gene expression values were normalized on housekeeping GAPDH mRNA and the values reported (copies per ml) represent the mean of the triplicate per sample. Bars: SD. Statistical significance was examined using Student t-test. (^*p < 0.05; ^**p < 0.01).

Figure 5

ANKRD44 silencing turn shANK (BT474) cells metabolism to a more intense glycolysis. BT474 cells with empty vector (shCTRL) and with ANKRD44 shRNA (shANK). Cells were treated (+) and untreated (–) with trastuzumab 21 μg/ml for 48 h. (A) Representative results of the metabolic potential of shCTRL and shANK cells measured using a Seahorse XF. Data shows OCR and ECAR, reflective of energy production by mitochondrial respiration and glycolysis, respectively, following sequential additions of oligomycin and FCCP. (B) Reactive oxygen species (ROS) production by shCTRL and shANK cells. Values represent green fluorescence measurement (520 nm) normalized on respective untreated cells. Each experiment was repeated in triplicate and each bar represents the mean ± SD of 3 independent experiments. Bars: SD. Statistical significance was examined using Student t-test. ^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

Figure 6

Snapshot of trastuzumab resistance pathway. Model of trastuzumab resistance acquisition in BT474 Her2+ breast cancer cell line due to ANKRD44 silencing. (A) In sensitive BT474 cells (shCTRL) without ANKRD44 silencing, trastuzumab blocks the downstream PI3K/Akt pathway (black dotted arrows) and prevents cell growth. (B) Thanks to ANKRD44 silencing BT474 cells (shANK) become partially resistant to the treatment due to an up-regulation of TAK1/Akt proteins which constitutively activate NF-kβ p65 subunit and leads to its nuclear accumulation (blue arrows). The nuclear p65 subunit activates the expression of target pro-inflammatory genes, survival genes and cell cycle regulators. Moreover, the TAK/Akt pathway activates mTORC1 complex which mediates cell proliferation and simultaneously increase glycolysis rate through an overexpression of Lactate dehydrogenase Beta protein (LDHB).