A novel isoform of the B cell tyrosine kinase BTK protects breast cancer cells from apoptosis

Abstract

Tyrosine kinases orchestrate key cellular signaling pathways and their dysregulation is often associated with cellular transformation. Several recent cases in which inhibitors of tyrosine kinases have been successfully used as anticancer agents have underscored the importance of this class of proteins in the development of targeted cancer therapies. We have carried out a large-scale loss-of-function analysis of the human tyrosine kinases using RNA interference to identify novel survival factors for breast cancer cells. In addition to kinases with known roles in breast and other cancers, we identified several kinases that were previously unknown to be required for breast cancer cell survival. The most surprising of these was the cytosolic, nonreceptor tyrosine kinase, Bruton's tyrosine kinase (BTK), which has been extensively studied in B cell development. Down regulation of this protein with RNAi or inhibition with pharmacological inhibitors causes apoptosis; overexpression inhibits apoptosis induced by Doxorubicin in breast cancer cells. Our results surprisingly show that BTK is expressed in several breast cancer cell lines and tumors. The predominant form of BTK found in tumor cells is transcribed from an alternative promoter and results in a protein with an amino-terminal extension. This alternate form of BTK is expressed at significantly higher levels in tumorigenic breast cells than in normal breast cells. Since this protein is a survival factor for these cells, it represents both a potential marker and novel therapeutic target for breast cancer.

Figure 1

An RNAi screen targeting tyrosine kinase genes in an ERBB2 positive breast cancer. A: BT474 breast cancer cells were transfected with 234 shRNA constructs targeting 83 protein tyrosine kinase genes. Three transfection mixes were produced for each shRNA and each was transfected into triplicate wells of BT474 cells for 96 hr. AlamarBlue was used to monitor cell proliferation and viability. The averages of the nine parallel cultures were calculated for each shRNA, normalized to transfection efficiency, presented as the % of control shRNA and sorted on the basis of effect. shRNAs that produced a greater than 50% decrease in AlamarBlue signal are shown in red and are presented in Table 1. B: siRNA knockdown of BTK in BT474 cells (48 hr) results in increased cleaved caspase-3 (Caspase-3) staining indicative of apoptosis. C: Degree of apoptosis due to BTK knockdown in BT474 and MCF-7 cells was calculated as a percentage of the total cellular population. The data were expressed as the mean of triplicate of the samples transfected with the BTK siRNA relative to scrambled siRNA control samples. Error bars represent standard deviation from the average of three replicates. Statistical significance between samples was calculated using the student’s t test, where (*) indicates a P value of <0.0001. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 2

An alternative form of the BTK transcript is present in BT474 breast cancer cells. A: Nucleotide sequence 1 – 395 bp from the published BTK sequence (accession # U13399) was aligned to the nucleotide sequence obtained from BT474 cells using 5′RACE. Identical sequence is highlighted in grey. The BT474 sequence obtained using 5′RACE encodes an additional 34 amino acid open reading frame (ORF) and contains two additional methionine codons, highlighted in grey, that are in frame with the methionine start codon from the published BTK gene (highlighted in grey and with an arrow). B: Schematic representation showing alternate splicing of alternative first exons from both isoforms inferred from sequence analysis. Sequences are identical from exons 2 through 19. C: Map of BTK on the X-chromosome. BTK-C exon 1 transcription initiates divergently 255 bp from the start site of the ribosomal protein L36a gene.

Figure 3

The BTK-C gene produces an 80 kD product. A: A schematic representation showing the domains of the BTK-A and predicted BTK-C protein. B: Total lysate from breast lines and Namalwa B-cells subjected to immunoblotting and probed with an anti-BTK antibody (BD Transduction Laboratory, 611116). C: HEK293 cells cotransfected with a BTK-C flag vector and either BTK-C siRNAs or Non-Target siRNA. Total lysate was prepared 96 hr post transfection and was used for immunoblotting with anti-Flag antibody. D: BT474 cells were transfected with two BTK-C specific siRNAs, non-target siRNA as a control and cotransfected with GFP to mark transfected cells. Transfected cells were counted at 24 and 96 hr and the 96 to 24 hr ratio was calculated and expressed as % of the control.

Figure 4

BTK-C is activated in BT474 cells. In BT474 cells both forms of the over-expressed BTK-C proteins are phosphorylated on tyrosine residue 223, which becomes autophosphorylated after activation. BT474 cells containing the stably integrated BTK-A-flag, the BTK-C-flag or control flag vector were treated with 100 μM LFM-A13 for 45 min. Tyrosine-phosphorylated BTK was assessed by immunoprecipitation (IP) using anti-Flag (Stratagene) and immunoblot analysis using anti-BTK Phospho (pY223) and anti-BTK antibody (BTK-E9 Santa Cruz). B: Inhibition of BTK autophosphorylation using LFM-A13 results in increased apoptosis. BT474 cells incubated with 35 μM LFM-A13 for 48 hr results in increased cleaved caspase-3 (Caspase-3) compared with control cells treated with DMSO. Apoptotic cells were calculated as a percentage of the total cellular population as in Figure 1C. C: BTK-C inhibits apoptosis induced by doxorubicin in MCF-10A cells. BTK-C expression in vector control (10A-Vec) and MCF-10A-BTK-C (10A-Btk-C) cells using anti-Flag antibody. GAPDH is used as a loading control. D: BTK-C expression reduces Doxorubicin-induced apoptosis as monitored by Cleaved caspase-3 signal. 10A-Vec or 10A-BTK-C cells were either treated with DMSO (Con) or with 35 μM LFM-A13 for 24 hr, after that the cells were washed with PBS for three times and added fresh medium with doxorubicin(1 μM) for 24 hr. Immunofluorescence was performed for cleaved caspased-3 signal; cell nuclei were stained with Hoechst 33342. E: Apoptotic cells were calculated as a percentage of the total cellular population, as indicated B. Error bars indicate the standard deviation from three individual experiments, *P < 0.01. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 5

BTK is more abundant in breast cancer cells compared to nontumorigenic breast cells. A: BTK protein levels were examined in normal, matched breast tissues and breast carcinoma tissue in tissue microarrays using immunofluorescence microscopy. DAPI staining of nuclei is shown in cyan false color; anti-BTK (ProSci) staining is red. Tissue samples and BTK classifications were (i) normal-low level; (ii) benign hyperplasia-low level; (iii) cancer-low-moderate/heterogenous; (iv) cancer-heterogenous with strong positives; (v) cancer-homogenous moderate with nuclear; (vi) cancer-negative. B: BTK-C message is more abundant than the BTK-A isoform in cancer cell lines. qPCR primers designed to target specifically the BTK-C message and cDNA from the breast cancer cell lines BT474, MCF7, MDA-MB-361, and two nontumorigenic breast cell lines, HMEC and MCF10a, were amplified using SYBR Green. Fold change was calculated using the delta, delta Ct method. Error bars represent standard deviation from the average of four replicates. Statistical significance between samples was calculated using the student’s t test, where (*) indicates a P value of < 0.005 and (**) indicates a P value of <0.0005. C: BTK-C message is more abundant than the BTK-A isoform in breast tumors. cDNA prepared from RNA isolated from human breast tissue was subjected to qPCR using primers specific for BTK-A and BTK-C isoforms. The same set of samples in another plate was used for detection of actin mRNA. The data represent relative mRNA levels of each BTK isoform normalized to actin. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 6

BTK-C promotes glucose uptake. A: LFM-A13 inhibits glucose uptake in breast cancer cell lines. MCF-7, MDA-MB-361, and BT474 cell lines. Cells were either treated with DMSO (Control) or with 35 μM LFM-A13 for 24 hr, at which point cells were washed with PBS three times and 100 μm 2-NBDG added for 15 min. Immunofluorescence images were acquired with an INCELL-1000; (B) Fluorescence intensity quantification. Error bars indicate standard deviation from three individual experiments, *P < 0.01. C: The effect of BTK-C on glucose uptake in breast cells. 10A-Vec and 10A-BTK-C cells cell lines were treated as in A. Fluorescence intensity is shown in (D), *P < 0.01. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]