Aurora kinase A is a target of Wnt/β-catenin involved in multiple myeloma disease progression

Jui Dutta-Simmons, Yunyu Zhang, Gullu Gorgun, Moshe Gatt, Mala Mani, Teru Hideshima, Kohichi Takada, Nicole E. Carlson, Daniel E. Carrasco, Yu-Tzu Tai, Noopur Raje, Anthony G. Letai, Kenneth C. Anderson and Daniel R. Carrasco

Data supplements

  • Supplemental materials for: Dutta-Simmons et al

    Files in this Data Supplement:

    • Table S1. Correlation between β-catenin and AurKA expression in human patient TMAs (PDF, 38.3 KB) -
      Serial consecutive sections were stained and analyzed for two independent observations. Expression was graded from 0 to 5, and samples showing expression over total number of samples are plotted.

    • Figure S1. Different β-catenin shRNAs specifically knockdown β-catenin and affect function in MM cells (JPG, 70.4 KB) -
      (A) Immunoblotting in MM1.S cells show specific ability of shRNAs #1 & #2 to knockdown β-catenin. (B) Proliferation assays showed both β-catenin specific shRNAs inhibit proliferation of MM1.S cells.

    • Figure S2. Apoptosis assays (JPG, 83.2 KB) -
      (A) Representative figures of sorted MM1.S or OPM1 cells, stably transduced with control or β-catenin shRNA, and assayed for cell death by Annexin V/7-AAD staining (left) and Caspase 3 cleavage (right) 8 days after infection. (B) Representative apoptosis assay by Annexin V/7-AAD staining (left) and Caspase 3 cleavage (right) in stably selected MM1.S control or AurKA shRNA cells 6 days after infection.

    • Figure S3. Effect of Wnt3A stimulation and AurKA knockdown in OPM1 cells is similar to MM1.S cells (JPG, 255 KB) -
      (A) Wnt3A CM stimulation of OPM1 cells show increased β-catenin and AurKA levels and enhanced proliferation (B). (C) Cell cycle profiling of control or AurKA shRNA OPM1 cells show increased G2/M phase. (D) AurKA knockdown in OPM1 decreased proliferation.

    • Figure S4. Decreased β-catenin in MM1.S cells induces an apoptotic signature shift (JPG, 123 KB) -
      BH3 profiling (100µM peptides) of MM1.S control or β-catenin shRNA cells demonstrates distinct apoptotic signatures. Control cells showed a BCL2 profile which shifted to an overlay of BCL2, MCL1 and BCLxL profiles in β-catenin shRNA cells. Data is representative of the mean of two independent experiments.

    • Figure S5. H&E stains of a representative section of a manually prepared bone marrow TMA (JPG, 31.6 KB) -
      Bone marrow core needle biopsies done on normal individuals and patients with MGUS or MM at the time of diagnosis were arrayed on arrayed on the same paraffin block and serially sectioned at 5 µm thick.

      Archival paraffin embedded tissue blocks from bone marrow core needle biopsies done on normal individuals and patients with MGUS or MM at the time of diagnosis were collected from the Pathology Department at Brigham and Women�s Hospital. All biopsies were performed during the year 2006 and processed under the same standard conditions which include: i) 16 hours fixation and decalcification in Zenker�s solution, ii) serial dehydratation in ethanol solutions and, ii) embedding into paraffin blocks. After clinical use, the tissue blocks were deparaffinized by heating at 60°C and the remaining of tissue core biopsies were re-embedded and arrayed on a single paraffin block. To accommodate a relatively large number of samples in the same block the core biopsies were oriented perpendicularly to the main axis of the paraffin block. The core biopsies were arrayed in three different rows: i) first row includes samples from normal individuals, ii) second row include samples from patients with MGUS and, iii) third row include sample from patients with MM. Paraffin blocks were serially sectioned and 5 µm thick consecutive tissue sections were mounted onto glass slides and stored at 4°C prior to use for IHC analysis (a representative H&E stained section from a BM-TMA is shown in Fig. S5).

      Immunohistochemical stains were performed as described in material and methods. Optimal working conditions including antigen retrieval and concentrations for anti β-catenin and anti AurKA antibodies were determined individually on separate sets of bone marrow biopsies. As positive controls we used breast and colon cancer tissue sections. As negative controls we incubated BM-TMA sections without the primary Ab but with the secondary Ab. When the appropriate working conditions were established, consecutive BM-TMA serial sections were stained, one with anti β-catenin Ab and the other with anti AurKA Ab. To avoid differences in protein expression between samples in the TMA that might have been related to variation in sample processing all samples were processed in parallel. The IHC slides were reviewed blindly (without knowing the identity of the samples) and independently by two Board-certified hematopathologists. An arbitrary visual scoring procedure of common use for trained pathologists was utilized to evaluate the results. The score is based on the intensity and/or size of the signal (brown nuclear color staining for AurKA Ab and brown perinuclear color staining for β-catenin Ab) as observed under bright light microscope that ranges from 0+ to 5+. A score of 0+ to 1+ was given for weak signal such as seen in normal BM plasma cells and a score of 5+ was given for stronger signals such as seen in myeloma cells (see representative case in Fig. 6A).

      Statistical analysis was done as follows, assessing the significance of the correlation coefficient determined for each sample set:

      1) NBM: not significant
      Pearson�s product-moment correlation data: d1 and d2
      t = 0.8823, df = 8, p-value = 0.4034
      alternative hypothesis: true correlation is not equal to 0 sample estimates: cor 0.297775

      Pearson�s product-moment correlation data: d1 and d2
      t = 2.22, df = 4, p-value = 0.0906
      alternative hypothesis: true correlation is not equal to 0 sample estimates: cor 0.742967

      3) MM
      Pearson�s product-moment correlation data: d1 and d2
      t = 8.901, df = 18, p-value = 5.191e-08
      alternative hypothesis: true correlation is not equal to 0 sample estimates: cor 0.9027002

Article Figures & Data


  • Figure 1

    β-catenin is aberrantly expressed in MM, and its down-regulation by the use of shRNA knockdown increases sensitivity to chemotherapeutic agents. (A) Affymetrix analysis of β-catenin mRNA expression in normal plasma cells (PC) and multiple myeloma primary tumors (MMPT). (B) Immunoblot of β-catenin protein in PC and MMPT (1-6) cells. Immunoglobulin heavy chain (IgH) and a nonspecific protein were used as loading markers. (C) Nuclear (N) and cytoplasmic (C) protein fractions of MM1.S and OPM1 cells. Lamin B was used as nuclear fraction loading control. (D) Immunofluorescent staining of total β-catenin (green) in MM1.S (60×) and OPM1 (100×) cells. DAPI is shown in blue. (E) Immunoblot of β-catenin knockdown by stable lentiviral shRNA transduction in MM1.S and OPM1 cells. GFP was used as a whole-cell lysate transduction efficiency marker, whereas Actin and Lamin B were used as whole-cell lysate loading and nuclear fraction loading controls markers, respectively. Note reduced β-catenin levels in both cytoplasmic (C) and nuclear fractions (N). (F) MM1.S control or β-catenin shRNA cells were cocultured with BM stromal cells and treated with different drugs followed by MTT assays to measure metabolism.

  • Figure 2

    β-catenin directly affects proliferation and cell cycle in MM. (A) Wnt3A stimulation of MM1.S (top) and OPM1 (bottom) cells increases proliferation. Cells were treated with control (Con CM) or Wnt3A-conditioned medium (Wnt3A CM) for 24 or 48 hours and assayed for 3H-thymidine incorporation. (B) β-catenin (β-Cat shRNA) knockdown decreased proliferation of OPM1 and MM1.S cells. (C) Partial rescue in proliferation of OPM1 or MM1.S β-catenin shRNA (β-Cat shRNA) cells after 24 hours treatment with Wnt3A CM. In panels B and C, cells had been stably transduced and sorted for GFP-positive cells within 4 days after infection, followed by a growth curve (B) or treatment with control or Wnt3A CM for 24 hours, followed by 3H-Thy proliferation assay (C). The difference in proliferation starts out modestly but is more apparent over a period of 6 days, as is shown in panel B, whereas in panel C the experiment was performed with the use of cells technically at the day 2 time point of panel B. (D) A representative immunoblot shows a slight-but-consistent increase in β-catenin after 24-hour Wnt3A CM treatment in β-catenin shRNA OPM1 cells. (E) β-catenin (β-Cat shRNA) knockdown inhibits proliferation in MM1.S, RPM18226, and MMS1 cells. (F) Immunoblot analysis confirms the reduction of β-catenin in the different MM cells stably transduced with β-catenin shRNA (β-cat). Proliferation assays were performed in triplicate and repeated at least twice. The average and SEM of 2 to 3 experiments are shown. (G) Primary MM patient samples with β-catenin knockdown (β-Cat shRNA) also show decreased proliferation. Proliferation assays were performed in triplicate. The error bars and statistical significance were calculated from the triplicate dataset. (H) Increase in G1 and G2/M cell-cycle phases and decrease in S phase in β-catenin shRNA MM1.S cells compared with control cells. The FACS data represent the average of triplicate analyses repeated twice.

  • Figure 3

    β-catenin knockdown reveals novel cell-cycle regulators as targets in MM. (A) Decreased Wnt reporter activity in β-catenin shRNA cells. MM1.S and OPM1 cells were transfected with TOPFLASH and assayed for luciferase activity after 48 hours. Control FOPFLASH levels were unaffected. Experiments were performed in triplicate and repeated twice. The results denote the average and SEM of triplicate assays. (B) GEP revealed 2 distinct sets of genes that are most up- or down-regulated by β-catenin knockdown compared with control shRNA MM1.S cells. (C) Microarray target validation by immunoblot show decreased expression of known and novel targets after β-catenin knockdown in MM1.S (left) and OPM1 cells (right). (D) GEP of β-catenin shRNA down-regulated genes containing LEF1/TCF4 binding sites (GCTTTGT/A). Only probe sets expressed in all 3 samples are shown. (E) Immunofluorescence of control or β-catenin shRNA MM1.S cells (100×) for AurKA (red) and DAPI (blue). (F) Diagram of the AurKA gene structure (adapted from GenBank, accession no. AL121914) and AurKA-luciferase reporter pGL1486 (adapted from Tanaka et al24). Red lines depict potential TCF4 binding sites, untranslated regions (open box), solid box (open reading frame). (G) Cells transfected with pGL1486 and increasing β-catenin amounts assayed similar to panel A show increased AurKA-luciferase activity. (H) Chromatin IP (ChIP) assays in MCF-7 cells demonstrated direct binding of β-catenin transcriptional complex to AurKA gene promoter sequences. The bar graph represents ChIP/immunoglobulin G (IgG) signal normalized to DNA concentration. A nonspecific primer (Neg1) and water were used as negative controls; known primers against the TBE1 region of the c-Myc promoter (c-Myc) was used as a positive control; AurKA primer pairs A and B were designed around regions of putative TCF4 binding sites of the AurKA promoter, described in panel F (bottom). The average fold difference in relative enrichment (ChIP/IgG) over the negative control is represented.

  • Figure 4

    Wnt/β-catenin-mediated effects on proliferation are affected by Aurora kinase A levels. (A) Wnt3A CM stimulation of MM1.S cells shows increased β-catenin and AurKA levels. MM1.S cells stably overexpressing β-catenin show increased AurKA levels (B) and enhanced proliferation (C). (D) Immunoblot of stable AurKA shRNA knockdown in MM1.S cells. (E) Cell-cycle profiling of control or AurKA shRNA MM1.S cells show increased G2/M phase. (F) AurKA knockdown in MM1.S decreased proliferation. (G) AurKA shRNA counteracts the positive proliferative effect of β-catenin overexpression or Wnt3A CM stimulation (H). Control (Control GFP) or β-catenin overexpressing cells (β-catenin GFP) were stably transduced with control (■) or AurKA shRNA ([graphic024]) and assayed for proliferation. Proliferation experiments were done in triplicate and repeated twice. The results denote the average and SEM of triplicate assays.

  • Figure 5

    β-catenin knockdown improves survival in a xenograft mouse model of MM. β-catenin knockdown improves survival in Kaplan-Meier survival curves (A) by decreasing tumor load and metastasis (B). Mice injected with control or β-catenin shRNA-GFP MM1.S cells were analyzed by whole-body imaging. Note decreased tumor GFP nodule number in the spine and liver of β-catenin shRNA mice compared with control mice (white arrows). Histologic and IHC analysis of tumors showed decreased tumor metastasis in the liver (C) and kidney (D), increased numbers of tingible body macrophages (E), and increased TUNEL staining (F), as well as decreased expression of β-catenin (G) and AurKA (H) in engrafted β-catenin shRNA MM1.S cells compared with control MM1.S xenografts. Hematoxylin & eosin stains (panels C, D, and E); IHC stains (panels G and H).

  • Figure 6

    β-catenin affects expression of target genes in patient sample tissue microarrays. (A) IHC analysis on patient sample tissue microarrays show that AurKA and CDC25B reflect increased β-catenin expression from MGUS to MM compared with normal BM. Costaining of β-catenin (red) and AurKA (brown) in top right panel inset. (B) Immunoblot analysis of AurKA expression normal plasma cells (PC) and MM primary tumors compared with β-catenin expression. Note a correlation between β-catenin and AurKA expression in patient samples 1, 2, 5, and 6.