The isolation and characterization of renal cancer initiating cells from human Wilms' tumour xenografts unveils new therapeutic targets†

There are considerable differences in tumour biology between adult and paediatric cancers. The existence of cancer initiating cells/cancer stem cells (CIC/CSC) in paediatric solid tumours is currently unclear. Here, we show the successful propagation of primary human Wilms' tumour (WT), a common paediatric renal malignancy, in immunodeficient mice, demonstrating the presence of a population of highly proliferative CIC/CSCs capable of serial xenograft initiation. Cell sorting and limiting dilution transplantation analysis of xenograft cells identified WT CSCs that harbour a primitive undifferentiated – NCAM1 expressing – “blastema” phenotype, including a capacity to expand and differentiate into the mature renal-like cell types observed in the primary tumour. WT CSCs, which can be further enriched by aldehyde dehydrogenase activity, overexpressed renal stemness and genes linked to poor patient prognosis, showed preferential protein expression of phosphorylated PKB/Akt and strong reduction of the miR-200 family. Complete eradication of WT in multiple xenograft models was achieved with a human NCAM antibody drug conjugate. The existence of CIC/CSCs in WT provides new therapeutic targets.


Movies S1 and S2: NCAM + cells show enhance in vitro motility compared to NCAMprimary WT cells.
Representative time lap-video microscopy movies of primary cultured, FACS-sorted Wilms' tumor NCAM + cells and NCAMcells showing enhance in vitro motility of the former. Supplemental figure S3B shows analysis of motility characteristics of 58 cells from 8 movies of NCAM + and NCAMcells.

Quantitative reverse transcription PCR analysis.
Quantitative reverse transcription PCR (qRT-PCR), to determine expression of renal

Immunohistochemical staining of primary WT and WT Xn.
Sections, 4-µm thick, were cut from primary WT and WT Xn for immunohistochemistry. Immunostainings were performed as previously described (Dekel et al., 2006b). Brifly, the sections were processed within 1 week to avoid oxidation of antigens. Before immunostaining, sections were treated with 10mM citrate buffer, PH 6.0 for 10 min at 97ºC in a microwave oven for antigen retrieval, followed by 3% H 2 O 2 for 10 min. The slides were subsequently stained using the labeled strepavidin-biotin (LAB-SA) method using a Histostain plus kit (Zymed, San Francisco, CA, USA). Anti human NCAM antibody (LifeSpan Biosciences, Inc. Seattle, WA, USA), anti human WT1 antibody and anti human Ki67 antibody, at a dilution of 1:50 were used. Controls were prepared by omitting the primary antibodies or by substituting goat IgG isotype for the primary antibodies. The immunoreaction was visualized by an HRP-based chromogen/substrate system (liquid DAB substrate kit -Zymed, San Francisco, CA, USA).

Ki67 immunostaining Analysis (Quantification).
Quantification of Ki67+ cells in blastema and non-blastemal components of two primary WTs (W013, W014) and their corresponding WT Xn (W013 Xn, W014 Xn) was performed as previously described (Ghanem, Van der Kwast et al, 2004). The slides were evaluated by two independent observers, using a standard light microscope with a ×60 objective and equipped with an ocular grid. Cells were considered positive regardless of the intensity or location of nuclear staining.
Quantification was performed by counting at least 1000 tumor cells in five randomly selected fields of view.
Cells were collected and homogenized in buffer containing 20 mM Tris-HCl(pH 7.5), 1 mM EDTA, 1 mM DTT, and 250 mM sucrose. The extract was centrifuged at 13000xg for 15 min. The supernatant subjected to ultracentrifugation for 16 hours at 100,000xg. The pellet was resuspended and loaded on a nondenaturing polyacrylamide gel using the protocol previously described (Tsvetkov et al., 2009).

Immunoblot analysis.
The Signals were detected using the Ez-ECL kit (Biological Industries, Israel).

Colony forming assay.
Cells were routinely cultured in IMEM medium supplemented with 10% FBS ("growth medium"). For assessment of colony forming ability (CFU), primary WT cells before and after treatment with lorvotuzumab mertansine, or NCAM + ALDH1 + and ALDH1sorted cells were plated in growth medium on matrigel-coated either 6 or 24 well plates at 1000 or 5000 cells/well in triplicate, respectively. The medium was changed twice a week. After two weeks both the number of colonies and the number of cells/colonies were determined, and means calculated.

Panorama® Antibody Array (Sigma-Aldrich, St. Louis).
Two frozen WT xenograft tissues from different sources were tested while AK and FK were used as comparative controls. The experiment was conducted according to the manufacturer's instructions: the tissues were homogenized on ice using a homogenizer and the proteins concentration were determined by the Bradford protein assay (Thermo scientific). Equal amounts of protein extracts (>1 mg/ml) were labeled using Cy3 monoreactive reactive dyes (Amersham Biosciences) as described

Treatment of WT cells with combination chemotherapy.
In order to determine the in-vitro IC50 of WT cells for each combination of the were found to be 0.24µM and 62µM, respectively. All further experiments evaluating the effects of these drugs on WT cells were performed at these concentrations.

In vitro effects of chemotherapeutic drugs on WT initiating cells.
In order to examine the effect of the chemotherapy regimens on the percentage of cells expressing NCAM and ALDH1, WT cells (from each source) were plated in 3x75T flasks for 72h. Following the indicated time, medium was removed and replaced by medium containing the first line drug combination for the first flask, the second line drug combination for the second flask and no drugs for the third flask.
The untreated flask was used as the baseline for NCAM and ALDH1 expression in each tumor examined. After treatment, cells were incubated for 48h, the medium was removed, cells were harvested using 0.05% Trypsin/EDTA, counted and analyzed by FACS for the percentage of cells expressing NCAM and ALDH1 as described above.
To study the in vitro effect of the two chemotherapeutic combinations on the CIC fraction we assessed the CFU capacity (as described above) of the cells after treatment in comparison to untreated control.

Treatment of WT cells with lorvotuzumab mertansine in vitro.
Lorvotuzumab mertansine (huN901-DM1, formerly IMGN901), a humanized version of the anti-CD56 antibody N901 conjugated to the highly cytotoxic maytansine derivative DM1 via a hindered disulfide linker (Wang et al., 2005)  showed the most precise correlation between cell death after treatment (as assessed by trypan blue exclusion assay and MTS proliferation assay) and NCAM expression prior to treatment as determined by FACS analysis. All further experiments evaluating the effect of this drug on WT cells in vitro were performed at this concentration.

Assessment of WT cell survival.
To assess the effect of lorvotuzumab mertansine on WT cell survival in comparison to the unconjugated mAb, WT cells from 3 WT Xn derived from distinct patients (W016 Xn, W027 Xn and W028 Xn) were seeded in 96 well plates in growth medium at 10 4 cells/well. 24 h later medium was changed with growth medium either containing 0.18 µM of lorvotuzumab mertansine or containing 1µM of unconjugated mAb (HuN-901) or with regular growth medium (control) in triplicates. Following 5-day exposure, cell survival was assessed by the MTS proliferation assay as described above.

Assessment of WT cell survival and NCAM expression following treatment with lorvotuzumab mertansine in correlation with initial NCAM expression.
Two assays were used to assess WT cell survival after treatment with the conjugate: MTS proliferation assay and trypan blue exclusion assay. For the MTS assay, cells from four WTs from distinct patients (primary tumors: W005, W006, W007 and freshly dissociated 3 rd generation W011 Xn cells: W011 Xn3) with different NCAM expression levels, as assessed by FACS prior to treatment, and grown, treated and assessed as described above. For the trypan blue exclusion assay, cultured cells from two primary WTs (W007 and W009) and two 3 rd generation WT xenografts (W011 Xn3 Cul and W013 Xn3 Cul) were seeded in 25T flasks at 10 5 cells/flask and treated with lorvotuzumab mertansine or with medium alone as described for the MTS assay.
Following 5 day-exposure, cells were harvested using 0.05% Trypsin/EDTA (Gibco, Grand Island, NY), and viable cells were counted with trypan blue as previously described (Songyang et al., 1997). Assessment of NCAM expression was performed on both treated and untreated cells by FACS analysis as described above.