- Open Access
GLI2 inhibition abrogates human leukemia stem cell dormancy
- Anil Sadarangani1, 8,
- Gabriel Pineda1,
- Kathleen M Lennon1,
- Hye-Jung Chun2,
- Alice Shih1,
- Annelie E Schairer1,
- Angela C Court1,
- Daniel J Goff1,
- Sacha L Prashad3,
- Ifat Geron1,
- Russell Wall1,
- John D McPherson4,
- Richard A Moore2,
- Minya Pu1,
- Lei Bao1,
- Amy Jackson-Fisher7,
- Michael Munchhof7,
- Todd VanArsdale7,
- Tannishtha Reya1,
- Sheldon R Morris1,
- Mark D Minden5, 6,
- Karen Messer1,
- Hanna KA Mikkola3,
- Marco A Marra2,
- Thomas J Hudson4 and
- Catriona HM Jamieson1, 8Email author
© Sadarangani et al.; licensee BioMed Central. 2015
Received: 25 January 2015
Accepted: 6 March 2015
Published: 21 March 2015
Dormant leukemia stem cells (LSC) promote therapeutic resistance and leukemic progression as a result of unbridled activation of stem cell gene expression programs. Thus, we hypothesized that 1) deregulation of the hedgehog (Hh) stem cell self-renewal and cell cycle regulatory pathway would promote dormant human LSC generation and 2) that PF-04449913, a clinical antagonist of the GLI2 transcriptional activator, smoothened (SMO), would enhance dormant human LSC eradication.
To test these postulates, whole transcriptome RNA sequencing (RNA-seq), microarray, qRT-PCR, stromal co-culture, confocal fluorescence microscopic, nanoproteomic, serial transplantation and cell cycle analyses were performed on FACS purified normal, chronic phase (CP) chronic myeloid leukemia (CML), blast crisis (BC) phase CML progenitors with or without PF-04449913 treatment.
Notably, RNA-seq analyses revealed that Hh pathway and cell cycle regulatory gene overexpression correlated with leukemic progression. While lentivirally enforced GLI2 expression enhanced leukemic progenitor dormancy in stromal co-cultures, this was not observed with a mutant GLI2 lacking a transactivation domain, suggesting that GLI2 expression prevented cell cycle transit. Selective SMO inhibition with PF-04449913 in humanized stromal co-cultures and LSC xenografts reduced downstream GLI2 protein and cell cycle regulatory gene expression. Moreover, SMO inhibition enhanced cell cycle transit and sensitized BC LSC to tyrosine kinase inhibition in vivo at doses that spare normal HSC.
In summary, while GLI2, forms part of a core HH pathway transcriptional regulatory network that promotes human myeloid leukemic progression and dormant LSC generation, selective inhibition with PF-04449913 reduces the dormant LSC burden thereby providing a strong rationale for clinical trials predicated on SMO inhibition in combination with TKIs or chemotherapeutic agents with the ultimate aim of obviating leukemic therapeutic resistance, persistence and progression.
Chronic myeloid leukemia (CML) is a myeloproliferative disease arising from the acquisition of the Philadelphia chromosome (Ph) by hematopoietic stem cells (HSC). This is characterized by the reciprocal translocation between the long arms of chromosomes 9 and 22, t (9;22) (q34;q11) which in turn gives rise to a fusion of c-ABL and BCR to generate the BCR-ABL fusion gene that encodes the constitutively active tyrosine kinase BCR-ABL1 . This results in the modulation of several key-signaling cascades leading to increased cell proliferation, induction of cellular transformation and blockage of apoptotic pathways. In CML, chronic phase occurs after BCR-ABL1 oncogene expression in hematopoietic stem cells (HSCs), that can then generate an expansion of a pool of myeloid GMP (CD34 + CD38 + Lin−) progenitors that aberrantly acquire the capacity to self-renew, leading to progenitor LSC generation and BC transformation .
Deregulation of critical stem cell processes, such as self-renewal and dormancy, in hematopoietic progenitors contributes to the generation of leukemia stem cells (LSC) that are impervious to therapies that target dividing cells [2-14]. In advanced human myeloid malignancies that progress from antecedent hematologic malignancies, such as blast crisis (BC) CML, LSC have a proclivity for becoming dormant in protective niches [2,5,7,11,15-18].
Targeted therapies for the treatment of CML using tyrosine kinase inhibitors (TKI) that block BCR–ABL activity have advanced in recent years to include not only imatinib but also second-generation TKIs such as dasatinib and nilotinib. While a striking reduction in CML-related mortality has occurred in patients treated early in chronic phase with BCR-ABL1 targeted tyrosine kinase inhibitor (TKI) therapy, the majority of patients relapse with detectable BCR-ABL1 transcripts following TKI discontinuation, which has been attributed to persistence of dormant BCR-ABL expressing hematopoietic stem and progenitor cells [19,20] Compelling studies, such as the STIM trial, in which a majority of CML patients relapsed within 12 months of TKI discontinuation, suggest that new therapeutic strategies will be required to achieve molecular cures .
Previous studies have established a role for aberrant activation of Hh signaling in a number of solid tumors as well as hematologic malignancies [16,17]. Hh proteins (ligands) bind to the receptor patched (PTCH), causing the release of inhibition by smoothened (SMO) and the subsequent activation of transcription factors GLI-1 and GLI-2 that then translocate to the nucleus and activate their target genes . As the main effectors of Hh pathway signaling following SMO activation, the GLI transcriptional factors, GLI1 and GLI2 [23,24] promote oncogenesis through up-regulation of genes encoding apoptosis inhibitors (e.g., Bcl-2, cFLIP) [25,26] and angiogenesis inducers (e.g., vascular endothelial growth factor) [27-33]. Thus, GLI down-regulation through SMO antagonism could reduce LSC maintenance driven by both Hh ligand dependent and ligand independent epigenetic activation mechanisms [34,35].
In CML Hh signaling is vital for maintenance of LSC through SMO-mediated GLI1 activation but in humans GLI2 increases with blastic transformation . Moreover, activation of the Hh stem cell pathway agonist, smoothened (SMO) and downstream glioma associated (GLI) zinc finger transcriptional regulators, promoted LSC maintenance in a mouse model of CML . Previously, we uncovered a GSK3β missplicing event that prevented degradation of both β-catenin and GLI transcription factors in serially transplantable LSC . Others studies have analyzed newly diagnosed cases of CML and found differential activation of Sonic Hh (SHH) signaling was seen in 50% of chronic phase, 70% of accelerated phase and over 80% of blast crisis cases. It was shown that auto-activation of SHH signaling provided survival and proliferative cues in CML progenitor cells through downstream β- catenin signaling and that by blocking the pathway with inhibitor or neutralizing antibodies, apoptosis could be induced .
While Hh is involved in embryonic pattern formation, including the hemogenic endothelial to hematopoietic transition  it also plays a pivotal role in the maintenance of adult neural and skin stem cell populations. However Hh signaling may be dispensable for maintenance of normal adult hematopoiesis [39-41] suggesting that a therapeutic window may exist for targeting human LSC compared with normal HSC and progenitor populations in the bone marrow and other hematopoietic niches.
Despite extensive transcriptional profiling of BC CML samples [11,13,14], the cell type and context specific role of Hh pathway activation in human LSC therapeutic resistance and cell cycle deregulation had not been elucidated. Moreover, diagnostic strategies that predict progression and therapies capable of abrogating human LSC dormancy while sparing normal HSC function have remained elusive. In this study, we performed RNA-seq, humanized stromal co-cultures, human LSC xenograft assays, lentiviral overexpression and multi-parameter cell cycle FACS analysis to investigate whether 1) Hh pathway and dormancy gene activation are biomarkers of myeloid leukemic progression and LSC generation; 2) to determine if GLI2 drives leukemic progenitor cell cycle deregulation in human LSC supportive niche assays and 3) if selective inhibition of SMO by PF-04449913 reduces niche-dependent LSC maintenance and therapeutic resistance in LSC in vivo models.
Bone marrow and peripheral blood banked samples
Normal cord blood and adult peripheral blood samples were purchased from All Cells. Myeloid malignancy peripheral blood and bone marrow samples were obtained from consenting patients enrolled in Institutional Review Board approved protocols. Human hematopoietic stem and progenitors cells were FACS Aria purified as described previously .
Whole transcriptome RNA sequencing
Progenitor cells (n = 50,000) derived from viably cryopreserved normal cord blood (n = 3), normal peripheral blood (n = 3), CP CML (n = 8) and BC CML (n = 9) samples, were FACS Aria sorted directly into RLT buffer (Qiagen) and LSC microarray studies were performed as described previously . For each sample, approximately 5 ng of total RNA was processed using smarttm cDNA synthesis protocol with smartscribe reverse transcriptase (clontech, #639536), which used a modified oligo (dt) primer to prime the first strand synthesis reaction and a template switching mechanism to generate full length single stranded cDNAs containing the complete 5’ end of the mRNA as well as universal priming sequences for end-to-end amplification by 20 cycles of PCR. The resultant cDNAs were fractionated and size-selected for fragments with 250-350 bp in length. The amplified cDNA was subjected to Illumina paired end library construction using NEBNext paired-end DNA sample Prep Kit (NEB, E6000B-25). Libraries sequenced on Illumina HiSeq2000 instruments with the average of approximately 163.8 million Chastity-passed read pair with 50 bp in read length. Spearman’s rank correlation coefficient was used as the distance metric. To test whether the grouping in the PC plots was statistically significant, permutation tests were performed by random sampling of gene sets with a sample size equal to our signature gene set from KEGG-annotated genes in MSigDB. Genes were excluded if expression levels were zero across all subjects; a total of 4617 genes remained. We drew 5000 random sets in total. A principal components analysis was performed on each random data set and an MANOVA model was built using the first 3 principal components as response variables and the group label as the predictor. A permutation p-value was defined as the proportion of the F-statistic from the original data set that was greater than the F-statistics from the random sets.
Illumina platform RNA-seq read alignment, coverage analysis and clustering: Using BWA software (version 0.5.7) , the Chastity-passed reads were aligned to the human reference genome (hg19/GRch37) plus exon junction sequences constructed from all known transcript models in RefSeq, EnsEMBL and known genes from the UCSC database as described previously . We used default parameter settings of BWA except for the option –s to disable Smith-Waterman alignment. After the alignment, the reads that aligned to exon junctions were repositioned in the genome as large-gapped alignments using repositioning software developed in-house. We removed adapter dimer sequences and soft-clipped reads that contained adapter sequences. The unambiguously aligned, filtered reads were then analyzed by in-house developed gene coverage analysis software to calculate the coverage over the total collapsed exonic regions in each genes as annotated in EnsEMBL (version 59), and the RPKM values were calculated to represent the expression level of exons and genes. We used the raw read counts for differential gene expression analysis using DESeq (v1.4.1) in R. The multiple hypothesis testing correction of p-values from Mann–Whitney U test was done using Benjamini-Hochberg method in p.adjust function in R. We performed unsupervised hierarchical clustering using complete linkage on log2-transformed RPKM values that were centered on the median. Spearman’s rank correlation coefficient was used as the distance metric for clustering samples and genes. The clustering results were visualized using Treeview (Java Cluster software). We used RPKM values and performed the non-negative matrix factorization clustering using NMF (v0.5.06), Biobase (v2.14.0) and cluster (v1.14.2) packages in R. We ran the NMF clustering using the default ‘brunet’ method for the NMF algorithm and set the seeding method to default ‘random’ and performed the clustering over 200 iterations. For the factorization rank survey, we performed clustering using 50 iterations for our dataset, and 25 iterations for a randomized dataset. The visualization of the consensus matrix heatmap and the cophenetic correlation coefficient plot was done using the plot function in R.
Transcriptome analysis of LSC isolated from PF-04449913 and vehicle treated mice
Sample preparation, library construction and sequencing
Neonatal RAG2−/−c−/− mice transplanted intrapheptically with 50,000 BC CML LSC were treated 8 weeks later with vehicle or PF-04449913 (100 mg/kg) for 2 weeks by oral gavage. Four mice were treated with vehicle and four mice were treated with SMO inhibitor (Additional file 1: Table S3). Mice were sacrificed and human leukemia stem cells (~50,000 cells/sample) were sorted from the liver into RLT buffer from the Qiagen Rneasy kit and RNA was extracted. Total RNA samples were treated once with Ribominus kit (Invitrogen, #A10837-08) to deplete ribosomal RNA. From the resulting RNA whole transcriptome libraries were prepared for SOLiD sequencing. Samples were sequenced in 2 different batches to produce 50 bp fragment (i.e. non-paired-end) reads with an average of approximately 118 million reads. Samples 1–7 were sequenced with SOLiD v3.5 chemistry and samples 8–12 were sequenced SOLiD v4 chemistry.
The limma method was used to test for main effects of PF-04449913 and Dasatinib, and their synergistic interaction among 41 genes. Null hypotheses were rejected at p = 0.05 significance level without adjusting for multiple comparisons.
We filtered RNA Seq data using the raw read counts per gene to included genes with at least 10 mapped reads in one or more samples. A total of 13,850 protein-coding genes were included in the analysis. The counts were normalized using upper-quartile normalization. Significance Analysis of Microarrays (SAM) was used to rank the genes according to their differences in expression levels between the four SMO inhibitor treated mice and the four vehicle mice. Gene set enrichment analysis (GSEA) was used to assess the effect of SMO treatment on cell cycle pathways. Of the eight a priori cell cycle pathways considered in the analysis, “Regulation of Cell Cycle” was significantly down regulated comparing treated mice (n = 4) to control mice (n = 4) (family-wise p value =0.02). Six additional pathways out of the 8 total were observed to be down regulated, although not significantly. Table columns are pathway name, number of genes included in the pathway, nominal p-value, FDR adjusted q-value and adjusted p-value controlling for the family-wise error rate. Initially, all 13,850 genes were ranked according to SAM score, and from this ranking the GSEA score for the pathway was computed. The significance of the GSEA score was assessed by randomly permuting gene labels 2000 times and a family wise p-value was computed, correcting for the 8 pathways tested.
For the GSEA enrichment plot for the down-regulated “Regulation of Cell Cycle” Pathway, all 13,850 genes were ranked in descending order based on the Significance Analysis of Microarrays (SAM) score comparing four treated mice vs. four control mice (horizontal heatmap). The GSEA enrichment score for the pathway (genes indicated as vertical bars) is indicated as the maximal excursion of the green line. This pathway has 41 genes; however one gene was filtered out due to low expression. Because the pathway is down regulated, the core enrichment subset consists of the right-most 18 among the 41 expressed genes in the pathway.
Quantitative PCR analysis
The following primers were used in reactions run with SYBR: BCR-ABL Forward: ctccagactgtccacagcat, BCR-ABL Reverse: ccctgaggctcaaagtcaga, HPRT Forward: cgtcttgctcgagatgtgatg, HPRT Reverse: tttatagccccccttgagcac. Relative levels of mRNA were determined according to standard curves. All values were then normalized to HPRT or RPL27 values from the same sample. TaqMan primer/probe sets (ABI) for FoxM1 (Mm00514924_m1), Gli1 (Mm00494645_m1), Gli2 (Mm01293117_m1), Mycn (Mm00476449_m1), Ptch1 (Mm00436026_m1), Ptch2 (Mm00436047_m1), Sfrp1 (Mm00489161_m1), Smo (Mm01162710_m1) on the ABI 7900HT instrument. Target gene expression levels were normalized to GAPDH and to vehicle treated mice to yield the relative quantitation (RQ) value. Expression qRT-PCR arrays (SA biosciences) were also utilized to quantify SHH pathway gene expression.
Stromal co-culture assays
Normal or BC CML CD34+ cells were plated on confluent mitomycin-C treated SL/M2 cells with vehicle, PF-04449913 (1uM), dasatinib (50nM), or combination treatment. Mouse bone marrow stromal cell lines, M2-10B4 (M2) and SL/SL (SL) (Stem Cell Technologies Inc, Vancouver, British Colombia) were treated with mitomycin-C (1 mg/ml) and plated in a 1:1 mixture at a total concentration of 100,000 cells/ml one day prior to co-culture with 10,000-20,000 CD34+ BC CML or normal progenitors. After 1 week of culture, progenitors were FACS sorted into hematopoietic progenitor assays and colonies were scored at 14 days as previously described . To assess survival of normal human hematopoietic stem and progenitor cells, irradiated (20 Gray) OP9 (M2 clone) stromal cells were co-cultured with 50,000 human CD34+ cord blood cells, vehicle or PF-04449913 in AlphaMEM (Gibco) with 20% Hyclone FBS, 1% pen strep glutamine and supplemented with 50 ng/ml SCF, 10 ng/ml thrombopoietin, and 10 ng/ml Flt3 and quantified by weekly FACS analysis.
Normal and leukemia stem cell xenograft assays
Neonatal immunocompromised RAG2−/−c−/− mice were transplanted intrahepatically with equal numbers of normal progenitors or BC LSC as previously described . Upon detection of human CD45+ cell peripheral blood engraftment, mice were treated daily by oral gavage with vehicle (50% 1,2 Propandiol, 50% HBSS or methylcellulose), PF-04449913 (100 mg/kg), Dasatinib (50 mg/kg), or the combination for 14 days followed by FACS to quantify human engraftment in hematopoietic niches. To assess effects on normal HSC function, 7 to 10 week old NOD.Cg-PrkdcSCID IL2R1Wjl/SzJ mice were sublethally irradiated, transplanted retro-orbitally with 100,000 CD34+ human cord blood cells and treated 8 weeks later with vehicle or PF-04449913 (100 mg/kg) for 14 days followed by FACS engraftment analysis.
Confocal fluorescence microscopic and nanoproteomic GLI2 analysis
Human BC LSC xenografted mouse spleens were embedded in OCT freezing media (Sakura, Torrance, CA) frozen at −80°C, sectioned and stained with anti-GLI2 (Sigma) and Alexa 647-conjugated anti-human CD45 (1:25, Serotec) antibodies followed by mounting with Prolong® Gold antifade with DAPI (Invitrogen). Confocal fluorescence images were acquired using an Olympus Fluoview FV10i microscope. Nanofluidic phospho-proteomic immunoassay (NPI) experiments were performed in triplicate with the Nanopro 1000 instrument (Cell Biosciences). After separation and photo-activated in-capillary immobilization, GLI2 was detected using a GLI2-specific antibody (Abcam). A β2-microglubulin-specific antibody (β2M; Upstate) was used to normalize the amount of loaded protein. The peaks were quantified by calculating the area under the curve (AUC).
Cell cycle FACS analysis
Single cell suspensions of bone marrow cells from LSC xenografted mice treated with vehicle or PF-04449913, were immunostained with Alexa647-conjugated anti-human CD45 (BioLegend) and the LIVE⁄DEAD® stain (Invitrogen). Cells were then immunostained with FITC-conjugated anti-Ki-67 (Abcam, 1:100) in 0.15% saponin/2% fetal bovine serum/PBS and incubated with 7- AAD (10 μg/mL in 0.1 M sodium citrate/5 mM EDTA pH8.0/0.15 M NaCl/0.5% BSA/0.02% saponin) followed by FACS Aria analysis.
Statistical analyses for pre-clinical studies were performed with Microsoft Excel and Graphpad Prism software. Continuous variables for each comparison group were assessed for distribution through univariate statistics. If the assumption of normal distribution could be supported, then the Student’s t test was performed for comparison of two samples with assessment of equality of variance with an F statistic. If the assumption of normal distribution was not supported, nonparametric testing was performed with the two samples Wilcoxon test using the t approximation for samples with N of less than 20. Differential gene expression from RNA-Seq data was detected using R package, DESeq, (v1.4.1) with a significance threshold at 5% false discovery rate (FDR). The Wilcoxon test was used for comparing means of genes expression and identifying genes with significant fold changes. P-values were corrected for multiple hypotheses testing using the Benjamini-Hochberg method. Unsupervised hierarchical clustering analysis was performed using complete linkage on log2-transformed RPKM values that were centered on the median.
Hh pathway deregulation signifies human leukemic progression and LSC generation
Selective SMO inhibition reduces GLI2-expressing LSC burden in supportive niches
While some studies suggest that Hh signaling is dispensable for normal mouse hematopoiesis , others demonstrate that GLI regulates hematopoietic stem cell (HSC) fate decisions . However, the role of Hh signaling in normal human HSC and progenitor cell maintenance had not been examined extensively in vitro or in primary sample xenograft models that permit robust engraftment. Thus, we compared the effects of PF-04449913 and vehicle treatment on normal hematopoietic stem and progenitor cells and demonstrated no change in differentiation capacity (Additional file 1: Figure S2a). When CD34+ cord blood engrafted NSG mice were treated with PF-04449913, the frequency of human CD45+ cells, progenitors and both myeloid and lymphoid cell fate commitment remained comparable to vehicle treated controls (Additional file 1: Figure S2b-d) indicating that unlike LSC, normal human HSC cell fate decisions are Hh pathway independent . These results highlight the important niche dependent effects of selective SMO inhibition that induce GLI2 downregulation in a cell type and context specific manner.
GLI2 induces cell cycle arrest in leukemic progenitors
SMO inhibition abrogates LSC dormancy in vivo
GLI2 inhibition enhances LSC cycling and TKI sensitivity
RNA-Seq gene expression analyses of functionally validated LSC in advanced human myeloid malignancies, such BC CML, demonstrate that activation of stem cell self-renewal and survival pathways signifies therapeutic resistance that could translate into poor clinical outcomes. These studies suggest that LSC profiling may represent an important patient prognostication strategy. Current evidence also suggests that deregulated stem cell dormancy, survival and self-renewal pathway activation leads to malignant reprogramming of progenitors into LSC that drive therapeutic resistance, relapse and myeloid leukemia progression [2,7,15,50-52]. However, selective and clinically effective dormant human LSC eradication strategies have remained elusive.
Because SMO signals primarily through GLI2, we hypothesized that selective SMO inhibition would reduce the dormant LSC burden particularly when combined with TKIs. Indeed, SMO treatment enhanced TKI sensitivity and inhibited LSC self-renewal in stromal co-culture and LSC xenograft models at doses that spared normal hematopoietic stem and progenitor cells. Moreover, SMO inhibition enhanced BC LSC sensitivity to a potent BCR-ABL1 inhibitor, dasatinib, leading to a reduction in self-renewal capacity. These studies suggest that cell cycle induction in dormant LSC in conjunction with targeted therapy may prevent leukemic persistence and progression secondary to therapeutic resistance.
Other key hedgehog related mechanisms of chemotherapeutic resistance have been identified in other myeloid malignancies, such as AML. A recent study demonstrated that GLI alone is sufficient to drive UGT1A-dependent glucuronidation of Ara-C in AML thereby leading to drug resistance. This resistance was overcome by genetic or pharmacologic inhibition of GLI using a SMO antagonist . Based on recent advances in leukemia clinical trials showing promise of a variety of molecular therapies, the key to eradicating an evolving population of cells lies in combination therapies that selectively inhibit multiple targets. Ideally, these targets will be specifically activated in the LSC compartment, and pharmacological inhibition will spare normal adult stem cells during the course of treatment.
While selective reduction in LSC self-renewal capacity provides strong pre-clinical evidence for a direct LSC inhibitory effect, the potential capacity of PF-04449913 to modify the malignant niche may also contribute to its anti-leukemic efficacy and will need to be examined in clinical correlative studies in ongoing clinical trials. Moreover, the clinical importance of SMO inhibitor mediated sensitization of dormant LSC to therapeutic modalities that target dividing cells will need to be assessed in Phase 2 clinical trials. Overall, our results inform the development of a clinically tractable strategy for obviating therapeutic resistance by selective SMO inhibitor targeting of dormant leukemia stem cells alone or in combination with standard of care in advanced hematologic malignancies.
Our work addresses the role of smoothened (SMO) activation of its downstream hedgehog pathway transcriptional activator, GLI2, in human myeloid leukemia progression, leukemia stem cell self-renewal and dormancy. RNA-Seq, nanoproteomic and qRT-PCR analyses of FACS purified progenitors revealed that distinctive expression patterns of both SHH and cell cycle regulatory genes, including GLI2cpg CDKN1A and CDKN2A, were harbingers of myeloid leukemic progression and LSC generation. Because of the niche dependency of Hh pathway signaling, the capacity of a selective SMO antagonist, PF-04449913, to inhibit human LSC was evaluated in stromal co-cultures and in various hematopoietic tissues in xenografts models. In these systems, LSC burden decreased commensurate with GLI2 and cell cycle regulatory gene repression and while GLI2 overexpression enhances dormant human leukemia stem cell generation, selective inhibition of SMO abrogates it and sensitizes them to tyrosine kinase inhibitors. While lentiviral overexpression of human GLI2 enhanced CML progenitor dormancy in stromal co-culture assays, the transcriptionally inactive form of GLI2 did not. In addition, cell cycle FACS analysis revealed a reduction in dormant LSC following PF-04449913 treatment in all hematopoietic niches. Our data suggest that GLI2 represents part of a core transcriptional program driving human myeloid LSC maintenance thereby providing the rationale for clinically assessing the capacity of a selective SMO antagonist alone or in combination with anti-proliferative agents to eradicate dormant therapy resistant LSC in advanced myeloid malignancies.
We are indebted to Dr. Dennis Carson for his indispensable advice as well as Kimberly Booth and Leisa Sutton for invaluable assistance with manuscript preparation. We would also like to thank Maria Sierra, Masaya Ueno and Yanling Wang for assistance with LSC experiments. We acknowledge expert technical assistance from the Library and Sequencing Cores at the Canada’s Michael Smith Genome Sciences Centre. PF-04449913 was provided through a MTA with Pfizer Inc. CHMJ, DJG and this research were supported by California Institute for Regenerative Medicine grants (RN2-00910-1, RS1-00228-1 DR1-01430, TG2-01154), Cancer Stem Cell Consortium funding from Genome Canada and the Ontario Genomics Institute (OGI-047); NIH NIGMS 5K12GM068524; NIH NCI 2P30CA023100-28; The Sanford Stem Cell Clinical Center; The San Diego Foundation; The Ratner Family Foundation; The Mizrahi Family Foundation; and the Canadian Institute of Health Research (CSC-105367). The authors wish to thank the Genome Sciences Centre Library Construction, Sequencing and Bioinformatics teams (BC Cancer Agency) and the UC San Diego Cancer Biology Training Grant. GP was supported by 5K12GM068524 from the NIGMS. SLP is a Howard Hughes Medical Institute Gilliam fellow. This work was also funded by the Ontario Institute for Cancer Research, through generous support from the Ontario Ministry of Research and Innovation and the Cancer Stem Cell Consortium with funding from Genome Canada, Genome British Columbia, the Ontario Genomics Institute (OGI-047), and the Canadian Institute of Health Research (CSC-105367).
- Tkachuk DC, Westbrook CA, Andreeff M, Donlon TA, Cleary ML, Suryanarayan K, et al. Detection of bcr-abl fusion in chronic myelogeneous leukemia by in situ hybridization. Science. 1990;250(4980):559–62.View ArticlePubMedGoogle Scholar
- Jamieson CH, Ailles LE, Dylla SJ, Muijtjens M, Jones C, Zehnder JL, et al. Granulocyte-macrophage progenitors as candidate leukemic stem cells in blast-crisis CML. N Engl J Med. 2004;351(7):657–67.View ArticlePubMedGoogle Scholar
- Levy JG, Dowding C, Mitchell D, Sorrenti R, Yip S, Jamieson C. Selective elimination of malignant stem cells using photosensitizers followed by light treatment. Stem Cells. 1995;13(4):336–43.View ArticlePubMedGoogle Scholar
- Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997;3(7):730–7.View ArticlePubMedGoogle Scholar
- Holyoake T, Jiang X, Eaves C, Eaves A. Isolation of a highly quiescent subpopulation of primitive leukemic cells in chronic myeloid leukemia. Blood. 1999;94(6):2056–64.PubMedGoogle Scholar
- Jin L, Hope KJ, Zhai Q, Smadja-Joffe F, Dick JE. Targeting of CD44 eradicates human acute myeloid leukemic stem cells. Nat Med. 2006;12(10):1167–74.View ArticlePubMedGoogle Scholar
- Abrahamsson AE, Geron I, Gotlib J, Dao KH, Barroga CF, Newton IG, et al. Glycogen synthase kinase 3beta missplicing contributes to leukemia stem cell generation. Proc Natl Acad Sci U S A. 2009;106(10):3925–9.View ArticlePubMed CentralPubMedGoogle Scholar
- Essers MA, Trumpp A. Targeting leukemic stem cells by breaking their dormancy. Mol Oncol. 2010;4(5):443–50.View ArticlePubMedGoogle Scholar
- Forsberg EC, Passegue E, Prohaska SS, Wagers AJ, Koeva M, Stuart JM, et al. Molecular signatures of quiescent, mobilized and leukemia-initiating hematopoietic stem cells. PLoS One. 2010;5(1):e8785.View ArticlePubMed CentralPubMedGoogle Scholar
- Saito Y, Kitamura H, Hijikata A, Tomizawa-Murasawa M, Tanaka S, Takagi S, et al. Identification of therapeutic targets for quiescent, chemotherapy-resistant human leukemia stem cells. Sci Transl Med. 2010;2(17):17ra19.View ArticleGoogle Scholar
- Eppert K, Takenaka K, Lechman ER, Waldron L, Nilsson B, Van Galen P, et al. Stem cell gene expression programs influence clinical outcome in human leukemia. Nat Med. 2011;17(9):1086–93.View ArticlePubMedGoogle Scholar
- Li L, Wang L, Wang Z, Ho Y, McDonald T, Holyoake TL, et al. Activation of p53 by SIRT1 inhibition enhances elimination of CML leukemia stem cells in combination with imatinib. Cancer Cell. 2012;21(2):266–81.View ArticlePubMed CentralPubMedGoogle Scholar
- Goff DJ, Court Recart A, Sadarangani A, Chun HJ, Barrett CL, Krajewska M, et al. A Pan-BCL2 inhibitor renders bone-marrow-resident human leukemia stem cells sensitive to tyrosine kinase inhibition. Cell Stem Cell. 2013;12(3):316–28.View ArticlePubMed CentralPubMedGoogle Scholar
- Jiang Q, Crews LA, Barrett CL, Chun HJ, Court AC, Isquith JM, et al. ADAR1 promotes malignant progenitor reprogramming in chronic myeloid leukemia. Proc Natl Acad Sci U S A. 2013;110(3):1041–6.View ArticlePubMed CentralPubMedGoogle Scholar
- Krivtsov AV, Twomey D, Feng Z, Stubbs MC, Wang Y, Faber J, et al. Transformation from committed progenitor to leukaemia stem cell initiated by MLL-AF9. Nature. 2006;442(7104):818–22.View ArticlePubMedGoogle Scholar
- Dierks C, Beigi R, Guo GR, Zirlik K, Stegert MR, Manley P, et al. Expansion of Bcr-Abl-positive leukemic stem cells is dependent on Hedgehog pathway activation. Cancer Cell. 2008;14(3):238–49.View ArticlePubMedGoogle Scholar
- Zhao C, Chen A, Jamieson CH, Fereshteh M, Abrahamsson A, Blum J, et al. Hedgehog signalling is essential for maintenance of cancer stem cells in myeloid leukaemia. Nature. 2009;458(7239):776–9.View ArticlePubMed CentralPubMedGoogle Scholar
- Heidel FH, Bullinger L, Feng Z, Wang Z, Neff TA, Stein L, et al. Genetic and pharmacologic inhibition of beta-catenin targets imatinib-resistant leukemia stem cells in CML. Cell Stem Cell. 2012;10(4):412–24.View ArticlePubMed CentralPubMedGoogle Scholar
- Nagase T, Nagase M, Machida M, Yamagishi M. Hedgehog signaling: a biophysical or biomechanical modulator in embryonic development? Ann N Y Acad Sci. 2007;1101:412–38.View ArticlePubMedGoogle Scholar
- Varjosalo M, Taipale J. Hedgehog signaling. J Cell Sci. 2007;120(Pt 1):3–6.PubMedGoogle Scholar
- Wang Y, McMahon AP, Allen BL. Shifting paradigms in Hedgehog signaling. Curr Opin Cell Biol. 2007;19(2):159–65.View ArticlePubMedGoogle Scholar
- Irvine DA, Copland M. Targeting hedgehog in hematologic malignancy. Blood. 2012;119(10):2196–204.View ArticlePubMedGoogle Scholar
- Bai CB, Auerbach W, Lee JS, Stephen D, Joyner AL. Gli2, but not Gli1, is required for initial Shh signaling and ectopic activation of the Shh pathway. Development. 2002;129(20):4753–61.PubMedGoogle Scholar
- Lipinski RJ, Gipp JJ, Zhang J, Doles JD, Bushman W. Unique and complimentary activities of the Gli transcription factors in Hedgehog signaling. Exp Cell Res. 2006;312(11):1925–38.View ArticlePubMedGoogle Scholar
- Regl G, Kasper M, Schnidar H, Eichberger T, Neill GW, Ikram MS, et al. The zinc-finger transcription factor GLI2 antagonizes contact inhibition and differentiation of human epidermal cells. Oncogene. 2004;23(6):1263–74.View ArticlePubMedGoogle Scholar
- Kump E, Ji J, Wernli M, Hausermann P, Erb P. Gli2 upregulates cFlip and renders basal cell carcinoma cells resistant to death ligand-mediated apoptosis. Oncogene. 2008;27(27):3856–64.View ArticlePubMedGoogle Scholar
- Berman DM, Karhadkar SS, Maitra A, Montes De Oca R, Gerstenblith MR, Briggs K, et al. Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours. Nature. 2003;425(6960):846–51.View ArticlePubMedGoogle Scholar
- Thayer SP, Di Magliano MP, Heiser PW, Nielsen CM, Roberts DJ, Lauwers GY, et al. Hedgehog is an early and late mediator of pancreatic cancer tumorigenesis. Nature. 2003;425(6960):851–6.View ArticlePubMed CentralPubMedGoogle Scholar
- Katoh M. WNT/PCP signaling pathway and human cancer (review). Oncol Rep. 2005;14(6):1583–8.PubMedGoogle Scholar
- Eichberger T, Sander V, Schnidar H, Regl G, Kasper M, Schmid C, et al. Overlapping and distinct transcriptional regulator properties of the GLI1 and GLI2 oncogenes. Genomics. 2006;87(5):616–32.View ArticlePubMedGoogle Scholar
- Mukherjee S, Frolova N, Sadlonova A, Novak Z, Steg A, Page GP, et al. Hedgehog signaling and response to cyclopamine differ in epithelial and stromal cells in benign breast and breast cancer. Cancer Biol Ther. 2006;5(6):674–83.View ArticlePubMed CentralPubMedGoogle Scholar
- Ji Z, Mei FC, Xie J, Cheng X. Oncogenic KRAS activates hedgehog signaling pathway in pancreatic cancer cells. J Biol Chem. 2007;282(19):14048–55.View ArticlePubMedGoogle Scholar
- Thiyagarajan S, Bhatia N, Reagan-Shaw S, Cozma D, Thomas-Tikhonenko A, Ahmad N, et al. Role of GLI2 transcription factor in growth and tumorigenicity of prostate cells. Cancer Res. 2007;67(22):10642–6.View ArticlePubMed CentralPubMedGoogle Scholar
- Siegelin MD, Siegelin Y, Habel A, Rami A, Gaiser T. KAAD-cyclopamine augmented TRAIL-mediated apoptosis in malignant glioma cells by modulating the intrinsic and extrinsic apoptotic pathway. Neurobiol Dis. 2009;34(2):259–66.View ArticlePubMedGoogle Scholar
- Zahreddine HA, Culjkovic-Kraljacic B, Assouline S, Gendron P, Romeo AA, Morris SJ, et al. The sonic hedgehog factor GLI1 imparts drug resistance through inducible glucuronidation. Nature. 2014;511(7507):90–3.View ArticlePubMed CentralPubMedGoogle Scholar
- Radich JP, Dai H, Mao M, Oehler V, Schelter J, Druker B, et al. Gene expression changes associated with progression and response in chronic myeloid leukemia. Proc Natl Acad Sci U S A. 2006;103(8):2794–9.View ArticlePubMed CentralPubMedGoogle Scholar
- Su W, Meng F, Huang L, Zheng M, Liu W, Sun H. Sonic hedgehog maintains survival and growth of chronic myeloid leukemia progenitor cells through beta-catenin signaling. Exp Hematol. 2012;40(5):418–27.View ArticlePubMedGoogle Scholar
- Kim PG, Albacker CE, Lu YF, Jang IH, Lim Y, Heffner GC, et al. Signaling axis involving Hedgehog, Notch, and Scl promotes the embryonic endothelial-to-hematopoietic transition. Proc Natl Acad Sci U S A. 2013;110(2):E141–50.View ArticlePubMed CentralPubMedGoogle Scholar
- McMahon AP, Ingham PW, Tabin CJ. Developmental roles and clinical significance of hedgehog signaling. Curr Top Dev Biol. 2003;53:1–114.View ArticlePubMedGoogle Scholar
- Gao J, Graves S, Koch U, Liu S, Jankovic V, Buonamici S, et al. Hedgehog signaling is dispensable for adult hematopoietic stem cell function. Cell Stem Cell. 2009;4(6):548–58.View ArticlePubMed CentralPubMedGoogle Scholar
- Hofmann I, Stover EH, Cullen DE, Mao J, Morgan KJ, Lee BH, et al. Hedgehog signaling is dispensable for adult murine hematopoietic stem cell function and hematopoiesis. Cell Stem Cell. 2009;4(6):559–67.View ArticlePubMed CentralPubMedGoogle Scholar
- Geron I, Abrahamsson AE, Barroga CF, Kavalerchik E, Gotlib J, Hood JD, et al. Selective inhibition of JAK2-driven erythroid differentiation of polycythemia vera progenitors. Cancer Cell. 2008;13(4):321–30.View ArticlePubMedGoogle Scholar
- Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25(14):1754–60.View ArticlePubMed CentralPubMedGoogle Scholar
- Morin R, Bainbridge M, Fejes A, Hirst M, Krzywinski M, Pugh T, et al. Profiling the HeLa S3 transcriptome using randomly primed cDNA and massively parallel short-read sequencing. Biotechniques. 2008;45(1):81–94.View ArticlePubMedGoogle Scholar
- Lou H, Dean M. Targeted therapy for cancer stem cells: the patched pathway and ABC transporters. Oncogene. 2007;26(9):1357–60.View ArticlePubMedGoogle Scholar
- Morton JP, Lewis BC. Shh signaling and pancreatic cancer: implications for therapy? Cell Cycle. 2007;6(13):1553–7.View ArticlePubMedGoogle Scholar
- Merchant A, Joseph G, Wang Q, Brennan S, Matsui W. Gli1 regulates the proliferation and differentiation of HSCs and myeloid progenitors. Blood. 2010;115(12):2391–6.View ArticlePubMed CentralPubMedGoogle Scholar
- Roessler E, Ermilov AN, Grange DK, Wang A, Grachtchouk M, Dlugosz AA, et al. A previously unidentified amino-terminal domain regulates transcriptional activity of wild-type and disease-associated human GLI2. Hum Mol Genet. 2005;14(15):2181–8.View ArticlePubMedGoogle Scholar
- Mahon FX, Rea D, Guilhot J, Guilhot F, Huguet F, Nicolini F, et al. Discontinuation of imatinib in patients with chronic myeloid leukaemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre Stop Imatinib (STIM) trial. Lancet Oncol. 2010;11(11):1029–35.View ArticlePubMedGoogle Scholar
- Perrotti D, Jamieson C, Goldman J, Skorski T. Chronic myeloid leukemia: mechanisms of blastic transformation. J Clin Invest. 2010;120(7):2254–64.View ArticlePubMed CentralPubMedGoogle Scholar
- Goardon N, Marchi E, Atzberger A, Quek L, Schuh A, Soneji S, et al. Coexistence of LMPP-like and GMP-like leukemia stem cells in acute myeloid leukemia. Cancer Cell. 2011;19(1):138–52.View ArticlePubMedGoogle Scholar
- Crews LA, Jamieson CH. Chronic myeloid leukemia stem cell biology. Curr Hematol Malig Rep. 2012;7(2):125–32.View ArticlePubMed CentralPubMedGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.