UvA-DARE (Digital Academic Repository)
Crossing Borders: Manipulating stemness through carcinogenesis
Zimberlin, C.D.
Publication date 2016 Document Version Final published version
Link to publication
Citation for published version (APA): Zimberlin, C. D. (2016). Crossing Borders: Manipulating stemness through carcinogenesis.
General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
Download date:02 Oct 2021 8 HDAC inhibitors sensitize colon cancer stem cells (CSCs) to apoptosis via FOXO4 regulation
Manuscript in preparation
TThesis_CD_Zimberlin.indbhesis_CD_Zimberlin.indb 143143 008/05/168/05/16 22:0522:05 Chapter 8
HDAC inhibitors sensitize colon cancer stem cells (CSCs) to apoptosis via FOXO4 regulation
C.D.Zimberlin*1, S.Colak*1,2, M.S.Roca*1,3, K.Cameron1, E.E.Santo4, S.R.van Hooff1, H.Rodermond1, S.Simmini1, M.Bots1, C.M.Grandela1,5, B.M.Burgering6,7, J.P.Medema1,7
* These authors contributed equally
1 LEXOR (Laboratory of Experimental Oncology and Radiobiology), Center for Experimental Molecular Medicine, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands
2 Department of Molecular Cell Biology, Leiden University Medical Center, Einthovenweg 20, 2333 ZC, Leiden, The Netherlands
3 Experimental Pharmacology Unit, Istituto Nazionale Tumori Fondazione G. Pascale - IRCCS, 80131 Naples, Italy
8 4 Department of Pathology and Laboratory Medicine , Weill Cornell Medicine , New York City , NY , USA
5 Current address: Pluriomics BV, Leiden 2333BD, The Netherlands
6 Department of Molecular Cancer Research, University Medical Center Utrecht, Universiteitsweg 100, 3584 CG, Utrecht, the Netherlands
7 Cancer Genomics Center, The Netherlands
Key words: Histone deacetylase, Tumor initiating cells (TIC), LBH-589, BCL-XL, FOXO, ABT-737, Colorectal cancer, HDACi
Running title: HDAC inhibitors target CSCs for cell death
144
TThesis_CD_Zimberlin.indbhesis_CD_Zimberlin.indb 144144 008/05/168/05/16 22:0522:05 Chapter 8
Abstract
In many cancer types cancer stem cells (CSCs) have been identifi ed and suggested to fuel tumor growth. CSCs are reportedly more resistant to chemotherapy and radiotherapy, providing a mechanism for disease relapse after therapy. Targeting CSC therefore provides a possible therapeutic window for the treatment of tumors. Using a single cell-based analysis of cell death and a small-scale compound screen, we studied sensitization of colon-CSCs towards conventional chemotherapy. We identifi ed HDAC inhibitors (HDACi) as potential candidates, which strongly enhanced oxaliplatin-induced apoptosis in colon-CSCs. Closer examination revealed that HDACi induced a change in cellular morphology and a striking increase in the level of differentiation markers with a concomitant decrease of stem cell markers. Importantly, de novo transcription was required for sensitization and subsequent gene-expression analysis pointed to the involvement of FOXO transcription factors, particularly FOXO4. Treatment with HDACi consistently increased FOXO4 expression levels and lowered the expression of BCL-XL. Importantly, knockdown of FOXO4 prevented this HDACi-induced decrease in BCL-XL, indicating that FOXO4 induction is instrumental in this regulation. More importantly, we fi nd that HDACi specifi cally decreased the mitochondrial apoptotic threshold in colon-CSCs as measured with ABT-737. Our results therefore show that HDACi sensitize colon-CSCs through a novel mechanism involving the loss of stemness and a FOXO4-dependent lowering of the apoptotic threshold. Both activities provide a rational for the combination of HDACi with chemotherapy in colorectal cancer. 8
145
TThesis_CD_Zimberlin.indbhesis_CD_Zimberlin.indb 145145 008/05/168/05/16 22:0522:05 Chapter 8
Introduction
Colorectal cancer (CRC) is one of the leading causes of cancer-related mortality worldwide. Therapeutic developments have increased 5-year patient survival rates to ±90%, however, once distance metastasis has occurred survival drops to less than 15% [1]. Patients with advanced stage CRC are routinely treated with a combination of fl uoropyrimidines (Capecitabine or Fluorouracil) and oxaliplatin or irinotecan. Despite the development of targeted therapies, expanding experimental evidence indicates robust signaling redundancy present within CRC lesions, leading to acquired secondary resistance [2]. Research efforts to effi ciently block tumor growth and adaptation should thus be prioritized in the development of new therapeutic strategies.
CRC lesions are reported to contain a cellular hierarchy, of which the colon cancer stem cells (CSCs) are at the apex. The CSCs are a small subset of cells, which drive tumor growth and progression. Colon-CSCs can be defi ned by a variety of different markers including Wnt pathway activity, LGR5, OLFM4 and CD133 (reviewed in [3]). Intriguingly, despite being proliferative, colon-CSCs been have been proven to display selective resistance to conventional chemotherapy, thus escaping chemotherapeutic insults and allowing for a repopulation of the tumor [4-6].
The resistance of CSCs appears to be an intrinsic property that is lost upon differentiation and exploiting the differences between CSCs and their more differentiated progeny may thus be the key to fi nding new ways to sensitize CSCs to chemotherapeutics. Indeed upon activation of the 8 Notch pathway using a DLL4 neutralizing antibody in vivo, CSCs undergo differentiation and can subsequently be targeted by chemotherapeutics [7]. Similarly, activation of the ER stress pathway [8] or exposure to BMP4 ligands [9] can induce differentiation of colon-CSCs and sensitize them to chemotherapy. Recently, we reported that, when compared to their differentiated progeny, colon-CSCs are less primed to mitochondrial cell death and can be sensitized by inhibiting the anti-apoptotic molecule BCL-XL, with small compound BH3 mimetics [4]. Further identifi cation of compounds that share the ability to modulate CSC resistance could potentially provide key mechanistic insight into CSC functioning, which can subsequently be exploited for therapeutic use.
We therefore performed a small-scale screen of selected compounds and identifi ed histone deacetylase inhibitors (HDACi) to effectively sensitize CSCs to chemotherapy, which involved de novo transcription/translation. Motif analysis of microarray data indicated a role for the FOXO family members, particulary FOXO4. HDACi upregulate FOXO4 expression, which negatively regulates BCL-XL levels in primary colon cells enriched for CSCs. In addition, we observed loss of stemness and increased differentiation-associated gene expression. Decreased levels of BCL- XL lowered the apoptotic threshold of colon-CSCs sensitizing them to chemotherapeutics. The
146
TThesis_CD_Zimberlin.indbhesis_CD_Zimberlin.indb 146146 008/05/168/05/16 22:0522:05 Chapter 8
transcriptional regulation of FOXO4 to alter the apoptotic threshold of colon-CSCs reveals a novel means to target these cells and enhance their chemotherapy sensitivity.
Results
HDAC inhibitors can sensitize colon-CSCs to oxaliplatin treatment
CRC develops in a step-wise manner, where aberration of the Wnt pathway by mutations in APC or CTNNB1 is viewed as an early event [10, 11]. Despite deregulation of the Wnt pathway in all CRC cells a heterogeneous expression of Wnt activity can be observed within a CRC lesion [12-14]. Previously we have shown that the heterogeneity in Wnt pathway activity can help defi ne different subset of cells, where high Wnt activity defi nes the clonogenic colon-CSC population and low Wnt activity defi nes their more differentiated progeny [12]. Using a single cell cloned primary spheroid culture, Co01, containing a TOP-GFP Wnt reporter, the Wnt pathway activity can be monitored by GFP expression. In this manner both differentiated cells (10% GFPlow) and CSCs (10% GFPhigh) can be examined within the same population. Recently we have described a fl ow cytometry technique, in which apoptosis based on activated caspase-3 can be measured in single cells allowing for an analysis of chemotherapy-induced cell death of differentiated and colon-CSCs under the same conditions. Based on this technique we showed colon-CSCs to be less sensitive to conventional chemotherapeutics than their differentiated counterparts due to a BCL-XL dependent roadblock in apoptotic signaling [4]. As BCL-XL targeting mimetics display relatively high toxicity we here opted for a strategy to fi nd novel means to sensitize colon-CSCs to oxaliplatin. To this end we performed a small-scale drug screen using our singe 8 cell-based apoptosis assay, testing 50 different therapeutic agents targeting multiple cellular processes (fi gure 1a-c, supplementary table 1). Interestingly, the highest sensitization of CSCs to chemotherapeutics was observed with two independent HDAC inhibitors (fi gure 1b-c and supplementary table 1). The establishment of chemosensitivity in colon-CSCs upon HDACi treatment was validated and expanded to various pan-HDAC inhibitors including panobinostat (LBH-589), vorinostat (SAHA), MS-275, valproic acid (VPA), Trichostatin A (TSA) and sodium butyrate (NaB) (fi gure 1d). Colon-CSCs hereby appear to be dependent on HDAC activity to maintain their chemoresistant state. In order to validate the effect of HDAC inhibition on chemotherapy-effi cacy, in vivo xenograft experiments were performed using oxaliplatin and the HDACi panobinostat. Colon-CSCs were injected in subcutaneously in mice and allowed to grow to a size of 100mm3. As reported before oxaliplatin alone at these low concentrations provides no survival benefi t to the mice due to a selective resistance of CSCs [5, 15]. Similar results were observed for HDACi alone. However, the combination provided a strong survival benefi t and even induced a complete regression in 2 of the 8 mice (Figure 1e). HDAC inhibition in combination with oxaliplatin thus signifi cantly improves chemotherapy response in vitro and in vivo.
147
TThesis_CD_Zimberlin.indbhesis_CD_Zimberlin.indb 147147 008/05/168/05/16 22:0522:05 Chapter 8
A. 1. Treat Primary CRC spheriods 2. Cell isolation + staining 3. Analysis of cell death
Wnt low % DeDeadad CellsCells Diffrentiated cells
10% Differentiated GFPlow Cells
% Dead Cells
10% GFPhigh Wnt high Stem Cells Stem cells
B. C. Cell death in CSCs in single vs combined treatment with oxaliplatin 100 Dif Ranking Inhibitor Fold induction Class Inhibits CSC 1 Vorinostat 3.9 Epigenetics HDACs 80 2 Panobinostat 3.7 Epigenetics HDACs 3 17-DMAG 2.5 Transc/Transl HSP90 60 4 SB203580 2.1 Signalling p38 MAPK 5 Roscovitine 1.8 Signalling CDK 40 6 NU1025 1.8 DNA repair PARP 7 Pepstatin A 1.7 Proteases Aspartic proteinases 20
% Caspase-3 active cells 8 SCD1 inhibitor 1.6 Metabolism Fatty acid synthase 9 Olaparib 1.6 DNA repair PARP 0 CtrlOxali Oxali + Oxali + Oxali + Oxali + 10 2-aminopurine 1.5 Transc/Transl PKR Vorino Pano 17-DMAG SB D. E. 80 100 Ctrl Oxali Pano 60 80 Pano + Oxali
60 8 40 40
20 % surviving mice p = 0.02 20 % Caspase-3 active cells
0 0 Oxali -++ -+-+-+-+-+- 0 35 70 105 140 CtrlPano Vorino MS-275 TSA VPA NaB Days after first treatment
Figure 1: HDAC inhibitors can sensitize CSCs to oxaliplatin treatment A) overview of the screening system used. B) Flow cytometry analysis of activated caspase-3 shows the outcome of the induced chemosensitivity in colon-CSCs (red) and differentiated cells (blue) when oxaliplatin (Oxali) is combined with: Vorinostat (Vorino) or Panobinostat (Pano) or 17-DMAG or SB203580 (SB). C) A table representing the top ten compounds that established the greatest fold change in cell death in colon-CSCs specifi cally when compared to oxaliplatin alone. A full list of all the dosages and fold change observed is provided in Sup. Table 1. D) Flow cytometry analysis of activated caspase-3 shows that multiple HDAC inhibitors sensitize colon-CSCs to cell death in combination with oxliplatin. Panobinostat (Pano), vorinostat (Vorino), MS-275, valproic acid (VPA), Trichostatin A (TSA) and sodium butyrate (NaB). E) Survival curves of colon- CSC xenografts treated with the vehicle (Ctrl), panbinostat (Pano), oxaliplatin (Oxali) and a combination of panobinostat and oxaliplatin (Pano + Oxali). The grey box shows the time-scale when mice were treated. n=6
148
TThesis_CD_Zimberlin.indbhesis_CD_Zimberlin.indb 148148 008/05/168/05/16 22:0522:05 Chapter 8
HDAC inhibitors induce chemosensitivity by transcriptional regulation
HDAC together with their counterpart’s histone acetylases (HATs) maintain the acetylation status of proteins within the cell. The acetylation status of proteins is crucial for the maintenance of homeostasis and a plethora of cellular processes. HDACs have been shown to induce transcriptional repression as deacetylation of lysine residues on histone proteins induces chromatin compaction [16, 17]. However, non-histone proteins, including transcription factors, are also regulated in their activity by HDAC-mediated deacetylation, which makes the effect of HDACi both pleiotropic and cell type dependent. Using Actinomycin D (ActD) and Cycloheximide (CHX) to block transcription and translation respectively, we did observe that HDACi-induced chemosensitization fully depended on de novo mRNA/protein synthesis (fi gure 2a).
In order to gain further insight into the mechanism underlying HDACi sensitization of colon- CSCs, gene expression profi les were generated from control or panobinostat-treated colon-CSCs or differentiated cells. Using a previously described CSC signature [18], panobinostat-treated colon-CSCs were found to have a decreased resemblance to colon-CSCs, gaining similarity to the more-differentiated cell population (fi gure 2b). Importantly, upon closer inspection panobinostat treatment decreased the expression levels of selected colon CSC markers LGR5, CD133, ASCL2, and EPHB2, while it resulted in increased expression levels of the differentiation markers; cytokeratin 20 (KRT20) and CEACAM5 (fi gure 2c and supplementary table 2). This decrease in CSC markers and concomitant increase in differentiation markers suggests that HDAC inhibition induces a differentiation-like phenotype, which may sensitize CSCs to 8 conventional therapeutics. Consistent with these fi ndings, panobinostat treatment revealed marked morphological changes in Co01 cells (supplementary fi gure 1). This morphological differentiation was even more evident when spheroid cultures were grown in Matrigel, where panobinostat induced the formation of strongly polarized structures, comparable to spheroids treated with FCS, a known differentiation factor [5] (fi gure 2d). These distinctive polarization changes were present in 90% of the structures observed upon panobinostat treatment (fi gure 2d) and coincided with increased and polarized expression of the differentiation marker cytokeratin 20 (KRT20), both on protein (fi gure 2d) and mRNA expression (fi gure 2e).
Microarray data also showed a signifi cant decrease in stem cell markers, which was confi rmed using qPCR-based analysis of LGR5 (fi gure 2f). Moreover, panobinostat treatment also decreased surface expression of stem cell markers LGR5 and CD133 (fi gure 2g). Importantly, these effects are not unique to this primary colon cancer culture, but could further be extended to several primary CRC spheroid cultures (supplementary fi gure 2). Combined the decrease in stem cell-associated markers and increase in differentiation markers, supports the notion that HDACi induces colon-CSCs to undergo differentiation.
149
TThesis_CD_Zimberlin.indbhesis_CD_Zimberlin.indb 149149 008/05/168/05/16 22:0522:05 Chapter 8
A. B.
50 1. 2. CSC_Ctrl 40 3. 1. 30 2. Dif_Ctrl 3. 20 1. 2. CSC_Pano
% Caspase-3 active cells 10 -1 0 1 3. C. LGR5 CD133 EPHB2 0 ActD - + -+--+-- -+- 12.5 9.8 CHX - - +--+-+- - -+ 11.0 9.7 Ctrl Oxali Pano Oxali+Pano 12.0 9.6 10.5 9.5 11.5
D. Expression 9.4 Ctrl 100 10.0 11.0 9.3 80 Pano - + - + - + - + - + - + CSC Dif CSC Dif CSC Dif ASCL2 KRT20 CEACAM5 60 13.0