Brief Report: Human Pluripotent Stem Cell Models of Fanconi Anemia Deficiency Reveal an Important Role for Fanconi Anemia Proteins in Cellular Reprogramming and Survival of Hematopoietic Progenitors†‡§
Sun K. Yung,
Institute of Genetic Medicine, Newcastle University, Newcastle, United Kingdom
NESCI, Newcastle University, Newcastle, United Kingdom
Author contributions: S.Y., K.T., and M.L.: performed experiments, data analysis, and final approval of manuscript; S.H., I.N., C.S., and G.S.: performed some of the experiments and final approval of manuscript; M.S.: conception and design, fund raising, and final approval of manuscript; L.A.: conception and design, manuscript writing, fund raising, and final approval of manuscript; S.P.: performed some of the experiments, collection and analysis of the data, contributed to manuscript writing, and final approval of manuscript; M.L.: conception and design, performed experiments, data analysis, manuscript writing, fund raising, and final approval of manuscript. S.K.Y. and K.T. contributed equally to this article.
Disclosure of potential conflicts of interest is found at the end of this article.
First published online in STEM CELLSEXPRESS December 27, 2012.
Fanconi anemia (FA) is a genomic instability disorder caused by mutations in genes involved in replication-dependant-repair and removal of DNA cross-links. Mouse models with targeted deletions of FA genes have been developed; however, none of these exhibit the human bone marrow aplasia. Human embryonic stem cell (hESC) differentiation recapitulates many steps of embryonic hematopoietic development and is a useful model system to investigate the early events of hematopoietic progenitor specification. It is now possible to derive patient-specific human-induced pluripotent stem cells (hiPSC); however, this approach has been rather difficult to achieve in FA cells due to a requirement for activation of FA pathway during reprogramming process which can be bypassed either by genetic complementation or reprogramming under hypoxic conditions. In this study, we report that FA-C patient-specific hiPSC lines can be derived under normoxic conditions, albeit at much reduced efficiency. These disease-specific hiPSC lines and hESC with stable knockdown of FANCC display all the in vitro hallmarks of pluripotency. Nevertheless, the disease-specific hiPSCs show a much higher frequency of chromosomal abnormalities compared to parent fibroblasts and are unable to generate teratoma composed of all three germ layers in vivo, likely due to increased genomic instability. Both FANCC-deficient hESC and hiPSC lines are capable of undergoing hematopoietic differentiation, but the hematopoietic progenitors display an increased apoptosis in culture and reduced clonogenic potential. Together these data highlight the critical requirement for FA proteins in survival of hematopoietic progenitors, cellular reprogramming, and maintenance of genomic stability. STEM CELLS 2013;31:1022–1029
Fanconi anemia (FA) is caused by mutations in genes involved in replication-dependant repair and is characterized by congenital abnormalities, bone marrow failure, increased risk of hematological malignancies, radiosensitivity, and premature ageing [1–4]. Fifteen complementation groups and the encoding genes have been indentified so far [5–8] and functionally studied in details [9–20].
Several mouse models with targeted deletions of Fanca [21, 22], Fancc [23, 24], Fancd2 , and Fancg  have been generated; however, none of these models exhibit human marrow aplasia, showing that they do not resemble closely the hematopoietic phenotype in FA. Human embryonic stem cell (hESC) differentiation recapitulates many steps of embryonic hematopoietic development and is a useful model system to investigate the early events of hematopoietic stem cell specification . Knockdown of FANCA and FANCD2 in hESC results in reduced numbers of hematopoietic progenitors, suggesting an important role for the FA genes during human embryonic hematopoiesis . It is now possible to derive human-induced pluripotent stem cells (hiPSC) from patients with various genetic disorders. This approach has been met with various degrees of success in the FA field. Raya et al. reported that generation of hiPSC lines from FA-A and FA-D2 patients was only possible after genetic complementation , while a more recent study indicated that reprogramming barrier of FA fibroblasts could be bypassed by reprogramming under hypoxic conditions . In this study, we report derivation of hiPSC lines from a FA-C patient under normoxic conditions and show that both the FA-C patient-specific hiPSC lines and the hESC with stable knockdown of the FANCC gene result in reduced number of clonogenic hematopoietic progenitors due to increased apoptosis in culture.
MATERIALS AND METHODS
Derivation of hiPSC lines was carried out using a polycystronic lentiviral system (Allele Biotech, San Diego, California, USA) as described in detail in Jiang et al. .
Generation of hESC lines with stable knockdown of FANCC was carried out using ready-made lentiviral particles (multiplicity of infection = 0.5) encoding the following shRNAs embedded within the pLKO.1-puro vector (Sigma, Dorset, UK): FANCC-shRNA 1: 5′-CCGGGTGAGAG AAATTGTCTGAGAACTCG-3′; FANCC-shRNA 2: CCGGCACGAGATCATTGGCTTTCTTCTCGA. Control hESCs were created by transfecting a β2M shRNA control (SHC008). Stable clones were isolated using puromycin selection (0.6 μg/ml). Downregulation of FANCC was confirmed using quantitative RT-PCR (FANCC forward: 5′-GGCCCTGCTGTGGCTCTTGG-3′ and FANCC reverse: 5′-GCACGGCCTTCACCTGGACC-3′) and Western blotting (FANCC antibody: Novus Biologicals, Cambridge, UK, Cat. Number: H00002176-M01) using methods described in Yung et al. . FANCD2 foci analysis was carried out as described in Raya et al. .
Apoptosis and proliferation of progenitors was carried out using a flow cytometric kit (cat. no. 562253; BD Biosciences, San Jose, California, USA) and following manufacturer's instructions. In brief, hESC and hiPSC were labeled with BrdU, then fixed, permeabilized, and treated with DNase I. Following this treatment, cells were simultaneously stained with PerCP-Cy5.5-labeled BrdU to detect cells entering and progressing through S-phase and Phycoerythrin-labeled anticleaved Poly (ADP-ribose) polymerase to detect apoptotic cells. Cell surface antibodies were added to select desired subpopulations and analyzed by flow cytometry using the LSRII flow cytometer. At least 10,000 events were recorded for each sample and results were analyzed using FACS DIVA software. Hematopoietic differentiation of hESC/hiPSC lines and characterization of hematopoietic progenitors was carried out as described by Yung et al. . Teratoma analysis was carried out as described in Jiang et al. . Double-strand break quantification by γH2A.X staining was performed as described in Saretzki et al. .
All results are shown as mean ± SEM. Differences between groups were assessed using two-tailed Student's t test.
RESULTS AND DISCUSSION
Primary fibroblasts from several FA patients (complementation groups A, C, D2, and G; supporting information Table S1) were transduced with lentiviruses expressing the polycystronic OSKM (OCT4, SOX2, KLF4, and c-MYC) cassette under normoxic conditions (Fig. 1A). Although hiPSC clones were derived from FA-A, FA-C, and FA-D2 patients (Fig. 1B; supporting information Table S1), only FA-C patient-specific hiPSC lines could be maintained in culture for an extended number of passages. These findings are consistent with the studies published by Raya et al.  and Müller et al.  which suggested that derivation and long-term maintenance of FA patient-specific hiPSC under normoxic conditions are unfeasible [29, 30]. However, derivation of FA-A and FA-C patient-specific hiPSC lines with normal karyotype under hypoxic conditions was reported by Müller et al. . This raises the question of how we were able to derive and maintain patient-specific FA-C hiPSC under normoxic conditions. The patient samples used in this manuscript were different to the ones used in the previous two studies, hence there may be differences between patients in the ability to reprogram to hiPSC even if they belong to the same FA complementation group depending on the nature of the mutation and residual DNA repair ability. For most FA groups, it was possible to work with only one patient-specific sample; however, in the case of FA-A group, we were able to obtain nine samples (age range 12–52 years old), but able only to derive two hiPSC clones from one sample (supporting information Table S1), suggesting that variability between patients samples is a contributing factor. The age of the FA donor and accumulation of chromosomal instabilities could be another factor behind the success rate of hiPSC derivation. This is based on the simple observation that the primary fibroblast cells from the FA-C patient which resulted in successful hiPSC derivation were obtained from a young donor (3 years old) when compared with the other three groups (13, 7, and 14-year old, respectively, refer to supporting information Table S1). It is also worth noting that both FA-C patient-specific hiPSC lines had a very abnormal karyotype, suggesting that successful reprogramming may also be due to occurrence of unknown genetic events that compensate for the deficient FA pathway.
In all cases, FA fibroblasts reprogrammed to hiPSC with lower efficiency when compared with controls (supporting information Table S1), indicating the relevance of a functional FA pathway during the reprogramming process. This was not due to differences in transduction efficiency as control transduction with a green fluorescent protein encoding lentiviral construct showed no significant differences between control and FA fibroblasts (supporting information Fig. S1). Addition of two factors (LIN28 and NANOG) or three (LIN28, NANOG, and TERT) to the polycystronic OSKM resulted in 1.9 and 7.4-fold increase, respectively, in reprogramming efficiency when compared with four factor reprogramming (data not shown).
Quantitative RT-PCR indicated silencing of the exogenous transgenes (Fig. 1C) in hiPSC clones. Both FA-C hiPSC clones behaved very similarly in teratomas formation tests, karyotype analysis, and hematopoietic differentiation assays. For this reason, we have either shown representative examples of one clone (Figs. 1F, 2A for example) or averaged the data from both hiPSC clones (Fig. 3C for example) and have indicated this in each figure legend. In parallel, hESC lines with stable knockdown of FANCC and control were generated using a combination of two different shRNA for FANCC and a single shRNA for β2m (β-2 macroglobulin), respectively (Fig. 1D, 1E).
All resulting hiPSC/hESC were morphologically indistinguishable and expressed all the tested pluripotency markers (Fig. 1E). DNA fingerprinting analysis (supporting information Fig. S2) confirmed the genetic identity of FANCC hiPSC to parent fibroblasts (supporting information Fig. S2). Suspension culture of hiPSC/hESC resulted in differentiation into cells derived from all three germ layers (Fig. 2A). Injection of FANCC-hiPSC cells into immunocompromised mice resulted in formation of masses, which were not characteristic of normal teratoma tissue and consisted primarily of immune cells (Fig. 2B, panels j–l). In contrast, the control hESC, control hiPSC, and hESC with stable knockdown of FANCC lines gave rise to typical teratoma containing cells derived from all three germ layers (Fig. 2B, panels a–i).
Karyotype analysis indicated an abnormal karyotype for FANCC fibroblasts (Fig. 3A). The resulting hiPSC clones (Fig. 3A) displayed the same abnormalities as the parent fibroblasts in addition to significant additional abnormalities (both clones were hyperploid). However, the hESC clones with stable knockdown of FANCC or β2m showed a normal karyotype up to 30 passages in culture (data not shown). Together these data suggest that the most likely cause for the occurrence of additional chromosomal abnormalities is the DNA damage associated with the reprogramming process as shown recently by Müller et al. .
Treatment with mitomycin C resulted in activation of FA pathway in control hESC and hiPSC leading to the formation of FANCD2 foci (supporting information Fig. 3); however, the number of FANCD2 foci was significantly reduced (93% and 90%, respectively, in the FANCC hiPSC or hESC with stable knockdown of FANCC when compared with their respective controls). Hypersensitivity of FANCC-deficient cells to mitomycin C also resulted in a significant increase in the percentage of cells with double strand breaks (DSBs) in the FANCC-deficient hESC/hiPSC clones (Fig. 3C), while the average number of DSB/nuclei remained similar between those and controls (Fig. 3D).
All hESC and hiPSC lines were differentiated in the presence of BMP4, basic fibroblast growth factor, and vascular endothelial growth factors as described in one of our recent publications . The emergence of hemato-endothelial progenitors (CD34+CD31+ CD45−), primitive (CD34+CD45+), and total blood cells (CD45+) was analyzed throughout embryoid body (EB) development. Hematopoietic specification from FANCC-deficient hESC and hiPSC lines was similar (Fig. 4A); however, analysis of clonogenic potential (Fig. 4B) indicated a significant reduction in the ability of FANCC-deficient hESC and hiPSC lines to form hematopoietic colonies. Flow cytometry analysis of cells taken from hematopoietic colonies arising in methylcellulose assays indicated no change in proliferation rate of committed progenitors, but an increase in apoptosis (31% and 27% increase in FANCC hiPSC or hESC with stable knockdown of FANCC when compared with their respective controls; data not shown).
Apoptosis assays of the hemato-endothelial progenitors (CD34+CD31+CD45−) indicated higher levels of apoptosis in FANCC-deficient hESC/hiPSC when compared with equivalent cells from respective controls (Fig. 4C). Hematopoietic progenitors derived from FANCC-deficient hESC/hiPSC lines showed an increased proliferation in contrast to other cell types within the EBs (Fig. 4D), suggesting perhaps a feedback loop between higher apoptosis in these cells and demand to replenish the progenitor cell population.
Together our data highlight an important role for FANCC gene in the maintenance of genomic stability during cellular reprogramming process and survival of emerging hematopoietic progenitors. They also highlight the usefulness of human pluripotent stem cell models in mimicking the disease phenotype and providing future platforms for high throughput screening of compounds that may ameliorate the FA phenotype.
We would like to thank Ian Dimmick and Dr. Owen Hughes for their help with the flow cytometric analysis, Complement Genomics plc. for carrying out the DNA fingerprinting analysis, Gavin Cuthbert and Alicia Madgwick for performing the karyotyping, Dr. Susana Navarro and Prof. Juan Bueren for carrying out the FANCD2 foci analysis, and Coriell Cell Repositories for providing the FA patient fibroblasts. This study was supported by Leukemia and Lymphoma Research Grant 10050, Fanconi Anemia Research Fund U.S., Fanconi Hope U.K., and funds for research in the field of Regenerative Medicine through the collaboration agreement from the Conselleria de Sanidad (Generalitat Valenciana) and the Instituto de Salud Carlos III (Ministry of Science and Innovation), Spain.
DISCLOSURE OF POTENTIAL CONFLICTS OF INTERESTS
The authors indicate no potential conflicts of interest.