Non-homologous end joining mediated DNA repair is impaired in the NUP98-HOXD13 mouse model for myelodysplastic syndrome

Abstract

Chromosomal translocations typically impair cell differentiation and often require secondary mutations for malignant transformation. However, the role of a primary translocation in the development of collaborating mutations is debatable. To delineate the role of leukemic translocation NUP98-HOXD13 (NHD13) in secondary mutagenesis, DNA break and repair mechanisms in stimulated mouse B lymphocytes expressing NHD13 were analyzed. Our results showed significantly reduced expression of non-homologous end joining (NHEJ)-mediated DNA repair genes, DNA Pkcs, DNA ligase4, and Xrcc4 leading to cell cycle arrest at G2/M phase. Our results showed that expression of NHD13 fusion gene resulted in impaired NHEJ-mediated DNA break repair.

1. Introduction

Chromosomal translocations (CT) are the hallmark features associated with many hematological malignancies [1]. The majority of CT are considered class II mutations resulting in impaired hematopoietic cell differentiation but require collaborating secondary mutations for complete malignant transformation [1–3]. Class II mutations target transcription factors and function predominately to impair hematopoietic differentiation and subsequently apoptosis [3,4]. However, the mechanisms by which the cell acquires secondary mutations are not completely understood. Increased DNA damage and/or impaired DNA repair mechanisms are widely accepted as mechanisms for genomic mutations [5,6]. An emerging concept is that CT might increase the chance for further DNA damage or that the fusion gene can somehow impair the DNA repair mechanisms [7–9]. Of the different types of DNA damage to occur, one of the most important types is the double strand breaks (DSB), which occurs in each dividing cell at an estimated rate of 10 breaks per day [10]. These breaks are typically repaired by non-homologous end joining (NHEJ) which is a tightly controlled process involving multiple factors [11]. Ironically, impaired NHEJ can lead to mutations including the formation of CT [12,13]. Delineating DNA break induction and repair pathways in cells with primary CT during leukemic progression will identify the potential mechanism by which CT induce secondary mutations. These mechanisms may be through direct DNA–DNA interactions, fusion protein–DNA interactions, or fusion protein–protein interactions. By doing so, identification of these mechanisms could reveal potential therapeutic targets.

Transgenic mice expressing the leukemic fusion gene NUP98-HOXD13 (NHD13) progress through myelodysplastic syndrome (MDS) and develop acute leukemia with the occurrence of collaborating mutations [14]. We have recently shown that expression of NHD13 in stimulated B-lymphocytes results in: (1) impaired B-cell differentiation at stages in the bone marrow that are dependent on VDJ recombination and (2) impaired class switch recombination (CSR) and antibody production [15]. During CSR naïve B lymphocytes undergo DNA breaks on immunoglobulin heavy chain gene and NHEJ-mediated break repair when stimulated with appropriate antigens and cofactors to produce different immunoglobulin isotypes [16]. DNA DSB results in the phosphorylation of histone H2AX by Ataxia telangiectasia mutated (ATM), which in turn results in the stabilization of break ends [17,18]. Proteins Ku70 and Ku80 bind to the break ends to form a heterodimer and recruits DNAP-Kcs [19]. DNAPKcs acts as a scaffold and DNA repair factors DNA ligase 4 and XRCC4 complex [20]. As CSR involves physiological DNA breaks and NHEJ mediated repair, expression of the NHD13 fusion in B lymphocytes is a suitable system for delineating the mechanism by which CT can induce DNA instability and result in secondary mutations. Based on these observations, we hypothesized that NHD13 enhances double strand break induction and impairs the DNA repair mechanism as an underlying reason for secondary mutations. Here we used CSR as a tool to induce DNA DSB repair in the presence of NHD13. Our results show that NHD13 fusion gene impairs expression of critical NHEJ repair factors, and impairs DNA repair mechanism resulting in cell cycle arrest.

2. Materials and methods

2.1. Animals

Five young (8–12 weeks) NHD13 mice and five wild type (WT) littermates on an FVB background were used for each experiment. Mice were bred and maintained at the AAALAC accredited animal facility at Virginia Maryland Regional College of Veterinary Medicine, Virginia Tech. All experiments were carried out as per NIH guidelines with prior approval from the Virginia Tech Institutional Animal Care and Use Committee.

2.2. In vitro class switch recombination

To induce in vitro class switch recombination, splenic B lymphocytes were harvested using anti-mouse IgM magnetic beads and magnetic assisted cell sorting (Milteny Biotec, Auburn, CA). A total of 2 × 10⁵ cells were treated with 5 μM CFSE and cultured with media containing E. coli Lipopolysaccharide (LPS, Sigma–Aldrich, St. Louis, MO) (25 μg/ml) and IL-4 (PeproTech, Rocky Hill, NJ) (25 ng/ml). To verify that CSR had occurred, cells were harvested at 72 h, labeled with anti-mouse IgG1 and IgE antibodies and analyzed by flow cytometry as previously published [15].

2.3. DNA break analysis by confocal microscopy

Splenic B lymphocytes were harvested and cultured to induce CSR as described above. Cells were harvested at 0, 24, 48 and 72 h and cytospun onto charged slides. Cells were fixed using 4% paraformaldehyde (Thermo Scientific, Rockford, IL) and incubated with rabbit anti-mouse γH2AX (Cell Signaling, Danvers, MA) at 1:500 dilution in TBS + 1% FBS followed by labeling with Alexafluor-488 conjugated anti-rabbit antibody (Cell Signaling, Danvers, MA). After labeling with DAPI (Cell Signaling), cells were visualized using LSM700 Carl Zeiss confocal microscope (Carl Zeiss Microimaging, Thornwood, NY) and LSM900^® software (Carl Zeiss).

2.4. DNA break analysis by flow cytometry

Splenic B lymphocytes cultured in the presence of LPS (Sigma–Aldrich) and IL-4 (Sigma–Aldrich) were harvested at 0, 24, 48 and 72 h. Cells were fixed with 70% ethanol and incubated with rabbit anti-mouse phosphoH2AX antibody (Cell Signaling) followed by incubation with Alexa-fluor 488 conjugated anti-rabbit antibody (Cell Signaling). Cells were analyzed using a FACScan flow cytometer (BD Biosciences).

2.5. Cell cycle analysis

B lymphocytes from NHD13 and WT mice were harvested and cultured in the presence of LPS and IL-4 as described above. Cells were harvested at 24 h intervals, fixed in 70% ethanol and labeled with propidium iodide (PI) (Sigma–Aldrich). Cells were analyzed by flow cytometry.

2.6. Gene expression analysis

Lymphocytes stimulated with LPS + IL-4 were harvested at 72 h of culture, total RNA was extracted and cDNA was synthesized. Gene expression was determined by reverse quantitative PCR using gene specific primers and Sybrgreen mastermix (Applied Biosystems). Gapdh was used as the internal control.

2.7. Data and statistical analysis

Flow cytometric data were analyzed using FlowJo software 7.6 (FlowJo, Ashland, OR). Confocal images were analyzed using Zen 2009^® software (Carl Zeiss Microimaging). Data was analyzed with GraphPad Prism 5.0^® (GraphPad Software, La Jolla, CA), using either two tailed t-test or ANOVA and Bonferroni post test; a p value < 0.05 was considered significant.

3. Results

3.1. B cells from NHD13 mice had reduced DNA double strand break repairs

Class switch recombination essentially involves DNA double strand breaks and NHEJ-mediated DNA recombination. To delineate the effect of NHD13 on the pattern of DNA DSB during CSR, we analyzed stimulated B cells from NHD13 and WT mice for DNA breaks using flow cytometry and confocal microscopy. Lymphocyte stimulation using E. coli lipopolysaccharide (LPS) and IL-4 results in expression of activation induced deaminase (Aid), which will initiate DNA DSB. These DNA DSB result in phosphorylation of histone H2AX (γH2AX), and can be employed as a marker for DNA DSB. Flow cytometric analysis for γH2AX revealed a comparable percentage of cells with DNA breaks in NHD13 and WT mice at 0, 24 and 48 h (Fig. 1A and B). However, at 72 h, NHD13 mice had increased DNA DSB as evidenced by flow cytometric analysis (Fig. 1A and C). We also analyzed the pattern of breaks at the single cell level using confocal microscopy (Fig. 1B). The DNA repair efficiency was estimated and expressed as the percentage reduction in break positive cells between 48 and 72 h. Our results showed a significantly lower DNA repair efficiency in NHD13 B cells (Fig. 1D). These results indicate that NHD13 B cells have comparable DNA double strand break induction following stimulation, but have reduced efficiency for break repair during physiological class switch recombination in B cells.

Fig. 1.

B cells from NHD13 mice have impaired DNA double strand break repair following stimulation. LPS and IL-4 stimulated splenic B cells from NHD13 and WT mice were harvested at 24 h intervals and labeled for γH2AX, a marker for DNA double strand breaks. (A) Representative flow cytometric pattern of DNA double strand breaks at 24 h intervals. (B) Confocal analysis of DNA double strand breaks showing DNA breaks in NHD13 and WT B-cells at 72 h. (C) Analysis of the percentage of DNA breaks using flow data showed significantly high levels in NHD13 B-cells at 72 h following stimulation. (D) DNA repair efficiency calculated based on the percentage of breaks repaired between 48 and 72 h showed significantly lower repair efficiency in NHD13 B-cells. n = 5, p < 0.05.

3.2. NHD13 B cells had a G2/M cell cycle arrest following stimulation

To better understand the impact of NHD13 on DNA repair, we determined the cell cycle kinetics of stimulated B cells from NHD13 mice at 24 h intervals. B cells were stimulated with LPS and IL-4, harvested at 24 h intervals and labeled with PI for flow cytometric analysis. Our results showed a comparable cell cycle pattern in NHD13 and WT cells at 0, 24 and 48 h following stimulation as indicated by a comparable percentage of apoptotic cells and cells in G1, S and G2/M phases (Fig. 2A and B). However, at 72 h of stimulation, NHD13 mice had a significantly higher percentage of cells in the G2/M stage (Fig. 2A and B) indicating cell cycle arrest at this stage. Cells in other stages, apoptotic, G1 and S phase were altered in NHD13 mice at 72 h, but not significant (Fig. 2B). We assayed expression of the cell cycle regulator gene p53 and found that its expression was higher at 72 h in NHD13 B cells (not statistically significant, Fig. 2C). Our results suggest that the NHD13 fusion gene impairs the DNA break repair mechanism and results in a G2/M check point arrest during the cell cycle in stimulated B-cells.

Fig. 2.

Cell cycle is arrested at G2/M phase in B-cells from NHD13 mice following stimulation. Isolated splenic B cells were stimulated with LPS and IL-4 to induce CSR. Cells were harvested at 24 h intervals and stained with propidium iodide to analyze cell cycle pattern. (A) Representative cell cycle plots from NHD13 and WT cells at 24 h intervals, 0 indicating prior to stimulation. (B) Statistical analysis of different phases of cell cycle pattern showed comparable kinetics at 0, 24 and 48 h. However, at 72 h, cells in G2 phase were significantly high in NHD13 mice suggesting a cell cycle arrest. (C) Expression analysis of p53 showed elevated levels in NHD13 lymphocytes 72 h following stimulation. n = 5, **p < 0.01.

3.3. DNA ligase 4, XRCC4, and DNAPKcs were downregulated in NHD13 B lymphocytes

Considering the essential role of NHEJ during DNA double strand repair, we hypothesized that NHD13 fusion gene might interfere with the NHEJ mechanism. To test this assumption, we harvested RNA from stimulated B cells at 72 h of culture and assayed the gene expression of critical genes involved in NHEJ mechanism. Our results showed comparable levels of Ku70 and Ku80 in cells from both NHD13 and WT mice (Fig. 3A). However, DNA ligase 4, Xrcc4, and DNAPKcs were significantly lower in NHD13 B lymphocytes. We also considered the possibility that Alternative End Joining (AEJ) could play a role during the DNA repair process. We evaluated expression of possible AEJ factors including Ligase 1 and Ligase 3α (Fig. 3B) and found that there was no statistical significance between transgenic and wild type cells. Our results indicate that expression of NHD13 fusion gene results in reduced expression of DNA ligase 4, Xrcc4, and DNAPKcs, leading to reduced DNA repair.

Fig. 3.

DNA ligase 4 and DNAPKcs expression is downregulated in NHD13 B cells following stimulation. LPS and IL-4 stimulated splenic B cells were harvested at 72 h following stimulation and, RNA was harvested for RQPCR. (A) Expression analysis of critical NHEJ genesf showed that Ku70, and Ku80 were comparable between NHD13 and WT B cells. However, DNA ligase 4, Xrcc4 and DNAPKcs were significantly low in NHD13 mice following stimulation. (B) Expression analysis of AEJ factors Ligase 1 and Ligase 3 showed comparable levels in both WT and NHD13 TG B lymphocytes. n = 5, p < 0.05.

4. Discussion

Chromosomal translocations are hallmark features of hematopoietic malignancies, which generally require collaborating mutations for malignant transformation [4]. Recent studies have proposed that CT can induce secondary mutations before malignant transformation [21,22]. A wide variety of factors can cause DNA damage, promote misrepairs and lead to the development of mutations [23]. Of these etiologies, DNA DSB [24] are frequent in mammalian cells as a result of physiological and pathological factors including reactive oxygen species, ionizing radiations or chemicals [25,26], and are repaired mainly by NHEJ [5]. Effective repair is essential for the maintenance of genomic stability and cell viability, whereas defective break repair results in cell cycle arrest, apoptosis, mutagenesis and malignant transformation [7,27]. Here, we analyzed the effects of a primary leukemic fusion gene, NHD13, on the DNA DSB repair pathways. Previous studies in transgenic mice expressing NHD13 have shown that leukemic progression is accompanied by random mutations in a wide variety of collaborating genes [14]. Taken together, it is reasonable to believe that the presence of a primary leukemic fusion gene can induce DNA instability by impairing the NHEJ pathway and lead to the formation of secondary mutations.

Analysis of lymphocyte development and gene recombination pattern in NHD13 mice suggests that the VDJ recombination mechanism is perturbed in NHD13 lymphocytes [28]. Furthermore our previous studies have shown that B lymphocyte differentiation and CSR are altered in NHD13 mice, with partial blocks occurring during gene recombination events [15]. Based on these findings, we used CSR as a tool to determine the role of NHD13 to induce DNA breaks and thereby analyze the NHEJ-mediated break repair pattern. As CSR involves DNA double strand break induction and NHEJ-mediated repair mechanisms, delineating the underlying molecular pathways of impaired CSR can have wider applications [29]. Our current study showed a comparable percentage of cells with DNA breaks prior to stimulation and at 24 and 48 h following stimulation (Fig. 1A and B). However, the percentage of cells with DNA breaks was significantly higher at 72 h of culture, suggesting an impaired DNA repair mechanism. Single cell analysis for DNA breaks using confocal microscopy showed a comparable amount of breaks in both WT and TG B cells, suggesting that the breaks are induced in a similar pattern. DNA break repair efficiency calculated based on the percentage of repair in a 24 h time frame (Puthiyaveetil et al., 2013, Mol Immunol, in press) showed a significantly lower repair in stimulated cells from NHD13 mice between 48 and 72 h (Fig. 1D), suggesting that DNA repair is impaired in NHD13 mice rather than break induction. Analysis of cell cycle pattern showed a G2/M stage cell cycle arrest in NHD13 B lymphocytes at 72 h of stimulation. Consistent with previous studies that have shown DNA repair defects will result in cell cycle arrest at G2/M checkpoint [30,31], our results also indicate that cell cycle arrest can be mediated by NHD13 leading to, or as a consequence of, defective or inefficient DNA repair.

Our analysis of expression of DNA repair factors showed comparable levels of Ku70, Ku80 and, but significantly lower levels of DNA Ligase 4, Xrcc4, and DNAPKcs at 72 h following stimulation. Reduced expression of repair factors has been shown to impair classical NHEJ pathway, leading to more error prone repair mechanisms, increasing the chance for secondary mutations [9]. During NHEJ-mediated DNA repair, Ligase 4 expression is often associated with increased AEJ activity. However, in our model, the C-NHEJ factors DNA Ligase 4, Xrcc4, and DNAPKcs are down regulated without any substantial increase in expression of AEJ factors, i.e., Ligase 1 and Ligase 3α. Considering the phosphorylation of H2AX, an early break detection process, is intact but that NHEJ factors are not up-regulated, we propose that NHD13 impairs DNA break repair induction. In conclusion, our results showed that the myeloid leukemic fusion gene NHD13 suppressed the expression of DNA Ligase 4, Xrcc4, and DNAPKcs, DNA double strand break accumulation and cell cycle arrest as probable underlying mechanisms for secondary mutations. Additional studies are warranted to more specifically determine the molecular mechanisms by which expression of NHD13 regulates the activity of DNA Ligase 4, XRCC4, and DNAPKcs. Doing so could reveal novel pathways or therapeutic targets useful in the treatment of leukemia and other types of cancer.