A novel mouse model identifies cooperating mutations and therapeutic targets critical for chronic myeloid leukemia progression

The introduction of highly selective ABL-tyrosine kinase inhibitors (TKIs) has revolutionized therapy for chronic myeloid leukemia (CML). However, TKIs are only efficacious in the chronic phase of the disease and effective therapies for TKI-refractory CML, or after progression to blast crisis (BC), are lacking. Whereas the chronic phase of CML is dependent on BCR-ABL, additional mutations are required for progression to BC. However, the identity of these mutations and the pathways they affect are poorly understood, hampering our ability to identify therapeutic targets and improve outcomes. Here, we describe a novel mouse model that allows identification of mechanisms of BC progression in an unbiased and tractable manner, using transposon-based insertional mutagenesis on the background of chronic phase CML. Our BC model is the first to faithfully recapitulate the phenotype, cellular and molecular biology of human CML progression. We report a heterogeneous and unique pattern of insertions identifying known and novel candidate genes and demonstrate that these pathways drive disease progression and provide potential targets for novel therapeutic strategies. Our model greatly informs the biology of CML progression and provides a potent resource for the development of candidate therapies to improve the dismal outcomes in this highly aggressive disease.Chronic myeloid leukemia (CML) is a chronic myeloproliferative neoplasm, resulting from a reciprocal translocation between chromosomes 9 and 22, t(9;22)(q34;q11). This lesion was the first recurrent chromosomal abnormality described in cancer (Nowell and Hungerford, 1960; Rowley, 1973) and generates the BCR-ABL oncoprotein, a constitutively activated protein tyrosine kinase (TK; Deininger et al., 2000). Mouse models and human data have demonstrated BCR-ABL expression to be causative in CML (Daley et al., 1990; Heisterkamp et al., 1990; Zhao et al., 2001; Ramaraj et al., 2004; Koschmieder et al., 2005), and this observation has led to the paradigmic development of potent small molecule inhibitors that selectively target ABL enzymatic function and interrupt its oncogenic TK activity. Imatinib mesylate, the prototypic ABL tyrosine kinase inhibitor (TKI), and subsequent second and third generation TKIs, have revolutionized CML treatment (Druker et al., 1996; 2006; Carroll et al., 1997; Heinrich et al., 2000; O’Brien et al., 2003), significantly improving cytogenetic and molecular response rates, keeping the majority of patients in chronic phase, and prolonging overall survival (Druker et al., 2001, 2006; Sawyers et al., 2002; Hughes et al., 2003). However, despite this vast improvement, significant clinical challenges still remain in CML therapy. CML stem cells appear relatively resistant to the effects of TKIs (Copland et al., 2006; Jørgensen et al., 2007; Konig et al., 2008) such that, in the majority of patients, CML is controlled rather than cured. In addition, resistance occurs and this, together with stem cell persistence, facilitates disease transformation. Three distinct phases of the disease have been described. The initial phase, in which ∼85–90% of patients are diagnosed, is the indolent chronic phase (CP), which is readily amenable to treatment. However, without adequate therapy, this almost inevitably progresses to an aggressive acute leukemia of myeloid or lymphoid phenotype (70 and 30%, respectively), termed blast crisis (BC), which may be preceded by an ill-defined intermediate or accelerated phase (AP; during which the levels of myeloblasts in the BM or peripheral blood (PB) are increased but remain <20%). 10–15% of patients present beyond CP and a small percentage of CP cases continue to transform even on TKI therapy. The frequency of transformation is recorded at 3–5% within the first few years of TKI therapy but drops to ∼1% per year thereafter in randomized trials (Druker et al., 2006), although these values have been found to be higher in population-based studies (de Lavallade et al., 2008; Gallipoli et al., 2011). Treatment options for AP and BC are very limited, with response rates to TKIs lower and much less durable. Other options involve highly toxic therapies, such as combination chemotherapy and BM transplantation, and are not available or appropriate for many patients with progression. Therefore, even in the TKI era, the median survival of patients with BC is still dismal at around 6 mo (Hehlmann and Saussele, 2008; Silver et al., 2009), defining it as an unmet clinical need.Although the chronic phase of CML appears almost entirely dependent on BCR-ABL and CML is regarded as an invaluable model of leukemic evolution, the molecular mechanisms underlying disease progression are still poorly annotated. It is generally accepted that additional mutations cooperate with BCR-ABL during progression to BC (Calabretta and Perrotti, 2004), as is demonstrated by the observation that >75% of BC patients harbor additional cytogenetic abnormalities (Mitelman and Levan, 1978; Radich, 2007). There is also good evidence that the BCR-ABL protein itself contributes to the acquisition of further mutations, through its effects on reactive oxygen species induction, DNA damage, DNA repair, apoptosis, and cellular growth (Perrotti et al., 2010; Nieborowska-Skorska et al., 2012; Bolton-Gillespie et al., 2013), and the levels of BCR-ABL protein can indeed increase in the transition from CP to BC (Gaiger et al., 1995). However, to date, only a small number of mutations in specific pathways have been associated with disease progression in CML. For example, mutations or deletions in TP53, ASXL1, and RUNX1 are commonly described in myeloid BC at frequencies ranging between 3 and 25% (Ahuja et al., 1989; Grossmann et al., 2011), 15 and 20% (Boultwood et al., 2010; Grossmann et al., 2011), and 13 and 33% (Grossmann et al., 2011; Zhao et al., 2012) of cases, respectively. Similarly, mutations or deletions in the CDKN2A/B and IKAROS genes have been reported in up to 50 and 80% of patients in lymphoid BCs, respectively (Sill et al., 1995; Mullighan et al., 2008). Modern sequencing technologies and lowered costs have refined the mutational landscape for many tumors (Pleasance et al., 2010a,b; Curtis et al., 2012; Cancer Genome Atlas Research Network, 2013), but as yet have only been used in a directed fashion in CML (Piccaluga et al., 2009; Boultwood et al., 2010; Grossmann et al., 2011). Therefore, the spectrum of mutations that cooperate with BCR-ABL and the majority of pathways and processes that are corrupted by these mutations during the progression of CML to advanced phases, particularly for myeloid transformation, have yet to be fully described.Mouse models have greatly informed cancer biology in general and CML in particular. Several models have been previously generated, in which transgenic BCR-ABL expression is driven by several different promoters after either germline or retroviral integration (Hariharan et al., 1989; Castellanos et al., 1997; Honda et al., 1998; Huettner et al., 2000, 2003; Koschmieder et al., 2005). However, many of these models have failed to recapitulate the human disease by either generating predominantly acute lymphoid leukemias that lacked a preceding chronic phase, or a very rapidly fatal myeloproliferative neoplasm (MPN)–like disease not resembling the human counterpart (Daley et al., 1990; Honda et al., 1998; Huettner et al., 2000; Huettner et al., 2003). Models of BC have also been reported, where BCR-ABL expression has been combined with a known second hit, such as p53 or Dok1/Dok2 loss, or NUP98-HOXA9 or Hes1 overexpression (Skorski et al., 1996; Honda et al., 2000; Dash et al., 2002; Yasuda et al., 2004; Neering et al., 2007). Although confirmatory of the cooperation of specific mutations with BCR-ABL, these models have not informed the broader biology of BC due to their directed nature. Previous attempts to model random secondary mutations using retroviral insertional mutagenesis have also proven of limited value, with two reported studies only documenting three common insertions (Notch 1, Zfp423, and BCR-ABL; Mizuno et al., 2008; Miyazaki et al., 2009). Furthermore, the majority of these models have generated lymphoid leukemias, mainly T-ALL, thereby reducing their relevance for human disease.We therefore set out to generate a novel mouse model of CML progression that would allow us to identify mechanisms of BC progression in an unbiased and tractable manner. Here, we have combined a mouse transposon-based insertional mutagenesis system with a published transgenic mouse model of chronic phase CML (Koschmieder et al., 2005). For the first time, we report a BC model that closely mimics the natural progression of human CML and faithfully recapitulates the cellular and molecular aspects of its biology. We have identified known and novel candidate genes and pathways that, in combination with BCR-ABL, drive disease progression and could act as potential therapeutic targets in BC. Our novel model therefore defines mechanisms of CML progression, identifies therapeutic targets and provides a translational resource to improve clinical outcomes in this aggressive disease.     

Loss of bone mineral density in Chinese pre-menopausal women with systemic lupus erythematosus treated with corticosteroids

The adverse effect of disease and chronic corticosteroid therapy on bone mineral density (BMD) in patients with systemic lupus erythematosus (SLE) has been reported in several studies of Caucasian populations. As the factors controlling bone homeostasis may be different in Asian populations, we measured BMD in 52 pre-menopausal Chinese women (mean age 34.1 +/- 8.0 yr) with SLE (mean disease duration 6.4 +/- 4.5 yr) treated with prednisone (mean daily dose 11.4 +/- 10.8 mg/day). Lumbar spine, hip (total and subregions) and total body BMDs were measured in the SLE patients using dual-energy X-ray absorptiometry (DEXA), and compared with those from healthy controls matched for age, sex and body mass index. Compared to controls, SLE patients were found to have lower BMD (g/cm2) at several sites: the lumbar spine (0.98 vs 0.90, P = 0.001), Ward's triangle (0.72 vs 0.67, P = 0.03), total body (1.04 vs 1.01, P = 0.04) and total hip (0.87 vs 0.82, P = 0.05). There was no correlation between BMD at any region and duration of disease, activity of disease or prednisone therapy (mean daily dose, cumulative dose or treatment duration). When BMDs were compared between controls and SLE patients, subgrouped according to those not on calcium and those arbitrarily receiving calcium supplements (1 g/day), significantly lower BMDs were found in those not on calcium compared to both controls and SLE patients on calcium. BMDs in SLE patients on calcium were not different from those in controls. The low prevalence of osteoporosis in our SLE patients (4-6%) suggests significant loss of BMD in Chinese SLE patients on corticosteroid therapy is less than that reported in Caucasians (12-18%).      

Intracranial haemorrhage in typhoid fever

Intracranial haemorrhage in typhoid fever is very rare. We report another case of non-traumatic intracranial hemorrhage in a 6-year-old boy suffering from typhoid fever, unconsciousness, seizure and non-coherent speech. Investigations revealed severe thrombocytopenia and prolonged prothrombin time. CT scan of brain showed intraparenchymal haemorrhage in frontal regions bilaterally with perilesional oedema, subarachnoid bleed and extension into the lateral ventricles. No aneurysm or arterio-venous malformation was seen on MR angiography. The patient recovered without any neurological deficit.     

Limited mutational heterogeneity in the LDLR gene in familial hypercholesterolemia in Tunisia

Familial hypercholesterolemia (FH) is an autosomal dominant disease caused by mutations in the low-density lipoprotein receptor (LDLR), apolipoprotein B (APOB), and proprotein convertase subtilisin/kexin type 9 (PCSK9) genes. In previous studies, we have identified novel mutations in Tunisian FH families. In this study, we have extended our investigation to additional families. Five unrelated probands were screened for mutations in the LDLR and APOB genes, using direct sequencing and enzymatic restriction. We identified two novel LDLR mutations: a missense mutation in exon 7: p.Gly343Cys (c.1027G>T), and a nonsense mutation in exon 17: p.Lys816X (c.2446A>T). Using the PolyPhen and SIFT prediction computer programs the p.Gly343Cys is predicted to have a deleterious effect on LDL receptor activity. The missense mutation we found in exon 3, p.Cys89Trp (c.267C>G), has previously been identified in patients from United Kingdom and Spain, and is reported here for the first time in the Tunisian population. Finally, the framshift mutation in exon 10, p.Ser493ArgfsX44, is reported here for the fourth and fifth time in Tunisian families. The latter is the most frequent FH-causing mutation in Tunisia. These LDLR gene mutations enrich the spectrum of mutations causing FH in the Tunisian population. The framshift mutation, p.Ser493ArgfsX44, seems to be a founder mutation in this population.     

Four weeks of near-normalisation of blood glucose improves the insulin response to glucagon-like peptide-1 and glucose-dependent insulinotropic polypeptide in patients with type 2 diabetes

Objective  The incretin effect is attenuated in patients with type 2 diabetes mellitus, partly as a result of impaired beta cell responsiveness to glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1). The aim of the present study was to investigate whether 4 weeks of near-normalisation of the blood glucose level could improve insulin responses to GIP and GLP-1 in patients with type 2 diabetes. Methods  Eight obese patients with type 2 diabetes with poor glycaemic control (HbA_1c 8.6 ± 1.3%), were investigated before and after 4 weeks of near-normalisation of blood glucose (mean blood glucose 7.4 ± 1.2 mmol/l) using insulin treatment. Before and after insulin treatment the participants underwent three hyperglycaemic clamps (15 mmol/l) with infusion of GLP-1, GIP or saline. Insulin responses were evaluated as the incremental area under the plasma C-peptide curve. Results  Before and after near-normalisation of blood glucose, the C-peptide responses did not differ during the early phase of insulin secretion (0–10 min). The late phase C-peptide response (10–120 min) increased during GIP infusion from 33.0 ± 8.5 to 103.9 ± 24.2 (nmol/l) × (110 min)⁻¹ (p < 0.05) and during GLP-1 infusion from 48.7 ± 11.8 to 126.6 ± 32.5 (nmol/l) × (110 min)⁻¹ (p < 0.05), whereas during saline infusion the late-phase response did not differ before vs after near-normalisation of blood glucose (40.2 ± 11.2 vs 46.5 ± 12.7 [nmol/l] × [110 min]⁻¹). Conclusions  Near-normalisation of blood glucose for 4 weeks improves beta cell responsiveness to both GLP-1 and GIP by a factor of three to four. No effect was found on beta cell responsiveness to glucose alone. ClinicalTrials.gov ID no.: NCT 00612950 Funding: This study was supported by The Novo Nordisk Foundation, The Medical Science Research Foundation for Copenhagen.     

Abeta(1-42) reduces synapse number and inhibits neurite outgrowth in primary cortical and hippocampal neurons: a quantitative analysis

Synaptic loss represents one of the earliest signs of neuronal damage and is observed within both Alzheimer's disease patients and transgenic mouse models of the disease. We have developed a novel in vitro assay using high content screening technology to measure changes in a number of cell physiological parameters simultaneously within a neuronal population. Using Hoechst-33342 to label nuclei, betaIII-tubulin as a neuron-specific marker, and synapsin-I as an indicator of pre-synaptic sites, we have designed software to interrogate triple-labelled images, counting only those synaptic puncta associated with tubulin-positive structures. Here we demonstrate that addition of amyloid beta peptide (Abeta(1-42)), to either primary hippocampal or cortical neurons for 4 days in vitro has deleterious effects upon synapse formation, neurite outgrowth and arborisation in a concentration-dependent manner. Control reverse peptide showed no effect over the same concentration range. The effects of Abeta(1-42) were inhibited by D-KLVFFA, which contains residues 16-20 of Abeta that function as a self-recognition element during Abeta assembly and bind to the homologous region of Abeta and block its oligomerisation. These effects of Abeta(1-42) on synapse number and neurite outgrowth are similar to those described within AD patient pathology and transgenic mouse models.