A knockout mutation associated with juvenile paroxysmal dyskinesia in Markiesje dogs indicates SOD1 pleiotropy.

A juvenile form of paroxysmal dyskinesia segregated in the Markiesje dog breed. Affected pups exhibited clinical signs of a severe tetraparesis, dystonia, cramping and falling over when trying to walk. In most cases, the presentation deteriorated within weeks and elective euthanasia was performed. Pedigree analysis indicated autosomal recessive inheritance. Genome-wide association and homozygosity mapping of 5 affected dogs from 3 litters identified the associated locus on chromosome 31 in the region of SOD1. The DNA sequence analysis of SOD1 showed that the patients were homozygous for a frameshift mutation in the fourth codon. None of the other analyzed dogs of the breed was homozygous for the mutation, indicating full penetrance of the genetic defect. Mutations in SOD1 are known to cause recessive degenerative myelopathy in middle-aged dogs with low penetrance and dominant amyotrophic lateral sclerosis in humans with variable age of onset. Our findings are similar to recent observations in human patients that a loss of function mutation in SOD1 leads to a juvenile neurologic disease distinct from amyotrophic lateral sclerosis.

A novel VWF variant associated with type 2 von Willebrand disease in German Wirehaired Pointers and German Shorthaired Pointers.

Von Willebrand disease (VWD), caused by deficiency of the von Willebrand factor (VWF), is the most common bleeding disorder in humans and dogs. The complete cDNA encoding VWF of a German Wirehaired Pointer with type 2 VWD was sequenced, and we found four variants that alter the amino acid sequence. These variants were: c.1657T>G corresponding to p.Trp553Gly; c.1777G>A (p.Glu593Lys); c.4937A>G (p.Asn1646Ser) and c.5544G>A (p.Met1848Ile). A haplotype of the c.1657G, c.1777A and c.4937G alleles co-segregated with the VWF antigen level in a four-generation pedigree with the disease. Healthy dogs of the breed were found that were homozygous for the c.1777A or the c.5544A allele, indicating that these variants do not cause VWD. Dogs that were homozygous for the c.4937G allele and had no signs of a bleeding disorder were observed in the Chinese Crested dog breed. Thus, only the c.1657G variant was found in the homozygous state exclusively in VWD affecteds, and this variant is the strongest candidate to be the cause of VWD type 2 in the German Wirehaired Pointer breed. A screen of German Shorthaired Pointers indicated that the variant also segregates with VWD in this breed.

The canine era: the rise of a biomedical model.

Since the annotation of its genome a decade ago, the dog has proven to be an excellent model for the study of inherited diseases. A large variety of spontaneous simple and complex phenotypes occur in dogs, providing physiologically relevant models to corresponding human conditions. In addition, gene discovery is facilitated in clinically less heterogeneous purebred dogs with closed population structures because smaller study cohorts and fewer markers are often sufficient to expose causal variants. Here, we review the development of genomic resources from microsatellites to whole-genome sequencing and give examples of successful findings that have followed the technological progress. The increasing amount of whole-genome sequence data warrants better functional annotation of the canine genome to more effectively utilise this unique model to understand genetic contributions in morphological, behavioural and other complex traits.

Tandem repeats modify the structure of the canine CD1D gene.

Among the CD1 proteins that present lipid antigens to T cells, CD1d is the only one that stimulates a population of T cells with an invariant T-cell receptor known as NKT cells. Sequencing of a 722 nucleotide gap in the dog (Canis lupus familiaris) genome revealed that the canine CD1D gene lacks a sequence homologous to exon 2 of human CD1D, coding for the start codon and signal peptide. Also, the canine CD1D gene contains three different short tandem repeats that disrupt the expected gene structure. Because canine CD1D cDNA lacks sequences homologous to human exon 2 and 3, the functionality of canine CD1d protein may be affected, and this could have consequences for the development and activation of canine NKT cells.

Phenotypic and genetic trends of patellar luxation in Dutch Flat-Coated Retrievers.

Canine patellar luxation has been described in various dog breeds, with high prevalence especially in smaller dogs. Most dogs suffer from medial displacement of the patella, although in larger dogs lateral displacement is also seen. A sex predisposition has been described for females. Patellar luxation is considered a polygenic, multifactorial disorder. From 1990 to 2007, in total 3834 Flat-Coated Retrievers were screened; 23.6% of those animals were affected with patellar luxation. Lateral displacement of the patella was most common in this breed (61% of cases), whereas medial (31% of cases) and lateral and medial (8% of cases) were less common. Unilateral involvement (51% of cases) was just as often observed as was bilateral involvement (49% of cases). Females were more often affected with patellar luxation (30% of all tested females) than were males (17% of all tested males). The heritability of patellar luxation was 0.17 ± 0.03 in this population, and breeding with one affected parent increased the prevalence of patellar luxation in offspring by 45% compared to that with two unaffected parents. Since the start of the screening program, there was an initial decrease from 28% to 18% in incidence, but this stagnated thereafter. The annual average estimated breeding values followed the same pattern. With approximately one quarter of the Dutch Flat-Coated Retrievers being affected with patellar luxation, this population shows unusually high prevalence compared with reports in other large-breed dogs. The heritability for patellar luxation in this population was moderate (0.17), indicating that environmental factors play a large role in the manifestation of the disorder. A screening program reduced the prevalence of patellar luxation in this breed, but improvement has recently stagnated. Inclusion of breeding values in the screening program could improve its effectiveness.

Subunits of the translation initiation factor eIF2B are mutant in leukoencephalopathy with vanishing white matter.

Leukoencephalopathy with vanishing white matter (VWM) is an inherited brain disease that occurs mainly in children. The course is chronic-progressive with additional episodes of rapid deterioration following febrile infection or minor head trauma. We have identified mutations in EIF2B5 and EIF2B2, encoding the epsilon- and beta-subunits of the translation initiation factor eIF2B and located on chromosomes 3q27 and 14q24, respectively, as causing VWM. We found 16 different mutations in EIF2B5 in 29 patients from 23 families. We also found two distantly related individuals who were homozygous with respect to a missense mutation in EIF2B2, affecting a conserved amino acid. Three other patients also had mutations in EIF2B2. As eIF2B has an essential role in the regulation of translation under different conditions, including stress, this may explain the rapid deterioration of people with VWM under stress. Mutant translation initiation factors have not previously been implicated in disease.

Severe combined immunodeficiency in Frisian Water Dogs caused by a RAG1 mutation.

Mortality of pups at 8-12 weeks of age was frequently observed in Frisian Water Dogs. Blood parameters and clinical signs of newborns from three litters were monitored. Three pups from two litters showed strongly reduced levels of immunoglobulins and lymphocytes. These dogs were euthanized after first display of disease. Concurrent clinical and pathological features were consistent with a diagnosis of severe combined immunodeficiency (SCID). Defective V(D)J recombination is one of the causes of SCID in humans and animals. Eight genes involved in V(D)J recombination were investigated by segregation analysis of closely located microsatellite markers and by DNA sequence analysis. A nonsense mutation in the gene coding for V(D)J recombination factor RAG1 was identified in DNA from the cases at a position similar to that of nonsense mutations found in human SCID. It was concluded that SCID due to a mutation of RAG1 led to the high mortality.

Evidence of inheritance of intrahepatic portosystemic shunts in Irish Wolfhounds.

BACKGROUND: The etiogenesis of congenital portosystemic shunt in dogs is not understood. In Irish Wolfhounds, intrahepatic portosystemic shunt (IHPSS) is thought to be hereditary, but the mode of inheritance is unknown. OBJECTIVES: To document the genetic background and investigate the potential mode of inheritance of IHPSS in Irish Wolfhounds. ANIMALS: Three mature, privately owned, affected siblings and their progeny produced in 2 litters. METHODS: Prospective, observational study. Two test matings of 1 affected sire with 2 of his affected sisters were used to determine the inheritance pattern. Affection status was determined by measuring venous blood ammonia concentrations, detection of the shunt by ultrasonography and confirmation during surgical attenuation of the intrahepatic shunting vessel. RESULTS: In 1 litter of 5 pups all had an IHPSS. In the other litter 5 of 11 pups were affected. Both left- and right-sided shunts occurred in both litters. No sex predisposition was evident among affected dogs. CONCLUSIONS AND CLINICAL IMPORTANCE: Our results show that IHPSS in Irish Wolfhounds is a familial disorder that is likely genetic. It is unlikely that the mode of inheritance is monogenic. A digenic, triallelic trait could explain the observed occurrence of IHPSS but other modes of inheritance cannot be excluded.

Expression and mutation analysis of Tpit in the canine pituitary gland and corticotroph adenomas.

Pituitary-dependent hyperadrenocorticism (PDH) in dogs is caused by a pituitary corticotroph adenoma. Although PDH is a common disorder in dogs, little is known about the underlying pathogenesis. In the pituitary glands of humans and mice, the pro-opiomelanocortin (POMC)-expressing cell lineages, the corticotrophs and melanotrophs, have a specific marker in common, the T-box transcription factor Tpit (Tbx19), which is obligate for POMC expression. Tpit also regulates the late differentiation of the corticotrophs and melanotrophs, and therefore may contribute to the pathogenesis of the corticotroph adenomas. The aim of this study was to perform an expression and mutation analysis of Tpit in the normal canine pituitary and in corticotroph adenomas. The distribution of the Tpit protein in the pituitary gland was studied with immunohistochemistry and the expression of the gene with RT-PCR. The coding region of Tpit cDNA from 14 dogs with PDH was screened for mutations. Tpit was expressed in corticotroph and melanotroph cells of the normal and adenomatous canine pituitary, and remained present in non-adenomatous corticotrophs of pituitaries from PDH dogs. No tumor-specific mutation in the Tpit cDNA from the corticotroph adenomas was found. However, a missense polymorphism in the highly conserved DNA-binding domain, the T-box, was discovered in one dog. It is concluded that Tpit can be used as a reliable marker for the corticotroph and melanotroph cells in the canine pituitary tissue and that mutations in the Tpit gene are unlikely to play a major role in the pathogenesis of canine corticotroph adenomas.

Canine NAPEPLD-associated models of human myelin disorders.

Canine leukoencephalomyelopathy (LEMP) is a juvenile-onset neurodegenerative disorder of the CNS white matter currently described in Rottweiler and Leonberger dogs. Genome-wide association study (GWAS) allowed us to map LEMP in a Leonberger cohort to dog chromosome 18. Subsequent whole genome re-sequencing of a Leonberger case enabled the identification of a single private homozygous non-synonymous missense variant located in the highly conserved metallo-beta-lactamase domain of the N-acyl phosphatidylethanolamine phospholipase D (NAPEPLD) gene, encoding an enzyme of the endocannabinoid system. We then sequenced this gene in LEMP-affected Rottweilers and identified a different frameshift variant, which is predicted to replace the C-terminal metallo-beta-lactamase domain of the wild type protein. Haplotype analysis of SNP array genotypes revealed that the frameshift variant was present in diverse haplotypes in Rottweilers, and also in Great Danes, indicating an old origin of this second NAPEPLD variant. The identification of different NAPEPLD variants in dog breeds affected by leukoencephalopathies with heterogeneous pathological features, implicates the NAPEPLD enzyme as important in myelin homeostasis, and suggests a novel candidate gene for myelination disorders in people.

A novel susceptibility locus for Hirschsprung's disease maps to 4q31.3-q32.3.

We report on a multigenerational family with isolated Hirschsprung's disease (HSCR). Five patients were affected by either short segment or long segment HSCR. The family consists of two main branches: one with four patients (three siblings and one maternal uncle) and one with one patient. Analysis of the RET gene, the major gene involved in HSCR susceptibility, revealed neither linkage nor mutations. A genome wide linkage analysis was performed, revealing suggestive linkage to a region on 4q31-q32 with a maximum parametric multipoint LOD score of 2.7. Furthermore, non-parametric linkage (NPL) analysis of the genome wide scan data revealed a NPL score of 2.54 (p = 0.003) for the same region on chromosome 4q (D4S413-D4S3351). The minimum linkage interval spans a region of 11.7 cM (12.2 Mb). No genes within this chromosomal interval have previously been implicated in HSCR. Considering the low penetrance of disease in this family, the 4q locus may be necessary but not sufficient to cause HSCR in the absence of modifying loci elsewhere in the genome. Our results suggest the existence of a new susceptibility locus for HSCR at 4q31.3-q32.3.

Population genetic analysis and genome-wide association study of patellar luxation in a Thai population of Pomeranian dogs.

The genetics of patellar luxation (PL) were investigated in Pomeranian dogs presented at the Small Animal Hospital, Faculty of Veterinary Science, Chulalongkorn University. A cohort of 339 Pomeranian dogs, part of a four-generation pedigree of 842 Pomeranians, was screened for PL from 2006 to 2013. PL was present in 77% of the screened dogs, with 84% having bilateral and 16% unilateral luxation. Medial PL was more common (95%) than lateral PL (2%) or bidirectional PL (3%). The risk of PL was similar in male and female dogs (female:male relative risk 1.11, 95% CI 0.98-1.25). The heritability of PL in the screened population was 0.44±0.04 using a threshold model. A genome-wide association study of PL (48 cases and 48 controls) using a high-density SNP array indicated the possible involvement of 15 chromosomal regions, of which CFA05 and CFA32 remained associated in a larger study involving an additional 128 cases and 7 controls. Candidate genes in these regions may be involved in the pathogenesis of PL in Pomeranian dogs.

Disease burden in four populations of dog and cat breeds compared to mixed-breed dogs and European shorthair cats.

Current public and professional opinion is that many dog breeds suffer from health issues related to inherited diseases or extreme phenotypes. The aim of this historical comparative observational study was to evaluate the breed-related disease burden in three purebred dog populations (Chihuahua, French bulldog, Labrador retriever) and one purebred cat breed (Persian cats) in the Netherlands by comparison to a control population of mixed-breed dogs and European Shorthair cats. A qualitative query was performed, consisting of a literature review and collecting the expert opinions of University veterinary specialists, to gather insight into potential diseases of the study population. Next, a referral clinic case control study of the patients referred to specific medical disciplines in the University Clinic was performed. The odds ratio (OR) was calculated to determine the likelihood of a patient referred to a particular medical discipline being a certain breed. Together, the qualitative query and the case control study resulted in a list of potentially relevant diseases limited to five organ systems per breed. These were analysed in data from primary practices. Patient files from ten primary practices over a period of two years were manually extracted and examined. Four-hundred individual patient records per breed as well as 1000 non-breed records were randomly selected from the 10 practices, weighted per practice size. Records were then examined and the presence or absence of certain diseases was identified. To evaluate the disease burden per breed, proportional difference (PD) was estimated, as well as the animal's age at presentation in months. The results of the referral clinic case control study showed an overrepresentation (Odds Ratio>1.5) of the selected breeds in several medical specialties, while median age at presentation was in some cases significantly lower than in the non-breed animals. Results of the practice-based extended cross-sectional study showed that only a few of the selected diseases contribute to the disease burden in these purebred populations, which was different from the expectations derived from the literature or expert opinion. Additional results included age difference at presentation, which may be interpreted as age of onset, and could indicate a higher disease burden for the individual animal. Also, only a small percentage of purebred dogs was registered with the national kennel club. Our final recommendation is that population-based data mining is needed to evaluate country-specific companion animal health and welfare.

Paroxysmal Dyskinesia in Border Terriers: Clinical, Epidemiological, and Genetic Investigations.

BACKGROUND: In the last decade, a disorder characterized by episodes of involuntary movements and dystonia has been recognized in Border Terriers. OBJECTIVES: To define clinical features of paroxysmal dyskinesia (PD) in a large number of Border Terriers and to study the genetics of the disease. ANIMALS: 110 affected and 128 unaffected client-owned Border Terriers. METHODS: A questionnaire regarding clinical characteristics of PD was designed at Utrecht University and the University of Helsinki. Thirty-five affected Border Terriers underwent physical examination and blood testing (hematology and clinical biochemistry). Diagnostic imaging of the brain was performed in 17 affected dogs and electroencephalograms (EEG) between episodes were obtained in 10 affected dogs. A genomewide association study (GWAS) was performed with DNA of 110 affected and 128 unaffected dogs. RESULTS: One hundred forty-seven questionnaires were included in the study. The most characteristic signs during episodes were dystonia, muscle fasciculations, and falling over. The majority of owners believed that their dogs remained conscious during the episodes. A beneficial effect of anti-epileptic therapy was observed in 29 of 43 dogs. Fifteen owners changed their dogs' diet to a hypoallergenic, gluten-free diet, and all reported reasonable to good improvement of signs. Clinical examinations and diagnostic test results were unremarkable. The GWAS did not identify significantly associated chromosome regions. CONCLUSIONS AND CLINICAL IMPORTANCE: The survey results and EEG studies provided further evidence that the observed syndrome is a PD rather than epilepsy. Failure to achieve conclusive results by GWAS indicates that inheritance of PD in Border Terriers probably is complex.

In vitro and in vivo effects of kisspeptin antagonists p234, p271, p354, and p356 on GPR54 activation.

Kisspeptins (KPs) and their receptor (GPR54 or KiSS1R) play a key-role in regulation of the hypothalamic-pituitary-gonadal axis and are therefore interesting targets for therapeutic interventions in the field of reproductive endocrinology. As dogs show a rapid and robust LH response after the administration of KP10, they can serve as a good animal model for research concerning KP signaling. The aims of the present study were to test the antagonistic properties of KP analogs p234, p271, p354, and p356 in vitro, by determining the intracellular Ca2+ response of CHEM1 cells that stably express human GPR54, and to study the in vivo effects of these peptides on basal plasma LH concentration and the KP10-induced LH response in female dogs. Exposure of the CHEM1 cells to KP-10 resulted in a clear Ca2+ response. P234, p271, p354, and p356 did not prevent or lower the KP10-induced Ca2+ response. Moreover, the in vivo studies in the dogs showed that none of these supposed antagonists lowered the basal plasma LH concentration and none of the peptides lowered the KP10-induced LH response. In conclusion, p234, p271, p354, and p356 had no antagonistic effects in vitro nor any effect on basal and kisspeptin-stimulated plasma LH concentration in female dogs.

The leukemia inhibitory factor receptor gene is not involved in the etiology of pituitary dwarfism in German shepherd dogs.

Pituitary dwarfism in German shepherd dogs is characterized by combined pituitary hormone deficiency (CPHD) and intrapituitary cyst formation. Activation of the leukemia inhibitory factor (LIF)-LIF receptor (LIFR) signal transduction pathway results in a similar phenotype in (transgenic) mice. We therefore assessed the role of the LIFR in the etiology of pituitary dwarfism in German shepherd dogs. A polymorphic microsatellite marker (UULIFR) was used to analyze the segregation of the LIFR gene in 22 German shepherd dogs from 4 pedigrees, each including one dwarf. There was no allelic association between UULIFR and the dwarfism phenotype. Based on our findings LIFR was excluded as a candidate gene for CPHD.

The effects of kisspeptin agonist canine KP-10 and kisspeptin antagonist p271 on plasma LH concentrations during different stages of the estrous cycle and anestrus in the bitch.

Kisspeptin (KP) plays a key role in the regulation of the hypothalamic-pituitary-gonadal axis via the release of GnRH. As normal KP signaling is essential for reproductive function, it could be an interesting new target for therapeutic interventions, e.g., nonsurgical contraception in dogs. The aims of the present study were to investigate the effect of KP-10 administration on plasma LH concentration in different stages of the reproductive cycle and to investigate the suitability of p271 as KP antagonist in the bitch. Two groups of six adult Beagle bitches were used. In one group, plasma LH concentration was determined before (40 and 0 minutes) and 10, 20, 40, and 60 minutes after the intravenous administration of 0.5-µg/kg body weight (BW) canine KP-10. In the other group, the bitches received a continuous intravenous infusion with p271 (50 µg/kg BW/h) for 3 hours, and 0.5-µg/kg BW canine KP-10 was administered intravenously 2 hours after the start of the p271 infusion. Their plasma LH concentration was determined before (-40 and 0 minutes) and 30, 60, 90, 120, 130, 140, 160, and 180 minutes after the start of the p271 infusion. In both groups, the experiments were performed during the follicular phase, the first and second half of the luteal phase, and during anestrus. Canine KP-10 induced an increase of plasma LH concentration during all estrous cycle stages and anestrus. There was no difference in LH response between the two groups. The lowest LH response was seen during the follicular phase and the highest response during anestrus. The area under the curve (AUC) for LH and LH increment in the follicular phase were lower than those in anestrus. The AUC LH and LH increment in the first half of the luteal phase were lower than those in the second half of the luteal phase and anestrus. The AUC LH and LH increment in the second half of the luteal phase were not different from those in anestrus. Continuous administration of the antagonist p271 did not alter basal plasma LH concentration and could not prevent or lower the LH response to KP-10 in any of the cycle stages and anestrus. It can be concluded that the LH response to KP-10 is dependent on estrous cycle stage and that peripheral administrated p271 cannot be used as KP antagonist in the dog. This provides new insight in reproductive endocrinology of the bitch, which is important when KP signaling is considered for therapeutic interventions, such as for estrus induction or nonsurgical contraception in the bitch.

Adenovirus DNA replication in vitro: duplication of single-stranded DNA containing a panhandle structure.

Adenovirus DNA replicates by displacement of one of the parental strands followed by duplication of the displaced parental single strand (complementary strand synthesis). Displacement synthesis has been performed in a reconstituted system composed of viral and cellular proteins, employing either the viral DNA-terminal protein complex as template or linearized plasmids containing the origin. Previously, evidence was obtained that in vivo complementary strand synthesis requires formation of a panhandle structure originating from hybridization of the inverted terminal repeats. To study the conditions for complementary strand synthesis in vitro, we have constructed an artificial panhandle molecule that contains a double-stranded inverted terminal repetition (ITR) region and a single-stranded loop derived from the left and right terminal XmaI fragments of Ad2. Such a molecule appeared to be an efficient template and could initiate by the same protein-priming mechanism as double-stranded DNA, employing the precursor terminal protein. The efficiency of both types of template was comparable. Like for replication of the duplex molecule initiation of panhandle replication was stimulated by nuclear factors I and III, proteins that bind to specific double-stranded regions of the ITR. The Ad DNA-binding protein is essential and the 39 kDa C-terminal domain of this protein that harbors the DNA-binding properties is sufficient for its function. These results support the hypothesis that panhandle formation is required for duplication of the displaced strand.

The canine sarcoglycan delta gene: BAC clone contig assembly, chromosome assignment and interrogation as a candidate gene for dilated cardiomyopathy in Dobermann dogs.

Dilated cardiomyopathy (DCM) is a common disease of the myocardium recognized in human, dog and experimental animals. Genetic factors are responsible for a large proportion of cases in humans, and 17 genes with DCM causing mutations have been identified. The genetic origin of DCM in the Dobermann dogs has been suggested, but no disease genes have been identified to date. In this paper, we describe the characterization and evaluation of the canine sarcoglycan delta (SGCD), a gene implicated in DCM in human and hamster. Bacterial artificial chromosomes (BACs) containing the canine SGCD gene were isolated with probes for exon 3 and exons 4-8 and were characterized by Southern blot analysis. BAC end sequences were obtained for four BACs. Three of the BACs overlapped and could be ordered relative to each other and the end sequences of all four BACs could be anchored on the preliminary assembly of the dog genome sequence (www. ensembl.org). One of the BACs of the partial contig was localized by fluorescent in situ hybridization to canine chromosome 4q22, in agreement with the dog genome sequence. Two highly informative polymorphic microsatellite markers in intron 7 of the SGCD gene were identified. In 25 DCM-affected and 13 non DCM-affected dogs seven different haplotypes could be distinguished. However, no association between any of the SGCD variants and the disease locus was apparent.

DNA methylation patterns in the calcitonin gene region at first diagnosis and at relapse of acute lymphoblastic leukemia (ALL).

Aberrant DNA methylation can occur early in neoplastic transformation and may lead to the development of cancer. We describe the alterations of methylation patterns at the DNA sequence level which occurred in the 5' region of the calcitonin gene in lymphoblasts from 14 pediatric patients with acute lymphoblastic leukemia (ALL). The DNA methylation status of 25 CpG sites was determined by sequence analysis after bisulfite treatment of the DNA. This method showed that 13 out of 14 patients had increased numbers of methylated CpG sites in the calcitonin gene region at initial diagnosis when compared to control DNA from healthy individuals. The 5' region of the calcitonin gene appears to be methylated to a significantly higher degree in T lineage ALL compared to B lineage ALL (P < 0.01). Each of six ALL patients who were investigated at initial diagnosis and at relapse showed alterations in DNA methylation between the two stages. These six cases were also investigated by Southern blot analysis with methylcytosine-sensitive restriction enzymes and this method showed an increase in DNA methylation in only four of the six cases. The DNA sequencing method thus appears to be better suited to assess alterations of DNA methylation than Southern blot analysis. There are marked regional differences in the frequency of methylation of individual CpG sites and in the frequency of alterations between the two stages. Our results show that alterations in DNA methylation continue to occur from the initial stage to the relapse stage of ALL, suggesting that aberrant DNA methylation may play a role in tumor progression.