D. Alan Underhill

Ltap , a mammalian homolog of Drosophila Strabismus/Van Gogh , is altered in the mouse neural tube mutant Loop-tail

Neural tube defects (NTDs) such as spina bifida and anencephaly are common congenital malformations in humans (1/1,000 births) that result from failure of the neural tube to close during embryogenesis. The etiology of NTDs is complex, with both genetic and environmental contributions; the genetic component has been extensively studied with mouse models. Loop-tail (Lp) is a semidominant mutation on mouse chromosome 1 (ref. 4). In the two known Lp alleles (Lp, Lpm1Jus), heterozygous mice exhibit a characteristic looped tail, and homozygous embryos show a completely open neural tube in the hindbrain and spinal region, a condition similar to the severe craniorachischisis defect in humans. Morphological and neural patterning studies indicate a role for the Lp gene product in controlling early morphogenesis and patterning of both axial midline structures and the developing neural plate. The 0.6-cM/0.7-megabase (Mb) Lp interval is delineated proximally by D1Mit113/Apoa2/Fcer1g and distally by Fcer1a/D1Mit149/Spna1 and contains a minimum of 17 transcription units. One of these genes, Ltap, encodes a homolog of Drosophila Strabismus/Van Gogh (Stbm/Vang), a component of the frizzled/dishevelled tissue polarity pathway. Ltap is expressed broadly in the neuroectoderm throughout early neurogenesis and is altered in two independent Lp alleles, identifying this gene as a strong candidate for Lp.

Dynamic Changes in Histone H3 Lysine 9 Methylations IDENTIFICATION OF A MITOSIS-SPECIFIC FUNCTION FOR DYNAMIC METHYLATION IN CHROMOSOME CONGRESSION AND SEGREGATION

Histone methylation is unique among post-translational histone modifications by virtue of its stability. It is thought to be a relatively stable and heritable epigenetic mark for gene-specific regulation. In this study, we use quantitative in situ approaches to investigate the cell cycle dynamics of methylated isoforms of histone H3 lysine 9. Contrary to the expected stability of trimethylated lysines, our results for trimethylated lysine 9 (tMeK9) of H3 demonstrate that the genomic content of this methylation undergoes significant changes as cells progress through mitosis. Unexpectedly, there is a loss of tMeK9 that appears to reflect a robust demethylase activity that is active during the period between anaphase and cytokinesis. Subsequent investigations of mitoses in tMeK9-deficient cells revealed defects in chromosome congression and segregation that are distinct from the increased cohesion at centromeres previously reported in association with the loss of tMeK9. Collectively, these results identify a mitosis-specific trimethylation of Lys9 in pericentromeric heterochromatin that functions in the faithful segregation of chromosomes.

Identification of α-tubulin as a granzyme B substrate during CTL-mediated apoptosis

Cytotoxic lymphocytes induce target cell apoptosis via two major pathways: Fas/FasL and granule exocytosis. The latter pathway has largely been defined by the roles of the pore-forming protein perforin and by the serine proteinases granzymes A and B. Upon entry into target cells, the granzymes cleave substrates that ultimately result in cell death. To gain further insight into granzyme B function, we have identified novel substrates. SDS-PAGE analysis of S100 cell lysates identified a 51 kDa protein that was cleaved by granzyme B. Mass spectrometry analysis revealed that this fragment was the microtubule protein, α-tubulin, which was confirmed by western blotting. In addition, two-dimensional gel analysis showed that the truncated form of α-tubulin had a more basic isoelectric point than the full-length molecule, suggesting that granzyme B removed the acidic C-terminus. Site-directed mutagenesis within this region of α-tubulin revealed the granzyme B recognition site, which is conserved in a subset of α-tubulin isoforms. Significantly, we showed that α-tubulin was cleaved in target cells undergoing apoptosis as induced by cytotoxic T lymphocytes. Therefore, in addition to its role in the activation of mitochondria during apoptosis, these results suggest a role for granzyme B in the dismantling of the cytoskeleton.

Poly(ADP-ribosyl)ation-dependent Transient Chromatin Decondensation and Histone Displacement following Laser Microirradiation

Chromatin undergoes a rapid ATP-dependent, ATM and H2AX-independent decondensation when DNA damage is introduced by laser microirradiation. Although the detailed mechanism of this decondensation remains to be determined, the kinetics of decondensation are similar to the kinetics of poly(ADP-ribosyl)ation. We used laser microirradiation to introduce DNA strand breaks into living cells expressing a photoactivatable GFP-tagged histone H2B. We find that poly(ADP-ribosyl)ation mediated primarily by poly(ADP-ribose) polymerase 1 (PARP1) is responsible for the rapid decondensation of chromatin at sites of DNA damage. This decondensation of chromatin correlates temporally with the displacement of histones, which is sensitive to PARP inhibition and is transient in nature. Contrary to the predictions of the histone shuttle hypothesis, we did not find that histone H1 accumulated on poly(ADP-ribose) (PAR) in vivo. Rather, histone H1, and to a lessor extent, histones H2A and H2B were rapidly depleted from the sites of PAR accumulation. However, histone H1 returns to chromatin and the chromatin recondenses. Thus, the PARP-dependent relaxation of chromatin closely correlates with histone displacement.

Comparative Analyses of SUV420H1 Isoforms and SUV420H2 Reveal Differences in Their Cellular Localization and Effects on Myogenic Differentiation

Background Methylation of histone H4 on lysine 20 plays critical roles in chromatin structure and function via mono- (H4K20me1), di- (H4K20me2), and trimethyl (H4K20me3) derivatives. In previous analyses of histone methylation dynamics in mid-gestation mouse embryos, we documented marked changes in H4K20 methylation during cell differentiation. These changes were particularly robust during myogenesis, both in vivo and in cell culture, where we observed a transition from H4K20me1 to H4K20me3. To assess the significance of this change, we used a gain-of-function strategy involving the lysine methyltransferases SUV420H1 and SUV420H2, which catalyze H4K20me2 and H4K20me3. At the same time, we characterized a second isoform of SUV420H1 (designated SUV420H1_i2) and compared the activity of all three SUV420H proteins with regard to localization and H4K20 methylation. Principal Findings Immunofluorescence revealed that exogenous SUV420H1_i2 was distributed throughout the cell, while a substantial portion of SUV420H1_i1 and SUV420H2 displayed the expected association with constitutive heterochromatin. Moreover, SUV420H1_i2 distribution was unaffected by co-expression of heterochromatin protein-1α, which increased the targeting of SUV420H1_i1 and SUV420H2 to regions of pericentromeric heterochromatin. Consistent with their distributions, SUV420H1_i2 caused an increase in H4K20me3 levels throughout the nucleus, whereas SUV420H1_i1 and SUV420H2 facilitated an increase in pericentric H4K20me3. Striking differences continued when the SUV420H proteins were tested in the C2C12 myogenic model system. Specifically, although SUV420H1_i2 induced precocious appearance of the differentiation marker Myogenin in the presence of mitogens, only SUV420H2 maintained a Myogenin-enriched population over the course of differentiation. Paradoxically, SUV420H1_i1 could not be expressed in C2C12 cells, which suggests it is under post-transcriptional or post-translational control. Conclusions These data indicate that SUV420H proteins differ substantially in their localization and activity. Importantly, SUV420H2 can induce a transition from H4K20me1 to H4K20me3 in regions of constitutive heterochromatin that is sufficient to enhance myogenic differentiation, suggesting it can act an as epigenetic ‘switch’ in this process.

Subnuclear localization and mobility are key indicators of PAX3 dysfunction in Waardenburg syndrome

Mutations in the transcription factor PAX3 cause Waardenburg syndrome (WS) in humans and the mouse Splotch mutant, which display similar neural crest-derived defects. Previous characterization of disease-causing mutations revealed pleiotropic effects on PAX3 DNA binding and transcriptional activity. In this study, we evaluated the impact of disease alleles on PAX3 localization and mobility. Immunofluorescence analyses indicated that the majority of PAX3 occupies the interchromatin space, with only sporadic colocalization with sites of transcription. Interestingly, PAX3 disease alleles fell into two distinct categories when localization and dynamics in fluorescence recovery after photobleaching (FRAP) were assessed. The first group (class I), comprising N47H, G81A and V265F exhibit a diffuse distribution and markedly increased mobility when compared with wild-type PAX3. In contrast, the G42R, F45L, S84F, Y90H and R271G mutants (class II) display evidence of subnuclear compartmentalization and mobility intermediate between wild-type PAX3 and class I proteins. However, unlike class I mutants, which retain DNA binding, class II proteins are deficient for this activity, indicating that DNA binding is not a primary determinant of PAX3 distribution and movement. Importantly, class I properties prevail when combined with a class II mutation, which taken with the proximity of the two mutant classes within the PAX3 protein, suggests class I mutants act by perturbing PAX3 conformation. Together, these results establish that altered localization and dynamics play a key role in PAX3 dysfunction and that loss of the underlying determinants represents the principal defect for a subset of Waardenburg mutations.

The PAX3 paired domain and homeodomain function as a single binding module in vivo to regulate subnuclear localization and mobility by a mechanism that requires base-specific recognition.

The transcription factor PAX3 is essential for myogenesis and neural crest development, and is one of several genes mutated in human Waardenburg syndrome. Analysis of disease-causing missense mutations in PAX3 has established the interdependence of its two DNA-binding domains, the paired domain (PD) and the homeodomain (HD), as well as defects in localization and mobility. Paradoxically, mutants that retained DNA binding activity exhibited the greatest defects in localization and mobility, regardless of the domain in which they reside. In the present study, structure-function analyses were used to determine the mechanistic basis of this effect. In the context of the isolated DNA-binding domains, HD mutants adopted an increase in mobility proportional to their loss in DNA binding, while PD mutants continued to display the inverse relationship observed in the full-length protein. At the structural level, this reflected an unexpected dependence on base-specific contacts in the PD, whereas HD mobility was more severely affected by loss of backbone contacts, as has been observed with other DNA-binding proteins. This requires that the HD switch to a base-specific mode in the full-length protein. Moreover, both domains underwent substantial reduction in mobility and altered localization when in a contiguous polypeptide with the endogenous linker segment. Notably, although the HD conferred localization to heterochromatin, this activity was masked when linked to the PD, despite the absence of determinants for subnuclear compartmentalization in the PD or linker. Last, the propensity for PAX3 heterochromatin localization was modulated by sequences at the amino and carboxy termini, supporting a model in which alternate conformations lead to unmasking of the HD. These data indicate that the PD and the HD functionally interact in vivo and behave as a single binding module whose mobility and localization are dependent on sequence-specific contacts.

Transcription mapping and expression analysis of candidate genes in the vicinity of the mouse Loop-tail mutation

Abstract. Loop-tail (Lp) is a semidominant mutation that maps to the distal portion of mouse Chromosome (Chr) 1 and is an established model for neural tube defects (NTDs). Homozygous embryos exhibit an open neural tube from the caudal midbrain to the tip of the tail that results from over-differentiation of the floor plate. To facilitate the positional cloning of the Lp gene, both cDNA selection and assignment of sequence-tagged-sites from the human transcript map have been used to identify genes within the Lp interval. Together with previous physical mapping, this has allowed the placement of 13 transcription units within an approximately 1-Mb region that spans the Lp genetic interval, and eight of these genes map to the nonrecombinant interval. This map includes genes that encode proteins involved in protein sorting and targeting (Tim23 and Copa), ion transport (Atp1a2, Atp1a4, and Girk3), transcription (Nhlh1), immune regulation (Cd48 and Fcer1α), cell adhesion (R88252), apoptosis (Pea15), and several of unknown function (H326, Kiaa0253, and Estm34). Expression analysis by Northern blotting indicated that a subset of these genes are expressed preferentially in the developing nervous system. Finally, this region of mouse Chr 1 represents a conserved linkage group with genes on human chromosome 1q21, a region that is frequently altered in human cancers and that harbors loci for several genetic conditions. Consequently, analysis of the Lp interval may provide important tools to understand how the corresponding region of human Chr 1 contributes to disease, in addition to defining a key gene product required for neurulation.

Targeting Epigenetic Pathways in the Treatment of Pediatric Diffuse (High Grade) Gliomas

Progress in the treatment of adult high-grade gliomas (HGG), including chemoradiation with concurrent and adjuvant temozolomide for glioblastoma, has not translated into significant therapeutic advances for pediatric HGG, where overall survival has plateaued at 15% to 20%, especially when considering specialized pediatric treatment in tertiary care centers, maximal safe neurosurgical resection, optimized delivery of involved field radiation, and improvements in supportive care. However, recent advances in our understanding of pediatric HGG, including the application of next-generation sequencing and DNA methylation profiling, have identified mutations in the histone variant H3.3 and canonical H3.1 genes, respectively. These mutations are relatively specific to neuroanatomic compartments (cortex, midline structures, thalamus, brainstem) and are often associated with other mutations, especially in specific growth factor receptor tyrosine kinases. Targeting epigenetic pathways affected by these histone mutations, alone or in combination with small molecule inhibitors of growth factor receptor signaling pathways, will inform new treatment strategies for pediatric HGG and should be incorporated into novel cooperative group clinical trial designs.

Identification of a novel, alternatively spliced isoform and single nucleotide polymorphisms in the murine Pea-15 gene

Pea-15 (a 15-kDa p hosphoprotein e nriched in astrocytes; Estelle ́s et al. 1996) belongs to a family of death-effector-domain (DED) proteins originally defined by the pro-apoptotic protein FADD (Muzio et al. 1996). The death-effector-domain functions as a protein-protein interface that couples polypeptides in the apoptotic pathway. In CD95 (Fas/APO-1)-mediated apoptosis, the FADD DED recruits FLICE, a member of the ICE/CED-3 family of cysteine proteases, and physically links the Fas/APO-1 receptor to this proapoptotic cysteine protease (Muzio et al. 1996). Pea-15 has an antagonistic role in Fas-mediated death signaling, and this involves interactions of the Pea-15 DED with homologous segments of FADD and FLICE (Condorelli et al. 1999; Estelles et al. 1999) that may prevent their association with the Fas/APO-1 receptor complex. In addition to a role in regulating apoptosis through the Fas/ APO-1 receptor, Pea-15 may also have roles in integrin activation and glucose transport. Integrins mediate cell-cell and cellextracellular matrix adhesion through a signaling cascade that is suppressed by the Ras-initiated MAP-kinase pathway. Pea-15 can inhibit Ras-mediated suppression of integrin activation, and this function requires the Pea-15 DED, in a role that is distinct from its association with apoptosis (Ramos et al. 1998). The PEA-15 gene is overexpressed in type 2 diabetes mellitus, and its transfection into L6 skeletal myoblasts increases expression of the Glut1 glucose transporter, while inhibiting the insulin-stimulated recruitment of the Glut4 transporter (Condorelli et al. 1998). As a result, elevated expression of PEA15 may contribute to insulin resistance in type 2 diabetes mellitus. In humans, the PEA-15 gene localizes to Chromosome (Chr) 1q21, which is in a region of conserved synteny with the telomeric portion of mouse Chr 1 (Hwang et al. 1997). We have previously shown that thePea-15gene is located in the candidate region for the mouse loop-tail ( Lp) mutation (Underhill et al. 1999). The mouseLp mutation is inherited in a semidominant fashion, with affected heterozygotes exhibiting a ‘looped-tail’ and wobbly head (Strong and Hollander 1949). In homozygous embryos, the neural tube remains open from the caudal midbrain to the tip of the tail. At the molecular level,Lp mice exhibit axial patterning defects that arise during gastrulation and are most clearly evidenced by the overexpression of Shhin the notochord and floor plate (Greene et al. 1998). Importantly, previous Northern blotting studies have indicated thatPea-15has a restricted pattern of expression that includes the developing brain and spinal cord (Estelle ́s et al. 1996), axial structures that are affected in Lp embryos. Moreover, both apoptosis and integrin function have important roles during embryonic development, prompting us to examine the integrity of the Pea-15transcript inLp embryos. To obtain more detailed expression data in the neural tube, the