Homozygous Disruption of the Tip60 Gene Causes Early Embryonic Lethality

Abstract

Tat-interactive protein 60 (Tip60) is a member of the MYST family, proteins in which are related by an atypical histone acetyltransferase (HAT) domain. Although Tip60 has been implicated in cellular activities including DNA repair, apoptosis and transcriptional regulation, its function during embryonic development is unknown. We ablated the Tip60 gene (Htatip) from the mouse by replacing exons 1–9 with a neomycin resistance cassette. Development and reproduction of wild-type and heterozygous animals was normal. However, homozygous ablation of the Tip60 gene caused embryolethality near the blastocyst stage of development, as evidenced by inability of cells in Tip60-null blastocysts to hatch and survive in culture. Monitoring cell proliferation and death by detecting EdU-substituted DNA and TUNEL labeling revealed suppression of cell proliferation concomitant with increased cell death as Tip60-null cells attempted to hatch from blastocysts. These findings indicate that Tip60 is essential for cellular survival during the blastocyst-gastrula transition of embryogenesis.

Introduction

Tip60 is a vertebrate member of the protein family termed MYST, an acronym coined in accord with the names of the family’s founding members MOZ, YBF2, SAS2 and Tip60 (Utley & Cole, review, 2003). MYST proteins are related by a ~300 amino acid domain containing atypical zinc finger and histone acetyltransferase (HAT) domains. Similar to MYSTs such as yeast Esa1 (Smith et al. 1998) and vertebrate Mof (Neal et al. 2000), Tip60 also contains a chromodomain, Drosophila homologues of which (heterochromatin protein-1, polycomb) repress transcription via histone H3 methylation. Although the HAT domain in Tip60 can acetylate histones (Yamamoto & Horikoshi, 1997), this activity is inefficient in comparison with other HATs; moreover, the Tip60 HAT domain has other functions, as indicated by its ability to acetylate non-histone proteins (Gaughan et al. 2002) as well as by the effects of mutating its acetyl-CoA binding sites, which results in inhibition of DNA repair (Kusch et al. 2004) and apoptosis (Ikura et al. 2000). Other evidence suggests that Tip60 is a negative regulator of cellular homeostasis. For example a pro-apoptotic role for Tip60 was recently documented by findings that Tip60 must acetylate a specific residue in p53 to enable p53’s apoptotic function (Sykes et al. 2006; Tang et al. 2006, 2008). In addition, Tip60 inhibits mesangial cell proliferation (Muckova et al. 2006), and, with p53 it up-regulates the p21 gene to induce cell-cycle arrest (Legube et al. 2004; Tang et al. 2006). These findings are consistent with a genome-wide siRNA screen indicating that Tip60 participates in p53-dependent cell-cycle arrest (Berns et al. 2004). Because these and other activities have been ascribed to Tip60 in context-dependent fashion (reviews: Sapountzi et al. 2006; Thomas & Voss 2007), including the co-repression and co-activation (DeRan et al. 2008) of target genes, its function is considered pleiotropic.

We previously reported that Tip60 protein is expressed at early stages of heart development (Lough, 2002; Kim et al. 2006). Because it was therefore of interest to evaluate the role of Tip60 during this process, we globally ablated the mouse Tip60 gene (Htatip). As reported here, although heterozygous Tip60^+/− mice are viable, exhibiting apparently normal development and reproduction, homozygous-null Tip60^−/− mice have never been observed at post-gastrulation stages of development; hence the role of Tip60 in heart development could not be informed. Investigation of the time and cause of embryonic death due to homozygous ablation of the Tip60 gene revealed that although an appropriate Mendelian complement of Tip60-null embryos is present among blastocysts at ED3.5, these die with 100% penetrance shortly thereafter, as indicated by the inability of Tip60^−/− blastocyst cells to hatch and survive in culture, concomitant with inhibition of cell proliferation and increased cell death. While these findings do not reveal the molecular cause of the Tip60 embryolethal phenotype, this is likely multi-factorial considering the pleiotropic function of Tip60 (reviews: Sapountzi et al. 2006; Thomas & Voss 2007), a likelihood supported by a genome-wide screen revealing that Tip60 is one of only six “hub” genes, which regulate multiple signaling pathways (Lehner et al. 2006).

Results

The murine Tip60 gene was ablated using the strategy depicted in Figure 1. Homologous recombination with a targeting vector having homology arms complementary to approximately 9 kbp of the Tip60 gene (Fig. 1A) resulted in ablation of exons 1 through 9, including approximately 200 bp of the promoter. Heterozygous matings revealed that among more than 300 weanlings, an approximate Mendelian ratio of two heterozygotes per wild-type was maintained (Table 1). No homozygous-null weanlings were ever detected. The average number of pups in each litter was 5.74 (N = 62 litters: SD ± 2.24; SEM ± 0.28), in contrast to an average of 10–11 pups/litter resultant from wild-type crossings of CD-1 mice. Genotyping embryos at ED7.5 (Fig. 1H) and ED8.5 (not shown) also failed to detect homozygous-null embryos (Table 1). Because this indicated that Tip60 gene ablation causes early embryolethality, 8-cell stage (ED2.5) and blastocyst stage (ED3.5) embryos from heterozygous matings were harvested and genotyped. Among 53 8-cell stage embryos evaluated, 15 wild-type, 24 heterozygous and 14 homozygous-null embryos were present (Fig. 1F & Table 1), consistent with the expected 1:2:1 Mendelian ratio (Chi square P=0.78). A similar genotypic ratio was observed among 69 ED3.5 blastocysts collected one day later, wherein the percentage of Tip60-null embryos did not differ from the Mendelian prediction of 25% (Chi square P=0.22; Fig. 1G & Table 1). Hence, Tip60-null embryos survive at least until the blastocyst stage of embryonic development.

Figure 1. Tip60 Gene Targeting.

(A), Map of the Tip60 gene (top), targeting vector (middle) and targeted allele (bottom). Shaded regions indicate the homology arms of the targeting vector; filled and diagonally-hatched boxes indicate exons used by major and minor isoforms of Tip60, respectively. The targeting vector replaced 2,787 bp, including 193 bp of the promoter and exons 1–9, with the neomycin resistance cassette of pPNT (1,849 bp; note that this is in reverse orientation). B=BamHI; E=EcoRI; H=HindIII; RV=EcoRV; Sc=ScaI; Sm=SmaI; X=XhoI. Positions of external probes A and B used for Southern hybridization are shown. Arrowheads denote the position and orientation of primers used for PCR genotyping. (B), PCR genotyping of targeted ES-R1 cells. PCR was performed using the two primers denoted by filled arrowheads beneath the WT allele, and the primer denoted by the filled arrowhead beneath the neo cassette, in panel A. PCR products of 1,682 bp and 1,410 bp were amplified from the respective WT and targeted alleles, as predicted. (C), Verification of targeting by Southern blotting: DNA from a gan/neo-selected ESC clone was digested with HindIII and hybridized with probe B, revealing conversion of a WT 7.6 kbp allele (+/+) to a ~13 kbp allele in targeted cells (+/−) due to removal of the HindIII site in exon 3. (D–E), Genotyping of offspring from heterozygous crosses. D: PCR genotyping using the primers denoted by open arrowheads beneath the wild-type allele and the filled arrowhead beneath the targeting vector in A were used to amplify 462 and 188 bp products from the WT and targeted alleles, respectively. (E) Southern blot verification of PCR genotyping: Tail tip DNA of weaned mice was digested with BamHI or HindIII followed by respective hybridization with probe A or probe B. (F–G) Genotypes of ED2.5 and ED3.5 embryos assessed by amplifying WT and targeted alleles; note that these PCR products were amplified using a single primer pair in separate PCR reactions. Genotypes that could not be determined are denoted “nd”. (H) Genotypes of ED7.5 embryos determined by multiplex genotyping, using three primers to amplify PCR products from both alleles in the same reaction. For details see Experimental Procedures. All images are from EtBr-stained agarose gels. Cumulative results of these determinations are presented in Table 1.

Table 1. Genotypes of embryonic and adult mice resultant from mating heterozygous (Tip60^+/− × Tip60^+/−) mice.

All genotypes were determined by PCR as described in Experimental Procedures. Embryos were at the indicated developmental stages; weaned mice were 1–2 months old. ND = genotypes that could not be determined. Chi-square analysis revealed that the proportion of genotypes at ED2.5 (P = 0.78) and ED3.5 (P = 0.22) did not significantly differ from Mendelian expectations.

Distribution of the Tip60-null Allele in Embryos and Weanlings following Tip60^+/− × Tip60^+/− Intercrosses.

	ED2.5		ED3.5		ED7.5		ED8.5		weaned
	#	%	#	%	#	%	#	%	#	%
+/+	15	28	11	16	8	31	32	28	107	34
+/−	24	45	39	56	18	69	81	72	207	66
−/−	14	27	19	28	0	0	0	0	0	0
[nd]	[7]	[12]	[4]	[5]	0	0	0	0	0	0
totals	53	100	69	100	26	100	113	100	314	100

To assess whether cells in Tip60-null blastocysts could hatch from the zona pellucida and expand in vitro, blastocysts from a heterozygous mating were cultured in separate wells of glass chamber slides. In this experiment, cultures were monitored for up to 21 days in order to assess whether a line of proliferating Tip60-null cells could be obtained. At the time of harvest, the presence of fewer cell numbers in Tip60-null embryos (i.e. embryos #105 and #107 in Fig. 2A) was a common observation, suggesting reduced proliferation. During the first 2–3 days in culture, most blastocysts attached to the gelatin substrate, followed on the fourth day by hatching of cells from the zona pellucida. Hatched cells that survived the first four days of the culture period expanded and became segregated into monolayered and multilayered aggregates that respectively resembled trophectoderm (TE) and inner cell mass (ICM) (Fig. 2B–C). As cells became apoptotic/necrotic during the 21-day in vitro period they were harvested and saved for genotyping. This revealed 19 wild-type, 45 heterozygous and 15 homozygous-null genotypes; although the proportion of homozygous-null cultures was not significantly different from Mendelian expectations (Chi square P=0.38), the slight under-representation of Tip60-null embryos suggested that onset of lethality might be detected by increasing statistical power. Among the 15 Tip60-null embryos, visual observations during the first day of culture had revealed that ten of these resembled embryos nos. 105 and 107 shown in Fig. 2A which were Tip60-null, and that the remaining five embryos contained small cells that obscured the blastocoel. Remarkably, all 15 Tip60-null genotypes represented dying embryos that were harvested during the first four days of the culture period (Fig. 2D), whereas all cells exhibiting wild-type or heterozygous genotypes were harvested after the fourth day of the 21-day culture period (Fig. 2E). These observations indicated that Tip60-null cells cannot hatch from the zona pellucid or survive in vitro.

Figure 2. Morphology and Genotype of Blastocysts and Blastocyst-derived Cells.

(A) Blastocysts harvested at ED3.5. The genotype of each embryo is indicated. (B–C) Phase-contrast images of cells derived from Tip60 WT and Tip60-heterozygous blastocysts after 21 days in culture. ICM, inner cell mass ICM; trophoectoderm, TE. (D) PCR genotyping of blastocyst-derived cells cultured for 3–4 days. WT and null alleles were amplified in separate reactions. Genotypes that could not be determined are denoted “nd”. (E) PCR genotyping of blastocyst-derived cells cultured for 5–21 days; no homozygous-null cells were detected during this interval. Panels D and E are EtBr-stained agarose gels.

The results of Figure 2 indicated that Tip60-null embryolethality may involve inability of the inner cell mass (ICM) to survive the transition from pluripotency to epiblast and hypoblast (primitive endoderm). To address this possibility, blastocysts from heterozygous matings were cultured for up to four days. As shown in Figure 3A, whereas wild-type and heterozygous blastocysts enlarged during growth in culture, followed by cellular hatching and differentiation, Tip60-null blastocysts did not become enlarged and cells therein did not hatch. During the four day culture period, embryos were harvested after two or four days to assess expression of Tip60 mRNA, as well as mRNAs for markers for primitive endoderm (GATA-4) and cell-cycle regulation (cyclin D2). As shown in Figure 3B–D, semi-quantitative RT/PCR revealed no detectable expression of Tip60 (Fig. 3B), GATA4 (Fig. 3C) or cyclinD2 (Fig. 3D) mRNAs in Tip60-null blastocysts that were cultured for either two or four days. While the absence of Tip60 mRNA was anticipated, the absence of GATA-4 mRNA indicated that Tip60-null cells were unable to differentiate into primitive endoderm, and, the absence of cyclinD2 indicated that Tip60-null cells were unable to undergo cell-cycle transit.

Figure 3. Morphology and Gene Expression in Cultured Blastocysts.

(A) Photomicrographs of wild-type, Tip60-heterozygous and Tip60-null blastocysts at time of harvest (ED3.5 embryo) and during subsequent cell culture (days 1–4). Tip60-null blastocysts did not increase in size, and, cells did not hatch from them during the 4-day culture period. The horizontal bar in the upper left panel = 50 µm; all photographs are at the same magnification. (B–D) RT-PCR after 2 and 4 days in culture. No evidence of Tip60 (panel B), GATA4 (panel C) or cyclinD2 (panel D) mRNAs were detected in Tip60-null cells as early as day 2. EB = embryoid body samples (WT) used for positive controls; nt = no template control; unlabeled lanes were empty.

To assess cell proliferation in Tip60-null blastocysts, cells cultured for two days were evaluated for cell-cycle transit per incorporation of EdU into replicating DNA using the Click-iT™ EdU cell proliferation assay. As shown in Figure 4A, whereas fluorescence indicative of significant EdU labeling was observed in wild-type and heterozygous cells, only background levels were observed in Tip60-null cells, indicating that cell proliferation was inhibited in Tip60-null embryos. To complement this assessment, survival of cultured Tip60-null cells at four days in culture was evaluated by TUNEL labeling, examples of which are shown in Figure 4B. Because it was not possible to subject each blastocyst to both TUNEL labeling and PCR genotyping, genotypes in this determination were inferred from co-immunostaining Tip60 protein; hence the presence of Tip60 protein denoted either a wild-type or heterozygous blastocyst, while the absence of Tip60 protein denoted a Tip60-null blastocyst. Results shown in Figure 4B indicated that cells in blastocysts expressing Tip60 protein (panels b & a/b) could hatch, as indicated by the DAPI-positive cellular monolayer (Fig. 4B panel a); hence these cells were either wild-type or heterozygous. TUNEL co-labeling revealed that the wild-type/heterozygous trophectoderm cellular monolayer, which was extensive in comparison with cells attempting to hatch from Tip60-null blastocysts (Fig. 4B panel c), contained relatively few dying cells (arrows in Fig. 4B panel a). By contrast with the low incidence of cell death in trophectoderm derived from Tip60-positive blastocysts, the relatively few cells attempting to hatch from Tip60-null blastocysts were preponderantly TUNEL-positive (Fig. 4B panel c). Although TUNEL-positive cells were noted within the multilayers of both Tip60-positive (Fig.4B panel a) and Tip60-null (Fig. 4B panel c) blastocysts per se, it was not possible to enumerate relative numbers within these structures. In summary, results shown in Figure 4 indicate that embryolethality in Tip60-null embryos is accompanied by suppression of cell-cycle transit, and increased cell death.

Figure 4. Cell Proliferation & Death in Cultured Blastocysts.

Panel A depicts cell proliferation. Blastocysts harvested at ED3.5 were cultured for two days and treated with 10 µM EdU for 2.5 hours (except in h, no EdU control), followed by detection of replicated DNA with Alexa Fluor 488 azide (Click-iT™ detection cocktail) as described in Experimental Procedures. All pictures are at the same magnification; the bar in a = 100 µm. Panel B depicts cell death. Blastocysts harvested at ED3.5 were cultured for four days and processed for TUNEL labeling to detect dying cells (a, c; TUNEL/DAPI merge). Co-immunostaining of Tip60 protein (b, d; Tip60 immunostain) was used to distinguish wild-type or heterozygous embryos (b) from Tip60-null (d) embryos. Magnified, confocal (Z-series) merged images of the WT or Het embryo in a–b, and the Tip60-null embryo in c–d, are respectively shown in a/b and c/d. Note that TUNEL signal in c/d (arrows) is diminished in a/b. The low level of Tip60 protein signal in c/d indicates background and/or remnant maternal Tip60 protein after 4 days in culture. Arrows in a, c and c/d denote dying cells. The bar in a–d = 100 µm; the bar in a/b and c/d = 47 µm.

Discussion

We globally ablated the Tip60 gene to assess whether Tip60 protein, which is expressed in the myocardium beginning at the earliest stages of heart development (Lough 2002; Kim et al. 2006), is crucial for that process. Because homozygous ablation caused completely penetrant embryolethality, apparently just after the blastocyst stage of embryogenesis, Tip60’s role in heart development could not be assessed. This result indicates that Tip60 is a vital protein whose function cannot be compensated by other members of the MYST family, including Mof (Neal et al. 2000) which is structurally similar to Tip60 and shares its ability to acetylate p53 in site-specific fashion (Sykes et al. 2006). The identity of Tip60 as an indispensable protein is consistent with our inability to isolate Tip60-null embryonic stem cells (ESCs) by increasing G418 concentration, as well as with recent reports in the literature. For example, Tip60 was recently shown to be essential for Drosophila embryogenesis (Zhu et al. 2007). In regard to pinpointing the molecular cause of embryolethality caused by Tip60 ablation, a genome-wide siRNA screen revealed that mys-1, the C. elegans homologue of Tip60, is one of only six “hub” genes; this distinction is informative because inactivation of Tip60 would predictably exacerbate the effects of mutation in a variety of downstream genes, thereby disrupting multiple signaling pathways (Lehner et al. 2006). Considering this possibility, as well as the pleiotropic function that has been ascribed to Tip60 (reviews: Sapountzi et al. 2006; Thomas & Voss, 2007), the cause of embryolethality observed in the mouse embryo is likely multi-factorial.

Although these findings do not reveal the molecular cause of embryolethality, that Tip60 ablation causes a severe growth defect phenotype at or shortly after the blastocyst stage of development is supported by morphological observations (Fig. 2–Fig 3), RT/PCR determinations showing that cyclin D2 mRNA (like Tip60 mRNA) is knocked-out (Fig. 3D), and by the observation that ability to incorporate EdU is inhibited (Fig. 4A). Extinguished expression of the cyclin D2 gene is consistent with previous findings indicating that Tip60 co-regulates the cyclin D2 promoter (Kioussi et al. 2002). Most significantly, the observation that Tip60 is required for blastocyst growth is consistent with results from a recent siRNA knockdown study demonstrating that Tip60 is required for the renewal of mouse ESCs (Fazzio et al. 2008). Curiously, Tip60 knockdown in the latter was accompanied by knockdown of p53, similar to our preliminary data indicating that expression of p53 as well as pluripotency factors (Oct4 & Nanog) are respectively inhibited >15-fold and >5-fold in Tip60-null blastocysts after two days in culture (not shown).

The observation of increased TUNEL-positive cells as Tip60-null cells attempt to hatch from the blastocyst (Fig. 4B panel c) suggests that cell death contributes to embryolethality. This is consistent with a previous study using these mice which demonstrated that lymphocytes in Tip60-haploinsufficient adults have a compromised DNA damage response (DDR; Gorrini et al. 2007), raising the possibility that a defective DDR in Tip60-null blastocysts contributes to embryolethality. However, increased cell death in the absence of Tip60 is seemingly inconsistent with its function as a pro-apoptotic molecule (Tang et al. 2006, 2008), a role shared by the closely related MYST family member Mof (Sykes et al. 2006). Recently, knockout of the mouse Mof gene was shown to cause embryolethality at a similar stage of embryonic development. The Mof phenotype, like the Tip60 phenotype, was accompanied by increased TUNEL labeling that was interpreted to reflect apoptosis; this paradox was reconciled by suggesting that Mof-induced apoptosis is restricted to differentiating cells (Thomas et al. 2008). The possibility that Mof and Tip60 have different functions in pluripotent and differentiating cells has precedence in that the major function of p53 (which is functionally related to both Tip60 and Mof) in pluripotent cells is to suppress Nanog, resulting in induction of differentiation whereupon p53 assumes its familiar cell-cycle inhibitory and pro-apoptotic functions.

Including Tip60, eight members of the MYST gene family have been identified, three in yeast and five in metazoan cells (Utley & Cole, review, 2003). Gene ablation and knockdown studies indicate that the MYST proteins have diverse functions. Among yeast MYSTs, disruption of the Esa1 gene abolishes cellular growth (Smith et al. 1998). On the other hand, although mutation of the related yeast MYST genes SAS2 and SAS3 (YBF2) inhibits their transcriptional repressor function, growth is unaffected (Reifsnyder et al. 1996). Two vertebrate-specific MYSTs, MOZ (MYST3) and MORF (QKF, MYST4), were discovered as chromosomal translocations involved in acute myeloid leukemia (Yang & Ullah, review, 2007). MOZ (monocytic leukemia zinc-finger protein) disruption in mice was recently shown to cause lethality at embryonic or neonatal stages due to inability to maintain the hematopoietic stem cell compartment (Katsumoto et al. 2006; Thomas et al. 2006). MORF disruption causes death at weaning, a phenotype accompanied by osteogenic and neurogenic deficits (Thomas et al. 2000; Merson et al. 2006). Among the remaining metazoan MYSTs, HBO1 (HAT bound to ORC1; MYST2) has apparently not been targeted. Interestingly, as mentioned above the recently reported effects of Mof (the homolog of Drosophila male-absent on the first; MYST1) gene targeting in mice described a phenotype that is very similar to the effects of Tip60 ablation reported here (Gupta et al. 2008; Thomas et al. 2008). Hence, despite their significant structural similarity and common function of acetylating p53, Tip60 and Mof (Neal et al. 2000; Sykes et al. 2006) are not functionally redundant. This is not surprising considering that the functions of Mof and Tip60 diverge, i.e., Mof site-specifically acetylates histone H4 (Akhtar & Becker, 2000) to effect transcriptional activation, cell proliferation and tumorigenesis (Gupta et al. 2008), whereas Tip60 has pleiotropic functions (reviews: Sapountzi et al. 2006; Thomas & Voss 2007) including the co-regulation of gene promoters (including cyclin D2, p21, histone genes), as well as the site-specific acetylation of p53 (Sykes et al. 2006; Tang et al. 2006, 2008) and participation in the DNA damage response (DDR; Gorrini et al. 2007). Taken together, results from targeting of the metazoan MYST genes indicate that these molecules non-redundantly maintain stem cell compartments, during both late stages of embryogenesis as in the instances of MOZ and MORF, and during early embryogenesis as in the instances of Mof and Tip60. This interpretation is consistent with a p53-associated role for Tip60 in maintaining genome stability in early embryonic cells, as well as with recently reported findings that Tip60 is required for ESC renewal (Fazzio et al. 2008).

Among HAT-encoding genes that have been disrupted, lethality caused by ablation of Tip60 and Mof occurs at the earliest stage of embryonic development. Disruption of the genes encoding the two most widely studied HATs, CBP and p300, causes embryolethality at mid-gestation, respectively due to defective vasculogenesis (Oike et al. 1999) and defective cell proliferation (Yao et al. 1998). CBP and p300 both interact with the histone acetyltransferases Gcn5 and Pcaf. Ablation of the gene encoding Gcn5 causes embryolethality by ED10.5 due to increased apoptosis, resulting in failure to form dorsal mesoderm (Xu et al. 2000). By contrast, ablation of PCAF, which is minimally expressed in the embryo but is strongly expressed in the adult heart, is without consequence (Xu et al. 2000). The findings discussed in this and in the preceding paragraph indicate that among MYST and non-MYST HATs, Tip60 and Mof are vitally important at the earliest stages of embryogenesis. Elucidation of Tip60’s role during formation of the primary germ layers, and in organs derived from them, will be clarified using a mouse line containing conditionally-targeted Tip60 alleles, which is in final stages of preparation.

Experimental Procedures

Targeting the Tip60 Gene

We previously reported the structure of the mouse Tip60 gene (Htatip; McAllister et al. 2002) per characterization of overlapping 12 kbp BamHI and 7.5 kbp HindIII fragments isolated from a BAC obtained from a mouse 129/SVJ genomic library (Incyte Genomics). Using these fragments, a targeting construct was created in pPNT plasmid (Tybulewicz et al. 1991) which contains pgk-neo and HSV-thymidine kinase expression cassettes for respective positive and negative selection. As shown in Figure 1A, the pgk-neo gene was flanked by a Tip60 longarm consisting of a blunted 4,931 bp EcoR1/SmaI sub-fragment of the HindIII fragment which was inserted into pPNT’s blunted Xho1 site. The shortarm consisted of a 1,165 bp ScaI/ScaI sub-fragment of the BamHI fragment cloned into pPNT’s blunted EcoR1 site. The homology arms are in opposite orientation to the pgk-neo gene. The targeting vector, linearized by cutting the NotI site of pPNT, was electroporated into R1 ESCs (129/Sv × 129/Sv-CP). Clones resistant to G418 and sensitive to gancyclovir were isolated and accuracy of targeting indicated by PCR was verified by Southern blotting. Targeted ESCs were genotyped by multiplex PCR using primers that amplify the Tip60 gene domains indicated by the filled arrowheads beneath the WT allele in Figure 1A (5’-GGCACAACTCAGAACTCACAAG-3’ [Tip60 fwd] and 5’-ACTCATCTTCGTTGTCCTGGTTGC-3’ [Tip60 rev]), plus a primer (5’-ACGAGATCAGCAGCCTCTGT-3’ [neo]) which anneals to the neomycin resistance cassette indicated by the filled arrowhead beneath the targeting vector in Figure 1A; this multiplex primer mixture respectively amplifies PCR products of 1,682 (which spans the entire shortarm) and 1,410 bp from the wild-type and targeted alleles as shown in Fig. 1B. Correctness of targeting was verified by Southern blotting performed by digesting DNA from the targeted ESC clone with BamHI or HindIII, followed by hybridization respectively using ³²P-labeled probe A or probe B which correspond to the domains shown in Figure 1A; Fig. 1C is an example of HindIII-digested DNA hybridized with probe B.

Genotype Verification of Chimeric Offspring

Approximately 15 trypsinized ESCs from a correctly targeted ES-R1 clone (#1F2A) were aggregated with pronase-treated (to remove the zona pellucida) 8 cell-stage (ED2.5) CD1 embryos from super-ovulated females, followed one day later (ED3.5) by implantation of recombinant blastocysts into the uterine horns of pseudo-pregnant CD1 female mice. Mating of chimeric offspring verified germline transmission. Initial offspring were genotyped by multiplex PCR (Fig. 1D) using Tip60 primers 5’-ATGATCTGAGTGACCGGCGT-3’ (Tip60 fwd) and 5’-ACTCATCTTCGTTGTCCTGGTTGC-3’ (Tip60 rev) (see open arrowheads beneath WT allele in Figure 1A), plus the above neomycin primer (filled arrowhead beneath targeting vector in Fig. 1A) to respectively amplify 462 bp and 188 bp products from the wild-type and targeted alleles. Southern blotting was used to verify correctness of targeting in offspring, examples of which are shown in Figure 1E including BamHI-digested DNA hybridized with probe A and HindIII-digested DNA hybridized with probe B.

Genotyping Post-Gastrulation Embryos & Adult Mice

Following verification of germline transmission and correctness of targeting, routine genotyping of embryos at ED7.5 and older as well as of adult animals was performed on DNA obtained from yolk sacs (embryos) or tail tips (adults). DNA was purified using the Wizard SV Genomic DNA Purification System (Promega #A2361), from which 50 ng was used as template for multiplex amplification using a mixture of the following primers: 5’-AAGCCTAAACATGATCTGAGTGACCGGCGT-3’ (Tip60 fwd), 5’-CACGCCACTCATCTTCGTTGTCCTGGTT-3’ (Tip60 rev), and 5-GGCCAGCTCATTCCTCCACTCATGATCTAT-3’ (neo). This primer mixture, denoted by the shaded arrowheads beneath the WT allele and targeting vector in Figure 1A, respectively amplifies 478 bp and 327 bp products from the WT and targeted alleles (Fig. 1H). With the exception of genotyping embryos at ED3.5 and younger as described in the following paragraph, PCR was performed using a reaction mixture including 1.5 mM MgCl₂, 200 µM each nucleotide and 0.5 µM each primer. Amplification was initiated by adding Taq DNA Polymerase (Promega #M1661) to a concentration of 25 units/ml, followed by 30 cycles of 94 °C denaturation for 30 seconds, 65 °C annealing for 1 minute and 72 °C extension for 1 minute, with a final 10 minute extension at 72 °C. PCR products were separated on a 1.7% agarose gel and stained with ethidium bromide.

Genotyping Morula & Blastocyst Stage Embryos

Reliable PCR genotyping of embryos at ED3.5 and younger required modification of the above conditions, the most pronounced being that alleles at these early embryonic stages were individually amplified, instead of performing amplification in multiplex using three primers. All embryos at ED3.5 and younger were harvested by flushing uteri with a 30 1/2-gauge needle attached to a 1 ml syringe containing M2 medium (Specialty Media/Chemicon #MR-015-D) that had been pre-warmed to 37 °C. Each embryo was individually washed through five drops of M2 medium to remove blood and cell debris. The zona pellucida was not removed. To genotype 8-cell stage (ED2.5) embryos, modifications used by Lombaert et al. (2008) were used wherein each embryo was transferred to 5.0 µl proteinase-K buffer containing containing SDS (17 µM; to penetrate the zona pellucida) and proteinase-K (400 ng/µl) for 60 minutes at 37 °C, followed by heating for 15 minutes at 95 °C and withdrawal of 2.0 µl samples for use as template. Similar conditions were used to genotype blastocysts (ED3.5) except that these were digested with proteinase-K (200 ng/µl) overnight at 55 °C using 0.5% Triton-X-100 instead of SDS. At both embryonic stages, alleles were individually amplified instead of performing multiplex PCR. The WT Tip60 allele was amplified using 5’-GACAGACTCGGCGTTCCTCCAATC-3’ (Tip60 fwd) and 5’-CGGCAGCCCTCGATTATCTC-3’ (Tip60 rev); this primer pair, which is nested within the domain amplified by the primer pairs used to amplify alleles in embryos at ED7.5 and older, amplified a 379 bp fragment. The null allele was detected using the same forward primer in combination with the neo reverse primer (5-GGCCAGCTCATTCCTCCACTCATGATCTAT-3’), yielding a 278 bp product. PCR reactions were performed in 25 µl total volume using GoTaq Green Master Mix (Promega #7121) in which components were present at the same concentration as above except that the final concentration of each primer was 0.4 µM. Amplification was performed using 45 cycles of 94 °C denaturation, 55 °C annealing and 72 °C extension, 45 seconds each, with a final 10 minute extension at 72 °C.

Culture of Blastocysts

Super-ovulation of young (3–4 week old) Tip60 heterozygous female mice was performed by intraperitoneal injection of 10 Units pregnant mare serum gonadotropin (PMSG), followed 48 hours later by intraperitoneal injection of 10 Units human chorionic gonadotropin (HCG) and immediate mating with Tip60 heterozygous stud males. Three days after observing a vaginal plug, blastocysts (ED3.5) were harvested and washed in pre-warmed M2 medium as described above and cultured as previously described (Herceg et al. 2001; Kwon et al. 2008). Briefly, blastocysts were individually suspended in 100 µl ES Cell Qualified DMEM supplemented with 15% FBS (Hyclone #SH30071.03) and 1× L-glutamine, aminoacids, sodium pyruvate, β-mercaptoethanol, penicillin-streptomycin and LIF (all from Specialty Media). Blastocysts were individually plated in chambers of a glass 16-well Lab-Tex Chamber Slide (Nunc #178599) pre-coated with 0.1% gelatin and cultured under 5% CO₂. To facilitate adherence to the gelatin substrate, medium was not replenished until three days (0+3 days) later, when an additional 100 µl fresh medium were added. At 3-day intervals thereafter, medium in all chambers was completely exchanged with 100 µl fresh medium. During the first week of culture, blastocysts were examined daily for grossly detectable evidence of apoptosis/necrosis as indicated by the presence of shriveled cells within the zona pellucid; dying cells were immediately harvested in 10 mM NaOH/1 mM EDTA and saved for genotyping at −20 °C.

Reverse Transcription/Polymerase Chain Reaction (RT/PCR)

Total RNA was isolated and purified from embryoid bodies (EBs, used as controls) and blastocysts using Trizol Reagent (Invitrogen #15596-026) according to the manufacturer’s recommendations. First-strand cDNA was reverse-transcribed by priming with oligo-dT (20 µg/ml) and extending in buffer containing 50 mM Tris (pH 8.3)/75 mM KCl/3 mM MgCl₂/10 mM dithiothreitol/500 µM each deoxynucleotide/25 µg/ml BSA; the reaction was catalyzed by adding 10,000 units/ml M-MLV reverse transcriptase (RNase H-minus; Promega #M530A). PCR amplification was performed on templates consisting of 1/20 of the RT product, using buffer that included 1.5 mM MgCl₂/200 µM each nucleotide/0.5 µM each primer. Primer pairs used to amplify the various cDNAs were as follows: GAPDH: 5’-ATGGTGAAGGTCGGTGTGAA-3’ (fwd) & 5’-TGGTGGTGCAGGATGCATTG-3’ (rev), 450 bp product; Cyclin D2: 5’-TCAAGTGCGTGCAGAAGGACA-3 (fwd) & 5’-TCGACGGCGGGTACATGGCAAACT-3’ (rev), 471 bp product; Gata4: 5’-CTGGAGGCGAGATGGGACGGGACACTAC-3’ (fwd) & 5’-CCGCAGGCATTACATACAGGCTCACC-3’ (rev), 207 bp product; Tip60: 5’-GCCTGGACGGAAGCGGAAATCTAAT-3’ (fwd) & 5’-AAACACTTGGCCAGAAGACACAG-3’ (rev), 420 bp product. Amplification was initiated by adding Taq DNA Polymerase (Promega #M1661) to a concentration of 25 units/ml, followed by 45 cycles of 94 °C for 30 seconds, 65 °C for 1 minute, and 72 °C for 1 minute with a final 10 minute extension at 72 °C. PCR products were separated on a 1% agarose gel and stained with ethidium bromide.

Cell Proliferation, TUNEL & Immunohistochemical Staining

Cell proliferation was monitored by adding EdU (5-ethynyl-2’-deoxyuridine) to blastocyst cultures at a final concentration of 10 µM for 2.5 hours, followed by detection of replicated DNA using Alexa Fluor 488 azide (Click-iT™ detection cocktail) according to the manufacturer’s protocol (Invitrogen #C35002). Cells undergoing apoptosis/necrosis were monitored by TUNEL labeling using the APO-BrdU TUNEL Assay Kit (Invitrogen #A23210) according to the manufacturer’s instructions. Tip60 immunostaining was performed using a rabbit anti-Tip60 antibody (diluted 1:100) which recognizes peptide EGCRLPVLRRNQDNE within the N-terminus of all known Tip60 isoproteins; there is no sequence homology between this peptide and the highly related MYST protein MOF. Specificity of this antibody was confirmed by its ability to detect a single band at the specified M_r of Tip60α and its Tip60β isoform in Western blots. The secondary antibody used to detect Tip60 was goat anti-rabbit IgG conjugated to rhodamine (1:200; Jackson ImmunoResearch Laboratories Inc #111-295-003). Nuclei were stained with TOPRO and DAPI (300 nM). Conventional fluorescent microscopy was performed using a Nikon TE300 fluorescent microscope. Confocal fluorescent imaging was performed using a Leica TCS SP2^® Laser Scanning Confocal Microscope.

Statistical Analysis

In order to determine whether embryos from heterozygous matings were obtained in Mendelian ratios, Chi-square analysis was used to compare observed with expected genotypes,