Depletion of Tip60 from In Vivo Cardiomyocytes Increases Myocyte Density, Followed by Cardiac Dysfunction, Myocyte Fallout and Lethality

Tat-interactive protein 60 (Tip60), encoded by the Kat5 gene, is a member of the MYST family of acetyltransferases. Cancer biology studies have shown that Tip60 induces the DNA damage response, apoptosis, and cell-cycle inhibition. Although Tip60 is expressed in the myocardium, its role in cardiomyocytes (CMs) is unclear. Earlier studies here showed that application of cardiac stress to globally targeted Kat5+/—haploinsufficient mice resulted in inhibition of apoptosis and activation of the CM cell-cycle, despite only modest reduction of Tip60 protein levels. It was therefore of interest to ascertain the effects of specifically and substantially depleting Tip60 from CMs using Kat5LoxP/-;Myh6-Cre mice in the absence of stress. We report initial findings using this model, in which the effects of specifically depleting Tip60 protein from ventricular CMs, beginning at early neonatal stages, were assessed in 2–12 week-old mice. Although 5’-bromodeoxyuridine immunostaining indicated that CM proliferation was not altered at any of these stages, CM density was increased in 2 week-old ventricles, which persisted in 4 week-old hearts when TUNEL staining revealed inhibition of apoptosis. By week 4, levels of connexin-43 were depleted, and its patterning was dysmorphic, concomitant with an increase in cardiac hypertrophy marker expression and interstitial fibrosis. This was followed by systolic dysfunction at 8 weeks, after which extensive apoptosis and CM fallout occurred, followed by lethality as mice approached 12 weeks of age. In summary, chronic depletion of Tip60 from the ventricular myocardium beginning at early stages of neonatal heart development causes CM death after 8 weeks; hence, Tip60 protein has a crucial function in the heart.


Introduction
Tip60 (Tat-interactive protein 60 kD) is an acetylase protein, which in cancer cells has been shown to induce the DNA Damage Response (DDR) and apoptosis by respectively acetylating ATM [1,2] and p53 [3,4]. Tip60 has also been shown to inhibit the cell-cycle [5]. These functions aren't mutually exclusive. Tip60 is encoded by the Lysine acetyltransferase-5 (Kat5) gene. We previously reported that Tip60 is a vital intracellular protein, since ablation of both Kat5 alleles causes embryolethality at the blastocyst stage of development. However, global heterozygous Kat5 +/mice reproduce normally, exhibiting no apparent haploinsufficient phenotype under normal conditions [6]. This laboratory wishes to elucidate Tip60's role in the adult heart, wherein it is relatively enriched [7]. We previously subjected Kat5 +/heterozygous adult hearts to stress of cMyc over-expression, and to pressure overload [8]. In accord with Tip60's role in cancer cells, both stressors reduced apoptosis while inducing re-entry of ventricular cardiomyocytes (CMs) into the cell-cycle, despite the retention of Tip60 protein at~80% of wildtype levels [9].
According to the TSGene database (http://bioinfo.mc.vanderbilt.edu/TSGene/index.html), Tip60 is a tumor suppressor. Because our previous work indicated that modest Tip60 reduction in combination with stress relieved inhibition of the CM cell-cycle [9], similar to the effects of depleting retinoblastoma [10], hippo pathway components [11], and Meis1 [12], it was of interest to ascertain the effects of substantially depleting Tip60 in CM-specific fashion from the mouse heart, in the absence of stress. Hence, we prepared a genetic model wherein most of the Kat5 coding region, including the Tip60 acetylase domain, is efficiently and specifically disrupted in mouse CMs in vivo. This is our first report using this model, wherein we describe the phenotype caused by CMspecific Tip60 depletion, beginning in the ventricles at early neonatal stages consequent to activation of a constitutively expressed Myh6-driven Cre-recombinase transgene, in 2, 4, 8 and 12 week-old mouse hearts. Proliferation of ventricular CMs was not altered at any age. However, an increase in CM density was observed in 2 week-old hearts, which persisted through week 4, when TUNEL staining indicated a transient reduction in apoptosis. By week 4, signs of deterioration were also observed, including connexin-43 depletion, increased hypertrophy marker expression, and interstitial myocardial fibrosis. By week 8 these changes became pronounced, leading to significantly impaired cardiac function that culminated in death of most Tip60-depleted mice prior to week 12, when extensive apoptosis and CM fallout were observed. Taken together, these findings indicate that although depletion of Tip60 from the ventricular myocardium may transiently permit CM expansion, its chronic depletion is incompatible with CM survival.

Animal Care and Use
This investigation adhered to the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH Pub. Nos. 85-23, Revised 1996). All protocols described in the authors' Animal Use Application (AUA #000225), which were approved by the Medical College of Wisconsin's Institutional Animal Care and Use Committee (IACUC), were adhered to in this study. The Medical College of Wisconsin IACUC has an Animal Welfare Assurance status from the Office of Laboratory Welfare (A3102-01). In this study, we observed that Tip60 depletion caused mortality between 8-12 weeks of age. Because this outcome was not anticipated, it was not reviewed and approved by the animal ethics committee as part of our IACUC-approved protocol. Despite daily monitoring to identify humane endpoints based on BCS (body condition scoring, [13]) criteria, neither a BCS score of "3" in a single category nor an aggregate score of "7" across categories was observed prior to death. All male mice (approximately 25 in this study) bearing the Kat5 LoxP/-;Myh6-Cre genotype spontaneously died at approximately 12 weeks of age. [Note: BCS criteria that were monitored included discomfort (assessed by hunching with disheveled fur), lethargy, rapid weight loss, labored breathing, and behavioral response to external stimuli.]. For details see Supporting Information (S1 File).

Generation of Mice with LoxP-Flanked Kat5 Alleles
The generation of mice with a globally-targeted (i.e. null) Kat5 allele, wherein the promoter and exons 1-9 of the Kat5 gene are replaced with a neomycin resistance gene [14], was described previously [6].
To generate mice with LoxP-flanked Kat5 alleles, a targeting vector was prepared by recombineering [15,16], resulting in the insertion of LoxP sites into introns 2 and 11 of the Kat5 gene (S1A Fig). Correctly targeted ESCs (line V6.5) were injected into C57/Bl6 blastocysts, which implanted and transmitted the targeted allele, termed Kat5 LoxP(FRTneoFRT)/+ , via germline. A neomycin phosphotransferase (neo) cassette in intron 2 was subsequently removed by mating with a mouse expressing a Flp recombinase transgene, resulting in the genotype Kat5 LoxP/+ , which was then bred to homozygosity (Kat5 LoxP/LoxP ). Genotyping performed by PCR and Southern blotting confirmed targeting accuracy (S1B and S1C Fig).
Mice containing the cardiac-specific Myh6-driven Cre-recombinase transgene, which is briefly activated during early development followed by strong and persistent expression in the ventricles beginning at early neonatal stages [17], were purchased from the Jackson Laboratory (Tg(Myh6-cre)1Jmk/J; #009074). This transgene was bred into the heterozygous Kat5 +/background, then backcrossed to produce a stud line that was heterozygous for Kat5 and homozygous for Myh6-Cre (i.e. Kat5 +/-;Myh6-Cre/Myh6-Cre ). The latter were bred with Kat5 LoxP/LoxP mice to obtain the experimental (Kat5 LoxP/-;Myh6-Cre ) and control (Kat5 LoxP/+;Myh6-Cre ) genotypes that were compared in this study. The experimental genotype (Kat5 LoxP/-;Myh6-Cre ) was chosen to increase efficiency of Tip60 protein removal, because only one Kat5 allele had to be ablated, and since our previous work revealed that Kat5 +/heterozygous mice (i) reproduce normally, (ii) have no apparent adult phenotype and (iii) express near-normal levels of Tip60 protein [9]. Although wild-type mice with either one or two copies of the Myh6-Cre transgene exhibit no phenotype [18], breeding was adjusted so that only one copy was inherited. Because the Myh6-Cre transgene is located on the X-chromosome, the extent of Tip60 depletion occurs in gender-specific fashion (S2A-S2C Fig

Echocardiography
Echocardiography was performed as described in Supporting Information (S1 File).

Genotyping
Genotyping was performed by PCR in 25.0 μl reactions that included 1x GoTaq G2 Green Mastermix (Promega #M7822), 0.5 μM each primer, and 4.0 μl template. Templates consisted of low-speed supernatants of extracted ear tissue samples (punches) that had been boiled for 10 minutes in 0.3 ml 10 mM NaOH/1 mM EDTA. Primer pairs used to amplify each allele are listed in Table 1. PCR products were amplified in an AB Applied Biosystems GeneAmp PCR System 9700 followed by separation and imaging of amplicons on ethidium bromide-stained 1% agarose gels.

Assessment of Cre-Mediated Kat5 Recombination
Cre-mediated recombination of the Kat5 allele was assessed by PCR on samples of left ventricle (LV) and liver, using the primer pair listed in Table 1. These primers, which anneal to introns 2 and 11, amplify 444 and 4,793 bp segments in the presence and absence of recombination, respectively.

Heart Tissue Processing
This was performed by anesthetizing mice with 4% isoflurane, followed by removal of the anterior chest wall and rapidly perfusing the heart by gravity with 20 ml ice-cold cardioplegic solution (25 mM KCl/5% dextrose/ PBS) introduced via the left ventricular apex with a 27-gauge needle. Hearts were rapidly removed and the LV was isolated, followed by transverse sectioning to obtain superior and inferior parcels that were respectively used for western blotting/mRNA analysis and immunostaining.

Quantitative RT-PCR (qPCR)
All qPCR reactions were performed per the manufacturers' recommendations. Heart tissue was disrupted with a Tissue Lyser II and total RNA was purified using the RNeasy Plus Mini Kit (Qiagen #74134) followed by quantitation using a Qubit Fluorometer Broad Range RNA

Western Blotting
Blots were prepared using total protein extracted from the superior half of the left ventricle, which was minced with scissors, suspended in lysis buffer (20mM HEPES [pH 7.5]/150mM NaCl/1mM EDTA/100mM dithiothreitol/1% Triton-X-100/5% glycerol plus 1x protease inhibitor (Roche #1836153) and phosphatase inhibitor cocktail) and sonicated. Protein concentration was determined using the Bradford assay (Bio-Rad #500-0006). Samples were stored at -80°C until electrophoresis. For electrophoresis, samples containing 10 μg total protein were suspended in SDS sample buffer (250 mM Tris-HCl [pH 6.8]/3.75% glycerol/0.005% bromphenol blue/2% SDS/100mM dithiothreitol), heated at 100°C for 5 minutes, and electrophoretically separated on 7.5% acrylamide:bis (30:0.8)/SDS gels (Bio-Rad #161-1118). Separated proteins were electroblotted onto nitrocellulose membrane (Bio-Rad #162-0146) and blocked overnight with NFDM/TBST (5% non-fat dry milk/10mM Tris-HCl (pH 7.6)/150mM NaCl/0.05% Tween-20). Primary and secondary antibodies and dilutions are listed in Table 2. Blots were reacted with primary antibody in blocking buffer overnight at 4°C. The affinity-pure anti-Tip60 was custom prepared (Bethyl Labs.) in rabbit, using the N-terminal peptide EGCRLPVLRRNQ DNE peptide as immunogen. This peptide is present in the N-terminus of all Tip60 isoproteins and is absent from all other proteins including the highly related MYST family member MOF. This antibody was characterized as previously described in detail [9]. Tip60 blots were simultaneously reacted with anti-GAPDH. Secondary antibodies were diluted in NFDM/TBST and applied for 60 minutes at RT. Blots were then covered with HRP-substrate (Amersham #RPN2232) for 1 minute at room temperature followed by antigen localization, which was accomplished by exposing the membrane to Hyperfilm-ECL (Kodak #8294985) for selected intervals up to 20 minutes and processing in a Kodak X-Omat 2000A developer. Films were scanned with a Hewlitt-Packard Scanjet G4010 scanner; densitometry was quantified using ImageJ software.

Immunostaining & Cell Counting
Two days before harvest, mice were intraperitoneally injected with 5'-bromodeoxyuridine (1.0 mg/100 μl; Sigma #B9285). Perfused hearts were fixed overnight in fresh 4% paraformaldehyde/PBS, processed through 70% EtOH, and embedded in paraffin. Sections (5 μm) on microscope slides were immunostained using the primary and secondary antibodies listed in Table 2. Cells were manually counted by scanning the entirety of each section at 600x magnification. BrdU-labeled cells were scored as positive only if stain was clearly confined to the nucleus (verified with DAPI), and was granular. Caspase-3-labeled cells were scored as positive only if clearly defined granular immunohistochemical product was observed in the cytoplasm. Cells at edges of sections were not scored, in order to avoid including edge-artifacts. Due to their large numbers, CM nuclei, which were efficiently identified using Cre-recombinase immunostaining, were enumerated by photographing six 400x fields in each section, followed by ImageJ particle analysis. Briefly, using ImageJ software, the red (594) channel containing Cre-recombinase-positive nuclei was isolated and smoothened 4x. This was followed by respectively adjusting brightness:contrast to 30:35. Particles (red-stained nuclei) were then analyzed by setting the range of pixel sizes to 250-infinity, and of circularity to 0.5-1.0. Numbers and average sizes of particles (red-stained nuclei) were recorded. Cells at the periphery of sections, including the epicardium, were excluded during both manual and automated counting.

TUNEL Labeling & Counting
Apoptosis was assessed using the DeadEnd Fluorometric TUNEL System (Promega #G3250) per the manufacturer's recommendations. Total numbers of TUNEL-positive nuclei in each section were manually counted at 400x magnification. TUNEL signal was counted only if it was confined to a DAPI-positive nucleus. Adherence to this convention likely underestimated total numbers of TUNEL-positive nuclei in 8 and 12 week-old hearts wherein nuclei were deteriorating. Nuclei were scored as TUNEL-positive only if at least 50% of the nucleus was filled with signal. Within each heart, 1,500-5,000 cells were enumerated to calculate percentages of labeled cells, relative to total numbers of DAPI-stained nuclei that were counted by automated scanning.

Results
The initial objective of this study was to compare the effects of depleting Tip60 from CMs in the left ventricle of experimental mice having the Kat5 LoxP/-;Myh6-Cre genotype, with control mice having the Kat5 LoxP/+;Myh6-Cre genotype, at 4, 8 and 12 weeks of age. These genotypes were selected because the only genetic variable was the presence or absence of the null allele; hence, the effects, if any, of Cre protein [19] were self-canceling. To increase efficiency of Tip60 depletion, experimental mice were bred to contain a null Kat5 allele instead of a second LoxPflanked allele, since only one Kat5 allele had to be recombined, and because our previous work had revealed that Kat5 +/heterozygous mice express Tip60 protein at near wild-type levels [9]. Breeding was adjusted so that only one copy of the Myh6-Cre transgene was present. Because this transgene is located on the X-chromosome, causing Tip60 to be depleted in gender-specific fashion, only male mice were evaluated except as shown in S2 Fig. All assessments were restricted to the ventricular myocardium. The LoxP sites located in introns 2 and 11 of the Kat5 gene (Fig 1, upper) were designed to undergo recombination induced by Cre-recombinase, the Myh6-Cre driver for which becomes strongly expressed in the ventricles near the time of birth [17]. This resulted in ablation of approximately two-thirds of the coding region, including Tip60's chromo and acetylase domains (Fig 1, right). As shown in Fig 1, PCR amplification of the Kat5 allele in Kat5 LoxP/-; Myh6-Cre mice revealed that the LoxP-targeted allele was disrupted in tissue-specific fashion, as revealed by the 444 bp product indicative of recombination in the heart, which was not detected in liver samples (Fig 1, lower).
The effect of Kat5 disruption on Tip60 mRNA and protein levels was respectively evaluated by RT-PCR at weeks 4, 8 and 12, and by western blotting at week 12. The former revealed that mRNA levels were reduced by approximately 75% at 4 and 8 weeks, followed by a less extensive decrease at 12 weeks (Fig 2A-2C). Tip60 protein levels relative to GAPDH were similarly depleted in 12 week-old Kat5 LoxP/-;Myh6-Cre hearts (Fig 2D). Note in Fig 2D that the normalized level of Tip60 protein depletion in 12 week-old Kat5 LoxP/-;Myh6-Cre hearts is likely over-estimated due increased GAPDH levels, which may reflect glycolytic/hypoxic changes or more likely the ingression of fibroblasts and inflammatory cells into the myocardium; this may also explain the apparent increase in mRNA levels in these hearts. Importantly, Tip60 levels were more extensively reduced in Kat5 LoxP/-;Myh6-Cre hearts ( Fig 2D) than in Tip60-haploinsufficient (Kat5 +/-) hearts [9], wherein Tip60 protein was maintained at near-normal levels. Finally, it is emphasized that because CMs constitute a minority cell type in the murine heart [20,21], complete depletion of Tip60 in these heterogeneous tissue samples was not anticipated, despite the complete depletion of cardiomyocytes from Kat5 LoxP/-;Myh6-Cre hearts described below.
As shown in Fig 3A, hearts in Tip60-depleted male mice were slightly diminished in size by 4 and 8 weeks of age. Remarkably, as shown in Fig 3B, these mice began to die after the 8 week timepoint, with few survivors remaining by 12 weeks of age. The echocardiographic assessments shown in Fig 4 revealed that although systolic function (%FS) was normal at 4 weeks, heart mass was reduced (Fig 4A), consistent with the assessment of heart weight/tibia length at that age (Fig 3A). By 8 weeks, systolic parameters were altered, culminating in significantly reduced fractional shortening (Fig 4B).  These changes in Tip60-depleted hearts of male mice were accompanied by increased expression of the pathologic hypertrophy marker atrial natriuretic factor (ANF) at 4 weeks, followed by significantly increased levels of ANF as well as β-myosin heavy chain (β-MHC) at 8 weeks (S3 Fig). Increased levels of interstitial fibrosis were also seen by 4 weeks (Fig 5). Perhaps most remarkably, content of connexin-43 (Cxn43) protein was substantially depleted in 4 and 8 week-old hearts (Fig 6E and 6F); these reductions were accompanied by dysmorphic patterning of Cxn43 expression in Tip60-depleted hearts (Fig 6A-6D).  The results shown in Figs 3-6 demonstrated that CM survival was crucially dependent on the presence of Tip60 protein. To ascertain Tip60's crucial role in the myocardium, it was decided to assess the effects of Tip60 depletion on CM apoptosis and proliferation at younger ages, beginning with 2 week-old hearts. Because Tip60 had previously been shown to induce apoptosis in cancer cells [3,4,22], and more recently in cultured CMs under stress [23], we attempted to assess, despite its normally low levels in non-stressed myocardium, whether the incidence of apoptosis was further reduced by Tip60 depletion. As revealed by TUNEL staining (Fig 7E), apoptosis was not altered in 2 week-old Kat5 LoxP/-;Myh6-Cre hearts. However, at 4 weeks of age, percentages of TUNELpositive nuclei were significantly and strongly reduced in the left ventricle. Thereafter, as anticipated by the onset of pathogenesis at week 4 (Figs 3-6), a trend towards significantly increased apoptosis was observed. Although caspase-3 immunostaining did not detect the decline in apoptosis revealed by TUNEL at week 4, observations of increased apoptosis at weeks 8 and 12 were corroborated (S4E Fig). Because Tip60 had previously been shown to inhibit the cell-cycle [4,24,25], we also assessed whether CM proliferation and/or CM density was increased in Tip60-depleted cells (Fig 8). Increased proliferation of total myocardial cells (5'-BrdU-positive cells in Fig 8Cc, lower), most of which were presumably fibroblasts (consistent with fibrosis; Fig 5) and/or inflammatory cells, was detected in 8 week-old Tip60-depleted hearts. Although increased CM proliferation was not observed at any timepoint (Fig 8Cc, upper), significantly increased CM density was seen in both 2 and 4 week-old Tip60-depleted hearts (Fig 8Ae). It may also be noteworthy that CM nuclei were significantly enlarged in 2 week-old Tip60-depleted hearts (Fig 8B). By 8 weeks, CM density was unchanged. Remarkably, this was followed by the complete fallout of CMs from Tip60-depleted hearts by 12 weeks of age (Fig 8Ae).

Discussion
Tip60, a pleiotropic protein, has been shown to induce the DNA damage response (DDR [2,26]), apoptosis [3,4], and cell-cycle regulation [5] in cancer cells. Previous work performed here [9] and in another laboratory [23] has indicated that Tip60 may have similar functions in cardiomyocytes (CMs). However, the effect of specifically and extensively removing Tip60 from in vivo CMs in the absence of stress remained unknown. This report describes initial experiments designed to address this question. Our findings indicate that protracted depletion of Tip60 from CMs in the in vivo left ventricle from the time of birth is incompatible with survival, resulting in complete CM fallout (Fig 8Ae), and death (Fig 3B) of the animal, between 8-12 weeks of age. Although this study focused on ventricular changes, it must also be considered that depletion of Tip60 from CMs in the developing atria, which is anticipated when employing the Myh6-driven cre transgene [18], may have contributed to the ventricular phenotype; for example it has been shown that mutations that specifically disrupt the developing atria also affect form and function in the developing ventricle [27,28]. In any event, however, our results show that as in pluripotent cells of the blastocyst [6], Tip60 is necessary to maintain survival of differentiated CMs.
We previously addressed this question using Tip60 heterozygous (Kat5 +/-) mice, wherein a single Kat5 allele was disrupted in all cells (i.e. global ablation). Unlike Kat5 -/mice which die as blastocysts [6], Kat5 +/mice maintain Tip60 protein at near-normal levels while exhibiting no reproductive or haploinsufficient defects at any stage [9]. To elicit a haploinsufficient response in the heart, stress was applied by over-expressing c-Myc (to induce the cell-cycle [10]) or by trans-aortic constriction (to induce CM hypertrophy). This resulted in cell-cycle activation and reduced apoptosis [9], consistent with Tip60's role in cancer cells. In the current study, with the exception of reduced TUNEL staining observed in 4 week-old hearts (Fig 7E), which was not corroborated by caspase-3 staining (S4 Fig), these effects were not phenocopied. We speculate that this indicates that the lethal effects of protracted and extensive Tip60 depletion, which culminate in complete CM fallout and death by 12 weeks (Figs 8Ae and 3B), overshadowed responses reflecting Tip60's function in CMs at all the stages examined here, even as early as 2 weeks. We further speculate that effects supporting the hypothesis, such as extended cell-cycle transit, may have transiently occurred at early neonatal stages that were not examined in this study. The latter possibility is consistent with the observation of increased CM density in 2 and 4 week-old hearts (Fig 8Ae).
The identification of endogenous factors that maintain CMs in proliferative senescence is currently of high interest (reviewed in [29]). To date, tumor suppressors including retinoblastoma [10] and components of the Hippo pathway [11], as well as Meis1 [30], have been implicated in CM cell-cycle inhibition. Whether Tip60, which has been accorded tumor suppressor status [31], is among these factors remains unresolved. Identification of these factors may be hampered by the CM's evolution of highly differentiated cellular architecture to accommodate cardiac function, including sarcomeres and intercalated discs, since recent evidence indicates that cell-cycle re-activation requires their de-differentiation ( [32,33], reviewed in [34]). A major component of the intercalated disc, connexin-43, which has been shown to inhibit CM proliferation [35,36], becomes strongly depleted several weeks prior to end-stage heart failure in afflictions ranging from myocarditis to ischemic heart disease (reviewed in [37]). It is therefore interesting to consider the possibility that the marked reduction of connexin-43 in heart disease, as well as during Tip60 depletion (Fig 6), indicates an abortive attempt of CMs to reenter the cell-cycle in order to mediate cardiac regeneration.
It was recently reported that the transition from an hypoxic to a normoxic environment at the time of birth causes oxidative DNA damage in CMs, in response to which the DDR is initiated, which in turn causes CM replicative senescence [38]. In that study, suppression of DNA damage until P7 by maintaining mice in an hypoxic environment, or by scavenging reactive oxygen species (ROS), increased neonatal CM proliferation; the latter treatment also caused a significant increase in CM numbers by P14, similar to the increase in CM density in Tip60-depleted hearts we observed at the same age (Fig 8Ae). Hence it is of extraordinary interest that a wealth of accumulating evidence now exists showing that cells initially respond to DNA damage by activating Tip60, which in turn induces the DDR by acetylating targets including ATM and histone H2AX [1,2,26]. It will therefore be of considerable interest to assess whether, at the time of birth, Tip60 responds to oxidative DNA damage by initiating the DDR, thereby inducing arrest of the CM cell-cycle.
In summary, these results demonstrate that Tip60 protein is necessary for long-term CM vitality and survival. It nonetheless remains possible that transient, less exhaustive depletion of this crucial protein from CMs permit cell-cycle transit while reducing apoptosis. We are addressing these possibilities at early neonatal stages using this model, and at adult stages following acute depletion of Tip60 from tamoxifen-treated Kat5 LoxP/LoxP;Myh6-MerCreMer mice.
Supporting Information S1 Fig. Targeting the Kat5 (Tip60) Gene. The targeting vector (Panel A) was prepared by recombineering to introduce LoxP sites into introns 2 and 11 of the Kat5 gene. This permits Cre-recombinase-mediated excision of exons 3-11, which consists of 71% of the exon structure including the acetyltransferase domain. Correctly targeted ESCs (line V6.5) were injected into C57Bl6 blastocysts, generating 100% chimeric males that transmitted the floxed allele via germline, providing the genotype Kat5 LoxP(FRTneoFRT)/+ . The neomycin phosphotransferase (neo) gene in intron 2, which was flanked by FRT sites, was removed by mating with a mouse expressing the Flp recombinase transgene. Kat5 LoxP/mice, obtained by mating Kat5 LoxP/LoxP mice with Kat5 +/mice, were used for all experiments reported here because this necessitates deletion of only one allele to maximally reduce Tip60 protein levels. Panel A, Restriction Map of the LoxP-targeted Kat5 allele. Bars denote Southern blot probes A-D. Arrowheads denote PCR primer sites. Panel B, PCR Genotyping: The primer pair denoted by green arrowheads (A) amplifies a 250 bp band of the WT allele, which does not contain the previously removed Neo cassette. The 5' green + 3' blue (in neo) arrowhead pair in A amplifies 580 bp of the LoxPmodified allele. The red-black arrowhead pair amplifies 2,301 bp of the LoxP-modified allele (note that the red primer anneals to the LoxP locus). Panel C, Southern Blot Genotyping. a, double digest with Hind III (H3) and XhoI (within the LoxP site) followed by hybridization with probe B reveals a 3391 band diagnostic of the floxed allele, and a 7,587 band diagnostic of the WT allele; b, digestion with KpnI and hybridization with probe C yields a 7270 bp band from the WT allele, and a 6362 bp fragment created by a KpnI site within the neo cassette of the floxed allele; c is the same as b except that hybridization was with probe A, an internal probe used to ensure that the targeting vector did not randomly insert into a non-endogenous target(s). . In E, entire sections from each heart were scanned at 60x, during which all caspase-3-positive cells in each section were manually enumerated, followed by normalization to the total number of nuclei assessed by automated scanning of DAPI-stained immediately adjacent sections. F/-;cre denotes hearts from Kat5 LoxP/-;Myh6-Cre mice; F/+;cre denotes hearts from Kat5 LoxP/+;Myh6-Cre mice. The number above each bar indicates the number of mouse hearts that were evaluated. Vertical lines = ±SEM. P-values were calculated using Student's T-Test (two-tailed). All CMs contained the Myh6-Cre transgene.