Introduction

GADD45γ is a member of a family of proteins (Growth arrest and DNA damage response) implicated in DNA repair and active DNA demethylation1,2, and the activation of mitogen activated protein kinase (MAPK) signalling3. Whilst initially thought to be dispensable for mouse development4, loss of function studies reveal a role for GADD45γ in fetal testis determination via its positive effects on p38 MAPK and Sry expression5,6,7. In the adult, GADD45γ has reported roles in cardiomyocyte apoptosis following myocardial infarction8, and in regulating the thermogenic capacity of brown adipose tissue9. Careful profiling of expression in the developing cerebral cortex also suggests roles for Gadd45a, Gadd45b and Gadd45g in neurogenesis10. However, the precise molecular roles played by GADD45γ in these different physiological contexts remain to be elucidated.

The study of gene function can be greatly enhanced by the use of fluorescent reporters that allow live imaging of cellular processes. A previous attempt at generating a reporter for Gadd45g used 1767 base-pairs (bp) of DNA immediately upstream of the gene’s transcription start-site to drive Venus expression in mice11. Whilst reporter expression was clearly detectable in neuronal cell-types, there was no expression detected in the developing gonads (ref.11 and Takumi Kawaue, pers. comm.). Fetal gonadal expression of Gadd45g underlies its role in testis determination and the absence of gonadal expression in these transgenic mice suggests that enhancers required for gonadal expression are not found within this 1767 bp region.

Introducing reporters or tags into endogenous genes has been made easier by genome editing technology, but any such alterations risk disrupting the function of a gene, especially when larger sections of heterologous DNA are introduced. This may result in haploinsufficiency and unwanted phenotypic effects. An alternative approach to ensuring that all regulatory elements required for expression are present is the use of large bacterial artificial chromosomes (BACs), which contain a gene of interest and commonly large regions of adjacent DNA (up to 500 kb)12. Transgenic mice can be readily made using microinjection of BAC DNA into 1-cell embryos13 or into metaphase II (MII) oocytes, at the same time as intracytoplasmic sperm injection (ICSI)14. Moreover, recombineering technology can be used to modify the gene carried by the BAC and thereby permit study of, amongst other things, gene expression and protein localisation15. The latter is particularly useful if a good antibody is not available. But prior evidence is required that expression from the BAC recapitulates the endogenous gene.

Here we characterise a novel mouse line carrying a Gadd45g BAC transgene. This transgene allows normal sexual development in the absence of the endogenous gene. Moreover, careful examination of expression from the BAC indicates that both gonadal and broader fetal expression matches that of the endogenous gene. This BAC clone is appropriate for modification to further explore the developmental roles of Gadd45g and may be suitable for investigations of adult function. Finally, we report the generation of a Gadd45g-mCherry reporter that can be used for a variety of imaging applications.

Results

To identify BAC clones containing Gadd45g we searched Ensembl and identified two potential candidates (Fig. 1A). We selected clone RP23–194E15 (henceforth E15), a 238.3 kb BAC, on the basis of the large amount of DNA flanking the Gadd45g gene on both sides: 59.1 kb 5′ of the gene and 177.4 kb 3′ of the gene. E15 DNA was prepared and a PCR analysis was performed to ensure that all exons were present. Coding regions were sequenced to determine whether any mutations existed. Once it had passed this quality control, DNA was used to inject 1-cell mouse embryos and generate transgenic founders. Five founder mice were found to contain E15 BAC vector sequences. These founders were bred to check whether the transgene was transmitted to offspring. Four founders were found to transmit. All studies below refer to a line derived from founder 5 (line 5). This was the only line found to be capable of phenotypic rescue in experiments described below.

Figure 1
figure 1

(A) Position of BACs on chromosome 13 containing Gadd45g; (B) qRT-PCR reveals elevated expression of Gadd45g in transgenic (XY BAC, XX BAC) gonads at 11.5 dpc (18 ts) compared to XY and XX wild-type controls; (C) WMISH of Gadd45g in XY and XX gonads from wild-type (+/+), null (−/−) and E15 BAC transgenic null (−/− BAC) embryos at 11.5 dpc. Transverse sectioning (lower panel) reveals that expression is absent from the coelomic epithelium (arrow head); (D) The presence of the E15 BAC rescues XY gonadal sex reversal caused by absence of endogenous Gadd45g (−/−) based on Sox9 (Sertoli cell) and Stra8 (meiotic entry germ cells) marker expression.

Full size image

We first determined whether embryos transgenic for E15 had elevated levels of Gadd45g in the gonad at the stage of sex determination (11.5 dpc). Quantitative reverse-transcription PCR (qRT-PCR) revealed that transgenic gonads had approximately 5-fold higher levels of Gadd45g (Fig. 1B). These data were consistent with wholemount in situ hybridisation (WMISH) showing that Gadd45g expression in XY and XX gonads at this stage was stronger in homozygous Gadd45g knockout embryos when the transgene was present than in wild-type controls (Fig. 1C). No expression was detected in homozygotes in the absence of the transgene. Expression was not detected in the coelomic epithelium, as previously reported6.

Given this gonadal expression, we then tested whether the presence of the E15 transgene could rescue testis determination defects in XY Gadd45g homozygous knockouts. In the absence of Gadd45g, XY gonads have an ovarian morphology, lack Sox9 and express Stra8, a marker of germ cell meiotic entry normally restricted to the ovary (Fig. 1D). In contrast, transgenic embryos that lacked the endogenous gene formed testes like XY wild-type controls (Fig. 1D). These had high levels of Sox9 in testis cords and no detectable Stra8 expression. Thus, rescue of the sex reversal phenotype is complete in the presence of the E15 transgene.

Given this functional rescue, we sought to determine whether the onset and cessation of the BAC-encoded Gadd45g was comparable to that of the endogenous gene. This is still important to test and will determine whether this BAC is suitable for studying the temporal dynamics and cellular specificity of Gadd45g expression. We performed WMISH analysis on XY samples at 11.75 dpc (21 tail-somites (ts) stage) and 14.5 dpc using two distinct lengths of time for staining to develop: short (0.5 hrs) and long (5 hours). Wild-type samples required the longer staining time to allow signal to be reliably detected (Fig. 2). This revealed strong expression at 11.5 dpc but no detectable signal at 14.5 dpc (Fig. 2A,B), consistent with previous reports5,6,7. In contrast, expression in transgenic 11.5 dpc embryos that lacked the endogenous gene was detectable after just 0.5 hours of staining (Fig. 2C). Importantly, no expression was detected at 14.5 dpc using the same short stain (Fig. 2D), suggesting that down-regulation of expression from the BAC transgene follows the same dynamics as the endogenous gene. After longer staining, transgenic Gadd45g expression could be detected at high levels at 11.5 dpc (Fig. 2E) and to a lesser degree at 14.5 dpc, in both the testis cords and Müllerian duct (Fig. 2F). No expression was detected in mutant homozygous tissue at either stage, whether with short or long staining (Fig. 2G–J). Similar observations were made using XX samples (Supplementary Information: Fig. S1).

Figure 2
figure 2

WMISH analysis of Gadd45g expression in XY wild-type (+/+) at 21 ts (A) and 14.5 dpc (B). Expression is also shown for transgenic null (−/− BAC) gonads at 21 ts following staining for 0.5 hours (C) or 5 hours (E) and similarly at 14.5 dpc (D,F). No expression is detectable in null (−/−) gonads at any stage or length of staining (G–J). No Gadd45g expression is seen at 6ts in embryos of any genotype (K–M). Expression is observed at 9 ts in wild-type (N) and more strongly in transgenic null embryos (P) but not in null embryos (O).

Full size image

The onset of expression from the BAC-encoded Gadd45g was also the same as the endogenous gene (Fig. 2K–P). No gonadal expression was detected at very early stages (6 ts, approximately 10.2 dpc)(Fig. 2K–M), but expression was detected in wild-type and transgenic embryonic gonads at 9 ts (around 10.5 dpc)(Fig. 2N,P). Expression was restricted to the centre of the gonad, as previously reported6, and was stronger in transgenic embryos.

We then determined whether extra-gonadal expression of the BAC-encoded gene matched that of the endogenous gene by examining expression in whole embryos at 10.5 dpc. Gadd45g expression in wild-type embryos was detected in the neural tube, the trigeminal and facial acoustic ganglia, the dorsal root ganglia, the forming tail somite, the olfactory epithelium and somites (Fig. 3A–D). These sites of expression are consistent with those previously reported16. All these sites were recapitulated in the presence of the transgene and in the absence of the endogenous gene (Fig. 3E–H). As predicted, levels of expression were higher in transgenic embryos. No expression was detected in homozygous knockout embryos lacking the transgene (Fig. 3I–L).

Figure 3
figure 3

WMISH analysis of Gadd45g in wild-type (+/+), transgenic null (−/− BAC) and null (−/−) embryos at 10.5 dpc. Transverse sections of all three genotypes are shown in (D,H,L) (at the level indicated in panel (A). No signal is detected in null embryos (I–L). Inset in panel A shows forming somite. mb, midbrain; fb, forebrain; hb, hindbrain; oe, olfactory epithelium; so, somite; drg, dorsal root ganglion; tg, trigeminal ganglion; fg, facial ganglion; nt, neural tube; fs, forming somite. Scale bar = 50 μm.

Full size image

Finally, as a proof of principle, we modified the E15 BAC using the Red/ET recombineering method17 in order to introduce a fluorescent mCherry reporter into the Gadd45g open reading frame after the first codon (Fig. 4A). Seven lines of founder mice were generated by microinjection of MII oocytes and these were used in timed mates to generate embryos between 10.5 and 11.5 dpc. Embryos from three founders exhibited Cherry fluorescence in the developing gonads of both XX and XY individuals (Fig. 4B–D and data not shown). Wholemount immunostaining of 11.5 dpc transgenic gonads with anti-PECAM antibody, which marks germ cells and endothelial cells, revealed Cherry localisation to most, but not all, gonadal somatic cells. Imaging of 10.5 dpc embryos revealed Cherry signal in the neural tube and other reported sites of expression at this stage (Fig. 4E–G). This reporter line can now be used for a variety of imaging and cell sorting applications.

Figure 4
figure 4

(A) Position of the mCherry reporter in the Gadd45g open reading frame of E15 BAC; (B) Detection of Cherry fluorescence in 11.5 dpc embryonic gonad. Green signal indicates location of PECAM. Blue is DAPI staining. (C) Cherry signal is detected in somatic cells of the gonad, which lack PECAM staining shown in (D). A few somatic cells (lacking PECAM) also lack Cherry signal (white arrows). (E) Lateral view of 10.5 dpc embryo showing Cherry fluorescence in developing neural tissue (forebrain (fb), midbrain (mb), hindbrain (hb)). (F) Dorsal view of same embryo reveals signal in neural tube (nt), trigeminal ganglion (tg) and facial ganglion (fg). (G) Section of embryo (in plane indicated by dotted line in (E)) shows neural tube and dorsal root gangion (drg) fluorescence. Scale bar = 50 μm.

Full size image

Discussion

We show here that the E15 BAC contains sequences that drive embryonic expression of Gadd45g in a pattern indistinguishable to that of the endogenous gene at the stages examined. This recapitulation extends to the temporal dynamics of Gadd45g expression in the gonad. The conservation of this profile suggested that E15 would make a good substrate for the generation of a reliable reporter of embryonic gonadal Gadd45g expression and probably other sites of expression. The Gadd45g-Cherry reporter described here, for example, could be used in live imaging of the behaviour of the supporting cell lineage in both XY and XX gonad development from the earliest stages, as has been performed for other lineages18.

BAC transgenic expression of Gadd45g was reliably detected at significantly higher levels to that of the endogenous gene. This increased expression did not result in any overt phenotypic abnormalities and may make it easier to detect sites of expression not easily detected in wild-type mouse embryos. However, caution should be exercised in how additional sites of expression in transgenic embryos are interpreted. For example, the Müllerian duct expression readily detected in E15 transgenic embryos at 14.5 dpc is much more difficult to detect in wild-type embryos (see Fig. 2B), but it is unclear whether this is due to the ease of detection of the elevated transcript levels of the transgenic tissue or whether occasional ectopic expression is observed, either due to missing inhibitory elements or disturbance to the transcriptional regulation of Gadd45g caused by the copy number of the BAC transgene. Independent verification of such novel sites would be required, but, if verified, they may also suggest that particular, additional phenotypic tests should be performed on the knockout.

The characterisation of the E15 BAC reveals it to be suitable for modification by recombineering as a means of studying the embryonic role of GADD45γ. It remains unclear as yet whether all adult sites of expression of Gadd45g are supported by this BAC; however, given the extensive neuronal gene expression detected in reporter mice transgenic for just 1767 bp of DNA upstream of the Gadd45g transcription start-site, it is likely that all the sites of expression reported by Kawaue et al.11 will be recapitulated in the E15 transgenic line. Definitively establishing that all sequences required for a particular function are present in the BAC requires rescue of the phenotype associated with loss of these sites of expression, as performed here for testis determination.

In summary, we identify here a BAC (E15) that supports functional Gadd45g transgenesis. All endogenous embryonic sites of expression tested appear to be conserved by expression from the BAC transgene and this clone represents a suitable tool for the generation of genetic modifications aimed at marking or tagging the gene. We describe the generation of a Gadd45g-mCherry reporter line using BAC recombineering that exhibits robust embryonic gonadal and extra-gonadal expression; other modifications, including the introduction of Cre recombinase or a highly efficient epitope tag, can be envisaged. Given the hypothesized role of changes to the GADD45g regulatory region in the evolution of human-specific traits19, a functional mouse Gadd45g transgene can be a very useful tool for further study of the regulation and function of this important gene.

Methods

Mouse strains and ethical approval

All mouse experimentation was approved by the Animal Welfare and Ethical Review Body (AWERB) at the MRC Harwell Institute, UK. Mice used were bred with licensed approval from the UK Home Office (PPLs 30/2877 and 70/8988). Mice were housed in individually ventilated cages (IVCs) in a specific pathogen-free (SPF) environment. Further details of micro- and macro-environmental conditions are available on request. Mice lacking Gadd45g have been previously described6. All lines were generated and maintained on C57BL/6 J (B6).

Generation of BAC transgenic mice

The RP23–194E15 (E15) BAC was initially characterised to ensure that all Gadd45g exons were present and free of mutation. Microinjection of single-cell B6 embryos with E15 BAC DNA was performed as previously described13,20. Co-injection of metaphase II oocytes with the modified E15-Cherry BAC and sperm was performed as described14. The E15 transgene was detected by amplification of the chloramphenicol resistance gene (Cmr)20. In order to identify E15-positive, Gadd45g−/− mice, Cmr detection was used to confirm presence of the BAC; a qPCR copycaller assay (Applied Biosystems) for neo was used to detect the Gadd45g null allele and a qPCR copycaller assay was also used to detect copies of Gadd45g, which were higher in the BAC transgenics. mCherry-positive mice were detected by PCR. All primer sequences are available on request.

BAC recombineering

BAC recombineering was performed using the counter-selection BAC modification kit (Gene Bridges) according to manufacturer’s instructions. Briefly, a rpsL-neo targeting cassette was generated that included Gadd45g homology arms. In the first step, this was electroporated into E15-positive E. coli cells to confer resistance to kanamycin and sensitivity to streptomycin. A construct containing mCherry, nuclear localization signal and SV40pA terminator derived from Addgene plasmid GW1-Peredox-mCherry-NLS (#32381, a kind gift from Gary Yellen) was used to replace the rspL-neo cassette in a second round of transformation, restoring streptomycin resistance. Unmodified, rpsL-neo integrant and mCherry recomineered BAC DNA samples were digested with ScaI and resolved on a 0.5% agarose gel to reveal 4.35 kb, 5.8 kb and 5.3 kb fragments, respectively, confirming fidelity of the two-step modification process. All other ScaI bands remained unchanged, suggesting that no unwanted modifications had occurred.

Generation of embryos

Noon on the day of the copulatory plug was counted as 0.5 dpc. Adult mice were humanely sacrificed by dislocation of the neck, confirmed by palpation, and embryos were decapitated in ice-cold, phosphate buffered saline solution. Embryos collected at 11.5 dpc were staged accurately based on the number of tail somites (ts).

Wholemount in situ hybridization

Wholemount in situ hybridization (WMISH) analysis of embryonic tissues and probes for Sox9 and Stra8 have been previously described6,21. A Gadd45g probe for exons 2–4, which are absent from any transcript from the targeted allele, was generated by RT-PCR from gonadal RNA. At least three independent biological samples from a given group were analysed with a particular marker.

Quantitative reverse transcription PCR (qRT-PCR)

Total RNA was extracted using RNeasy plus micro kit (Qiagen) from gonads separated from the mesonephros. Reverse transcription (RT) was carried out with 200 ng of total RNA using the High capacity cDNA RT kit (Applied 24 Biosystem). Quantitative RT-PCR (qRT-PCR) was performed with Fast SYBR Green Master Mix (Life technologies) on a 7500 Fast Real-Time PCR system (Applied Biosystems). RNA expression levels were normalized to those of Hrpt1 (endogenous control) using the ΔΔCt method. At least three samples for each genotype were analysed. Primer sequences are available on request.