|Home | About | Journals | Submit | Contact Us | Français|
We generated transgenic mouse line C57BL/6-Tg(Hspa2-cre)1Eddy/J (Hspa2-cre), which expresses cre-recombinase under the control of a 907-bp fragment of the heat shock protein 2 (Hspa2) gene promoter. Transgene expression was determined using Gt(ROSA)26Sortm1Sor/J (ROSA26) and Tg(CAG-Bgeo/GFP)21Lbe/J (Z/EG) reporter strains and RT-PCR and immunohistochemistry assays. Hspa2-cre expression mimicked the spermatogenic cell-specific expression of endogenous HSPA2 within the testis, being first observed in leptotene/zygotene spermatocytes. Expression of the transgene also was detected at restricted sites in the brain, as occurs for endogenous HSPA2. Although the results of mating the Hspa2-cre mice to mice with a floxed Cdc2a allele indicated that some expression of the transgene occurs during embryogenesis, the Hspa2-cre mice provide a valuable new tool for assessing the roles of genes during and after meiotic prophase in pachytene spermatocytes.
Spermatogenesis is a complex, coordinated developmental process characterized by mitotic proliferation of spermatogonia, dramatic transformation in nuclear content and chromatin organization in meiotic spermatocytes, and differentiation of postmeiotic haploid spermatids into spermatozoa capable of fertilization. The use of targeted mutagenesis in embryonic stem cells to generate gene knockout mice has contributed significantly to understanding how these processes are regulated. However, this approach has been limited to genes that are not also essential during development or for maintenance of viability during fetal and adult life. Use of cre-LoxP recombinase technology has overcome this problem for many systems, but the lack of appropriate cre-expressing mice has limited studies of genes involved in regulating cell cycle processes or other essential cellular functions during meiosis in the male. Transgenic mice exist in which cre is driven by promoters from the synaptonemal complex protein 1 (Sycp1) (Vidal et al, 1998) or phosphoglycerate kinase 2 (Pgk2) (Ando et al, 2000; Bhullar et al, 2001) genes, but weak and/or aberrant excision (Rassoulzadegan et al, 2002; Rasoulpour and Boekelheide, 2006; Ando et al, 2000; Bhullar et al, 2001) considerably limits their usefulness for studying the roles of most genes during meiosis in males.
The Hspa2 gene encodes a member of the HSP70 family of heat-shock proteins that serve as molecular chaperones. Unlike most other members of the HSP70 family, Hspa2 is developmentally regulated and expressed predominantly in spermatocytes and spermatids (Rosario et al, 1992; Dix et al, 1996a; Eddy, 1999). Translation occurs immediately after transcription in leptotene/zygotene spermatocytes and the HSPA2 protein is present at high levels in primary spermatocytes (O'Brien, 1987; Dix et al, 1997). In addition, previous studies had characterized the Hspa2 gene promoter in transgenic mice using LacZ as the reporter (Dix et al, 1996b). This led us to generate an Hspa2-cre line that expresses cre in spermatocytes to overcome the limitations of the other transgenic lines.
A 907-bp region of the mouse Hspa2 gene proximal promoter was used for constructing the Hspa2-cre transgene (Fig. 1a). It contained a 640-bp fragment 5' to the Hspa2 transcription start codon shown previously to contain the minimal promoter region sufficient to drive expression of a lacZ transgene in spermatocytes (Dix et al, 1996b). In an attempt to reduce positional effects on transgene expression, we added to each end of the transgene a 53-bp region of the SP-10 (acrosomal vesicle protein 1, Acrv1) spermatid-specific promoter that had been shown to function as an insulator in somatic cells (Reddi et al, 2003). Additional studies will be required to determine if the addition of insulators was beneficial.
Pronuclear microinjection resulted in the production of seven independent transgenic lines carrying the Hspa2-cre transgene. Founders were identified by PCR and Southern analysis of genomic DNA (Fig. 1b and data not shown). Four of the lines efficiently transmitted the transgene and line 54 was selected for further analysis based on the level and tissue specificity of transgene expression seen in preliminary studies. Transcriptional activity of the Hspa2-cre transgene was assessed by RT-PCR using cDNA from brain, heart, liver, kidney, spleen, testis, and ovary. Expression of the Hspa2-cre transgene was robust in testis and detectable in brain (Fig. 1c), consistent with previous RT-PCR results for Hspa2 expression (Dix et al, 1996b).
The ROSA26 reporter strain was used to verify that Hspa2-cre activity was targeted to spermatogenic cells in the testis. Tissues were collected from postnatal day 20 Hspa2-cre;ROSA26 compound transgenic males and wild type littermates (Fig. 2). In the presence of cre activity, the loxP-flanked DNA STOP sequence located within the lacZ gene of ROSA26 reporter mice is excised and lacZ expression is detected by X-gal staining (Soriano, 1999). X-gal staining occurred in the testis in Hspa2-cre;ROSA26 males (Fig. 2b), but not in wild type littermates (Fig. 2a). Enzymatic activity was present in spermatocytes in day 20 animals and by day 30 was present additionally in spermatids, but not in spermatogonia at the periphery of the seminiferous tubule, consistent with previous reports of Hspa2 expression (Rosario et al, 1992, Dix et al, 1996a). Expression of Hspa2-cre was detected in the brain of some transgenic animals by X-gal staining (Fig. 2f), but not in the brain of wild type animals (Fig. 2e), consistent with RT-PCR results. Endogenous HSPA2 expression in brain of juvenile animals (Figs. 2g and 2h) was detected by immunohistochemistry. HSPA2 was observed primarily within the ependymal cells of the lateral ventricles, confirming other recent findings with an Hspa2-GFP transgene (unpublished observations). Transgene expression was not detected in heart, muscle, lung or liver (data not shown). Enzymatic activity was detected by X-gal staining in the spleen (not shown) and kidneys in both Hspa2-cre;ROSA26 (Fig. 2d) and wild type mice (Fig. 2c). However, Hspa2-cre transcripts were not detected in these tissues by RT-PCR, indicating that the staining was due to endogenous β-galactosidase.
To determine more precisely when cre recombinase expression occurs during spermatogenesis and the recombination efficiency, Hspa2-cre males were mated to Tg(CAGBgeo/GFP)21Lbe/J (Z/EG) reporter females that express enhanced green fluorescent protein (EGFP) after cre-mediated excision (Novak et al, 2000). Cre expression was determined on sections of testes from juvenile and adult males whose genotype indicated they inherited both the cre-expressing and reporter expressing transgenes (Hspa2-cre;Z/EG) by immunohistochemistry with an antibody to GFP. GFP was expressed in all offspring with the Hsp2a-cre transgene and the Z/EG reporter, indicating a recombination efficiency of 100% (n=13). At postnatal day 14, GFP immunostaining was detected during the first wave of spermatogenesis within some seminiferous tubules of transgenic mice (Fig. 3a). Non-specific staining in the interstitium was caused by the second antibody (data not shown). Using an antibody to HSPA2, it was determined that GFP expression was coincident with HSPA2 expression in leptotene/zygotene spermatocytes (Fig. 3b). The images shown are from sections of the same testis. Neither GFP nor HSPA2 were detected in spermatogonia at the periphery of the seminiferous tubules and GFP was not detected in any germ cells in wild type animals (data not shown).
At postnatal day 20, GFP was detected in pachytene spermatocytes (Fig. 3c). The localization of cre expression during the meiotic phase of spermatogenesis in Hspa2-cre;Z/EG males and not in spermatogonia at the periphery of the seminiferous tubule, again coincided with the pattern of HSPA2 expression (Fig. 3d). These results confirmed that the Hspa2-cre transgene was transcribed and translated similarly to the endogenous Hspa2 gene during spermatogenesis.
Upon mating Hspa2-cre transgenic mice to mice with a floxed Cdc2a allele, we discovered recombination-mediated excision events were not restricted to spermatocytes. Excision of the floxed gene was detected by PCR analysis of tail biopsies of offspring resulting from matings of Cdc2a mice with one wild-type allele and one floxed allele (Cdc2a+/floxed) with Hspa2-cre male or female transgenic mice (data not shown). This generated offspring with one wild-type allele, one disrupted Cdc2a allele and the Hspa2-cre transgene (Cdc2a+/-;Hspa2-cre). Excision of the floxed allele occasionally was incomplete, allowing the detection of all three Cdc2 allele forms (Cdc2a+/floxed/-;Hspa2-cre) within somatic tissues. In most cases, the excised allele appeared predominant over the floxed allele. RT-PCR analyses determined that Hspa2-cre expression did not occur at detectable levels in somatic tissues or ovary (Fig.1c), suggesting excision of the floxed allele was a result of transgene expression during embryogenesis. This was confirmed by analysis of Mendelian ratios of the genotypes of offspring resulting from mating studies. When Cdc2a+/floxed/-;Hspa2-cre male mice were crossed with Cdc2afloxed/floxed females, 25% of offspring were expected to be Cdc2afloxed/-;Hspa2-cre mice, but only 7.8% were of this genotype (Table 1). Similar ratios were observed when female Cdc2a+/-;Hspa2-cre mice were mated with Cdc2afloxed/floxed males (data not shown). These results demonstrate Hspa2-cre transgene expression occurs during embryogenesis and confirm a previous report that CDC2A is required for embryonic development (Santamaría et al, 2007). These results also indicate that the Hspa2-cre transgene is either not expressed in some embryos or is expressed at a low level during embryogenesis that allows some floxed Cdc2a alleles to escape excision.
Expression of an EGFP transgene driven by the rat Hspa2 (Hst70) gene promoter during embryogenesis was reported recently (Rupik et al, 2006). EGFP expression coincided with the period of organogenesis, but was noted primarily in tissues involved in the development of the nervous system. Initial characterization of the rat Hspa2 promoter in transgenic mice indicated expression occurred similarly to HSPA2 in the mouse (Widlak et al, 1995), but later studies reported that expression occurred in a wider variety of tissues including the ovary, oviduct and uterus (Scieglińska et al, 1997). While the rat promoter might be regulated differently in the mouse or the expression pattern seen might be the result of ectopic expression, this suggests a closer examination of Hspa2 expression during mouse embryogenesis is warranted.
Although expression of the Hspa2-cre transgene during embryonic development limits the usefulness of these mice for studying cell cycle genes essential for both embryonic development and meiosis, they represent a useful tool for defining the roles of genes expressed at different times during spermatogenesis or expressed in spermatogenic cells and in other tissues in the adult.
A 907-bp promoter fragment, corresponding to the mouse Hspa2 genomic sequence-932/-25 relative to the start codon, (Dix et al, 1996b) was amplified from 129 SvEv genomic DNA using primers designed with a SpeI recognition site at the 5' end of the forward primer (5'-CCGACTAGTAGGAAAGCCGAGGGAGAAAGTT) and an XhoI recognition site on the 5' end of the reverse primer (5'-TACCTCGAGAACGTTAGGACGAAAGCGTCAG). The resulting product was cloned into the pBS185 plasmid 5' of the cre recombinase coding sequence, replacing the hCMV promoter. The Hspa2-cre fragment was then amplified and subcloned into pBluescript II KS(+) (Stratagene, La Jolla, CA) using engineered HindIII and PstI sites. A 3' bovine growth hormone (BGH) polyadenylation sequence was added to the 3' end. In an attempt to reduce positional effects on expression of the transgene, the 53-bp Acrv1 insulator sequence (Reddi et al, 2003) was amplified from the CMVmin-91 Luc plasmid and placed 5' of the Hspa2 promoter and 3' of the BGH polyadenylation sequence (Fig. 1a).
A 2.8-kb fragment produced by digestion with AhdI and PvuII was purified and microinjected into pronuclei of fertilized eggs from C57BL/6 mice. Transgenic animals were produced by the NIH Transgenic Mouse Development Facility (University of Rochester, contract # NO1-DE-12634). The resulting transgenic line was named C57BL/6-Tg(/Hspa2-cre)1Eddy/J (Hspa2-cre) and has been transferred to the Jackson Laboratory (Bar Harbor, ME) for future distribution (JAX Stock Number 008870).
To assess cre expression, transgenic mice were crossed with Gt(ROSA)26Sortm1Sor/J (ROSA26) and Tg(CAG-Bgeo/GFP)21Lbe/J (Z/EG) reporter mice. All animal studies were approved by the NIEHS Institutional Animal Care and Use Committee and carried out according to U.S. Public Health Service (USPHS) guidelines.
Transgenic founders were identified by PCR screening of genomic DNA isolated from tail biopsies using the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA) with primer pairs IHGI 5089F (5'-TTGAAGCTACCCCCTAACACACTA)/cre 1 4 8 7 R ( 5 '-TTGCCCCTGTTTCACTATC); and cre 484F (5'-AATGTCCAATTTACTGACCGT)/IHGI 2148R (5'-TTGAAGCTACCCCCTAACACACTA). Southern analysis was used to confirm transgene integration. Briefly, genomic DNA isolated from the testis was digested with HindIII and SpeI, separated on a 1.0% agarose (w/v) gel, and transferred to Hybond-N nylon membrane (GE Healthcare-Life Science, Piscataway, NJ). A 1003-bp probe, corresponding to the cre recombinase-coding region was amplified with primers cre 484F and cre 1487R. For routine genotyping the generic cre standard PCR protocol (version 1) from The Jackson Laboratory was used. Products were separated on 2% (w/v) agarose gels.
To assess the level of cre transcription, total RNA was extracted from brain, heart, liver, kidney, spleen, testis and ovary using Trizol Reagent (Invitrogen, Carlsbad, CA) according to manufacturer's instructions. Total RNA (1 μg) was treated with RNase-free DNase I and reverse transcribed using oligo dT primers. A 1003-bp fragment, indicative of cre expression, was amplified with primers cre 484F and cre 1487R. Gapdh was amplified as a control transcript. Products were separated on 2% (w/v) agarose gels.
Tissues were fixed in Bouins fixative (Sigma-Aldrich, St. Louis, MO), embedded in paraffin, and sectioned at 6 μm thickness using standard procedures. Slides were deparaffinized and incubated with a GFP antibody (Abcam, Cambridge, MA) or HSPA2 antiserum (Rosario et al, 1992). Immunolocalization was detected using the Vectastain Elite ABC kits (Vector Laboratories, Burlingame, CA) with 3,3'-diaminobenzidine tetrahydrocholoride (Sigma-Aldrich), and slides were counterstained with Hematoxylin QS (Vector Laboratories). Images were recorded using an Axioplan microscope (Carl Zeiss, Thornewood, NJ,) and QImaging camera and software (QImaging, Tucson, AZ).
Tissues were fixed overnight in 4% paraformaldehyde (PFA) in PBS followed by immersion in 20% sucrose in PBS for 16 hours at 4°C. Frozen sections were cut 8-10 μm in thickness, washed in PBS, immersed in rinse buffer [0.1M phosphate buffer (pH 7.4), 2mM MgCl2, 0.01% sodium deoxycholate, 0.02% NP-40], and stained at 37°C overnight in buffer containing 2 mg/ml X-gal (Sigma-Aldrich), 5 mM K3Fe(CN)6, and 5 mM K4Fe(CN)6. After staining, the slides were washed in PBS, post-fixed in 4% PFA, and observed using Nomarski interference microscopy.
The authors would like to thank Mr. Clyde Rogers and Mr. Linwood Koonce for their help in caring for the animals, Drs. Manas Ray and Chris Geyer for helpful suggestions on the manuscript, Ms. Pat Stockton for assistance in preparing the paraffin and frozen sections, Drs. Yuji Mishina and Yoshi Komatsu for providing the ROSA26 mice, Dr. Manas Ray for providing the pBS185 plasmid, and Dr. Prabhakara Reddi for the CMVmin-91 Luc plasmid containing the Acrv1 insulator sequence. This research was supported by the Intramural Research Program of the NIH, National Institute of Environmental Health Sciences.