Kaibo Lin,Shikun Zhang,Jieli Chen,Ding Yang,Mengyi Zhu,Eugene Yujun Xu

State Key Laboratory of Reproductive Medicine,Nanjing Medical University,Nanjing,Jiangsu 211166,China.

Abstract The highly conserved RNA binding protein PUF(Pumilio/FBF)family is present throughout eukaryotes from yeast to mammals,with critical roles in development,fertility and the nervous system.However,the function of the mammalian PUF family members remains underexplored.Our previous study reported that a gene-trap mutation of Pum2 results in a smaller testis but does not impact fertility and viability.Although the gene-trap mutation disrupted the key functional domain of PUM protein–PUM-HD(Pumilio homology domain),but still produced a chimeric Pum2-β-geo protein containing part of PUM2,raising a question if such a chimeric protein may provide any residual function or contribute to the reproductive phenotype.Here,we report the generation of a conditional PUM2 allele,when knocked out,producing no residual PUM2 and hence a complete loss-of-function allele.We also uncovered small but significant reduction of male fertility and viability in the mutants,suggesting requirement of PUM2 for male fertility and viability.

Keywords:RNA-binding protein,Pumilio2,PUF,fertility,viability

Introduction

ThePumilio2(Pum2)gene is a mammalian member of the classical RNA binding protein family,the PUF family(Pumilio and FBF).The PUF family genes are widely present in eukaryotes,from yeast to mammals.ThePumiliogene was first identified in Drosophila by Lehmann and Nüsslein-Volhard in 1987[1].InC.elegans,FBF(for fem-3 binding factor)was reported as ahomolog of DrosophilaPumilioin 1997[2].FBF and Pumilio homologs hence constitute a large and evolutionarily conserved RNA binding protein family,thereafter called the PUF family.The PUF family is characterized by its highly conserved RNA binding domain PUM-HD(Pum Homolog Domain).PUM-HD consists of eight Puf repeats with about 40 amino acids eachrepeat and is essential for RNA binding by the PUF protein[2–4].Furthermore,PUM-HD is reported to interact with other proteins[5–6].Members of the PUF family modulate target gene expression at posttranscriptional levelsviabinding to the PBE(PUM Binding Element)motif(UGUAHAUA)in the 3′UTR of target transcripts,and thus regulate developmental processes among various organisms[7–8].

Invertebrate members of the PUF family play significant roles in embryogenesis and germ cell development.DrosophilaPumilio protein controls anterior-posterior axis formation,maintenance of germ-line stem cells and neuron developmentviaposttranscriptional regulation[9–14].InC.elegans,FBF1 and FBF2 are involved in the regulation of sperm-oocyte switch,maintenance of germline stem cell and the size of germline mitotic region[2,5,15–16].These findings suggest that highly conserved mammalian PUM proteins may also play important roles in development[8,17].

Previously mousePum2XE772/XE772mutation was reported to affect nervous system function and testis size but not fertility or viability in mice[18–19].ThisPum2mutation resulted from a gene trap mutation where the β-geo gene was inserted between exon10 and exon11ofPum2.The insertion mutation led to a PUM2-β-geo fusion protein,removing most part of PUM2 protein after exon 10 including the most conserved PUM-HD domain[18].While thisPum2allele should disrupt the key function of PUM2 protein in its ability to bind its target sequences,it remains unknown if all the other functions of PUM2 were also disrupted and if the chimeric PUM2-β-geo fusion protein might impact the phenotype ofPum2allele.Furthermore,recent studies reported that PUM2 together with PUM1 regulates neurogenesis[20],indicating a need to generate a conditionalPum2allele for simultaneous study of bothPum1andPum2in specific tissues of interest.We hence seek to generate a new loss-of-functionPum2allele,which removes all parts of PUM2 protein and could be used for conditionally removingPum2in any tissues for future characterization of PUM family protein function.

We succeeded in establishing a complete loss-offunction allele ofPum2and the usability for conditional knockout ofPum2in the presence of Cre recombinase.Whole body knockout ofPum2led to reduced bodyweight and reduced testis size,similar to the previousPum2gene trap allele.However,this new mutation also reduced male fertility and viability slightly,suggesting key roles of PUM2 in maintaining normal male fertility and viability.

Materials and methods

Animals

Mice were housed and maintained on a 12 h-light-12 h-dark cycle with free access to water and food under the specific pathogen-free conditions in Animal Core Facility of Nanjing Medical University,China.All animal experiments were approved by the Institutional Animal Care and Use Committee(IACUC)of Nanjing Medical University and Northwestern University.Pum2mutant mice used for analysis were on C57BL/6(Jackson Laboratory,#000664)background.

Generation of Pum2 mutant mice

ThePum2strain was created from ES cell clones Pum2(EPD0270)B04,Pum2 F03,obtained from KOMP repository and generated by Welcome Trust Sanger Institute[21].The cassette is composed of an FRT site followed by lacZ sequence and a loxP site.This first loxP site is followed by neomycin under the control of the human beta-actin promoter,SV40 polyA,a second FRT site and a second loxP site.A third loxP site is inserted at the position of 207 bp downstream of the targeted exons 6 and 7.The target region,exon6 and exon7,was flanked by two loxP sites and the selection cassette was flanked byFRTsites(Fig.1A).Blastocyst injection and derivation of chimera were done at Northwestern University Transgenic Core facility.We first injected the ES cell clone into C57BL/6 blastocyst to get a chimeric mouse and backcrossed the chimeric mice to establish a colony of mice that transmitted the targeting constructPum2nfgene in the germ line.Then,thePum2nf/+mice were mated with CAG- flip(C57BL/6)(to delete the selection cassette)and Ella-Cre(C57BL/6)(to delete exons 6 and 7)to generate mouse with thePum2FandPum2–alleles,respectively.All offspring were genotyped by PCR using primers shown in Fig.1A.

Genotyping of Pum2 mutant mice by genomic PCR

Genomic DNA was extracted from tail tip for genotyping.Pum2-F1,Pum2-R1 and Pum2-FrtR were used for detecting the wild-type allele(Pum2+,175bp)and the Frt-Neo-Frt-loxP cassette-containing allele(Pum2nf,310bp).Genotyping primers for the wildtype allele(Pum2+,132bp),the loxP-Exon6-Exon7-LoxP(Pum2LoxPhereafter referred to asPum2F,271bp)and the deletion allele(Pum2–,210bp)are Pum2-F3,Pum2-R2,Pum2-KOR and Pum2-loxR.Target regions were amplified by GoTaq Green master mix(Promega).All primers for genotyping are presented in Table 1.

RNA was extracted by using TRIzol Reagent(Invitrogen)according to the manufacturer’s instruc- tions and measured by Nanodrop 2000(Thermo).Reverse transcription(RT)was carried out following standard procedures using random primers and Prime-Script RT Master Mix(Takara)(Table 2).

Table 1 PCR primers for genotyping

Fig.1 Generation of Pum2 knockout mice.A:Strategy for generation of Pum2 knockout allele(Pum2–).Schematic representation of genomic alleles of wild type Pum2 locus,targeting allele or nf(LacZ-Neo-LoxP)allele, floxed allele and knockout allele.Primers used for genotyping are indicated with arrows.Flp for CAG- flipase mice used in the cross,Cre indicates cross with Ella-CRE mice.B:Genotyping results of Pum2nf,Pum2Fand Pum2– allele with different combinations of primers.C:Detection of different segments of Pum2 transcripts from Pum2–/–testes by RT-PCR reveals absence of exons 6 and 7 in all the residual transcripts,confirming a complete loss of wildtype Pum2 transcript.

Testes were collected from adultPum2+/+,Pum2–/–,Pum2nf/nf,andPum2XE772/XE772mice.All samples were lysed in RIPA buffer containing protease inhibitor(5 mg/mL final concentration).After incubation for 15 min on ice,the tissue lysates were centrifuged at 20,000gfor 30 min at 4°C and then the supernatant was carefully collected.The lysates were subjected to SDSPAGE followed by immunoblotting.Proteins were separated by 12%SDS-PAGE gels and transferred to PVDF membranes(Bio-Rad).The membrane was blocked with 10%(w/v)skim milk powder in TBS(100 mmol/L Tris-HCl,pH7.3)containing 0.1%(v/v)Tween 20(TBST)for 1 h and then incubated with anti-PUM2 antibody(1:500,Bethyl laboratory)or anti-βactin antibody(1:2,000,Sigma)at 4°C overnight.After three washes(5 min each)with 1×TBST,the membranes were incubated with a secondary antibody conjugated with horseradish peroxidase for 1 h at room temperature,followed by washing.Protein bands were detected by using ECL Western blotting reagents(Thermo)for exposure.

Bodyweight and fertility measurement

Bodyweight was measured weekly from newborn to adult stage for the three genotypes(Pum2+/+,Pum2+/–,Pum2–/–).Eight-week-old males or six-week-old females were housed singly with wildtype ICR females or ICR males of proven fertility for at least 6 months.A total of 6Pum2–/–,4Pum2+/–and 3Pum2+/+mice were tested in both males and females.The litter sizes were recorded continuously.

Table 2 RT-PCR primers for Pum2 mRNA

Histology analyses

Tissues were fixed for 24 h in Hartman’s fixative(Sigma).Paraffin embedded sections(5 μm)were stained with hematoxylin and eosin(H&E)and images were obtained with microscope(Zeiss,Axioscope)and processed using the AxioVision LE software.

Sperm count and sperm motility assay

The number of sperm was countedviaa hemocytometer as follows:one caudal epididymis was collected and dissected into numerous smaller pieces in 1 mL prewarmed(37°C)1 × PBS and followed by incubation at 37°C for 30 min to allow sperm release.Sperm motility was analyzed by using the computer-assisted sperm analysis(CASA,HamiltonThorne,TOX IVOS)system after 5 min incubation at 37°C.

Results

Generation of a conditional Pum2 allele

Pum2ES clone containing the targetPum2allele was injected in C57BL/6 blastocyst,and the resulting chimera was crossed with C57BL/6 mice to generatePum2nfallele(LacZ-Neo-LoxP allele),a knockout allele containing neomycin cassette and loxP insertions in thePum2locus(Fig.1A).Pum2nf/+mice were crossed to CAG- flip mice to generatePum2Fallele(Flox allele)at the same time.Pum2+/–mouse was generated after crossing Ella-CRE withPum2F/Fmouse,confirming the usability of floxedPum2allele for conditional knockoutviaCre recombinase(Fig.1A).Pum2–/–,Pum2nf/nf,Pum2f/fand their heterozygotes could be distinguishedviaPCR genotyping(Fig.1B).

To confirm the deletion of exon6 and exon7 inPum2homozygotes,we extracted RNA and protein from adultPum2+/+andPum2–/–testes.No transcripts containing exon 6 or 7 were present in the mutant tissues as RTPCR using primers from exon 6 or 7 failed to detect any product(Fig.1C).Despite deletion of exons 6 and 7,Pum2–/–mice still produced a shorterPum2transcript.However this mutant transcript is predicted to only produce a greatly truncated PUM2(191 aa;MNHDFQALALESRGMGELLPTKKFWEPD DSTKDGQKGIFLGDDEWRETAWGTSHHSMSQP IMVQRRSGQSFHGNSEVNAILSPRSESGGLGVS MVEYVLSSSPADKLDSRFRKGTFGTRDAETDG PEKGDQKGKASPFEEDQNRDLKQDDEDSK INGRGLPNGMDADCKDFNWCILSRPCSSCICAKS IYY*)without most part of PUM2 including PUM-HD,due to a reading frame shift and a premature stop codon after exon 5.Thus,we generated a novelPum2allele lacking wildtypePum2transcripts.

We next determined if PUM2 was completely knocked out.We examined PUM2 expression in the testes of wildtype and three Pum2 mutants.ThePum2nfallele was the initial knockout allele produced from chimeric founder mice.It is similar to the gene trap allele,disrupting the expression ofPum2but without signs of fusion protein.Indeed,PUM2 protein was completely absent inPum2-/-andPum2nf/nftestes whereas the PUM2-β-geo fusion protein was abundant inPum2XE772/XE772(Fig.2A).Further examination of PUM2 in other tissues(brain,thymus,liver,spleen,lung,kidney and ovary)validated complete loss of PUM2 in all the tissues examined(Supplementary Fig S1,available online).Given our antibodies recognize only theN-terminus of PUM2(100 to 150 amino acids of PUM2),failure to detect any PUM2 bands of the predicted truncated protein on Western bots indicated that no truncated or residual PUM2 protein was produced from the newPum2mutant tissues.We hence generated a complete loss-of-functionPum2allele.

Pum2 is required for normal growth in mice

Previous study reports thatPum2XE772/XE772mice show reduced bodyweight[18].We found that the newPum2–/–homozygous mice were similar to wildtype in overall morphology(data not shown)and also showed reduced bodyweight from 1 week to 12 weeks of age in males(Fig.2B).These data suggest that loss ofPum2impacted male bodyweight during postnatal mouse growth.The difference in sexes could be due to sample size and can be further examined with a larger number of animals.Indeed,for mice 20-week old or older,both male and femalePum2–/–mice showed small yet significant reduction of bodyweight than their wild type counterparts,confirming bodyweight reduction was not restricted to males.

Fig.2 Bodyweight reduction in Pum2 homozygotes.A:PUM2 is completely absent in the testes from Pum2–/– and Pum2nf/nfmutant,but a chimeric protein bigger than PUM2 is detectable in gene trapPum2XE772/XE772.Arrow indicates PUM2-β-geo fusion protein.B:Bodyweight was measured for Pum2+/+,Pum2+/– and Pum2– /– mice(n≥9 at each time point)from the first week to adult stage.Data are presented as the mean ±SD.n≥9 at each time point.Student's t-test:*P＜0.05,**P＜0.01,***P＜0.001.

We also noticed slightly reduced number ofPum2–/–(18.4%)among the pups produced from heterozygote interbreeding(Table 3).Although the number of offspring analyzed remained small,this significant deviation from Mendelian ratio may suggest a role of PUM2 in growth control.

Reduced fertility of Pum2–/– mutant males

AdultPum2–/–males exhibited smaller testes th an those ofPum2+/+,similar to tPum2XE772/XE772(Fig.3A).Testis weight of bothPum2+/–andPum2–/–was reduced(Fig.3B).To exclude the effect from the overall bodyweight reduction,we measured the ratio of testis weight over bodyweight and found that it was also significantly reduced inPum2+/–andPum2–/–mice than that of wildtype mice(Fig.3C).Histological examination of mutant testes showed a small increase in the number of degenerating tubules in the testis,suggesting a possible cause of testis weight reduction(Fig.3D).Examination of apoptotic cells in the testis did not reveal any significant difference between wildtype andPum2–/–(Supplementary Fig.2A,B and C,available online).Other than degenerated tubules,spermatogenesis stages were similar in both wildtype andPum2–/–based on histological staging of the testes(Fig.3E).

We further examined the fertility ofPum2–/–males,by mating 8-week-oldPum2+/+,Pum2+/–andPum2–/–males with wildtype females.Our results showed that loss ofPum2also reduced male fertility(Fig.4A).Pum2–/–males showed reduced sperm count(Fig.4B);however,sperm motility and sperm morphologyPum2homozygotes appeared normal(Fig.4C and Supplementary Fig.2D,available online).These data suggested thatPum2was required for normal male fertility.

Table 3 Pum2 genotype analysis from heterozygous intercrosses

Pum2 is dispensable for female fertility

Pum1,another member of Puf family,is an important translational regulator for mammalian female germ cell development[22].To analyze roles ofPum2in female mice,6-week-oldPum2mutant female mice were crossed to wildtype males.There was no significant difference in female fertility(Supplementary Fig.3A,available online)and ovary weight(Supplementary Fig.3B,available online).The number of developing follicles and corpus luteum in adultPum2–/–ovaries was comparable to age-matchedPum2+/+ovaries(Supplementary Fig.3C,available online).Further supporting this notion,superovulation experiments revealed that the number of MII oocytes retrieved from 3-month-oldPum2–/–females was similar to age-matchedPum2+/+females(Supplementary Fig.3D,available online),suggesting that Pum2 is dispensable for female fertility,consistent with the previous report[22].

Our work suggests thatPum2,a member of the conserved mammalian PUF family,is needed for attaining normal viability and male fertility,but is not essential for female fertility.

Discussion

Fig.3 Reduced testis weight in Pum2–/– male mice.A:Representative images of adult Pum2+/+,Pum2+/–,and Pum2–/– testis.B and C:Comparison of testis weight and testis/bodyweight ratio amongadult mice of three genotypes(Data were presented asmean±SEM.Pum2+/+,n=12;Pum2+/–,n=12;Pum2–/–,n=15.Student’s t-test:*P ＜ 0.05,**P ＜ 0.01,***P ＜ 0.001).D:H&E staining of adult Pum2+/+and Pum2–/–testis sections revealed increased degenerated tubules in Pum2–/–.Scale bar=50 μm.(Data are presented as mean±SEM.Pum2+/+,n=4;Pum2–/–,n=4.Student's t-test:*P ＜ 0.05,**P ＜ 0.01,***P ＜ 0.001).Arrowhead denotes degenerating tubules in testis sections.E:Spermatogenesis in most tubules appeared unaffected in the mutant testes.Three representative stagesshow that the mutant is not different from wildtype,suggesting normal spermatogenesis in most tubules.Scale bar=50 μm.

Fig.4 Pum2 null males exhibited subfertility.A:Fertility assay of Pum2+/+,Pum2+/– and Pum2–/–males revealed significantly reduced fertility(Data are presented as mean±SD.Pum2+/+,n=3;Pum2+/–,n=4;Pum2–/–,n=6.Student’s t-test:*P ＜ 0.05,**P ＜ 0.01,***P ＜ 0.001).B and C:Comparison of sperm counts(B)and sperm motility(C)in adult Pum2 males(Data were presented as mean±SEM.For(B),Pum2+/+,n=12;Pum2+/–,n=12;Pum2– /–,n=15;For(C),Pum2+/+,n=7;Pum2+/–,n=7;Pum2–/–,n=8.Student’s t-test:*P ＜ 0.05,**P ＜ 0.01,***P＜0.001).

RNA binding proteins have been found to play important roles in diverse physiologic processes as well as in human diseases.Recently,highly conserved RNA binding proteins-PUMILIO proteins were reported to be important for neurogenesis,germline development,genomic stability and hematopoietic stem cell proliferation[20,22–25].Physiologic function ofPumiliohas not been fully characterized yet.Our understanding of PUM2 has mainly come from a gene trap mutation,which disrupted the critical RNA binding ability of PUM2 but still produced a PUM2-β-geo fusion protein containing N-terminal PUM2[18].Thus,we constructed anotherPum2mutant allele-Pum2–/–,producing no PUM2.Complete absence of wildtypePum2transcripts and absence of any PUM2 residual protein indicated that we have generated aPum2null allele.

We confirmed the previous reports thatPum2is not essential for female fertility but is required for attaining normal bodyweight and testis weight[18–19,22].The small effect thatPum2mutation has on viability and male fertility revealed potential roles of PUM2 in growth and spermatogenesis.We failed to detect any increase in apoptosis inPum2mutant testes,suggesting the reduction of testis weight may not result from apoptosis.Further investigation may be needed to pinpoint the contribution of other causes such as reduced proliferation toward reduced male fertility.PUM1 was reported to play an important role in safeguarding spermatogenesis[26]andPum1mutants are reduced in bodyweight[26–27].We hence hypothesize thatPum1andPum2may both regulate viability and male fertility together.Removal of bothPum1andPum2altogether using conditional alleles in particular tissues such as the testis could help uncover fundamental roles of the conserved PUF family proteins.

Overall,Pum2–/–mutant mice is similar toPum2XE772/XE772in that they both exhibit reduced bodyweight,testis weight and sperm count[18–19],validating that the gene trapPum2XE772allele is a strong loss-of-function allele.However,Pum2–/–males displayed male subfertility and underrepresentation of homozygotes among the progeny.Such difference could result from the allelic difference due to the presence of part of PUM2 in the chimeric protein of the gene trap mutant;alternatively,it could result from a difference in mouse strain background asPum2–/–is in C57BL/6 background,but the gene trap allele used for fertility analysis was in mixed B6 and 129 background[18].Phenotypic comparison of transheterozygotes and individual homozygotes could distinguish the two possibilities.Regardless,generation of this conditional nullPum2allele established the functional requirement of mousePum2gene and provided a key model for future characterization of the role of the PUF family proteins in diverse biological processes.

Acknowledgments

We would like to thank Dr.Takeshi Kurita for discussion and assistance throughout this project.We also thank Welcome Trust Sanger Institute and Northwestern Targeting and Transgenic Core facility for assistance in the generation of mutant mice.This work was supported by National Basic Research Program of China(973 program,2013CB945201 and 2015CB-943002);National Science Foundation of China(81270737);Natural Science Foundation of Jiangsu Province(BK2012838);Provincial Innovation and Entrepreneurship Grant as well as NIH grant U01 HD045871.Funding for open access charge:Provincial Shuangchuang Program.