shi et al, diaphragmatic respiratory failure in nf90(-/-) mice nf90 …peterkao/kao...

M4:11034R2 Shi et al, Diaphragmatic respiratory failure in NF90(-/-) mice

1

NF90 regulates cell cycle exit and terminal myogenic differentiationby direct binding to the 3’-untranslated region of MyoD and

p21WAF1/CIP1 mRNAsLingfang Shi1, Guohua Zhao1, Daoming Qiu1, Wayne R. Godfrey2, Hannes Vogel3, Thomas A.

Rando4, Hong Hu1and Peter N. Kao1

From 1The Division of Pulmonary and Critical Care Medicine, Stanford University Medical Center,Stanford, CA 94305-5236; 2Cellular Immunology, Dendreon Corp., 3005 First Ave., Seattle, WA

98121; the 3Department of Pathology, 300 Pasteur Drive, Stanford, CA 94305; and the 4Departmentof Neurology and Neurological Sciences, Stanford University, and the Neurology Service, Veterans

Affairs Palo Alto Health Care Systems, Palo Alto, CARunning title: Diaphragmatic respiratory failure in NF90(-/-) mice

Address correspondence to: Peter N. Kao, M.D., Ph.D., Pulmonary and Critical Care Medicine, StanfordUniversity Medical Center, Stanford, CA 94305-5236. Tel (650) 725-0570, FAX (650) 725-5489. E-mail:[email protected]

NF90 and splice variant NF110/ILF3/NFARare double-stranded RNA-binding proteinsthat regulate gene expression. Mice withtargeted disruption of NF90 were engineered:NF90(-/-) mice were born small and weak andsuccumbed to perinatal death within 12 h dueto neuromuscular respiratory failure. Lunginflation and morphology were normal inNF90(-/-) mice. The diaphragm and otherskeletal muscles in NF90(-/-) micedemonstrated disorganized arrangement andpaucity of myofibers, evidence of myocytedegeneration and increased apoptosis. Theexpression of myogenic regulators, MyoD,myogenin and p21WAF1/CIP1, was severelydecreased in NF90(-/-) mice. These myogenictranscription factors and cell cycle inhibitorare regulated in part throughposttranscriptional mRNA stabilization.Northwestern blotting revealed that NF90 isthe principal and specific p21WAF1/CIP1 andMyoD 3’-untranslated region RNA-bindingprotein in developing skeletal muscles. NF90regulates transcription factors and a cell-cycleinhibitor essential for skeletal muscledifferentiation and for survival.

Keywords: Transcription, NF45, RNA stability,skeletal muscle

Introduction

Proteins that contain double-stranded RNAbinding motifs (dsRBM) contribute to generegulation at multiple levels. The dsRBMconsists of 65-75 amino acids that adopt anα−β−β−β−α conformation that confers theability to bind structured RNAs. The dsRBM isevolutionarily conserved from E.coli (RNAse III)through Drosophila (Staufen), to humans (NF90,PKR, ADAR). Proteins with dsRBMs contributeto RNA metabolism at multiple levels includingtranscription (NF90 and RNA helicase A), RNAsplicing (NF90), RNA editing (ADAR), RNAexport (NF90), subcellular RNA localization(NF90, Staufen) and translation initiation (NF90,PKR) (reviewed in (1-3)). Posttranscriptionalgene silencing through RNA interferenceinvolves enzymatic cleavage of RNA by thedsRBM protein, DICER (4).

NF90 and its splice variant, NF110, are proteinswith 2 dsRBMs, which are localizedpredominantly in the nucleus of diverse cells andtissues (5). Alternative names conferred on theNF90 family include human mitotic phasephosphoprotein 4 (MPP4) (6), Xenopus 4F.1 and4F.2 (7), murine interleukin enhancer bindingfactor (ILF)-3 (8), translation control protein(TCP)-80 (9), double-stranded RNA-bindingprotein (DRBP)-76 (10), and nuclear factor

JBC Papers in Press. Published on March 3, 2005 as Manuscript M411034200

Copyright 2005 by The American Society for Biochemistry and Molecular Biology, Inc.


2

associated with RNA (NFAR)-1 and 2 (11). Weoriginally purified NF90 and NF45 in a complexwith lupus autoantigens Ku70 and Ku80 from thenucleus of activated Jurkat T-cells (12). TheNF90/NF45/Ku complex bound specifically tothe antigen receptor response element/nuclearfactor of activated T-cells (NF-AT) binding sitein the human IL-2 promoter (5,12,13). NF90 andNF45 stabilize the association between thecatalytic subunit of DNA-dependent proteinkinase, DNA-PKcs and DNA-targeting subunits,Ku70 and Ku80 (14). NF90 and NF110 activateor repress transcription depending on thepromoter context (11,15). NF90 and NF45specifically transactivate the IL-2 promoter inactivated T-cells (16)(Zhao and Kao, unpublishedobservations). The arginine-glycine rich C-terminal domain of NF110/ILF3 is methylated invitro by protein arginine methyltransferase 1(17), and in vivo methylation may modulatetranscriptional regulation and chromatinreorganization (18).

The Xenopus homolog of NF90 and NF110 istermed CCAAT-box transcription factor (CBTF)and regulates the transcriptional activation of thehematopoietic regulatory factor, GATA-2 (19).In early development, CBTF is retained in thecytoplasm through interactions with maternalRNA. At the midblastula transition, degradationof maternal RNA releases CBTF for translocationinto the nucleus, followed 4h later bydevelopment of CCATT DNA-binding activityand transcriptional activation of GATA-2 (20).The dsRBMs of CBTF are capable of bindingboth DNA and RNA (21), and intact dsRBMs arerequired for transcriptional control (15).

In addition to roles in transcriptional regulation,NF90 contributes to RNA splicing as acomponent of the spliceosome (11,22). NF90specifically binds to AU-rich elements in the 3’UTR of IL-2 mRNA, stabilizing IL-2 mRNAagainst rapid degradation, and contributing to theupregulation of IL-2 gene expression in activatedT-cells (23). NF90 mediates nuclear export of IL-2 mRNA to the cytoplasm through the interactionof NF90’s nuclear export signal with the nuclear

expor t karyopherin, exportin-5 (24).Redistribution of NF90/MPP4 from the nucleusto the cytoplasm during mitosis is associated withincreased phosphorylation (6,13).

In the cytoplasm, NF90 is associated both withtranslationally quiescent ribonucleoprotein(RNP) complexes and with ribosomes (25). Likemany other mRNP components, NF90 is anautoantigen in human and experimental systemiclupus erythematosis (26). DsRBM proteins suchas NF90/ILF3 are highly expressed in the testis,where they contribute to gene regulation byrepressing the translation of nascent mRNAs (8).The mRNA encoding acid β-glucosidase isspecifically translationally repressed byNF90/TCP80 (9). A search for other livermRNAs specifically bound and translationallyrepressed by NF90/TCP80 identified aldolase B,complement protein 8 and fibronectin receptor β1(9). The prototypic dsRBM protein, staufen,contributes to localized translation of specificmRNAs in mammalian neurons (27). Tau is amicrotubule-associated protein that ispreferentially localized in axons, and abnormalaccumulations of tau form neurofibrillary tanglesthat contribute to the pathology of Alzheimer’sdementia (28). The 3’ UTR of tau mRNAcontains AU-rich elements that contribute toaxonal targeting; the axonal targeting element isspecifically bound by NF90 and NF110/ILF3(29). The interactions between cytoplasmicNF90/TCP80, translation elongation initiationfactors α, β and γ (14) and PKR (10,11,25,30) arelikely to regulate translation of specifically-bound mRNAs.

NF90 has been implicated in host antiviralresponses (25,30) (31,32). Virally encodedstructured RNAs interact specifically with NF90dsRBMs, including the HIV TAR loop, theadenovirus VA2 RNA (30), hepatitis B RNA (33)and the 5’ and 3’ UTRs of the bovine virusdiarrhea RNA virus (32). Specific binding ofNF90 and NF45 to both the 5’ and 3’ UTRs ofBVDV RNA virus was proposed to regulate thebalance between viral replication and translation(32).


3

To further characterize the role of NF90 inregulation of gene expression and development,we have generated mice with targeted disruptionof NF90. Our observation of perinatal lethalityassociated with muscle weakness and respiratoryfailure in NF90(-/-) mice revealed an unexpectedcontribution of NF90 to myogenic differentiation.Muscle development is regulated through thecoordinated expression of myogenic regulatoryfactors that include MyoD, Myf5, myogenin andMrf4, and cyclin-dependent kinase inhibitors ofthe p21WAF1/CIP1 family (34,35). MyoD and Myf5direct primary myocyte differentiation, myogenindirects secondary differentiation, and p21WAF1/CIP1

expression promotes cell cycle exit and terminaldifferentiation. In differentiating myoblasts andregenerating adult muscle, upregulation ofMyoD, myogenin and p21WAF1/CIP1 expressionoccurs concurrently through transcriptionalmechanisms and posttranscriptional mRNAstabilization by specific RNA-binding proteins(35,36). Here, we demonstrate that MyoD,myogenin and p21WAF1/CIP1 are novel target genesessentially regulated by NF90. Our results lead usto propose that the phenotype of muscleweakness and increased apoptosis in NF90(-/-)mice is explained by defects in skeletal myocytedifferentiation arising from insufficientexpression of myogenic regulatory factors.

Experimental Procedures

Genomic cloning and targeting vectorgeneration: The murine genomic NF90/ILF3 wasisolated from a 129SvJ library using the humanNF90 cDNA probe (performed at IncyteGenomics). The nucleotide sequence and intron-exon structure of the 29085 bp gene wasdetermined and deposited in GenBank(AF506968).

A targeting vector was designed for constitutivedisruption of NF90 exons 2-4 (Figure 1A). The5’ targeting arm was a 6.75 kB BamH1restriction fragment digested from a genomicNF90 subclone, and the 3’ targeting arm was 2.8kB generated with PCR primers that introducedflanking EcoR1 restriction sites (sense primer b3:

5’-GGAGTGGGTCTTCCTGGCCAGTTGCG-3’ and anti-sense primer c3: 5’-GGCGAATTCTCTGCCAGCTATCAAGGA-3’). These targeting arms were cloned into pKOScrambler NKTV-1903 targeting vector, whichcontains a positive selection neomycin resistancegene and a herpes thymidine kinase negativeselection gene (Stratagene). Embryonic stemcells (R1) were transfected by electroporationand selected for resistance to G418 andganciclovir.

Genotype analysis by Southern hybridization andPCR: Genomic DNA was isolated from ES cellsor tail biopsy specimens, digested with HindIII(3’ analysis) or EcoR1 (5’ analysis), andseparated on a 0.7% agarose gel followed bytransfer to nitrocellulose membrane. Southernhybridization of the specific wild type andtargeted genomic fragments was performed usingrandomly-primed 32P-labeled probes outside thetargeting arms (Figure 1a), or an internal probethat hybridized to the neomycin gene, accordingto standard methods.

Genotyping was also performed by PCR acrossthe 3’ targeting arm, using LA TaqTM(TakaraBio) with locus pr imers , b3- ,5’-GGAGTGGGTCTTCCTGGCCAGTTGCGGTGG - 3 ’ & C 4 - * , 5 -GGTCCAGCGTCCCTGACTGGATAAGCTAAGAGGGC-3’, or target primers, Neo-1179,5’-TATCAGGACATAGCGTTGGCTACCCGTGATATT-3’ & C4-*. PCR amplification wasconducted according to the manufacturer’sdirections consisting of an initial incubation at94°C for 2 min, followed by 35 cycles at 98°Cfor 20 s, 68°C for 3 min.

Expression analysis by Northern and Westernhybridizations. Total RNA was extracted fromnewborn mice using Trizol reagent (Invitrogen),and 20 µg was fractionated in a 1% denaturingagarose gel then transferred to nitrocellulose.Northern hybridization was performed using arandomly primed 32P-labeled probe for NF90.


4

Extracted total proteins (20 µg) from newbornmice were fractionated by SDS-PAGE (8%separating gel) and transferred to nitrocellulosemembranes. Western immunoblotting wasperformed using rabbit polyclonal antisera(1:1000 dilution) generated against recombinantNF90 protein (5), followed by secondary goatanti-rabbit antibody (1:3000 dilution) anddetection using enhanced chemiluminescence(Amersham). Immunoblotting analysis of wildtype murine tissue extracts was performed usinga monoclonal antibody that recognizes NF90(anti-DRBP76, Transduction Laboratories612154), followed by a goat anti-mousesecondary antibody (Santa Cruz SC-2031).

Histology, immunohistochemistry and in situTUNEL analysis: Whole embryos or newbornmice with skin removed were fixed in 10%neutral buffered formalin followed by paraffin-embedding and tissue sectioning at 4-µmthickness. Staining with hematoxylin and eosinwas performed according to standard procedures.

For immunodetection of NF90, sections weredeparaffinized and rehydrated, then boiled for 20min in 10 mM citric acid in a microwave oven.Blocking of nonspecific mouse antigens wasperformed by incubation with 3% H2O2 for 5mins followed by incubation with the VectorM.O.M. immunodetection kit (Vectorlabs) for 1h. Anti-DRBP76 monoclonal antibody specificfor NF90/ILF3/DRBP76 (TransductionLaboratories) was incubated at 1:100 dilution for35 min at room temperature. The signal wasamplified by using VECTASTAIN Elite ABCReagent (Vector labs) and visualized by a HRPreaction using stable diaminobenzidine(Invitrogen).

In situ apoptosis was analyzed by terminaldeoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL; TACS 2 TdT,Trevigen).

Reverse Transcription Polymerase ChainReactions. RT-PCR analysis was performed todetect mRNA expression of myogenic regulatory

factors and cell cycle control protein p21WAF1/CIP1.Total RNA from skeletal muscles of newbornmice (2-5 µg) was used as template for firststrand cDNA synthesis using Superscript reversetranscriptase (Invitrogen), according to themanufacturer’s directions. PCR amplificationswere subsequently performed using 20% of thefirst-strand cDNA mixture, and specific primersfor p21WAF1/CIP1 (NM_007669) 511+ 5’-GGCAGACCAGCATGACAGATTTC (with onemismatched base from human p21WAF1/CIP1

u n d e r l i n e d ) a n d 8 8 3 - 5 ’ -CCCTCCCAACCCAAGCTCTG (372bp),m y o g e n i n ( B C 0 4 8 6 8 3 ) 5 ’ -GAGCGCGATCTCCGCTACAGAGG and 5’-CTGGCTTGTGGCAGCCCAGG (380bp), andM y o D ( N M _ 0 1 0 8 6 6 ) 5 ’ -AGGCTCTGCTGCGCGACC and 5’-TGCAGTCGATCTCTCAAAGCACC (489bp).Control amplification of murine β-actin mRNAwas performed using primers 5’-AGACGGGGTCACCCACACTGTGCCCATCTA a n d 5 ’ -CTAGAAGCACTTGCGGTGCACGATGGAAGGG (671bp). The PCR cycling conditions were24 cycles of denaturation at 95˚C for 30”,annealing at 55˚C for 30” and extension at 72˚C.

Northwestern Blotting. Partial cDNAs of murineMyoD, myogenin and p21WAF1/CIP1 were amplifiedby RT-PCR from skeletal muscle RNA. Nestedamplifications were performed with a forwardprimer containing a T7 RNA polymeraser e c o g n i t i o n s e q u e n c e , T 7 ,CCAAGCTTCTAATACGACTCACTATAGGGAGA. Primers to generate DNA templates forT7-directed in vitro transcription of the 3’ UTRswere: MyoD forward (T7-1241+), 5’ T7-GACACTCTTCCCAACTGTCC and MyoDr e v e r s e ( 1 7 6 2 - ) , 5 ’ -GCACTACACAGCATGCCT (36), p21 WAF1/CIP1

forward, T7-511+ and 883- (above). RadiolabledRNA probes were generated by in vitrotranscription using T7 RNA polymerase, in thepresence of 32P-labeled UTP (36), according tothe manufacturer’s directions (Promega).


5

Murine tissue extracts (testis, brain, skeletalmuscle, 80 µg each) or recombinant NF90 andNF45 proteins (5) (10µg) were fractionated by10% SDS-PAGE, and t rans fe r redelectrophoretically (25 mM Tris-HCl pH 8.3, 192mM glycine, 20% methanol) to nitrocellulosemembranes. The proteins were renatured byincubating the membranes overnight at roomtemperature in Northwestern buffer (10 mM Tris-HCl pH 7.5, 150 mM KCl, 5 mM MgCl2 0.1%gelatin, 0.1% Tween20, containing salmon spermDNA (50µg/ml). The membranes werehybridized with the radiolabeled RNA probes (2-4 x 107 cpm) for 2 h at room temperature in thepresence of yeast tRNA (50µg/ml) (29,37),washed twice for 5 min at room temperature andonce at 45˚C for 2 h, followed byautoradiography.

Results

Generation of mice with targeted disruption ofNF90 gene. The targeting strategy for NF90 wasto disrupt Exons 2-4, eliminating the ATGtranslation initiation codon and introducing atranslational frameshift that disrupts the openreading frame (Figure 1A). The targeting vectorwas introduced into R1 embryonic stem cells byelectroporation and transfected ES cells weresubjected to positive and negative selection.Twelve out of 110 G418/gancyclovir-resistant EScell clones showed appropriate gene targeting.The targeted allele shows a Southernhybridization signal at 9 kB, in addition to thewild type allele with a hybridization signal of 7.4kB (Figure 1B, lane 4). Targeted ES cell cloneswere further analyzed for correct integrationusing an internal probe for the neomycinresistance gene (data not shown). Threeindependently derived ES cell lines were injectedinto blastocyts (C57BL/6), and chimeric micewere derived. Two separate NF90-targeted EScell lines transmitted through the germline. Inmice examined through 18 months of age,NF90(+/-) heterozygote mice demonstrate noabnormal phenotypes.

Gross phenotype of NF90(-/-) mice. Followingheterozygous intercrossing, homozygote NF90(-/-) mice were present in utero at the Mendelianfrequency (Table, Figure 1C, lanes 4, 8). At thetime of delivery, all NF90(-/-) mice were smallerand weighed 35% less than their littermates(Table, Figure 1D). Approximately half of theNF90(-/-) mice died immediately after birth. Thesurviving NF90(-/-) mice appeared weak andshowed almost no spontaneous movements.When lifted by their tails, they emitted a veryweak cry compared to their littermates. Over thecourse of several hours, the NF90(-/-) micebecame increasingly cyanotic compared to theirlittermates, and oxygen saturation was noted tobe 65% in a cyanotic NF90(-/-) mouse, comparedto 85-98% in littermates (Figure 1D). Videoanalysis of breathing revealed that NF90(-/-)mice displayed more rapid and shallow tidalbreathing, with no sighs, compared to theirheterozygote and wild-type littermates. TheNF90(-/-) mice never suckled for milk, and wereignored by their mothers. In over 400 live births,no homozygote NF90(-/-) mice animals survivedbeyond 12 h after delivery. Thus, the NF90 geneis essential for perinatal survival.

Gene expression analysis in NF90(-/-) mice. Weanalyzed the expression of NF90 mRNA andprotein in Day 1 NF90 homozygote, heterozygoteand wild type littermates. The Northern analysisshows that wild type mice express a single NF90mRNA migrating at 3.4 kB (Figure 2A, lanes 2,4, 5), and heterozygote mice express twoisoforms of NF90 mRNA, migrating at 3.4 and3.1 kB (Figure 2A, lanes 1, 3). These mRNAsizes are consistent with transcription of bothwild type and targeted NF90 alleles. The NF90(-/-) mice demonstrate expression of a singleisoform of NF90 mRNA migrating at 3.1 kB(Figure 2A, lanes 6, 7), and this size is consistentwith transcription just of targeted alleles, whichlack exons 2-4. The prominent intensity of thetruncated 3.1 kB targeted NF90 mRNA inNF90(-/-) mice indicates that the endogenousNF90 promoter is functional following deletionof Exons 2-4. Furthermore, the increase inintensity of the 3.1 kB truncated form of NF90


6

mRNA compared to the 3.4 kB wild type form(Figure 2A, lanes 6, 7 vs. 2, 4, 5) suggests thatthe null phenotype (absent NF90 protein, Figure2B) is associated with increased transcription ofthe NF90 gene. Because the targeting strategyintroduced a translational frameshift downstreamof exon 4, there is no NF90 protein synthesizedin NF90(-/-) mice (see below and Figure 2B).Our prior studies demonstrated that transgenicexpression of histidine-tagged NF90 proteinpotently suppressed the expression ofendogenous NF90 protein (5). Here, the absenceof NF90 protein in NF90(-/-) mice is associatedwith increased mRNA expression of the 3.1 kBtargeted NF90 allele, through transcriptional orposttranscriptional upregulation.

We analyzed protein expression of NF90 inextracts prepared from newborn wild type,heterozygote and homozygote NF90 mice byWestern immunoblotting (Figure 2B). In NF90wild type and heterozygote mice we observed aprominent immunoreactive band at 90 kDa, and afainter band at 110 kDa, similar to the principalsize observed in human Jurkat T-cell nuclearextracts (5) (Figure 2B, lanes 1-5). The level ofNF90 protein expression in heterozygote NF90mice was slightly decreased from that in wildtype mice (Figure 2B, lanes 1 and 3 vs. 2, 4, 6).The NF90(-/-) mice showed complete absence ofthe 90 and 110 kDa bands (Figure 2B, lanes 6, 7vs. 1-5), demonstrating that the translationalframeshift in the knockout phenotype wasassociated with complete absence of NF90 andNF110 protein expression. Examination of thetissue-specific expression of NF90 and NF110proteins in adult wild type mice showed thegreatest expression in testis, brain and skeletalmuscles (Figure 2C, lanes 8, 2, 6), and moderateexpression in heart, spleen, lung, liver and kidney(Figure 2C, lanes 1, 3, 4, 5 and 7).

Histological analysis of NF90(-/-) mice. W eperformed careful histology on hematoxylin andeosin-stained sections of developing andnewborn wild type, heterozygote andhomozygote NF90 mice and did not discern anydramatic organ-specific developmental

abnormalities. Consistent with the overall sizereduction of the homozygote NF90-deficientmice, we observed a proportionate reduction inthe size of all the internal organs, compared towild type and heterozygote littermates.

We characterized the tissue-specific expressionand subcellular localization of NF90 protein byimmunohistochemistry with a monoclonalantibody that specifically recognizes NF90(Figure 3). Wild-type mice show widespreadexpression of NF90 protein, predominantly in thenucleus of cells, including cardiomyocytes(Figure 3A), lung epithelial and endothelial cells(Figure 3B), certain hepatocytes (Figure 3C), andmost epithelial cells lining the stomach (Figure3D). We observed the strongest expression ofNF90 in cerebral (Figure 3E) and cerebellar(Figure 3F) neurons, and in skeletal muscles ofthe back (Figure 3G) and diaphragm (Figure 3H)in wild type mice. Skeletal muscles (Figure 3G),certain neurons (Figure 3F) and epithelial cells(Figure 3B, D) of NF90(+/+) mice revealedmoderate NF90 immunostaining in thecytoplasm. For comparison, NF90immunostaining was completely absent inNF90(-/-) mice in cerebral (Figure 3I) andcerebellar (Figure 3J) neurons, and in skeletalmuscles of the back (Figure 3K) and diaphragm(Figure 3L). These results support the specificityof the monoclonal antibody in recognizing NF90,and confirm that NF90(-/-) mice express nodetectable NF90 protein.

As the cause of death of the homozygote NF90-deficient mice appeared to be respiratory failure,we examined the lung histology in detail. Wefound no evidence of lung developmentalabnormalities, insufficient lung inflation,pulmonary edema or hyaline membraneformation in the NF90(-/-) mice that mightexplain perinatal death due to respiratory failure.We detected no differences in the expression oflung surfactant genes SP-A, B and C, and a 50%decrease in SP-D expression by RT-PCR inNF90(-/-) mice compared to wild-type littermates(data not shown). Surfactant protein-D functions


7

principally in lung host defense, and mice lackingSP-D are viable after birth (38).

Blood, serum and lymphocytes subset analyses ofNF90(-/-) mice. Complete blood count analysesof the neonatal mice were difficult to performdue to the low amount of blood recoverable atsacrifice. However, there was a consistentsuggestion of anemia in the NF90(-/-) mice (datanot shown). White blood cell counts anddifferentials were similar between NF90(-/-)mice and wild-type littermates. There were nosignificant differences in the clinical chemistrymeasurements of hepatic or renal functionbetween homozygote, heterozygote and wild typemice NF90 mice (data not shown).

Flow cytometry analyses of thymic and spleniclymphocyte subsets from neonatal NF90(-/-) andheterozygous littermates revealed no differencesin the expression of CD4 and CD8 doublepositive and single positive thymocytes betweenNF90(+/-) and NF90(-/-) mice (data not shown).

NF90(-/-) mice show pathological maturation ofskeletal muscles. Searching for a potential causeof perinatal respiratory failure that would beindependent of the lung parenchyma, wecarefully examined the skeletal muscle fibers ofthe diaphragm and discovered that NF90(-/-)mice exhibited a relative paucity of muscle fibersand generally thinner diaphragm muscles thanheterozygote or wild type littermates (Figure 4Bvs. A, and Figure 3L vs. H). The abnormalfeatures present in the diaphragm were alsopresent in intercostal and back muscles ofhomozygote NF90(-/-) mice, including overallpaucity of muscle fibers, wider variation ofmuscle fiber sizes, and some evidence ofmyofibers with strongly eosinophilic andhomogeneous cytoplasm (Figure 4B vs. A).These abnormal fibers were likelyhypercontracted, possibly as a consequence ofdisrupted integrity of the sacrolemma inducingunstable ion fluxes. Nuclei within skeletalmyofibers of NF90(-/-) mice appeared larger withmore granular chromatin visible compared tonuclei from myofibers of heterozygote and wild

type littermates. We observed greater numbers ofmultinucleated giant fusing myocytes in theNF90(-/-) mice (Figure 4D vs. C), suggesting thatskeletal muscles in the NF90(-/-) mice aredevelopmentally delayed, or that ongoing muscledegeneration may be triggering efforts atregeneration.

TUNEL analysis of NF90-deficient mice. Thesmall size of the NF90(-/-) mice and their internalorgans may be a consequence of a combinationof generalized growth retardation, developmentaldelay and/or increased apoptosis duringdevelopment. To further understand the causes ofsmall size and perinatal death in NF90-deficientmice, we performed in situ TUNEL staining(terminal deoxynucleotidyl transferase-mediateddUTP nick end labeling). TUNEL-positive cellswere more prominent in E19 and newbornNF90(-/-) mice compared to wild-type orheterozygote littermates, throughout multipleorgans. The greatest numbers of TUNEL-positivecells were present in cortical neurons and skeletalmuscles of the diaphragm in NF90(-/-) animals(Figure 5B, D vs. 5A, C). The increased presenceof TUNEL-positive cells in NF90(-/-) animalscorrelated with the observed reduction in size andweight.

Reduced expression of myogenic regulatoryfactors, MyoD, myogenin and p21WAF1/CIP inNF90(-/-) mice. We analyzed neonatal skeletalmuscle expression of MyoD, myogenin andp21WAF1/CIP by semi-quantitative RT-PCR (Figure6A). Expressio of MyoD mRNA expression wasgreatly inhibited, and myogenin and p21WAF1/CIP

expression was reduced approximately 50% inNF90(-/-) compared to wild-type andheterozygote littermates (Figure 6A, lanes 4-6 vs.2, 3 and 1). We discerned slight decreases in themyogenin and p21WAF1/CIP RT-PCR products inthe heterozyogote compared to the wild-typelittermates (Figure 6A, lane 1 vs. 2, 3). Weconfirmed the reduction in myogenic regulatoryfactors in NF90(-/-) mice by Western analysis ofskeletal muscle extracts (Figure 6B). Proteinexpression of MyoD and p21WAF1/CIP1 is markedlyreduced in NF90(-/-) mice (Figure 6B, lanes 4-6


8

vs. 2, 3), which corresponds well with the resultsof mRNA expression (Figure 6A).

p21WAF1/CIP1 and MyoD 3’ UTR RNAs interactspecifically with NF90. The 3’-UTRs of themurine and human p21WAF1/CIP1 mRNA sequencesshow significant conservation in the proximal350 nt region that contains multiple AU-richbinding elements (39). This proximal 3’ UTR ispredicted to adopt secondary structures thatincludes regions of double-stranded RNA andhairpin loops (Figure 7A) (40). The MyoD 3’UTR is also predicted to adopt extensivesecondary structure (Figure 7B). These RNAsecondary structures are binding targets forspecific regulatory RNA-binding proteins such asHuR (36) and NF90 (this work). We usedNorthwestern blotting to characterize specificinteractions between a radioactively labeledmurine p21WAF1/CIP1 3’-UTR probe and testis,brain and muscle extracts from adult wild typemice, as well as brain and muscle extracts fromneonatal NF90(-/-) mice (Figure 7C). Thep21WAF1/CIP1 3’-UTR RNA probe specificallylabels a 90 kDa protein that is present in wildtype tissues (Figure 7C, lanes 1, 2, 3), iscompletely absent in extracts of NF90(-/-) mice(Figure 7C, lanes 4, 5), and that comigrates withlabeled, recombinant NF90 protein (Figure 7C,lane 6). The MyoD 3’ UTR probe alsospecifically labels a 90 kDa protein in wild-typetissues (Figure 7D, lanes 1, 3) and recombinantNF90 protein (Figure 7D, lane 6), and shows nolabeling in extracts of NF90(-/-) mice (Figure 7D,lanes 4, 5). Both 3’ UTR probes showed nolabeling of recombinant NF45 protein (Figure7C, D, lanes 7). Taken together, these resultsidentify NF90 as the principal specificp21WAF1/CIP1 and MyoD 3’-UTR RNA bindingprotein present in skeletal muscles.

Discussion

Our study is the first to establish that NF90, adsRNA-binding protein, is essential for perinatalsurvival in mice. The gross phenotype of NF90(-

/-) mice is small size, weak cry, lack ofspontaneous movements and death due to rapidlyprogressive respiratory failure. The relativecontributions of diaphragmatic weakness,neurologic abnormalities or metabolicderangements to perinatal death are difficult todistinguish. Nevertheless, the immediate cause ofdeath less than 12 h after delivery appears to bediaphragm muscle weakness and respiratoryfailure. We observed morphological correlationswith respiratory failure of disorganizeddiaphragm architecture and muscle fibers withinclusions suggestive of apoptosis in the NF90(-/-) mice. A similar phenotype of perinatal deathdue to respiratory failure and decreased mass ofdiaphragm and intercostals muscles was observedin mice deficient in the activin- and transforminggrowth factor β-modulating protein, follistatin(41).

Targeted disruptions of other dsRBM-bindingproteins in mice have been associated withvarying phenotypes. RNA helicase is a nucleartranscriptional regulator that is the mammalianhomolog of the Drosophila X-chromosomedosage-compensating factor, maleless. Targeteddisruption of RNA helicase resulted in embryoniclethality before gastrulation, and was associatedwith enhanced TUNEL-positive staining (42). Zfris a nuclear zinc finger RNA-binding proteinwith homology to NF90, expressed at highestlevels in testis, ovary and brain (43). Targeteddisruption of Zfr resulted in early embryoniclethality at day 8 - 9 as a consequence ofdevelopmental delay and enhanced apoptosis(44). Spnr is a cytoplasmic dsRNA-bindingprotein localized to microtubules in the testis,brain and ovary that may contribute totranslational activation (45). A hypomorphicallele of Spnr produced by a gene trap strategyresulted in small size, increased lethality andneurological and reproductive defects (46). Twoindependent knockouts of the cytoplasmicdsRNA-activated protein kinase, PKR, resultedin no obvious phenotype, and this may reflect thefinding that each represents only a partialknockout (47). Near complete targeted disruptionof the nuclear ADAR2 RNA-editing enzyme


9

resulted in perinatal lethality by P12, due toincreased seizure activity (48). Targeteddisruption of nuclear ADAR1 resulted inembryonic lethality before E12, associated withwidespread apoptosis (49). Together, theseresults suggest that nuclear dsRNA-bindingproteins serve essential roles such astranscriptional regulation and RNA splicing.Cytoplasmic dsRNA binding proteins contributeto mRNA stabilization, subcellular localizationand regulation of translation.

Essential and redundant contributions ofmyogenic regulatory factors (MRFs), MyoD,Myf5 and myogenin, and p21WAF1/CIP1 familymembers to embryonic muscle development havebeen elucidated through gene targeting studies.Either MyoD or Myf5 can direct myogenicspecification (50), whereas myogenin (51), andeither p21WAF1/CIP1 or p57KIP2 are essential fornormal myogenic differentiation (52). Anessential role for MyoD in the development ofthe skeletal muscle of the diaphragm wasrevealed by the failure to obtain viable miceduring backcrossing with mdx mice (lackingdystrophin) (53). Reductions in myogenin geneexpression in mice resulted in embryonic orneonatal lethality characterized by severereductions in skeletal muscles (51,54,55). Recentevidence suggests that Mrf4, an MRF previouslyconsidered to be involved in myogenicdifferentiation, may in fact be an early myogenicspecification factor, serving a redundant functionto that of MyoD and Myf5 (56). Upregulation ofp21WAF1/CIP1 promotes cell cycle exit and terminaldifferentiation in skeletal muscles, characterizedby the fusion of myoblasts into myotubes andincreased synthesis of muscle contractile proteins(57). P21WAF1/CIP1(-/-) mice develop normallyhowever (58), and this was proposed to be due toredundant cell cycle regulation by p57KIP2 (52).Mice that lacked both p21WAF1/CIP1 and p57KIP2

were not viable due to severely compromisedmuscle development as manifested by diminisheddiaphragm muscles. The absence of both cellcycle inhibitors resulted in overproliferation andincreased apoptosis leading to a marked decreasein mature skeletal muscle fibers in vivo. The

absence of p21WAF1/CIP1 was associated withenhanced programmed cell death indifferentiating myoblasts in vitro (59).

Mechanisms of posttranscriptional mRNAstabilization of MyoD, myogenin and p21WAF1/CIP1

have been studied in differentiating C2C12myoblasts (36,60). Ultraviolet crosslinking ofradiolabeled 3’UTR RNA probes to cytosolicextracts of C2C12 cells labeled proteins of 37, 86and 116 kDa (36). The 37 kDa protein wasidentified as the embryonic lethal abnormalvision (elav) protein, HuR, that binds AU-richelements (AREs) and mediates RNA-stabilizationand nuclear-to-cytoplasmic export. Knockdownof HuR protein expression by RNA interferencein C2C12 cells inhibited myogenesis in vitro(60). However, in MDA-468 human breastcancer cells, although HuR bound p21WAF1/CIP

RNA, it was not a major modulator p21WAF1/CIP

expression or growth inhibition (39). HuRprotein expression in young adult mice wasprominent in intestine, thymus and spleen,followed by the liver, and was almostundetectable in skeletal muscle, brain, kidney,lung and heart (61). This low to absentexpression of HuR in skeletal muscles raises aquestion about its contributions toposttranscriptional stabilization of MyoD,myogenin and p21WAF1/CIP1 during myogenesis invivo.

Here, we demonstrate that NF90 is the principaland specific RNA-binding protein in skeletalmuscles for MyoD and p21WAF1/CIP1. Gene-targeted NF90(-/-) mice, completely lackingNF90 protein, demonstrate severely reducedexpression of MyoD, myogenin and p21WAF1/CIP1.We propose that NF90(-/-) mice are deficient inposttranscriptional stabilization of MyoD,myogenin and p21WAF1/CIP1 mRNAs. Thephenotype of impaired muscle formationassociated with increased apoptotic myonuclei islikely the direct result of this coordinateddownregulation of multiple myogenic regulators,including those involved both in myogenicspecification, cell cycle withdrawal, anddifferentiation. Because MyoD(-/-)(50) and


10

p21WAF1/CIP1(-/-)(58) mice are each viable throughcompensatory expression of myogenic regulators,we reasoned that our phenotype of perinataldeath due to diaphragmatic respiratory failuremust arise from effects on more than one of thesemyogenic regulators. Mice deficient in MyoD(62), or in cell cycle inhibitors p21WAF1/CIP1 andp57KIP2 (52), exhibited overproliferation andincreased programmed cell death. This isconsistent with our observations of increasedapoptosis in NF90(-/-) mice, especiallyprominent in skeletal muscles and brain. Theoverall growth stunting of NF90(-/-) mice (40%lower weight than littermates) is likely aconsequence of abnormal myogenesis and

organogenesis related to defects in exiting thecell cycle.

Future studies will be directed at identifyingother structured RNA targets involved indifferentiation that are critically regulatedthrough specific binding to NF90.

AcknowledgementsWe thank Tushar Desai and Mark Krasnow forhelpful discussions. Supported by NIH R01-AI39624 and R01-HL62588 and the Donald E.and Delia B. Baxter Foundation.

References

1. Fierro-Monti, I., and Mathews, M. B. (2000) Trends Biochem Sci 25, 241-2462. Saunders, L. R., and Barber, G. N. (2003) Faseb J 17, 961-9833. Tian, B., Bevilacqua, P. C., Diegelman-Parente, A., and Mathews, M. B. (2004) Nat Rev Mol Cell

Biol 5, 1013-10234. Zamore, P. D., Tuschl, T., Sharp, P. A., and Bartel, D. P. (2000) Cell 101, 25-335. Kao, P. N., Chen, L., Brock, G., Ng, J., Kenny, J., Smith, A. J., and Corthesy, B. (1994) Journal of

Biological Chemistry 269, 20691-206996. Matsumoto-Taniura, N., Pirollet, F., Monroe, R., Gerace, L., and Westendorf, J. M. (1996) Mol

Biol Cell 7, 1455-14697. Bass, B. L., Hurst, S. R., and Singer, J. D. (1994) Curr Biol 4, 301-3148. Buaas, F. W., Lee, K., Edelhoff, S., Disteche, C., and Braun, R. E. (1999) Mamm Genome 10, 451-

4569. Xu, Y. H., and Grabowski, G. A. (1999) Mol Genet Metab 68, 441-45410. Patel, R. C., Vestal, D. J., Xu, Z., Bandyopadhyay, S., Guo, W., Erme, S. M., Williams, B. R., and

Sen, G. C. (1999) J Biol Chem 274, 20432-2043711. Saunders, L. R., Perkins, D. J., Balachandran, S., Michaels, R., Ford, R., Mayeda, A., and Barber,

G. N. (2001) J Biol Chem 276, 32300-3231212. Corthesy, B., and Kao, P. N. (1994) Journal of Biological Chemistry 269, 20682-2069013. Aoki, Y., Zhao, G., Qiu, D., Shi, L., and Kao, P. N. (1998) Am J Physiol 275, L1164-L117214. Ting, N. S. Y., Kao, P. N., Chan, D. W., Lintott, L. G., and Lees-Miller, S. P. (1998) Journal of

Biological Chemistry 273, 2136-214515. Reichman, T. W., Muniz, L. C., and Mathews, M. B. (2002) Mol Cell Biol 22, 343-35616. Zhao, G., Shi, L., Qiu, D., Hu, H., and Kao, P. N. (2005) Exp Cell Res in press17. Tang, J., Kao, P. N., and Herschman, H. R. (2000) J Biol Chem 275, 19866-1987618. Fischle, W., Wang, Y., and Allis, C. D. (2003) Curr Opin Cell Biol 15, 172-18319. Orford, R. L., Robinson, C., Haydon, J. M., Patient, R. K., and Guille, M. J. (1998) Mol Cell Biol

18, 5557-556620. Brzostowski, J., Robinson, C., Orford, R., Elgar, S., Scarlett, G., Peterkin, T., Malartre, M., Kneale,

G., Wormington, M., and Guille, M. (2000) Embo J 19, 3683-3693


11

21. Scarlett, G. P., Elgar, S. J., Cary, P. D., Noble, A. M., Orford, R. L., Kneale, G. G., and Guille, M.J. (2004) J Biol Chem 279, 52477-52455

22. Zhou, Z., Licklider, L. J., Gygi, S. P., and Reed, R. (2002) Nature 419, 182-18523. Shim, J., Lim, H., J, R. Y., and Karin, M. (2002) Mol Cell 10, 1331-134424. Brownawell, A. M., and Macara, I. G. (2002) J Cell Biol 156, 53-6425. Langland, J. O., Kao, P. N., and Jacobs, B. L. (1999) Biochemistry 38, 6361-636826. Satoh, M., Shaheen, V. M., Kao, P. N., Okano, T., Shaw, M., Yoshida, H., Richards, H. B., and

Reeves, W. H. (1999) J Biol Chem 274, 34598-3460427. St Johnston, D. (2001) Embo J 20, 6169-617928. Avila, J., Lucas, J. J., Perez, M., and Hernandez, F. (2004) Physiol Rev 84, 361-38429. Larcher, J. C., Gasmi, L., Viranaicken, W., Edde, B., Bernard, R., Ginzburg, I., and Denoulet, P.

(2004) Faseb J 18, 1761-176330. Liao, H. J., Kobayashi, R., and Mathews, M. B. (1998) Proc Natl Acad Sci U S A 95, 8514-851931. Krasnoselskaya-Riz, I., Spruill, A., Chen, Y. W., Schuster, D., Teslovich, T., Baker, C., Kumar, A.,

and Stephan, D. A. (2002) AIDS Res Hum Retroviruses 18, 591-60432. Isken, O., Grassmann, C. W., Sarisky, R. T., Kann, M., Zhang, S., Grosse, F., Kao, P. N., and

Behrens, S. E. (2003) Embo J 22, 5655-566533. Shin, H. J., Kim, S. S., Cho, Y. H., Lee, S. G., and Rho, H. M. (2002) Arch Virol 147, 471-49134. Kablar, B., and Rudnicki, M. A. (2000) Histol Histopathol 15, 649-65635. Puri, P. L., and Sartorelli, V. (2000) J Cell Physiol 185, 155-17336. Figueroa, A., Cuadrado, A., Fan, J., Atasoy, U., Muscat, G. E., Munoz-Canoves, P., Gorospe, M.,

and Munoz, A. (2003) Mol Cell Biol 23, 4991-500437. Chen, X., Sadlock, J., and Schon, E. A. (1993) Biochem Biophys Res Commun 191, 18-2538. Botas, C., Poulain, F., Akiyama, J., Brown, C., Allen, L., Goerke, J., Clements, J., Carlson, E.,

Gillespie, A. M., Epstein, C., and Hawgood, S. (1998) Proc Natl Acad Sci U S A 95, 11869-1187439. Giles, K. M., Daly, J. M., Beveridge, D. J., Thomson, A. M., Voon, D. C., Furneaux, H. M.,

Jazayeri, J. A., and Leedman, P. J. (2003) J Biol Chem 278, 2937-294640. Zuker, M. (2003) Nucleic Acids Res 31, 3406-341541. Matzuk, M. M., Lu, N., Vogel, H., Sellheyer, K., Roop, D. R., and Bradley, A. (1995) Nature 374,

360-36342. Lee, C. G., da Costa Soares, V., Newberger, C., Manova, K., Lacy, E., and Hurwitz, J. (1998) Proc

Natl Acad Sci U S A 95, 13709-1371343. Meagher, M. J., Schumacher, J. M., Lee, K., Holdcraft, R. W., Edelhoff, S., Disteche, C., and

Braun, R. E. (1999) Gene 228, 197-21144. Meagher, M. J., and Braun, R. E. (2001) Mol Cell Biol 21, 2880-289045. Schumacher, J. M., Lee, K., Edelhoff, S., and Braun, R. E. (1995) J Cell Biol 129, 1023-103246. Pires-daSilva, A., Nayernia, K., Engel, W., Torres, M., Stoykova, A., Chowdhury, K., and Gruss,

P. (2001) Dev Biol 233, 319-32847. Baltzis, D., Li, S., and Koromilas, A. E. (2002) J Biol Chem 277, 38364-3837248. Higuchi, M., Maas, S., Single, F. N., Hartner, J., Rozov, A., Burnashev, N., Feldmeyer, D.,

Sprengel, R., and Seeburg, P. H. (2000) Nature 406, 78-8149. Wang, Q., Miyakoda, M., Yang, W., Khillan, J., Stachura, D. L., Weiss, M. J., and Nishikura, K.

(2004) J Biol Chem 279, 4952-496150. Rudnicki, M. A., Schnegelsberg, P. N., Stead, R. H., Braun, T., Arnold, H. H., and Jaenisch, R.

(1993) Cell 75, 1351-135951. Hasty, P., Bradley, A., Morris, J. H., Edmondson, D. G., Venuti, J. M., Olson, E. N., and Klein, W.

H. (1993) Nature 364, 501-50652. Zhang, P., Wong, C., Liu, D., Finegold, M., Harper, J. W., and Elledge, S. J. (1999) Genes Dev 13,

213-224


12

53. Inanlou, M. R., Dhillon, G. S., Belliveau, A. C., Reid, G. A., Ying, C., Rudnicki, M. A., andKablar, B. (2003) Dev Biol 261, 324-336

54. Venuti, J. M., Morris, J. H., Vivian, J. L., Olson, E. N., and Klein, W. H. (1995) J Cell Biol 128,563-576

55. Vivian, J. L., Gan, L., Olson, E. N., and Klein, W. H. (1999) Dev Biol 208, 44-5556. Kassar-Duchossoy, L., Gayraud-Morel, B., Gomes, D., Rocancourt, D., Buckingham, M., Shinin,

V., and Tajbakhsh, S. (2004) Nature 431, 466-47157. Parker, S. B., Eichele, G., Zhang, P., Rawls, A., Sands, A. T., Bradley, A., Olson, E. N., Harper, J.

W., and Elledge, S. J. (1995) Science 267, 1024-102758. Deng, C., Zhang, P., Harper, J. W., Elledge, S. J., and Leder, P. (1995) Cell 82, 675-68459. Wang, J., and Walsh, K. (1996) Science 273, 359-36160. van der Giessen, K., Di-Marco, S., Clair, E., and Gallouzi, I. E. (2003) J Biol Chem 278, 47119-

4712861. Lu, J. Y., and Schneider, R. J. (2004) J Biol Chem 279, 12974-1297962. Kablar, B., Krastel, K., Tajbakhsh, S., and Rudnicki, M. A. (2003) Dev Biol 258, 307-318


13

Genotype analysis of progeny from NF90+/- heterozygous intercrosses

Age NF90(+/+)

Number (%) Weight

NF90(+/-)

Number (%) Weight

NF90(-/-)

Number (%) Weight

< 2 h 32 (25) 1.42 + 0.24 65 (50.8) 1.39 + 0.25 31 (24.2) 0.94 + 0.11

Table


14

Figure Legends

Figure 1. Targeted disruption of the NF90 gene results in growth retardation and perinatal lethality.(A) Genomic organization of the NF90 locus, and targeting strategy to delete exons 2 – 4 by homologousrecombination. Restriction sites are shown for BamHI (B1), EcoRI (E1), HindIII (H3), NotI (N1). (B)Southern blot analysis of Hind III restricted R1 ES cell DNA. The 7.44-kb fragment represents the wild-type allele, and the 9.05-kb fragment represents the targeted allele (due to elimination of the Hind III sitedownstream of Exon 4). (C) Photograph of newborn mice from F1 intercross of NF90-targetedheterozygotes. (D) Genotype analysis of NF90-targeted E19 embryos by Southern hybridization.

Figure 2. Expression analysis of NF90 and NF110 in gene-targeted mice. (A) Northern hybridization ofNF90 RNA expression in gene-targeted littermates. NF90(+/+) mice show expression of a single transcriptof 3.4 kB, NF90(-/-) mice show prominent expression of a single targeted transcript of 3.1 kB andNF90(+/-) mice show mRNA transcription of both sizes. Constant expression of glyceraldehyde phosphatedehydrogenase (GAPDH) is shown in the lower panel. (B) Western immunoblot of NF90 proteinexpression in gene-targeted littermates. Equal amounts of extracted total embryonic proteins (20 µg) wereanalyzed. (C) Tissue expression of NF90 and NF110 in wild type mice. Equal amounts of total proteinwere analyzed (80 µg) by immunoblotting.

Figure 3. Immunohistochemical expression analysis of NF90 in wild-type NF90(+/+) and NF90(-/-)mice. NF90(+/+) staining is shown in (A – H) and NF90(-/-) staining is shown in panels (I – L). (A) Heart,(B) lung, (C) liver, (D) stomach; (E and I) cerebral cortex, (F and J) cerebellum, (G and K) diaphragmmuscle, and (H and L) skeletal muscle (latissimus).

Figure 4. Skeletal muscle histopathology. Comparative histology by hematoxylin and eosin staining ofdiaphragm muscle (A, B) and skeletal muscle (latissimus)(C, D) between NF90(+/+) mice (A, C) andNF90(-/-) mice (B, D).

Figure 5. TUNEL analysis of apoptosis in NF90(-/-) mice. TUNEL-positive nuclear stainingpredominated in cerebral and muscle tissues of NF90(-/-) mice compared to wild-type mice. Sections forTUNEL analysis were prepared from E19 specimens.

Figure 6. Myogenic Regulatory Factor expression is reduced in NF90(-/-) mice. (A) MyoD, myogeninand p21WAF1/CIP1 and b-actin mRNA were detected in neonatal skeletal muscle by reverse-transcription PCR.(B) Western immunoblot of p21WAF1/CIP1 expression; actin expression is shown as a loading control.

Figure 7. Northwestern analysis reveals that p21WAF1/CIP1 and MyoD 3’ UTR RNAs interactspecifically with NF90. RNA secondary structure predictions were performed using Zucker’s mfoldprogram (40) for (A) p21WAF1/CIP1 3’ UTR (511 – 883nt), and (B) MyoD (1241-1762 nt). Northwestern blotswere performed using in vitro transcribed, radiolabeled 3’-UTR RNA probes for murine p21WAF1/CIP1 (C) orMyoD (D). Murine tissue extracts (80 µg each) (lanes 1, 2, 3: adult NF90(+/+), lanes 5, 6: neonatal NF90(-/-)) and recombinant histidine-tagged NF90 and NF45 proteins (lanes 7, 8: 10 µg each) (5) werefractionated by SDS-PAGE, transferred to nitrocellulose and hybridized with probed 32P-lableled 3’-UTRprobes.

+/+ +/+ +/+ +/+ +/+ +/+ +/++/-

1 2 3 4 5 6 7 8

7.4kb WT

9.0kb MU

A

B

Figure 1

-/- +/+

+/- -/- +/+ +/+ -/-+/- +/- +/-

1 2 3 4 5 6 7 8

9.0kb MU

7.4kb WT

D

C

Neo HSV-TK

Neo

R1

R1

B1 B1

targeting construct

wild-type allele

targeted allele

5'-probe

9kb

9.8kb

StuI

H3H3 H3

H3

H3 H3

3'-probe

9kb

7kb 2.8kb

7.5kbR1 R1

R1R1 R1

R1

R1

NF90

G3PDH

+/- +/+ -/-+/- +/+ +/+ -/-WT 3.4kb--MU 3.1kb--

1 2 3 4 5 6 7

NF90

1 2 3 4 5 6 7

A

B

Figure 2

NF110/ILF3

90 kDa

110 kDa

+/- +/+ -/-+/- +/+ +/+ -/-

NF90

NF110/ILF3

90 kDa

110 kDa

Hea

rt

Bra

in

Sp

leen

Lu

ng

Liv

er

Ske

lMu

scle

Kid

ney

Test

is

1 2 3 4 5 6 7 8

C

+/+

+/+

-/-

A B C

E G

D

H

I J K L

F

Figure 3

-/-+/+

D

BA

C

Diaphragm

Latissimus

Figure 4

cerebralcortex

skeletalmuscles

+/+ -/-

A B

C D

Figure 5

MyoD

Myogenin

β-actin

p21WAF1/Cip1

+/+ +/++/- -/- -/- -/-

1 2 3 4 5 6

1 2 3 4 5 6

p21WAF1/Cip1

actin

A

B +/+ +/++/- -/- -/- -/-

Figure 6

MyoD

G

G

CAG

A

C

C

A

G

CC

U

GACAG

A

U

UU

C

UA

U

CAC

U

C

C

AA G

CG

C

A

G

A

UU

G

G

U

C

U U

C

U

G

C

A

A

G

A

G

A

A

A

A

C

C

C

UG

AA

GU

GCCCACGGG

AG

CCCC

G

C

C

CU C U

U

C UG

C U G U G G G U

C

A

G

G

A

G

G

CC

UCUU

CC

C

C

A

U

C

U

U

C

GG

CC

U

U A

GC

CC

UC

A

CU

C

U GU

G

UG

U C U U A A U U A UU

AU

UU

GUG

UUUUA

AU

UU

AA

AC

G

U

C

U

C

C

U

G

UA U

A

UA

CG

CU

G

CC U

GC

C CU

CU C

CCAG

U

C

U

C

CA

A A CU

U

A

A

A

GU

U

A UU

U

A A

A

AAA

A

GAA

CA A A A C A A A

A C AA A A A A A

AACCAA

AACAAAACAAAC C U A

A A

U

UAG

UAGGA

C

GG

UAG

GGC

CCU

UA

GU

GU

G

G

G

G

G

A

U

U

U

C

UA

UU

AU

G

U

A

G

A

UU

A UU

A

U

U

AU

UUA

A

G

C

C CC

U

C

C

C

A

AC

CCAA

G

C

U

C

U

G

50

100

150

200

250

300

350

Murine p21WAF1/CIP1 3'-UTRFree energy = -72 kJ/mol

A

G

A

C

A

C

UC

U

U

C

C

C

A

A

C

U

G

U

CC

U

U

U

CG

A

A

G

C

CG

U

UC

U

U

C

C

A

G

A

G

GGAAGGGAAGAGC

A

GA

A

G

U

C

U

G

UC

CU

A

G

A

UCC

AGCC

C

CA

AAG

AA

AG

GA

C

AU

A GU

CC

UU

UU

UG

UUGUUGUU

GUUGUA

G

UCC

U

U

CAG

UU

G

U

U

UG

U

U U G

U

UU

U UU

C

A

U G C G G CU C

A C A G C G AA

G

G

C

C

A

C

U

U

G

CA

CU

CU

GG

CUG

CAC

C

U

C A

C

UG

GG

CC

AG

AG C

U

G

AU

C

C

UU

G

A

G

U

G

G

C

C

A

G G C

G C U C U U C C U U U C C

U

C A U A G C A

C

A G

G

G

G

UG

A

G

C

C

U

U

GC A

C

A

C

C

U

AA

G

CC

C

U

G

CC

C

U

C

C

AC

AU

C

CU U U

U

G

U

U

U

G

U

C

A

C

UU

U C UG

G

AG

C

C

C

U C

C

UG

G

C

AC

C

C

A

C

U

U

U

UC

CCCA

C

A

G

C

UU

G

CG

G

A

G

G C

C

AC

U

C

A

G

G

UC

U

CA

G

G

U

G

UA

A

CA

G G U G U

AA

C

CAUACC

C

CAC

U

C

UC

C

C

CC

U

UC

CC

GC

GG

U UC

AG

GA C C

A

CU

UA

UU U

U

U

UUA

U

AU

AA

G

AC

UU

UU

GU

A A UC

U

A

U

U

CGU

G

U

A

A

AU

AA

GA

GUU

G

CU

UG

GC

C

AG

A

GC

GG

GA

G

C

CC

C

UU

G

G

GC

U

A

U

A

UU

U

A

UC

U

C

C

C

A

G

G

CA

UGCUGUG

U

A

G

U

G

C

50

100

150

200

250

300

350

400

450

500

MyoD 3' UTRFree energy = -156 kJ/mol

B

90 kDa 90 kDa

Test

is

Test

is

Bra

in

Bra

in

Mu

scle

Mu

scle

Bra

in

Mu

scle

Bra

in

Mu

scle

+/+ +/+-/- -/-

rNF

90

rNF

90

rNF

45

rNF

45

1 2 3 4 5 6 7 1 2 3 4 5 6 7

C D

Probe: p21 3'-UTR Probe: MyoD 3'-UTR

Figure 7

shi et al, diaphragmatic respiratory failure in nf90(-/-) mice nf90 …peterkao/kao...

Documents