THE JOURNAL OF EXPERIMENTAL ZOOLOGY 282:261–271 (1998)

© 1998 WILEY-LISS, INC. †This article is a US Government workand, as such, is in the public domain in the United States of America.

HSP70-2 Heat-Shock Protein of MouseSpermatogenic Cells

E.M. EDDY*Gamete Biology Section, Laboratory of Reproductive and DevelopmentalToxicology, National Institute of Environmental Health Sciences, NationalInstitutes of Health, Research Triangle Park, North Carolina 27709

ABSTRACT The HSP70 heat-shock proteins are molecular chaperones that assist other pro-teins in folding, transport, and assembly into complexes. The genes for these proteins are eitherconstitutively expressed (Hsc70, Grp78), or their expression is induced by heat shock and otherstresses (Hsp70-1, Hsp70-3). Two additional genes encode proteins that are developmentally regu-lated and expressed specifically in spermatogenic cells (Hsp70-2, Hsc70t). The HSP70-2 protein issynthesized during the meiotic phase of spermatogenesis and is abundant in pachytene spermato-cytes. Studies in transgenic mice indicated that the region between nucleotides –640 and +1 con-tains promoter sequences necessary for expression of Hsp70-2 in spermatocytes. Because of thepattern of gene expression, it was hypothesized that HSP70-2 is a chaperone necessary for comple-tion of meiosis in spermatogenic cells. The gene knockout approach was used to test this hypoth-esis, and it was found that male mice homozygous for the mutation were infertile, whereashomozygous females were fertile. Spermatogenesis was disrupted, with the nuclei of late pachytenespermatocytes often appearing fragmented and spermatids being absent. Disruption of spermato-genesis occurred at the G2-M phase transition in prophase of meiosis I, and all pachytene sperma-tocytes underwent apoptosis. It was demonstrated that HSP70-2 is a chaperone for Cdc2, withtheir association allowing Cdc2 to acquire the necessary conformation to form a heterodimer withcyclin B1, leading to changes in Cdc2 phosphorylation and the development of kinase activitynecessary for the G2-M phase transition. This appears to be the first demonstration that interac-tion between an HSP70 protein and a cyclin-dependent kinase is necessary for progression of thecell cycle. J. Exp. Zool. 282:261�271, 1998. © 1998 Wiley-Liss, Inc.†

CHAPERONES AND HEAT-SHOCKPROTEINSChaperones

Molecular chaperones assist other proteins intheir folding, transport, and assembly into com-plexes. They prevent the premature misfolding ofnascent polypeptide chains as they emerge fromribosomes by associating with exposed hydropho-bic domains. They also associate with foldedpolypeptides during their “molten globule” stateby ATP-dependent cycles of release and rebind-ing, until the polypeptides achieve their correctlyfolded state. Some cytosolic proteins completetheir folding within chaperonins, barrel-shapedcomplexes of two stacked rings of seven to ninesubunit proteins (Hendrick and Hartl, ’95; Rassowet al., ’97). The chaperonins in mammals areheterooligomeric complexes that are referred toas “chaperonin-containing TCP-1” (CCT) or “TCP-1 ring complex” (TRiC) structures (Frydman etal., ’92; Kubota et al., ’95). The subunit proteins

in mice are encoded by nine different genes(Kubota et al., ’97). The first of these reported wasTCP-1 and was identified from studies on mousespermatogenic cells (Silver and White, ’82). Manyof the subsequent studies characterizing CCT com-plexes also have been done in spermatogenic cells(Kubota et al., ’94; Hynes et al., ’96). Most of theCCT proteins appear to be present in all cell types,but the Cctz-2 gene is expressed only in spermato-genic cells (Kubota et al., ’97).

HSP70 heat-shock proteinsThe best characterized group of chaperones are

the 70-kDa HSP70 heat-shock proteins, which arehighly homologous from Escherichia coli to hu-mans and among the most conserved proteins

*Correspondence to: Dr. E.M. Eddy, Laboratory of Reproductiveand Developmental Toxicology, C4-01, National Institute of Environ-mental Health Sciences, National Institutes of Health, Research Tri-angle Park, NC 27709.

Received 27 March 1998; Accepted 27 March 1998

262 E.M. EDDY

known (Hunt and Morimoto, ’85). HSP70 is theproduct of a single-copy gene in E. coli, whereasHSP70 proteins in eukaryotes are encoded bymultigene families. There are at least five and pos-sibly as many as ten closely homologous genes forHSP70 proteins in mammals (Hunt et al., ’93).The murine HSP70 proteins are encoded by genesthat are either constitutively expressed, or theirexpression is induced by heat shock and otherstresses in most or all cell types. Furthermore,two HSP70 proteins encoded by developmentallyregulated genes are expressed specifically in sper-matogenic cells.

Constitutive heat-shock proteinsThe Hsc70 and Grp78 genes are constitutively

expressed in spermatogenic cells. The Hsc70 geneencodes the HSC70 (also referred to as hsc71 orhsp73) heat-shock protein (Giebel et al., ’88).HSC70 is the major cytosolic molecular chaper-one of mammalian cells (Hendrick and Hartl, ’93).It is an ATP-binding protein like other membersof the HSP70 family, and one of its specific rolesis the removal of clatherin protein from coatedpits (Chappell et al., ’86). The protein has beenidentified in developing germ cells from thepreleptotene to the condensing spermatid stagesof mouse spermatogenesis and in sperm (Allen etal., ’88b; Maekawa et al., ’89). It was found in boththe nuclear and cytoplasmic fractions of mousespermatogenic cells (Allen et al., ’96). The Grp78gene is on chromosome 2 of the mouse (Hunt etal., ’93) and encodes a “glucose-regulated” heat-shock protein that is constitutively expressed.Higher levels of GRP78 (also referred to as hsc74,BiP, or grp80) expression are induced by glucosestarvation and a variety of chemicals, but not byheat shock (Shiu et al., ’77; Olden et al., ’79; Welchet al., ’83; Resendez et al., ’85). The GRP78 pro-tein is present in the endoplasmic reticulum, andin lymphocytes it is a chaperone for immunoglo-bulin heavy chain protein (Munro and Pelham,’86). It is relatively abundant in mouse pachytenespermatocytes and round spermatids (Allen et al.,’88a), and it is present in the cytoplasmic fractionof spermatogenic cells (Allen et al., ’96). The Grp75gene encodes a mitochondrial HSP70 (Mizzen etal., ’89) that is present in testis (Massa et al., ’95)and probably in spermatogenic cells.

Inducible heat-shock proteinsExpression of the Hsp70-1 and Hsp70-3 genes

is induced by heat shock or other stresses. Thesegenes are located in tandem on mouse chromo-

some 17, within the major histocompatibility com-plex (Hunt et al., ’93). There have been differentopinions on whether these genes are induced inspermatogenic cells by heat shock. The initialstudies indicated that synthesis of a new HSP70protein was induced in isolated mouse pachytenespermatocytes and round spermatids subjected toheat stress, as determined by biosynthetic label-ing and analysis by two-dimensional (2D) SDS-PAGE and autoradiography (Allen et al., ’88a,b).Furthermore, the peptide map of HSP70 proteininduced by heat stress in pachytene spermatocyteswas comparable with that of HSP70 protein in-duced by heat stress in 3T3 cells (Allen et al.,’88b). However, other investigators found only lowlevels of heat-inducible HSP70 protein and mRNAin isolated mouse spermatogenic cells and sug-gested that these were produced by contaminat-ing somatic cells (Zakeri et al., ’90). More recentstudies have indicated that heat stress of mousepachytene spermatocytes causes increased phos-phorylation of the HSF1 heat-shock transcriptionfactor and an accumulation of heat-inducibleHSP70 protein (Sarge, ’95). Additional studiesfound that spermatocytes exhibit a lower tempera-ture threshold for activation of HSF1 and de novoHSP70 synthesis relative to other mouse cell types(Sarge and Cullen, ’97). Although most of the evi-dence indicates that a heat-shock response occursin spermatogenic cells, it is unknown if one or bothof the inducible Hsp70 genes are activated.

Developmental heat-shock proteinsThe Hsp70-2 and Hsc70t genes are developmen-

tally regulated members of the HSP70 family thatare expressed specifically in spermatogenic cells.The Hsp70-2 gene is on mouse chromosome 12(Hunt et al., ’93), and the HSP70-2 protein is syn-thesized mainly during the meiotic phase of sper-matogenesis (O’Brien, ’87; Allen et al., ’88a,b;Rosario et al., ’92; Dix et al., ’96b). The Hsc70tgene is on mouse chromosome 17, adjacent to theHsp70-3 gene (Hunt et al., ’93), and is expressedonly during the postmeiotic phase of spermatoge-nesis (Zakeri and Wolgemuth, ’87; Maekawa etal., ’89; Matsumoto and Fujimoto, ’90; Matsumotoet al., ’93). The Hsp70-2 and Hsc70t genes aremembers of a category that has been referred toas “chauvinist genes” because male germ cells fa-vor their expression with a strong prejudice (Eddy,’95). Like other genes expressed specifically inspermatogenic cells, Hsp70-2 and Hsc70t aremembers of gene families. The developmentalregulation of such genes presumably is due to ac-

HSP70-2 OF SPERMATOGENIC CELLS 263

tivation by spermatogenic cell-specific combina-tions of transcription factors, some of which maybe unique to these cells.

IDENTIFICATION OF HSP70-2HEAT-SHOCK PROTEIN

Identification of P70 proteinSpermatogenic cells of most mammals are main-

tained 5 to 7°C below body temperature by twoanatomic and physiologic specializations: locationof the testis in the scrotum outside the body cav-ity and a countercurrent heat-exchange mecha-nism that cools the blood entering the testis(Kandell and Swerdloff, ’88). The process of sper-matogenesis is sensitive to modest elevation oftemperature and to other environmental insults(Chowdhury and Steinberger, ’64; Creasy, ’97).This suggested that differences might occur be-tween the heat-shock proteins present in sper-matogenic cells and in other cell types. To examinethis possibility, heat-shock proteins of mouse 3T3tissue culture cells were compared with those ofspermatogenic cells. 3T3 cells were cultured at37°C and their HSP70 proteins examined by 2DSDS polyacrylamide gel electrophoresis (2D SDS-PAGE). The constitutively expressed HSC70 andGRP78 heat-shock proteins were found to be rela-tively abundant in these cells (Allen et al., ’88a).When 3T3 cells were heat-shocked at 42.5°C for10 min and then cultured at 37°C for 6 hr, theinducible HSP70 protein also was present. Ex-tracts of mixed spermatogenic cells maintained at32°C also contained relatively abundant amountsof HSC70 and GRP78. However, they containedan additional protein (referred to as P70) of thesame pI and mass as the HSP70 protein in heat-shocked 3T3 cells (Allen et al., ’88a). It was de-termined by 2D SDS-PAGE with purified cells thatP70 was present only in the spermatogenic cellsof the testis (O’Brien, ’87; Allen et al., ’88b).

These results suggested that the P70 protein ofspermatogenic cells was related to the HSP70 pro-teins. P70 was further examined to see if this wastrue. A characteristic feature of HSP70 proteinsis that they bind ATP and can be purified on anATP affinity column (Welch and Feramisco, ’85).It was found that P70 also binds to an ATP affin-ity column (Allen et al., ’88a). In addition, a mono-clonal antibody to Drosophila Hsp70 (Kurtz et al.,’86) that recognizes many HSP70 family proteinsin a wide range of species (Munro and Pelham,’86) also recognized P70 in mouse spermatogeniccell extracts. Although these results suggested

that P70 was related to HSP70, it was not clearif it was equivalent to HSP70 or was a uniqueHSP70 protein of spermatogenic cells. If P70 wasthe same as HSP70, the gene might be inducedby endogenous stresses in spermatogenic cells.However, if it was a different protein, spermatoge-nic cells might have their own heat-shock protein.When peptide map analyses were performed, thepatterns for HSP70 from 3T3 cells and P70 fromspermatogenic cells were similar but not identical,suggesting that P70 and HSP70 were related butnot the same protein (Allen et al., ’88a).

To determine when P70 is synthesized duringmouse spermatogenic cell development, isolatedpopulations of spermatogenic cells were incu-bated in medium containing 35S-labeled me-thionine and analyzed by 2D SDS-PAGE andfluorography. It was found that P70 synthesisoccurs mainly during meiosis by beginning inleptotene spermatocytes, reaching a high levelin pachytene spermatocytes, and then becom-ing low in round spermatids (O’Brien, ’87). Thisindicated that the synthesis of P70 was develop-mentally regulated and suggested that the pro-tein serves a role during the meiotic phase ofspermatogenesis.

Isolation and characterization of P70 cDNAThe genes encoding HSP70 family proteins have

high sequence homology (Hunt and Morimoto, ’85),and identification of the gene for P70 was impor-tant for determining its relationship to this fam-ily. A genomic clone for a member of the HSP70gene family (Hsp70-2) was reported to hybridizewith 2.7-kb transcripts abundant in pachytenespermatocytes of the testis but not detected inother tissues (Zakeri et al., ’88). This suggestedthat the Hsp70-2 gene might encode the P70 pro-tein. To examine this possibility, an antiserum wasproduced against a peptide sequence of the de-duced HSP70-2 protein that was not present inother HSP70 family members or known proteins.The antiserum was specific for P70 on Westernblots of mixed germ cell proteins separated by 2DSDS-PAGE (Rosario et al., ’92). It did not bind toother proteins on Western blots of somatic or sper-matogenic cells, and binding to P70 was elimi-nated by preincubation of antiserum with thepeptide. In addition, cyanogen bromide fragmentsof the isolated P70 protein were of the size pre-dicted from the location of methionine residues ofthe protein encoded by Hsp70-2 (Rosario et al.,’92). When the antiserum was used to localize P70in mouse testis by immunohistochemistry, it was

264 E.M. EDDY

found to be abundant in spermatogenic cells fromthe early part of meiotic prophase until the endof spermatid development (Rosario et al., ’92). Thiswas consistent with earlier results that P70 wassynthesized in the leptotene and pachytene sper-matocyte stages of spermatogenesis (O’Brien, ’87).Since P70 protein synthesis is low in spermatids(O’Brien, ’87), the immunostaining in round andcondensing spermatids indicated that P70 mustbe a stable protein to persist throughout the 2-week postmeiotic phase of spermatogenesis.

The antiserum also was used to screen a λgt11expression vector library prepared with mRNAfrom mouse pachytene spermatocytes. Twelve ofthe 15 clones that were isolated contained cDNAinserts that cross-hybridized and ranged from 900to 1650 base pairs in length. The longest cDNAswere sequenced and found to contain nucleotidesequences virtually identical to those in theHsp70-2 genomic clone. Northern analysis of to-tal RNA samples determined that these cDNAshybridized with a 2.7-kb transcript that is abun-dant in pachytene spermatocytes and round sper-matids but not in the somatic tissues examined.Northern analysis of RNA from testes of juvenilemice at different ages first detected the 2.7-kbtranscript at day 10. This coincides with the ear-liest appearance of leptotene spermatocytes in thenearly synchronous first wave of spermatogenesisof the prepuberal mouse (Bellvé et al., ’77). Thetranscript level was high on day 14, when pri-mary spermatocytes reach the early pachytenestage of development. Since these results weresimilar to those reported for the Hsp70-2 gene(Zakeri et al., ’88), they support the conclusionthat P70 is the product of the Hsp70-2 gene(Rosario et al., ’92).

Hsp70-2 geneThe Hsp70-2 gene was first isolated from a

mouse genomic library using a Drosophila HSP70gene probe (Zakeri et al., ’88). Southern analysisindicated that Hsp70-2 is a single-copy gene(Zakeri et al., ’88). Sequence analysis identifiedthe coding region and indicated that the Hsp70-2gene contains a single unspliced open readingframe capable of encoding a protein of 634 aminoacids. The 5´ upstream region was reported to con-tain an untranslated 121-base-pair leader se-quence and other promoter sequence motifs,including a TATA box at nucleotide -153 bp, aninverted CCAAT box at nucleotide -192, and a par-tial heat-shock element (HSE) at nucleotide -186,relative to the translation start codon (Zakeri et

al., ’88). Since heat-inducible HSP70 genes usu-ally have multiple overlapping HSEs of high ho-mology (Lindquist, ’86), it was concluded that theHsp70-2 HSE sequence probably was nonfunc-tional with respect to heat inducibility. The mouseHsp70-2 gene was found to have 79% nucleotide(83% amino acid) similarity to the heat-inducibleHSP70 genes in both human and mouse. Thenucleotide similarity of Hsp70-2 with the consti-tutively expressed Hsc70 gene in the human is73% (86% amino acid) and in the rat is 74% (87%amino acid) (Zakeri et al., ’88).

Fourteen nucleotide differences were found be-tween the cDNA sequences (Rosario et al., ’92) andthat of the Hsp70-2 genomic fragment (Zakeri etal., ’88), but only two were within the coding re-gion and did not alter the predicted amino acidsequence. Since the cDNAs and the genomic clonewere isolated from libraries derived from differ-ent mouse strains, these nucleotide differencesmay represent allelic variations. However, thecDNA sequences indicated that the transcriptended at nucleotide 3031 and that the polya-denylation signal was probably at nucleotide 2967or 2994 (Rosario et al., ’92) rather than at posi-tion 3350, as suggested earlier (Zakeri et al., ’88).Primer extension analysis indicated that putativetranscription start sites for Hsp70-2 were presentas far upstream as nucleotide –353 (Dix et al.,’96b) rather than at nucleotide –121, as suggestedearlier (Zakeri et al., ’88). These studies found apreviously unidentified 239-base-pair intron 5´ ofthe Hsp70-2 coding region when cDNA and ge-nomic sequences were compared (Dix et al., ’96b).

Northern analysis with an antisense-orientationRNA probe for Hsp70-2 demonstrated that a 2.7-kb transcript was present at high levels in mousetestis and low levels in several other tissues, in-cluding brain, epididymis, prostate, seminalvesicles, and extraembryonic tissues of mid-gestation embryos (Murashov and Wolgemuth,’96a). In addition, Northern analysis with a sense-orientation RNA probe for the same sequencefound a 2.8-kb “antisense” transcript in mouseprostate, seminal vesicles, and brain but not tes-tis (Murashov and Wolgemuth, ’96a,b). It shouldbe noted that the 230-nt RNA probes were pro-duced from an Hsp70-2 genomic clone and con-tained 195 nt of intronic sequence that was notpresent in Hsp70-2 cDNA from spermatogeniccells (Dix et al., ’96b). This suggests either thatthe short region of sequence homology of theprobes and transcripts was sufficient for hybrid-ization or that sense and antisense Hsp70-2 tran-

HSP70-2 OF SPERMATOGENIC CELLS 265

scripts containing the intron sequence are presentin some tissues. Low levels of Hsp70-2 transcriptsalso were detected in mouse somatic tissues bythe reverse transcriptase polymerase chain reac-tion (RT-PCR) assay (Dix et al., ’96b). The PCRprimers used were located in exons 1 and 2 andshould have detected both processed and unproc-essed Hsp70-2 transcripts if they were present,but only processed transcripts were found. In ad-dition, low levels of Hsp70-2 transcripts were de-tected in some mouse somatic tissues on Northernblots prepared with poly(A+) RNA and subjectedto long exposures, using a hybridization probe forthe 3´ untranslated region (Dix et al., ’96b). Al-though low levels of Hsp70-2 transcripts arepresent in somatic tissues, gene knockout studiesindicate that the protein has a significant role onlyin spermatogenic cells (see below).

Regulation of Hsp70-2 expressionAlthough the protein encoded by Hsp70-2 is

similar to other Hsp70s, the developmentally regu-lated expression of this gene is unlike that of mostmembers of this gene family. However, there area number of other “chauvinist” genes expressedspecifically in spermatogenic cells (Eddy, ’95), andsome of these are expressed only during meiosis(Eddy and O’Brien, ’98). This suggests that thereare elements in the promoter region of Hsp70-2and other genes, activated by transcription fac-tors present in pachytene spermatocytes, thatregulate their cell type–specific and developmen-tal stage–specific expression. A commonly usedapproach to identify the promoter regulatory se-quences is to perform transient transfection stud-ies with different promoter constructs. Thesestudies typically are carried out in a tissue cul-ture line that is capable of expressing the geneand thus has the appropriate combination of tran-scription factors required for activating transcrip-tion. However, there are no germ cell tissueculture lines available to allow these approachesto be used for analysis of the Hsp70-2 gene pro-moter. Fortunately, promoter-reporter constructscan be used in transgenic mice to define the 5´flanking region that contains sequence elementsinvolved in regulating expression of such a gene.

To define the region required for regulating ex-pression of the Hsp70-2 gene, a 2.8-kb genomicsegment immediately 5´ to the coding region ofthe Hsp70-2 gene was ligated to the bacterial β-galactosidase (lacZ) reporter gene. This was di-gested with one of three restriction enzymes toproduce different length promoter regions. The

three transgenes were injected into pronuclei ofmouse zygotes to generate multiple transgeniclines for each of the promoter constructs (Dix etal., ’96b). Transgene expression was detected byβ-galactosidase histochemistry on whole testes orwith the O-nitrophenyl-β-D-galactopyranoside(ONPG) assay on tissue homogenates. The lacZgene was expressed in the testes of mice with atransgene containing a promoter fragment extend-ing to nucleotide –640 upstream of the transla-tion start codon but not when the transgenecontained a 318-base-pair promoter fragment. His-tologic analysis of the testis indicated that in bothjuvenile and adult mice the β-galactosidase prod-uct was first synthesized in early pachytene sper-matocytes, coincident with HSP70-2 proteinsynthesis, as determined by immunostaining. Thissuggested strongly that the region lying betweennucleotides –640 and –318 contains promoter se-quences that are necessary to trigger expressionof the Hsp70-2 gene in spermatogenic cells (Dixet al., ’96b). However, other sequences betweennucleotide –318 and the translation start codonalso may be required for Hsp70-2 expression. Com-parisons of this region with promoters of othergenes expressed in spermatocytes did not revealcommon sequence elements, but the promoter re-gions for few of these genes have been well char-acterized.

The Hsp70-2 gene promoter region contains asequence similar to the palindromic estrogen-re-sponse element (ERE). The estrogen receptor is aligand-dependent transcription factor that regu-lates the expression of some genes by binding toERE motifs in their promoters. However, HSP70-2 was present in spermatogenic cells of male micehomozygous for a mutation in the major estrogenreceptor (ER-α) gene (Eddy et al., ’96). This con-firmed that the ERE-like sequence does not havea significant role in regulating this gene (Kraw-czyk et al., ’92).

Two heat-shock transcription factors participatein the regulation of heat-shock gene expression,HSF1 and HSF2. HSF1 is primarily responsiblefor mediating the cellular stress response, whereasHSF2 mainly regulates heat-shock gene expres-sion under nonstress conditions, including devel-opment and differentiation (Morimoto et al., ’92).The Hsf2 transcript is present at higher levels intestis than in other tissues of the mouse, increasesbetween days 14 and 21 of postnatal development,and is more abundant in pachytene spermatocytesand round spermatids than in germ cells at otherstages of development (Sarge et al., ’94). Two

266 E.M. EDDY

HSF2 isoforms are produced by alternative tran-script splicing: a larger HSF2α isoform that pre-dominates in the testis and appears to be the morepotent transcriptional activator and a smallerHSF2β isoform found elsewhere (Goodson et al.,’95). The Hsp70-2 promoter contains a partialHSE, and HSF2 was reported to bind to theHsp70-2 promoter sequence in an in vitro assay(Sarge et al., ’94). However, HSP70-2 protein syn-thesis begins during spermatocyte developmentbefore the increase in Hsf2α transcription occurs,suggesting that HSF2 does not regulate Hsp70-2expression.

ROLE OF HSP70-2 IN SPERMATOGENESIS

Hsp70-2 gene knockoutExpression of the Hsp70-2 gene and synthesis

of the HSP70-2 protein both begin early in themeiotic phase of spermatogenesis, suggesting thatthe protein has a role in meiosis. Unique processesthat occur during meiosis include chromosomecondensation, pairing of homologous chromo-somes, formation of the synaptonemal complexes(SCs), and genetic recombination (Moens, ’94). Itwas found that HSP70-2 is associated with theSC in pachytene spermatocytes (Allen et al., ’96;Dix et al., ’96a). These observations suggested thatthe HSP70-2 protein is required for completion ofmeiotic prophase during spermatogenesis. Sincethe HSP70 proteins are chaperones, it was hy-pothesized that HSP70-2 was necessary for thefolding, transport, or assembly of protein com-plexes required for completion of meiosis. To testthis hypothesis and to determine the role ofHSP70-2, homologous recombination was used todisrupt the Hsp70-2 gene (Dix et al., ’96a).

Male and female mice that were homozygous(Hsp70-2�/�) or heterozygous (Hsp70-2+/�) for themutation were mated with wild-type mice (Hsp70-2+/+) to determine if the mutation had an effecton fertility. The heterozygous male and femalemice and Hsp70-2�/� female mice were fertile. How-ever, Hsp70-2�/� male mice were infertile (Dix etal., ’96a). There were no significant differences inthe weight of androgen-dependent organs inHsp70-2�/� males, indicating that infertility wasprobably not due to an endocrine disorder. How-ever, the testis weight in Hsp70-2�/� mice was lessthan one-half that in wild-type mice, and histo-logic examination of the testes revealed that sper-matogenesis was disrupted. Spermatogonia andspermatocytes were present, but spermatids wereabsent. Late pachytene spermatocytes were infre-

quent, and the nuclei of pachytene spermatocytesoften appeared fragmented. The fertility of Hsp70-2�/� female mice indicated that the HSP70-2 pro-tein is not required for meiosis during oogenesisand was consistent with other observations thatthe Hsp70-2 gene does not appear to be expressedat significant levels in germ cells of the female(Allen et al., ’96; Dix et al., ’96b).

Apoptosis in spermatocytes of Hsp70-2–/–

The presence of fragmented nuclei in pachytenespermatocytes and the absence of spermatids sug-gested that spermatocyte development is arrestedand that the cells undergo programmed cell deathin the absence of HSP70-2. Although apoptosis isa common event in spermatogenic cells, mostapoptotic cells are spermatogonia and few arespermatocytes (Allen et al., ’87; Billig et al., ’95;Mori et al., ’97). Use of the TUNEL method onsections indicated that chromosomal fragmenta-tion, a hallmark of apoptosis, was occurring inmany late pachytene spermatocytes in Hsp70-2�/�

mice. DNA ladders were detected by Southernanalysis, verifying that internucleosomal chroma-tin degradation typical of apoptosis was occurring(Dix et al., ’96a; Mori et al., ’97).

The constitutive GRP78 and HSC70 proteinsare present in pachytene spermatocytes(O’Brien, ’87; Allen et al., ’88a,b, ’96) but do notcompensate for the loss of HSP70-2 in Hsp70-2�/� mice. This suggests that the role of HSP70-2 in preventing apoptosis in pachytenespermatocytes is relatively specific. It has beenshown that heat shock and other stresses trig-ger apoptosis in somatic cells, that precondition-ing cells with mild heat stress to induce Hsp70reduces apoptosis, and that antisense oligomersof heat-inducible Hsp70 enhance initiation ofapoptosis (Wei et al., ’94, ’95). It also has beendetermined that DNA strand breaks can trig-ger p53-dependent apoptosis (Clarke et al., ’94)and that p53 is relatively abundant in sperma-tocytes (Schwartz et al., ’93). This suggestedthat spermatocytes may be poised to undergoapoptosis triggered by the failure of DNA re-pair or recombination processes. Proteins in-volved in DNA repair or recombination orinhibitors of apoptosis that require HSP70-2chaperone activity to fold or assemble into func-tional complexes may be disrupted in pachytenespermatocytes of Hsp70-2�/� mice, thereby lead-ing to apoptosis (Dix et al., ’96a). However, theassociation between HSP70-2 and apoptosis re-mains to be determined.

HSP70-2 OF SPERMATOGENIC CELLS 267

Meiosis and HSP70-2Spermatogenic cells appeared to develop nor-

mally to postnatal day 15 in Hsp70-2�/� mice, butthe frequency of TUNEL-positive cells was signifi-cantly higher at day 17, indicating that apoptosiswas occurring in late pachytene spermatocytes(Mori et al., ’97). Pachytene spermatocytes are inprophase of meiosis I, and disruption of spermato-genesis in Hsp70-2�/� mice occurred at the transi-tion from G2 to M phase of the cell cycle. The G2-Mtransition requires Cdc2 kinase activity (Draettaand Beach ’88; Draetta et al., ’88), which is ac-quired when the cyclin B regulatory subunit bindsto the Cdc2 catalytic subunit and triggers changesin Cdc2 phosphorylation (Solomon, ’94; Morgan,’95). Recent studies in mice found that cyclin B1is present at high levels in pachytene spermato-cytes (Chapman and Wolgemuth, ’94) and thatCdc2 transcripts are abundant in late pachyteneand diplotene spermatocytes preparing to undergothe first meiotic division (Rhee and Wolgemuth,’95). In addition, Cdc2 kinase activity is presentmainly in pachytene spermatocytes and low or un-detectable in somatic cells and early germ cells ofthe testis (Chapman and Wolgemuth, ’94). Thisled to the hypothesis that HSP70-2 is a molecu-lar chaperone required for Cdc2 activation, withtheir interaction being necessary for spermatoge-nic cells to proceed to the M phase of meiosis I(Zhu et al., ’97).

It was found by immunoprecipitation and West-ern blotting approaches that HSP70-2 associateswith Cdc2 but not with cyclin B1 or the Cdc2–cyclin B1 heterodimer in the testis of wild-typemice. In contrast, although cyclin B1 and Cdc2 arepresent in the testis of Hsp70-2�/� mice, they donot assemble into a heterodimer. Furthermore,Cdc2 in the testis of Hsp70-2�/� mice lacked ki-nase activity and did not undergo the changes inphosphorylation that occurred in wild-type mice.These results strongly suggested that the pres-ence of HSP70-2 was required for Cdc2 to form acomplex with cyclin B1 and become an active ki-nase. This was verified by showing that addition ofrecombinant HSP70-2 protein to a homogenate oftestis from Hsp70-2�/� mice restored the ability ofCdc2 to form a heterodimer with cyclin B1 and tobecome an active kinase (Zhu et al., ’97). These re-sults strongly suggest that HSP70-2 is a chaper-one for Cdc2, with their association allowing Cdc2to acquire the necessary conformation to form aheterodimer with cyclin B1. Cdc2–cyclin B1 asso-ciation then leads to changes in Cdc2 phosphory-

lation and development of the kinase activity nec-essary for the G2-M transition to occur in prophaseI of meiosis in pachytene spermatocytes. Thesestudies are the first to demonstrate that interac-tion between an HSP70 protein and cyclin-depen-dent kinase (CDK) is necessary for progression ofthe cell cycle in any cell type.

Other roles of HSP70-2It is likely that proteins in addition to Cdc2 re-

quire chaperoning by HSP70-2. By biosyntheticlabeling of proteins synthesized in isolated sper-matogenic cells, followed by immunoprecipita-tion with antiserum to HSP70-2, it was foundthat multiple proteins coimmunoprecipitatewith HSP70-2 (Allen et al., ’96). One or moreof these is probably a protein of the SC. HSP70-2 was seen on meiotic spreads to be associatedwith SCs of pachytene spermatocytes, and disas-sembly of the SCs at the end of meiotic prophaseis disrupted in Hsp70-2�/� mice (Allen et al., ’96;Dix et al., ’96b). SC proteins that might be asso-ciated with HSP70-2 include those with structuralroles, such as SCP1 (Moens and Syropoulos, ’95),COR1 (Dobson et al., ’94), and SC65 (Chen et al.,’92). In addition, proteins are associated with theSC that have roles in DNA recombination and re-pair, including RAD51 (Haaf et al., ’95), PMS2(Baker et al., ’95), MLH1 (Baker et al., ’96;Edelmann et al., ’96), DNA topoisomerase II(Moens and Ernshaw, ’89; Cobb et al., ’97), BRCA1(Scully et al., ’97), ATM, ATR (Keegan et al., ’96),and UBC9 (Kovalenko et al., ’96). If HSP70-2 is achaperone for one or more of these proteins dur-ing SC disassembly, absence of HSP70-2 couldlead to failed desynapsis and fragmentation ofSCs. In addition, SCP1 is a major component ofthe transverse filaments of the SC and containsa carboxy-terminal basic domain that is a poten-tial target site for phosphorylation by Cdc2(Meuwissen et al., ’92; Dobson et al., ’94). Sincephosphorylation can lead to disassembly of pro-tein complexes (Draetta and Beach, ’88), failureof SC desynapsis in Hsp70-2�/� mice might becaused by a lack of Cdc2 protein kinase activityneeded for this process.

A rat sperm 68-kDa protein referred to as SLIP1 (sulfolipid immobilized protein 1) that specificallybinds to sulfolipids in vitro (Lingwood, ’85) colo-calizes on sperm with SGG (sulfogalactoglycero-lipid), a major acidic sulfogalactolipid of the testis,brain, and kidney (Lingwood, ’86). Defects in tes-ticular SGG biosynthesis can result in sterility(Lingwood et al., ’85). SLIP 1 is restricted in dis-

268 E.M. EDDY

tribution to the testis, brain, and oocytes (Law etal., ’88) and is present on the surface of spermato-genic cells (Lingwood, ’85) and sperm (Lingwood,’86; Tanphaichitr et al., ’93). SLIP 1 contains anATP-binding site (Lingwood and Nutikka, ’91) andis immunologically related to HSP70 family pro-teins, including HSP70-2 (Boulanger et al., ’95).Incubation of mouse sperm in antiserum to SLIP1 inhibited sperm-egg binding, as determined byin vivo (Tanphaichitr et al., ’92) and in vitro fer-tilization assays, and addition of purified SLIP 1to in vitro fertilization medium decreased mousesperm-egg binding (Tanphaichitr et al., ’93).

These studies suggest that HSP70-2, or otherproteins with which it is associated, is a spermsurface component with an important role infertilization. However, it remains to be provedthat SLIP 1 is HSP70-2, and a direct interac-tion between HSP70-2 and SGG has not beendemonstrated.

Other speciesThe rat Hst70 gene is highly expressed in the

testis and encodes a protein that differs frommouse HSP70-2 in only 4 of 633 amino acids(Wisniewski et al., ’90). The transcript was de-tected in the testis of 21-day-old rats (Krawczyket al., ’87) and was localized to pachytene sper-matocytes and spermatids (Krawczyk et al., ’88).The rat HST70 protein was detectable in the tes-tis of 22-day-old rats, pachytene spermatocytes,and spermatids isolated from adult rats but notin sperm or in any somatic tissue examined (Raabet al., ’95). A 253-base-pair region of the Hst70promoter adjacent to the translation start codonwas sufficient to produce testis-specific and de-velopmentally regulated expression of a reportergene in transgenic mice (Widlak et al., ’95). Anantiserum that recognized rat HST70 also recog-nized a protein of the same apparent mass in tes-tes of human, boar, guinea pig, and rooster (Raabet al., ’95). The antiserum to SLIP 1, which is ap-parently HST70, also recognized a protein of thesame mass as HSP70-2 in testes of hamster,guinea pig, rabbit, dog, pig, bull, Xenopus laevis,cod fish, and rooster (Law et al., ’88).

The human HSPA2 gene has 91.7% homologyat the nucleotide level in the coding region and98.2% homology at the amino acid level to themurine Hsp70-2 gene (Bonnycastle et al., ’94). Thededuced HSPA2 protein differs in four amino ac-ids from mouse HSP70-2 and five amino acidsfrom rat HST70 and contains a six-amino-acid se-quence near the 3´ end of the coding region not

present in mouse or rat. The gene is located onhuman chromosome 14 and was mapped to 14q-24.1 (Bonnycastle et al., ’94) or 14q22 (Roux etal., ’94). A 2.9-kb transcript was present at highlevels in testis but also was relatively abundantin skeletal muscle and detectable in ovary, smallintestine, colon, and brain (Bonnycastle et al., ’94).There are regions of homology in the promoterregions of the human HSPA2, mouse Hsp70-2, andrat Hst70 genes, and sequences around the exon-intron boundaries in the mouse gene are conservedin the human and rat genes (Bonnycastle et al.,’94; Dix et al., ’96b).

These studies suggest that genes homologousto mouse Hsp70-2 are expressed in spermatoge-nic cells in diverse vertebrate species. It will beinteresting to learn if the homologous genes havesimilar patterns of developmental expression andif the proteins they encode have functional rolessimilar to HSP70-2. The difference between hu-mans and mice in the level of expression of HSPA2and HSP70-2 in somatic tissues suggests that thisgene may be evolving rapidly and may have some-what different functions in lower vertebrates andmammals.

LITERATURE CITEDAllen, D.J., B.V. Harmon, and J.F.R. Kerr (1987) Cell death

in spermatogenesis.In: Perspectives on Mammalian CellDeath. C.S. Potter, ed. Oxford University Press, Oxford, pp.229–258.

Allen, J.W., B.W. Collins, B.A. Merrick, C. He, J.K. Selkirk,P. Poorman-Allen, M.E. Dresser, and E.M. Eddy (1996)HSP70-2 is part of the synaptonemal complex in mouse andhamster spermatocytes. Chromosoma, 104:414–421.

Allen, R.L., D.A. O’Brien, and E.M. Eddy (1988a) A novelhsp70-like protein (P70) is present in mouse spermatoge-nic cells. Mol. Cell. Biol., 8:828–832.

Allen, R.L., D.A. O’Brien, C.C. Jones, D.L. Rockett, and E.M.Eddy (1988b) Expression of heat shock proteins by isolatedmouse spermatogenic cells. Mol. Cell. Biol., 8:3260–3266.

Baker, S.M., C.R. Bronner, L. Zhang, A.W. Plug, M. Robatzek,G. Warren, E.A. Elliott, J. Yu, T. Ashley, N. Arnheim, R.A.Flavell, and R.M. Liskay (1995) Male mice defective in theDNA mismatch repair gene PMS2 exhibit abnormal chro-mosome synapsis in meiosis. Cell, 82:309–319.

Baker, S.M., A.W. Plug, T.A. Prolla, C.R. Bronner, A.C. Har-ris, X. Yao, D.-M. Christie, C. Monell, N. Arnheim, A. Brad-ley, T. Ashley, and R.M. Liskay (1996) Involvement of mouseMlh1 in DNA mismatch repair and meiotic crossing over.Nature Genet., 14:336–342.

Bellvé, A.R., J.C. Cavicchia, C.F. Millette, D.A. O’Brien, Y.M.Bhatnagar, and M. Dym (1977) Spermatogenic cells of theprepubertal mouse: isolation and morphological character-ization. J. Cell Biol., 74:68–85.

Billig, H., I. Furuta, C. Rivier, J. Tapanainen, M. Parvinen,and A.J.W. Hsueh (1995) Apoptosis in testis germ cells: de-velopmental changes in gonadotropin dependence and lo-calization to selective tubule stages. Endocrinology, 136:5–12.

HSP70-2 OF SPERMATOGENIC CELLS 269

Bonnycastle, L.L.C., C.-E. Yu, C.R. Hunt, B.J. Trask, K.P.Clancy, J.L. Weber, D.P. Patterson, and G.D. Schellenberg(1994) Cloning, sequencing, and mapping of the human chro-mosome 14 heat shock protein gene (HSPA2). Genomics,23:85–93.

Boulanger, J., D. Faulds, E.M. Eddy, and C.A. Lingwood(1995) Members of the 70 kDa heat shock protein fam-ily specifically recognize sulfoglycolipids: role in gameterecognition and mycoplasma-related infertility. J. Cell.Physiol., 165:7–17.

Chapman, D.L., and D.J. Wolgemuth (1994) Regulation ofM-phase promoting factor activity during development ofmouse germ cells. Dev. Biol., 165:500–506.

Chappell, T.G., W.J. Welch, D.M. Schlossman, K.B. Palter,M.J. Schlesinger, and J.E. Rothman (1986) Uncoating AT-Pase is a member of the 70 kilodalton family of stress pro-teins. Cell, 45:3–13.

Chen, Q., R.E. Pearlman, and P.B. Moens (1992) Isolationand characterization of a cDNA encoding a synaptonemalcomplex protein. Biochem. Cell Biol., 70:1030–1038.

Chowdhury, A.K., and E. Steinberger (1964) A quantitativestudy of the effect of heat on germinal epithelium of rattestes. Am. J. Anat., 115:509–524.

Clarke, A.R., S. Gledhill, M.L. Hooper, C.C. Bird, and A.H.Wyllie (1994) p53-dependence of early apoptotic and prolif-erative responses within the mouse intestinal epitheliumfollowing gamma irradiation. Oncogene, 9:1767–1773.

Cobb, J., R.K. Reddy, C. Park, and M.A. Handel (1997) Analy-sis of expression and function of topoisomerase I and II dur-ing meiosis in male mice. Mol. Reprod. Dev., 46:489–498.

Creasy, D.M. (1997) Evaluation of testicular toxicity in safetyevaluation studies: the appropriate use of spermatogenicstaging. Toxicol. Pathol., 25:119–131.

Dix, D.J., J.W. Allen, B.W. Collins, C. Mori, N. Nakamura, P.Poorman-Allen, E.H. Goulding, and E.M. Eddy (1996a) Tar-geted gene disruption of Hsp70-2 results in failed meiosis,germ cell apoptosis, and male infertility. Proc. Natl. Acad.Sci. USA, 93:3264–3268.

Dix, D.J., M. Rosario-Herrle, H. Gotoh, C. Mori, E.H.Goulding, C.V. Barrett, and E.M. Eddy (1996b) Develop-mentally regulated expression of Hsp70-2 and a Hsp70-2/lacZ transgene during spermatogenesis. Dev. Biol., 174:310–321.

Dobson, M.J., R.E. Pearlman, A. Karaiskakis, B. Spyropoulos,and P.B. Moens (1994) Synaptonemal complex proteins: oc-currence, epitope mapping and chromosome disjunction. J.Cell Sci., 107:2749–2760.

Draetta, G., and D. Beach (1988) Activation of cdc2 proteinkinase during meiosis in human cells: cell cycle-dependentphosphorylation and subunit rearrangement. Cell, 54:17–26.

Draetta, G., H. Piwnica-Worms, D. Morrison, B. Druker, T.Roberts, and D. Beach (1988) Human cdc2 protein kinaseis a major cell-cycle regulated tyrosine kinase subunit. Na-ture, 336:738–744.

Eddy, E.M. (1995) Chauvinist genes of male germ cells: geneexpression during mouse spermatogenesis. Reprod. Fertil.Dev., 7:695–704.

Eddy, E.M., and D.A. O’Brien (1998) Gene expression duringmammalian meiosis. In: Meiosis and Gametogenesis. M.A.Handel, ed. Current Topics in Developmental Biology, Vol.37, Academic Press, San Diego, pp. 141–200.

Eddy, E.M., T.F. Washburn, D.O. Bunch, E.H. Goulding, B.C.Gladen, D.B. Lubahn, and K.S. Korach (1996) Targeted dis-

ruption of the estrogen receptor gene in male mice causesalteration of spermatogenesis and infertility. Endocrinology,137:4796–4805.

Edelmann, W., P.C. Cohen, M. Kane, K. Lau, B. Morrow, S.Bennett, A. Umar, T. Kunkel, G. Cattoretti, R. Chaganti,J.W. Pollard, R.D. Kolodner, and R. Kucherlapati (1996)Meiotic pachytene arrest in MLH1-deficient mice. Cell,85:1125–1134.

Frydman, J., E. Nimmesgern, H. Erdjument-Bromage, J.S.Wall, P. Tempst, and F.-U. Hartl (1992) Function in proteinfolding of TRiC, a cytosolic ring complex containing TCP-1and structurally related subunits. EMBO J., 11:4767–4778.

Giebel, L.B., B.P. Dworniczak, and E.K.F. Bautz (1988) De-velopmental regulation of a constitutively expressed mousemRNA encoding a 72-kDa heat shock-like protein. Dev. Biol.,125:200–207.

Goodson, M.L., O.-K. Park-Sarge, and K.D. Sarge (1995) Tis-sue-dependent expression of heat shock factor 2 isoformswith distinct transcriptional activities. Mol. Cell. Biol.,15:5288–5293.

Haaf, T., E.I. Golub, G. Reddy, C.M. Radding, and D.C. Ward(1995) Nuclear foci of mammalian Rad51 recombination pro-tein in somatic cells after DNA damage and its localizationin synaptonemal complexes. Proc. Natl. Acad. Sci. USA,92:2298–2302.

Hendrick, J.P., and F.-U. Hartl (1993) Molecular chaperonefunctions of heat-shock proteins. Annu. Rev. Biochem.,62:349–384.

Hendrick, J.P., and F.-U. Hartl (1995) The role of molecularchaperones in protein folding. FASEB J., 9:1559–1569.

Hunt, C., and R.I. Morimoto (1985) Conserved features ofeukaryotic HSP70 genes revealed by comparison with thenucleotide sequence of human HSP70. Proc. Natl. Acad. Sci.USA, 82:6455–6459.

Hunt, C.R., D.L. Gasser, D.D. Chaplin, J.C. Pierce, and C.A.Kozak (1993) Chromosomal localization of five murineHSP70 gene family members: Hsp70-1, Hsp70-2, Hsp70-3,Hsc70t, and Grp78. Genomics, 16:193–198.

Hynes, G., C.W. Sutton, S. U, and K.R. Willison (1996) Pep-tide mass fingerprinting of chaperonin-containing TCP-1(CCT) and copurifying proteins. FASEB J., 10:137–147.

Kandell, F.R., and R.S. Swerdloff (1988) Role of temperaturein regulation of spermatogenesis and the use of heating asa method for contraception. Fertil. Steril., 49:1–23.

Keegan, K.S., D.A. Holtzman, A.W. Plug, E.R. Christenson,E.E. Brainerd, G. Glaggs, N.J. Bentlye, E.M. Taylor, M.S.Meyn, S.B. Moss, A.M. Carr, and M.F. Hoekstra (1996) TheAtr and Atm proteins associate with different sites alongmeiotically pairing chromosomes. Genes Dev., 10:2423–2437.

Kovalenko, O.V., A.W. Plug, T. Haaf, D.K. Gonda, T. Ashley,D.C. Ward, C.M. Radding, and E.I. Golub (1996) Mamma-lian ubiquitin-conjugating enzyme Ubc9 interacts withRad51 recombination protein and localizes in synaptone-mal complexes. Proc. Natl. Acad. Sci. USA, 93:2958–2963.

Krawczyk, Z., N. Szymik, and J. Wisniewski (1987) Expres-sion of hsp70-related gene in developing and degeneratingrat testis. Mol. Biol. Reprod., 12:35–41.

Krawczyk, Z., P. Mali, and M. Parvinen (1988) Expression ofa testis-specific hsp70 gene-related RNA in defined stagesof rat seminiferous epithelium. J. Cell Biol., 107:1317–1323.

Krawczyk, Z., W. Schmid, P. Harkonen, and P. Wolnirzek(1992) The ERE-like sequence from the promoter region ofthe testis specific hsp70-related gene is not estrogen respon-sive. Cell Biol. Int. Rep., 16:937–948.

270 E.M. EDDY

Kubota, H., G. Hynes, A. Carne, A. Ashworth, and K. Willison(1994) Identification of six Tcp-1-related genes encoding di-vergent subunits of the TCP-1-containing chaperonin. Curr.Biol., 4:89–99.

Kubota, H., G. Hynes, and K. Willison (1995) The chaperonincontaining t-complex polypeptide 1 (TCP-1). Multisubunitmachinery assisting in protein folding and assembly in theeukaryotic cytosol. Eur. J. Biochem., 230:3–16.

Kubota, H., G. Hynes, S.M. Kerr, and K.R. Willison (1997)Tissue-specific subunit of the mouse cytosolic chaperonin-containing TCP-1. FEBS Lett., 402:53–56.

Kurtz, S., J. Rossi, L. Petko, and S. Lindquist (1986) An an-cient developmental induction: heat-shock proteins inducedin sporulation and oogenesis. Science, 231:1154–1156.

Law, H., O. Itkonnen, and C.A. Lingwood (1988) Sul-fogalactolipid binding protein SLIP 1: a conserved functionfor a conserved protein. J. Cell. Physiol., 137:462–468.

Lindquist, S. (1986) The heat-shock response. Ann. Rev.Biochem., 55:1151–1191.

Lingwood, C.A. (1985) Developmental regulation of galacto-glycerolipid and galactosphingolipid sulphation during mam-malian spermatogenesis. Biochem. J., 231:393–400.

Lingwood, C.A. (1986) Colocalization of sulfogalactosyl-acylalkylglycerol (SGG) and its binding protein during sper-matogenesis and sperm maturation: Topology of SGG definesa new testicular germ cell membrane domain. Biochem. CellBiol., 64:984–992.

Lingwood, C.A., and A. Nutikka (1991) Studies on the sper-matogenic sulfogalactolipid binding protein SLIP 1. J. Cell.Physiol., 146:258–263.

Lingwood, C.A., H.W. Kunz, and T.J.I. Gill (1985) Deficiencyin the regulation of testicular galactolipid sulphotransferasein rats carrying the growth and reproduction-complex (grc)gene. Biochem. J., 231:401–405.

Maekawa, M., D.A. O’Brien, R.L. Allen, and E.M. Eddy (1989)Heat-shock cognate protein (hsc71) and related proteins inmouse spermatogenic cells. Biol. Reprod., 40:843–852.

Massa, S.M., F.M. Longo, J. Zuo, S. Wang, J. Chen, and F.R.Sharp (1995) Cloning of rat grp75, an hsp70-family mem-ber, and its expression in normal and ischemic brain. J.Neurosci. Res., 40:807–819.

Matsumoto, M., and H. Fujimoto (1990) Cloning of a hsp70-related gene expressed in mouse spermatids. Biochem.Biophys. Res. Commun., 166:43–49.

Matsumoto, M., S.-I. Kurata, H. Fujimoto, and M. Hoshi(1993) Haploid specific activations of protamine 1 and hsc70tgenes in mouse spermatogenesis. Biochim. Biophys. Acta,1174:274–278.

Meuwissen, R.L.J., H.H. Offenberg, A.J.J. Dietrich, A.Riesewijk, M. Van Iersel, and C. Heyting (1992) A coiled-coil related protein specific for synapsed regions of meioticprophase chromosomes. EMBO J., 11:5091–5100.

Mizzen, L.A., C. Change, J. Garrels, and W.J. Welch (1989)Identification, characterization and purification of two mam-malian stress proteins present within mitochondria: one re-lated to hsp 70 and the other to the bacterial GroEL protein.J. Biol. Chem., 264:20664–20675.

Moens, P.B. (1994) Molecular perspectives of chromosomepairing at meiosis. Bioessays, 16:101–106.

Moens, P.B., and W.C. Earnshaw (1989) Anti-topoisomeraseII recognizes meiotic chromosome cores. Chromosoma,98:317–322.

Moens, P.B., and B. Spyropoulos (1995) Immunocytology ofchiasmata and chromosomal disjunction at mouse meiosis.Chromosoma, 104:175–182.

Morgan, D.O. (1995) Principles of CDK regulation. Nature,374:131–134.

Mori, C., N. Nakamura, D.J. Dix, M. Fujioka, S. Naka-gawa, K. Shiota, and E.M. Eddy (1997) Morphologicalanalysis of germ cell apoptosis during postnatal testisdevelopment in normal and Hsp70-2 knockout mice. Dev.Dyn., 208:125–136.

Morimoto, R.I., K.D. Sarge, and K. Abravaya (1992) Tran-scriptional regulation of heat shock genes. A paradigm forinducible genomic responses. J. Biol. Chem., 267:21987–21990.

Munro, S., and H.R.B. Pelham (1986) An Hsp70-like proteinin the ER: identity with the 78 kd glucose-regulated pro-tein and immunoglobulin heavy chain binding protein. Cell,46:291–300.

Murashov, A.K., and D.J. Wolgemuth (1996a) Distinct tran-scripts are recognized by sense and antisense riboprobesfor a member of the murine HSP70 gene family, HSP70.2,in various reproductive tissues. Mol. Reprod. Dev., 43:17–24.

Murashov, A.K., and D.J. Wolgemuth (1996b) Sense andantisense transcripts of the developmentally regulated mu-rine hsp70.2 gene are expressed in distinct and only par-tially overlapping areas in the brain. Mol. Brain Res.,37:85–95.

O’Brien, D.A. (1987) Stage-specific protein synthesis by iso-lated spermatogenic cells throughout meiosis and early sper-miogenesis in the mouse. Biol. Reprod., 37:147–157.

Olden, K., R.M. Pratt, C. Jaworski, and K.M. Yamada (1979)Evidence for role of glycoprotein carbohydrates in membranetransport: specific inhibition by tunicamycin. Proc. Natl.Acad. Sci. USA, 76:791–795.

Raab, L.S., K.L. Polakoski, L.W. Hancock, and D.W. Hamilton(1995) Characterization of the heat shock protein P70 inrat spermatogenic cells. Mol. Reprod. Dev. 40:186–195.

Rassow, J., O. von Ahsen, U. Bomer, and N. Pfanner (1997)Molecular chaperones: towards a characterization of theheat-shock protein 70 family. Trends Cell Biol., 7:129–133.

Resendez, E., J.W. Attenello, A. Grafsky, C.S. Chang, and A.S.Lee (1985) Calcium ionophore A23187 induces expressionof glucose-regulated genes and their heterologous fusiongenes. Mol. Cell. Biol., 5:1212–1219.

Rhee, K., and D.J. Wolgemuth (1995) Cdk family genes areexpressed not only in dividing but also in terminally differ-entiated mouse germ cells, suggesting their possible func-tion during both cell division and differentiation. Dev. Dyn.,204:406–420.

Rosario, M.O., S.L. Perkins, D.A. O’Brien, R.L. Allen, andE.M. Eddy (1992) Identification of the gene for the devel-opmentally expressed 70 kDa heat-shock protein (P70) ofmouse spermatogenic cells. Dev. Biol., 150:1–11.

Roux, A.-F., V.T.T. Nguyen, J.A. Squire, and D.W. Cox (1994)A heat shock gene at 14q22: Mapping and expression. Hum.Mol. Genet., 3:1819–1822.

Sarge, K.D. (1995) Male germ cell-specific alteration in tem-perature set point of the cellular stress response. J. Biol.Chem., 270:18745–18748.

Sarge, K.D., and K.E. Cullen (1997) Regulation of hsp ex-pression during rodent spermatogenesis. Cell. Mol. Biol.Res., 53:191–197.

Sarge, K.D., O.-K. Park-Sarge, J.D. Kirby, K.E. Mayo, andR.I. Morimoto (1994) Expression of heat shock factor 2 inmouse testis: potential role as a regulator of heat-shock pro-tein gene expression during spermatogenesis. Biol. Reprod.,50:334–1343.

HSP70-2 OF SPERMATOGENIC CELLS 271

Schwartz, D., N. Goldfinger, and V. Rotter (1993) Expressionof p53 protein in spermatogenesis is confined to the tetrap-loid pachytene primary spermatocytes. Oncogene, 8:1487–1494.

Scully, R., J. Chen, A. Plug, Y. Ziao, D. Weaver, J. Feunteun,T. Ashley, and D.M. Livingston (1997) Association of BRCA1and Rad51 in mitotic and meiotic cells. Cell, 88:265–275.

Shiu, R.P.C., J. Pouyssegur, and I. Pastan (1977) Glucosedepletion accounts for the induction of two tranformation-sensitive membrane proteins in Rous sarcoma virus-trans-formed chick embryo fibroblasts. Proc. Natl. Acad. Sci. USA,74:3840–3844.

Silver, L.M., and M. White (1982) A gene product of the mouset complex with chemical properties of a cell surface-associ-ated component of the extracellular matrix. Dev. Biol.,91:423–430.

Solomon, M.J. (1994) The function(s) of CDK, the p34cdc2-ac-tivating kinase. Trends Biochem. Sci., 19:498–500.

Tanphaichitr, N., A. Tayabali, C. Gradil, S. Juneja, M.C.Léveillé, and C.A. Lingwood (1992) Role of a germ cell-spe-cific sulfolipid-immobilizing protein (SLIP1) in mouse in vivofertilization. Mol. Reprod. Dev., 32:17–22.

Tanphaichitr, N., J. Smith, S. Mongkolsirikieart, C. Gradil,and C.A. Lingwood (1993) Role of a gamete-specific sulfo-glycolipid immobilizing protein on mouse sperm-egg bind-ing. Dev. Biol. 156:164–175.

Wei, Y., Y. Kariya, H. Fukata, K. Teshigawara, and A. Uchida(1994) Induction of apoptosis by quercetin: involvement ofheat shock protein. Cancer Res., 54:4952–4957.

Wei, Y., X. Zhao, Y. Kariya, K. Teshigawara, and A. Uchida(1995) Inhibition of proliferation and induction of apoptosisby abrogation of heat-shock protein (HSP) 70 expression intumor cells. Cancer Immunol. Immunother., 40:73–78.

Welch, W.J., and J.R. Feramisco (1985) Rapid purification ofmammalian 70,000-dalton stress proteins: Affinity of theproteins for nucleotides. Mol. Cell. Biol., 5:1229–1237.

Welch, W.J., J.I. Garrels, P.G. Thomas, J.J.C. Lin, and J.R.Feramisco (1983) Biochemical characterization of the mam-malian stress proteins and identification of two stress pro-teins as glucose- and Ca2+-ionophore-regulated proteins. J.Biol. Chem., 258:7102–7111.

Widlak, W., M. Markkula, Z. Krawczyk, K. Kananen, andI. Huhtaniemi (1995) A 252 bp upstream region of therat spermatocyte-specific hst70 gene is sufficient to pro-mote expression of the hst70-CAT hybrid gene in testisand brain of transgenic mice. Biochim. Biophys. Acta.,1264:191–200.

Wisniewski, J., T. Kordula, and Z. Krawczyk (1990) Isolationand nucleotide sequence analysis of the rat testis-specificmajor heat-shock protein (HSP70)-related gene. Biochim.Biophys. Acta., 1048:93–99.

Zakeri, Z.F., and D.J. Wolgemuth (1987) Developmental-stage-specific expression of the hsp70 gene family during differ-entiation of the mammalian male germ line. Mol. Cell. Biol.,7:1791–1796.

Zakeri, Z.F., D.J. Wolgemuth, and C.R. Hunt (1988) Identifi-cation and sequence analysis of a new member of the mouseHSP70 gene family and characterization of its unique cel-lular and developmental pattern of expression in the malegerm line. Mol. Cell. Biol., 8:2925–2932.

Zakeri, Z.F., W.J. Welch, and D.J. Wolgemuth (1990) Charac-terization and inducibility of hsp 70 proteins in the malemouse germ line. J. Cell Biol., 111:1785–1792.

Zhu, D., D.J. Dix, and E.M. Eddy (1997) HSP70-2 is requiredfor CDC2 kinase activity in meiosis I of mouse spermato-cytes. Development (in press).