The EMBO Journal Vol.19 No.6 pp.1335–1345, 2000 Perinatal synthetic lethality and hematopoietic defects in compound mafG::mafK mutant mice Ko Onodera, Jordan A.Shavit, to this element, referred to as nuclear factor-erythroid 2 Hozumi Motohashi1, Masayuki Yamamoto1,2 (NF-E2), was shown to be distinct from AP-1 through and James Douglas Engel2 analysis of the erythroid-specific porphobilinogen deamin- ase gene promoter (Mignotte et al., 1989). Over the past Department of Biochemistry, Molecular Biology and Cell Biology, decade, this same sequence motif has been identified Northwestern University, Evanston, IL 60208-3500, USA and within cis regulatory elements of numerous erythroid 1Department of Molecular and Developmental Biology, Center for TARA and Institute of Basic Medical Sciences, University of Tsukuba, genes. Despite the accumulated wealth of information Tsukuba 305-8577, Japan demonstrating the importance of this binding site in erythroid gene regulation, there has been no formal demon- 2Corresponding authors e-mail: [email protected] stration of which transcription factor actually elicits responses from this cis-element in erythroid cells. K.Onodera, J.A.Shavit and H.Motohashi contributed equally to this work After its purification and cloning by reverse genetics, NF-E2 was found to be a heterodimeric basic region plus Prior studies exploring the mechanisms controlling leucine zipper transcription factor. The larger subunit of erythroid gene regulation implicated MARE (Maf this complex, called p45 (Andrews et al., 1993a), displayed recognition element) cis-elements as crucial to the all the hallmarks of a hematopoietic cell-restricted tran- transcriptional activity of many erythroid genes. scription factor, while the smaller subunit, called p18 or Numerous transcription factors can elicit responses MafK (Andrews et al., 1993b; Igarashi et al., 1994), through MAREs, including not only the AP-1 family consisted of little more than an amphipathic dimerization proteins, but also a growing list of factors composed of motif and a positively charged domain. MafK shared Cap-N-Collar (CNC)–small Maf heterodimers. While greatest sequence homology with the family of proteins these factors can activate transcription from MAREs related to the chicken v-Maf oncoprotein (Nishizawa et al., in co-transfection assays, mouse germline mutations in 1989), while the p45 subunit was most closely related to cnc genes tested to date have failed to reveal primary the Drosophila nuclear regulatory Cap-N-Collar (CNC) erythroid phenotypes. Here we report that after com- protein (Mohler et al., 1991). bining the mafK and mafG targeted null alleles, mutant Since those early observations, the number of proteins animals display several synthetic phenotypes, including has expanded considerably that can form (either productive erythroid deficiencies. First, compound homozygous or unproductive) homo- or heterodimers, or that can small maf gene mutants survive embryogenesis, but specifically bind to the extended AP-1 sequence motif die postnatally. Secondly, compound mutant animals (called a MARE, for Maf recognition element; Kataoka develop severe neurological disorders. Thirdly, they et al., 1994b; Motohashi et al., 1997) present in the exhibit an exacerbated mafG deficiency in megakaryo- regulatory sequences of most erythroid as well as many poiesis, specifically in proplatelet formation, resulting non-erythroid genes. This group of MARE binding factors in profound thrombocytopenia. Finally, the compound now includes four large Maf proteins (Swaroop et al., mutant animals develop severe anemia accompanied 1992; Kataoka et al., 1993, 1994a; Ogino and Yasuda, by abnormal erythrocyte morphology and membrane 1998), six CNC family members (Chan et al., 1993, 1996; protein composition. These data provide direct evidence Itoh et al., 1995; Johnsen et al., 1996; Oyake et al., 1996; that the small Maf transcription factors play an import- Kobayashi et al., 1999) and three small Maf proteins ant regulatory role in erythropoiesis. (Fujiwara et al., 1993; Kataoka et al., 1995), in addition Keywords: anemia/cytoskeleton/platelet/small Maf/ to all of the previously identified members of the AP-1 spherocytosis (Jun/Fos) transcription factor families. Thus, the complex- ity of regulatory responses that can be controlled through the MARE element is vast: the factors capable of binding Introduction to these sites are pervasive, and the consequences of their binding have been shown to elicit transcriptional responses The chicken ε/β-globin gene enhancer was the first distant ranging from activation to repression (Engel, 1994; transcriptional control element identified that was proven Kataoka et al., 1995; Motohashi et al., 1997). to be crucial for tissue-specific control over erythroid gene Three of the six currently known cnc family members regulation (Choi and Engel, 1986; Hesse et al., 1986). have been examined by germline mutagenesis. p45 muta- Analysis of enhancer activity by transfection of clustered tion conferred permanently impaired platelet formation as array mutants into erythroid cells showed that one particu- well as transient neonatal erythrocyte abnormalities, which lar DNA sequence motif, specifying an AP-1 binding site, were proposed to be an indirect consequence of thrombo- conferred the greatest contribution to enhancer activity cytopenia since adult mutant animals had normal (Reitman and Felsenfeld, 1988). The protein that bound erythropoiesis (Shivdasani and Orkin, 1995; Shivdasani © European Molecular Biology Organization 1335 K.Onodera et al. et al., 1995). Nrf1 mutant mice had defective definitive hematopoiesis, although it was non-cell autonomous (Farmer et al., 1997), while Nrf2 mutants displayed no erythroid phenotype (Itoh et al., 1997; Kuroha et al., 1998). Thus, no clear relationship between CNC family member–small Maf heterodimers and hematopoiesis had been established. The ominous question naturally arose as to whether or not the bona fide regulatory protein(s) that activates transcription from erythroid MAREs was among this group of transcription factors. We recently embarked on an analogous strategic approach to this same question, initiating experiments to determine whether or not any of the heterodimeric partners of the CNC proteins, the small Maf transcription factors (MafF, MafG and MafK), exhibited erythroid phenotypes Fig. 1. Small maf mutant detection strategy. PCR screening of the mafG and mafK loci. Owing to the high similarity between the mafG after germline gene targeted ablation. We (Shavit et al., and mafK loci, both are represented on a single diagram. (A)A 1998) and others (Kotkow and Orkin, 1996) reported common 5Ј primer was used with two distinct 3Ј primers, one that targeted disruption of mafK led to no discernible corresponding to sequences within the second intron of the wild-type phenotype, and very recently we also found that mafF allele (top) and a 3Ј lacZ primer for the mutant allele (bottom). germline ablation similarly caused no apparent disturbance (B) Two percent agarose gel electrophoresis showing a typical distribution of wild-type, heterozygous and homozygous mutant in embryonic or adult development (Onodera et al., animals from a compound heterozygous mutant intercross. 1999). However, mafG germline mutation led to mild thrombocytopenia, weakly phenocopying the p45 CNC mutation (Shavit et al., 1998). Once again, the small maf never survive postnatally. These experiments show for the mutant animals displayed no erythroid deficiencies. first time an in vivo requirement for the small Maf family In order to characterize pathologies engendered by small proteins in erythropoiesis, and they thus provide a key maf gene mutations more fully, we wished to determine link in identifying trans-acting factor(s) that acts at whether mafG mutant phenotypes were exacerbated by this demonstrably crucial cis regulatory transcriptional loss of other small maf alleles. If more profound deficien- control element. cies were encountered in the compound mutants than in either single gene homozygous mutant animal, this would Results constitute strong prima-facie evidence for interallelic com- plementation among this family of factors. To this end, Synthetic perinatal lethality as a consequence of we intercrossed the small maf germline mutant animals to combining homozygous mafK and mafG mutant generate compound mutants. alleles The results of the compound mutant loss-of-function We first intercrossed mafGϩ/– with mafK–/– mutant mice analyses shed significant new insight into the cellular (129/CD1 mixed hybrid background) to generate mafGϩ/– processes controlled by small Maf–CNC heterodimer ::mafKϩ/– compound mutants; the compound hetero- activity. First, compound homozygous maf mutant animals zygotes were recovered at the expected Mendelian frequen- survive gestation. Animals missing both the mafF and cies. Since those animals exhibited no apparent mafK genes suffer virtually no abnormalities, while com- dysfunction, we intercrossed them and genotyped progeny pound mutants missing both mafF and mafG differ only between postnatal day 7 (P7) and P14. Genomic tail DNA slightly from those missing mafG alone (our unpublished was analyzed by PCR using common 5Ј mafG- and mafK- observations). However, unlike any of the individual small specific as well as separate 3Ј primers to distinguish maf mutant or the p45 mutant animals, mafK::mafG between the wild-type and mutant alleles (Figure 1). Of compound homozygous mutants succumb immediately 279 pups,