Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Growth retardation and early death of β‐1, 4‐galactosyltransferase knockout mice with augmented proliferation and abnormal differentiation of epithelial cells

Growth retardation and early death of β‐1, 4‐galactosyltransferase knockout mice with augmented... The EMBO Journal Vol.16 No.8 pp.1850–1857, 1997 Growth retardation and early death of β-1,4-galactosyltransferase knockout mice with augmented proliferation and abnormal differentiation of epithelial cells of dolichophosphate synthesis, inhibits blastocyst form- Masahide Asano, Kiyoshi Furukawa , 1 2 ation of mouse embryos, suggesting that cell surface Masahiro Kido , Satoshi Matsumoto , 2 3 carbohydrates play an important role in early embryo- Yoshinori Umesaki , Naohisa Kochibe and 4 genesis (Iwakura and Nozaki, 1985). Several lines of Yoichiro Iwakura evidence suggest that galactose (Gal)-containing complex Laboratory Animal Research Center, Institute of Medical Science, N-glycans are particularly important in these processes. University of Tokyo, Minato-ku, Tokyo 108, Department of Biosignal For example, stage-specific embryonic antigen-1 (SSEA-1) Research, Tokyo Metropolitan Institute of Gerontology, Itabashi-ku, (Solter and Knowles, 1978), which contains poly-N- Tokyo 173, Yakult Central Institute for Microbiological Research, 3 acetyllactosamine structure and is expressed specifically on Kunitachi, Tokyo 186 and Faculty of Education, Gunma University, Maebashi 371, Japan pre-implantation embryos and undifferentiated embryonal carcinoma cells, has been suggested to be involved in Corresponding author cellular interactions during morula compaction and implantation, since oligosaccharides containing this anti- Carbohydrate chains on a glycoprotein are important genic structure can competitively inhibit these processes not only for protein conformation, transport and (Fenderson et al., 1984). The observation that treatment stability, but also for cell–cell and cell–matrix inter- of early embryos with endo-β-galactosidase (Rastan et al., actions. UDP-Gal:N-acetylglucosamine β-1,4-galacto- 1985) or UDP-Gal (Shur et al., 1979) inhibited compaction syltransferase (GalT) (EC 2.4.1.38) is the enzyme which also suggests the importance of Gal-containing molecules transfers galactose (Gal) to the terminal N-acetyl- in these processes. glucosamine (GlcNAc) of complex-type N-glycans in The involvement of Gal-containing carbohydrates in the Golgi apparatus. In addition, it has also been other processes has also been suggested. Lower galactosyl- suggested that this enzyme is involved directly in cell– ation of serum IgGs is characteristic of patients with cell interactions during fertilization and early embryo- rheumatoid arthritis, suggesting involvement in the patho- genesis through a subpopulation of this enzyme dis- genesis of this disease (Parekh et al., 1985; Rademacher tributed on the cell surface. In this study, GalT-deficient x a et al., 1988). Sialyl Le and sialyl Le , which contain mice were produced by gene targeting in order to NeuAcα2→3Galβ1→4[Fucα1→3]GlcNAc and Neu- examine the pathological effects of the deficiency. GalT- Acα2→3Galβ1→3[Fucα1→4]GlcNAc, respectively, are deficient mice were born normally and were fertile, but suggested to be ligands for selectins which mediate they exhibited growth retardation and semi-lethality. interactions between leukocytes and vascular endothelial Epithelial cell proliferation of the skin and small and platelet cells (reviewed in Lasky, 1992). Recently, intestine was enhanced, and cell differentiation in galectins, which particularly recognize β-galactoside sugar intestinal villi was abnormal. These observations sug- chains, were shown to be important for immune recogni- gest that GalT plays critical roles in the regulation of tion (Perillo et al., 1995; Yang et al., 1996). Furthermore, proliferation and differentiation of epithelial cells after a b x some Gal-containing carbohydrates such as Le ,Le ,Le , birth, although this enzyme is dispensable during embryonic development. Le antigens and their derivatives were shown to be Keywords: epithelial cell differentiation/epithelial cell associated with tumors (reviewed in Hakomori, 1989). It growth/galactosyltransferase/glycosyltransferase/ was shown that the expression of these antigens and knockout mouse tumor progression were closely correlated, suggesting that aberrant glycosylation at the cell surface perturbs normal cell–cell and cell–matrix interactions which results in uncontrolled cell growth, invagination and metastasis of Introduction tumor cells (reviewed in Hakomori, 1991). However, so far, the pathological effects of the deficiency in Gal- N-linked sugar chains are associated with most proteins containing carbohydrates on an animal and its embryonic found on the mammalian cell surface and in secretions. development have not been examined. Oligosaccharides on secreted glycoproteins are shown to UDP-Gal:N-acetylglucosamine β-1,4-galactosyltrans- be important for their biological activities, transport, ferase (GalT) is the only enzyme which is known to stability and clearance from the circulation, while oligo- transfer Gal from UDP-Gal to terminal N-acetylglucos- saccharides on cell surface glycoproteins are suggested to amine (GlcNAc) of complex N-glycans to form a be involved in various cellular functions including cell– cell and cell–matrix interactions during embryogenesis, Galβ1→4GlcNAc structure in the Golgi apparatus. In the immune reactions and tumor development (reviewed in mammary gland, GalT is known to synthesize lactose Iwakura, 1989; Hakomori, 1991; Furukawa and Kobata, by binding to α-lactalbumin (α-lac) which modifies the 1992; Varki, 1993). acceptor specificity of the transferase from GlcNAc to Previously, we showed that tunicamycin, an inhibitor glucose (reviewed in Kuhn, 1983). In addition to a 1850 © Oxford University Press Effects of β-1,4-galactosyltransferase deficiency biosynthetic role for GalT, it was suggested that a sub- population of this enzyme is present at the cell surface and is involved directly in cellular interactions by binding to terminal GlcNAc of glycoconjugates at the opposing cell surface and matrix (reviewed in Shur, 1993). In this context, it was suggested that this enzyme is involved directly in mouse gamete interaction (Miller et al., 1992), neurite extension (Begovac and Shur, 1990), mesenchymal and neural crest cell migration (Runyan et al., 1986) and late morula compaction (Bayna et al., 1988). In the present study, we generated mice deficient in the GalT gene by homologous recombination in order to examine the pathological effects of the deficiency in vivo. We found that homozygous GalT-deficient mice can be obtained at almost the expected ratio in crosses between heterozygous mice. However, approximately half of them died before weaning, with growth retardation, indicating the importance of GalT for survival after birth. Results –/– Generation of GalT-deficient (GalT ) mice The targeting vector was constructed by replacing the GalT exon 1 with a PGKneobpA cassette (Soriano et al., 1991), in order to disrupt both the long and short forms of the GalT gene (Shaper et al., 1988; Hollis et al., 1989). In this construct, the translation initiation sites for both forms and the Golgi retention signal (Nilsson et al., 1991) in exon 1 were deleted (Figure 1A). The targeting vector was electroporated into embryonic stem (ES) cells and G418-resistant colonies were picked up. Twelve positive clones were obtained out of 180 Neo clones by PCR, and the homologous recombination was confirmed in nine clones by Southern blot analysis using an external 3 probe (Figure 1B) and 5 probe (data not shown). Chimeric Fig. 1. Targeted disruption of the GalT gene by homologous mice were produced using three targeted ES cell clones recombination. (A) Targeting strategy. Coding and non-coding exons and the targeted allele was transmitted through a germ of the GalT gene are shown by closed and open boxes, and the –/– line in two independent ES cell clones. GalT mice PGKneobpA cassette and the diphtheria toxin A fragment gene /– were generated by inter-crossing of GalT mice. The cassette (DT-A) by dotted and hatched boxes, respectively. PCR primers (P1 and P2) used for screening are shown by arrowheads. phenotypes described here were identical in the two mouse –/– 5-Probe and 3-probe show the positions of external probes used for lines. GalT mRNA was not detected in GalT mice, Southern blot analysis and the expected HindIII fragments are indicating disruption of this gene (Figure 1C). indicated by arrows. BH, BamHI; Hd, HindIII; RI, EcoRI; Sa, SacII; No, NotI. (B) Southern blot analysis of ES cell clones. Genomic DNA GalT activity and N-linked carbohydrate chains in (10 μg) from parent ES cells (R1) and targeted ES cell clones were –/– digested with HindIII and hybridized with the 3 probe. The expected GalT mice DNA fragments for the mutant allele (Mt) and wild-type allele (Wt) When GalT activity was determined using the liver homo- –/– are indicated. (C) Northern blot analysis of GalT mice. Poly(A) genates as an enzyme source and asialo (As)- and agalacto RNA (1 μg) from the liver was hybridized with exon 1 and 2 probes (Ag)-transferrin as an acceptor, the transferase activity in of the GalT gene, and rehybridized with the β-actin probe. /– / GalT mice was just half of that in GalT mice, while –/– / that in GalT mice was 5% of that in GalT mice (Figure 2A). However, it should be pointed out that this position of the GlcNAc of AsAg-transferrin (data not –/– level in GalT mice was significantly higher than in the shown). control without enzyme. A similar result was obtained In order to determine whether or not the Galβ1→ using the spleen homogenates as an enzyme source (data 4GlcNAc structure is expressed in N-linked sugar chains –/– not shown). In order to verify the galactosyltransferase of GalT mouse glycoproteins, serum glycoproteins were activity found in this assay, the galactosylated product analyzed by lectin blotting. When filters were stained with isolated by Bio-Gel P-6 column chromatography was Coomassie brilliant blue (CBB), no significant difference digested with diplococcal β-galactosidase which cleaves was detected between the three genotypes (Figure 2B, the Galβ1→4GlcNAc linkage but not the Galβ1→ CBB). When the filters were incubated with Ricinus 3GlcNAc linkage (Paulson et al., 1978). When the digested communis agglutinin (RCA)-I, which binds to oligosacch- material was analyzed by Bio-Gel P-4 chromatography, arides which terminate with Galβ1→4GlcNAc/Glc groups the radioactive product was eluted at the same position as (Baenziger and Fiete, 1979), RCA-I binding was observed / /– galactose, indicating that Gal was transferred at the C-4 for most protein bands in GalT and GalT mice, 1851 M.Asano et al. / /– –/– Fig. 2. (A) β-1,4-galactosyltransferase activity in liver homogenates from 10-week-old GalT (d), GalT (m) and GalT (j) mice. (B) Lectin / /– –/– blot analysis of serum glycoproteins from 10-week-old GalT (/), GalT (/–) and GalT (–/–) mice. The bars indicate the positions of molecular weight markers. CBB indicates filters stained with Coomassie brilliant blue to detect protein bands. RCA-I and PVL indicate filters / /– –/– incubated with RCA-I and PVL, respectively. (C) Lactose concentration in milk of five GalT (/), five GalT (/–) and two GalT (–/–) mice. The values indicate averages with standard deviations. –/– –/– while no binding was observed in a GalT mouse, with No lactose was detected in GalT mouse milk, while /– the exception of a few weak bands (Figure 2B, RCA-I). its concentration in the milk of GalT mice was about RCA-I binding was completely lost upon digestion of half as much as that of wild-type mice (Figure 2C). These glycoprotein samples with diplococcal β-galactosidase or results confirm that GalT is a component of lactose –/– N-glycanase prior to the analysis (data not shown), show- synthetase (Kuhn, 1983). As a result, these GalT mothers ing that these bands actually represented N-glycans con- could not rear their offspring (data not shown). taining β-linked Gal. In contrast, when filters were incubated with Psathyrella velutina lectin (PVL), which –/– Semi-lethality of GalT mice binds to oligosaccharides which terminate with β-linked –/– GalT mice were born healthy at almost the expected GlcNAc (Endo et al., 1992), PVL did not bind to most / /– ratio of 20% in heterozygous mating, suggesting normal protein bands in GalT and GalT mice, but did so in –/– –/– embryonic development, although GalT mice were a GalT mouse (Figure 2B, PVL). No PVL binding was /– / slightly smaller than GalT and GalT mice at birth observed upon digestion of glycoprotein samples with –/– (body weight of GalT mice was 80–90% of that of jack bean β-N-acetylhexosaminidase or N-glycanase prior control littermates at birth). Most of them, however, to the analysis (data not shown), showing that these developed severe skin lesions within a few days after bands actually represented N-glycans containing β-linked birth (Figure 3A). Horny cells continued to come off from GlcNAc. These results indicate that N-linked sugar chains –/– the flaky skin for ~2 weeks, and the skin lesions eventually of most serum glycoproteins from GalT mice are not –/– healed. GalT mice also showed growth retardation and galactosylated and terminated with β-GlcNAc, although a nearly half of them died before 4 weeks of age (Figure few protein bands were RCA-I positive and contained –/– 3B). The body weight of GalT survivors was only 60– β-1,4-linked Gal in their N-linked sugar chains. In order 70% of that of GalT mice at 4 weeks old, while to rule out the possibility that the RCA-I-positive glyco- /– –/– GalT mice were normal. After weaning, GalT mice proteins were produced by bacteria in the gut, we analyzed –/– gained weight normally, but they still continued to die, proteins from intestinal bacteria in GalT mice and found and only 20% of them survived beyond 16 weeks (Figure that these bacteria have no RCA-I-reactive glycoproteins –/– upon lectin binding (data not shown). In addition, no 3B). Both mature female and male GalT mice were –/– change has been detected in the RCA-I-reactive glyco- fertile (Table I). Sperm from GalT males was able to –/– protein pattern on lectin blotting after treatment of these fertilize. Although GalT females appeared to be able to mice with antibiotics (data not shown). ovulate, and maintain pregnancy to term, the litter size 1852 Effects of β-1,4-galactosyltransferase deficiency Fig. 4. Histological analysis (HE staining) of the skin of 10-day-old / –/– pups. (A and C) A GalT mouse; (B and D) a GalT mouse. Note that the prickle cell layer and granular layer are much thicker in the –/– GalT mouse than in the normal mouse. Ep: epidermis. Original magnification: 100 (A and B), 400 (C and D). cell layer and granular layer was observed, and the horny –/– layer was also thickened in GalT mice (Figure 4C and D). Many keratohyalin granules were deposited not only in the granular layer but also in the prickle cell layer in –/– GalT mice (Figure 4C and D). These histological observations indicate that acanthosis and hyperkeratosis –/– occurred in the skin of GalT mice. This result suggests that proliferation of epithelial stem cells located in the –/– basal layer of the epidermis is augmented in GalT mice. –/– In the small intestine, the crypts of GalT mice were found to be enlarged 3–4 times compared with those of –/– –/– Fig. 3. Semi-lethality in GalT mice. (A) A 10-day-old GalT pup wild-type (Figure 5A and B), indicating that proliferation –/– (left) with a control littermate (right). Note that the GalT pup is of epithelial stem cells located in the crypt is also smaller than the normal littermate and the hair is lost from the back –/– –/– with the presence of flaky skin. (B) Survival ratio of GalT mice enhanced in GalT mice. RCA-I staining showed that after birth. The genotype ratios at birth in heterozygous mating were Galβ1→4GlcNAc-containing sugar chains were not 20% (–/–), 49% (/–) and 31% (/) from 59 pups analyzed. The –/– expressed in the crypts and villi of GalT mice, while number of mice at birth is taken as 100% and the ratios of survivors the surface of crypt and villus cells was stained strongly in 30 mice are shown. /– with RCA-I in GalT mice (Figure 5C and D). The number of metaphase cells was counted after vincristine injection to confirm the augmented proliferation of epi- Table I. Litter sizes of –/–/– matings thelial cells. The numbers of metaphase cells per crypt per 90 min were 21.7  5.6 (–/–) versus 5.5  1.4 (/–) in Femalemale Litter size (No. of litters examined) the duodenum and 6.9  1.8 (–/–) versus 2.0  1.1 (/–) in the ileum, indicating that the mitotic indices of the –/–/– 5.1  2.1 (24) /––/– 6.2  2.3 (51) small intestine crypt cells were increased ~4 times in // 7.2  2.6 (53) –/– /– GalT mice as compared with those of GalT mice. These results suggest that the Galβ1→4GlcNAc structure P0.001. may be involved in growth control of epithelial cells in P0.05. the small intestine and skin. However, these abnormalities –/– were less prominent in adult GalT mice. was slightly smaller compared with that of the wild-type Abnormal differentiation of small intestinal villus control (Table I). cells Interestingly, differentiation of small intestine villus cells –/– Enhanced proliferation of epithelial cells in the was also found to be abnormal in GalT mice using skin and small intestine disaccharidases as markers. Although lactase was Histological examination revealed that the epidermis was expressed strongly over the entire surface of villus epi- –/– /– thickened in the skin lesions of GalT mice (Figure 4A thelial cells of 7-day-old GalT mice (Figure 6A), lactase –/– and B). At higher magnification, hyperplasia of the prickle localization was restricted to the top of the villi in GalT 1853 M.Asano et al. Fig. 5. Histological analysis of the small intestine of pups. Fig. 6. Analysis of small intestine disaccharidases. (A and B) HE staining; (C and D) RCA-I staining. (A) A 10-day-old Immunohistochemical localization of lactase (A and D), maltase (B / –/– GalT mouse; (B) a 10-day-old GalT mouse; (C) a 7-day-old and E) and the sucrase–isomaltase complex (C and F)inthe /– –/– GalT mouse; (D) a 7-day-old GalT mouse. Note that the crypt /– –/– duodenum of GalT (A, B and C) and GalT (D, E and F) mice. –/– size is much larger in the GalT mouse than in the control mouse –/– Note that the expression of lactase is much weaker in the GalT and RCA-I-reactive sugar chains are not expressed in crypt and villus mouse compared with the normal mouse, whereas maltase and sucrase –/– cells of the GalT mouse. Cr: crypt. Original magnification: –/– expression is stronger in the GalT mouse. Original magnification: 400 (A and B), 200 (C and D). Table II. Disaccharidase activities (nmol/mg protein/min) in 7-day-old mice (Figure 6D). Additionally, maltase and the sucrase– mice –/– isomaltase complex appeared precociously in GalT mice (Figure 6E and F). These enzymes had not yet been /– –/– /– expressed on day 7 in GalT mice (Figure 6B and C). Lactase 2.80, 2.27 0.86, 0.78 Since no RCA-I binding was detected in villus cells of Maltase 1.89, 1.17 9.62, 5.56 –/– GalT mice (Figure 5D), it is suggested that the lack of Isomaltase 0.00, 0.00 0.35, 0.76 Galβ1→4GlcNAc structure is responsible for the abnormal differentiation of villus cells. Consistent with the immuno- –/– histochemical findings, lactase activity in GalT small intestine homogenates was markedly low, while maltase galactosyltransferase in porcine trachea was suggested –/– and isomaltase activities were higher in GalT mice previously (Sheares and Carlson, 1984). However, we /– compared with GalT mice (Table II). Sucrase activity cannot rule out the possibilities that galactosyltransferases, was barely detectable in both types of mice at day 7, and which are involved in glycolipid or glycosaminoglycan –/– was still not detected in GalT mice by day 16, in synthesis, or other linkage specificity, might be responsible /– contrast to GalT mice in which the activity was detected for this activity. We are now trying to discriminate between at day 16 (data not shown). these possibilities. Homozygous GalT-deficient mice were obtained at almost the expected ratio from crosses between hetero- Discussion –/– zygous mice, suggesting that the development of GalT In this study, we have generated GalT-deficient mice and embryos is normal. This is in clear contrast to N-acetyl- shown that 95% of the galactosyltransferase activity glucosaminyltransferase (GlcNAcT)-I knockout mice was inactivated and that most Galβ1→4GlcNAc residues which die on day 10.5 of embryonic development due to on glycoprotein carbohydrate chains were absent in these defects in neural tube formation and vascularization (Ioffe mice. The origin of the low residual galactosyltransferase and Stanley, 1994; Metzler et al., 1994). Furthermore, we –/– activity is not clear. Since the translation initiation sites found that both male and female GalT mice were fertile. and Golgi retention signal were deleted in the targeting The litter size from homozygous mothers, however, was construct, the galactosyltransferase activity of the GalT slightly smaller compared with normal mice, suggesting gene is considered to be lost. It is unlikely that GalT a possible deficiency in uterus function. These observations activity and these Gal-containing glycoproteins are derived suggest that neither the carbohydrate structures containing –/– from colostrum because they were detected in GalT Gal nor GalT itself are absolutely necessary for cell–cell mice even at 10 weeks of age. One possibility is that this interactions during fertilization and embryonic develop- residual activity could derive from previously unknown ment, although previous studies suggested the involvement galactosyltransferases encoded by genes distinct from of Gal-containing oligosaccharides and/or GalT in these GalT. In this context, the presence of another β-1,4- processes (Eggens et al., 1989; Shur, 1993). However, 1854 Effects of β-1,4-galactosyltransferase deficiency besides GalT, two other proteins, sp56 (Cheng et al., case, our results indicate the importance of GalT in the 1994) and a 95 kDa peptide (Leyton et al., 1992), are regulation of cellular growth and differentiation. –/– known as zona-binding proteins during fertilization, and Lactose was not detected in the milk of GalT mothers –/– E-cadhelin/uvomorulin is also suggested to play an import- as was also the case in the milk of α-lac mothers (Stinnakre et al., 1994; Stacey et al., 1995). This is ant role in morula compaction (Vestweber and Kemler, because GalT is an essential component of lactose synthet- 1985). Thus, it is possible that these molecules might ase together with α-lactalbumin (Kuhn, 1983). The milk have substituted for the function of GalT in these processes. –/– of α-lac mothers has been reported to be too viscous to It is also possible that embryonic Gal-containing carbo- be sucked by pups because lactose regulates the osmotic hydrates such as SSEA-1 are synthesized by enzymes pressure and volume of the milk (Stinnakre et al., 1994; other than GalT or that enzymes which are usually not Stacey et al., 1995). This is probably the reason why involved in the synthesis of this structure substitute for –/– GalT mothers could not rear their offspring. GalT and allow development of sperm and embryos. We –/– No apparent arthritis developed in GalT mice, sug- are now examining these possibilities. gesting that galactose deficiency per se is not arthritogenic. GalT-deficient mice showed semi-lethality after birth, However, it is possible that the life span of the mutant and almost 80% of them died before 4 months of age. mice is too short to develop arthritis. Detailed analyses Thus, it was shown that GalT plays essential roles in the of the carbohydrate chains of IgG, joint pathology and growth and survival of young mice. autoantibody production are necessary to evaluate this Interestingly, epithelial cells of the skin and small issue fully. intestine of GalT-deficient mice showed marked enhance- ment of proliferation. Moreover, the age dependency of the disaccharidase expression pattern in intestinal epithelial Materials and methods cells was abnormal, suggesting abnormal differentiation. Thus, it was suggested that the Galβ1→4GlcNAc structure Targeting vector construction The GalT gene isolated from a 129/SvJ genomic library (Stratagene) is important for the control of growth and differentiation was used to construct the targeting vector. The PGKneobpA cassette of epithelial cells. In agreement with this notion, Metzler (Soriano et al., 1991), in which the neomycin resistance gene was ligated et al. (1994) reported that proliferation of neural epithelial under the phosphoglycerate kinase I promoter and the polyadenylation cells was enhanced in GlcNAcT-I-deficient mice in which site from bovine growth hormone was ligated downstream of the neo gene, was inserted between the SacII and NotI sites in GalT exon 1 for the expression of Gal on N-linked sugar chains is perturbed positive selection. The DT-A cassette (Yagi et al., 1993), in which the due to the lack of an acceptor GlcNAc for GalT. diphtheria toxin A fragment gene was ligated under the MC1 promoter, The abnormalities of the intestine, especially the lack was ligated at the 5 end of the targeting vector for negative selection. of lactase in suckling mice, may cause malnutrition of Most of exon 1 (420 bp from SacII to NotI) was deleted, and homologous regions at the 5 and 3 ends were of 5.7 and 1.7 kb, respectively. infant mice and result in their growth retardation and early death, although these mice did not show obvious diarrhea –/– Generation of GalT mice which might be expected in lactase deficiency. Analysis The linearized targeting vector (20 μg) was electroporated (250 V, of nutritional absorption efficiency as well as feeding of 500 μF) into 10 R1 ES cells (Nagy et al., 1993) and selected with these mice with nutrients other than lactose will clarify 180 μg (active form)/ml G418 (Gibco/BRL) for 7–10 days. Homologous recombinants were screened by PCR and confirmed by Southern blot this point. However, the cause of semi-lethality of adult –/– hybridization. The forward primer (P1) in the PGKneobpA cassette was GalT mice is not clear because these abnormalities were CTCTATGGCTTCTGAGGCGGAAAG and the reverse primer (P2) not prominent in the adult animals. outside the targeting vector was CACAGTCCCTCATATTTCAGCAGG. The mechanism for the growth augmentation of epi- PCR was carried out for 40 cycles at 94°C for 1 min, 60°C for 2 min and 72°C for 3 min in a volume of 50 μl containing 10 mM Tris–HCl thelial cells is not known at present, although several (pH 9.0), 50 mM KCl, 0.1% Triton X-100, 1.5 mM MgCl , 0.2 mM possibilities are conceivable. One possibility is that the dNTPs, 1 μM P1 and P2 primers and 2.5 U of Taq DNA polymerase. deficiency in N-linked sugar chains might cause activation Chimeric mice were generated by the aggregation method (Nagy et al., of epithelial cell growth factors or their receptors due 1993) with some modifications. Chimeras were mated with C57BL/6J females, and homozygous mutant mice were generated by inter-crossing to conformational changes. It is also possible that the of heterozygotes. Genotypes were determined by tail DNA dot-blot Galβ1→4GlcNAc group in N-linked sugar chains of hybridization using the GalT gene exon 1 as a wild-type allele-specific a cell surface has suppressive effects on growth and probe and the neo gene as a mutant allele-specific probe. Mice were differentiation of opposing cells through receptors that kept under specific pathogen-free conditions in an environmentally controlled clean room at the Laboratory Animal Research Center, recognize terminal Gal residues. In this context, β-linked Institute of Medical Science, University of Tokyo. The experiments terminal Gal in N-linked sugar chains of a cell surface were conducted according to institutional ethical guidelines for animal glycoprotein, contactinhibin, is suggested to be involved experiments and safety guidelines for gene manipulation experiments. in the contact-dependent inhibition of human lung fibroblast cell growth (Wieser et al., 1991). Southern and Northern blot analysis Another possible explanation is that GalT by itself Genomic DNA prepared from ES cells was digested with restriction enzymes, electrophoresed through a 0.7% agarose gel and transferred to participates in the regulation of cell growth. The cell nylon membranes (Gene Screen Plus, NEN). Total RNA was prepared surface GalT has been shown recently to deliver a growth from the liver by the acid guanidinium thiocyanate–phenol–chloroform inhibitory signal using cell lines which overexpress either method (Chomczynski and Sacchi, 1987) and poly(A) RNA was the intact or truncated GalT gene (Hilton et al., 1995). In purified using a QuickPrep Micro mRNA Purification Kit (Pharmacia Biotech). Poly(A) RNA was electrophoresed through a 0.8% denatured addition, involvement of this enzyme in cell growth signal agarose gel and transferred to nylon membranes. Hybridization was transduction was suggested by its direct association with carried out according to standard methods (Sambrook et al., 1989) using the epidermal growth factor receptor (Hilton et al., 1995) P-labeled DNA probes made by Multiprimed DNA labeling system and CDC2-related kinase (Bunnell et al., 1990). In any (Amersham). 1855 M.Asano et al. β-1,4-galactosyltransferase assay Bunnell,B.A., Adams,D.E. and Kidd,V.J. (1990) Transient expression of The transferase assay was conducted in a final volume of 50 μl containing a p58 protein kinase cDNA enhances mammalian glycosyltransferase 20 mM MES buffer (pH 6.5), 100 μM UDP-[ H]Gal (323 μCi/mmol), activity. Biochem. Biophys. Res. Commun., 171, 196–203. 0.05% NP-40, 3 mM 5-AMP, 5 mM 2,3-dimercapto-1-propanol, 10 mM Cheng,A., Le,T., Palacios,M., Bookbinder,L.H., Wassarman,P.M., MnCl with 100–400 μg of liver homogenate proteins as an enzyme Suzuki,F. and Bleil,J.D. (1994) Sperm–egg recognition in the mouse: source and 250 μg of human AsAg-transferrin as an acceptor which was characterization of sp56, a sperm protein having specific affinity for described previously (Furukawa et al., 1990). After the mixtures were ZP3. J. Cell Biol., 125, 867–878. incubated at 37°C for 60 min, the galactosylated product was separated Chomczynski,P. and Sacchi,N. (1987) Single-step method of RNA by paper electrophoresis in the presence of borate, and radioactivity was isolation by acid guanidinium thiocyanate–phenol–chloroform determined using a liquid scintillation counter. extraction. Anal. Biochem., 162, 156–159. Dahlqvist,A. (1964) Method for assay of intestinal disaccharidases. Anal. Measurement of lactose concentration in milk Biochem., 7, 18–25. Milk samples were collected from oxytocin-injected (0.2 U/kg weight) Eggens,I., Fenderson,B., Toyokuni,T., Dean,B., Stroud,M. and x x mothers between days 3 and 6 of lactation. The lactose concentration Hakomori,S. (1989) Specific interaction between Le and Le was determined using a lactose/D-galactose analysis kit (Boehringer determinants. A possible basis for cell recognition in preimplantation Mannheim) according to the manufacturer’s instructions. embryos and in embryonal carcinoma cells. J. Biol. Chem., 264, 9476–9484. Lectin blot analysis Endo,T., Ohbayashi,H., Kanazawa,K., Kochibe,N. and Kobata,A. (1992) Mouse serum and homogenized tissues were defatted with acetone and Carbohydrate binding specificity of immobilized Psathyrella velutina then with chloroform and methanol mixtures (2:1 and 1:2, v/v) twice lectin. J. Biol. Chem., 267, 707–713. for each. The defatted protein samples were lyophilized, subjected Fenderson,B.A., Zehavi,U. and Hakomori,S. (1984) A multivalent lacto- to SDS–polyacrylamide gel (7.5%) electrophoresis and transferred to N-fucopentose III-lysyllysine conjugate decompacts preimplantation nitrocellulose filters. Each lane contained ~10 μg of protein. Since most mouse embryos, while the free oligosaccharide is ineffective. J. Exp. Gal residues of serum glycoproteins are sialylated, the filters initially Med., 160, 1591–1596. were treated with A.ureafaciens sialidase prior to the analysis. Lectin Furukawa,K. and Kobata,A. (1992) Protein glycosylation. Curr. Opin. blot analysis using RCA-I or PVL was performed as described previously Biotechnol., 3, 554–559. (Sato et al., 1993). Furukawa,K., Matsuta,K., Takeuchi,F., Kosuge,E., Miyamoto,T. and Kobata,A. (1990) Kinetic study of a galactosyltransferase in the B Histological analysis cells of patients with rheumatoid arthritis. Int. Immunol., 2, 105–112. Tissues were fixed in 4% formaldehyde, dehydrated and embedded in Goda,T., Bustamante,S., Thornburg,W. and Koldovsky,O. (1984) Dietary- paraffin according to the standard procedure. Sections of 6 μm were induced increase in lactase activity and immunoreactive lactase in made and stained with hematoxylin–eosin (HE). adult rat jejunum. Biochem. J., 221, 261–263. Hakomori,S. (1989) Aberrant glycosylation in tumors and tumor- Measurement of mitotic indices associated carbohydrate antigens. Adv. Cancer Res., 52, 257–331. Ten-day-old pups were injected with vincristine (1 mg/kg weight) Hakomori,S. (1991) Possible functions of tumor-associated carbohydrate intraperitoneally and tissues were collected after 90 min. Samples were antigens. Curr. Opin. Immunol., 3, 646–653. fixed with Carnoy’s solution and stained with Shiff’s solution (Okada Hilton,D.A., Evans,S.C. and Shur,B.D. (1995) Altering the expression et al., 1994). The number of cells in metaphase was counted in 20 crypts of cell surface β1,4-galactosyltransferase modulates cell growth. Exp. per sample. Cell Res., 219, 640–649. Hollis,G.F., Douglas,J.G., Shaper,N.L., Shaper,J.H., Stafford-Hollis,J.M., Immunohistochemistry and lectin staining Evans,R.J. and Kirsch,I.R. (1989) Characterization of the full length Immunohistochemistry was performed as described previously cDNA for murine β-1,4-galactosyltransferase. Biochem Biophys. Res. (Matsumoto et al., 1992). In brief, cryo-sections were fixed with acetone Commun., 162, 1069–1075. and then incubated with a rabbit polyclonal antibody against lactase Ioffe,E. and Stanley,P. (1994) Mice lacking N-acetylglucosaminyl- (Goda et al., 1984), sucrase–isomaltase complex or maltase (Umesaki transferase I activity die at mid-gestation, revealing an essential role et al., 1982) for 90 min at room temperature. Finally, sections were for complex or hybrid N-linked carbohydrates. Proc. Natl Acad. Sci. visualized with 3-amino-9-ethylcarbazole. For RCA-I staining, fixed USA, 91, 728–732. cryo-sections were reacted with biotinated RCA-I and visualized with Iwakura,Y. (1989) Mechanisms of blastocyst formation of the mouse streptavidin–peroxidase. embryo. Dev. Growth Differ., 31, 523–529. Iwakura,Y. and Nozaki,M. (1985) Effects of tunicamycin on Assay of disaccharidase activities preimplantation mouse embryos: prevention of molecular The entire small intestine of 7-day-old mice was homogenized in ice- differentiation during blastocyst formation. Dev. Biol., 112, 135–144. cold phosphate-buffered saline by using a Potter–Elvejhem homogenizer Kuhn,N.J. (1983) The biosynthesis of lactose. In Mepham,T.B. (ed.), and the disaccharidase activities were measured as described previously Biochemistry of Lactation. Elsevier, Amsterdam, pp. 159–176. (Dahlqvist, 1964). The enzyme activities of each of two mice are shown. Lasky,L.A. (1992) Selectins: interpreters of cell-specific carbohydrate information during inflammation. Science, 258, 964–969. Leyton,L., LeGuen,P., Bunch,D. and Saling,P.M. (1992) Regulation of Acknowledgements mouse gamete interaction by a sperm tyrosine kinase. Proc. Natl The authors thank Drs S.Mori and T.Kuroki for their valuable comments, Acad. Sci. USA, 89, 11692–11695. and S.Kakuta and N.Uetani for their help in the experiments. We also Matsumoto,S., Setoyama,H. and Umesaki,Y. (1992) Differential thank Drs A.Nagy, R.Nagy and W.Abramow-Newerly for R1 cells, Dr induction of major histocompatibility complex molecules on mouse T.Yagi for the DT-A cassette, Dr T.Goda for anti-lactase antibody, and intestine by bacterial colonization. Gastroenterology, 103, all the members of Laboratory Animal Research Center for excellent 1777–1782. animal care. This work was supported by a grant from the Ministry of Metzler,M., Gertz,A., Sarkar,M., Schachter,H., Schrader,J.W. and Education, Science, Sports and Culture of Japan. Marth,J.D. (1994) Complex asparagine-linked oligosaccharides are required for morphogenic events during post-implantation development. EMBO J., 13, 2056–2065. References Miller,D.J., Macek,M.B. and Shur,B.D. (1992) Complementarity between sperm surface β-1,4-galactosyltransferase and egg-coat ZP3 mediates Baenziger,J.U. and Fiete,D. (1979) Structure of the complex sperm–egg binding. Nature, 357, 589–593. oligosaccharides of fetuin. J. Biol. Chem., 254, 9795–9804. Nagy,A., Rossant,J., Nagy,R., Abramow-Newerly,W. and Roder,J. (1993) Bayna,E.M., Shaper,J.H. and Shur,B.D. (1988) Temporally specific Derivation of completely cell culture-derived mice from early-passage involvement of cell surface β-1,4 galactosyltransferase during mouse embryonic stem cells. Proc. Natl Acad. Sci. USA, 90, 8424–8428. embryo morula compaction. Cell, 53, 145–157. Begovac,P.C. and Shur,B.D. (1990) Cell surface galactosyltransferase Nilsson,T., Lucocq,J.M., Mackay,D. and Warren,G. (1991) The membrane mediates the initiation of neurite outgrowth from PC12 cells on spanning domain of β-1,4-galactosyltransferase specifies trans Golgi laminin. J. Cell Biol., 110, 461–470. location. EMBO J., 10, 3567–3575. 1856 Effects of β-1,4-galactosyltransferase deficiency Okada,Y., Setoyama,H., Matsumoto,S., Imaoka,A., Nanno,M., Kawaguchi,M. and Umesaki,Y. (1994) Effects of fecal microorganisms and their chloroform-resistant variants derived from mice, rats, and humans on immunological and physiological characteristics of the intestines of ex-germfree mice. Infect. Immun., 62, 5442–5446. Parekh,R.B. et al. (1985) Association of rheumatoid arthritis and primary osteoarthritis with changes in the glycosylation pattern of total serum IgG. Nature, 316, 452–457. Paulson,J.C., Prieels,J.P., Glasgow,L.R. and Hill,R.L. (1978) Sialyl- and fucosyltransferases in the biosynthesis of asparaginyl-linked oligosaccharides in glycoproteins. Mutually exclusive glycosylation by β-galactoside α2→6 sialyltransferase and N-acetylglucosaminide α1→3 fucosyltransferase. J. Biol. Chem., 253, 5617–5624. Perillo,N.L., Pace,K.E., Seilhamer,J.J. and Baum,L.G. (1995) Apoptosis of T cells mediated by galectin-1. Nature, 378, 736–739. Rademacher,T.W., Parekh,R.B. and Dwek,R.A. (1988) Glycobiology. Annu. Rev. Biochem., 57, 785–838. Rastan,S., Thorpe,S.J., Schudder,P., Brown,S., Gooi,H.C. and Feizi,T. (1985) Cell interactions in preimplantation embryos: evidence for involvement of saccharides of the poly-N-acetyllactosamine series. J. Embryol. Exp. Morphol., 87, 115–128. Runyan,R.B., Maxwell,G. and Shur,D.B. (1986) Evidence for a novel enzymatic mechanism of neural crest cell migration on extracellular glycoconjugate matrices. J. Cell Biol., 102, 432–441. Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sato,T., Furukawa,K., Greenwalt,D.E. and Kobata,A. (1993) Most bovine milk fat globule membrane glycoproteins contain asparagine-linked sugar chains with GalNAcβ1→4GlcNAc groups. J. Biochem., 114, 890–900. Shaper,N.L., Hollis,G.F., Douglas,J.G., Kirsch,I.R. and Shaper,J.H. (1988) Genomic structure of murine β-1,4-galactosyltransferase.J.Biol.Chem., 263, 10420–10428. Sheares,B.T. and Carlson,D.M. (1984) Two distinct UDP-galactose: 2- acetamido-2-deoxy-D-glucose 4β-galactosyltransferases in porcine trachea. J. Biol. Chem., 259, 8045–8047. Shur,B.D. (1993) Glycosyltransferases as cell adhesion molecules. Curr. Opin. Cell Biol., 5, 854–863. Shur,B.D. Oettgen,P. and Bennett,D. (1979) UDP-galactose inhibits blastocyst formation in the mouse. Dev. Biol., 73, 178–181. Solter,D. and Knowles,B.B. (1978) Monoclonal antibody defining a stage- specific mouse embryonic antigen (SSEA-1). Proc. Natl Acad. Sci. USA, 75, 5565–5569. Soriano,P., Montogomery,C., Geske,R. and Bradley,A. (1991) Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell, 64, 693–702. Stacey,A., Schnieke,A., Kerr,M., Scott,A., McKee,C., Cottingham,I., Binas,B., Wilde,C. and Colman,A. (1995) Lactation is disrupted by α- lactalbumin deficiency and can be restored by human α-lactalbumin gene replacement in mice. Proc. Natl Acad. Sci. USA, 92, 2835–2839. Stinnakre,M.G., Vilotte,J.L., Soulier,S. and Mercier,J.C. (1994) Creation and phenotypic analysis of α-lactalbumin-deficient mice. Proc. Natl Acad. Sci. USA, 91, 6544–6548. Umesaki,Y., Tohyama,K. and Mutai,M. (1982) Biosynthesis of microvillus membrane-associated glycoproteins of small intestinal epithelial cells in germ-free and conventionalized mice. J. Biochem., 92, 373–379. Varki,A. (1993) Biological roles of oligosaccharides: all of the theories are correct. Glycobiology, 3, 97–130. Vestweber,D. and Kemler,R. (1985) Identification of a putative cell adhesion domain of uvomorulin. EMBO J., 4, 3393–3398. Wieser,R.J., Scutz,S, Tschank,G., Thomas,H., Dienes,H.-P. and Oesch,F. (1991) Isolation and characterization of a 60–70-kD plasma membrane glycoprotein involved in the contact-dependent inhibition of growth. J. Cell Biol., 111, 2681–2692. Yagi,T., Nada,S., Watanabe,N., Tamemoto,H., Kohmura,N., Ikawa,Y. and Aizawa,S. (1993) A novel negative selection for homologous recombination using diphtheria toxin A fragment gene. Anal. Biochem., 214, 77–86. Yang,R.-Y., Hsu,D.K. and Liu,F.-T. (1996) Expression of galectin-3 modulates T-cell growth and apoptosis. Proc. Natl Acad. Sci. USA, 93, 6737–6742. Received on October 29, 1996; revised on December 31, 1996 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Growth retardation and early death of β‐1, 4‐galactosyltransferase knockout mice with augmented proliferation and abnormal differentiation of epithelial cells

Download PDF
8 pages

Loading next page...
 
/lp/springer-journals/growth-retardation-and-early-death-of-1-4-galactosyltransferase-uEC005ORii

References (72)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1997
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1093/emboj/16.8.1850
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal Vol.16 No.8 pp.1850–1857, 1997 Growth retardation and early death of β-1,4-galactosyltransferase knockout mice with augmented proliferation and abnormal differentiation of epithelial cells of dolichophosphate synthesis, inhibits blastocyst form- Masahide Asano, Kiyoshi Furukawa , 1 2 ation of mouse embryos, suggesting that cell surface Masahiro Kido , Satoshi Matsumoto , 2 3 carbohydrates play an important role in early embryo- Yoshinori Umesaki , Naohisa Kochibe and 4 genesis (Iwakura and Nozaki, 1985). Several lines of Yoichiro Iwakura evidence suggest that galactose (Gal)-containing complex Laboratory Animal Research Center, Institute of Medical Science, N-glycans are particularly important in these processes. University of Tokyo, Minato-ku, Tokyo 108, Department of Biosignal For example, stage-specific embryonic antigen-1 (SSEA-1) Research, Tokyo Metropolitan Institute of Gerontology, Itabashi-ku, (Solter and Knowles, 1978), which contains poly-N- Tokyo 173, Yakult Central Institute for Microbiological Research, 3 acetyllactosamine structure and is expressed specifically on Kunitachi, Tokyo 186 and Faculty of Education, Gunma University, Maebashi 371, Japan pre-implantation embryos and undifferentiated embryonal carcinoma cells, has been suggested to be involved in Corresponding author cellular interactions during morula compaction and implantation, since oligosaccharides containing this anti- Carbohydrate chains on a glycoprotein are important genic structure can competitively inhibit these processes not only for protein conformation, transport and (Fenderson et al., 1984). The observation that treatment stability, but also for cell–cell and cell–matrix inter- of early embryos with endo-β-galactosidase (Rastan et al., actions. UDP-Gal:N-acetylglucosamine β-1,4-galacto- 1985) or UDP-Gal (Shur et al., 1979) inhibited compaction syltransferase (GalT) (EC 2.4.1.38) is the enzyme which also suggests the importance of Gal-containing molecules transfers galactose (Gal) to the terminal N-acetyl- in these processes. glucosamine (GlcNAc) of complex-type N-glycans in The involvement of Gal-containing carbohydrates in the Golgi apparatus. In addition, it has also been other processes has also been suggested. Lower galactosyl- suggested that this enzyme is involved directly in cell– ation of serum IgGs is characteristic of patients with cell interactions during fertilization and early embryo- rheumatoid arthritis, suggesting involvement in the patho- genesis through a subpopulation of this enzyme dis- genesis of this disease (Parekh et al., 1985; Rademacher tributed on the cell surface. In this study, GalT-deficient x a et al., 1988). Sialyl Le and sialyl Le , which contain mice were produced by gene targeting in order to NeuAcα2→3Galβ1→4[Fucα1→3]GlcNAc and Neu- examine the pathological effects of the deficiency. GalT- Acα2→3Galβ1→3[Fucα1→4]GlcNAc, respectively, are deficient mice were born normally and were fertile, but suggested to be ligands for selectins which mediate they exhibited growth retardation and semi-lethality. interactions between leukocytes and vascular endothelial Epithelial cell proliferation of the skin and small and platelet cells (reviewed in Lasky, 1992). Recently, intestine was enhanced, and cell differentiation in galectins, which particularly recognize β-galactoside sugar intestinal villi was abnormal. These observations sug- chains, were shown to be important for immune recogni- gest that GalT plays critical roles in the regulation of tion (Perillo et al., 1995; Yang et al., 1996). Furthermore, proliferation and differentiation of epithelial cells after a b x some Gal-containing carbohydrates such as Le ,Le ,Le , birth, although this enzyme is dispensable during embryonic development. Le antigens and their derivatives were shown to be Keywords: epithelial cell differentiation/epithelial cell associated with tumors (reviewed in Hakomori, 1989). It growth/galactosyltransferase/glycosyltransferase/ was shown that the expression of these antigens and knockout mouse tumor progression were closely correlated, suggesting that aberrant glycosylation at the cell surface perturbs normal cell–cell and cell–matrix interactions which results in uncontrolled cell growth, invagination and metastasis of Introduction tumor cells (reviewed in Hakomori, 1991). However, so far, the pathological effects of the deficiency in Gal- N-linked sugar chains are associated with most proteins containing carbohydrates on an animal and its embryonic found on the mammalian cell surface and in secretions. development have not been examined. Oligosaccharides on secreted glycoproteins are shown to UDP-Gal:N-acetylglucosamine β-1,4-galactosyltrans- be important for their biological activities, transport, ferase (GalT) is the only enzyme which is known to stability and clearance from the circulation, while oligo- transfer Gal from UDP-Gal to terminal N-acetylglucos- saccharides on cell surface glycoproteins are suggested to amine (GlcNAc) of complex N-glycans to form a be involved in various cellular functions including cell– cell and cell–matrix interactions during embryogenesis, Galβ1→4GlcNAc structure in the Golgi apparatus. In the immune reactions and tumor development (reviewed in mammary gland, GalT is known to synthesize lactose Iwakura, 1989; Hakomori, 1991; Furukawa and Kobata, by binding to α-lactalbumin (α-lac) which modifies the 1992; Varki, 1993). acceptor specificity of the transferase from GlcNAc to Previously, we showed that tunicamycin, an inhibitor glucose (reviewed in Kuhn, 1983). In addition to a 1850 © Oxford University Press Effects of β-1,4-galactosyltransferase deficiency biosynthetic role for GalT, it was suggested that a sub- population of this enzyme is present at the cell surface and is involved directly in cellular interactions by binding to terminal GlcNAc of glycoconjugates at the opposing cell surface and matrix (reviewed in Shur, 1993). In this context, it was suggested that this enzyme is involved directly in mouse gamete interaction (Miller et al., 1992), neurite extension (Begovac and Shur, 1990), mesenchymal and neural crest cell migration (Runyan et al., 1986) and late morula compaction (Bayna et al., 1988). In the present study, we generated mice deficient in the GalT gene by homologous recombination in order to examine the pathological effects of the deficiency in vivo. We found that homozygous GalT-deficient mice can be obtained at almost the expected ratio in crosses between heterozygous mice. However, approximately half of them died before weaning, with growth retardation, indicating the importance of GalT for survival after birth. Results –/– Generation of GalT-deficient (GalT ) mice The targeting vector was constructed by replacing the GalT exon 1 with a PGKneobpA cassette (Soriano et al., 1991), in order to disrupt both the long and short forms of the GalT gene (Shaper et al., 1988; Hollis et al., 1989). In this construct, the translation initiation sites for both forms and the Golgi retention signal (Nilsson et al., 1991) in exon 1 were deleted (Figure 1A). The targeting vector was electroporated into embryonic stem (ES) cells and G418-resistant colonies were picked up. Twelve positive clones were obtained out of 180 Neo clones by PCR, and the homologous recombination was confirmed in nine clones by Southern blot analysis using an external 3 probe (Figure 1B) and 5 probe (data not shown). Chimeric Fig. 1. Targeted disruption of the GalT gene by homologous mice were produced using three targeted ES cell clones recombination. (A) Targeting strategy. Coding and non-coding exons and the targeted allele was transmitted through a germ of the GalT gene are shown by closed and open boxes, and the –/– line in two independent ES cell clones. GalT mice PGKneobpA cassette and the diphtheria toxin A fragment gene /– were generated by inter-crossing of GalT mice. The cassette (DT-A) by dotted and hatched boxes, respectively. PCR primers (P1 and P2) used for screening are shown by arrowheads. phenotypes described here were identical in the two mouse –/– 5-Probe and 3-probe show the positions of external probes used for lines. GalT mRNA was not detected in GalT mice, Southern blot analysis and the expected HindIII fragments are indicating disruption of this gene (Figure 1C). indicated by arrows. BH, BamHI; Hd, HindIII; RI, EcoRI; Sa, SacII; No, NotI. (B) Southern blot analysis of ES cell clones. Genomic DNA GalT activity and N-linked carbohydrate chains in (10 μg) from parent ES cells (R1) and targeted ES cell clones were –/– digested with HindIII and hybridized with the 3 probe. The expected GalT mice DNA fragments for the mutant allele (Mt) and wild-type allele (Wt) When GalT activity was determined using the liver homo- –/– are indicated. (C) Northern blot analysis of GalT mice. Poly(A) genates as an enzyme source and asialo (As)- and agalacto RNA (1 μg) from the liver was hybridized with exon 1 and 2 probes (Ag)-transferrin as an acceptor, the transferase activity in of the GalT gene, and rehybridized with the β-actin probe. /– / GalT mice was just half of that in GalT mice, while –/– / that in GalT mice was 5% of that in GalT mice (Figure 2A). However, it should be pointed out that this position of the GlcNAc of AsAg-transferrin (data not –/– level in GalT mice was significantly higher than in the shown). control without enzyme. A similar result was obtained In order to determine whether or not the Galβ1→ using the spleen homogenates as an enzyme source (data 4GlcNAc structure is expressed in N-linked sugar chains –/– not shown). In order to verify the galactosyltransferase of GalT mouse glycoproteins, serum glycoproteins were activity found in this assay, the galactosylated product analyzed by lectin blotting. When filters were stained with isolated by Bio-Gel P-6 column chromatography was Coomassie brilliant blue (CBB), no significant difference digested with diplococcal β-galactosidase which cleaves was detected between the three genotypes (Figure 2B, the Galβ1→4GlcNAc linkage but not the Galβ1→ CBB). When the filters were incubated with Ricinus 3GlcNAc linkage (Paulson et al., 1978). When the digested communis agglutinin (RCA)-I, which binds to oligosacch- material was analyzed by Bio-Gel P-4 chromatography, arides which terminate with Galβ1→4GlcNAc/Glc groups the radioactive product was eluted at the same position as (Baenziger and Fiete, 1979), RCA-I binding was observed / /– galactose, indicating that Gal was transferred at the C-4 for most protein bands in GalT and GalT mice, 1851 M.Asano et al. / /– –/– Fig. 2. (A) β-1,4-galactosyltransferase activity in liver homogenates from 10-week-old GalT (d), GalT (m) and GalT (j) mice. (B) Lectin / /– –/– blot analysis of serum glycoproteins from 10-week-old GalT (/), GalT (/–) and GalT (–/–) mice. The bars indicate the positions of molecular weight markers. CBB indicates filters stained with Coomassie brilliant blue to detect protein bands. RCA-I and PVL indicate filters / /– –/– incubated with RCA-I and PVL, respectively. (C) Lactose concentration in milk of five GalT (/), five GalT (/–) and two GalT (–/–) mice. The values indicate averages with standard deviations. –/– –/– while no binding was observed in a GalT mouse, with No lactose was detected in GalT mouse milk, while /– the exception of a few weak bands (Figure 2B, RCA-I). its concentration in the milk of GalT mice was about RCA-I binding was completely lost upon digestion of half as much as that of wild-type mice (Figure 2C). These glycoprotein samples with diplococcal β-galactosidase or results confirm that GalT is a component of lactose –/– N-glycanase prior to the analysis (data not shown), show- synthetase (Kuhn, 1983). As a result, these GalT mothers ing that these bands actually represented N-glycans con- could not rear their offspring (data not shown). taining β-linked Gal. In contrast, when filters were incubated with Psathyrella velutina lectin (PVL), which –/– Semi-lethality of GalT mice binds to oligosaccharides which terminate with β-linked –/– GalT mice were born healthy at almost the expected GlcNAc (Endo et al., 1992), PVL did not bind to most / /– ratio of 20% in heterozygous mating, suggesting normal protein bands in GalT and GalT mice, but did so in –/– –/– embryonic development, although GalT mice were a GalT mouse (Figure 2B, PVL). No PVL binding was /– / slightly smaller than GalT and GalT mice at birth observed upon digestion of glycoprotein samples with –/– (body weight of GalT mice was 80–90% of that of jack bean β-N-acetylhexosaminidase or N-glycanase prior control littermates at birth). Most of them, however, to the analysis (data not shown), showing that these developed severe skin lesions within a few days after bands actually represented N-glycans containing β-linked birth (Figure 3A). Horny cells continued to come off from GlcNAc. These results indicate that N-linked sugar chains –/– the flaky skin for ~2 weeks, and the skin lesions eventually of most serum glycoproteins from GalT mice are not –/– healed. GalT mice also showed growth retardation and galactosylated and terminated with β-GlcNAc, although a nearly half of them died before 4 weeks of age (Figure few protein bands were RCA-I positive and contained –/– 3B). The body weight of GalT survivors was only 60– β-1,4-linked Gal in their N-linked sugar chains. In order 70% of that of GalT mice at 4 weeks old, while to rule out the possibility that the RCA-I-positive glyco- /– –/– GalT mice were normal. After weaning, GalT mice proteins were produced by bacteria in the gut, we analyzed –/– gained weight normally, but they still continued to die, proteins from intestinal bacteria in GalT mice and found and only 20% of them survived beyond 16 weeks (Figure that these bacteria have no RCA-I-reactive glycoproteins –/– upon lectin binding (data not shown). In addition, no 3B). Both mature female and male GalT mice were –/– change has been detected in the RCA-I-reactive glyco- fertile (Table I). Sperm from GalT males was able to –/– protein pattern on lectin blotting after treatment of these fertilize. Although GalT females appeared to be able to mice with antibiotics (data not shown). ovulate, and maintain pregnancy to term, the litter size 1852 Effects of β-1,4-galactosyltransferase deficiency Fig. 4. Histological analysis (HE staining) of the skin of 10-day-old / –/– pups. (A and C) A GalT mouse; (B and D) a GalT mouse. Note that the prickle cell layer and granular layer are much thicker in the –/– GalT mouse than in the normal mouse. Ep: epidermis. Original magnification: 100 (A and B), 400 (C and D). cell layer and granular layer was observed, and the horny –/– layer was also thickened in GalT mice (Figure 4C and D). Many keratohyalin granules were deposited not only in the granular layer but also in the prickle cell layer in –/– GalT mice (Figure 4C and D). These histological observations indicate that acanthosis and hyperkeratosis –/– occurred in the skin of GalT mice. This result suggests that proliferation of epithelial stem cells located in the –/– basal layer of the epidermis is augmented in GalT mice. –/– In the small intestine, the crypts of GalT mice were found to be enlarged 3–4 times compared with those of –/– –/– Fig. 3. Semi-lethality in GalT mice. (A) A 10-day-old GalT pup wild-type (Figure 5A and B), indicating that proliferation –/– (left) with a control littermate (right). Note that the GalT pup is of epithelial stem cells located in the crypt is also smaller than the normal littermate and the hair is lost from the back –/– –/– with the presence of flaky skin. (B) Survival ratio of GalT mice enhanced in GalT mice. RCA-I staining showed that after birth. The genotype ratios at birth in heterozygous mating were Galβ1→4GlcNAc-containing sugar chains were not 20% (–/–), 49% (/–) and 31% (/) from 59 pups analyzed. The –/– expressed in the crypts and villi of GalT mice, while number of mice at birth is taken as 100% and the ratios of survivors the surface of crypt and villus cells was stained strongly in 30 mice are shown. /– with RCA-I in GalT mice (Figure 5C and D). The number of metaphase cells was counted after vincristine injection to confirm the augmented proliferation of epi- Table I. Litter sizes of –/–/– matings thelial cells. The numbers of metaphase cells per crypt per 90 min were 21.7  5.6 (–/–) versus 5.5  1.4 (/–) in Femalemale Litter size (No. of litters examined) the duodenum and 6.9  1.8 (–/–) versus 2.0  1.1 (/–) in the ileum, indicating that the mitotic indices of the –/–/– 5.1  2.1 (24) /––/– 6.2  2.3 (51) small intestine crypt cells were increased ~4 times in // 7.2  2.6 (53) –/– /– GalT mice as compared with those of GalT mice. These results suggest that the Galβ1→4GlcNAc structure P0.001. may be involved in growth control of epithelial cells in P0.05. the small intestine and skin. However, these abnormalities –/– were less prominent in adult GalT mice. was slightly smaller compared with that of the wild-type Abnormal differentiation of small intestinal villus control (Table I). cells Interestingly, differentiation of small intestine villus cells –/– Enhanced proliferation of epithelial cells in the was also found to be abnormal in GalT mice using skin and small intestine disaccharidases as markers. Although lactase was Histological examination revealed that the epidermis was expressed strongly over the entire surface of villus epi- –/– /– thickened in the skin lesions of GalT mice (Figure 4A thelial cells of 7-day-old GalT mice (Figure 6A), lactase –/– and B). At higher magnification, hyperplasia of the prickle localization was restricted to the top of the villi in GalT 1853 M.Asano et al. Fig. 5. Histological analysis of the small intestine of pups. Fig. 6. Analysis of small intestine disaccharidases. (A and B) HE staining; (C and D) RCA-I staining. (A) A 10-day-old Immunohistochemical localization of lactase (A and D), maltase (B / –/– GalT mouse; (B) a 10-day-old GalT mouse; (C) a 7-day-old and E) and the sucrase–isomaltase complex (C and F)inthe /– –/– GalT mouse; (D) a 7-day-old GalT mouse. Note that the crypt /– –/– duodenum of GalT (A, B and C) and GalT (D, E and F) mice. –/– size is much larger in the GalT mouse than in the control mouse –/– Note that the expression of lactase is much weaker in the GalT and RCA-I-reactive sugar chains are not expressed in crypt and villus mouse compared with the normal mouse, whereas maltase and sucrase –/– cells of the GalT mouse. Cr: crypt. Original magnification: –/– expression is stronger in the GalT mouse. Original magnification: 400 (A and B), 200 (C and D). Table II. Disaccharidase activities (nmol/mg protein/min) in 7-day-old mice (Figure 6D). Additionally, maltase and the sucrase– mice –/– isomaltase complex appeared precociously in GalT mice (Figure 6E and F). These enzymes had not yet been /– –/– /– expressed on day 7 in GalT mice (Figure 6B and C). Lactase 2.80, 2.27 0.86, 0.78 Since no RCA-I binding was detected in villus cells of Maltase 1.89, 1.17 9.62, 5.56 –/– GalT mice (Figure 5D), it is suggested that the lack of Isomaltase 0.00, 0.00 0.35, 0.76 Galβ1→4GlcNAc structure is responsible for the abnormal differentiation of villus cells. Consistent with the immuno- –/– histochemical findings, lactase activity in GalT small intestine homogenates was markedly low, while maltase galactosyltransferase in porcine trachea was suggested –/– and isomaltase activities were higher in GalT mice previously (Sheares and Carlson, 1984). However, we /– compared with GalT mice (Table II). Sucrase activity cannot rule out the possibilities that galactosyltransferases, was barely detectable in both types of mice at day 7, and which are involved in glycolipid or glycosaminoglycan –/– was still not detected in GalT mice by day 16, in synthesis, or other linkage specificity, might be responsible /– contrast to GalT mice in which the activity was detected for this activity. We are now trying to discriminate between at day 16 (data not shown). these possibilities. Homozygous GalT-deficient mice were obtained at almost the expected ratio from crosses between hetero- Discussion –/– zygous mice, suggesting that the development of GalT In this study, we have generated GalT-deficient mice and embryos is normal. This is in clear contrast to N-acetyl- shown that 95% of the galactosyltransferase activity glucosaminyltransferase (GlcNAcT)-I knockout mice was inactivated and that most Galβ1→4GlcNAc residues which die on day 10.5 of embryonic development due to on glycoprotein carbohydrate chains were absent in these defects in neural tube formation and vascularization (Ioffe mice. The origin of the low residual galactosyltransferase and Stanley, 1994; Metzler et al., 1994). Furthermore, we –/– activity is not clear. Since the translation initiation sites found that both male and female GalT mice were fertile. and Golgi retention signal were deleted in the targeting The litter size from homozygous mothers, however, was construct, the galactosyltransferase activity of the GalT slightly smaller compared with normal mice, suggesting gene is considered to be lost. It is unlikely that GalT a possible deficiency in uterus function. These observations activity and these Gal-containing glycoproteins are derived suggest that neither the carbohydrate structures containing –/– from colostrum because they were detected in GalT Gal nor GalT itself are absolutely necessary for cell–cell mice even at 10 weeks of age. One possibility is that this interactions during fertilization and embryonic develop- residual activity could derive from previously unknown ment, although previous studies suggested the involvement galactosyltransferases encoded by genes distinct from of Gal-containing oligosaccharides and/or GalT in these GalT. In this context, the presence of another β-1,4- processes (Eggens et al., 1989; Shur, 1993). However, 1854 Effects of β-1,4-galactosyltransferase deficiency besides GalT, two other proteins, sp56 (Cheng et al., case, our results indicate the importance of GalT in the 1994) and a 95 kDa peptide (Leyton et al., 1992), are regulation of cellular growth and differentiation. –/– known as zona-binding proteins during fertilization, and Lactose was not detected in the milk of GalT mothers –/– E-cadhelin/uvomorulin is also suggested to play an import- as was also the case in the milk of α-lac mothers (Stinnakre et al., 1994; Stacey et al., 1995). This is ant role in morula compaction (Vestweber and Kemler, because GalT is an essential component of lactose synthet- 1985). Thus, it is possible that these molecules might ase together with α-lactalbumin (Kuhn, 1983). The milk have substituted for the function of GalT in these processes. –/– of α-lac mothers has been reported to be too viscous to It is also possible that embryonic Gal-containing carbo- be sucked by pups because lactose regulates the osmotic hydrates such as SSEA-1 are synthesized by enzymes pressure and volume of the milk (Stinnakre et al., 1994; other than GalT or that enzymes which are usually not Stacey et al., 1995). This is probably the reason why involved in the synthesis of this structure substitute for –/– GalT mothers could not rear their offspring. GalT and allow development of sperm and embryos. We –/– No apparent arthritis developed in GalT mice, sug- are now examining these possibilities. gesting that galactose deficiency per se is not arthritogenic. GalT-deficient mice showed semi-lethality after birth, However, it is possible that the life span of the mutant and almost 80% of them died before 4 months of age. mice is too short to develop arthritis. Detailed analyses Thus, it was shown that GalT plays essential roles in the of the carbohydrate chains of IgG, joint pathology and growth and survival of young mice. autoantibody production are necessary to evaluate this Interestingly, epithelial cells of the skin and small issue fully. intestine of GalT-deficient mice showed marked enhance- ment of proliferation. Moreover, the age dependency of the disaccharidase expression pattern in intestinal epithelial Materials and methods cells was abnormal, suggesting abnormal differentiation. Thus, it was suggested that the Galβ1→4GlcNAc structure Targeting vector construction The GalT gene isolated from a 129/SvJ genomic library (Stratagene) is important for the control of growth and differentiation was used to construct the targeting vector. The PGKneobpA cassette of epithelial cells. In agreement with this notion, Metzler (Soriano et al., 1991), in which the neomycin resistance gene was ligated et al. (1994) reported that proliferation of neural epithelial under the phosphoglycerate kinase I promoter and the polyadenylation cells was enhanced in GlcNAcT-I-deficient mice in which site from bovine growth hormone was ligated downstream of the neo gene, was inserted between the SacII and NotI sites in GalT exon 1 for the expression of Gal on N-linked sugar chains is perturbed positive selection. The DT-A cassette (Yagi et al., 1993), in which the due to the lack of an acceptor GlcNAc for GalT. diphtheria toxin A fragment gene was ligated under the MC1 promoter, The abnormalities of the intestine, especially the lack was ligated at the 5 end of the targeting vector for negative selection. of lactase in suckling mice, may cause malnutrition of Most of exon 1 (420 bp from SacII to NotI) was deleted, and homologous regions at the 5 and 3 ends were of 5.7 and 1.7 kb, respectively. infant mice and result in their growth retardation and early death, although these mice did not show obvious diarrhea –/– Generation of GalT mice which might be expected in lactase deficiency. Analysis The linearized targeting vector (20 μg) was electroporated (250 V, of nutritional absorption efficiency as well as feeding of 500 μF) into 10 R1 ES cells (Nagy et al., 1993) and selected with these mice with nutrients other than lactose will clarify 180 μg (active form)/ml G418 (Gibco/BRL) for 7–10 days. Homologous recombinants were screened by PCR and confirmed by Southern blot this point. However, the cause of semi-lethality of adult –/– hybridization. The forward primer (P1) in the PGKneobpA cassette was GalT mice is not clear because these abnormalities were CTCTATGGCTTCTGAGGCGGAAAG and the reverse primer (P2) not prominent in the adult animals. outside the targeting vector was CACAGTCCCTCATATTTCAGCAGG. The mechanism for the growth augmentation of epi- PCR was carried out for 40 cycles at 94°C for 1 min, 60°C for 2 min and 72°C for 3 min in a volume of 50 μl containing 10 mM Tris–HCl thelial cells is not known at present, although several (pH 9.0), 50 mM KCl, 0.1% Triton X-100, 1.5 mM MgCl , 0.2 mM possibilities are conceivable. One possibility is that the dNTPs, 1 μM P1 and P2 primers and 2.5 U of Taq DNA polymerase. deficiency in N-linked sugar chains might cause activation Chimeric mice were generated by the aggregation method (Nagy et al., of epithelial cell growth factors or their receptors due 1993) with some modifications. Chimeras were mated with C57BL/6J females, and homozygous mutant mice were generated by inter-crossing to conformational changes. It is also possible that the of heterozygotes. Genotypes were determined by tail DNA dot-blot Galβ1→4GlcNAc group in N-linked sugar chains of hybridization using the GalT gene exon 1 as a wild-type allele-specific a cell surface has suppressive effects on growth and probe and the neo gene as a mutant allele-specific probe. Mice were differentiation of opposing cells through receptors that kept under specific pathogen-free conditions in an environmentally controlled clean room at the Laboratory Animal Research Center, recognize terminal Gal residues. In this context, β-linked Institute of Medical Science, University of Tokyo. The experiments terminal Gal in N-linked sugar chains of a cell surface were conducted according to institutional ethical guidelines for animal glycoprotein, contactinhibin, is suggested to be involved experiments and safety guidelines for gene manipulation experiments. in the contact-dependent inhibition of human lung fibroblast cell growth (Wieser et al., 1991). Southern and Northern blot analysis Another possible explanation is that GalT by itself Genomic DNA prepared from ES cells was digested with restriction enzymes, electrophoresed through a 0.7% agarose gel and transferred to participates in the regulation of cell growth. The cell nylon membranes (Gene Screen Plus, NEN). Total RNA was prepared surface GalT has been shown recently to deliver a growth from the liver by the acid guanidinium thiocyanate–phenol–chloroform inhibitory signal using cell lines which overexpress either method (Chomczynski and Sacchi, 1987) and poly(A) RNA was the intact or truncated GalT gene (Hilton et al., 1995). In purified using a QuickPrep Micro mRNA Purification Kit (Pharmacia Biotech). Poly(A) RNA was electrophoresed through a 0.8% denatured addition, involvement of this enzyme in cell growth signal agarose gel and transferred to nylon membranes. Hybridization was transduction was suggested by its direct association with carried out according to standard methods (Sambrook et al., 1989) using the epidermal growth factor receptor (Hilton et al., 1995) P-labeled DNA probes made by Multiprimed DNA labeling system and CDC2-related kinase (Bunnell et al., 1990). In any (Amersham). 1855 M.Asano et al. β-1,4-galactosyltransferase assay Bunnell,B.A., Adams,D.E. and Kidd,V.J. (1990) Transient expression of The transferase assay was conducted in a final volume of 50 μl containing a p58 protein kinase cDNA enhances mammalian glycosyltransferase 20 mM MES buffer (pH 6.5), 100 μM UDP-[ H]Gal (323 μCi/mmol), activity. Biochem. Biophys. Res. Commun., 171, 196–203. 0.05% NP-40, 3 mM 5-AMP, 5 mM 2,3-dimercapto-1-propanol, 10 mM Cheng,A., Le,T., Palacios,M., Bookbinder,L.H., Wassarman,P.M., MnCl with 100–400 μg of liver homogenate proteins as an enzyme Suzuki,F. and Bleil,J.D. (1994) Sperm–egg recognition in the mouse: source and 250 μg of human AsAg-transferrin as an acceptor which was characterization of sp56, a sperm protein having specific affinity for described previously (Furukawa et al., 1990). After the mixtures were ZP3. J. Cell Biol., 125, 867–878. incubated at 37°C for 60 min, the galactosylated product was separated Chomczynski,P. and Sacchi,N. (1987) Single-step method of RNA by paper electrophoresis in the presence of borate, and radioactivity was isolation by acid guanidinium thiocyanate–phenol–chloroform determined using a liquid scintillation counter. extraction. Anal. Biochem., 162, 156–159. Dahlqvist,A. (1964) Method for assay of intestinal disaccharidases. Anal. Measurement of lactose concentration in milk Biochem., 7, 18–25. Milk samples were collected from oxytocin-injected (0.2 U/kg weight) Eggens,I., Fenderson,B., Toyokuni,T., Dean,B., Stroud,M. and x x mothers between days 3 and 6 of lactation. The lactose concentration Hakomori,S. (1989) Specific interaction between Le and Le was determined using a lactose/D-galactose analysis kit (Boehringer determinants. A possible basis for cell recognition in preimplantation Mannheim) according to the manufacturer’s instructions. embryos and in embryonal carcinoma cells. J. Biol. Chem., 264, 9476–9484. Lectin blot analysis Endo,T., Ohbayashi,H., Kanazawa,K., Kochibe,N. and Kobata,A. (1992) Mouse serum and homogenized tissues were defatted with acetone and Carbohydrate binding specificity of immobilized Psathyrella velutina then with chloroform and methanol mixtures (2:1 and 1:2, v/v) twice lectin. J. Biol. Chem., 267, 707–713. for each. The defatted protein samples were lyophilized, subjected Fenderson,B.A., Zehavi,U. and Hakomori,S. (1984) A multivalent lacto- to SDS–polyacrylamide gel (7.5%) electrophoresis and transferred to N-fucopentose III-lysyllysine conjugate decompacts preimplantation nitrocellulose filters. Each lane contained ~10 μg of protein. Since most mouse embryos, while the free oligosaccharide is ineffective. J. Exp. Gal residues of serum glycoproteins are sialylated, the filters initially Med., 160, 1591–1596. were treated with A.ureafaciens sialidase prior to the analysis. Lectin Furukawa,K. and Kobata,A. (1992) Protein glycosylation. Curr. Opin. blot analysis using RCA-I or PVL was performed as described previously Biotechnol., 3, 554–559. (Sato et al., 1993). Furukawa,K., Matsuta,K., Takeuchi,F., Kosuge,E., Miyamoto,T. and Kobata,A. (1990) Kinetic study of a galactosyltransferase in the B Histological analysis cells of patients with rheumatoid arthritis. Int. Immunol., 2, 105–112. Tissues were fixed in 4% formaldehyde, dehydrated and embedded in Goda,T., Bustamante,S., Thornburg,W. and Koldovsky,O. (1984) Dietary- paraffin according to the standard procedure. Sections of 6 μm were induced increase in lactase activity and immunoreactive lactase in made and stained with hematoxylin–eosin (HE). adult rat jejunum. Biochem. J., 221, 261–263. Hakomori,S. (1989) Aberrant glycosylation in tumors and tumor- Measurement of mitotic indices associated carbohydrate antigens. Adv. Cancer Res., 52, 257–331. Ten-day-old pups were injected with vincristine (1 mg/kg weight) Hakomori,S. (1991) Possible functions of tumor-associated carbohydrate intraperitoneally and tissues were collected after 90 min. Samples were antigens. Curr. Opin. Immunol., 3, 646–653. fixed with Carnoy’s solution and stained with Shiff’s solution (Okada Hilton,D.A., Evans,S.C. and Shur,B.D. (1995) Altering the expression et al., 1994). The number of cells in metaphase was counted in 20 crypts of cell surface β1,4-galactosyltransferase modulates cell growth. Exp. per sample. Cell Res., 219, 640–649. Hollis,G.F., Douglas,J.G., Shaper,N.L., Shaper,J.H., Stafford-Hollis,J.M., Immunohistochemistry and lectin staining Evans,R.J. and Kirsch,I.R. (1989) Characterization of the full length Immunohistochemistry was performed as described previously cDNA for murine β-1,4-galactosyltransferase. Biochem Biophys. Res. (Matsumoto et al., 1992). In brief, cryo-sections were fixed with acetone Commun., 162, 1069–1075. and then incubated with a rabbit polyclonal antibody against lactase Ioffe,E. and Stanley,P. (1994) Mice lacking N-acetylglucosaminyl- (Goda et al., 1984), sucrase–isomaltase complex or maltase (Umesaki transferase I activity die at mid-gestation, revealing an essential role et al., 1982) for 90 min at room temperature. Finally, sections were for complex or hybrid N-linked carbohydrates. Proc. Natl Acad. Sci. visualized with 3-amino-9-ethylcarbazole. For RCA-I staining, fixed USA, 91, 728–732. cryo-sections were reacted with biotinated RCA-I and visualized with Iwakura,Y. (1989) Mechanisms of blastocyst formation of the mouse streptavidin–peroxidase. embryo. Dev. Growth Differ., 31, 523–529. Iwakura,Y. and Nozaki,M. (1985) Effects of tunicamycin on Assay of disaccharidase activities preimplantation mouse embryos: prevention of molecular The entire small intestine of 7-day-old mice was homogenized in ice- differentiation during blastocyst formation. Dev. Biol., 112, 135–144. cold phosphate-buffered saline by using a Potter–Elvejhem homogenizer Kuhn,N.J. (1983) The biosynthesis of lactose. In Mepham,T.B. (ed.), and the disaccharidase activities were measured as described previously Biochemistry of Lactation. Elsevier, Amsterdam, pp. 159–176. (Dahlqvist, 1964). The enzyme activities of each of two mice are shown. Lasky,L.A. (1992) Selectins: interpreters of cell-specific carbohydrate information during inflammation. Science, 258, 964–969. Leyton,L., LeGuen,P., Bunch,D. and Saling,P.M. (1992) Regulation of Acknowledgements mouse gamete interaction by a sperm tyrosine kinase. Proc. Natl The authors thank Drs S.Mori and T.Kuroki for their valuable comments, Acad. Sci. USA, 89, 11692–11695. and S.Kakuta and N.Uetani for their help in the experiments. We also Matsumoto,S., Setoyama,H. and Umesaki,Y. (1992) Differential thank Drs A.Nagy, R.Nagy and W.Abramow-Newerly for R1 cells, Dr induction of major histocompatibility complex molecules on mouse T.Yagi for the DT-A cassette, Dr T.Goda for anti-lactase antibody, and intestine by bacterial colonization. Gastroenterology, 103, all the members of Laboratory Animal Research Center for excellent 1777–1782. animal care. This work was supported by a grant from the Ministry of Metzler,M., Gertz,A., Sarkar,M., Schachter,H., Schrader,J.W. and Education, Science, Sports and Culture of Japan. Marth,J.D. (1994) Complex asparagine-linked oligosaccharides are required for morphogenic events during post-implantation development. EMBO J., 13, 2056–2065. References Miller,D.J., Macek,M.B. and Shur,B.D. (1992) Complementarity between sperm surface β-1,4-galactosyltransferase and egg-coat ZP3 mediates Baenziger,J.U. and Fiete,D. (1979) Structure of the complex sperm–egg binding. Nature, 357, 589–593. oligosaccharides of fetuin. J. Biol. Chem., 254, 9795–9804. Nagy,A., Rossant,J., Nagy,R., Abramow-Newerly,W. and Roder,J. (1993) Bayna,E.M., Shaper,J.H. and Shur,B.D. (1988) Temporally specific Derivation of completely cell culture-derived mice from early-passage involvement of cell surface β-1,4 galactosyltransferase during mouse embryonic stem cells. Proc. Natl Acad. Sci. USA, 90, 8424–8428. embryo morula compaction. Cell, 53, 145–157. Begovac,P.C. and Shur,B.D. (1990) Cell surface galactosyltransferase Nilsson,T., Lucocq,J.M., Mackay,D. and Warren,G. (1991) The membrane mediates the initiation of neurite outgrowth from PC12 cells on spanning domain of β-1,4-galactosyltransferase specifies trans Golgi laminin. J. Cell Biol., 110, 461–470. location. EMBO J., 10, 3567–3575. 1856 Effects of β-1,4-galactosyltransferase deficiency Okada,Y., Setoyama,H., Matsumoto,S., Imaoka,A., Nanno,M., Kawaguchi,M. and Umesaki,Y. (1994) Effects of fecal microorganisms and their chloroform-resistant variants derived from mice, rats, and humans on immunological and physiological characteristics of the intestines of ex-germfree mice. Infect. Immun., 62, 5442–5446. Parekh,R.B. et al. (1985) Association of rheumatoid arthritis and primary osteoarthritis with changes in the glycosylation pattern of total serum IgG. Nature, 316, 452–457. Paulson,J.C., Prieels,J.P., Glasgow,L.R. and Hill,R.L. (1978) Sialyl- and fucosyltransferases in the biosynthesis of asparaginyl-linked oligosaccharides in glycoproteins. Mutually exclusive glycosylation by β-galactoside α2→6 sialyltransferase and N-acetylglucosaminide α1→3 fucosyltransferase. J. Biol. Chem., 253, 5617–5624. Perillo,N.L., Pace,K.E., Seilhamer,J.J. and Baum,L.G. (1995) Apoptosis of T cells mediated by galectin-1. Nature, 378, 736–739. Rademacher,T.W., Parekh,R.B. and Dwek,R.A. (1988) Glycobiology. Annu. Rev. Biochem., 57, 785–838. Rastan,S., Thorpe,S.J., Schudder,P., Brown,S., Gooi,H.C. and Feizi,T. (1985) Cell interactions in preimplantation embryos: evidence for involvement of saccharides of the poly-N-acetyllactosamine series. J. Embryol. Exp. Morphol., 87, 115–128. Runyan,R.B., Maxwell,G. and Shur,D.B. (1986) Evidence for a novel enzymatic mechanism of neural crest cell migration on extracellular glycoconjugate matrices. J. Cell Biol., 102, 432–441. Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sato,T., Furukawa,K., Greenwalt,D.E. and Kobata,A. (1993) Most bovine milk fat globule membrane glycoproteins contain asparagine-linked sugar chains with GalNAcβ1→4GlcNAc groups. J. Biochem., 114, 890–900. Shaper,N.L., Hollis,G.F., Douglas,J.G., Kirsch,I.R. and Shaper,J.H. (1988) Genomic structure of murine β-1,4-galactosyltransferase.J.Biol.Chem., 263, 10420–10428. Sheares,B.T. and Carlson,D.M. (1984) Two distinct UDP-galactose: 2- acetamido-2-deoxy-D-glucose 4β-galactosyltransferases in porcine trachea. J. Biol. Chem., 259, 8045–8047. Shur,B.D. (1993) Glycosyltransferases as cell adhesion molecules. Curr. Opin. Cell Biol., 5, 854–863. Shur,B.D. Oettgen,P. and Bennett,D. (1979) UDP-galactose inhibits blastocyst formation in the mouse. Dev. Biol., 73, 178–181. Solter,D. and Knowles,B.B. (1978) Monoclonal antibody defining a stage- specific mouse embryonic antigen (SSEA-1). Proc. Natl Acad. Sci. USA, 75, 5565–5569. Soriano,P., Montogomery,C., Geske,R. and Bradley,A. (1991) Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell, 64, 693–702. Stacey,A., Schnieke,A., Kerr,M., Scott,A., McKee,C., Cottingham,I., Binas,B., Wilde,C. and Colman,A. (1995) Lactation is disrupted by α- lactalbumin deficiency and can be restored by human α-lactalbumin gene replacement in mice. Proc. Natl Acad. Sci. USA, 92, 2835–2839. Stinnakre,M.G., Vilotte,J.L., Soulier,S. and Mercier,J.C. (1994) Creation and phenotypic analysis of α-lactalbumin-deficient mice. Proc. Natl Acad. Sci. USA, 91, 6544–6548. Umesaki,Y., Tohyama,K. and Mutai,M. (1982) Biosynthesis of microvillus membrane-associated glycoproteins of small intestinal epithelial cells in germ-free and conventionalized mice. J. Biochem., 92, 373–379. Varki,A. (1993) Biological roles of oligosaccharides: all of the theories are correct. Glycobiology, 3, 97–130. Vestweber,D. and Kemler,R. (1985) Identification of a putative cell adhesion domain of uvomorulin. EMBO J., 4, 3393–3398. Wieser,R.J., Scutz,S, Tschank,G., Thomas,H., Dienes,H.-P. and Oesch,F. (1991) Isolation and characterization of a 60–70-kD plasma membrane glycoprotein involved in the contact-dependent inhibition of growth. J. Cell Biol., 111, 2681–2692. Yagi,T., Nada,S., Watanabe,N., Tamemoto,H., Kohmura,N., Ikawa,Y. and Aizawa,S. (1993) A novel negative selection for homologous recombination using diphtheria toxin A fragment gene. Anal. Biochem., 214, 77–86. Yang,R.-Y., Hsu,D.K. and Liu,F.-T. (1996) Expression of galectin-3 modulates T-cell growth and apoptosis. Proc. Natl Acad. Sci. USA, 93, 6737–6742. Received on October 29, 1996; revised on December 31, 1996

Journal

The EMBO Journal – Springer Journals

Published: Apr 15, 1997

Keywords: epithelial cell differentiation; epithelial cell growth; galactosyltransferase; glycosyltransferase; knockout mouse

There are no references for this article.