Complementation of a Defect in the Asparagine-Linked Glycosylation of a Mouse FM3A Mutant G258 Cell Line by Spheroplast Fusion of a Human Mega YAC Clone 923F5

Biosci. Biotechnol. Biochem., 76 (1), 108–114, 2012

Complementation of a Defect in the Asparagine-Linked Glycosylation of a Mouse FM3A Mutant G258 Cell Line by Spheroplast Fusion of a Human Mega YAC Clone 923f5

y Takahisa MASUDA,1 Masayuki MORIYA,1 Kensuke KATAOKA,1 and Yoshihisa NISHIKAWA1;2;

1Department of Applied Chemistry, School of Engineering, Tokai University, 4-4-1 Kitakaname, Hiratsuka, Kanagawa 259-1292, Japan 2Institute of Glycoscience, Tokai University, 4-4-1 Kitakaname, Hiratsuka, Kanagawa 259-1292, Japan

Received August 4, 2011; Accepted October 1, 2011; Online Publication, January 7, 2012 [doi:10.1271/bbb.110569]

Mouse G258 mutant stopped both cell growth and genes were isolated mainly in yeast by complementation the synthesis of lipid-linked oligosaccharide at the of mutants in cell growth, and later in mammals by 1) Man3GlcNAc2-P-P-Dolichol at a restricted temperature homology cloning. with a single gene mutation. To clarify the lesion in the Until recently, the gene for two steps of the 14 steps G258 mutant, we isolated human genomic DNA trans- in the synthesis of the lipid-linked oligosaccharides formants of the G258 mutant, which recovered from (the mannosyltransferase II gene and the mannosyl- both defects by way of cell hybridization with X-ray transferase IV gene) remained to be clarified for a long irradiated HeLa cells. We detected a common 1.3-kb time, but in 2006, Imperiali’s group2) found that the product by inter-human specific sequence in the L1 ALG2 protein (originally attributed to mannosyltrans- (L1Hs) PCR in the transformants (Kataoka et al., ferase III) and the ALG11 protein (originally attributed Somat. Cell Mol. Genet., 24, 235–243 (1998)). In the to mannosyltransferase V) have dual functions. The present study, we screened a human mega yeast ALG2 protein has both mannosyltransferase II and artificial chromosome (YAC) library by PCR with mannosyltransferase III activities, and the ALG11 primers designed according to the 1.3-kb DNA, and protein has both mannosyltransferase IV and mannosyl- selected YAC clone 923f5. Moreover, we found by transferase V activities. Thus all the genes involved in spheroplast fusion that YAC clone 923f5 complemented the 14 steps in the synthesis of the lipid-linked both defects of the G258 mutant. Since the human oligosaccharide were elucidated. counterpart of the yeast ALG11 gene is localized in the We have been attempting to isolate the human and region, the G258 mutant might have a defect in the mouse genes encoding the GDP-mannose dependent mouse ALG11 gene. mannosyltransferase involved in the first half of the lipid-linked oligosaccharide synthesis in the cytoplasmic Key words: mannosyltransferase; asparagine-linked face of the rough ER membrane by expressional glycosylation; lipid-linked oligosaccharide; cloning3) and homology cloning.4,5) yeast artificial chromosome (YAC) library; As for the expression cloning approach, previously spheroplast fusion we isolated temperature-sensitive asparagine-linked glycosylation mutant G258 from the FM3A mouse 3 The lipid-linked oligosaccharide Glc3Man9GlcNAc2- mammary carcinoma cell line by [2- H] mannose P-P-Dol serves as a precursor in the biosynthesis of the suicide selection. The G258 mutant has temperature asparagine-linked oligosaccharide of glycoprotein. The sensitivities in both cell growth and asparagine-linked biosynthesis of lipid-linked oligosaccharide is highly glycosylation6) due to a single gene mutation at a step conserved among eukaryotes and is carried out by 14 in the synthesis of the lipid-linked oligosaccharide.7) glycosyltransferases in an ordered stepwise manner. The Since the G258 mutant can synthesize the full-sized first half of the assembly is done in the cytoplasmic face lipid-linked oligosaccharide (Glc3Man9GlcNAc2-P-P- of the rough ER membrane by glycosyltransferases, Dol) at 33 C but at 39 C, the mutant cells cannot using UDP-GlcNAc and GDP-Man as donors. The latter elongate the lipid-linked oligosaccharide beyond half is done in the luminal face of the rough ER Man1-3(Man1-6)Man1-4GlcNAc1-4GlcNAc1-P-P- membrane by glycosyltransferases, using Dol-P-Man Dol. Therefore the defect in the G258 mutant might and Dol-P-Glc as donors.1) Since it is difficult to obtain reside in the activity or the expression of GDP- acceptor for each enzyme, genetic studies have been Man:Man1-3(Man1-6)Man1-4GlcNAc1-4GlcNAc1- performed to elucidate the process. Mutants that have a P-P-Dol 1,2-mannosyltransferase (mannosyltransferase defect in these steps have been isolated in yeast mainly IV) activity.3,6,7) Since both temperature sensitivities as temperature-sensitive mutants, by the tritium-suicide were due to a single gene mutation, a cDNA or a method and the synthetic lethal method.1) The respective genomic DNA fragment that can complement the

y To whom correspondence should be addressed. Present address: 2-18-7 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan; Tel/Fax: +81-3-5689-2764; E-mail: [email protected] Abbreviations: Dol, Dolichol; FBS, fetal bovine serum; ER, endoplasmic reticulum; YAC, yeast artificial chromosome; L1Hs, human specific sequence in the L1 A Human YAC Complemented a Defect in Asparagine-Linked Glycosylation of a Mouse G258 Mutant 109 temperature sensitivity in cell growth of G258 mutant hygromycin B at 37 C. cells might contain the mannosyltransferase IV gene Mammalian cells were grown in a humidified 5% CO2 atmosphere (putatively designated mat-4).6,7) in a CO2 incubator. Cell viability was measured by trypan blue exclusion. Hence we attempted to isolate the human mannosyl- Yeast cells harboring human mega YAC clone 923f5 were obtained transferase IV gene, Hmat-4, that complements the from CEPH (Paris, France), and were grown at 30 CinSD temperature sensitivity of lipid-linked oligosaccharide URATRP medium (Funakoshi, Tokyo). synthesis of G258 mutant cells by combining the expression cloning and the positional cloning approach. YAC library screening. Human mega YAC libraries (CEPH, Paris, As described in a previous report,3) we isolated the France) were screened by a PCR-based method using LPP1-2F and LPP1-4R primers designed on the basis of a unique sequence in the human genomic DNA transformants of the G258 mutant 7) 0 for several reasons: (i) The population of cDNAs 1.3-kb sequence: LPP1-2F (5 -GCCCAAGTCAAAATACTAG- CAGG-30), LPP1-4R (50-TTGAGGTATAACATAGATGGTAAG-30) encoding the proteins involved in lipid-linked oligosac- (Greiner Japan, Tokyo). All PCR amplifications were done using 50 ng charide synthesis in the cDNA libraries is very low, and of template DNA in a reaction volume of 50 mL including 10 mM Tris– the gene expression involved in lipid-linked oligosac- HCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2,1mM LPP1-2F, 1 mM LPP1- charide synthesis is low in yeast and mammalian 4R, 200 mM dNTP, and 2.5 units of Taq polymerase (Gibco, Rockville, cells8–12) (ii) The G258 mutant shows very low effi- MD). The reactions were carried out through 30 cycles of 94 C for ciency of transfection by ordinary methods such as 45 s (denaturation), 55 C for 1 min (annealing), and 72 C for 1 min calcium phosphate-DNA co-precipitation, electropora- (elongation) in an Astec Thermal Cycler (Program Temp Control System PC700, Fukuoka, Japan). PCR products were analyzed by 1% tion, and lipofection. Even on transfection of a select- agarose gel electrophoresis. EcoT14I-digested phage DNA was used able marker (G418 resistance), the efficiency of trans- as the molecular weight standard. The size of the positive PCR product fection was found to be less than 1 104 (unpublished was 551 bp (from G648 to A1198, Fig. 5 of reference 3). data). (iii) However, it is likely that the population of genomic DNAs encoding proteins involved in lipid- Transformation of the G258 mutant with YAC clone 923f5. linked oligosaccharide synthesis is similar to those of Spheroplast fusion was done by the method of Blunt et al.14) Yeast cells harboring YAC clone 923f5 were grown in SD URATRP genomic DNAs encoding most other proteins. In a 7 3) medium to late log phase (about 2 10 cells/mL) and then diluted previous study, we isolated human genomic DNA with the medium to two-thirds and grown through two doubling cycles fragment transformants of the G258 mutant by the at 30 C. The yeast cells (total, 1 108 cells) were harvested, washed radiation hybrid method using X-ray-irradiated HeLa twice with 1 M sorbitol, and suspended in 5 mL of 1 M sorbitol-0.1 M cells. The resulting transformants (KK-1, KK-3, and sodium citrate-10 mM EDTA-30 mM 2-mercaptoethanol, pH 5.8 (SCE). KK-4) showed recovery from both temperature-sensitive Then 25 mL of Zymolyase (Seikagaku Kogyo, Tokyo, 20 mg/mL in SCE) was added to the suspension with gentle stirring. The suspension growth and temperature-sensitive asparagine-linked glycosylation. Moreover, we detected a common 1.3-kb was incubated at 30 C until 90% of the cells formed spheroplasts as determined by lysis in 5% SDS. The spheroplasts were harvested by inter-L1Hs PCR sequence in the transformants and centrifugation at 4 C at 1,200 rpm for 7 min, washed twice with 5 mL determined the nucleotide sequence of the 1.3-kb inter- of 1 M sorbitol-10 mM Tris–HCl, pH 7.5 (ST), and suspended with L1Hs PCR product of the KK-3 cells. That sequence 500 mL of ice-cold ST on ice. The suspension was added to G258 cells contained 1,279 bases, including one Alu sequence, one (total, 1 107 cells) suspended in serum-free RPMI1640 medium. The mixed cells were harvested by centrifugation at 1,200 rpm for 3 min at (GT)21 repeated sequence, and one Xba I site. Moreover, room temperature, resuspended in 0.5 mL of 50% polyethyleneglycol we detected a unique sequence containing about 700 bp in the 1.3-kb sequence (Fig. 1). The 1.3-kb sequence 1500, and allowed to fuse for 1 min at 37 C. Then the suspension was carefully diluted 2-fold with RPMI1640 + 5% FBS several times. The presumably localizes near the gene that complements diluted cell mixtures were centrifuged at 800 rpm for 5 min at room the temperature-sensitive defects in both cell growth and temperature. The precipitates were suspended in 10 mL of RPMI1640 lipid-linked oligosaccharide synthesis in the G258 medium supplemented with 5% heat-inactivated FBS, streptomycin mutant, and can be used as a marker DNA to isolate (40 mg/mL), and penicillin G (40 U/mL), and were grown at 33 C for the Hmat-4 gene.3) 3 d to allow them to express the genes of the YAC clone 923f5 in the 5 In the present study, by PCR using primers based on YAC transformant. Then about 1 10 cells were plated on each of several soft agar plates, and the plates were incubated at 39 C for 2 the unique sequence in the 1.3-kb sequence, we found a weeks. The surviving colonies were removed with toothpicks and human mega YAC clone 923f5 that contains the 1.3-kb streaked on the soft agar plates. The plates were incubated at 39 C sequence, and found that the YAC clone 923f5 com- for further 2 weeks. Single colony isolation was repeated 3 times. plemented the defects of G258 mutant. Randomly selected YAC clone 923f5 transformants (GYS-1, -2, -14, -17, -20, and -25) were maintained at 39 C throughout the study. The Materials and Methods isolates were checked for the presence of the 1.3-kb DNA sequence by PCR. Five out of the 6 candidate transformants showed the 1.3-kb Cell lines and culture. C3H/He mouse mammary carcinoma cell DNA sequence throughout the study, as explained below under line FM3A clone F28-713) and its temperature-sensitive mutant ‘‘Results.’’ G258,3,6,7) which is defective in the synthesis of lipid-linked oligosaccharides, were routinely maintained at 33 C in RPMI1640 medium PCR analysis of DNAs from mouse A9-human monochromosome (RPMI1640 no. 2, Nissui, Tokyo) supplemented with 5% heat- hybrid cell lines and the YAC clone 923f5 transformants of the G258 inactivated fetal bovine serum (FBS), streptomycin (40 mg/mL), and mutant. DNAs from mouse A9-human monochromosome hybrid cell penicillin G (40 U/mL). Candidate YAC clone 923f5 transformants of lines (JCRB2201, a mouse A9-human monochromosome 1 hybrid cell the G258 mutant were routinely maintained at 39 C (the restrictive line, and JCRB2213, a mouse A9-human monochromosome 13 hybrid temperature for G258 mutants) in the same medium. A mouse A9- cell line), and candidate YAC clone 923f5 transformants of G258 human monochromosome 1 hybrid cell line (JCRB2201) and a mouse mutant were prepared as previously described.15) The presence of the A9-human monochromosome 13 hybrid cell line (JCRB2213) were 1.3-kb DNA in the DNAs of JCRB2201 and JCRB2213 was analyzed purchased from the Japanese Collection of Research Bioresources by PCR using the LPP1-2F primer and the LPP1-4R primer, as (JCRB, Sennan, Osaka, Japan), and were grown in the Dulbecco’s described above. For the analysis of the 1.3-kb DNA in the DNAs of Modified MEM (DMEM) containing 10% FBS and 1 mg/mL of the YAC clone 923f5 transformants of the G258 mutant, the PCR 110 T. MASUDA et al. products (551 bp), using the LPP1-2F primer and the LPP1-4R primer, 5–10 Mb were digested with Xba I. The Xba I digestion yielded both the 439-bp 1.3-kb DNA sequence and the 112-bp fragments. Only the 439-bp fragment was detected on Hmat-4 agarose gel electrophoresis.

Colony formation on soft agar plates. In soft agar culture, LPP1-2F LPP1-4R appropriately diluted cells were plated on RPMI1640 soft agar plates 551 bp containing 0.5% Difco Agar Noble (Becton-Dickinson, Franklin Lakes, NJ), 5% heat-inactivated FBS, streptomycin (40 mg/mL), and L1 3’ region Alu (GT) Xba I penicillin (40 U/mL). The plates were incubated at 33 C and 39 Cin 21 L1 3’ region aCO2 incubator. 100 bp PCR primer

Asparagine-linked glycosylation. The asparagine-linked glycosyla- Fig. 1. Schematic Representation of the Hmat-4 Gene and the 1.3-kb tion of FM3A cells, G258 mutants, and some candidate YAC clone DNA Sequence. 923f5 transformants of the G258 mutant was measured as previously described.3,6,7)

Protein determination. Protein was assayed by the method of Lowry et al.,16) using bovine serum albumin (BSA) as standard.

Radioisotopes. D-[2-3H] mannose (1 mCi/mL in 50% ethanol) was purchased from Moravec Biochemicals (La Brea, CA). 551 bp YAC end isolation and analysis. Five mL of exponentially grown YAC clone 923f5 was harvested, suspended in 5 mL of SCE, and treated with Zymolyase (0.2 mg/mL) for 30 to 60 min at 30 C. DNA was purified from the lysate as previously described.17) The right end of YAC clone 923f5 was amplified from YAC clone 1 2 3 4 5 923f5 DNA by PCR by the Expand Long Template PCR System (Roche Diagnostics, Tokyo). PCR was done from 30 ng of YAC Fig. 2. The 1.3-kb Sequence in DNAs of Mouse A9 Human clone 923f5 genomic DNA in a reaction volume of 100 mL including Monochromosome Hybrid Cells. 1 PCR buffer (1.75 mM MgCl2), 350 mM dNTP mix, 0.3 mM YAC4R PCR analysis was done using PCR primers LPP1-2F and LPP1- (50-GGAAGCTTGGCGAGTCGAGTCGAACGCCCGATCTCAAG-30 4R, as described in ‘‘Materials and Methods.’’ Lane 1, EcoT14I- containing the HindIII site, underlined) and 0.3 mM Alu primer 278 (50- digested DNA (marker); lane 2, DNA from mouse A9-human CCGAATTCGCCTCCCAAAGTGCTGGGATTACAG-30 containing monochromosome 1 hybrid cell line JCRB2201; lane 3, DNA from the EcoRI site, underlined) (Greiner Japan, Tokyo). The reaction was mouse A9-human monochromosome 13 hybrid cell line JCRB2213; carried out through 30 cycles of 94 C for 1 min (denaturation), 60 C lane 4, G258 DNA (negative control); lane 5, HeLa DNA (positive for 45 s (annealing), and 72 C for 1 min (elongation) in the thermal control). cycler. The PCR product was isolated by agarose gel electrophoresis, and was purified with a High Pure PCR Product Purification Kit (Roche Diagnostics, Tokyo). The purified PCR product was ligated Among these, we analyzed two clones whose originated with Sma I-digested pUC118 vector at 16 C overnight using Ligation chromosomes were known at the time: YAC clone 879f4 Kit Ver. 2 (Takara, Kyoto, Japan). The ligated construct was originated from the chromosome 1 and YAC clone transfected to competent JM109. The transformants were plated on 923f5 (1.06 Mb) from the chromosome 13 region q14.3. LB agar containing ampicillin (50 mg/mL), X-Gal, and IPTG. The Hence we did PCR analysis using LPP1-2F and LPP1- transformants forming white colonies were isolated, and their DNAs 15) 4R PCR primers on DNAs from mouse A9-human were prepared as previously described. The size of their integrated monochromosome hybrid cell lines (JCRB2201, con- DNA was confirmed by agarose gel electrophoresis after PvuII digestion. The resulting right end of YAC clone 923f5 was sequenced taining human chromosome 1, and JCRB2213, contain- by a method described elsewhere.3) ing human chromosome 13). In the control experiment, we detected the 551 bp PCR product not in the DNAs of Database search. The Entrez Nucleotides database (http://www.ncbi/ the mouse G258 mutant cells, but those of human HeLa nlm/nih.gov/entrez/query) was used to obtain the nucleotide sequences cells. Thus, we detected the 551-bp PCR product using of the cloned glycosyltransferases. A BLAST search was done using both cellular DNA not from a mouse A9-human monochro- NCBI (http://www.ncbi.nlm.nih.gov/BLAST/index.shtml)andDDBJ/ mosome 1 hybrid cell line, but from a mouse A9-human EMBL/GenBank databases (http://blast.ddbj.nig.ac.jp/top-e.html). The Ensembl Genome Browser (http://www.sanger.ac.uk/resources/databases/ monochromosome 13 hybrid cell line (Fig. 2, lanes 2 ensembl.html) was used to obtain a BAC contig map of YAC clone and 3). Moreover, the 551-bp PCR product on the 923f5. cellular DNA was detected not in G258 mutant cells, but in HeLa cells (Fig. 2, lanes 4 and 5). The 551-bp Results sequence corresponded to the estimated distance between the PCR primer sequences (Figs. 1 and 2). Hence The human mega YAC clone 923f5 contained the 1.3- we concluded that the human mega YAC clone 923f5 kb sequence contains the 1.3-kb sequence. We designed PCR primers (LPP1-2F and LPP1-4R, see ‘‘Materials and Methods’’) based on a unique Isolation of candidate YAC clone 923f5 transformants sequence within the 1.3-kb sequence (Fig. 1),3) and of the G258 mutant screened a human mega YAC library by PCR with The results described above suggest that the human these primers. We detected six positive YAC clones, gene, putatively designated Hmat-4 (human mannosyl- 616e2, 628c3, 682f3, 816e1, 879f4, and 923f5, yielding transferase IV) at the time, that is expected to comple- approximately 551-bp PCR product (data not shown). ment the defect in the G258 mutant, might also localize A Human YAC Complemented a Defect in Asparagine-Linked Glycosylation of a Mouse G258 Mutant 111 on human chromosome 13 region q14.3 in YAC clone A 923f5 (1.06 Mb, 0.03% of the human genome in size). 1 X 106 ab1 X 106 To test this possibility, we transfected the 1.06 Mb 1 X 105 1 X 105 human genomic DNA fragment in the YAC clone 923f5 into G258 mutant cells by spheroplast fusion, and 135 135 isolated candidate transformants (GYS strains) that showed recovery from temperature-sensitive growth, as 1 X 106 c 1 X 106 de1 X 106 described above in ‘‘Materials and Methods.’’ This 5 5 method was based on the fact that the G258 mutant is 1 X 10 1 X 10 1 X 105 temperature sensitive as to both cell growth and lipid 135 135 135 intermediate synthesis, as well as asparagine-linked 7) 1 X 106 f 1 X 106 g 1 X 106 h glycosylation, due to a single gene mutation. / mL Cell number The efficiency of isolation of the transformants was 2:27:7 105. On the other hand, yeast YPH501, 1 X 105 1 X 105 1 X 105 which does not have human mega YAC genomic DNA, did not complement the defect in the G258 mutant under 135 135 135 Time in culture (d) the same conditions. We obtained 28 candidate transformants. Among them, six candidates (GYS-1, -2, -14, B 100 100 -17, -20, and -25) were randomly selected and analyzed 80 further. 80 60 ab60 Recovery of cell growth in the candidate YAC clone 135 135 923f5 transformants of the G258 mutant 100 100 100 Growth recovery was tested by both liquid culture and soft agar culture, as described in ‘‘Materials and 80 80 80 Methods.’’ Growth and viability in the liquid culture at 60 c 60 de60 33 C and 39 C are shown in Fig. 3. The FM3A cells Viability (%) 135 135 135 grew faster at 39 C than at 33 C (Fig. 3A a) and their 100 100 100 viability decreased as they reached the stationary phase (Fig. 3B a). At 33 C, the G258 mutants grew at a rate 80 80 80 similar to the FM3A cells grown at that temperature. 60 f 60 g 60 h However, at 39 C, the growth of the G258 mutants stopped before reaching the maximum cell number 135 135 13 5 achieved at 33 C (Fig. 3A b). The viability of the G258 Time in culture (d) mutants decreased rapidly after a temperature shift to Fig. 3. Cell Growth of YAC Clone 923f5 Transformants of G258 39 C (Fig. 3B b). On the other hand, all the candidate Mutants in Suspension Culture. transformants (GYS-1, -2, -14, -17, -20, and -25) Cells were grown as described in ‘‘Materials and Methods.’’ showed growth patterns in both growth curve and A, growth curve; B, viability. a, FM3A cells; b, G258 mutants; viability, similarly to the FM3A cells at both 33 C and c, GYS-1 cells; d, GYS-2 cells; e, GYS-14 cells; f, GYS-17 cells; 39 C (Fig. 3A c to h and Fig. 3B c to h). Growth on soft g, GYS-20 cells; h, GYS-25 cells. Symbols: ------,at33 C; ------,at39C. agar plates at 33 C and 39 C is shown in Table 1. The FM3A cells showed colony forming activity at 39 C similar to that at 33 C. The G258 mutant cells formed Table 1. Cell Growth of Candidate YAC Clone 923f5 Transformants colonies at 33 C but not at 39 C. All the candidate of the G258 Mutant on Soft Agar Plates transformants showed comparable efficiency of colony Average colony number Candidate Relative plating formation at 33 C and 39 C. transformants 33 C39C efficiency Recovery from temperature-sensitive asparagine- GYS-1 518 276 0.53 linked glycosylation in the candidate YAC clone 923f5 GYF-2 568 646 1.14 transformants GYS-14 816 512 0.63 GYS-17 548 428 0.78 Of the candidate transformants, the GYS-1, -14, -17, GYS-20 658 741 1.13 and -20 cells were further analyzed for reversion of GYS-25 270 142 0.53 temperature-sensitive asparagine-linked glycosylation, FM3A 697 702 1.01 as shown in Fig. 4. The FM3A cells showed increased G258 524 36 0.06 incorporation of [2-3H] mannose into the acid-insoluble fraction at 39 C as compared to 33 C (Fig. 4a). The average colony number at 39 C/average colony number at 33 C G258 mutants showed decreased incorporation of [2-3H] mannose into the acid-insoluble fraction at 39 Cas Occurrence of the 1.3-kb sequence in the candidate compared to 33 C and to FM3A cells at both temper- YAC clone 923f5 transformants of the G258 mutant cells atures (Fig. 4b). All the candidate YAC clone 923f5 Among the candidate YAC clone 923f5 transformants transformants of the G258 mutants showed recovery of of the G258 mutant cells tested, GYS-1, -14, -17, and [2-3H] mannose incorporation at 39 C, although the -20 showed reversion of both temperature-sensitive levels of recovery were different among them (Fig. 4, growth (Table 1) and asparagine-linked glycosylation c to f). (Fig. 4). It was necessary to show the presence of the 112 T. MASUDA et al.

XbaI 20 a 15 b

10 5 g protein)

µ 246 246 551 bp 439 bp

40 20 c d 123456 78 910111213141516171819

20 Fig. 5. The 1.3-kb Marker DNA in the YAC Clone 923f5 Trans- 10 formants. DNAs from mouse G258 mutants (negative control), HeLa cells

H-mannose (dpm / (positive control), and candidates YAC clone 923f5 transformants 3 246 24 6 (GYS-1, -2, -14, -17, -20, and -25) were analyzed by PCR using primers LPP1-2F and LPP1-4R. Aliquots of the PCR products (551-bp) were also digested with Xba I. Lanes 2, 4, 7, 9, 11, 13, 15, 80 efand 17 show PCR products of DNAs from the G258 mutant, HeLa 10 cells, GYS-1, GYS-2, GYS-14, GYS-17, GYS-20, and GYS-25 respectively. Lanes 3, 5, 8, 10, 12, 14, 16, and 18 show the Xba I- Incorporated 40 5 digested PCR products (439-bp) of the G258 mutant, HeLa cells, GYS-1, GYS-2, GYS-14, GYS-17, GYS-20, and GYS-25 respectively. Lanes 1, 6, and 19 show EcoT14I digested DNAs. 24 6 246 Labeling time (h) Table 2. Summary of the Characterization of YAC Clone 923f5 Transformants of G258 Mutant Fig. 4. Incorporation of [2-3H] Mannose into the TCA-Insoluble Fraction of YAC Clone 923f5 Transformants of G258 Mutant. Candidate Recovery of Recovery of Occurrence of a, FM3A cells; b, G258 mutants; c, GYS-1 cells; d, GYS-14 cells; transformants cell growth N-glycosylation 1.3-kb DNA e, GYS-17 cells; f, GYS-20 cells. Symbols: ------,at33C; ------,at39C. GYS-1 + + + GYF-2 + ND + GYS-14 + + + 1.3-kb sequence to confirm that they are the YAC clone GYS-17 + + + 923f5 transformants of G258 mutant cells. Hence the GYS-20 + + + GYS-25 + ND occurrence of the 551-bp PCR product and the cleav- ability of the 551-bp PCR product by Xba I digestion G258 were tested in these GYS strains (Fig. 5). The G258 ND, not determined mutant DNA (negative control) did not yield the 551-bp PCR product (lanes 2 and 3). On the other hand, HeLa DNA (positive control) yielded the 551-bp PCR product succeeded in amplifying the right end of YAC clone and its Xba I-digested 439-bp fragment (lanes 4 and 5). 923f5 by PCR using YAC4R and the Alu primer The DNAs from GYS-1, -14, -17, and -20 also yielded (primers 278, 517, etc.). The PCR products showed the the 551-bp PCR product and its Xba I-digested 439-bp expected sizes based on the positions of the primers in fragment (lanes 7, 8, 10 to 16). The DNAs from GYS-2 the Alu sequence. The PCR product using the YAC4R yielded a faint Xba I-sensitive 551-bp PCR product, but and 278 primers was ligated with the pUC118 and the DNAs from GYS-25 did not yield the Xba I- sequenced. The entire human genomic DNA sequence in sensitive 551-bp PCR product (lanes 9, 10, 17 and 18). the PCR product was almost the same as the partial sequence (from A36252 to G36643) of BAC clone RP11- Identification of the YAC clone 923f5 transformants of 431O22 (AL139085), which appears in the human the G258 mutant genomic DNA database (Fig. 6). On the other hand, The results of the present study are summarized in we found that BAC clone RP11-64P12 (AL158066) Table 2. Judging from the reversion of both temper- contained the 1.3-kb DNA. Based on the size of the ature-sensitive cell growth and temperature-sensitive human genomic DNA in YAC clone 923f5 (1.06 Mb), asparagine-linked glycosylation and from the presence we made a BAC contig map of YAC clone 923f5. In of the 1.3-kb sequence, we concluded that GYS-1, -14, addition, we found that the YAC clone 923f5 can -17, and -20 are genuine YAC clone 923f5 transform- complement the defect in the asparagine-linked glyco- ants of the G258 mutant. In other words, the YAC clone sylation of G258 mutant. Hence we concluded that the 923f5 complemented the defect of lipid-linked oligosac- Hmat-4 gene, originally defined as human mannosyl- charide synthesis in the G258 mutant. transferase IV gene, might reside in the human genomic DNA of YAC clone 923f5. YAC end isolation and analysis of it To localize the 1.3-kb sequence in the YAC clone Discussion 923f5, we attempted to obtain both ends of the insert in YAC clone 923f5 by PCR using the sequence of the Although yeast genetic studies were mainly per- cloning sites of YAC vector YAC4 (YAC4R and formed to analyze the steps in the synthesis of lipid- YAC4L) and the Alu sequences. As a result, we linked oligosaccharides, mutants with defects in man- A Human YAC Complemented a Defect in Asparagine-Linked Glycosylation of a Mouse G258 Mutant 113 of the Saccharomyces cerevisiae mannosyltransferase V gene (ALG11)20) also localized in BAC clone RP11- 248G5 (AL139082), right next to the BAC clone RP11- 64P12 (AL158066). It would be interesting to determine whether the human counterpart (Hmat-5) of yeast ALG11 and the human gene (Hmat-4) that complements the defect of the G258 mutant is the same. In this context, it is worthy of note that the G258 mutant stops the synthesis of lipid-linked oligosaccharide at Man1- 3(Man1-6)Man1-4GlcNAc1-4GlcNAc1-P-P-Dol at 3,6,7) ALG11 Fig. 6. Nucleotide Sequence of the Right End of YAC Clone 923F5. 39 C, while the yeast gene disruptant The sequences for forward primer YAC4R and reverse primer 278 can synthesize Man1-2Man1-3(Man1-6)Man1- are underlined. 4GlcNAc1-4GlcNAc1-P-P-Dol but not Man1-2Man1- 2Man1-3(Man1-6)Man1-4GlcNAc1-4GlcNAc1-P- nosyltransferase II, mannosyltransferase VI, and man- P-Dol, making it possible to synthesize a truncated form, 20) nosyltransferase V had not been isolated at the time we Glc3Man8GlcNAc2-P-P-Dol. Hence the yeast ALG11 isolated the mouse G258 mutant, which has a temper- gene was originally defined as the mannosyltransferase ature-sensitive defect in both the synthesis of lipid- V gene.20) linked oligosaccharide and cell growth due to a single Although we could not draw any conclusion at the gene mutation. The lesion in the G258 mutant appeared time, we suggested a possible relationship between reside in the mannosyltransferase IV activity. Therefore Hmat-5 and Hmat-4 based on the available data and the G258 mutant might be a useful tool for the study of information, as follows: Case 1-Both the Hmat-4 gene mannosyltransferase IV.3,6,7) and the Hmat-5 gene are the same gene that is localized Previously we isolated by the radiation hybrid method in the YAC clone 923f5. The gene product has both human genomic DNA transformants that recovered mannosyltransferase IV activity and mannosyltransfer- from temperature sensitivities in both cell growth and ase V activity. This means that mannosyltransferase V asparagine-linked glycosylation of the G258 mutants, can acts not only on Man1-2Man1-3(Man1- and we detected a 1.3-kb DNA product due to inter- 6)Man1-4GlcNAc1-4GlcNAc1-P-P-Dol, but also on L1Hs PCR in most of the transformants. Man1-3(Man1-6)Man1-4GlcNAc1-4GlcNAc1-P-P- In the present study, we attempted to use the 1.3-kb Dol as a -1,2 mannosyltransferase. Moreover, the sequence as a marker DNA to isolate and characterize Hmat-4/Hmat-5 (Hmat-4/5) gene (mannosyltransferase human gene Hmat-4, which complements the defect IV/V gene) can complement the temperature-sensitive of lipid-linked oligosaccharide synthesis in the G258 defect of the G258 mutant. However, there is a mutant. Hence we screened a human mega YAC library discrepancy between the sizes of the lipid-linked by PCR using primers designed on the unique region in oligosaccharides of the G258 mutant synthesized at the 1.3-kb DNA sequence. We detected YAC clone 39 C and that of the yeast ALG11 gene disruptant. Case 923f5 (1.06 Mb), which has human genomic DNA 2-The genes are different from each other. In this case, derived from human chromosome 13q14.3. By spher- there are several possibilities. (a) The Hmat-5 gene and oplast fusion of YAC clone 923f5 with the G258 mutant, the Hmat-4 gene are co-localized separately (independ- we isolated the YAC clone 923f5 transformants of the ently) in YAC clone 923f5. However, according to G258 mutant. The transformants showed recovery from bioinformatic research, we did not detect any manno- the temperature-sensitive defects in both cell growth and syltransferase-like sequences in this region, besides the asparagine-linked glycosylation shown by the G258 Hmat-5 gene. (b) The Hmat-5 gene is localized in YAC mutant (Figs. 3 and 4, Tables 1 and 2). clone 923f5 and the Hmat-4 gene is localized elsewhere. Moreover, among the tested transformants, GYS-1, This means the G258 mutant might have a mutation not -14, -17, and -20 showed the presence of the 1.3-kb in the Mmat-4 gene, but in the Mmat-5 gene. DNA (Fig. 5, Table 2). These results suggest that the To clarify this, enzymological or genetic studies YAC clone 923f5 contains the gene that complements are needed. In 2006, Imperiali’s group showed that the the defect in the synthesis of lipid-linked oligosaccha- yeast ALG11 protein has dual functions, and has both ride of the G258 mutants. We concluded at the time that mannosyltransferase IV and mannosyltransferase V the Hmat-4 gene might reside in YAC clone 923f5, activities.2) Thus the G258 mutant might have defects which has 1.06 Mb of human genomic DNA. It is worthy in the expression or enzyme activity of the Mmat-4/5 of note that complementation studies by spheroplast gene. Complementary DNA- and genomic DNA- fusion have been few in number.18,19) sequencing studies of the Mmat-4/5 gene of the G258 By BLASTn search, we detected the nucleotide are in progress. sequence of the 1.3-kb DNA in BAC clone RP11- 64P12 (AL158066). Also we detected the right-end Acknowledgments sequence of YAC clone 923f5 in BAC clone RP11- 248G5 (AL139082). Then we constructed a BAC contig This study was supported in part by the Grants-in-Aid map around the YAC clone 923f5 by BLASTn search; from the Ministry of Education, Culture, Sports, the centromere, (RP11-327P2)-(RP11-381L18)-(RP11- Science, and Technology of Japan, the Kihara Memorial 248G5)-(RP11-64P12)-(RP11-245D16)-(RP11-78J21)- Foundation, the Astellas Foundation for Research on (RP11-93H24)-(RP11-431O22). When making the con- Medicinal Resources (formerly the Fujisawa Founda- tig map, we noticed that the human counterpart (Hmat-5) tion), and the New Energy and Industrial Technology 114 T. MASUDA et al. Development Organization (NEDO) of Japan as part 8) Albright CF and Robbins PW, J. Biol. Chem., 265, 7042–7049 of the R&D Project of the Industrial Science and (1990). Technology Frontier Program (to Y.N.). We express 9) Jackson BJ, Kukuruzinska MA, and Robbins P, Glycobiology, 3, 357–364 (1993). thanks to Professor Yoshitaka Nagai (Riken, Wako, 10) Lennon K, Pretel R, Kesselheim J, te Heesen S, and Japan) and Professor Naoyuki Taniguchi (Osaka Uni- Kukuruzinska MA, Glycobiology, 5, 633–642 (1995). versity School of Medicine, Osaka, Japan), for encour- 11) Lehrman MA, Zhu X, and Khounlo S, J. Biol. Chem., 263, agement, to Dr. Katsuyuki Hashimoto (Division of 19796–19803 (1988). Genetic Resources, National Institute of Infectious 12) Mudgapalli A, Roy KS, Holmes EH, and Vijay IK, J. Biol. Diseases, Tokyo) for a supply of DNAs from mouse Chem., 269, 11327–11336 (1994). A9-human monochromosome hybrid cell lines, and 13) Koyama H and Kodama H, Cancer Res., 42, 4210–4214 (1982). 14) Blunt T, Taccioli GE, Priestley A, Hafezparast M, Mcmillan T, to Dr. T. Takahashi, Tokai University, Hiratsuka, Liu J, Cole CC, White J, Alt FW, Jackson SP, Schurr E, Kanagawa, Japan for his contribution to the early part Lehmann AR, and Jeggo PA, Genomics, 30, 320–328 (1995). of this study. 15) Sambrook J, Fritsch EF, and Maniatis T, ‘‘Molecular Cloning, a Laboratory Manual’’ 2nd ed., Cold Spring Harbor Laboratory References Press, Cold Spring Harbor (1989). 16) Lowry OH, Rosebrough NJ, Farr AL, and Randall RJ, J. Biol. 1) Burda P and Aebi M, Biochim. Biophys. Acta, 1426, 239–257 Chem., 193, 265–275 (1951). (1999). 17) Reeves RH, Cabin DE, and Lamb B, Curr. Protoc. Hum. Genet., 2) O’Reilly MK, Zhang G, and Imperiali B, Biochemistry, 45, 5.12.1–5.12.23 (2001). 9593–9603 (2006). 18) Ju¨licher K, Marquitan G, Werner N, Bardenheuer W, Vieten L, 3) Kataoka K, Takahashi T, Ayusawa D, and Nishikawa Y, Somat. Bro¨cker F, Topal H, Seeber S, Opalka B, and Schu¨tte J, J. Natl. Cell Mol. Genet., 24, 235–243 (1998). Cancer Inst., 91, 1563–1568 (1999). 4) Takahashi T, Honda R, and Nishikawa Y, Glycobiol., 10, 321– 19) Murakami Y, Nobukuni T, Tamura K, Maruyama T, Sekiya T, 327 (2000). Arai Y, Gomyou H, Tanigami A, Ohki M, Cabin D, 5) Simono N, Nishimura Y, Kamiguchi H, and Nishikawa Y, Frischmeyer P, Hunt P, and Reeves RH, Proc. Natl. Acad. Biosci. Biotechnol. Biochem., 75, 1451–1455 (2011). Sci. USA, 95, 8153–8158 (1998). 6) Nishikawa Y, J. Cell. Physiol., 119, 260–266 (1984). 20) Cipollo LF, Trimble HB, Chi JH, Yan Q, and Dean N, J. Biol. 7) Nishikawa Y, Biochim. Biophys. Acta, 1091, 135–140 (1991). Chem., 276, 21828–21840 (2001).