Germ Cell Survival Through Carbohydrate-Mediated Interaction with Sertoli Cells Tomoya O. Akama, Hiroaki Nakagawa, Kazuhiro Sugihara, Sonoko Narisawa, Chikara Ohyama, Shin-Ichiro Nishimura, Deborah A. O'Brien, Kelley W. Moremen, José Luis Millán, and Michiko N. Fukuda

Construction of targeting vector

A mouse genomic P1 clone 5911 (Genome Systems, St. Louis, MO) encoding Man2a2 (SVJ129 strain) was isolated by PCR-based screening. P1 plasmid DNA was digested by Avr II and Eco RI, subcloned into pBluescript vector, and sequenced. To prepare the knockout vector, a Hind III-Xba I fragment containing exons 1 to 8 of Man2a2 gene was cloned into pBluescript II. PGK neo was replaced with the region of exon 2, which contains the translation initiation codon. A diphtheria toxin (DTA) sequence was added at the 5´ region of Man2a2. The vector DNA was linearized with Sal I and used to transfect the mouse ES cell line CJ7.

Production of MX mutant mice and genotyping

Cloned ES cells selected as homologous recombinants were injected into C57BL/6 blastocysts. The blastocysts were returned to the uterus of pseudopregnant CD-1 females at 3.5 dpc. Resultant chimeric male mice were mated with wild-type 129SV/J female mice to produce heterozygous offspring. Genotyping was carried out by PCR using DNA prepared from tail biopsies, using primers for Man2a2: 5´-GTGTGTGGGGCTGCTATCTTC-3´ and 5´-ATTGCTGCTTTTCACTTCCTG-3´. Amplification reactions were carried out in a PTC-100 Thermal Cycler (MJ Research, Watertown, MA) by 10-min denaturation at 94°C before cycling, 35 cycles of denaturation at 94°C for 30 s, annealing at 58°C for 30 s, extension at 72°C for 45 s, and further extension at 72°C for 5 min. PCR reactions and primers for neo were described previously (6).

TUNEL assay

Testes were fixed in 10 % formalin-PBS and embedded in O.C.T. compound (Sakura Finetek, Torrance, CA). Nine-micron sections were stained by the TUNEL method (5) using ApopTag Peroxidase in situ Apoptosis Detection Kit (Intergen Company, Purchase, NY). Counterstaining was performed using methyl green.

Antibody against mouse MX

Polyclonal antibodies against mouse MX were raised by immunizing rabbits with a synthetic peptide SRIKDSVLELTANAEGC. This sequence corresponds to amino acid residues 69 to 84 of mouse MX, with a cysteine residue added to the COOH-terminus for conjugation to keyhole limpet hemocyanin. Antibody was purified on a protein A Sepharose column followed by affinity purification using Sulfo-link gel (Pierce, Rockford, IL) conjugated with the synthetic peptide used for the immunization.

Structural analysis of N-glycans

Methods for two-dimensional HPLC analysis for N-glycans (7, 8) were applied. Mouse testes (~50 mg) were homogenized in 2 ml of 0.1 M NH₄HCO₃ on ice and boiled for 10 min. The sample was treated twice with a mixture of chloroform/methanol, 2/1 (v/v), and insoluble materials were lyophilized. The sample was digested with 0.1 mg each trypsin and chymotrypsin per 10 mg lyophilized sample, at 37°C for 16 hours, and was boiled for 10 min. N-glycans were released from the glycopeptides by incubating with glycoamidase F (1 Unit/mg) (Roche, Indianapolis, IN) at 37°C for 24 hours. The released N-glycans were purified on a Bio-Gel P-4 column (1.0 × 38 cm) equilibrated with water and were conjugated with 2-aminopyridine by using sodium cyanoborohydride (9). After desalting by Sephadex G-15 gel filtration and desialylation with mild acid hydrolysis, PA-oligosaccharides were analyzed by an HPLC (Shimazu, model LC-10A) equipped with a reversed-phase column (Shimazu, HRC-ODS, 6 × 150 mm) equilibrated with 80% solvent A (10 mM Na-phosphate buffer, pH 3.8) and 20% solvent B (0.5% 1-butanol added to solvent A) (v/v). Samples were eluted at 55°C in a gradient from 20% to 60% of solution B over 80 min at flow rate of 1.0 ml/min. PA-oligosaccharides were detected by fluorescence using excitation at 320 nm and emission at 400 nm. Each oligosaccharide eluted from the ODS column was further separated by an amide-adsorption column (Amide-80, 4.6 × 250 mm, TOSOHAAS), in a gradient from 100% solvent C (35% of 0.5 M acetic acid-triethylamine buffer, pH 7.3, and 65% of acetonitrile) to 40% solvent C and 60% solvent D (50% of 0.5 M acetic acid-triethylamine buffer, pH 7.3, and 50% of acetonitrile) over 30 min at 40°C at a flow rate of 1.0 ml/min. The HPLC columns were calibrated using a mixture of PA-isomalto-oligosaccharides (degree of polymerization, 4 to 15) (Seikagaku, Tokyo, Japan). N-glycan standards were prepared from chicken ovalbumin, human transferrin, human ₁-acid glycoprotein, human serum IgG, and porcine thyroglobulin (all from Sigma) as described (7, 8). Each N-glycan was digested by appropriate exoglycosidases as described previously (10). The enzymes used in this study were -galactosidase, -N-acetylhexosaminidase and -mannosidase from jack bean (Seikagaku), and -L-fucosidase from beef kidney (Roche).

In vitro assay for germ cell-Sertoli cell adhesion

Methods described for rat and mouse spermatogenic cells and Sertoli cell adhesion assays (11, 12) were applied. Although the reported adhesion assays were performed at 32°C, we performed the assay at 37°C, because small numbers of MX^-/- spermatogenic cells adhered to Sertoli cells at 32°C presumably by an additional mechanism independent from MX defect. Sertoli cells were prepared from 3-week-old mouse testis. Spermatogenic cells were prepared from 8-week-old mouse testis, and enriched by Percoll density gradient centrifugation as described (13). The spermatogenic cells were metabolically labeled with [³H]galactose (10 Ci/ml DME low glucose medium) at 37°C for 1 hour. No differences were detected in the labeling efficiency between MX^+/+ and MX^-/- spermatogenic cells. Adhesion assay was performed using Sertoli cells cultured in 3.5-cm tissue culture dishes. [³H]-labeled spermatogenic cells 50,000 cpm (~2 × 10⁵ cells) were added to the Sertoli cells, incubated at 37°C for 4 hours. After washing unbound cells with PBS, radioactivity remaining on the Sertoli cells was counted in a scintillation counter. Adhesion inhibition assay with glycopeptides was performed using Sertoli cells cultured in a 24-well plate. [³H]-labeled spermatogenic cells 25,000 cpm (~1 × 10⁵ cells) were added to each well together with the glycopeptides, and incubated at 37°C for 4 hours. After washing the wells, remaining radioactivity was counted. For adhesion inhibition assay with purified PA-oligosaccharides, [³H]-labeled spermatogenic cells 2400 cpm (~1 × 10⁴ cells) with or without 10 pmol PA-oligosaccharide in 5 l -MEM were added to Sertoli cells grown on 3-mm² cover slips (SurgiPath, Richmond, IL). After incubating at 37°C for 2 hours, nonadhering spermatogenic cells were washed with PBS, and radioactivity remaining on Sertoli cells was counted.

References for Supplemental Information

1. R. Kornfeld, S. Kornfeld, Annu. Rev. Biochem. 54, 631 (1985).
2. M. Oh-eda et al., Eur. J. Biochem. 268, 1280 (2001).
3. F. Nantel et al., Nature 380, 159 (1996).
4. C. G. Print et al., Proc. Natl. Acad. Sci. U.S.A. 95, 12424 (1998).
5. W. Gorczyca, J. Gong, Z. Darzynkiewicz, Cancer Res. 53, 1945 (1993).
6. K. Sugihara et al., Oncogene 17, 3427 (1998).
7. N. Takahashi, N. Tomiya, in Handbook of Endoglycosidases and Glycoamidases, N. Takahashi and T. Muramatsu, Eds. (CRC Press, Boca Raton, 1992), pp. 199-332.
8. H. Nakagawa, Y. Kawamura, K. Kato, I. Shimada, N. Takahashi, Anal. Biochem. 226, 130 (1995).
9. S. Hase, T. Ibuki, T. Ikenaka, J. Biochem. (Tokyo) 95, 197 (1984).
10. K. Yamashita, N. Koide, T. Endo, Y. Iwaki, A. Kobata, J. Biol. Chem. 264, 2415 (1989).
11. R. M. DePhilip, D. G. Danahey, Biol. Reprod. 37, 1271 (1987).
12. S. Kohno, E. Ziparo, L. F. Marek, K. S. Tung, J. Reprod. Immunol. 5, 339 (1983).
13. M.-C. Hoffmann, S. Narisawa, R. Hess, J.-L. Millan, Exp. Cell Res. 201, 417 (1992).

Supplemental Figure 1. Involvement of MX in N-glycan biosynthesis. Complex-type N-glycans are produced from high-mannose-type precursors through hybrid type N-glycans (1). In CHO cells overexpressing MX, levels of Man₆GlcNAc₂ were reduced and levels of Man₄GlcNAc₂ were increased, suggesting an involvement of MX in N-glycan processing (2). MX potentially provides an alternate pathway independent of MII. In MX^-/- testis, HF5.11 (a specific substrate for MII) is reduced and 110.2 (the product of MII) is increased (Fig. 3C in the text of this report), consistent with the idea that MX inhibits MII (2). This study shows that Man2a2 null mutation affects synthesis of GlcNAc-terminated N-glycans but not Gal-terminated N-glycans in the mouse testis. GlcNAc-terminated glycans may be produced primarily in spermatogenic cells, whereas Gal-terminated glycans may be synthesized in nonspermatogenic cells. It is also possible that GlcNAc-terminated glycans and Gal-terminated glycans are carried by distinct proteins, and they are differentially accessible to MX or MII. These possibilities should be defined by future studies.

Medium version | Full size version

Supplemental Figure 2. Targeted disruption of the mouse Man2a2, genomic analyses of ES cells and mutant mice, and Western blot analysis of MX proteins in wild-type and mutant mouse testes. (A) Partial genomic structure of the mouse Man2a2, including exons 1 through 15. The translation initiation ATG codon is present in exon 2. The targeting vector was constructed by replacing the genomic region containing the translation initiation site with the neomycin resistance gene (neo^r), and the diphtheria toxin gene (DTA) was used as a negative selection for homologous recombination. The genomic DNA from a homologous recombinant ES clone gives 6- and 12-kb fragments upon digestion with Eco RV, while the wild-type gene produces a single 18-kb fragment. These Eco RV fragments can be distinguished by the 5´ probe. Each pair of arrowheads shows PCR primers for genotyping shown in (C). (B) Southern blot analysis of genomic DNAs from untransfected ES cells (+/+) and a homologous recombinant ES clone (+/-) using a 5´ probe [see (A)]. (C) Genotyping of mouse offspring by PCR. Mouse tail DNAs were subjected to PCR for the neo gene (upper row) and Man2a2, the MX gene (lower row). Wild-type (neo^-, Man2a2⁺), heterozygote (neo⁺, Man2a2⁺) and homozygote (neo⁺, Man2a2^-) are shown. The positions of PCR primers are shown by arrowheads in (A). (D) Western blot analysis of lysates (2 g protein/lane) from MX^+/+ and MX^-/- mouse testes probed with polyclonal antibodies against mouse MX.

Medium version | Full size version

Supplemental Figure 3. Weights of testes from MX^+/+, MX^+/-, and MX^-/- mice. The wet weight of each pair of testes was measured at 8 weeks of age, and the weights were averaged.

Medium version | Full size version

Supplemental Figure 4. Multinucleated giant cells appeared in the MX-deficient testis and increased apoptosis in MX^-/- spermatogenic cells. (A) The testes from MX^-/- mice exhibited multinucleated giant cells, which are indicated by arrows. Similar multinucleated cells were seen in the testes of CREM-deficient mice (3) and Bcl-w-deficient mice (4), which showed elevated apoptosis. Scale bar, 50 m. (B) A TUNEL assay (5) on MX^+/+ and MX^-/- littermates testes detected elevation of apoptosis in MX^-/- testis. Apoptotic cells in stage XII were stained in brown color. This degree of apoptosis in MX^-/- testis is not as extensive as that in Bcl-w-deficient testis (4). Scale bar, 50 m.

Medium version | Full size version