An Arabidopsis Thaliana Cdna Complements the N-Acetylglucosaminyltransferase I Deficiency of Cho LEC1 Cells

Biochemical and Biophysical Research Communications 261, 829 – 832 (1999) Article ID bbrc.1999.1117, available online at http://www.idealibrary.com on

An Arabidopsis thaliana cDNA Complements the N-Acetylglucosaminyltransferase I Deficiency of CHO Lec1 Cells Hans Bakker,* ,1 Arjen Lommen,† Wilco Jordi,‡ Willem Stiekema,* and Dirk Bosch* *Department of Molecular Biology, Centre for Plant Breeding and Reproduction Research (CPRO-DLO), P.O. Box 16, 6700AA Wageningen, The Netherlands; †Department of Natural Constituents, Residues and Contaminants, State Institute for Quality Control of Agricultural Products (RIKILT-DLO), P.O. Box 230, 6700AE Wageningen, The Netherlands; and ‡DLO-Research Institute for Agrobiology and Soil Fertility (AB-DLO), P.O. Box 14, 6700AA Wageningen, The Netherlands

Received July 6, 1999

N-Acetylglucosaminyltransferase I (GlcNAcT-I, EC 2.4.1.101) is the enzyme which initiates the formation of complex N-linked glycans in eukaryotes by transforming GlcNAc to the oligo-mannosyl acceptor Man 5GlcNAc 2-Asn. The enzymatic activity and the structure that is synthesised by this enzyme are found in animals and plants but not in yeast. cDNAs encoding the enzyme have already been cloned from several mammals and the nematode Caenorhabditis elegans. In this article the cloning of an Arabidopsis thaliana GlcNAcT-I cDNA with homology to animal cDNAs is described. By expression of the plant cDNA in CHO Lec1 cells, a mammalian cell line deficient in GlcNAcT-I, it was shown that it encodes an active enzyme with the same enzymatic activity as the animal homologue. It has already been shown that a human GlcNAcT-I can complement an A. thaliana mutant (cgl-1). Here it is shown that the reverse is also true, the plant glycosyltransferase is able to complement a mammalian mutant (Lec1) deficient in GlcNAcT-I. © 1999 Academic Press

Animal and plant glycoproteins contain both highmannose and complex N-linked glycans, while complex N-linked glycans are absent in yeast. The enzyme initiating the formation of complex N-linked glycans is N-acetylglucosaminyltransferase I (GlcNAcT-I, EC 2.4.1.101). After the transfer of GlcNAc from UDPGlcNAc to the oligomannosyl acceptor Man 5GlcNAc 2Asn by this enzyme, complex N-linked glycans are formed by the successive action of several other glycosyltransferases (1). The enzyme activity and the structure formed by GlcNAcT-I have been found in invertebrates (2– 4), vertebrates (5) and plants (6, 7), but are 1

Corresponding author. Fax: 31-317-418094. E-mail: h.bakker@ cpro.dlo.nl.

absent in yeast (8). Mammalian cell lines deficient in GlcNAcT-I (9 –11) are unable to form complex type glycans, but the cellular metabolism does not seem to be hampered by this defect. However, the same defect in mice causes severe developmental disorders and embryos do not survive beyond 10.5 days (12). Surprisingly, an Arabidopsis thaliana mutant (cgl-1) deficient in GlcNAcT-I activity does not show any obvious abnormalities (6), indicating that complex glycans are not essential for normal development of this plant. The A. thaliana mutant could be complemented by a human GlcNAcT-I cDNA (13). Genes and cDNA’s encoding GlcNAcT-I have been cloned from several mammalian species (5, 14 –16) and C. elegans (17). In mammals only one gene encoding the enzyme has been found for each species while three genes have been identified in C. elegans. At least two of these three genes encode an active GlcNAcT-I (17). In this report we describe the cloning of an A. thaliana GlcNAcT-I cDNA. 2 By expression of this cDNA in Lec1 Chinese hamster ovary cells (9, 10), a cell line that is deficient in GlcNAcT-I, it was proved to encode an active GlcNAcT-I. MATERIALS AND METHODS Construction of an Arabidopsis thaliana cDNA library. Arabidopsis thaliana (var. columbia) RNA was isolated from young seliques according to Brusslan and Tobin (18). Poly(A) 1 RNA was isolated using an mRNA purification kit (Pharmacia) and oligo(dT) primed cDNA was synthesised by Invitrogen’s copy kit. After addition of BstXI linkers (Invitrogen) to the cDNA the size fraction of .800 base pairs was prepared by agarose gel electrophoresis and ligated into the mammalian expression vector pABE (19), a derivative of pCDM8 (Invitrogen) containing b-lactamase instead of SupF. The ligation mix was electroporated into E. coli TOP10 cells (Invitro-

2

The A. thaliana cDNA sequence has been submitted to the EMBL/GenBank database under Accession No. AJ243198.

829

0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

Vol. 261, No. 3, 1999

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 1. Complementation of CHO-Lec1 cells by Arabidopsis thaliana GlcNAcT-I cDNA. Left: Lec1 cells transiently transfected with an unrelated control cDNA. Right: Lec1 cells transfected with A. thaliana GlcNAcT-I cDNA. Cells were stained with PHA as described in “Materials and Methods.”

gen). The library, containing a total of about 3 3 10 5 independent clones, was plated out in portions of 500-1000 clones that were stored as separate pools. Cloning of the GlcNAcT-I cDNA. Based on an A. thaliana genomic DNA sequence (BAC end; GenBank acc. B24856), two oligonucleotide primers (forward: GACCAATTCAAGGTTCTGACCT and reverse: TGGCCCAACCAGGAATACAC) were designed that potentially amplified a fragment of 209 base pairs. Initially pools of 5,000 different clones of the cDNA library were screened by PCR using this primer pair. The size and orientation of the cDNA insert of positive pools was determined by using the same primers in combination with primers located in the vector arms. One pool of cDNA clones that gave fragments of the expected size in the PCR reactions was further screened. First subpools of 500 clones (the pool of 5000 was composed of 10 smaller pools) were screened by PCR. E. coli cells from a positive subpool were plated and 4 3 384 individual colonies were picked in 384-well plates using a Flexus (Genomic solutions) colony-picking robot. Cells collected from each plate were again screened by PCR. From a positive plate, cells combined from rows were screened followed by cells from individual wells. This way a single cDNA clone positive in the PCR reaction was designated. The derived clone was subcloned in vector pSK 1 (Stratagene) and completely sequenced in both directions using vector and specific primers. Transient transfection and lectin staining of CHO cells. Chinese hamster ovary (CHO) cell line Lec1-15B (10) was maintained in

a-MEM (Life-Technologies) containing 8% newborn and 2% foetal bovine serum. Cells were seeded in 6 well plates and transfected using DOSPER transfection reagent (Roche) according the manufacturer’s protocol. Three days after transfection cells were fixed in glutaraldehyde (2.5% in TBS) blocked with blocking reagent (NEN Life Science Products) in TBS and lectin stained with biotinylated Phasseolus vulgaris agglutinin (PHA-E 1 L, 5 mg/ml) followed by alkaline-phosphatase conjugated streptavidin (5 mg/ml) (both Vector Laboratories) with intermediate washing in PBS 0.05% Tween-20. For colour development Fast-Red substrate (Sigma) was used according to instructions.

RESULTS AND DISCUSSION Cloning of Arabidopsis thaliana GlcNAcT-I cDNA An Arabidopsis thaliana BAC end sequence of 636 bases present in the EMBL/GenBank database showed sequence similarity to mammalian GlcNAcT-I (15). To isolate a cDNA clone corresponding to this genomic sequence, an A. thaliana cDNA library in the mammalian expression vector pABE was screened by PCR using oligonucleotide primers based on this sequence. Twenty-four pools of the A. thaliana cDNA library each

830

Vol. 261, No. 3, 1999

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 2. Alignment of amino acid sequences of Arabidopsis thaliana and human GlcNAcT-I. Dotted lines indicate putative transmembrane regions. Amino acids marked with * are identical in the human and plant sequence while bold and underlined amino acids are conserved between all GlcNAcT-I sequences known from mammals, C. elegans (17) and A. thaliana.

containing about 5,000 different clones were screened with the two primers that amplified a fragment of 209 base pairs. In two of the pools an amplified fragment of the right size was detected. By performing a PCR reaction using the two genomic based primers in combination with primers in the vector arms, it was shown that the cDNA in one pool was inserted in the right orientation behind the promoter. The combined size of the amplified fragments also suggested that the cDNA insert contained the whole coding region. By the method described in the material and methods section, a single clone (pABE-AtGN1) was isolated containing the cDNA sequence corresponding to the A. thaliana genomic sequence. Complementation of the Lec1 Mutant To assay the functionality of the selected cDNA, we took advantage of the fact that the A. thaliana cDNA library was constructed in a mammalian expression vector. Clone pABE-AtGN1 was transiently expressed in a CHO cell line deficient in GlcNAcT-I (Lec1-15B) (10) to determine if the A. thaliana cDNA was able to complement the deficiency. CHO cells were stained with PHA, a lectin that binds complex N-linked glycans (15). The surface of glutaraldehyde fixed wild type CHO cells is clearly recognized by PHA. However, because of the absence of complex glycans, no binding of this lectin is detectable in Lec1 cells. After transfection of Lec1 cells with clone pABE-AtGN1, part of the cells (not all cells become transfected) show a PHA staining comparable to wild type cells (Fig. 1). This demonstrates that the cells transfected with the cDNA now

express complex glycans and that the GlcNAcT-I deficiency has been complemented. Thus, the enzyme encoded by the cloned A. thaliana cDNA has the same activity as animal GlcNAcT-I. It also shows that functional complementation can work both ways: human GlcNAcT-I complements a plant deficiency (13) and our plant glycosyltransferase expressed in mammalian cells complement their GlcNAcT-I deficiency. Sequence of Arabidopsis thaliana GlcNAcT-I The isolated clone of 1801 nucleotides contains an open reading frame encoding a polypeptide of 445 amino acids that shows 34% identity to human GlcNAcT-I (15) and a similar percentage to other mammalian GlcNAcT-I sequences. The percentage identity to the three cloned C. elegans GlcNAcT-I homologues (17) was only slightly lower. The A. thaliana GlcNAcT-I sequence forms the outermost branch in a phylogenetic tree comprised of the mammalian and C. elegans sequences (not shown). The alignment clearly shows conserved boxes between animal and plant sequences, especially in the middle of the catalytic domain (Fig. 2) (20). Notably, the only two cysteine residues that are conserved between mammals and C. elegans (17) are absent in A. thaliana. Although there is hardly any sequence conservation in the putative cytoplasmic, transmembrane and stem region, the presence of a transmembrane helix is predicted to be in the same position in both the A. thaliana and human sequences (Fig. 2) (http://www.enzim.hu/hmmtop) (21).

831

Vol. 261, No. 3, 1999

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Recently, the complete A. thaliana genomic sequence corresponding to the GlcNAcT-I cDNA has become available (BAC clone F20D10, GenBank accession: AL035538, predicted gene: F20D10.360). Alignment with our cDNA sequence shows that the A. thaliana GlcNAcT-I gene consists of 18 exons. Six of the introns are located at identical positions to one or more of the three C. elegans GlcNAcT-I genes (17) that have between 12 to 14 exons. One of the predicted introns does not have the normal consensus splice site GT-AG but is an AT-AC intron (nucleotide 121110-120996 in clone F20D10), a rare type of intron that is, however, found in both plants and animals (22). The cDNA sequence is identical to the corresponding genomic sequences. As the cloned cDNA encodes an active GlcNAcT-I, the A. thaliana GlcNAcT-I mutant cgl-1 is expected to have a deficiency in this gene or in its regulation. However, this has still to be shown. In summary we can conclude that expression of an A. thaliana cDNA in mammalian cells is an efficient way to determine its function. By making a plant cDNA library in vector pABE the clones can be directly expressed in mammalian cells. If no prior sequence information of a target gene is available, the library could also be used to screen for A. thaliana cDNA clones that induce a detectable phenotypic change in CHO or other mammalian cells. ACKNOWLEDGMENTS The authors thank Dr. Jaap Keijer and Patrick Koks for hospitality in their cell culture facility and Dr. Rita Gerardy-Schahn for the gift of vector pABE. This work was supported by a concern-SEO grant from DLO.

REFERENCES 1. Schachter, H. (1991) Glycobiology 1, 453– 461. 2. Kubelka, V., Altmann, F., and Ma¨rz, L. (1995) Glycoconj. J. 12, 77– 83.

3. Van Kuik, J. A., Sijbesma, R. P., Kamerling, J. P., Vliegenthart, J. F., and Wood, E. J. (1987) Eur. J. Biochem. 169, 399 – 411. 4. Altmann, F., Kornfeld, G., Dalik, T., Staudacher, E., and Glossl, J. (1993) Glycobiology 3, 619 – 625. 5. Schachter, H., Hull, E., Sarkar, M., Simpson, R. J., Moritz, R. L., Ho¨ppener, J. W., and Dunn, R. (1991) Biochem. Soc. Trans. 19, 645– 648. 6. von Schaewen, A., Sturm, A., O’Neill, J., and Chrispeels, M. J. (1993) Plant Physiol. 102, 1109 –1118. 7. Lerouge, P., Cabanes Macheteau, M., Rayon, C., FischetteLaine´, A. C., Gomord, V., and Faye, L. (1998) Plant Mol. Biol. 38, 31– 48. 8. Dean, N. (1999) Biochim. Biophys. Acta 1426, 309 –322. 9. Stanley, P., Narasimhan, S., Siminovitch, L., and Schachter, H. (1975) Proc. Natl. Acad. Sci. USA 72, 3323–3327. 10. Gottlieb, C., Baenziger, J., and Kornfeld, S. (1975) J. Biol. Chem. 250, 3303–3309. 11. Opat, A. S., Puthalakath, H., Burke, J., and Gleeson, P. A. (1998) Biochem. J. 336, 593–598. 12. Ioffe, E., and Stanley, P. (1994) Proc. Natl. Acad. Sci. USA 91, 728 –732. 13. Gomez, L., and Chrispeels, M. J. (1994) Proc. Natl. Acad. Sci. USA 91, 1829 –1833. 14. Sarkar, M., Hull, E., Nishikawa, Y., Simpson, R. J., Moritz, R. L., Dunn, R., and Schachter, H. (1991) Proc. Natl. Acad. Sci. USA 88, 234 –238. 15. Kumar, R., Yang, J., Larsen, R. D., and Stanley, P. (1990) Proc. Natl. Acad. Sci. USA 87, 9948 –9952. 16. Kumar, R., Yang, J., Eddy, R. L., Byers, M. G., Shows, T. B., and Stanley, P. (1992) Glycobiology 2, 383–393. 17. Chen, S., Zhou, S., Sarkar, M., Spence, A. M., and Schachter, H. (1999) J. Biol. Chem. 274, 288 –297. 18. Brusslan, J. A., and Tobin, E. M. (1992) Proc. Natl. Acad. Sci. USA 89, 7791–7795. 19. Eckhardt, M., Mu¨hlenhoff, M., Bethe, A., and Gerardy Schahn, R. (1996) Proc. Natl. Acad. Sci. USA 93, 7572–7576. 20. Sarkar, M., Pagny, S., Unligil, U., Joziasse, D., Mucha, J., Glossl, J., and Schachter, H. (1998) Glycoconj. J. 15, 193–197. 21. Tusnady, G. E., and Simon, I. (1998) J. Mol. Biol. 283, 489 – 506. 22. Tarn, W. Y., and Steitz, J. A. (1997) Trends Biochem. Sci. 22, 132–137.

832