Organization, structure and alternate splicing of the murine RFC-1 gene encoding a folate transporter

Gene 189 (1997) 1–7

Organization, structure and alternate splicing of the murine RFC-1 gene encoding a folate transporter Berend Tolner a, Krishnendu Roy a, F.M. Sirotnak a,b,* a Laboratory for Molecular Therapeutics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA b Graduate School of Medical Sciences, Cornell University, New York, NY 10021, USA Received 13 June 1996; received in revised form 16 August 1996; accepted 18 August 1996; Received by C.M. Kane

Abstract The structural organization of the murine RFC-1 gene encoding a folate transporter has been determined. The entire nucleotide sequence of the L1210 cell RFC-1 cDNA, the 3∞- and 5∞-untranslated regions and the coding sequence were found to be distributed in eight exons, including six primary exons and alternates to exon 1 and exon 5, spanning 10.4 kb. Splice variants were identified in an L1210 cell cDNA library. The most common incorporates exons 1 through 6, encoding a 58-kDa polypeptide. The two least common incorporate exons 1 and 2, a truncated version of exon 3 and exons 4 through 6; or exons 1 through 4, an alternate to exon 5, and exon 6, encoding polypeptides of 53.6 and 43.4 kDa, respectively. A fourth variant reported earlier (GenBank/EMBL accession No. L36539) by others incorporates what we have found to be an alternate of exon 1 and exons 2 through 6. A relatively GC-rich region of the genome just 5∞ of exon 1 as well as exon 1a appears to be distinctly promoter-like and encodes a number of putative cis-acting elements. The findings pertaining to alternates of exon 1 suggest that the transcription of RFC-1 variants results from two different promoters. © 1997 Elsevier Science B.V. All rights reserved. Keywords: Murine RFC-1 gene; Folate transport

1. Introduction Mammalian cells have an absolute requirement for exogenous folates (Blakley, 1969; Kisliuk, 1984) for growth and macromolecular biosynthesis. Transport of folate compounds into mammalian cells can occur via carrier-mediated (Sirotnak, 1985; Goldman and Matherly, 1986; Ratnam and Freisheim, 1990) as well as receptor-mediated ( Kamen and Capdevila, 1986; Ratnam and Freisheim, 1990; Anderson et al., 1992) mechanisms depending upon the relative level at which each is operable in the plasma membrane. In tumor cells, the same carrier-mediated mechanism is involved (Sirotnak, 1985; Goldman and Matherly, 1986; Wester* Corresponding author at the Laboratory for Molecular Therapeutics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA. Fax: +1 212 7944342; e-mail: [email protected] Abbreviations: IFC, intestinal folate carrier; kb, kilobase(s); kDa, kilodalton; PCR, polymerase chain reaction; RFC, reduced folate carrier; UTR, untranslated region(s); IFCm, mouse intestinal folate carrier; RFCmh, RFCm, RFCr, hamster, mouse and rat reduced folate carrier, respectively. 0378-1119/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S 03 7 8 -1 1 1 9 ( 9 6 ) 0 0 6 76 - 2

hof et al., 1995) in the internalization of both folates and cytotoxic folate analogues, thus, the level of expression of this system has potential pharmacological relevance. Also, in many tumor cells a decrease in mediated entry of folate analogues by this system has been associated (Sirotnak et al., 1981; Sirotnak, 1987; Schuetz et al., 1989) with acquired resistance to these agents. Recently, a murine cDNA clone (RFC-1) was isolated (Dixon et al., 1994) from an L1210 cell cDNA library by expression cloning. This cDNA appears to code for a transporter with many of the properties of the onecarbon, reduced folate transporter. About the same time, a genomic fragment from a hamster homologue of the RFC-1 gene was also cloned ( Williams et al., 1994) from a cosmid library and used to isolate a hamster cDNA homologue of RFC-1. Also, a murine cDNA clone that extends the 5∞- and 3∞-UTR of the original RFC-1 clone was reported (Brigle et al., 1995). Subsequently, a human cDNA homologue of RFC-1 was isolated in four laboratories (Moscow et al., 1995; Prasad et al., 1995; Williams and Flintoff, 1995; Wong et al., 1995) using one of a variety of cloning approaches.

2

B. Tolner et al. / Gene 189 (1997) 1–7

Although it appears from the available information that these murine, hamster and human cDNA clones encode the one-carbon, reduced folate transporter, there remain unresolved questions, particularly, in regard to the molecular mass of the transporter encoded by RFC-1. The predicted mass of the transporter encoded (Dixon et al., 1994) by the murine RFC-1 gene is 58 kDa, while the size of the transporter determined (Price and Freisheim, 1987; Schuetz et al., 1988; Yang et al., 1988; Chiao et al., 1995; Wong et al., 1995) by a variety of methodologies is 43–46 kDa. Moreover, similar methodologies have shown (Matherly et al., 1992; Wong et al., 1995) that the molecular mass of the glycosylated human transporter is in the range of 80–90 kDa, while that detected (Moscow et al., 1995) with anti-human RFC-1 peptide antibodies is in the range of 50–60 kDa, a value somewhat lower, as well, than the molecular mass of 65 kDa predicted (Moscow et al., 1995; Prasad et al., 1995; Williams and Flintoff, 1995; Wong et al., 1995) by the human cDNA sequence. While there are possible explanations for these discrepancies that are consistent with the notion that RFC-1 encodes the one-carbon, reduced folate transporter, these still remain to be validated by the appropriate experimental data. In order to ultimately shed further light on this question it will be necessary to provide information at the level of the gene. Consequently, we now report data on the complete organization and structure of the murine RFC-1 gene using restriction mapping and nucleotide sequence analysis. Prior to this time, only a very limited amount of information has been available (Moscow et al., 1995) on the organization of the RFC-1 gene, in this case, from studies of the human genome.

selected for further study on the basis of hybridization with the region-specific probes. The fragments of interest after extraction were subcloned into Bluescript SK+ (Stratagene, Menaska, WI ) for sequencing. 2.2. Screening for murine RFC-1 splice-variants Murine RFC-1 cDNA clones were obtained by radioactive hybridization screening ( Hendricks et al., 1984; Roy et al., 1995) of an L1210/R83 custom-made cDNA library in pTZ18R-B (Invitrogen, San Diego, CA) using a 295-bp 5∞-PCR-fragment of the RFC-1 cDNA (Dixon et al., 1994; Brigle et al., 1995) as a probe (primers: TGAGACCTGGGCAACATG and AGACTGGCTTGTATCGCA). Polymerase chain reactions were carried out with Vent DNA polymerase (New England Biolabs) using the recommended buffer, 20 ng of primers, 50 ng vector DNA and 400 mM dNTPs in a total volume of 100 ml. After an initial denaturation step of 5 min at 94°C, 35 cycles of 1 min at 94°C, 1 min at 50°C and 1 min at 72°C, respectively, were carried out. Finally, the reactions were extended for 10 min at 72°C. The hybridization screening of the cDNA library was carried out under conditions described earlier (Roy et al., 1995). The radioactive 5∞-PCR-fragment was prepared by random priming (Boehringer Mannheim) with [a-32P]dCTP. The DNA insert of 34 positive clones was purified and characterized by restriction map comparison relative to the full-length IFC-1 and RFC-1 cDNAs. Subsequently, seven variant cDNA clones (designated RFC-Variant 1–7, respectively) were subjected to sequence analysis. 2.3. Sequencing of DNA fragments, intron size determination and primer extension

2. Materials and methods 2.1. Isolation of genomic clones by hybridization screening A mouse liver genomic library in vector Lambda FIX II (Stratagene, Menaska, WI ) was screened with radioactive ([a-32P]dCTP) murine RFC-1 cDNA (which was a generous gift of Dr. Cowan) that was prepared by random priming (Boehringer Mannheim). After screening of 106 plaques and purification, six positive clones were obtained. The DNA insert from these clones was purified and characterized by restriction mapping and Southern hybridization using fragments encompassing sequences from the most 5∞- and 3∞-ends of the RFC-1 cDNA (Dixon et al., 1994; Brigle et al., 1995) and a variant (IFC-1) derived from murine intestinal epithelium (Said et al., 1996). Two nonidentical clones designated l mRFC-2 and -6, were selected by the above procedure for further analysis. Restriction fragments of these clones generated with NotI and/or EcoRI were

Double-stranded DNA was sequenced in both directions according to the dideoxy method of Sanger et al. (1977) using Sequenase version 2.0 ( US Biosciences, Cleveland, OH ). Oligonucleotide primers based on the mouse RFC-1 cDNA (see above and Dixon et al., 1994; Brigle et al., 1995) sequence and that of the published IFC-1 (Said et al., 1996) were used initially. Additional oligonucleotide primers were prepared on the basis of the sequence data generated when necessary for extending the sequencing. Exon/intron junctions were determined by direct sequencing across these junctions using primers based upon the mouse RFC-1/IFC-1 cDNA sequences. Intron size was determined by sequencing through the region in question or by restriction mapping. The nucleotide sequence of the cloned genomic IFC-1 and RFC-1 intron/exon fragments and the RFC-1cDNA ( Variant I, see below) were submitted to GenBank (GenBank/EMBL accession Nos. U57780-86 and U66103, respectively). Primer extension was essentially performed as described previously (Sambrook et al.,

B. Tolner et al. / Gene 189 (1997) 1–7

1989). Briefly, an oligonucleotide primer complementary to the 5∞-end of the RFC/IFC-1 mRNA (CACCATGTTGCCCAGGTCCT; positions −15 to +6 relative to the translational start-site) was end-labeled with [c-32P]ATP. The labeled product was purified by ammonium acetate/ethanol precipitation. The pellet was washed 2 times with ethanol, dried and dissolved in SuperScript II reverse transcriptase buffer and the reverse transcription reaction was performed according to the manufacturer (Gibco-BRL). Extension products were purified by Sephadex G25 column chromatography, separated by electrophoresis in a 6% PAA, 8 M urea gel and visualized by autoradiography. 2.4. Materials All radioactive isotopes used for the above studies were obtained from NEN ( Waltham, MA). DNA restriction enzymes were purchased from Boehringer Mannheim, Indianapolis, IN. Nitrocellulose was purchased from Schleicher and Schuell and oligonucleotide primers were synthesized by Gene Link, Inc. ( Thornwood, NY ). All other materials were reagent grade.

3. Results and discussion 3.1. Organization and structure of the murine RFC-1 gene

3

enzyme mapping (Fig. 1). Clone l mRFC-6 which was 15 kb in length contained the region corresponding to the 5∞-end of the murine RFC-1 cDNA and approximately 5 kb of a sequence beyond, in addition to 10 kb of a sequence homologous to the more 5∞-end of the cDNA. Clone l mRFC-2 which was 14 kb in length, lacked the region containing the most 5∞-sequence of the cDNA, overlapped with approximately 5 kb of l mRFC-6 and contained 9 kb of additional sequence extending beyond the 3∞-UTR of the cDNA. A number of restriction fragments of l mRFC-2 and -6 were subcloned and sequenced in each direction, and the sequence of the gene including introns 1a and 5a was obtained. The lengths of the other introns were estimated by restriction mapping. Sequencing of the most 3∞-region of the RFC-1 gene extended the nucleotide sequence to an adjacent collagen gene (Rehn et al., 1996). The organization of the murine RFC-1 gene is shown in Fig. 1. The gene consists of eight exons including six primary exons and alternates of exon 1 (1a) and 5 (5a), with an overall length of 10.4 kb. Most intron/exon splice junctions shown in Table 1 conform to the GT-AG rule. The 5∞-UTR and one of two apparent leader peptide sequences is incorporated in either exon 1 or exon 1a. However, both apparent leader sequences would actually remain untranslated since each contains a stop codon at position −3646 (exon 1a) and −3112 (exon 1). Exons 2–5 incorporate only coding regions while exon 6 contains coding regions and the 3∞-UTR. 3.2. Differential splicing and alternate exon usage

A total of six genomic clones with relevant nucleotide sequence were identified by screening with an L1210 cell RFC-1 cDNA. Two of these clones designated l mRFC-2 and l mRFC-6 were overlapping and found to be nonidentical after characterization by restriction

Hybridization screening revealed three different splice variants in the L1210 cDNA library ( Fig. 2). Initially the most abundant (#95%) species ( Variant I ) was sequenced, which revealed a cDNA composed of

Fig. 1. Organization of the murine RFC-1 gene showing a restriction map and exon/intron structure. The two clones shown had inserts of 14 (l mRFC-2) and 15 kb (l mRFC-6), respectively. The lengths of the exons and introns identified are shown to scale.

4

B. Tolner et al. / Gene 189 (1997) 1–7

Table 1 Intron/exon splice junctions of the murine IFC1/RFC1 gene Exon No.

1a 1 2 3 3a 4 5a 5 6

Exon length (nt)a

40 89 230 755 635 192 79 142 1137

5∞-Junction

3∞-Junction

Intron sequence

Exon sequence

cagatcaag gtagtctgc tccttccag cgtccacag cgtccacag cctcccagg actatatag atctcccag ttctttcag

CAATCT GTGTGC GTGGAG GTGACT GTGACT CGCCAT ACCAAG TTTTCA TTCCGT

... ... ... ... ... ... ... ... ...

Intron length (nt)b

Exon sequence

Intron sequence

GGACGG TGCAAG GAGCAG GCTGAG CCTCTG TGCCAC TACACG GATCAG AACTAC

gtagggggc gtacggttc gcatgttgc tatgcaccg gtgggtctt gtgagtgaa gtgagtttg gtaagcacc cttgggctt

481 3500 2600 450 570 900 40 2600

aExon length determined by sequencing, except for exons 1a, 1 and 6, whose lengths were based on the longest cDNA clones available. The 3∞-ends of exon 6 of IFC-1 and RFC-1 are based on Said et al. (1996) and EMBL/GenBank entry U66103, respectively. bIntron length determined by restriction mapping of relevant subclones, except for introns 1a and 5a, which were sequenced.

Fig. 2. Exon composition of the splice-variants of the murine RFC-1 gene. Exon 1 through 6 are primary exons, whereas 1a and 5a are alternates of exon 1 and 5, respectively. Exon 3a is truncated form of exon 3; 120 bp appear to be deleted relative to exon 3.

2320 bp. The transcript includes a poly(A) tail of 25 adenylate residues which is indicative of the 3∞-UTR end. This cDNA extends the 5∞- and 3∞-UTRs of previously published sequences by Dixon et al. (1994) and Brigle et al. (1995). Except for an insertion of a G residue between residues 29 and 30 relative to the former and a CG to GC substitution at positions 12–13 relative to the latter sequence, our sequence was otherwise identical. Our data were confirmed by sequencing of genomic RFC-1 DNA. Primer extention analysis (data not shown) identified multiple transcription start sites just 5∞ of exon 1 which are consistent with different lengths of the alternate 5∞-ends identified in Fig. 2. Furthermore, the length of the cloned transcripts coincides with the size obtained by Northern blotting as was shown in several publications (Brigle et al., 1995; Said

Fig. 3. Alignment of the 5∞-region of rodent RFC-1 cDNAs. The 5∞-regions of mouse IFC-1 (IFCm), rat RFC-1 (RFCr) and hamster RFC-1 (RFCh) cDNA are aligned relative to the mouse RFC-1 (RFCm). Stop codons are underlined and in boldface. Symbols: *, identical residues; ,, splice-site between exons 1(a) and 2. Shaded box, start codon of the respective RFC cDNA species.

B. Tolner et al. / Gene 189 (1997) 1–7

5

Fig. 4. Nucleotide sequence of the 5∞-region comprising the putative promoter of the murine RFC-1 gene. The DNA sequence is numbered from the ATG start codon as found in exon 2. Since the intervening 3-kb EcoRI fragment was not sequenced, the numbering of the promoter-like region was reset arbitrarily at −3000 as indicated in the figure. Potential regulatory elements, as indentified according to Boulikas (1994) and a database search ( Transciption Factor Binding Sites Database, version 3, The Japanese Genome Center), are in bold face and identified by the appropriate symbols.

et al., 1996). Thus the RFC-1 splice variants discussed here are full-length transcripts, including 5∞- and 3∞-ends. Variant I is the most common and consists of exon 1 plus exons 2–6 encoding a polypeptide of 58 kDa. A very rare variant ( Variant II ) also consists of exon 1 plus exons 2–6. However, in this variant, 120 bp of cDNA was deleted at the most 3∞-end of exon 3. Another rare variant ( Variant III ) comprises exon 1 plus exons 2, 4, and an apparent alternative to exon 5 (exon 5a) in addition to exon 6. Variants II and III encode truncated

polypeptides of 53.6 and 43.4 kb, respectively. As a consequence, Variants II and III encode only 11 and 10 membrane spanning regions, respectively, compared to 12 encoded by Variants I and IV. The significance of variants II and III is unknown at this time, particularly since they are represented at extremely low frequencies (#5%). A fourth variant ( Variant IV ), reported recently by Said et al. (1996), was derived from a cDNA library prepared from small intestine of the mouse. This variant, which may be the predominant variant in this mouse

6

B. Tolner et al. / Gene 189 (1997) 1–7

tissue, incorporates an alternate to exon 1 (exon 1a) but is otherwise similar in composition to Variant I. Comparison of the above data with the hamster ( Williams et al., 1994) and rat (GenBank/EMBL accession No. U38180) cDNA sequences reveals (Fig. 3) considerable diversity in the region 5∞ of exon 2. In contrast, the nucleotide sequence of exon 2 in this gene is very similar among these species. 3.3. Sequence analysis of the 5∞-region The region of the genomic DNA immediately 5∞ of exon 1a contains (Fig. 4) a nucleotide sequence which is distinctly promoter like. This region is relatively GC rich and contains many putative Ets-1 and SP-1 binding sites, three GC boxes, two inverse CCAAT boxes and a di-nucleotide repetitive element which could act as a putative enhancer element (Hamada et al., 1984), but lacks TATA-like sequence motifs. Interestingly, a number of these same potential binding sites, and an Ap2 and a PEA3 binding site, were also found throughout the region spanned by exon 1a and intron 1a. Since all splice variants (I–III ) found in L1210 cells appear to use exon 1 and the major transcript in murine intestine (variant IV, Said et al., 1996) uses an alternative to exon 1, these finding suggests that the transcription of RFC-1 variants results from two distinct promoters. Tissue and/or developmental stage-specific gene expression via transcripts which originate from the same gene but incorporate an alternative 5∞-exon have previously been shown to be under the control of multiple promoters (Schibler and Sierra, 1987). We also noted from the sequence data that sequences homologous to a transposon-like element (GenBank/EMBL accession No. U17087-95) and recombination hot spots (GenBank/ EMBL accession Nos. D38457 and D38446) exist approximately 3.5 and 3.0 kb, respectively, upstream of exon 1a. As this is the first detailed information reported in the literature concerning the organization and structure of this gene in mammalian cells, it will be of interest to eventually compare its structure between this and other mammalian species. The TATA-less nature of this region of the gene that was identified as a possible promoter and the multiplicity of potential Sp-1 sites suggests that the RFC-1 gene is of the ‘housekeeping’ variety which includes genes that are differentially expressed in various tissues and during embryonic development. It was of interest as well to identify a putative transposon-like element and recombination hot spots approximately 3.5 and 3.0 kb, respectively, upstream of exon 1a. These elements are associated with gene amplification and genomic instability. The finding of stop codons within the exons encoding putative leader sequences ( Fig. 3) of the mouse indicate that these leaders are not translated as previously suggested (Brigle et al., 1995)

and that this region of the gene has a regulatory function. It is of interest to note that these stop codons also were identified in the same region (Fig. 3) of the rat homologue of RFC-1. The homologue of mouse RFC-1 in hamster lacks this stop codon; however, the 5∞-sequence data of this cDNA that is available ( Williams et al., 1994) is limited. The most common RFC-1 cDNA species isolated from L1210 cells ( Variant I ) encodes a polypeptide of 58 kDa. This size does not correlate with the size of the transporter determined (Price and Freisheim, 1987; Schuetz et al., 1988; Yang et al., 1988; Chiao et al., 1995; Wong et al., 1995) by a variety of methodologies to be 43–46 kDa. However, in all of the experimental procedures involved (except Yang et al., 1988) the membrane proteins were sized by SDSPAGE. It has to be noted that aberrant electrophoretic behavior is often observed for integral membrane proteins (e.g., Poolman et al., 1989; VanderRest et al., 1990; Wallace et al., 1990; Tolner et al., 1992; Pourcher et al., 1995; Gaillard et al., 1996; Knol et al., 1996) and is most likely explained by increased binding of sodium dodecyl sulfate due to the hydrophobic nature of these proteins. Nevertheless, in a recent preliminary report (Chiao et al., 1997) the l RFC-1 transcript from mouse small intestine was unequivocally correlated with a 58-kDa protein.

Acknowledgement Supported in part by grants CA08748 and CA56517 from the National Cancer Institute and the Elsa U. Pardee Foundation.

References Anderson, R.G.W., Kamen, B.A., Rothberg, K.G. and Lacey, S.W. (1992) Protocytosis: sequestration and transport of small molecules by caveolae. Science 255, 410–411. Blakley, R.L. (1969) In: The Biochemistry of Folic Acid and Related Ptridines. North-Holland, Amsterdam, pp. 3–87. Boulikas, T. (1994) A compilation and classification of DNA binding sites for protein transcription factors from vertebrates. Crit. Rev. Eukar. Gene Express. 4, 117– 321. Brigle, K.E., Spinella, M.J., Sierra, E.E. and Goldman, I.D. (1995) Characterization of a mutation in the reduced folate carrier in a transport defective L1210 murine leukemia cell line. J. Biol. Chem. 270, 22974–22979. Chiao, J.D., Yang, C.-H., Roy, K., Pain, J. and Sirotnak, F.M. (1995) Ligand-directed immunoaffinity purification and properties of the one-carbon, reduced folate transporter. Interspecies immuno-crossreactivity and expression of the native transporter in murine and human tumor cells and their transport-altered variants. J. Biol. Chem. 270, 29698–29704. Chiao, J.D., Roy, K., Yang, C.H., Tolner, B. and Sirotnak, F.M. (1997) RFC-1 gene expression regulates folate absorption in mouse small intestine. Submitted. Dixon, K.H., Lanpher, B.C., Chiu, J., Kelly, K. and Cowan, K.H.

B. Tolner et al. / Gene 189 (1997) 1–7 (1994) A novel cDNA restores reduced folate carrier activity and methotrexate sensitivity to transport deficient cells. J. Biol. Chem. 269, 17–20. Gaillard, I., Slotboom, D.J., Knol, J., Lolkema, J.S. and Konings, W.N. (1996) Purification and reconstitution of the glutamate carier GltT of the thermophilic bacterium Bacillus stearotherophilus. Biochemistry 35, 6150–6156. Goldman, I.D. and Matherly, L.H. (1986) The cellular pharmacology of methotrexate. Int. Encyl. Pharmacol. Ther. 118, 293–302. Hamada, H., Seidman, M., Howard, B.H. and Gorman, C.M. (1984) Enhanced gene expression by the poly(dT-dG)-poly(dC-dA) sequence. Mol. Cell. Biol. 4, 2622–2630. Hendricks, S.M., Cohen, D.I., Nielsen, E.A. and Davis, M.M. (1984) Isolation of cDNA clones encoding T-cell-specific membrane-associated proteins. Nature 308, 149–153. Kamen, B.A. and Capdevila, A. (1986) Receptor-mediated folate accumulation is regulated by the cellular folate content. Proc. Natl. Acad. Sci. USA 83, 5983–5987. Kisliuk, R.L. (1984) The Biochemistry of Folic Acid. In: Sirotnak, F.M., Burchall, J.J., Ensminger, W.D. and Montogmery, J.A. ( Eds.), Folate Antagonists as Therapeutic Agents, Vol. 1. Academic Press, New York, pp. 2–55. Knol, J., Veenhoff, L., Liang, W.J., Henderson, P.J.F., Leblanc, G. and Poolman, B. (1996) Unidirectional reconstitution into detergent-destabilized liposomes of the purified lactose transport system of Streptococcus thermophilus. J. Biol. Chem. 271, 15358–15366. Matherly, L.H., Angeles, S.M. and Czajkowsk, C.A. (1992) Characterization of transport-mediated methotrexate resistance in human tumor cells with antibodies to the membrane carrier for methotrexate and tetrahydrofolate cofactors. J. Biol. Chem. 267, 23253–2360. Moscow, J.A., Gong, M., He, R., Sgagias, M.K., Dixon, K.H., Anzick, S.L., Meltzer, P.S. and Cowan, K.H. (1995) Isolation of a gene encoding for a human reduced folate carrier (RFC1) and analysis of its expression in transport-deficient, methotrexate-resistant human breast cancer cells. Cancer Res. 55, 3790–3794. Poolman, B., Royer, T.J., Mainzer, S.E. and Schmidt, B.F. (1989) Lactose transport system of Streptococcus thermophilus: a hybrid protein with homology to the melibiose carier and enzyme III of phosphoenolpyruvate-dependent phosphotransferase systems. J. Bacteriol. 171, 244–253. Pourcher, T., Leclercq, S., Brandolin, G. and Leblanc, G. (1995) Melibiose carrier of Escherichia coli: large scale purification and evidence for H+, Na+, and Li+ sugar symport is catalyzed by a single polypeptide. Biochemistry 34, 4412–4420. Prasad, P.D., Ramamoorthy, S., Leibach, F.H. and Ganapathy, V. (1995) Molecular cloning of the human placental folate transporter. Biochem. Biophys. Res. Commun. 206, 681–687. Price, E.M. and Freisheim, J.H. (1987) Photoaffinity analogues of methotrexate as folate antagonist binding probes. 2. Transport studies, photoaffinity labeling and identification of the membrane carrier protein for methotrexate from murine L1210 cells. Biochemistry 26, 4757–4763. Ratnam, M. and Freisheim, J.H. (1990) In: Picriano, M.F. ( Ed.), Contemporary Issues in Clinical Nutrition. Wiley-Liss, New York, pp. 91–120. Rehn, M., Hintikka, E. and Pihlajaniemi, T. (1996) Characterization of the mouse gene for the alpha-1 chain of type XVIII collagen (COL18A1) reveals that the three variant N-terminal polypeptide forms are transcribed from two widely separated promoters. Genomics 32, 436–446. Roy, K., Mitsugi, K., Sirlin, S., Shane, B., and Sirotnak, F.M. (1995) Different antifolate-resistant L1210 cell variants with either increased or decreased folylpolyglutamate synthetase gene expression at then level of mRNA transcription. J. Biol. Chem. 270, 26918–26922.

7

Said, H.M., Nguyen, T.T., Dyer, D.L., Cowan, K.H. and Rubin, S.A. (1996) Intestinal folate transport: identification of a cDNA involved in folate transport and the functional expression and distribution of its mRNA. Biochim. Biophys. Acta 1281, 164–172. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd Edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Sanger, F., Nicklen, S. and Coulson, A.R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463–5467. Schuetz, J.D., Matherly, L.H., Westin, E.H. and Goldman, I.D. (1988) Evidence for a functional defect in the translocationof the methotrexate transport carrier in a methotrexate-resistant murine L1210 leukemia cell line. J. Biol. Chem. 263, 9840–9847. Schuetz, J.D., Westin, E.H., Matherly, L.H., Pincus, R., Swerdlow, P.S. and Goldman, I.D. (1989) Membrane protein changes in an L1210 leukemia cell line with a translocation defect in the methotrexate-tetrahydrofolate cofactor transport carrier. J. Biol. Chem. 264, 16261–16267. Schibler, U. and Sierra, F. (1987) Alternative promoters in developmental gene expression. Annu. Rev. Genet. 21, 237–257. Sirotnak, F.M. (1985) Obligate genetic expression in tumor cells of a fetal membrane property mediating ‘folate’ transport: biological significance and implications for improved therapy of human cancer. Cancer Res. 45, 3992–4000. Sirotnak, F.M. (1987) Determinants of resistance to antifolates: biochemical phenotypes, their frequency of occurrence and circumvention. NCI Monographs 5, 27–35. Sirotnak, F.M., Moccio, D.M., Kelleher, L.E. and Goutas, L.J. (1981) Relative frequency and properties of transport-defective phenotypes among methotrexate-resistant L1210 clonal cell lines derived in vivo. Cancer Res. 41, 4447–4452. Tolner, B., Poolman, B. and Konings, W.N. (1992) Characterization and functional expression in Escherichia coli of the sodium/proton/ glutamate symport proteins of Bacillus stearothermophilus and Bacillus caldotenax. Mol. Microbiol. 6, 2845–2856. VanderRest, M.E., Schwartz, E., Oesterhelt, D. and Konings, W.N. (1990) DNA sequence of a citrate carrier of Klebsiella pneumoniae. Eur. J. Biochem. 189, 401–407. Wallace, B., Yang, Y., Hong, J. and Lum, D. (1990) Cloning and sequencing of a gene encoding a glutamate and aspartate carrier of Escherichia coli K12. J. Bacteriol. 172, 3214–3220. Westerhof, G.R., Schornagel, J.H., Kathman, A.L., Rosowsky, A., Forsch, R.A., Hynes, J.B., Boyle, F.T., Peters, G.J., Pinedo, H.M. and Jansen, G. (1995) Carrier- and receptor-mediated transport of folate antagonists targeting folate-dependent enzymes: correlates of molecular structure and biological activity. Mol. Pharm. 48, 459–471. Williams, F.M.R. and Flintoff, W.F. (1995) Isolation of a human cDNA that complements a mutant hamster cell defective in methotrexate uptake. J. Biol. Chem. 270, 2987–2992. Williams, F.M.R., Murray, R.C., Underhill, T.M. and Flintoff, W.F. (1994) Isolation of a hamster cDNA clone coding for a function involved in methotrexate uptake. J. Biol. Chem. 269, 5810–5816. Yang, C.-H., Sirotnak, F.M. and Mines, L.S. (1988) Further studies on a novel class of genetic variants of the L1210 cell with increased folate analogue transport inward. Transport properties of a new variant, evidence for increased levels of a specific transport protein, and its partial characterization following affinity labeling. J. Biol. Chem. 263, 9703–9209. Wong, S.C., Proefke, S.A., Bhushan, A. and Matherly, L.H. (1995) Isolation of human cDNAs that restore methotrexate sensitivity and reduced folate carrier activity in methotrexate transport-defective Chinese hamster ovary cells. J. Biol. Chem. 270, 17468–17475