Differential Expression of Transporters for Norepinephrine and ...

9 downloads 132 Views 3MB Size Report
Chemicals-L-Alanosine and N-(phosphonacety1)-L-aspartic acid. (PALA)' were gifts of the Developmental Therapeutics Program of the National Cancer Institute ...
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1993 by The American Society for Biochemistry end Molecular Biology, Inc

Vol. 268, No. 32, Issue of November 15, pp. 23891-23897,1993 Printed in U.S. A.

Differential Expressionof Transporters forNorepinephrine and Glutamate in Wild Type, Variant, andWNT1-expressing PC12Cells* (Received for publication, April 2, 1993)

Baskaran Ramachandran, Karl Houben, Yanina Y. Rozenberg, Julian R. Haigh, Arousiak Varpetian, and BruceD. Howard From the Department of Biological Chemistp, School of Medicine, University of California, Los Angeles, California 90024

Wild type PC12 pheochromocytoma cells express a Na+-dependent norepinephrine transporter that operates in the uptake of catecholamines, including dopamine. This transporter is not expressed in two spontaneously occurring flat cell variants of PC12 or in two other flat cell variants whose phenotype was induced by expression of the Wnt-1 oncogene. However, each of the flat cell variants, including those that express Wnt-1, exhibit a Na+-dependent, C1”independent glutamate/aspartate transporter activity that not is present in wild type PC12 cells. The flat cell variants took up glycine by a Na+-dependent process as well as did wild type cells. All of the flat cell variants have decreased levels of norepinephrine transporter mRNA but normal levels of glycine transporter mRNA. Glutamate/aspartate transportermRNA was detected only in thevariantsthat exhibitedglutamate/aspartate transporter activity, and thenucleotide sequence of a partial glutamate/aspartate transporter cDNA from these cells demonstrated that it was the glial form of the transporter that was expressed. These variants were more sensitive than waswild type PC12 to alanosine, a toxic aspartate analog that enters cells by a transporter-mediated system such as the glutamate/ aspartate transporter; however, these variants were as sensitive as wild type cells to anothertoxic aspartate analog, N-(phosphonacety1)-L-asparticacid, which is believed to enter cells by endocytosis. We suggest that the Wnt-1 gene product, or a homolog, may beinvolved in glial differentiation and that the mechanisms that alter the expression of the norepinephrine and glutamate/aspartate transporters in wild type and variant PC12 cells may also operate to regulate neurotransmitter transporterexpression in vivo.

PC12 is a clonal line of rat pheochromocytoma cells that are spherical endocrine cells under standard conditions of culture but can be induced to differentiate into neurons by nerve growth factor or certain oncogenes (Greene and Tischler, 1976; Noda et al., 1985; Guerrero et al., 1988; Sugimoto et al., 1988; Sassone-Corsi et al., 1989). Inthispaper we describe some PC12 variants whose transport properties suggest that PC12 can also be induced to differentiate into a glial-like phenotype. * This work wassupported by United States Public Health Services Grant MH38633. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) L19564 and L19565.

Cultures of PC12 cells commonly contain a small percentage of flat cell variants, which derive their namebecause they adhere very tightly to the Petri dish and thus flatten out (Bothwell et al., 1980; Andersen et al., 1990). The flat cell variants fail to exhibit manyof the neuron-like properties or gene expression of wild type PC12; for example, they do not differentiate in response to nerve growth factor, synthesize dopamine, or express the gene for tyrosine hydroxylase (Bothwell et al., 1980; Bitler et al., 1986; Andersen et al., 1990). However, karyotypeanalysishas shown that the flat cell variants are true derivatives of wild type PC12 and are not contaminants (Andersen et al., 1990). Recently, flat cell variants have beenfound to beproduced by introducing and constitutively expressing the Wnt-I oncogene in wild type PC12 cells (Shackleford et al., 1993). The Wnt-1 oncogene was discovered as a gene whose expression was frequently induced when murine mammary tumor virus caused mammarytumors(Nusse et al., 1984; Nusseand Varmus, 1992). The protein product of this gene is an autocrine and paracrine factor in that it is secreted by the cell and bindstothesame cell and to neighboring cells, where it induces a signaling pathway that leads to transformation of the cells (Jue et al., 1992). Wnt-I is not expressed in the adult organism outside of the testes, but isitexpressed in embryonic brain; recent knockout experiments have shown that Wnt-I expression is necessary fordevelopment of large parts of the brain (McMahon and Bradley, 1990; Thomas and Capecchi, 1990). This resultsuggested that expression of Wnt-I in PC12 would causeneuronaldifferentiation of PC12 just as the expression of certain other oncogenes is known to do (Noda et al., 1985). Surprisingly, Wnt-1expression makes PC12cells unable to differentiate in response nerve to growth factor and toacquireotherproperties of flat cell variants of PC12; expression of an inactive frame-shift mutant Wnt-I in PC12 did not cause the flat cell phenotype (Shacklefordet al., 1993). Bitler et al. (1986) found one PC12 flat cell variant, M P T l (formerly called MPTr),tobe markedlydeficientin the uptake of catecholamines. Wild type PC12cells have aplasma membrane catecholamine transporter that transports both norepinephrine and dopamine into the cells by a Na+-dependent process (Greene and Rein, 1977; Koide et al., 1986). The PC12 catecholamine transporter islike those that are present in many nerve terminals where they function to transport certain neurotransmitters back into thenerve terminals from which they had been released (Iversen, 1970). This transport reduces the concentration of the neurotransmitter at postsynaptic sites and thereby terminates neurotransmitter action. Although norepinephrine and dopamine are equally good substrates for the PC12 transporter, the relative sensitivity of the transporter to certain inhibitors demonstrates that it is more like a true neuronal norepinephrine transporter than

23891

23892

Transporter Expression

in Wild Type and Variant

a dopamine transporter (Koide et al., 1986).However, for various technical reasons dopamine is commonlyused as t h e substrate in PC12 transport studies. The cDNA for the human norepinephrine transporter has recently been cloned (Pacholcyzk et al., 1991). The transporter is a protein of 617 amino acid residues i n length, and it has 12 transmembrane regions and six extracellular loops, one of which is quite large. It exhibits significant homology with other Na+-dependent transporters of neurotransmitters, including the dopamine transporter (Giros et al., 1991; Kilty et al., 1991; Shimada et al., 1991), glycinetransporter (Guastella et al., 1992; Smith et al., 1992), and the y-aminobutyric acid transporter (Guastella et al., 1990). In this paper we reportthat four differentflat cell variants, including two that were induced by expression of the Wnt-1 oncogene, are deficient in norepinephrinetransport and have diminished levelsof norepinephrine transportermRNA. However, all of these flat cell variants express a Na+-dependent glutamate/aspartate transporter that is not expressed bywild type PC12 cells. The glutamate/aspartatetransporterexpressed by the flat cell variants belongs to a family of three recently cloned glutamate/aspartatetransporters (Kanai and Hediger, 1992; Pines et al., 1992; Storck et al., 1992). These glutamate/aspartate transporters have 6-10 transmembrane regions and are not homologous to members of the transporter family that includes the norepinephrine transporter and glycine transporter. Based on in situ hybridization studies, one of these glutamate/aspartate transporters has been localized to glia (Storcket al., 1992), and another appears tobe neuronal (Kanai and Hediger,1992).Sequenceanalysis of partial cDNAs obtained from the PC12 flat cells demonstrate it t o be a glial form of the glutamate/aspartate transporter. EXPERIMENTALPROCEDURES

Chemicals-L-Alanosine and N-(phosphonacety1)-L-asparticacid (PALA)' were gifts of the Developmental Therapeutics Program of the National Cancer Institute. [3H]Dopamine(47 Ci/mmol) and [3H] glycine (16 Ci/mmol) were obtained from Amersham (Arlington Heights, IL), and [3H]aspartate (23 Ci/mmol) and [3H]glutamate(17 Ci/mmol) were obtained from Du Pont-New England Nuclear. The partial cDNA clone for the rat glycine transporter was a gift of Dr. John Guastella, California Institute of Technology. Cells and Cell Culture-Wild type PC12 cells were obtained from Dr. David Schubert, Salk Institute.The PC12 variants, Int4 and42b, were obtained from Dr. Greg Shackleford, University of Southern California. Other PC12 variants were isolated in this lab as described in the text under "Results." Cells were cultured in Dulbecco's modified Eagle's medium containing 10% horse serum, 5% fetal calf serum, 50 pg of streptomycin/ml, and 50 units of penicillin/ml. Transport of Dopamine and Amino Acids-Cells were incubated with radioactive substrate at 37 "C, while the cells were still attached to 60-mm plastic dishes in a cell incubation buffer consisting of 125 mM NaCl, 4.8 mMKC1, 1.2 mM potassium phosphate, 1.3 mM CaC12, 1.2 mM MgSOI, 5.6 mM glucose, and 25 mM Tris HCI, pH 7.3. For Na+-free buffer the NaCl was replaced by 125 mM choline chloride. For chloride-free buffer, each chloride salt was replaced by the corresponding nitrate salt. For measurement of cellular uptake of radioactive substrate, the cell culture medium was aspirated from the dishes, and after the cells were washed twice with 2.5 ml ofincubation buffer, they were incubated in 1.5 ml of incubation buffer containing a radioactive substrate for the time periods indicated under"Results." Following the incubation, the cells were washed twice with 2.5 ml of incubation buffer and solubilized in 1 ml of 0.1% Triton X-100. One sample was taken for counting of the radioactivity by scintillation spectroscopy, and another sample was taken for the determination of the amount of protein using the Micro bicinchoninic acid protein assay kit from Pierce Chemical Co. Polymerase Chain Reaction (PCR) and Nucleotide Sequencing-

PC12 Cells

Two partial cDNA clones of the PC12 norepinephrine transporter were obtained by PCR. Wild type PC12 poly(A+) RNA (1 pg) was reverse-transcribed using a 768-fold degenerate primer corresponding to nucleotide positions 931-953 of the human norepinephrine transporter (Pacholczyk et al., 1991). The resulting first strand cDNA was subjected to 35 amplification cycles consisting of 2-min denaturation at 95 " c , 2-min annealing at 62 "C, and 3-min extension at 72 "C. For the amplification in one case we used degenerate oligonucleotide primers corresponding to the 547-566 and 931-953 nucleotide segments of the human norepinephrine transporter (Pacholczyk et al., 1991); these primers had 512- and 768-fold degeneracy, respectively. To obtain the second partial cDNA clone, we used oligonucleotide primers corresponding to the 232-251 and 647-673 nucleotide segments of the human norepinephrine transporter; the first of these primers was degenerate (512-fold)and thesecond was nondegenerate. DNA fragments of -410 and 450 base pairs (bp), respectively, were gel-purified and cloned into pBluescript KS(-) T vector (Marchuk e.? al., 1991). A partial cDNA clone of the M P T l glutamate/aspartate transporter was obtained by PCR after first reverse transcribing MPTl poly(A+) RNA (1 pg) using a 128-fold degenerate oligonucleotide primer corresponding to amino acids 303-310 of the rat brain glutamate/aspartate transporter cloned by Storck et al. (1992). Amplification was done as above using a 192-fold degenerate primer corresponding to amino acids 152-159 of the ratbrain glutamate/aspartate transporter and thesame primer used for cDNA synthesis. A cDNA fragment of -477 bp was cloned into pBluescript KS(-) T. Nucleotide sequencing of the transporter partial cDNA clones was performed in both directions by dideoxy chain termination using the Sequenase Version 2.0 kit from United States Biochemical (Cleveland, OH). Northern Blot Analysis-PC12 RNA was extracted with guanidine isothiocyanate as described by Chirgwin et al. (1979), and poly(A+) RNA was selected by the method of Aviv and Leder (1972). Approximately 4 pg of poly(A+) RNA was electrophoresed through a 1.5% agarose gel containing formaldehyde (Maniatis et al., 1982), transferred to nitrocellulose paper as described (Thomas, 1980), and hybridized to 32P-labeledsingle-stranded antisense RNA probes (3.5-4 X 10' dpm/pg), which had been synthesized from transporter partial cDNA clones. Autoradiography was at -70 "C using Kodak XAR-5 film and an intensifying screen. RNase Protection Ass~y-~~P-Labeled single-stranded antisense RNA probes (3.5-4 X 10' dpm/pg) were synthesized from the transporter partial cDNA clones described above, gel purified, and hybridized to 1 pg of poly(A+) RNA at 42 "C for 12-14 h. RNase digestion conditions (2.3 units/ml RNase A and 91 units/ml RNase T1 for 30 min at 37 "C) andfurther sample processing were as described in the directions for the RPA I1 kit (Ambion). CellToxicityAssays-Cellswere platedon 12-well dishes and cultured in standard medium containingvarious amountsof alanosine or PALA. At the indicated times the surviving cells were dislodged and counted in a hemocytometer. RESULTS

Uptake of Glycine, Aspartate, and Glutamate by PC12 Flat Cell and Wntl -expressing Variants-We have previously reported(Andersen et al., 1990) that severalspontaneously occurring flat cell variantsof PC12 had lost the ability of wild type PC12 to synthesize catecholamines and do not express tyrosine hydroxylase, an important enzyme of catecholamine synthesis; furthermore, at least one these variants exhibited no Na+-dependent uptake of dopamine (Bitler et al., 1986). Fig. 1shows that other PC12 flat cell variants, including those that were induced by expression of t h e Wnt-1 oncogene, are unable to take up dopamine. MPTl is a subclone of PC12 cells that weselected as cells resistant to the neurotoxin MPTP (Denton and Howard, 1987). 2068was isolated by US nonselectively as a flat cell subclone of PC12. I n t 4 and 4% were derived as subclones of wild type PC12 cells that had beeninfectedwith a retrovirus that expressed the Wnt-1 oncogene (Shackleford et al., 1993). We have confirmed that The abbreviations used are: PALA, N-(phosphonacety1)-L-aspar- PC12 cells, that express the Wnt-I gene, acquire the flat cell of thesefour flat cell tic acid PCR, polymerase chain reaction; bp, base pair(s); kb, kilo- variantphenotype.Althoughnone variants was able to take up dopamine, they all transported base(s).

Transporter Expression in Wild Type and Variant PC12 Cells

23893

aspartate, andglutamate into the other flat cell variants (not shown). The K , values for glutamate transport into the flat cell variants are comparable to that for glutamate transport in various preparations from mammalian brain (Schousboe, 1981; Nicholls and Attwell, 1990). As shown in Table I, the MPTluptake of aspartate is also similar to its uptake of glutamate in that each was strongly inhibited by L-aspartate, D-aspartate, L-glutamate, L-cysteate, and alanosine, and neither uptakewas dependent on chloride or markedly affected by D-glutamate, glutamine, cysteine, or PALA. These are characteristicsexhibited by the glutamate/ aspartate transporters from various tissues (Schneider et al., 1980; Sips etal., 1982;Kanner and Schuldiner, 1987) suggesting that in MPTl cells one transporter transports both asMinutes partate and glutamate. What little glutamate transport there FIG. 1. Time course of uptake of dopamine ( D A ) ,aspartate was into wild type PC12 cells, however, wasinhibited only by (asp),glutamate (glu),and glycine (gly)by wild type (WT) L-glutamate (Table I) and was not dependent on Na+. These and variant PC12 cells, which were incubated with 0.2 NM results indicate that the transporter responsible for glutamate ’H-labeled substrate for 2.6,or 10 min in standard buffer (0) uptake by wildtype PC12 cells is distinct from the glutamate/ or in Na+-free buffer (0). aspartate transporter of the flat cell variants. As shown in Table 11, glycine uptake by wild type PC12 was markedly inhibited by sarcosine but only poorly if at all TABLEI N

0.3 -\>

0

0.8

-\>

0.6

0.3 -\>

0.2

0.2

4

4

12 I

103

20

2

6

1/, x102

10

2

6

10

1/, *102

FIG. 2. Double reciprocal plots of substrate uptake as a function of substrate concentration for uptake after a 2-min incubation of the cells with substrate. V,, is picomoles/min/mg protein. WT-gly, glycine uptake by wild type PC12 cells; MPTI-asp, aspartate uptake by MPTl cells; MPTI-glu, glutamate uptake by MPTl cells.

aspartate and glutamate by a Na+-dependent process. Wild type PC12 cells did not exhibit any uptake of aspartate and only a very limited amount of glutamate (el%the rate of uptake into MPTl cells). It was not simply the inability of the flat cell variants to take up dopamine that enabled them to transport aspartate and glutamate. As shown in Fig. 1, essentially no uptake of aspartate or glutamatewas exhibited by two other PC12 variants, Guan’ and B8, that aremarkedly deficient in dopamine transport, but are not flat cell variants, and have all theother characteristics ofwild type PC12 including catecholamine synthesis and the ability to differentiate into neurons upon exposure to nerve growth factor (Bitler et al., 1986). Guan’ was selected by its resistance to guanethidine, and B8 was isolated nonselectively by subcloning wild type cells (Bitler et al., 1986). All of the variantstook up glycine as well as did wild type cells (Fig. 1). Other Properties of Aspartate, Glutamate, and Glycine Transport-The K,,, and Vmaxfor dopamine uptake by wild type PC12 was previously shown to be 0.4 pM and 7 pmol/ min/mg protein, respectively (Koide et al., 1986). The double reciprocal plots (Fig. 2) of glycine uptake by wild type cells and of aspartate and glutamate uptakeby the MPTl variant measured after a 2-min incubation with substrate revealed the apparent K,,, and Vmax,respectively, to be as follows: glycine (125 pM, 750 pmol/min/mg protein); aspartate (32 phi, 290 pmol/min/mg protein); glutamate (35 pM, 835 pmol/ min/mg protein). Similar values were obtained, respectively, for the transport of glycine into MPTl cells and glycine,

Effect of various compounds on the uptake of aspartate and glutamate by M P T l and wild type cells Cells were incubated as described under “Experimental Procedures” in incubation buffer containing the indicated additives, each at 200 PM. Where indicated, Na’ and C1- in the buffer were replaced, respectively, by choline and NO,. The values, which are percent of uptake in control buffer, are the means f S.D. for duplicate incubations. The means uptake for control MPTl cells was 12.8 f 0.2 pmol of aspartate and 38.5 2 0.8 pmol of glutamate/mg protein/5 min, and for control wild type (WT) cells it was 0.14 f 0.01 pmol glutamate/5 min/mg motein. MPTl WT Additive

Aspartate uptake

Glutamate uptake

100 1821 28 f 2 38 1 5 93 t 1 88t7 26 5 2 86 t 3 1 611 f 1 74 f 6 4 f 2 164 f 9

100 14 f 1 16 f100 1 24 f 3 72 f 3 108 f 9 19k 1 101 f 10 f1 60 t 4 521 92 f 10

None L-Aspartate D-Aspartate L-Glutamate D-Glutamate L-Glutamine L-Cysteate L-Cysteine L-Alanosine PALA Omit Na’ Omit C1-

Glutamate uptake

100 94 f 2 f1 23 f 5 89 t 3

ND” 93 f 4 ND 66 f 9 93 f 2 95 f 4 61 f 3

ND, not determined.

TABLEI1 Effect of various compounds or chloride ion on the uptake ofglycine by wild type PC12 Wild type PC12 cells were incubated as described under “Experimental Procedures” in incubation buffer containing the indicated additives, each at 2 mM. In some cases, C1- in the buffer was replaced by NO;. The values, which are percent of control, are the means f S.D. for duplicate incubations. The mean uptake in control buffer was 9.1 +. 0.4 pmol of glycine/5 min/mg protein. Buffer

Control

+ Glycine + AIB + MeAIB + Sarcosine + Alanine + Glutamine - Chloride

Uptake

100 12 f 1 78k 3 68 f 2 15 f 1 56 2 5 74 f 2 30 f 2

23894

Transporter Expression

in Wild Type and Variant PC12

Cells

by a-aminoisobutyricacid,methyla-aminoisobutyric acid, rine transporter. However, the PC12 norepinephrine transalanine, or glutamine. These results indicate that glycine is porterhas onlytwo consensus glycosylation sitesinthis transported into PC12by a transporter that canbe classified region, whereas the human norepinephrine transporter has as theGly type system accordingto the criteria (Christensen, three (Pacholczyk et al., 1991). We also employed PCR to isolate a partial cDNA for the 1984) used foramino acid transporters of non-neuronal cells. Isolation of Partial cDNA Clones for the PC12 Norepineph- glutamate/aspartate transporter from MPT1, Int4 cells, and rine and Glutamate/Aspartate Transporters-In order to de- 42b cells. Because theamino acidsequence, butnotthe nucleotidesequence, of theratbrainglutamate/aspartate termine whether the differences in transporter activities in wild type and variant PC12cells are due todifferences in the transporter (Storck et al., 1992) was available to us at the levels of transporter mRNA,we isolated partial cDNA clones time, we used mRNA from these flat cell variants and oligoof the rat brain for the PC12 norepinephrine and glutamate/aspartate trans- nucleotide primers corresponding to segments porters. The inferred amino acid sequences are given in Fig. glutamate/aspartate transporter. This PC12 flat cell glutamate/aspartate transporter cDNA segment, which is 477 bp 3. Twopartial cDNAclones of thePC12norepinephrine in length, is identical sequence in to thatof the corresponding transporter were isolatedby PCRusing wild typePC12 mRNA andoligonucleotide primers corresponding to the hu- segment of theratbrainglutamate/aspartatetransporter mannorepinephrinetransporter (Pacholczyk et al., 1991). clone described by Storck et al. (1992). The predicted amino acid sequence of the MPTl glutamate/aspartate transporter These PC12cDNAs,which are 408 and 442 bp in length, respectively, overlap with one another, and the segment theycloneisshown in Fig. 3; the 42b and Int4 sequences are cover together has 96% amino acid identity and88% nucleo- identical. For thefollowing studies we also useda partial cDNA clone tide identity to the corresponding segmentbetween the first of the first to sixth transmembrane regions of the rat brain and sixth transmembraneregions of the human norepinephglycine transporter, which was obtained from J. Guastella. Expression of the Norepinephrine,Glutamate/Aspartate, A * and Glycine Transporters-We measured transporter mRNA PC-NE-T : NVWRFPYLCfKNGGGAFLIPYTLFLIIAGMPLFYMELALGQFTiR levels in wild type and variant PC12 using antisense RNA Hu-NE-T : H V W R F P Y L C Y K N G C E L A L G Q Y N R synthesized from the partial cDNA clones of the norepineph* rine, glycine, and glutamate/aspartate transportersdescribed PC-NE-T : EGAATVWKICPFFKGVGYAVILIALYVGFYYNVIIAWSLYYLFA above. The Northern blotsof Fig. 4 showthat wild type PC12 Hu-NE-T: EGAATVWKICPFFKGVGYAVILIALYVGFYYNVIIAWSLYYLFS containsnorepinephrinetransporter mRNA and glycine 5.8 and 3.3 kb insize, respectively. transporter mRNA that are * * 0 0* * PC-NE-T : SFTLNLPWTNCGHAWNSPNCTDPKLLNASVLGDHTKYSKYKFTP The flat cell variant, MPT1,also contains glycine transporter Hu-NE-T: SPTLNLPWTDCGHTWNSPNCTDPKLLNGSVLGNHTKYSKYKFTP mRNA but not norepinephrine transporter mRNA. In con** * trast, MPTl contains glutamate/aspartate transporter PC-NE-T: AAEFYERGVLHLHESSGIHDIGLPQWQLLLCLMWIWLYVSLW mRNA (4.7 kb in size), but this message was not found in Hu-NE-T: AAEFYERGVLHLHESSGIHDIGLPQWQUCLMWVIVLYFSLW wild type PC12 cells. A Northern blot of Int4 and 42b RNA revealed that these cellsalso contain a 4.7-kb glutamate/ PC-NE-T : KGVKTSGKWWITATLPYFVLFVLLVHGVTLPGASNGINAYLHI Hu-NE-T: ~KTSGKVVWITATLPYFVLFVLLVHGVTLPGASNGINAYLHI aspartate transporter mRNA (data not shown). Fig. 5 compares the levels of the mRNAs for these three transporters in wild type PC12 and in the MPT1, Int4,42b, PC-NE-T: DFYRLKEATVWIDAATQIFF Guan', and B8 variants as determined by RNase protection Hu-NE-T: DFYRLKEATVWIDAATQIFF assays. In these experiments, mRNA from these cells were hybridized to 32P-labeled transporter antisense RNAs, diB PC-GLU-T: KENMYREGKIVQVTAADAFLDLIRNMFPPNLVEACFKQFKTSYE br-GLU-T: KENMYREGKIVQVTAADAFLDLIRNMFPPNLVEACFKQFKTSYE PC-GLU-T: KRSE'KVPIQANETLLGAVINNVSEAMETLTRIREEMVPVPGSVN br-GLU-T. KRSFKVPIQANETLLGAVINNVSEAMETLTRIREEMVPVPGSU 5.8 kb

PC-GLU-T: GVNALGLVVFSMCFGFVIGNMKECGQALREFFDSLNEAIMRLVA br-GLU-T : GVNALGLVVFSMCFGFVIGNMKEQGQALREFFDSLNEAIMRU

-

0 4.7Lb-

0 3.3tb-

*

-

-

285

0 -485

PC-GLU-T: VIMWYAPLGILFLIAGKILEMEDMGVI br-GLU-T: VIMWYAPLGILFLIAGKILEMEDMGVI

FIG.3. Inferred amino acid sequences of the partial cDNA transporter clones from PC12 wild type and flat cell variant cells. A, norepinephrine transporter obtained from wild type PC12 (PC-NE-T) compared to the corresponding segment (amino acids 78-318) of the human norepinephrine transporter(Hu-NE-2') cloned by (Pacholczyk et al., 1991). B, glutamate/aspartate transporter obtained from the MPTl variant(PC-GLU-T)compared to the corresponding segment (amino acids152-310) of the rat brain glutamate/ aspartatetransporter (br-GLU-T) cloned byStorck et al. (1992). Underlinedportions indicatethetransmembranedomains of the previously published transporters. *, a position of nonidentity; 0, a glycosylation site of the PC12 norepinephrine transporter. The nucleotide sequences of the partial cDNAclones of the PC12 transporters have been submitted to the Gen Bank data base.

111

FIG.4. Northern blots of the mRNAs for the norepinephrine transporter (NE-T), glutamatelaspartate transporter (GLUT ) , or glycine transporter (GLY-n in wild type (WT) PC12 and the MPTl variant. Samples of the same preparationof mRNA was used for all three blots. Autoradiography was for 36 h. Below eachblot is a photograph of thecorresponding gel stained with ethidium bromide.

Transporter ExpressionWild in

Type and Variant

PC12 Cells

23895

of the corresponding transporter. Guan' cells displayed somewhat greater norepinephrineuptake thandid B8 cells but had a lower level of norepinephrine transporter mRNA. Furthermore, the glutamate/aspartate transporter mRNA levels of 442 bp Int4 and 42b cells were substantially greater than that in M P T l cells, although there was little difference in the uptake of glutamate or aspartateby these cells. Finally, Figs. 4 and 5 both showed the level of glycine transporter mRNA to be greater in M P T l cells than in wild type cells although the latter exhibited greater uptake of glycine; this difference in FIG. 5. Levels of the mRNAs for the norepinephrine trans- mRNA levels was observed repeatedly (the mRNA preparaporter (NE-T), glutamate/aspartate transporter (GLU-!#?'), or tions used for the experiments of Fig. 4 were different from in wild type (W!#?') PC12 and in those used for Fig. 5). glycine transporter (GLY-!#?') the MPT1, 42b, Int4, B8, and Guan' variants. TheRNase Differential Sensitiviy of Wild Type andFlat Cells to Alanprotectionassays were performedusingsingle-stranded antisense osine and PALA-Alanosine and PALA are aspartateanalogs RNA probes as described under "Experimental Procedures." Auto- that aretoxic to cells because they inhibit nucleotide syntheradiography was for 24-36 h. C-I control lanescontained the intact antisense RNA probe hybridized to yeast RNA but undigested. The sis. Alanosine competitively inhibits aspartate uptake into undigested RNA probes included part of the vector sequencebetween murine lymphoblasts and is believed to enter cells by an the phage promoter and the cDNA insert and hence are longer than aspartate transporter (Anandaraj et dl., 1980; Tyagi and Coothe corresponding protected RNAfragments whose approximate sizes ney, 1984), whereas PALA is believed to enter cells by a are given next to the arrows. C-2 control lanescontained the RNA transporter-independent endocytosis process (White and probes hybridized to yeastRNA and digested withRNase. The Hines, 1984). The fact that alanosineinhibited aspartate absence of any protection demonstrates thespecificity of the assay. uptake into MPTl cells, whereas PALA did not (Table I), is consistent with the postulated mechanisms of entry for alangested with RNase, and subjected to gel electrophoresis and osine and PALA. We have obtained further evidence for these autoradiography. Because the transporter antisense RNAs mechanisms of entry by finding that MPTl and Int4 cells, were synthesized from partial cDNA clones, the protected which exhibited aspartate transport, were much more sensiprobes were shorterthanthe corresponding transporter tive to alanosine than were wild type cells (Fig. 6). Whereas mRNA to which they hybridized. The protected portion of wild type PC12 was unaffected by 3 p~ alanosine, multiplithe antisense probe was-442 bp for the norepinephrine cation of MPTl and Int4 cells was greatly reduced by as little transporter, -477 bp for the glutamate/aspartate transporter, as 0.1 pM alanosine; 0.3 p~ alanosine killed 80-90% of the and -710 bp for the glycine transporter, respectively. Samples MPTl and Int4cells within 6 days, and less than 1%of the of the same preparation of mRNA were used with all three MPTl and Int4 cells survived a 6-day exposure to 1-30 p~ transporter antisense RNAs. The autoradiographs were overexposed in an attempt todetect transporter mRNA in cells that were deficient in the corresponding transporter activity. pM ahnosine WT The abundance of norepinephrine transporter mRNA in wild type PC12 cells was strikingly greater than in any of the variants, all of which exhibited decreased catecholamine transporter activity. Similar results were obtained when we used the second of the two partial cDNAs for the norepinephrine transporter (data not shown). We have found that the 2068 variant also failed to express norepinephrinetransporter mRNA (data notshown). Only those variants that exhibited Na+-dependent uptake of glutamate and aspartate (i.e. the flat cell variants) contained glutamate/aspartate transporter mRNA. The major protected RNA was -477 bp in length as expected. Smaller sized RNA fragments were also protected to some degree; they could represent degradation products formed i n vivo. Glutamate/aspartate transporter mRNA was not detected in wild type PC12 or in the Guan' or B8 variants, none of which exhibited glutamate/aspartate transporter activity. However, there were comparable levels of glycine transporter mRNA in wild type and variant cells, consistent with the finding that wild type and variantcells all hadcomparable levels of glycine transporter activity. The data ofFig. 5 were obtained by RNaseprotection experiments, but similar conclusions were reached by Northern blots probed with 32P-labeled transporter cDNAs (data Days not shown). These results indicate that the observed differFIG. 6. Effect of varying concentrations of alanosine on the ences in the norepinephrine and glutamate/aspartate transporter activities in wild type and variant PC12 are due to viability of wild type ( WT) and flat cell variant (MPTI and PC12 cells. As described under "Experimental Procedures," differences in the expression of these two transporter genes. Znt4) cells were incubated in the presence of alanosine at theconcentration However, in those cells that exhibited a t least some activity (micromolar) indicated to theright of each curve, and the number of of a particular transporter there was not a strict correlation surviving cells was determined. The results are the means for duplibetween the rate of substrate uptake and thelevel of mRNA cate incubations.

23896

Transporter Expression in Wild Type and Variant PC12 Cells I

10 -

In

I

I

I

pM PALA

WT

MPT I

’Oo0 I

I

2

4

I

I

6

8

1I

Days

FIG. 7. Effect of varying concentrations of PALA on the viability of wild type (WT) and MPTl variant PC12 cells. As described under “Experimental Procedures,”cells were incubated in the presence of PALA at the concentration (micromolar) indicated to the right of each curve, and the number of surviving cells was determined. The results are the means for duplicate incubations.

alanosine. In contrast,wild type cells were killed by PALA as readily as were M P T l cells (Fig. 7). DISCUSSION

We haveshown that PC12 flat cell variants contain no detectablenorepinephrinetransportermRNA,but,unlike wild type cells, theydocontainmRNA fora glutamate/ aspartate transporter and they exhibit glutamate/aspartate transport activity. The transporter properties of flat cells induced by the Wnt-1 oncogene are the same as those of spontaneously occurring flat cells. We do not understand therole of the Wnt-1 gene product in producing the PC12 flat cell phenotype. Both of the PC12 subclones into which the Wnt-1 gene had been introduced (Int4 and 42b) contain Wnt-1 mRNA(Shackleford et al., 1993). Itseem unlikely that the phenotypeof the MPTl and 2068 variant cells is caused by Wnt-1, because several other spontaneously occurring flat cell variants do not express Wnt1 (Shackleford et al., 1993). However, M P T l cells have reWnt-3,’ and itmay be that other cently been found to express spontaneously occurring flatcell variants also expressa Wnt1 homolog. Glutamate/aspartate transporters are expressed both by certain neurons and by glia (Schousboe, 1981; Kanner and Schuldiner, 1987). We have concluded that the glutamate/ aspartate transporter expressed by the PC12 flatcell variants is the glial form of this transporter. The partial cDNA of the flat cell variant glutamate/aspartate transporter is identical in nucleotide sequence to the corresponding segment of the rat glutamate/aspartate transporter cloned by Storck et al. (1992). Frog oocytes injected with mRNA for that transporter showed an increased Na+-dependent uptakeover control oocytes of both aspartate and glutamate but not of several other amino acids examined (Storcket al., 1992). That transporter was found by Northern analysis to be expressed in rat brain but not inliver, kidney heart, or skeletalmuscle; in brain, the S. Zheng and B. D. Howard, unpublished results.

transporter was localized to glia by in situ hybridization studies (Storck et al., 1992). The glutamate/aspartate transporter mRNA from MPT1,42b, or Int4 cells is 4.7 kb, similar in size to the rat brain mRNA (4.5 kb) towhich the glutamate/ aspartate transporter cDNA of Storck et al. (1992) hybridizes. There is no striking homology of the MPT1, 42b, or Int4 glutamate transporter cDNA segments to either of the other two cloned glutamate transporters (Kanai andHediger, 1992; Pines et al., 1992), at least oneof which appears to be neuronal based on in situ hybridization experiments (Kanai and Hediger, 1992). These other two glutamate transporter cDNAs hybridize t o mRNAs of 3.5 or 11 kb, respectively (Kanai and Hediger, 1992; Pines et al., 1992). Although the MPT1, 42b, and Int4 variants express a glial-type glutamate transporter, we have not been able to detect the expression of another glial marker, glial fibrillary acidic protein, in these cells.3 The fact that the glutamate/aspartate transporter found in PC12 variants expressing Wnt-1 is of the glial type suggests that one role of Wnt-1 in brain development might be the induction of glial cell differentiation.Since glial cells are known to be essential for neuronal viability, inadequate glial formation may have been responsible for the abnormal brain development observed inWntl-deficient mice (McMahon and Bradley, 1990; Thomas and Capecchi, 1990). Wnt-1 or anotherWnt homolog may alsofunction toinduce glial differentiation during the development of neural crest derivatives that become peripheral neurons, Schwann cells, and adrenalmedullary cells among others (for review of neural crest developmentsee Anderson (1989) and Patterson(1990)). There is evidence that at least some cells of the neural crest so as to lineage have the potential to change their phenotype adopt the characteristics of other derivatives. For example, mature adrenalmedulla chromaffin cells can differentiate into neurons when exposed to nerve growth factor in the absence of steroids(Unsicker et al., 1978). Stemple and Anderson (1992) have shown that asingle ratneuralcrest cell can differentiate into either neurons or glia. PC12 are adrenal medullary cells and thus are derived from the neural crest. The flat cell PC12 variants may represent a reversal of the neural crest differentiation pathway toa glial-precursor form that does not exhibit allof the characteristicsof glia, but does have glial glutamate/aspartate transporter activity. In this regardit is of interestthatSchwann cellsisolatedfrom neonatal rat sciatic nerve and maintained in culture for several passages also exhibit Na+-dependent uptake of aspartate and glutamate.4 The process by which the flat cell variants of PC12 terminate expression of the norepinephrine transporter may be similar to that operating in vivo in mammalian brain where specific and localized down-regulation of neurotransmitter transporters occurs under various conditions. For example, Bannon et al. (1992) demonstrated a precipitous decrease in the level of dopamine transporter mRNA in the substantia nigra of normal humans older than 57 years. This decrease was not thought to be due only to death of dopaminergic neurons, because the level of mRNA for tyrosine hydroxylase, a key enzyme of dopamine synthesis, did not exhibit the same precipitous decrease. It is possible that such adecrease in transporter expressionin vivo could be induced by the abnormal production of an autocrine or paracrine factor that acts analogously to Wnt-1and that exerts its effect on a neighboring population of neurons. This mechanism would explain the sudden onset and circumscribed localization of the diminished C. F. Wu and B. D. Howard, unpublished results. B. Ramachandran, K. Houben, Y. Y. Rozenberg, J. R. Haigh, A. Varpetian, and B. D. Howard, unpublished results.

Transporter Expression Wild in Type

and Variant PC12

Cells

23897

Giros, B., Mestikawy, S. E., Bertrand, L., and Caron, M. G. (1991) FEBS Lett. dopamine transporter expression occurring in humans over 2 9 5 , 149-154 Greene, L. A. (1978) J. Cell Biol. 7 8 , 747-755 57 years of age (Bannon et al., 1992). Greene, L. A,, and Rein, G. (1977) Brain Res. 1 2 9 , 247-263 Specific down-regulation of neurotransmitter transporter Greene, L. A., and Tischler, A. S. (1976) Proc. Natl. Acad. Sci. U. S. A . 7 3 , 2424-2428 activity may also be responsible for some neurodegenerative J., Nelson, N., Nelson, H., Czyzyk, L., Keynan, S., Miedel, M. C., diseases. Rothstein et al. (1992) showed that synaptosomes Guastella, Davidson, N., Lester, H. A,, and Kanner, B. I. (1990) Science 2 4 9 , 13031 qnfi prepared fromthe spinal cord, motor cortex, or somatosensory Guastella, J., Brecha,N., Weigmann, C., Lester, H. A., and Davidson, N. (1992) cortex taken at autopsy from individuals with amyotrophic Proc. Natl. Acad. Sei. U. S. A. 89. 7189-7193 Guerrero, L., Pellicer, A. & Burstein, D. E. (1988) Biochem.Biophys.Res. lateralsclerosis exhibited a decreaseduptakeofglutamate Cornmun. 1 6 0 , 1185-1192 (decreased Vmar;normal I C m ) , whereas uptake of y-aminobu- Iversen, L. L., (1970) Adu. Biochern. Psychopharmncol. 2,109-132 tyric acid by synaptosomes from the motor cortex was normal. Jue, S. F., Bradley, R. S., Rudnicki, J. A,, Varmus, H. E., and Brown, A. M. C. (1992) Mol. Cell. Biol. 12,321-328 Although death of glial cells or glutaminergic neurons could Kanai, Y., and Hediger, M. A. (1992) Nature 360,467-471 not beruled out as the principlecause of the decreased Kanner, B. I., and Schuldiner, S. (1987) CRC Crit. Reu. Biochem. 2 2 , 1-38 Kilty, J. E., Lorang, D., and Amara, S. G. (1991) Science 264,578-579 glutamate transport, the authors suggestedthe possibility that Koide, M., Cho, A. K., and Howard, B. D. (1986) J. Neurochem. 47,1277-1285 decreased clearance of glutamate from affected synaptic clefts Maniatis,T.,Fritsch, E. F., and Sambook, J. (1982) Molecular Cloning: A Laboratory Manual, pp. 1-545, Cold Spring Harbor Laboratory,Cold Spring could increase the likelihood of neuronal degeneration. It is Harbor, NY known that overstimulation of glutamate receptors by gluta- Marchuk. D.. Drumrn., M... Saulino.. A,.. and Collins. F. S. (1991) . . Nucleic Acids Res. 1 9 , 1154 mate and related compounds can be toxic to neurons (Choi, McMahon, A. P., and Bradley, A. (1990) Cell 62,1073-1085 1992). Nicholls, D., and Attwell, D. (1990) Trends Pharmncol. Sci. 1 1 , 462-468 M., KO, M., Ogura, A,, Liu, D., Amano, T., Takano, T., and Ikawa, Y. The studies reported here addto the growing evidencethat Noda, (1985) Nature 318.73-75 changes in the activities of transporters for neurotransmitter Nusse, R., and Varmus, H. E. (1992) Cell 6 9 , 1073-1087 Nusse, R., van Ooyen, A., Cox, D., Fung, Y. K., and Varmus, H. (1984) Nature reuptakecanserve as useful markersof developmentand 307,131-136 degeneration of the nervous system. Pacholcyzk, T.,Blakely, R. D., and Amara, S. G. (1991) Nature 3 6 0 , 350-353 " "

Acknowledgements-We thank Dr. Greg Shackleford for donating the Wnt-expressing PC12 variants and for sharing the information of the manuscript with us before its publication, and we thank Dr. John Guastella for the glycine transporter cDNA. We also thank Drs. Robert Edwards and Ali Roghani for advice on the PCR experiments, and Steve Chen, Sandra Cheng, and Sy Vu for help with some of the experiments. REFERENCES Anandaraj, S. J., Jayaram, H. N., Cooney,D. A., Tyagi,A. K., Han, N.,Thomas, J. H., Chitnis, M., and Montgomery, J. A. (1980) Biochem. Pharmacol. 2 9 , 227-245 Anderson, D. J. (1989) Neuron 3, 1-12 Andersen, J. K., Zhang, M.-b., Zhong, X.-h., Rozenberg, Y. Y., and Howard, B. D. (1990) J. Neurochem. 5 5 , 559-567 Aviv, H, and Leder, P. (1972) Proc. Natl. Acad. Sci. U. S. A. 6 9 , 1408-1412 Bannon, M. J., Poosch, M. S., Xia, Y. Goebel, D. J., Cassin, B., and Kapatos, G. (1992) Proc. Natl. Acad. Sci. U. i.A . 89,7095-7099 Bitler, C. M., Zhang, M.-B., and Howard, B. D. (1986)J. Neurochem. 47,12861293 Bothwell, M. A., Schechter, A. L., and Vaughn, K. M. (1980) Cell 21,857-866 Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. (1979) Biochemistry 18,5294-5299 Choi, D. W. (1992) Science 2 6 8 , 241-243 Christensen, H. N. (1984) Biochim. Biophys. Acta 7 7 9 , 255-269 Denton, T., and Howard, B. D. (1987) J. Neurochem. 9,622-630

Patterson, P. H. (1990) Cell 62,1035-1038 Pines, G., Danbolt, N. C., Bjoras, M., Zhang, Y., Bendahan, A,, Eide, L., Koepsell, H., Storm-Mathisen, J., Seeberg, E., andKanner, B. I. (1992) Nature 360,464-467 Rothstein, J. D., Martin, L. J., and Kunci,R. W. (1992) N . Engl. J. Med. 3 2 6 , 1464-1468 Sassone-Corsi, P., Der, C. J., and Verma, I. M. (1989) Mol. Cell. Biol. 9,317421R2

G., Hammerman, M. R., and Sacktor, B. (1980) J . Biol. Chem.

--- r""

Shimada, S., Kitayama, S., Lin, C.-L., Patel, A,, Nanthakumar, E., Gregor, P. Kuhar, M., and Uhl, G. (1991) Science 264,576-578 Sips, H. J., De Graaf, P. A., and Van Dam, K. (1982) Eur. J. Biochem. 1 2 2 ,

"-3.59-3fiA ".

Smith, K. E., Borden, L. A., Hartig, P. R., Branchek, T., and Weinshank, R. L. (1992) Neuron 8 , 927-935 Stemple, D. L., and Anderson, D. J. (1992) Cell 71,973-985 Storck, T., Schulte, S., Hofmann, K., and Stoffel, W. (1992) Proc. Natl. Acad. Sci. U. S. A. 89,10955-10959 Sugimoto, Y., Noda, M., Kitayama, H., and Ikawa, Y. (1988) J. Biol. Chem. 263,12102-12108 Thomas, K. R., and Capecchi M. R. (1990) Nature 346,847-850 Thomas, P. (1980) Proc. Natl: Acad. Sci. U. S. A . 77,5201-5205 Tyagi, A. L., and Cooney, D. A. (1984) Adu. Pharmacol. Chemother. 2 0 , 69n l 1

1Ll

Unsicker, K., Krisch, B., Otten, U., and Thoenen, H. (1978) Proc. Natl. Acad. Scz. U. S. A. 76,3498-3502 White, J. C., and Hines, L. H. (1984) Cancer Res. 44,507-513