SUBCELLWAR LOCALIZATION OF GAMMA-AMINOBUTYRATE
METABOLISM AND TRANSPORT OF GAMMA-AMINOBUTYRATE IN PLANTS
A Thesis
Presented to
The Faculty of Graduate Studies
of
The University of Guelph
by
KEVW EDMUND BREITKREUZ
In partial fulfillment of requirements
for the degree of
Doctor of Philosophy
January, 1999
O Kevin Edmund Breitkreuz, 1999 National Library Bibliothèque nationale I*m of Canada du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395, rue Wellington Ottawa ON K1A ON4 Ottawa ON K1A ON4 Canada Canada Your fi& Votre réference
Our ala Notre reftirence
The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Librw of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in rnicrofom, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfichelfilm, de reproduction sur papier ou sur format électronique.
The author retaùis ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fbm it Ni la fhèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. ABSTRACT
SUBCELLULAR LOCALIZATION OF GAMMA-AMINOBUTYRATE METABOLISM AND TRANSPORT OF GAMMA-AMINOBUTYRATE IN PLANTS
Kevin Edmund Breitkreuz Advisors: University of Guelph, 1999 Professor Barry J. Shelp Professor Wolf B. Frornmer
This study examined the subcellular localization of the enzymes involved in the
metabolism of 4-aminobutyrate or gamma-arninobutyrate (GABA) and the transport of
this compound across membranes of higher plant cells. The subcellular localization of
GABA shunt enzymes was investigated in protoplasts prepared from developing soybean cotyledons (Glycinemar (L.) Memll cv. Maple Arrow). Protoplast lysate was
fractionated by differential and continuous Percoll-gradient centrifugation to separate organelle fractions. Glutamate decarboxylase (EC 4.1.1.15) was found exclusively in the cytosol, whereas GABA:pyruvate transaminase (EC 2.6.1.19) and succinic sernialdehyde dehydrogenase (EC 1.S. 1.16) were exclusively associated with the rnitochondrial
Fractions. Mitochondrial fractions also catabolized W-'"C]GABA to labeled succinate.
These result provide conclusive evidence that GABA synthesized in the cytosol is transported across the rnitochondrial membranes into the matrix. In an effort to identifi proteins capable of transporùng GABA, heterologous complernentation and characterization of known amino acid transporters were perfonned using a GABA- transport-deficient Saccharomyces cerevisiae mutant. AAP3 and ProT2 from
Arabidopsis thaliana (L.) Heynh (Landsberg erecta ecotype) were identified as H+- coupled GABA transporters with Michaelis constants of 12.9 t 1.7 and 1.7 I0.3 rnM, respectively. The effect of extemal pH on the simultaneous transport of [1-14C]GABA and [2. 3-3~]prolineinto yeast expressing Pr072 provided evidence that the charged state of GABA is an important parameter in substrate transport. Pro=-mediated [1-
14 C]GAE!A transport was inhibited by proline, choline and quatemary ammonium compounds (QACs) such as glycine betaine. Direct evidence was provided for the transport of [methyl-'%]choline. Although different metabolic routes are involved in the syntheses of these stress-related compounds, ProT2 may be a comrnon carrier in plants for these metabolites. These acknowledgments conclude my requisite as a graduate student at the University of Guelph. During my tenure, under the supervision of Dr. Barry Shelp (University of
Guelph) and Dr. Wolf Frommer (University of Tübingen, Gerrnany), 1was able to pursue my interests in a special ubiquitous, four-carbon compound that accumulates in plants in response to stress ...... I am gratefully indebted to the combined efforts of my menton who have left me with a most memorable experience, teaching me how to think, read and write cntically. The repertoire of scientific methods 1 have learned are invaluable. Their collective kindness and generosity will not be forgotten. Thank-you both.
1 would also Iike to thank Drs. Judith Strornrner, Dave Wolyn and Alan Bown (Brock
University) for their valuable imput and criticisms as members of my advisory cornmittee. Special thanks to Dr. Edwardo Blumwald (University of Toronto) for his excellent and thorough job as the extemal examiner.
Many thanks also, to al1 my many laboratory colleagues on both sides of the Atlantic, who over the years have offered advice, support and most importantly, good times. These include (in no particular order): Rainer Schwacke, Brent Kaiser. Catharine Scott-Taggart,
Hanjo Hellrnmn, Wayne Snedden, Henrik Buschmann, Mechthild Tegeder, Dons
Rentsch, Owen Van Cauwenberghe, Sonia Gazzarrini and Mike McLean. I would also like to heartily thank Nice Barker for her assistance, cottage loans and nice foods.
Most importantly, I'd iike to thank Eddy, Ruth, Katie, Linda and Brian, my family.
Their continuous love and confidence was the reason for my perseverance. This book is for them. vii
2.1. Introduction ......
2.2. Stress-related Nitrogenous Compounds in Plants ......
I . GABA ......
I . GABA Metabolism ......
2 . Regulation of GABA Levels ......
3 . Subcellular Localization of the GABA Shunt ......
4 . Roles of GABA ......
1. Biochernical pH Stat ......
2 . Plant Defense ......
3 . Krebs Cycle Bypass ......
4 . Nitrogen Storage and Transport Compound ......
5. Plant Development ......
6 . Compatible Osmolyte ...... 1. Proline Metabolism ......
2 . Regulation of Proiine Levels ......
3. Roles of Proline ......
3 . Glycine Betaine ......
2.3. Amino Acid Transporters ......
I . Amino Acid Transporters in Plants ......
1. Physiological and Biochemical Descriptions ......
2 . Molecular Approaches ......
2 . GABA Transportes in Eukaryotes and Prokaryotes ......
1. Animal GAB A Transporters ......
2 . Fungal GABA Transporters ......
3. Bacterial GAB A Transporters ......
4 . Evidence for GABA Transporters in Plants ......
2.4. Research Objectives ......
CHAPTERTHREE . Subcellular compartmentation of the 4-aminobutyrate shunt in
protoplasts from developing soybean cotyledons ......
3.1. Introduction ......
3.2. Material and Methods ......
1. Plant Material ......
2 . Protoplast Isolation ......
3 . Subcellular Fractionation ......
4 . Enzyme Assays ......
5 . Radiolabelhg Study ...... 3.3. Results ......
3.4. Discussion ......
CHA~RFOUR . Identification and characterization of GAB A, proline and QAC
transporters frorn Arabidopsis ......
4.1. Introduction ......
4.2. Matends and Methods ......
1. Plant Growth ......
2 . MoIecular Cloning ......
3 . Yeast Growth, Transformation and Selection ......
4. DNA Sequencing ......
5 . Yeast Growth Assay ......
6 . Yeast Transport Assay ......
4.3. Results ......
Growth of Arabidopsis on GABA ......
Heterologous Complementation of a GABA Transport Mutant ......
Growth Assay of Yeast CompIernents ......
Characterization of AAP3 and ProT2 as GABA Transporters ......
Characterization of ProT2 as a QAC Transporters ......
4.4. Discussion ......
CHAPTERFIVE . GENERALDISCUSSION ......
REFERENCES ......
.iv- Table Page
1.1. Animai, fun@ and bacteriai GABA transporters and their substrate
affinities ......
2.1. Families of Arabidopsis thaliana amino acid transporters ......
3.1. Distribution of protein. organelle markers, and GABA shunt enzymes in
various fractions resulting from differential centrifugation of lysed
protoplas ts......
4.1. Effect of pH on the zwitterionic composition of substrate and the
simultaneous transport of [l-"CIGABA and [2, 3-I4C]prolineinto yeast
cells expressing Pm72 ...... Figure Page
The GABA shunt in relation to the Krebs cycle ......
Chernical structures of some amino acids, glycine betaine and precursors and
O ther betaines ......
Distribution of marker and GABA shunt enzymes among organelle fractions
Growth of Arobidopsis and yeast complements on media supplemented with
GABA as the sole nitrogen source ...... -......
Concentration dependence of GABA and proline transport by AAP3'
and ProT2 ......
B iochemical characterization of ProT2 ...... -.....
Time course of [methyl-'%]choline uptake into yeast cells expressing
Pro TS ......
Sumrnary of current knowledge of GABA metabolism and transport in
plants ...... ACC 1-aminocyclopropane 1-carboxylic acid
AAP amino acid permease
2-ABA 2-aminobutyric acid
3-ABA 3-aminobutyric acid
4-ABA 4-aminobutyric acid (GABA)
ADH alcohol dehydrogenase
ADP adenosine 5'-diphosphate
Arnpso 3-[( 1.1 -dimethyl-2-hydroxyethyl)amino]-2-hydroxy-prop~esulfonate
AOA aminooxyacetate
3-APA 3-aminopropionic acid (p-aianine)
ATP adenosine 5'-triphosphate
BADH betaine aldehyde dehydrogenase
Bet glycine betaine (nicotinic acid N-methytbetaine)
BetAld glycine betaine aldehyde
Beton betonicine (4-hydroxy-proline betaine) bp base pair
BS A bovine serum alburnin
CaM calmodulin
Cam camitine (4-arnino-3-hydroxybutyrïcacid aimethyl betaine)
Cat catalase cDNA complementary DNA
-vii- Cho1 choline
CM0 choline monoxygenase
Cyt c cytochrome c
Cyt c ox cytochrome c oxidase
DW dry weight
Ecto ectoine
FOA 5-fluoro-oro tic acid
FW fresh weight
GABA 4-aminobutyric acid
GAB A- Ald GABA aldehyde
GAI3 A-T GABA transaminase
GAD glutamate decarboxylase
GAP 1 general amino acid permease
GDH glutamate dehydrogenase
GOGAT glutamine-oxoglutarate amidotransferase
GOT glutamate-oxaloacetate transaminase
GP glutamic y-phosphate
GS glutamine synthetase
GSA glutarnic y-semialdehyde
GPT glutamate-pyruvate transaminase
HF'LC high pressure liquid chromatography
IDH isocitrate dehydrogenase
Km Michaelis constant 2-ME 2-mercaptoethanol
Mes 2-IN-morpholinoJethanesulfonicacid
Mops 3-m-morpholino]p:opanesulfonic acid
NAD(P)' nicotinamide adenine dinucleotide (phosphate) - oxidized
NAD(P)H nicotinamide adenine dinucleotide (phosphate) - reduced
NCBI National Center for Biotechnology Information
NMR nuclear magnetic resonance
OD optical density
ODH 2-oxoglutarate dehydrogenase
P5C A'-pyrroline-5-carboxylate
PSCDH A'-pyrroline-5-carboxylate dehydrogenase
P5CR A'-pyrroline-5-carboxylate reductase
PSCS A'-pyrroline-5-carboxylatesynthetase
PDH proline dehydrogenase
PLP pyridoxal5'-phosphate
ProT proline transporter
PUT4 proIine/GABA pemease
QAC quaternary ammonium compound
S.D. standard deviation of mean
SK succinyl-CoA kinase
SSA succinic serniaidehyde
SSADH succinic semialdehyde dehydrogenase
Tricine N-Tris [hydroxymethy11-methylglycine
-ix- Trig trigonelline (nicotinic acid N-methylbetaine)
Tris Tris[hydroxymethyl] aminomethane
UGA4 GABA permease
URA3 uracil synthase
UTR untranslated region
Vnmr maximum reaction velocity w7 N-(6-aminohexyl)-5-ckloro-1 -napthalenesulfonamide
WT wild-type CHAPTERONE
Introduction
4-Aminobutyrate or y-arninobutyrate (GABA) is a ubiquitous, four carbon, non- protein arnino acid found in higher plants, animais, fungi and bacteria (Bown and Shelp,
1989; Satya Narayan and Nair, 1990; Bown and Shelp, 1997). GABA is synthesized almost exclusively by the irrevenible a-decarboxylation of L-glutamate by glutamate decarboxylase (GAD; EC 4.1.1.15) (Chung et al., 1992; Tuin and Shelp, 1994).
Subsequently, GABA is catabolized by GABA transaminase (GABA-T; EC 2.6.1.19) and succinate serniaidehyde dehydrogenase (SSADH; EC 1.2.1.16) to succinate, an important
Krebs cycle metabolite (Tuin and Shelp, 1994; Tuin and Shelp, 1996). The conversion of glutamate carbon to succinate by these three reactions is referred to as the GABA shunt
(Fig. 1.1) and facilitates the entry of glutamate carbon into the Krebs cycle (Dixon and
Fowden, 196 1; Streeter and Thompson, 1972~1,b; Satya Narayan and Nair, 1986) and glutamate nitrogen into GABA and alanine (Tsushida and Murai, 1987).
The concentration of GABA in plants is markedly stimulated by a variety of stress conditions, including hypoxia, temperature shock, mechanical manipulation and damage, water stress and phytohormones, but little is known about the physiological role of
GABA and the regulation of its metabolisrn (Bown and Shelp, 1989; Satya Narayan and
Nair, 1990; Bown and Shelp, 1997). Although GABA accumulation is probably rnediated via an activation of GAD (Snedden et al., 1995; Bown and Shelp, 1997), decreased catabolism by GABA-T and SSADH (Satya Narayan and Nair, 1990) or changes in intracellular or intercellular transport may also contribute to changes in GABA
-1- Figure 1.1. Simplified metabolic diagram of the GABA shunt (bold) in relation to the
Krebs cycle. GAD, glutamate decarboxylase; GABA-T, GABA transaminase; GDH, glutamate dehydrogenase; GOT, glutamate-oxaloacetate transaminase; GPT, glutamate- pyruvate transaminase; ODH, 2-oxoglurate dehydrogenase; SK, succinyl-CoA kinase;
SSADH, succinic serniddehyde dehydrogenase (based on Bown and Shelp, 1997).
concentrations.
In plants, the subcellular localization of enzymes that constitute the GABA shunt
remains controversial, especially the localization of GABA-T and SSADH (Bown and
Shelp, 1989). Previous studies have suggested that GABA-T is rnitochondrial (Tokunaga
et al., 1976), or both cytosolic and mitochondrial (Dixon and Fowden, 1961; Wailace et
al., 1984). A dual cytosolic and mitochondrial localization has also been suggested for
SSADH (Satya Narayan and Nair, 1986). It is possible that the appearance of GABA-T
and SSADH activities in the soluble/cytosolic fractions is a consequence of mitochondrial
breakage during extraction and fractionation procedures. In animals, GABA-T and
SSADH are found in mitochondria and GAD is localized to the cytosol (Hearl and
Churchich, 1984). Separation of GAD activity from GABA-T and SSADH activities has
important implications for the regulation of GABA accumulation.
By using improved techniques for organelle fractionation that circumvent many of the
problematic methods used in previous studies (Le. tissue homogenization and differential
centrifugation), the hypothesis that the GAD is localized in the cytosol, whereas GABA-T
and SSADH are localized in mitochondria, was tested. Organelles were isoIated from
protoplasts of developing soybean (Glycine max (L.) Memll cv. Maple Arrow)
cotyledons and purified by continuous Percoll-gradient centrifugation; then, the
subcellular localization of the GABA shunt enzymes was detennined in organelles.
Evidence in support of the hypothesis, in conjunction with a survey of the animal,
fungal and bacterial literature, prompted the formulation of a second hypothesis that proteins transport GABA across plant membranes. GABA transport proteins, which Vary
in sequence homology, localization and substrate specificities, have been isolated and
-4- well charactenzed in representative species from three kingdoms (Table i .l). Most use
GABA, but also transport other compounds such as glycine betaine, proline, 3- arninopropionic acid @-alanine) and taurine. Heterologous complementation of a
GABA-transport-deficient yeast mutant (Grenson et al., 1987) and analysis of previously isolated amino acid transporters (see Fischer et al., 1998 for review) were used to test the second hypothesis using Arabidopsis thaliana (L.) Heynh (Landsberg erecta ecotype) as a mode1 system.
Table 1.1. (Cont'd)
Bet, 3-APA Borden et al. (1995) Bet (0.2), 3-APA Lopez-Corcuera et a\. (1 992)
VGAT R. irorvegicus McIntire et al. (1997) C.elegms McIntire et al. (1 997)
Fungal UGA4 S. cerevisiae n.a. Grenson et al. (1987) PUT4 S. cerevisiae Pro (0.02) Jauniaux et al. (1 987) GAPI S. cerevisiue tnost AA, Cit Grenson et al. (1987)
I Y Bacterial GabP B. subtilis Niegemann et al. (I 993) E. coli Ferson et al. (1 996)
a This table is representative and not coinplete. For animal systems, only the first gene cloned in each species is shown. b GAT, GABA transporter; BGT, betaine-GABA transporter; VGAT, vesicular GABA transporter. C R., Rattus; H, , Homo; M., Mus; T,Totpedo; C., Cariis; C., Caenorhabditis; S., Sacchai-oniyces;E., Escherichia; B, Bacillus. d Reported as either Kr?,or Ki; na, not available from the literature; 3-APA, 3-aminopropionate (P-Ala); Tau, taurine;
Bet, glycine betaine; AA, amino acids; Cit, citrulline; 3-ABA, 3-aminobutyrate. CHAPTERTWO
Literature Review
2.1. Introduction
In the lirerature review, the metabolism, function(s) and implications to plants of important stress-related nitrogenous compounds such as GABA, proline and glycine betaine are discussed, with particular attention given to GABA. This is followed with discussions of general arnino acid transport in plants and GABA transporters in animals, fungi and bacteria.
2.2. Stress-reIated Nitrogenous Compounds in Plants
2.2.1. GABA
GABA is a four-carbon amino acid which possesses an amino group on the fourth or y-carbon. The location of the amino group relative to the a-carboxyl group prevents the utilization of GABA as a protein amino acid. GABA exists in a free unbound form, and is zwitterionic in nature, with pK values of 4.03 and 10.56 (Anonymous, 1972). GABA is highly soluble in water (Baliantyne and Chamberlin, 1994). Stnicturally, GABA is a very flexibIe molecuIe which can assume a number of conformations in solution including a cyclic structure (Tanaka et al., 1978; Christensen et al., 1994; King et al., 1995; Brechtel et al., 1996). The stability of the cyclic or ring-like conformation is likely dependent on ionic interactions between the arnino and carboxyl groups; these interactions are strongest in the zwitterion. According to Chnstensen et al. (1994), cyclic GABA is structurally similar to proline and may, like 3-aminopropionic acid, be viewed as a biological analog.
-8- The structures of cyclic GABA, proline, 3-aminopropionic acid, and two structural isomers of GABA, 2- and 3-aminobutyric acid (2-ABA and 3-ABA), are shown in Fig.
2.1 .A.
Typicaliy, GABA levels in plant tissues are low ranging from 0.03-2 prnol g-' FW
(Rhodes et al., 1986; Satya NaMyan and Nair, 1990, Fougère et al., 1991) and increase several fold in response to many diverse stimuli (reviewed in Bown and Shelp, 1989,
1997; Satya Narayan and Nair, 1990). For example, mechanical or cold stimulation of soybean leaves increases GABA concentrations by 20- to 40-fold within 5 min to 1 to 2 pmol g-' FW (Wallace et al., 1984). Even higher GABA levels have been documented in suspension cells subjected to water deficits. Tomato ceils adapted to medium containing
25% polyethylene glycol contained 13 pmol GABA g-' FW (- 13 mM) (Rhodes et al.,
1986) and 32.9 mM GABA (Handa et al., 1983); respective increases of 5- and 15-fold greater, in cornparison to non-adapted control cells. Sirnilarly, tobacco cells adapted to
428 mM NaCl accumulated GABA to levels to 6.72 rnM, a value 7-fold higher than control cells (Binzel et al., 1987) . Other conditions such as heat shock (Mayer et al.,
1990), mechanical stimulation (Ramputh and Bown, 1996). hypoxia (Streeter and
Thompson, L972a; Tsushida and Murai. 1987; Reggiani et al., 1988; Aurisano et al.,
199%; Ratcliffe, 1995) and phytohormones (Ford et al., 1996)-
2.2.1.1. GABA Metabolism
The major route of GABA synthesis is by the direct and irreversible a- decarboxylation of glutamate by glutamate decarboxylase (GAD, EC 4.1. l. 15, Reaction 1)
(Bown and Shelp, 1997).
-9- Figure 2.ï. Chernical structures of various amino acids (A), glycine betaine and precursors (B), and other betaines (C). 2-ABA; a- or 2-aminobutyric acid: 3-ABA; P- or
3-aminobutyric acid; 4-ABA, 4-arninobutyrate (GABA); 3-APA, 3-aminopropionic acid
@-alanine);Bet, glycine betaine; BetAld, glycine betaine aldehyde; Beton, betonicine;
Cam, carnitine; Chol, choline; Ecto, ectoine; Pro, proline; Trig, tngonelline Pro
Ecto
Bet
Beton Trig Reaction 1: Glutamate + H' - GABA + CO2
Reaction HA: GABA + pynivate * succinic semialdehyde + alanine
Reaction W: GABA + 2-oxoglutarate * succinic semialdehyde + glutamate
Reaction III: succinic semialdehyde + NAD' - succinate + NADH + H+
Evidence for this decarboxylation reaction by GAD is provided by the in vivo conversion of [1-"C]giutamate to "CO,and unlabeled GABA (Chung et al., 1992; Tuin et al., 1994). GAD activity has been identified and characterïzed in many plant species and tissues (Bown and Shelp, 1989 and refs. therein; Snedden et al., 1992; Ling et al.,
1994; Snedden et al., 1995). It is specific for L-glutamate and exhibits a sharp acidic pH optimum of about 5.8 (Tsushida and Murai, 1987; Snedden et ai., 1992, 1995; Shelp et al., 1995). It is pyridoxal5'-phosphate (PLP) dependent and is inhibited by reagents known to react with sulfhydryl groups (Satya Narayan and Nair, 1985; Tsushida and
Murai, 1987). A cDNA encoding GAD was cioned by screening a petunia cDNA expression library using a "s-recombinant calmodulin probe (Baum et al., 1993), providing the first evidence that GAD possesses a calrnodulin (CaM) binding domain.
Subsequent studies demonstrated the involvement of this CaM binding domain in the regulation of GAD activity (See Section 221.2.; Ling et al, 1994; Arazi et al., 1995;
Snedden et al., 1995, 1996). GAD homotogues from tomato (Gallego et al., 1995), tobacco (Yun and Oh, 1998) and Arabidopsis (Turano and Fang, 1998; Wc et al., 1998) have been recently cloned. Regulation and the subcellular localization of GAD is discussed in Sections 2.2.1.2. and 2.2.1.3.
The second enzyme of the GABA shunt, GABA transaminase (GABA-T, EC
-12- 2.6.1.19) catalyzes the reversible conversion of GABA to succinic semialdehyde (SSA) using either pyruvate (Reaction IIA) or 2-oxoglutarate (Reaction W) as amino acceptors.
Plant GABA-T appears to use pyruvate preferentially, showing 1-25 times more activity with pyruvate than with 2-oxoglutarate in vitro (Dixon and Fowden, 1961; Streeter and
Thompson, 1972a; Wallace et al., 1984; Satya Narayan and Nair, 1986; Shelp et al.,
1995; Van Cauwenberghe, 1998); this conuasts with various bacterial and animal GABA-
T sources which only use 2-oxoglutarate (Scott and Jakoby, 1959; Schousboe et al.,
1973). Partially punfied GABA-T from tobacco possesses K,s of 1.5 rnM for GABA and
0.3 mM for pyruvate (Van Cauwenberghe, 1998). Both pynivate- and 2-oxoglutarate- dependent activities are present in crude extracts and unpurified mitochondnai preparations, and can be separated from each other by ion chromatography. GABA-T exhibits a broad pH optimum ranging frorn 8 to 10 (Streeter and Thompson, 1972a; Satya
Narayan and Nair, 1990; Shelp et al., 1995; Van Cauwenberghe, 1998). Plant GABA-T has not been punfied to homogeneity and no cDNAs encoding this protein have been isolated.
The last step of the GABA shunt, catalyzed by SSA dehydrogenase (SSADH, EC
1.2.1.16), irreversibly oxidizes SSA to succinate (Reaction [II). The partidly punfied plant enzyme has an alkaline pH optimum of approximately 9 (Streeter and Thompson,
1972a; Shelp et al., 1995) and is up to 20 times more active with NAD' than with
NADP+ (Streeter and Thompson, l972a; Yamaura et al., 1988, Satya Narayan and Nair,
1989; Shelp et al., 1995). The apparent K,s of SSADH for NAD' and SSA are 166-460 pM and 5- 15 FM, respectively (Yarnaura et al., 1988, Satya Narayan and Nair, 1990;
Walton, 1993). Interestingly, mamrnalian SSADH and GABA-T forrn a stable multi-
-13 - enzyme complex in vitro that may prornote substrate channeling (Hearl and Churchich,
1984). Like GABA-T, plant SSADH has not been purified to homogeneity and no
cDNAs encoding this protein have been isolated.
Although synthesis of GABA by GAD is thought to be the predominant pathway in
plants, GABA cm also be derived from the breakdown of putrescine, an important
polyamine (Flores et al., 1989). Diamine oxidase (EC 1.4.3.6) and 6aminobutyraldehyde
dehydrogenase (EC 1.2-1.19) cataboiize the conversion of putrescine to GABA via
pyrroline. The latter enzyme may be the betaine aldehyde dehydrogenase (BADH) that is
involved in glycine betaine synthesis (Trossat et al, 1997; see Section 2.2.3., Reaction
XIII). GABA and urea production via the hydrolysis of 4-guandinobutyric acid has aIso
been reported in plants; however, the contribution of this pathway and the putrescine
catabolic pathway in the overall production of GPLBA is believed to be rninor (Steward
and Durzan, 1965).
2.2.1.2. Regulation of GABA levels
Control of GABA accumulation is likely rnediated primarily via an activation of
GAD, the first cornmitted enzyme of GABA synthesis (Bown and Shelp, 1997). It is now well established that cab is involved in regulating GABA synthesis, thereby providing support for the early speculation by Wallace et al. (1984) that the induction of GABA synthesis is associated with stress treatments which influence Ca" fluxes. GAD possesses a CaM binding domain (Baum et al., 1993) and is activated in vitro by Ca-
CaM (Ling et al., 1994; Snedden et al., 1995). GAD activity increases by 276% in fava bean root (Ling et al., 1994). 250 to 800% in soybean leaf, seed coat, root and cotyledon
-14- (Snedden et al., 1995). and 212% in petunia petal (Arazi et al., 1995) in response to the addition of Ca2+-CaM. Ca2+-C~Malso increases the catalytic efficiency (V,,,JKm) of GAD by increasing the Vma and decreasing the Km(Snedden et al., 1995). At pH 7, partially punfied GAD from soybean seed coat has a Km of 2 1 mM and an V,, of 0.32 pmol min-' gALprotein in the absence of Ca2+-ca~.In the presence of Ca2'-CaM, the Km and V,, change to 9 rnM and 1.1 pmol min-' g-' protein. respectively, increasing the catalytic efficiency by 7-fold. Neither Ca2+nor CaM alone affects the enzyme activity of punfied recombinant petunia GAD at neutral pH (Snedden et al., 1996). Evidence for in vivo activation of GAD by Ca2+-CaMwas provided by inhibitor studies using Asparagus sprengeri Regel inesophyll cells (Cholewa et al., 1997). Abrupt cold shock (20°C to
1OC) stimulates GABA levels by 100% from 2.7 nmol GABA./1O6cells to 5.6 nmol
GABA/106 cells within 15 min. The addition of lanthanurn, a Ca" channel blocker, or
W7, a CaM antagonist reduces the cold-induced stimulation to only 28-32%. Use of a fluorescent Ca" indicator (Ruo-3) dernonstrated that cytosolic Ca" increases within 2 s of cold shock, followed by a return to initial levels within 25 S. Treatment with a Ca" ionophore, A23 187. increases GABA levels by 6 1% over control values of 5.13 nmol
GABA/~O~cells to 8.26 nmol GABA/~O~cells. Another study reported that ca2+-caM activates GAD and inhibits proteolysis in rice roots subjected to anoxia (Aurisano et al.,
1995b). Thus, Ca" is involved in the activation of CaM-dependent GAD activity and
GABA synthesis (Cholewa et al., 1997).
In addition to fine metabolic control of enzymatic GAD activity, molecular biological evidence suggests that plants possess multiple forms of GAD (Chen et al., 1994; Turano and Fang, 1998; Zik et al., 1998). For exarnple, Arabidopsis possesses two GAD
-15 - isoforms (GADI and GAD2) which are differentidy regulated. The GAD2 transcript,
encoded protein and specific activity are higher in leaves of Arnbidopsis supplied with 10
rnM NH,Cl, 5 mh.l NH,NO,, 5 mM glutamate or 5 mM glutamine as the sole nitrogen
source, than in leaves treated with 10 rnM KNO, (Turano and Fang, 1998). Moreover,
different peninia organs display different expression patterns of GAD mRNA and protein,
suggesting that GAD is transcnptionally and translationally regulated (Chen et al., 1994).
Although GAD is likely the major factor affecting GABA accumulation, decreased catabolism by GABA-T and SSADH (Bown and Shelp, 1989; Satya Narayan and Nair,
1986) or changes in intracellular or intercellular transport cannot be mled out. Activity ratios of GADGABA-T are in the order of 15-20: 1 in vitro (Streeter and Thompson,
1972b; Wallace et al., 1984; Satya Nair and Nair, 1986; Shelp et al., 1995) and GABA-T and GAD have very different pH optima in vitro (see Section 2.2.1.1.). Therefore, it has been suggested that GABA-T may restrict GABA metabolisrn in vivo, thereby contributing to its accumulation (Streeter and Thompson, 1972b; Bown and Shelp, 1989).
It is possible that during conditions that lirnit the recycling of reductant, such as anoxia and hypoxia (Raymond et al., 1987), low levels of NAD' may limit SSADH activity, which may in turn feedback inhibit GABA-T activity (Shelp et al., 1995). In vitro experiments using partially punfied tobacco GABA-T provide evidence that GABA-T is indeed inhibited by SSA (Van Cauwenberghe, 1998). Separation of GABA synthesis from GABA catabolism by subcellular compartmentation is another potential mechanism for regulating GABA levels and is discussed below (Section 2.2.1.3.). No data on the gene expression of either GABA-T or SSADH are available. 2.2.1.3. Subceliular Localization of GABA Shunt
An important feature of arnino acid metabolism is cornpartmentation, especially
within membrane-bound organelles (Ballantyne and Chamberlin, 1994). There is
considerable evidence suggesting that GAD is exclusively cytosolic (Dixon and Fowden,
1961: Wallace et al., 1984; Satya Naraya:: and Nair, 1986). In contrast, the subcellular
locations of GABA-T and SSADH are uncertain (Bown and Shelp, 1989). Previous
studies suggested that GABA-T in callus cells hmsoybean cotyledon is mitochondrial
(Tokunaga et al., 1976), whereas in pea seedling (Dixon and Fowden, 1961), soybean
leaves (Wallace et al., 1984) and potato tuber (Satya Narayan and Nair, 1986) it is both
cytosolic and mitochondrial. Satya Narayan and Nair (1986) suggested that both
GAI3A:pyruvate-T and SSADH share a dual mitochondrial and cytosolic location in a 1:3
ratio, respectively. In these studies, unpurified organelle fractions obtained by differential
centrifugation of homogenized tissues were used, with no assessment of contamination or
organelle breakage. It is possible, therefore, that the appearance of GABA-T and SSADH
activities in the soluble/cytosolic fractions is a consequence of mitochondrial breakage during the extraction and fractionation procedures. According to Douce et al. (1987), homogenization is detrimental to mitochondrial integrity and produces envelope-free rnitochondria and rnitochondria that reseal following rupture and loss of rnatrix content.
Currently, immunolocalization studies using specific antibodies are not possible since plant GABA-T and SSADH have not been purified to hornogeneity. In mamrnals, both
GABA-T and SSADH are confined to the mitochondria (Head and Churchich, 1984).
Since the accuracy of the intracellular localization of enzyme activity relies on pure and intact organelle fractions, many of the problerns associated with previous studies
-17- (such as organelle breakage and contamination) cari be circumvented by isolating organelles from gently lysed protoplasts rather than the conventional homogenization of tissue (Nishimura et al., 1982; Douce et al., 1987). Separation of organelles from protoplast lysate by sucrose or silica sol (Percoll) density gradient centrifugation, rather than differential centrifugation, provides an effective mechod to isolate pure and intact plant organelles (Jackson et al., 1979; Moreau and Romani, 1982).
2.2.1.4. Roles of GABA
2.2.1.4.1. Biochemical pH Stat
GABA accumulates in response to hypoxia and anoxia (Streeter and Thompson,
1972a; Aunsano et al., 199%; Ratcliffe, 1995) and mechanicai darnage (Wallace et al.,
1984). GABA levels increased 1.3-fold from 1.4 to 3.4 mM in rice roots exposed to 3 h of anaerobic treatment (Aurisano et al., 1996). Hypoxia decreases cytosolic pH values by
0.3 to 0.6 pH units (Roberts et al., 1984), and increases the accumulation of organic acids
(e.g. pymvate and lactate) (Demis et al., 1997). Mechanical darnage may also cause cytosolic acidification by rupturing acidic vacuoles (Bown and Shelp, 1989). Regardless of the nature of the perturbations, a decrease in cytosolic pH from the equilibrium set- point, stimulates the activity of GAD which has a sharp pH optimum of 5.8 (Snedden et al., 1992, 1995; Shelp et al., 1995). Since the conversion of glutamate to GABA consumes a H+ (see Reaction 1, Section 2.2.1.1 .) this reaction may help to ameliorate cytosolic acidosis as part of a metabolic pH-stat mechanism (Crawford et al., 1994;
Ratcliffe, 1995). Crawford et al. (1994) used a fluorescent pH indicator (BCECF) to monitor the cytosolic pH of A. sprengeri rnesophyll suspension cells in response to weak
-18- acid loads. Within several seconds of treatment with butyric acid, a decline of cytosolic pH of 0.6 units was observed; after 15 s, GABA levels increase by approximately 250% to 6 nrnol GABA/1O6 cells. The H+-consumingGABA production can account for approximately 50% of the imposed acid load after 45 s of weak acid treatment (Crawford et al., 1994; Bown and Shelp, 1997). Ford et al. (1996), using in vivo "P- and lS~-NMR spectroscopy, found that the cytosolic pH of carrot root cultures declines by 0.2-0.3 units after 6-12 hours of hypoxic conditions. When such roots are exposed to 15~-ammonium, the production of 15N-glutarnineand LS~-glutamateis reduced, together with an enhanced accumulation of 1S~-GA13~and '*N-alanine compared to control values. Similar observations with carrot cells (Carroll et al., 1994) and maize roots (Roberts et al., 1992) showed that GABA and alanine increased during cytosolic acidifying oxygen stress, together with a decrease in glutamine (Carroll et al., 1994) or glutamine and glutamate
(Roberts et al., 1992). Taken together, these results show that rapid or gradua1 (Le. seconds to hours) acidification of the cytosol, stimulates GABA synthesis and this reaction can offset the decrease in pH, consistent with a role for glutamate decarboxylation as a metabolic pH-stat mechanism.
2.2.1.4.2. Plant Defense
GABA is a potent inhibitory neurotransrnitter in both vertebrates and invertebrates, and stimulates CI' influx through GABA-gated Cl- channels, thereby causing a hyperpolarization of neurons (Kanner, 1994). This reduces the generation and propagation of action potentials. Invertebrates, unlike vertebrates, lack a 'blood brain barrier' and neuromuscular activity may be directly affected by ingested GABA levels
-19- (Bown and Shelp, 1997). Rarnputh and Bown (1996) investigated the hypothesis that
mechanical stimulation or damage resulting from the activities of phytophagous insects
stimulates GABA accumulation, which in turn inhibits insect growth and development.
Mechanical damage increases soybean leaf GABA levels 10- to 25-fold within 1 to 4 min
to values of approximately 2 pmol GABA g-' FW. The increase in GABA levels are
likely attributable to increased cytosolic Ca2+and H+ levels, thereby stimulating GAD
activity (See Section 2.2.1.2.). When this level of GABA is introduced into a synthetic
diet, the growth, developmental and survival rates of cultured oblique-banded leaf roller
larvae, an important fruit crop pest, are reduced. The larvae prefer developing leaf tissue,
which produce lower GABA levels when damaped, than mature leaves when damaged.
Mechanical stimulation and darnage, such as that produced by herbivory, rnay disrupt normal cellular organization and increase cytosolic Ca" and H' levek, thereby causing the rapid production and accumulation of GABA (Wallace et al., 1984; Ramputh and
Bown, 1996). Thus, GABA accumulation in response to wounding may play a role in plant defense against pests (Ramputh and Bown, 1996; Bown and Shelp, 1997).
2.2.1.4.3. Krebs Cycle Bypass
Glutamate is involved in a number of rnetabolic pathways, some of which are depicted in Fig. 1.l. Glutamate carbon may enter the Krebs cycle as succinate, produced by the GABA shunt, or as 2-oxoglutarate, produced either by deamination (glutamate dehydrogenase, GDH) or transamination reactions (glutamate:oxaloacetate transaminase,
GOT or g1utarnate:pyruvate transaminase, GR). Two-oxoglutarate is then converted to succinyl-CoA and succinate via 2-oxoglutarate dehydrogenase (ODH) and succinyl-CoA
-20- kinase (SK).
Satya Narayan and Nair (1986) suggested originally that carbon flow via the GABA shunt assumes significance when the activity of ODH is restncted. Metabolic studies using developing soybean cotyledons indicate that GABA shunt activity is not a response to low levels of enzymes involved in Zoxoglutarate metabolism or glutamate dearninatiodtransarnination (Shelp et al., 1995; Tuin and Shelp, 1996). The sarne physiologicd conditions that lunit ODH activity afso Limit SSADH activity. Both
SSADH and ODH require NAD', a cofactor that is Iirniting during hypoxia (Raymond et al., 1987). During such a stress, increasing levels of SSA can feedback inhibit GABA-T
(Van Cauwenberghe. 1998), and contribute increased GABA levels. Following removal of the stress, GABA nitrogen and carbon is converted to alanine or glutamate and succinate (Streeter and Thompson, 1972b). respectively. Subsequent succinate catabolism generates reducing equivalents (NADH) and ATP, cornponents essential for recovery from Suess and for repair of stress-induced damage (Hare and Cress, 1997).
According to Bown and Shelp (1997), the metabolism of glutamate to succinate via the GABA shunt (1 NADH) is energetically less favorable than via the Krebs cycle (1
NADH plus 1 ATP), and even less favorable than via GDH and the Krebs cycle (2
NADH plus 1 ATP) (Fig. 1.1). If the refixation of NH,' (1 ATP) generated by GDH is taken into consideration, glutamate metabolism via GDH results in a net production of one NADH in comparison to glutamate metabolism via the GABA shunt. Perhaps the metabolism of glutamate via the GABA shunt eliminates potential problems associated with NH,' production, such as MI,+toxicity (Vance, 1997) or NH, loss as a volatile gas
(Canvin and Salon, 1997). This may be relevant dunng darkness when the activities of
-21- chloroplastic glutamine synthetase (GS)and g1utamine:oxoglutarate amidotransferase
(GOGAT) are depressed (Canvin and Salon, 1997).
2.2.1.4.4. Nitrogen Storage and Transport Cornpound
GABA is a stable, neutrd compound that rnay serve as a nitrogen storage compound during plant stress (Bown and Shelp, 1997). Studies using isolates vacuoles showed that
50% of glutamate, GABA and alanine rnay be located inside the vacuole (Wagner, 1979;
Yarnaki, 1982), a result consistent with the cytosolic location of GAD (Bown and Shelp,
1997). Thus, GABA production and accumulation may occur in the cytosol (Bown and
Shelp, 1989). However. feeding studies with ['4C]glutamate suggest that the location of
GABA production and accumulation in A. sprengeri mesophyll cells are not identical, indicating that GABA is sequestered within organelles (Chung et al., 1992). GABA also undergoes passive and specific efflux fiom isolated cells into the incubation medium
(Secor and Schrader, 1985; Chung et al., 1992). Long-distance transport of GABA occurs prïmarily via the xylem (Bown and Shelp, 1989). GABA transport mechanisms are discussed in greater detail below (Section 2.3.2.).
2.2.1.4.5. Plant Development
GABA may regulate plant development (Baum et al., 1996; Bown and Shelp, 1997).
Transgenic tobacco plants overexpressing a mutant petunia GAD cDNA lacking the CaMr binding domain (GADAC plants) exhibit severe morphological abnorrnalities, such as shorter stems and more branching than wild-type plants and transgenic plants expressing the full-length GAD (GAD plants) (Baurn et ai., 1996). The GADAC plants have
-22- relatively higher levels of GABA and lower levels of glutamate than control plants; in stems of GADAC, GAD and wild-type plants, respectively, GABA is 45, 17 and 7 mol % of the total amino acids, and glutamate is 2, 15 and 32 mol % (Baum et al., 1996).
Presumably, in the GADAC plants, the intracellular Ca" pools are unaltered and growth inhibition may result fiom dtered GABA metabolism (Bown and Shelp, 1997). It is unclear whether the gross phenotypic changes exhibited by the GADdC mutants are the result of unregulated GABA synthesis siphoning glutamate carbon and nitrogen from proiein production or from pathways that synthesize products essential for growth and developrnent, such as growth hormones (Baum et al., 1996).
Another investigation reponed that exogenous GABA at a concentration of 100 mM stimulates ethylene production in excised sunflower tissues by 14-fold (Kathiresan et al.,
1997). Ethylene is a hormone that has been implicated in many developmental processes in plants ranging from seed germination to senescence of flowers and leaves (McCourt and Keith, 1997). GABA increases the levels of 1-aminocyclopropane-1-carboxylicacid
(ACC), the direct precursor of ethylene, as well as ACC synthase mRNA and protein,
ACC oxidase mRNA and ACC oxidase activity (Kathiresan et al., 1997). Another investigation found that the hormones kinetin and a-napthdene acetic acid, standard tissue culture chernicals used to induce de-differentiation of root tissue, reduce glutamate
IeveIs and elevate GABA levels in transforrned root cultures of Datura stramonium line
Dl56 (Ford et al., 1996). The possibility exists that GABA rnay be responsible for inducing de-differentiation of the tissue through lirnited alkaloid biosynthesis.
2.2.1.4.6. Compatible Osmolyte Compatible osmolytes refer to compounds that play important roles in counteracting ad preventing some of the deletenous effects of drought, salinity and freezing (Yoshiba et al., 1997). During water deficit conditions, perturbing solutes, such as inorganic ions, readily enter the hydration sphere of proteins and favor protein denaturation (Rhodes and
Hanson, 1993). In contrast, compatible osmolytes are prefereatially excluded fî-om the protein surface and its irnrnediate hydration sphere, thereby stabilizing folded protein structures. It has been suggested that bullcy compatible osrnolytes, unlike smaller water molecules, are unable to effectively cluster in close proximity to proteins (Yancey, 1994).
This possibility, in conjunction with the ability of compatible osmolytes to bind water molecules preferentially over other molecules, may permit proteins to possess a large hydration shell during conditions of low cellular water levels. In addition to these protein-stabilizing or osmoprotectant roles, compatible osrnolytes play a crucial role in cellular osmotic adjustment in response to osmotic stress (Rhodes and Hanson, 1993).
By accumulating within the cell, they reduce the water potential below that of the extemal environment and enable water to enter and remain within the ce11 (Yancey, 1994).
Compatible osmolytes in plants include quaternary ammonium compounds or QACs
(e.g. glycine betaine; see Section 2.2.3.), tertiary sulfoniurn compounds (e.g. 3- dirnethylsulfonioproprionate), polyols (e.g. glycerol) and certain arnino acids. Of amino acids, proline is the most ubiquitous compatible osmolyte in water-stressed plants (See
Section 2.2.2.). Proline, as well as glycine, alanine, 3-arninopropionic acid and taurine are classified as osmolytes in other organisrns (Yancey, 1994). Of particular interest is the classification of GABA, in addition to ectoine, pipercolate, N-acetyl-lycine and B- amino acids, as probable osmolytes in eubacteria and rnethanogenic archeabacteria.
-24- In plants, GABA may also function as a compatible osmolyte, like its proline and glycine betaine counterparts. Several studies showed that GABA levels increase in plants exposed to water deficit (Handa et al., 1983; Rhodes et al., 1986; Binzel et al., 1987;
Fougère et al., l99 1; Bolarin et al., 1995, Serraj et al., 1998). For example, in osmotically adapted suspension cells, GABA concentrations between 6 to 39 mM were documented, a
5- to 15-fold increase over non-adapted control cells (Handa et al., 1983; Rhodes et al..
1986; Binzel et al., 1987). Chernicaiiy, GABA has no net charge at neutral pH (pKs4.03 and 10.56; Anonymous, 1972) and is highly soluble in water (Yancey, 1994). Since
GABA accumulates in plants and is considered a stab1.e nitmgen storage compound
(Bown and Shelp, 1997; see Section 2.2.1.4.4.). it likely contributes no toxic effects to the cell. In fact, in vitra experiments showed that GABA, at concentrations between 25 to
200 mM, exhibits cryoprotective properties, exceeding those of proline, by stabilizing and protecting isolated thylakoids against freezing darnage in the presence of salt (Heber et al., 197 1). Evidence that GABA possesses hydroxyl radical scavenging activity. exceeding that of proline and glycine betaine at the sarne concentrations (16 mM), has also been demonstrated in vitro (Smimoff and Cumbes, 1989).
2.2.2. Proline
Proline is a cyclic protein imino acid that accumulates in virtually al1 plants in response to drought and salinity (Hare and Cress, 1997). For example, proline accumulates in leaf tissues and apical meristems of shoots and roots of plants experiencing water stress (Jones et al., 1980; Voetberg and Sharp, 199l), in desiccating pollen (Lansac et al.. 19!36), and in suspension-cultured plant cells adapted to water stress
-25- (Rhodes et al., 1986; Binzel et al., 1987) or salt stress (Thomas et al., 1992). In tobacco ce11 cultures adapted to 428 mM NaCI, proline accumulates to an average concentration of 129 rnM and represents more than 80% of the free amino acid pool as compared to an average of 0.29 mM and about 4% of the pool in unadapted cells (Binzel et al., 1987). In the apex of maize roots expenencing water deficit, proline represents a major solute, reaching cellular concentrations of 120 rnM (Voetberg and Sharp, 199 1). AIthough rnaize root tips are known to synthesize proline, it is still unclear whether increased deposition of proline in the apical region is a consequence of increased transport to the apex via the phloem, or de novo synthesis of proline at the apex. OsLier conditions, in addition to drought and salinity, induce proline accumulation; these include temperanire stress, pathogen infection, anoxia, nutrient deficiency, UV irradiation and heavy metal toxicity (Hare and Cress, 1997).
2.2.2.1. Proline Metabolism
Proline is produced from glutamate via two successive reductions. The bifunctional enzyme A'-pyrroline-5-carboxylatesynthetase (PSCS. EC 2.7.2.1 1EC 1.2.1.4 1) catalyzes the phosphorylation of glutamate to glutarnic y-phosphate (GP; Reaction IVA) and the subsequent reduction of GP to glutamic y-semialdehyde (GSA; Reaction NB) @elauney and Venna, 1993).
Reaction NA: Glutamate + ATP - glutarnyl y-phosphate + ADP
Reaction IVB: Glutarnyl y-phosphate + NADPH + H+- glutamic y-semiaidehyde
+ NADP + Pi
-2 6- Reaction V: Glutamic y-sernialdehyde ==AL-pyrroline-5-carboxylate
Reaction VI: A'-pyrroline-5-carboxylate + NAD(P)H + H+ - proline + NAD(P)+
PSCS, the rate-limiting enzyme for proline synthesis, was cloned from moth bean and expressed in tobacco (Kavi Kishor et al., 1995). The transgenic plants produce 10- to 18- fold more proline than control plants, arnounting to about 56.4 pmol proline g-l FW
(approximately 56 mM) under drought-stress conditions. GSA undergoes spontaneous conversion to Al-pyrroline-5-carboxylate(PX; Reaction V), which is then reduced to proline by P5C reductase (PSCR, EC 1.5.1.2, Reaction VI) (Delauney and Verma, 1993).
A cDNA encoding PSCR was cloned by direct cornplementation of an E. coli proline auxotroph with a soybean nodule cDNA expression library (Delauney and Verma, 1990), thereby facilitating the isolation of PSCR homologues from Pisurn sarivum (Williamson and Slocum, 1992) and Arabidopsis (Verbruggen et al., 1993). Both PSCS and PSCR transcripts increase in response to osmotic stress (Verbruggen et al., 1993).
Proline synthesis and accumulation of storage root tissue of red beet (Bera vulgaris
L.) occurs primarily in the cytosol (Leigh et al., 1981). There is also the possibility that proline synthesis from ornithine occurs in mitochondria; however, little is known about this pathway and controvers y surrounds its relative importance (Hare and Cress, 1997).
Following the removal of stress, proline may be oxidized to PSC by proline dehydrogenase (PDH, EC 1.4.3, Reaction VU), with the resulting product being further oxidized to glutamate by PXdehydrogenase (PSCDH, EC 1.5.1.12, Reaction VIII).
Reaction MI: Proline + %O2+ FAD - AI- pyrroline-5-carboxy1ate + FADH2 +H,O
Reaction VIII: AL-pyrroline-5-carboxylate + NAD(P)' - glutamate + NAD(P)H +
H+
Both PDH and PSCDH are localized to the rnitochondna (Rayapati et al., 1989), suggesting that proline must be transported across the mitochondrial membranes. A cDNA encoding PDH has been cloned from Arabidopsis and shown to be strongly repressed in response to osmotic stress, over-nding the induction by proiine under non- stressed conditions (Kiyosue et al., 1996). The gene encoding PSCDH has not ken cloned from plants, but the protein has been extensively purified and characterized
(Forlani et al., 1997).
2.2.2.2. Regulation of Prolhe Levets
Proline accumulation induced by stress conditions is mediated by both increased synthesis and reduced oxidation of the arnino acid (Hare and Cress, 1997). Transcnpts encoding the proline biosynthetic enzymes, PSCS and PSCR. increase in abundance in response to osmotic stress (Verbruggen et al., 1993). The catabolism of proline is also strictiy regulated; the gene encoding PDH is induced by proline under non-stressed conditions, but is strongly repressed in response to osmotic stress (Kiyosue et ai., 1996).
Thus, the reciprocal regulation of PSCS and PDH may account for the rapid accumulation of proline in response to osmotic stress and its rapid catabolisrn upon stress relief (Peng et al., 1996).
In addition to metabolic events involved in proline accumulation, the stimulation of
-28- proline transport processes may also increase cellular proline levels. Using heterologous yeast complernentation of a mutant lacking SHR3, a protein specificdly required for correct targeting of plasma membrane amino acid permeases, two genes encoding proline transporters (Pro7') frorn Arabidopsis were identified (Rentsch et al.. 1996). Pron and
Proîï are expressed ubiquitously in al1 organs, with the highest levels of ProTl vanscripts beirig found in roots, stems and flowee. Expression of Pr072 is strongly induced with water or salt stress, irnplying that ProT2 plays an important role in nitrogen distribution dunng water stress. Both ProTl and ProT2 exhibit a relatively high affinity for proline (K,,,= 360 I15 PM) and do not transport other protein amino acids; GABA as a substrate was never tested.
2.2.23. Roles of Probe
It is generally accepted that proline is a compatible osmolyte by virtue of its chernical and physiologicai properties (Hue and Cress, 1997). Chernically. proline possesses no net charge at neutral pH (pK 1.95 and 10.64; Anonymous, 1972) and is highly soluble in water (Yancey, 1994). Under conditions of water deprivation or extreme salinity, proline accumulates in the cytosoI (Leigh et al., 1981) and reduces the cellular water potential below the extemal water potential, enabling water to move into and rernain in the ce11
(Rhodes and Hanson, 1993). The ability of proline to accumulate to non-toxic levels, imposing little or no perturbing effects on macromolecule-solvent interactions, is another important attnbute (Yancey, 1994).
In addition to its role as a compatible osmolyte in higher plants, proIine synthesis may alleviate cytosolic acidosis by consuming H' (See Section 2.2.2.1., Reactions IVB and
-29- VI), and may maintain a favorable NADP/NADPH ratio (Hare and Cress, 1997).
Altematively, proline rnay function as a stable, non-toxic nitrogen storage molecule.
Rapid catabolism of proline upon stress relief may provide reducing equivalents that
support rnitochondriai oxidative phosphorylation and the generation of ATP (Hare and
Cress, 1997). Proline is also a substrate for protein synthesis or is converted to other
amino acids via glutamate. ProIine rnay also be modified to proline betaine (stachydrine)
(Rhodes and Hanson, 1993) and 4-hydroxyproline betaine (betonicine) (Hanson et al.,
1994), quaternary ammonium compounds (QACs) which function as compatible
osmolytes (See Sections 2.2.1.4.6. and 2.2.3.).
2.2.3. Glycine Betaine
Glycine betaine (N, N, N,-trimethyl glycine) is an important osmolyte that
accumuIates in many organisms, including bactena, cyanobacteria. algae, animals and
higher plants in response to sait and drought stress (Rhodes and Hanson, 1993; Yancey,
1994). In certain plant genera (e-g. Spinacia, Zea, Sorghurn, Tnifolirrm),species may
accumulate glycine betaine to 4-40 prnol g" DW under water deficit conditions compared
to OS-10 pmol g-l DW under non-stressed conditions; non-accumulating genera (e.g.
Oryza, Nicoriann, Lycopersicon) generally contain CO. 1 pmol glycine betaine g-' DW (for review see Rhodes and Hanson, 1993).
Glycine betaine is characterized as a QAC due to its fully-methyl-substituted N atom that creates a permanent positive charge on the N-moiety. This property renders glycine betaine and other plant QACs (e.g. proline betaine, hydroxy-proline betaine, fbaianine betaine, choline-O-sulfate, trigonelline) highly soluble in water with no net charge at
-30- neutral pH (Yancey, 1994). Stmcturally, glycine betaine is similar to probe (See Section
Fig. 2.1 ; Holmberg and BUlow, 1998).
The ability of some plant genera to accumulate QACs in cornparison to non- accumulators, is likely due to differences in biosynthetic capabilities. In general, glycine betaine and other QACs are derived fiorn amino acid precursors in a series of methylation reactions catalyzed by S-adenosylmethionine (SAM)-dependent N-methyltransferases
(Rhodes and Hanson, 1993). Unlike the synthesis of other amino acid betaines which are derived directly kom the N-methylation of the parent amino acid, glycine betaine is synthesized from senne-denved choline. The biosynthetic route for choline synthesis is highly cornplex involving one of three parallel and interconnected N-methylation reaction pathways. Following the synthesis of choline, glycine betaine is synthesized by two oxidation steps catalyzed by ferredoxin (Fd)-dependent choline monooxygenase (CMO,
Reaction XII) and betaine aldehyde dehydrogenase (BADH, EC 1.2.1.8, Reaction Xm).
Reaction XII: Choline + O2+ 2Fdd -O betaine aldehyde + 2Fd,,
Reaction XIII: Betaine aldehyde + NAD' - glycine betaine + NADH + H'
Both reactions are predorninantly localized in the chloroplast, although in spinach leaves approximately 10% of the BADH may exist as a cytosolic isoform (Weretilnyk and
Hanson, 1988). BADH is reportedly specific for betaine aldehyde; however, recent research shows that it acts on other aldehyde substrates including Caminobutyraldehyde
(Trossat et al., 1997). CM0 and BADH cDNAs from spinach have been cloned
(Weretilnyk and Hanson, 1990; Rathinasabapathi et al., 1997), and transgenic tobacco
-31- plants constinitively expressing the spinach CM0 in chloroplasts have been created
(Nuccio et al., 1998). These plants do not overproduce glycine betaine, presumably due to inadequate choline synthesis and transport of choline into the chloroplast. Glycine betaine is localized predorninantly in the chloroplasts of salinized spinach leaves where it provides osmotic adjustrnent (Robinson and Jones, 1986). The structures of choline, betaine ddehyde and glycine betaine are shown in Fig. 2.1 .B; several other cornmercially rivailable betaines which include carnitine, trigoneHine, and betonicine are shown as well
(Fig. 2.1 .C).Unlike mamrnals which catabolize glycine betaine, the fate of this compound in plants remains unknown (Rhodes, 1987).
2.3. Amino Acid Transporters
2.3.1. Amino Acid Transporters in Plants
2.3.1.1. Physiological and Biochemical Descriptions
Arnino acids are actively transported across the plasma membrane of higher plant cells by proton-coupled symports that utilize the electrochemical gradient for H+ to drive the uphill transport of arnino acids (Li and Bush, 1991, 1992; Bush, 1993). Evidence for arnino acid transport across the plasma membrane of higher plants was initially provided by radiotracer studies with intact tissues and suspension cells (Harrington and Henke,
1981; Kinraide and Etherton, 1980; McCutcheon et al., 1988). Bush and Langston-
Unkefer (1988), using isolated plasma membrane vesicles from zucchini, provided the first in vitro evidence for H+-aminoacid symport activity. Subsequent uptake studies in plasma membrane vesicles from sugar beet leaves revealed the activity of at Ieast four different H+-coupled amino acid transporters specific for acidic, basic and neutral amino
-32- acids (Li and Bush. 1990). Alanine, leucine, glutamine, glutamate, isoleucine and arginine are accumulated in a pH-dependent manner against a concentration gradient and exhibit saturable transport kinetics. Alanine transport is electrogenic as demonstrated by the effect of membrane potential on ApH-dependent influx. Li and Bush (1992) subsequently used several amino acid analogues to identiw structural features that are important in molecular recognition by the two neutral transport systems. The data suggested that the positional relationship between the a-amino group and carboxyl group is an important parameter in substrate recognition.
2.3.1.2. Molecular Approaches
The recent development of new molecular tools, such as functional expression of plant genes in Xenopiu oocytes and in yeast cells, circumvent many of the difficulties in identifying, isolating and characterizing transport proteins using biochemical approaches
(Fromrner and Ninnemann, 1995). Using various yeast amino acid transport mutants lacking uptake systems for certain amino acids, more than 25 amino acid transport genes from eight different plant species have been identified; Arabidopsis alone possesses more than 15 different genes (Table 2.1; Fischer et ai., 1998: Rentsch et al., 1998). Based on sequence similarities, the amino acid transporters from Arabidopsis have been classified into the ATF (amino acid transporter farnily) and the APC (arnino acid-polyamine- choline transporter) superfamilies (Fischer et al., 1998).
The ATF superfamily is the largest and best characterized group, and is compnsed of four families: the AAPs (amino acid permease), the ProTs (proline transporter), the LHT- related (lysinehistidine transporter) and the AUX1 -related (auxin transporter) proteins.
-33- TABLE 2.1. a Amino acid transportersfrom Arabidopsis thaliana
b Superfamily Gene Function of encoded protein Accession no.
--
ATF AAPl Low-affinity transporter of neutral and acidic arnino acids Low-affinity transporter of neutral and acidic amino acids Low-affinity general amino acid transporter Low-affinity transporter of neutrai and acidic amino acids Low-affinity general amino acid transporters High-affiity transporter of neutral and acidic amino acids AAP7 Putative amino acid transporter LHTI Lysinehistidine transporter LHT2 Putative amino acid transporter LHT3 Putative amino acid transporter Pro TI Proline transporter Pro T2 Proline transporter A WXI Putative auxin transporter A UX2 Function unknown A WX3 Function unknown
MC CATI High-affinity transporter of basic arnino acids Putative amino acid transporter a Modified fkom Fischer et al. (1998). b ATF, amino acid transporter family; APC, amino acid-polyamine-choline transporter. The AAPs represent the Iargest family of the ATF superfarnily with at least 7 members in
Arabidopsis (Table 2.1, Fischer et al., 1998). The AAPs recognize a large spectrum of protein amino acids, and cm be grouped into two subfarnilies based on their differential selectivity toward basic amino acids (Fischer et al., 1995; Fischer et al., 1998). AAPI,
AAP2, AAP4 and AAP6 from Arabidopsis are considered neutral and acidic carriers because of their preferences for these arnino acids. AAP3 and AAPS are general arnino acid permeases due to their abilities to transport neutral and acidic amino acids, in addition to basic arnino acids (Fischer et al., 1995; Rentsch et al., 1996; Fischer et al.,
1998). Amino acid transport rnediated by al1 the AAPs increases with decreasiiig pH, occurs against a concentration gradient, and is sensitive to an uncoupler of mitochondnal oxidative phosphorylation known as 2, Cdinitrophenol, suggesting active transport via a
H+-symport mechanism (Fischer et al., 1995; Rentsch et al., 1996). Recent electrophysiological studies revealed that amino acid transport by AAPS expressed in oocytes, has a H+:arnino acid stoichiometry of 1: 1 regardless of the net charge on the amino acid (Boorer and Fischer, 1997).
Although mernbers of the AAP family share overlapping substrate specificities, they show differential expression patterns (Fischer et al., 1995; Rentsch et al., 1996; Fischer et al., 1998: Hirner et al., 1998). For example. the expression of AAPI is specific to the endospems and embryos of developing seeds (Hirner et al., 1998), whereas AAP5 is expressed throughout most of the plant, dong with AAP2, AAP3 and AAP6 in the root,
AAP2 and AAP4 in the source leaf and stem, AAPI and AAP2 in the fruit, and AAP4 in the flower (Fischer et al., 1995; Fischer et al., 1998). Many AAP homologues have since been identified from other plants species including potato, tobacco, rice, castor bean and
-35- rape (Rentsch et al., 1998).
The ProTs comprise the second arnino acid transport family which exclusively
transport proline in a H+-dependent manner over al1 other protein amino acids (Fischer et
al., 1998; Rentsch et al., 1998). ProTl and Pro72 from Arabidopsis are the first proline
transporters cloned from plants and they have K,s of 360 2 15 FM (Rentsch et al., 1996).
PoTI is expressed in dl organs, but the highest levels are found in roots, stems and
flowers. Expression in flowers is highest in the florai stalk phloem that enters the carpels
and is down-regulated after fertilization, indicating a specific role in supplying the ovules
with proline. ProT2 transcnpts are found ubiquitousiy thraughout the pIant, but expression is strongly induced under water or salt stress. The role of ProTs in proline accumulation during water stress is discussed in Section 2.2.2.2. The LHT-related family includes only one functionally characterized transporter (LHTI)that prefers lysine and histidine (Chen and Bush, 1997), and two putative transporters (LHRand LHT3)
(Fischer et d.,1998). A member of the fourth fmily, the AUX 1-related proteins, is defective in an auxin-resistant mutant of Arabidopsis (Bennett et al., 1996); the functions of two related genes (A (BIR2 and A (IXR3) remain unknown (Fischer et al., 1998).
Based on computer-based hydropathy analyses of the aligned AAP family members. the predicted structure for these integral proteins is believed to have 10 membrane- spanning domains (Fischer et al., 1998; Rentsch et al., 1998). According to this rnodel, the characterized AAPs contain a number of putative phosphorylation sites on the cytosolic surface of the membrane and no N-glycosylation sites on the extemal side.
However, preliminary experimental evidence indicates that AAPs have 11 membrane- spanning domains, with the N-terminus exposed to the cytosol (Chang and Bush, 1997;
-36- Fischer et al., 1998; Rentsch et ai., 1998).
Only two transport genes comprking the APC superfarnily of plant amino acid
transporters have been identified. AtCATl (originally called AtAATl), the first member
of the APC superfarnily, was identified by complementation of a yeast histidine transport
mutant (Frornrner et al., 1995). This transporter shares the most homology with the
marnmalian CAT (cation amino acid transporter) family (Frommer et al., 1995; Fischer et
al.. 1998; Rentsch et al., 1998). Another member (AtCA72) has been identified by the
Arabidopsis sequencing project, but its function remains unknown [Fischer et al., 1998).
Like their mammalian counterparts, AtCATl and AtCAT2 possess 14 predicted
membrane-spanning domains (Kim et al., 1991; Fischer et al.. 1998; Rentsch et al., 1998).
2.3.2. GABA Transporters in Eukaryotes and Prokaryotes
2.3.2.1. Animal GABA Transporters
Molecular cloning revealed the existence of two distinct families of GABA
transporters, termed GAT (GABA transporter) and VGAT (vesicular GABA transporter)
(Table 1.1). The first and largest family contains four distinct members, termed GAT1,
GAT2, GAT3 and BGTl (betaine-GABA transporter) that have affïnitiesfor GAB A in
the micromoIar range (see refs. in Table 1.1), but differ in substrate specificities for compounds such as 3-aminopropionic acid that are stnicturally related to GABA
(Christensen et al., 1994). Stoichiometric studies with rnembers of the GAT family
revealed that every GABA molecule is cotransported with 2 Na' and 1 Cl-(Liron et al.,
1988). The structure of mammalian transporters of the GAT fdlyhas 12 membrane-
spanning domains (Liu et al., 1993).
-37- Members belonging to the VGAT famiiy function in the transport of GABA into
synaptic vesicles. The first cDNAs encoding these transporters were cloned from
Caenorhabditis elegans (UNC-47) and Rattris norvegicus (rVGAT) (McIntire et al.,
1997). Their open reading frarnes encode hydrophobic proteins, each with 12 potential
transmembrane helices and a large hydrophilic loop between helices 1and II. Kinetic studies with rVG44Tindicated that this transporter has a moderate af15nity for GABA of approximately 5 mM (see Table 1.1 ; McIntire et al., 1997). Members of the VGAT family have sequence homology to some plant amino acid transporters belonging to the
ATF superfamily. Like the plant amino acid transporters, members of the VGAT family dso rely on a H+-electrochemicalgradient as the driving force for the transport of their substrates, however, GABA transport occurs via a H'-antiport mechanism, unlike the H'- syrnport mechanism in plants (McIntire et al., 1997).
2.3.2.2. Fungal GABA Transporters
The yeast S. cerevisiae can utilize GABA as its sole source of nitrogen. Three distinct permeases capable of transporting GABA have been identified: GAPl (general amino acid permease), PUT4 (proline permease) and UGA4 (specific GABA permease) (Table
1.1, Grenson et al., 1987). Although no Kmfor GABA has been reported for GAP 1,
PUT4 and UGA4 possess high affinities for GABA with K,s of 50 pM and 110 PM, respectively. These permeases, together with al1 other yeast arnino acid penneases, are members of the APC (amino acid-polyamine-choline transporter) superfarnily (Fischer et al., 1998). This large superfamily may be subdivided into the GAP1-related transporter family and the GABA-permease-related transporter family. GAPl and PUT4 belong to
-38- the former farnily which includes nine characterized permeases and seven uncharacterized
permeases identified by the genome sequencing project. The second family includes
UGA4, a choIine transporter (CTRI). two methionine transporters (MUPI and MUP3)
and several other members.
Regdation of amino acid transport in yeast occurs maidy by transcriptional control of
dnoacid permease genes, and by targeting and turnover of the proteins (Vissers et ai.,
1989, 1990; Grenson, 1992). Expression of several permease genes, such as GAPI, are
subject to nitrogen repression and are either down-regulated in the presence of preferred
nitrogen sources (NH4,',glutamine or asparagine) or up-regulated in the presence of
nitrogen sources that support oniy lirnited growth rates (urea or proline) (see Grenson,
1992 for review). Expression of the UGA4 pemease gene, in addition to UGAI and
UGA2 (encoding GABA-T and SSADH, respectively), is induced by GABA (Vissers et
al., 1990, Talibi et al.. 1995). This induction requires two positive control factors
encoded by the UGA3 and (/GA35 genes, respectively. A negative regulatory factor,
encoded by the UGA43 gene, has also been identified (Vissers et al., 1989, 1990). It
regdates the expression of UGA4, but has no effect on the expression of GABA-T or
SSADH (Vissers et al., 1990, Talibi et al., 1995). Both UGA3 and UGA35 contain a
Cys,-Zn, binuclear zinc finger that serves as a DNA-binding domain (Talibi et al., 1995).
2.3.2.3. Bacterial GABA Transporters
Escherichia coli and Bacillus srtbtilzts each possess a GABA perrnease called GabP.
These permeases exhibit considerable selectivity for GABA, with Michaelis constants
ranging from 12-37 pM (Niegemann et al., 1993; Ferson et ai., 1996). The sequences of
-39- these permeases are highly homologous to other transporter genes, including the aromatic amino acid carriers (AroP) from E. coli and Corynebacterium glutarnicum, the proline
(PUT#),arginine (CANI) and histidine (HPl)carriers from S. cerevisiae, and the proline-specific penneases (ProY) from Salmonella iyphimrtncrn and Aspergillus nidrrlans. The E. coli protein is 466 residues in size and likely possesses 12 putative membrane spanning domains (King et al., 1995). Recently, Hu and King (1998) identified an arnphipathic domain in the coding sequence of this gene (codons 339-365) that is highly conserved in many fungai and animal GABA transporters belonging to the
APC and GAT farnilies. This region may be involved in pore formation, shielding the highiy charged GABA moiety from the lipid environment.
2.3.2.4. Evidence for GABA Transporters in Plants
Many studies have compared the xylem and phloem contents from a variety of plant species and indicated that GABA is always present in xylem, but is sometimes absent from phloem (see Bown and Shelp, 1989 for review). In unstressed plants, GKBA levels in xylem sap are typically 0.1 mM (Shelp et al., 1987; Serraj et al., 1998) and are increased (230%)during drought (Serra. et al., 1998). In contrast to other neutral amino acids such as asparagine, glutamine, valine and senne, GABA apparently does not undergo direct xylem-to-phloem transfer. Rather, it is extensively metabolized in cells of mature leaves, resulting in indirect xylem-to-phloem transfer of irs carbon and nitrogen
(see Bown and Shelp, 1989 for review). The metabolism of 14C-GABA results in the recovery of 14C-radioactivityin the phloern as glutamine, senne, aspartate and glutamate.
Aithough cellular GABA efflux has been demonstrated with isolated soybean leaf cells
-40- (Secor and Schrader. 1985) and A. sprengen mesophyll cells (Chung et al.. 1992), the
mechanism of GABA retrieval from apoplastic compartments into cells of mature leaves
remains unknown. Like other amino acids, GABA uptake into a celi is likely mediated
by transport proteins. With the exception of AtAAPl (Frommer et al., 1993), none of the
plant amino acid transporters isolated thus far were tested to see whether GABA is a
suitable substrate. The ability of AtAAP 1 to recognize and transport a wide range of
amino acids, but not GABA, is Mcely due to ciifferences in substrate structure (See
Section 2.2.1 .). Although GABA transport across the mitochondrial (Satya Narayan and
Nair, 1990), organellar (Chung et al., 1992) and plasma membr=-..er (Secor and Schrader,
1985; Chung et al., 1992) has been suggested, proteins capable of transporting GABA
have not been identified.
2.4. Research Objectives
The accumulation of GABA in plants is probably mediated via an activation of GAD,
the first committed enzyme of GABA synthesis, as a result of increased cytosolic Ca-
CaM or H' Ievels. However, decreased GABA catabolism by GABA-T and SSADH,
either by metabolic regulation or separate compartmentation from GABA synthesis, may
aIso Iead to changes in GABA levels. Akhough research has extensively investigated the
subcel!ular localization of GABA shunt enzymes, the localization of GABA-T and
SSADH remains uncertain. Therefore, in the present study the hypothesis that GABA-T and SSADH are localized exclusively in the mitochondrion was tested. Improved organellar-fractionation techniques, including gentle protoplast disruption and continuous
Percoll-gradient centrifugation, were used to isolate organelles from developing soybean
-41- (G. max) cotyledons.
In addition to de novo synthesis or decreased catabolism, GABA accumulation may also be a result of intracellular or intercellular GABA transport. Experimental support for the first hypothesis, in conjunction with a survey of pertinent animal, fungal, and bacterial literature, prompted the formulation of a second hypothesis that plants possess proteins capable of transporthg GABA across membranes. This hypothesis was tested using heterologous complementation of a GABA-transport-deficient yeast mutant and analysis of previously isolated amino acid transporters from A. thaliana. CHAPTERTHREE
Subcellular Compartmentatioo of the CAminobutyrate Shunt in Protoplasts
from Developing Soybean Cotyledonsl
3.1. Introduction
GABA is widely distributed in all higher plants and represents a significant fraction of the soluble amino acid pool. Although GABA accumulation has been associated with a variety of environmental stress conditions, including hypoxia, tempenture shock and mechanical manipulation (Streeter and Thompson, 1972a. b; Wallace et al., 1984; Satya
Narayan and Nair, 1990; Shelp et al., 1993, little is known about the physiological role of GABA and the regulation of its metabolism (Bown and Shelp, 1989). The major route of GABA synthesis is by the direct and irreversible a-decarboxylation of glutamate by
GAD. Subsequent reversible transarnination of GABA with pyruvate (or 2-oxoglutarate) in a reaction catalyzed by GABA-T, results in the production of succinic semialdehyde and alanine (or glutamate). The former product can then be oxidized by SSADH to form succinate, an important Krebs cycle intermediate. The conversion of glutamate carbon to succinate by these three reactions is referred to as the GABA shunt.
To date, GAD is the best characterized enzyme in this pathway. It possesses a calmodulin-binding domain (Baum et al., 1993), and is activated in vitro by Ca2+-CaM
(Ling et al., 1994; Snedden et al., 1994). There is considerable evidence suggesting that
GAD is exclusively cytosolic (Dixon and Fowden, 1961 ; Wallace et al., 1984; Satya
'Breitkreuz KE, Shelp BI (1995) Plant Physiol 108: 99-103
-43- Narayan and Nair, 1986). In contrast to GAD, less is known about the enzyrnatic properties of GABA-T and SSADH, and their subcellular locations are controversial
(Bown and Shelp, 1989). Previous studies suggested that GABA-T in callus cells fiom soybean cotyledon is rnitochondrial (Tokunaga et al., 1976), whereas in pea seedling
(Dixon and Fowden, 196 1), soybean leaves (Wallace et al., 1984), and potato tuber (Satya
Narayan and Nair, 1986) it appears to be both cytosolic and rnitochondrial. Satya
Narayan and Nair (1986) suggested that both GABA:pynivate-T and SSADH share a dual mitochondnal and cytosolic location in a 1 to 3 ratio, respectively. In these studies, unpurified organelle fractions obtained by differential centrifugation of homogenized tissues were used, with no assessrnent of contamination or organelle breakage. It is possible, therefore, that the appearance of GABA-T and SSADH activities in the soluble/cytosolic fractions was a consequence of rnitochondnal breakage during the extraction and fractionation procedures. As a result, the subcellular distribution of enzymes in the GABA shunt, especially GABA-T and SSADH, are questionable.
This study tested the hypothesis that enzymes of GABA catabolism are located exclusively in the mitochondrion. By using an improved cellular-fractionation technique, the subcellular localizations of the GABA shunt enzymes were determined in organelles that were derived from developing soybean cotyledonary protoplasts and purified by continuous Percoll-gradient centrifugation.
3.2. Materials and Methods
3.2.1. Plant Material
Soybean plants (Glycine max (L.) Merrill cv. Maple Arrow) were grown in a naturally
-44- lighted greenhouse in 9-1 pots (5-6 plants per pot) containing Pro-mix BX (Les Tourbières
Premier Lteé, Rivière du Loup, Quebec, Canada). Natural lighting was supplemented
with high-intensity sodium vapor Iamps yielding a 16-h photopenod and a photosynthetic
photon flux density of 60 pmol m-' s-l at pot level. Each pot was supplied twice weekly
with one quarter-strength Hoagland-type nutrient solution (Hoagland and Amon, 1938)
containing 8 mM N, and alternatively with tap water as required.
Fruits were harvested approximately 20 d after anthesis and cotyledons with a length
of 8- 10 mm were collected and used immediately for protoplast isolation-
3.2.2. ProtopIast Isolation
The cotyledons were sliced into cl mm sections and suspended in 20 ml of holding
solution consisting of 0.5 M mannitol, 0.1% BSA (w/v) and 20 rnM Mes (pH 6.0). Mer
30 min, the holding solution was decanted and replaced with 20 ml of 0.45-pm-filtered digestion solution which contained 1.0% (w/v) Cellulase Onozuka R-10,0.3% (w/v)
Pectolyase Y-23 (Kanematsu USA Inc., Los Angeles), 0.3% (w/v) Drieselase (Sigma
Chemicals), 0.5 M mannitol. 0.1% (w/v) BSA and 20 mM Mes (pH 5.8). The tissue pieces were incubated in the dark in a non-shaking water bath set at 30°C. After 3 h, the enzyme solution was decanted and replaced with 20 ml of holding solution. Protoplasts released by gentle swirling were sequentially filtered through two nylon meshes (100 and
53 pn) and gently pelleted by centrifugation at 50 g for 30 min.
3.2.3. SubceUular Fractionation
The protoplast pellet was resuspended in two volumes of 0.2 M sucrose and passed
-45- through a 15-pm pore-size nylon mesh covering the aperture of a 3-mi syringe. The resulting lysate (-12 ml) was initially centrifuged at 50 g to settle, but not rupture. large starch-containing amyloplasts. After 20 min, the centrifuga1 force was increased linearly to 2.500 g over 6 min to settle smaller amyloplasts and starch panicles. The resultant supernatant (-9 mi), which was enriched in mitochondria and microbodies relative to amyloplasts, was placed ont0 a 2-28% (v/v) linear Percoll (Pharmacia) gradient (26 ml) containing 0.25 M sucrose, 0.1% BSA and 30 mM Mops (pH 7.0). Centrifugation was at
40,000 g for 1 hr in a Beckman SW28 rotor. Fractions were collected from the top of the gradient using an automated fraction collector.
3.2.4. Enzyme Assays
Al1 enzymes were assayed at 30°C in a final 1-ml volume containing ü.û2% (v/v)
Triton X- 100. Alcohol dehydrogenase (ADH, EC 1.1.1.1) was measured according to
Kirnmerer (l987), isocitrate dehydrogenase (IDH, EC L. 1.1.4 1) according to Bergman et al. (1980), catalase (Cat, EC 1.1 1.1.6) according to Aebi (1984) and cytochrome c oxidase (Cyt c ox; EC 1.9.3.1) according to Schnarrenberger st al. (1971). Integrity of the outer mitochondrial membrane was estimated using the Cyt c impermeabilty assay
(Neuburger, 1985). GAD activity was determined as the L-[l-'4C]glutarnate-dependent production of ''CO2 essentially as described by Snedden et al. (1992). GABA-T activity was determined as the GABA-dependent production of alanine (or glutamate), with pyruvate (or 2-oxoglutarate) as an amino acceptor (Shelp et aL.1995). The assay mixture contained 100 mM Tris-HCl (pH 8.5), 2 mM GABA, 20 pMPLP, 400 pl of sarnple and either 10 mM pyruvate or 10 rnM 2-oxoglutarate. Fractions 1 to 7. which represented
-46- cytosol, were pooled together and desalted using a Sephadex G-25 colurnn (Pharrnacia)
equilibrated with LOO rnM Tris buffer, 40 pM PLP and 14 rnM 2-ME. This step
decreased the production of alanine by g1utarnate:pyruvate-transaminase. The reaction
mixture was incubated for 3 h, then sulfosalicyclic acid was added to a final concentration
of 60 m.and the precipitate removed by centrifugation. The supernatant was
neutralized with 1 N NaOH, then diluted with HPLC-grade H20 for amino acid analysis.
For each fraction, a control reaction not containing GABA was used. SSADH activity
was measured as the difference in NADH production before and after the addition of SSA
(Shelp et al., 1995). The final assay mixture contained 100 rnM Arnpso buffer (pH 9.3,
0.5 rnM SSA, 0.5 rnM NAD', 14 mM 2-ME and 200 pl of sarnple.
The density of the Percoll gradient was determined using density marker beads
(Sigma Chernicals) calibrated for 0.25 M sucrose. Starch in ethanol-insoluble residues
was quantified as described by Taylor et al. (1988). Protein was detemùned according to
the method of Bradford (1976), with BSA as the standard.
3.2.5. Radiolabeling Study
L-[U-14C]G~B~(209 Ci mol") was obtained from Amersham. The shipping medium
was evaporated under a gentle stream of ultrapure N, and the arnino acid was resuspended
in HPLC-grade water. To remove '4C-succinate and other contaminating organic acids from m-lJC]G~BA,the preparation was passed through a column containing AGSOW-
X8 (H') resin (BioRad Laboratories). Organic acids were washed from the column with
HPLC-grade water and GABA eluted with 4 N NH40H (Atkins and Canvin, 1971).
Subsequently, the washed GABA was dned under a gentle stream of filtered air,
-47- resuspended in HPLC-grade water and the purity checked by HPLC as descnbed by Shelp
et al. (1995).
GABA metabolism by cytosolic (desalted pooled fractions 1 to 7), microbody
(fraction 19) and mitochondrial (fraction 25) fractions from the PercoIl gradient was determined as the incorporation of 14C-radioactivityfrom L-[U-'4C]G~B~into succinate.
A 1-ml assay mixture contained 100 rnM Tricine (pH 8.5), 3 rnM maionate, 1 mM CU-
'"]GABA (1 mCi), 0.5 m.NAD', 80 pM PLP, 0.02% Triton X-100 (vfv), 10 mM pyruvate and 200 pl of the cytosolic fraction or main microbody and mitochondrial fraction in the presence or absence of 1 mM AOA. The reaction mixtürz was incubated for 1 hr at 30°C in a shaking water bath and the reaction was terminated with 5 ml of boiling 95% ethanol. The samples were subsequently dried under vacuum, resuspended in 2 ml of HPLC-grade water and partitioned against an equal volume of chloroform.
After centrifugation at 2,500 g for 5 min, the aqueous fraction was rernoved and dned under a gentle Stream of filtered air and resuspended in 0.5 ml of HPLC-grade water.
Labeled organic acids were analyzed by HPLC according to Tuin and Shelp (1994).
3.3. Results
Protoplasts prepared kom soybean cotyledons were gently lysed and used as a starting material to separate various organelles. Table 3.1 shows the recovery and distribution of protein, selected organelle markers and GABA-shunt enzymes from lysed protoplasts pnor to and following centrifugation at 2,500 g. Eighty-two % of the starch (arnyloplast marker) was recovered in the pellet, whereas 73% of the catalase activity (a rnicrobody marker) and 55% of the Cyt c oxidase (a mitochondrial marker) were recovered in the
-48- TABLE 3.1.
Distribution of protein, organelle markers, and GABA-shunt enzymes in Zysed pmtoplasts, and supernatant and pellet fractions resultingfront dgerential cen rnfuation of lysed protoplasts
Distribution Parameter Proioplast lysate Pellet Supernatant
mg (% of total recovered) Protein 144 S tarch -I - 2 nmol min nmol min (% of total recovered) 3 Catalase (x IO ) 1901 Cyt c oxidase 1492 GAD 124 SSADH GABA: pymvate-T an.d., not determined. supematant. Thus, differential centrifugation enriched the supematant fraction in
mitochondria and microbodies relative to amyloplasts. SSADH and GABA:pymvate-T
showed a sirnilar distribution to Cyt c oxidase in the pellet and supematant.
The supematant fraction obtained from the dserential spin was centrifuged through a
pre-fc~edcontinuous Percoll gradient to separate the organelles (Fig. 3.1). No
definitive starch peaks were recovered in the gradient, but fractions 24 to 27 near the
bottom of the tube contained a srnall amount of starch (-27 mg; data not shown). Protein
was primarily recovered near the top of the gradient, together with 95% of the alcohol dehydrogenase activity (cytosolic marker). OnIy about 20% of the catalase activity was associated with this cytosolic fraction, indicating minor rnicrobody breakage. The
majority of the catalase activity (65%),occurred as a separate broad peak in the middle of the gradient. The remaining catalase activity occurred as a second minor peak at higher
Percoll density (fractions 24 to 27). This second catalase peak was coincident with peak mitochondrial-marker activities (Cyt c oxidase and isocitrate dehydrogenase), indicating some microbody contamination of the main rnitochondrial fractions. The mitochondria from fractions 24 to 26, had an outer membrane intactness of >go%, as indicated by the
Cyt c impermeabilty assay . Negligibie mitochondnal-marker activity was found in the cytosolic fractions.
The distribution of GAD activity was similar to that of the cytosolic marker with no activity in the rnicrobody or mitochondnal fractions. In contrast, GABA-T and SSADH shared activity profiles very sirnilar to the rnitochondrial markers, with negative measured activity (drawn as zero) recovered in the cytosolic fraction. The distribution of GABA:2- oxoglutarate-T was identica! to that of GABA:pyruvate-T, but its activity was only 10%
-50- Figure 3.1. Distribution of protein (A), marker enzymes (B) and GABA-shunt enzymes
(C) arnong organelle fractions separated by continuous Percoll-density centrifugation of the supernatant Fraction from lysed soybean-cotyledon protoplasts. ADH, alcohol dehydrogenase; Cat, catalase; Cyt c ox, Cyt c oxidase; GAD, glutamate decarboxylase;
DH,isocitrate dehydrogenase; Pyr:GABA-T, pyruvate:GABA transaminase; SSADH, succinic semialdehyde dehydrogenase fi-
Fraction Number (bottom) (data not shown). Only SSADH showed Iatency, as the activity was stimulated
approximately three times by the addition of 0.02% Triton X-100.
To funher demonstrate that mitochondna are capable of metabolizing GABA, [U-
"'CJGABA and unlabeled pyruvate were fed to desalted cytosolic (fractions 1-7) and peak
microbody (hction 19) and rnitochondnal (fraction 25) fractions. After 1 h, only the
osmotically-shocked mitochondrial and rnicrobody fractions catabolized &J-14C]G~B~,
showing malonate-induced accumulation of IabeIed succinate at rates of 13.6 and 6.4
nmol ha'fraction-', respectively (data not shown). Thus, the rate of succinate
accumulation by the rnicrobody fraction was only 47% of the rnitochondrial fraction.
This is in agreement with the extent of mitochondrid contamination of the microbody
fraction (43% of the mitochondrial fraction). Succinate formation was not observed in
the presence of AOA, an inhibitor of the PLP-dependent GABA-T (Tuin and SheIp,
1994).
3.4. Discussion
Here, a gentle and efficient method was developed to isolate and purify organelle
fractions from a protoplast preparation of soybean cotyledons. The prelirninary
differential centrifugation of the protoplast lysate removed the majority of amyloplasts
and enriched the supernatant with mitochondria and microbodies (Table 3.1).
Centrifugation of this supernatant through a continuous PercolI gradient separated cytosolic, microbody and mitochondnal activities (Fig. 3.1). This technique resulted in
very low organelle darnage as indicated by the high yield of intact rnitochondria and rnicrobodies. The distribution of GAD was not associated with any organelle fraction, but was
found exclusively in the cytosolic fractions (Fig. 3.1). This result agrees with previous
work (Dixon and Fowden, 1961; Wallace et al., 1984; Satya Narayan and Nair, 1986). In
contrast, the distributions of GABA:pyruvate-T and SSADH suggest that these activities
were exclusively located in the mitochondria and not in the cytosol (Fig. 3.1). The fact
that negative rates were obtained for GABA-T and SSADH in the cytosolic fractions
(shown as zero in Fig. 3.1), perhaps suggests that the substrates for these reactions (Le.
GABA and succinate semialdehyde, respectively) inhibit background reactions. For
example, Good and Muench (1992) dernonstrated that 1 mM GABA inhibits alanine
aminotransferase by 24%. Furthemore, accumulation of succinate semiaidehyde is
known to be toxic to cells (Hearl and Churchich, 1984) and may possibly inhibit other
dehydrogenases.
Feeding studies with w-14C]GABA complemented the findings that GABA-T and
SSADH are mitochondrid, because only the mitochondrial fraction (and microbody
fraction contaminated with rnitochondria) and not the cytosolic fraction, catabolized
GABA to succinate in the presence of malonate, an inhibitor of succinate dehydrogenase
(Journet et al., 1982). Labeled succinate was not detected in the presence of AOA, a potent inhibitor of transaminase reactions (John et ai., 1978), demonstrating that GABA catabolism involved GABA-T and SSADH.
This work agrees in part with a mitochondnal localization for GABA:2-oxoglutarate-
T as proposed by Tokunaga et al. (1976), but it contradicts the dual cytosolic- rnitochondrial distribution suggested for GABA:pyruvate-T (Dixon and Fowden, 196 1;
Wallace et al., 1984) and the 1 to 3 ratio for GABA:pyruvate-T and SSADH in the
-54- mitochondria and cytosol as suggested by Satya Narayan and Nair (1986). The latency of
SSADH suggests that this enzyme may be membrane-bound. In mammalian systerns, both GABA-T and SSADH are confined to the rnitochondria and form a stable enzymatic complex to promote substrate channeling (Hearl and Churchlich, 1984). Although the possibility of species- or tissue- dependent isozymes rernains, it is likely that the duai localization suggested by previous workers is a direct result of mitochondrial breakage during tissue maceration. According to Douce et al. (1987), maceration is detrimental to rnitochondrial integrity, producing envelope-free mitochondria and rnitochondria that reseal following rupture and loss of matrix content. 1 is clear from the present study that
GABA-T and SSADH in developing soybean cotyledons are exclusiveiy confined to the rnitochondria, and therefore GABA produced from glutamate via GAD in the cytosol must be transported across the mitochondrial membrane. CHAPTERFOUR
Identification and Characterization of GABA, Prohe, and QAC Transporters
from Arabidopsis thaliana
4.1. Introduction
In plants, the concentration of GAE3A. a ubiquitous four-carbon non-protein arnino
acid, is markedly stimulated by a variety of stress conditions, including hypoxia,
temperature shock, mechanical manipulation and darnage, water stress and
phytohormones (Bown and Shelp, 1989; Satya Narayan and Nair, 1990; Bown and Shelp,
1997). GABA accumulation is probably mediated via an activation of glutamate
decarboxylase, the first committed enzyme of GABA synthesis. as a result of increased
cytosolic ~a"-C~Mor H' levels (Snedden et al., 1995; Bown and Shelp, 1997).
However, decreased catabolism by mitochondrial enzymes (Breitkreuz and Shelp, 1995)
or changes in intracellular or intercellular transport may also contribute to changes in
GABA levels.
Evidence is available for GABA efflux from isolated cells of A. sprengen and soybean Ieaf (Secor and Schrader, 1985; Bown et al., 1989; Chung et al., 1992). Other
research has shown that GABA is always present in xylem, but is sometirnes absent from phloem; this xylem-borne GABA undergoes extensive metabolism in cells of mature
leaves, resulting in transfer of its carbon and nitrogen to other molecules (Bown and
Shelp, 1989). Like other amino acids, GABA import into the ce11 across the plasma membrane is likely coupled to the downhill flow of H' (Li and Bush, 1991; Li and Bush,
1992.: Bush, 1993). In the cell, GABA must be îransported across the rnitochondrial
-56- membranes where it is metabolized to succinate (Breitkreuz and Shelp, 1995). Although plants possess multiple systems for the transport of proteinaceous amino acids (Frornmer. et al., 1994; Fischer et al., 1995; Rentsch et al., 1996; Fischer et al., 1998), the existence of GABA transporters in plants has not been reported.
A number of proteins that transport GABA in animal, fungal and bacterial systems have been identified (Table 1.1 and refs. therein). In animais, where GABA functions as an important neuroinhibitory compound (Kanner, 1994). different transporters, which
Vary in sequence homology, localization and substrate specificities have been characterized. Most use GABA, but also transport other related compounds such as 3- aminopropionic acid, glycine betaine, a quatemary ammonium compound (QAC), and taurine. In Saccharomyces cerevisiae, three different proteins are involved in GABA transport: one is specific for GABA (UGA4), a second uses both GABA and proline
(PUTLC), and a third uses most arnino acids (GAPl) (Jauniaux et al., 1987; Talibi et ai.,
1995).
In the present study, heterologous cornplementation of a GABA-transport-deficient yeast mutant (Fromrner and Ninnemann, 1995) and analysis of previously isolated amino acid transporters (Fromrner et al., 1994; Fischer et al., 1995; Rentsch et al., 1996) were used to identify GABA transporters in A. dialiana. Three known amino acid transporters, amino acid pemease 3 (AAP3) (Fischer et al., 1995) and the proline transporters 1 and 2
(ProTl and 2) (Rentsch et al., 1996), were capable of GABA transport. Since Pron expression is inducible by drought and salt stress, a condition that also increases GABA levels in plants (Rhodes et al., 1986; Bolafin et al., 1995; Serraj et al., 1998), it was characterized in further detail. Like the animal, fungal and bacterial GABA transport
-57- systems (Table 1. l), ProT2 effectively transported other related compounds.
4.2. Material and Methods
4.2.1. Plant Growth
A. thaliana (L.)Heynh (C24 ecotype) were gerrninated and grown to maturity on AM
medium according to Schmidt and Willmitzer (1988). Plates either contained no
nitrogen, or were supplemented with 20 rnM GABA or 10 rnM N in the form of
(M4)$O4and KNO, (66% MI,' and 33% N0,J.
4.2.2. Molecular Cloning
Al1 molecular biology reagents were from MBI Fermentas GmbH (Germany).
4.2.3. Yeast Growth, Transformation and Selection
The S. cerevisiae strain 22574d (tcra3-l, gapl-1, put4-1, uga4-l),deficient in GABA transport (Jauniaux et al., 1987), was transformed with an A. rhaliana (L.) Heynh
(Landsberg erecta ecotype) cDNA expression library in the yeast expression vector pFL61. which possesses CIRA3 as a selectable marker (Dohmen et al., 1991; Minet et al.,
1992). Transfonnants were selected on SD medium (Rose et al., 1990): washed from the plates in nitrogen-free liquid SD medium and re-selected on nitrogen-free SD medium supplemented with 20 mM GABA as the sole nitrogen source. Although slight background growth of the control transformants was evident, effective screening of
Arabidopsis cDNA candidates was easily performed. Background growth by this strain on GABA is possibly due to yet a fourth endogenous GABA permease (Grenson et al.,
-58- 1987).
The selection process on GABA was complicated by either a high suppressor activity or reversion frequency, which resulted in many fdse positives (reversion frequency
0.88%). 5-Fluoro-orotic acid (FOA) was used to select for these revertants by a modification of a method normaIIy used to select for mutagenized yeast cells (Rose et al.,
1990). Putative candidates were streaked ont0 nitrogen-fiee SD plates containing 0.1%
FOA, 20 mM GABA and uraciI(40 yg/ml), thereby allowing ody the revertants to grow.
Conversion of FOA to a highly toxic denvative (5-fluorouracil) by URA3 (carried on pFL61 dong with any putative plant GABA transporters) limits the gro:vth of al1 yeast that are dependent on the expression of the shuttle vector. With this negative screening method, four clones were identified as positive candidates. To confirrn these results, recombinant plasmids from these clones were rescued, arnplified, re-introduced into the mutant and re-selected on GABA. Further characterization of these plasrnids was performed by restriction digest analysis and full or partial sequencing of up to 500 bp at both 3'- and S'-ends.
4.2.4. DNA Sequencing
The cDNAs isolated by complementation were sequenced using the AB1 PRISM Dye
Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer) on the AB1 PRISM
Sequencer Mode1 3 10 or 377 (ferkin Elmer).
4.2.5. Yeast Growth Assay
The yeast strain 22574d was transfonned with the amino acid permease cDNAs
-59- (AAPI-6 and ProTl -2) in the yeast expression vector pFLdl. The transformed yeast strains and the clones found by complementation were then grown in liquid SD medium
&Othe same optical density (OD) in the logarithrnic phase, washed in nitrogen-fiee medium, and equai amounts streaked ont0 plates containing varying concentrations of proline and GABA. The plates were grown for 4 days at 28 OC.
4.2.6. Yeast Transport Assay
Yeast cells were grown in liquid SD medium at 28 OC to the logarithmic phase and harvested at an OD, of 0.5. Cells were washed twice and resuspended in 0.3 M sorbitol to a calculated OD, of 20. Prior to the transport rneasurernents, 100-p1 aliquots of cells were supplemented with 100 mM glucose and incubated for 10 minutes at 28 OC. To start the reaction, each aliquot was added to an equal volume of 100 mM potassium phosphate buffer containing 0.3 M sorbitol, the appropriate radiolabeled substrate (either
18.5 kBq of [ I -14C]G~BA,[U-"Clproline, [methyl-14C]cholineor [U-"Cladenine, or 37 kBq of [2,3 -)H]proline; simultaneous transport experiments utilized both [L1"C]GABA and [2,3 -3H]proline; Sigma), and 1 mM of the respective unlabeled substrate. The pH of the transport buffer was 4.0 (Fischer et al., 1995), unless stated otherwise. Where the pH was altered, the relative proportion of zwitterionic GABA and proline was calculated using the Henderson-Hasselbalch equation (pH = pK + log ([A']/[HA]) and pK, values of
4.03 for GABA and 1.95 for proline (Anonymous, 1972). Samples (45 pl) were taken after 1,2,3 and 4 min, filtered on GFfC glass fiber fdters (Whatman), and washed irnmediately with 8 ml of ice-cold wash solution. To reduce non-specific binding, the wash solution was supplemented with 5 rnM GABA, proline, choline or adenine,
-60- depending on the experiment. Radioactivity (14C- and 3~-labeledsubstrates) accumulated in the cells on the filter was measured by liquid scintillation spectrometry (Beckrnan LS
6800). Cornpetition for GABA transport was performed in the presence of a 5-fold molar excess of other amino acids, glycine betaine and precursors, other betaines, taurine or polyamines. Al1 kinetic data represent the mean of' three independent expenments k S.D. performed on separate days. Kinetic constants were calculated using non-linear regression anaiysis.
4.3. Results
4.3.1. Growth of ArabzXopsis on GABA
Arabidopsis seeds germinated and grew on nitrogen-free medium supplemented with nitrogen in the form of NH,' and NO, or GABA; seeds germinated on nitrogen-free medium did not grow (Fig. 4.1 .A). These observations indicate that external GABA, when supplied as the sole nitrogen source, can support the growth of Arabidopsis, and suggest that plants contain GABA transporters, at least in roots.
4.3.2. Heterologous Complementation of a GABA Transport Mutant
The S. cerevisiae strain 22574d, which carries mutations in the general arnino acid, proline and GABA pemease genes, grows poorly on minimal media supplemented with proline or GABA as the sole nitrogen source (Grenson et al., 1987). To identifj plant permeases capable of transporting GABA, it was transformed with an expression library derived from Arczbidopsis seedling rnRNA (Minet et al., 1992). After transformation, the cells were plated under selective conditions and four independent
-61- Figure 4.1. Growth of Arabidopsis (A)and yeast complements (B) on media supplemented with GABA as the sole nitrogen source. A. thaliana ecotype C24 seeds were sown on media containing either 10 mM N in the form of (M4),SO4and KNO,
(66% NI&+ and 33% NO,') (left), 20 mM GABA (rniddle) or no nitrogen (right; panel A).
The S. cerevisiae mutant 22574d transformed with pFL6I (control), or with the same vector carrying previously identified transporters (AAP3, ProTl and ProT2; Fischer et al..
1995; Rentsch et al., 1W6), or isolated complements (AAP3' and ZFPI). Transformants were streaked on medium containing either 10 rnM (m4)ZS04,1 rnM proline, or 1, 10 and 20 rnM GABA (panel B). clones that grew on GABA were found. Retransformation of the mutant with the recombinant plasmids isolated from these four clones confmed that growth was not the result of reversion or second-site mutations. Restriction digests and full-length or partial
3'- and S'-end sequencing of plasmid DNAs reveaied two classes. Two cDNAs (narned
ZFPI' and ZFPI for zinc finger protein), differing in length (-0.9 and 1. I kb, respectively), encode proteins that show homology (>50% identity) to RMAl (Matsuda and Nakano, 1998) and RZF (Zou and Taylor, 1997), two zinc finger proteins from
Arabidopsis; these were not pursued because they represent putative transcription factors that probably activate a silent endogenous GABA transporter in yeast. The other two cDNAs encode the previously characterized arnino acid transporter AAP3 (Fischer et al.,
1995). One of these, designated AAP3', has a 5'-untranslated region 6 1 bp shorter than
AAP3, and a coding sequence that differed from AAP3 by the addition of three nucleotides, thereby altenng the coding sequence by 5 amino acids. This difference in the coding sequence was an error in the earlier reported sequence (Fischer et al., 1995), which has been updated (NCBI Accession No. X77499).
4.3.3. Growth Assay of Yeast Complements
The growth of the mutant yeast strain 22574d expressing previously cloned amino acid permeases (AAPI-6 and ProTI-2) and two of the four cDNAs found by cornplementation in this study (AAP3' and ZFPI) were compared on various concentrations of GABA or 1 rnM proline. Growth media supplemented with 10 mM
(NH,),SO, served as a control; wild type yeast, al1 complements and the mutant 22574d transformed with the different cDNAs or pFUI grew equally well. Al1 cDNAs mediated
-64- growth on 1 mM proline, however, growth on GABA was variable. With the exception of AAPI. al1 AAPs and ProTs mediated growth on 20 mM GABA (Fig. 4.1.B of the
AAPs, only 22574d-AM3 is shown). However, 22574d expressing AAP3' and Pro72 grew more vigorously on lower GABA concentrations than cells expressing AAP3 and
ProTI, respectively.
Cells which grew on 20 mM GABA (Fig. 4.1 .B, 22574d-AAP2-6, ProTl-2. AAP3' and ZFPI) were tested for their ability to take up 1 mM [l-'%]GABA (data not shown).
Of these, only cells expressing AAP3', ProTl and Proï2 significantly accumulated '% within 5 min, to levels higher than the mutant 22574d transfomec! with pFL6I. Since
ProT 1 and ProT2 share similar kinetic properties (Rentsch et al., 1996), only ProT2 and
AAP3' were characterized in greater detail.
4.3.4. Characterization of AAP3 and ProT2 as GABA transporters
GABA or proline transport by AAP3' and ProT2 was linear for at least 5 min and exhibited saturable, concentration dependent uptake (Fig. 4.2). At pH 4.0, the apparent
Km (mean t S.D.) of AAP3' for GABA was estimated to be 12.9 + 1.7 rnM with an apparent V,, of 3.22 t 0.25 pmol min-' g-' yeast DW (Fig. 4.2.A). In contrast, the aKinities of ProT2 for GABA and proline were higher, with apparent K,ns of 1.70 * 0.34 mM and 0.42 -c 0.06 rnM, respectively (Fig. 4.2.B). The apparent Vdof ProT2 for
GABA (2.3 1 + 0.39 pmol min-' g-' yeast DW) and proline (1.74 I0.21 pmol min-' g-' yeast DW) were not simcantly different (Fig. 4.2.B).
The simultaneous transport of [l-'4C]GABA and [2, 3-3H]prolineby yeast expressing Pron was measured in response to increasing pH (Table 4.1). The rates were
-65- Figure 4.2. Concentration dependence of GABA and proline transport by A-3' (A) ad
ProT2 (B). Rates of [1-'%]GABA uptake plotted as a function of GABA concentration,
by the S. cerevisiae mutant 22574d transformed with either pFL6I (open square) or with
the sarne vector carrying AAP3' (closed square; panel A) and rates of [l-14C]GABA
uptake (open circles) or [2, 3-3~]prolineuptake (closed circles) as a function of substrate concentration, by the S. cerevisiae mutant 22574d expressing Pro72 (panel B). The figures show one of three independent uptake experirnents performed at pH 4.0. Rate of Su bstrate Transport Rate of GABA Transport (nmol min-'mgœ1yeast DW) (nmol rnin-'rng-l yeast DW) optimum between pH 4.7 and pH 5.2, and are consistent with the pH optimum of 5 for the transport of GABA (McKelvey et al., 1990) and other substrates into S. cerevisiae
(Stambuk et ai., 1996). While the ratio of GABA to proline never exceeded 0.48, it did increase with increasing pH, a trend associated with the predicted increase in zwitterionic
GABA.
To assess the biochemical characteristics of ProT2, a variety of compounds that arc stnicturaliy similar to GABA or proline, or are transported by animal, fungal or bacterial GABA transport systems (reviewed in Table 1. l), were chosen. Al1 structural analogs of GABA, which differ in the position of their arnino group with respect to the a- carboxylate group, inhibited 11-"CIGABA transport (Fig. 4.3), with DL-3-aminobutyrate giving the strongest inhibition (86%), followed by 4- and DL-2-isoforms (77% and 30%, respectively). Each of the L- and D-isoforms of proline inhibited [l-lJC]GABAtransport by 91%; these were the most effective amino acids tested. In contrast, 3-aminopropionic acid and ectoine inhibited GABA transport very slightly (approximately 20%).
4.3.5. Characterization of ProTl as QAC Transporter
Transport of [~-'"C]GABAwas also inhibited by al1 QACs tested (Fig. 4.3).
Glycine betaine (comrnonly referred to as betaine) and its direct metabolic precursors
(choline and betaine aldehyde) (Rhodes and Hanson, 1993) inhibited [1-14C]GABA transport by 96 to 99%. Other betaines such as camitine (3-hydroxy-GABA betaine), trigonelline (nicotinic acid N-methylbetaine) and betonicine (Chydroxy-proline betaine) also significantly inhibited GABA transport (9 1% to 72%, respectively). Taurine, and the polyamines, putrescine and spemiine, had no inhibitory effects (data not shown).
-69- Figure 4.3. Biochernical characterization of ProTî. Cornpetition of [ 1-lJC]G~BA uptake by the S. cerevisiae mutant 22574d expressing Pro72 in the presence of a 5-fold excess of the respective competitors. The uncompeted uptake rate of 1 rnM GABA served as the control. Data represents the mean of three independent experiments t S.D. performed at pH 4.0. ABA, aminobutyrate; 3-APA, 3-aminopropionate, Bet, glycine betaine; BetAld, gIycine betaine aldehyde; Beton, betonicine; Cam, cwnitine; Choi, choline; Ecto, ectoine; Pro, proline; Trig, trigonelline Control 0 Amino Acids O T GlvBet and Precursors W Other Betaines E T
T
Competitor (5 mM) Since choline was the most effective inhibitor of GABA transport, a üme course of [methyl-"C]choline transport into the mutant 22574d transformed with pFL61 or with the sarne vector carrying Pron was performed (Fig. 4.4). Expression of the endogenous yeast choline transporter (CTR1) was partially repressed by the addition of 1 rnM choline to standard yeast growth media (Nikawa et al., 1990). With corrections for [methyl-
'"Clcholine transport by the contro.ol, the mutant expressing ProR had a net transport rate of 180 nmol min-' g-' yeast DW, a rate 3-fold higher than the control and in a same order of magnitude as that for [l-'"C~GABAtransport in the absence of inhibitor (Fig. 4.3).
Choline did not affect yeast vitality since the transport of w-"~~adenineby endogenous yeast Hi-cotransport systems (André, 1995) was uninhibited by the presence of 0,s or 25 mM choline (783 t 97,778 t 63 and 800 + 34 nmol adenine min-' g" yeast DW).
4.4. Discussion
Growth of Arabidopsis on medium supplemented with GABA as the sole nitrogen source, provided preliminary evidence that plants possess proteins capable of transporting
GABA (Fig. 4.1 .A). Expression of Arabidopsis amino acid transporters (AAP2-6 and
ProT 1-2) (Frommer et al., 1994; Fischer et al., 1995; Rentsch et al., 1996) allowed the growth of a GABA-transport-deficient yeast mutant on high concentrations of GABA.
Previously, ProTl and ProT2 were identified as H+-coupled transporters that exhibit considerable selectivity for proline and are able to discriminate against many of the cornrnon a-amino acids (Rentsch et al., 1996); GABA as a potential substrate was never tested. In the present study, uptake experiments with [1-'"CIGABA demonstrated that
ProT2, and to a lesser extent ProTl, could mediate GABA transport. However,
-72- Figure 4.4. Time course of [methy 1- "C]choline uptake into yeast ceiis expressing ProE.
Uptakes were measured with the S. cerevisiae mutant 22574d transfonned with pK61
(open symbol) or with the same vector canying Pr072 (closed symbol). Yeast were grown in standard media supplemented with 1 rnM choline to suppress the endogenous choline transporter. Data represents the mean of three independent experiments IS.D. performed at pH 4.0. Time (min) calculation of the catalytic efficiency (VdK,) using the kinetic panmeters (Fig. 4.2) for proline (4.19 + 0.94) and GABA (1.39 I0.32), simultaneous transport studies of [l-
L4C]G~~~and [2, 3-3~proline (Table 4.1)' and competition of [ 1-"C]GABA upiake by
D- or L-proline versus GABA (Fig. 4.3), indicated that ProT2 preferred proline over
GABA. Furthemore, the ratio of GABA to proline transport increased with the pH- dependent availability of zwittenonic GABA. Therefore, GABA, like proline (which is present in the fuliy ionized state under the experimental conditions used), was preferentially transported in the zwitterionic or neutral form. The proline/GABA transporter (PUT4) from S. cerevisiae also displays preferential trz~sportof proline over
GABA (Table 1.1).
Since ProT2 had a preference for zwitterionic GABA (Tables 4.2), the apparent K, for neutrd GABA at pH 4.0 can be recalculated as 0.82 mM, one-half of the value for total GABA and double that for proline. With respect to the animal, fungal and bacterial
GABA transporters, the affinity of ProT2 for GABA is between those of the Low-affinity
VGAT transporters and the high-aîfinity GAT and APC transporters (Table 1.1). Recent studies showed that GabP, the GABA transporters frorn E. coli and B. subtilis, also transport zwitterionic GABA (King et al., 1995; Brechtel et al., 1996). The E. coli homologue exhibits a preference for analogs that rnimic a cyclic conformation of GABA, in contrast to the B. subtilis homologue which prefers analogs that mimic an extended or linear conformation (Brechtel et al., 1996). GABA is a very flexible molecule and can assume the cyclic structure depicted in Fig. 2.1 (Tanaka et al., 1978; King et al., 1995;
Brechtel et al., 1996). The stability of this ring-like conformation is likely dependent on ionic interactions between the amino and carboxyl groups; this is strongest in the
-75- zwitterionic form of GABA. According to Christensen et al. (1994), cyclic GABA is stmcturally sirnilar to proline and may be viewed as a structural analog.
DL-3-arninobutyrate proved to be a more effective inhibitor of ProT2-rnediated [l-
"CIGABA transport than 3-aminopropionic acid. This is consistent with results obtained for the E. coli GabP (Brechtel et al., 1996), but is different from the transport properties of the GAT family mernbers (Table 1.1 and refs. therein). The most effective inhibitors of GABA transport by RoT2 were betaines, which also accumulate in plants in response to stress (Rhodes and Hanson, 1993). Glycine betaine and its immediate metabolic precursors betaine aldehyde and choline were strong inhibitors (Fig. 4.3); other betaines, including carnitine, trigonelline (found in legumes) (Tramontano and Jouve, 1997) and betonicine (Hanson et al., 1994), also inhibited GABA transport, but were not as effective. As some of these compounds are achiral, and both D- or L-proline inhibited
GABA uptake, it suggested that Pro72 cannot discriminate between stereoisomers (Li and Bush, 1991; Li and Bush, 1992). Despite the structural dissimilarity between these compounds, the separation distance between polar functional groups may be the main factor determining substrate recognition. For example, the extra methyl group in 3- aminobutyrate prevents the charge separation in comparison to 3-arninopropionic acid, thereby improving the efficacy of its transport by E. coli GabP (Brechtel et ai., 1996). The ability of ProT2 to mediate significant [methyl-lJC]choiinetransport (Fig. 4.4) suggests that the inhibition of GABA transport by choline (Fig. 4.3) reflects competition at the binding site, rather than inhibition at an allosteric site. Regardless of the detailed mechanism, the data presented here support a stnicture/function mode1 in which molecules having different sizes and shapes are capable of interaction with Pro= either
-76- as transportable substrates or inhibitory ligands. Glycine betaine and other QACs, including p-alanine betaine. choline-O-sulfate, betonicine and stachydnne (proline betaine) (Rhodes and Hanson, 1993; Hanson et al., 1994) accumulate in many plant species under the same drought and salt stress conditions that induce the expression of
Pro72 (Rentsch et al., 1996) and the accumulation of proline (Rhodes and Hanson, 1993;
Yoshiba et al., 1997) and GABA (Rhodes et al., 1986; Bolarin et al., 1995; Serraj et al.,
1998). Thus. the ProTs may represent general carriers that facilitate the transport of a variety of stress-related compounds which can act as osmolytes or osmoprotectants.
Heterologous complementation of a GABA-transport-deficient yeast strain identified a previously reported AAP3 (AAP3') as a putative GABA transporter. Even though the affinity of AAP3' was about an order of magnitude lower than ProT2 and several orders of magnitude lower than mernbers of the GAT and APC farnilies (Table 1.1). yeast cells expressing AAP3' grew better than those expressing other amino acid transporters, including AAP3 and Pro= (Fig. 4.1.B). One possible explanation for this result is that the growth of S. cerevisiae on glucose results in acidification of unbuffered medium through an activation of the plasma membrane H+-ATPase (Carme10 et al., 1997)- thereby decreasing the availability of zwittenonic GABA, the preferred substrate for ProT2.
AAP3 is less affected by the charge of the side group of the arnino acid and transports a wide variety of acidic, neutral and basic amino acids (Fischer et al, 1995; Fischer et al,
1998). Lysine, a good transportable substrate of AAP3 that carries a positively charged
R-group (Fischer et al, 1995), resembles the charge symmetry of dibasic GABA.
Altematively, the 5'-UTR may act as a negative regulator of gene expression (Shantz et al., 1994), thereby providing the possibility that the shorter 5'-UTR of AAP3', compared
-77- to AAP3, increasing the heterologous transcription or other post-transcriptional events.
In planfa, the apoplastic pH is about 5.5 (Shelp et al., 1987), implying that 95% of
GABA found in the xylem is in the zwitterionic form. Furthemore, in unstressed plants, the GABA levels in xylem sap are about 0.1 m.(Shelp et al., 1987; Serraj et al.. 1998) and are dramatically increased (230%) dunng drought conditions (Serraj et ai., 1998).
GABA concentrations in the range of 6-39 mM in ce11 suspension cells adapted to water stress have been dociunented (Handa et al., 1983; Rhodes et al., 1986; BinzeI et al.,
1987). Zn Arabidopsis, the expression of AAP3 and Pro22 occurs in roots (Fischer et al.,
1995; Rentsch et aI., 1995) a result consistent with the growth of Arabidopsis on GABA as the sole N source (Fig. 4.1 .A). Together, these data suggest that AAP3 and ProT2 are physiologically significant in the transport of GABA within the plant.
Although significant progress has been made in the area of plant arnino acid transport in the last half decade (Fischer et al., 1998), little is known about the expression and regdation of transporter activity. In yeast, expression of a GABA-specific pemease
UGA4 is strongly induced by its substrate through the concerted efforts of two positive trans-acting regulatory proteins (UGA3 and UGA35) (Talibi et al., 1995). Both UGA3 and UGA35 contain a Cys,-Zn, binuclear zinc finger that serves as a DNA-binding domain in many tram-acting factors. A gene encoding a putative zinc-finger protein
(ZFPI) identified through heterologous complementation probably activates an unknown endogenous yeast GABA transporter (Grenson et al., 1987), suppressing the yeast mutant phenotype. This finding is rerniniscent of another isolated zinc-finger protein which enables yeast mutants to grow on sucrose (Kühn and Frommer, 1995). New transport genes were not identified using heterologous complernentation or transformation of the
-78- yeast mutant with a full length Arabidopsis cDNA (NCBI Accession # AF019637) homologous to the yeast GABA-specific transporter UGA4 (29% identity, 48% similarity) (Breitkreuz et al., unpublished). Therefore. we assume that the AAPs and
ProTs represent the major functional GABA transporters that can be identified by yeast complementation.
In conclusion, AAP3 and ProT2, amino acid transporters that are representative of the
AAP and ROTgene families from Arabidopsis recognize and transport GABA, albeit with different apparent afinities and maximal velocities. Recognition of the zwittenonic state of GABA seems to be an important parameter for substrate recognition and tr%nsport by ProT2. In addition to [I-'~C]GABAand [2,3-'~Iproline transport, hoTZ also transported [methyl-'JC]choline, providing evidence that the inhibition of [l-14C]GABA uptake by choline and QACs was the result of cornpetition for the binding site. Although different rnetabolic routes are involved in the synthesis of these compounds, ProT2 may be a common carrier in plants for these stress-related compounds. CHAPTERFIVE
General Discussion
Previously, the localization of GABA-T and SSADH, the enzymes responsible for the catabolism of GABA, was uncertain. Howevcr, with the use of organelles derived from developing soybean cotyledonary protoplasts and improved cellular fractionation techniques, it was clearly demonstrated that the enzymes involved in the GABA shunt are spatially segregated. GAD is localized to the cytosol, whereas GABA-T and SSADH are confined to the mitochondria, results in agreement with mamdian localization studies
(Hearl and Churchich, 1984). These findings indicate that GABA, like proline, which is synthesized in the cytosol, is transported into the rnitochondrion where it is catabolized
(Fi5 1) Although biochemical characterization of GABA transport into isolated mitochondria was unsuccessful (data not shown) and no molecular techniques are available to isolate transporter genes specifically from organelles of plants, animals or fungi (B. André and W. Frommer, persona1 coinmunication), it was possible to identify other plant proteins that hnction as GABA transporters. Characterization of previously isolated Arabidopsis amino acid transporters, in combination with heterologous yeast complementation, led to the identification of AM3 and ProT2 as GABA transport proteins (Fig. 5.1). Although the cellular location of these proteins has not been demonstrated in planta, it is speculated that AAP3 and ProT2 are targeted to the plasma membrane by their ability to complement a yeast mutant that lacks plasma membrane
GABA transporters. Regulation of these import proteins, in conjunction with a putative mitochondrial GABA transporter, may affect cytosolic GABA levels by contmlling the
-80- Figure 5.1. Summary of current knowledge conceming GABA metabolism and transport
in plants. Environmental stress signds, perceived by unknown plant rnechanisms,
increase of cytosolic Ca2+-CaMor H+ levels, which in tum stimulate the production of
GABA by GAD (A). GABA accumulation may also result fiom decreased transport into
the mitochondrion (B) or catabolism by GABA-T and SSADH (C),increased import (D), or synthesis via other sources (E). GABA may also be sequestered intracellulady (F) or exported (G). AAP, amino acid permease; CaM, calmodulin; GAD, glutamate decarboxylase; GABA-Ald, GABA aldehyde, ProT, proline transporter.
flux of GABA either into, or within the cell. Thus, cellular GABA accumulation may be
a result of increased synthesis, decreased catabolism by mitochondrial enzymes andfor
intra- or intercellular transport.
Since AM3 and Pro72 are constitutively expressed in roots (Fischer et al., 1995) and al1 tissues (Rentsch et al., 1996), respectively, and Pr072 expression is strongly is
induced by saIt and drought stress, it is possible that GABA undergoes intercellular andor Long-distance transport during nomal and water deficit conditions. Several studies have docurnented GABA accumuIations during water deficit (Rhodes et al., 1986;
Binzel et al., 1987, Fougère et al.. 1991; Bolarln et al., 1995; Serraj et al., 1998).
Although it is likely that increased synthesis is the major factor contnbuting to increased
GABA levels (Serraj et al., 1998), it is also possible that decreased catabolism andor increased transport into cells rnay affect GABA levels. Several studies compaiing the xylem and phloem contents from a variety of plant species suggest that GABA is present in xylem, but is sometimes absent from phloem (see Bown and Shelp, 1989 for review;
Serraj et al., 1998); the level of this amino acid in phloem during severe plant stress conditions has never been determined. Since GABA produced in source tissue such as mesophyll cells cm efflux out of the ce11 into the apoplastic space (Secor and Schrader,
1985; Chung et al., 1992), it is possible that GABA is loaded into the phloem and delivered to sink tissues such as roots or developing tissues, such as soybean cotyledons which readily metabolize GABA to succinate (Tuin and Shelp, 1994; Shelp et al., 1995).
GABA catabolism during water deficits may provide C, N and reducing equivalents essential for proline synthesis. GABA catabolism by mitochondrial2-oxoglutarate- dependent GABA-T and SSADH produces glutamate (Bown and Shelp, 1997; Van
-83- Cauwenberghe, 19981, succinate and NADH; the latter two compounds support rnitochondrid oxidative phosphorylation and the generation of ATP. Glutamate,
NAD(P)H and ATP are required for probe synthesis (Hare and Cress. 1997).
In addition to transporting I4c-labeled GABA and proline, which were transported in the zwitterionic form, ProT2 also transported L4C-labeledchc!In2. Evidence was also provided that ProT2 transports other stress-related compounds such as glycine betaine and other QACs, which function as compatible osmolytes during plant water deficit conditions (Rhodes and Hanson. 1993). Thus, the finding that a saltldrought inducible transporter is responsible for the common transport of GABA, proline, choline and possibly QACs, allows one to speculate about further roles of GABA during water deficit conditions.
By the classic definition, GBA, like proline and glycine betaine, fits the criteria of being a compatible osmolyte (Yancey, 1994). Chernically, GABA has no net charge at neutral pH (Anonymous, 1972) and is highly soluble in water (Yancey, 1994). Since
GABA rnay accumulate to low millimolar concentrations (Handa et al., 1983; Rhodes et al., 1986; Binzel et al., 1987) and possibly functions as a stable nitrogen storage compound (Bown and Shelp, 1997), it likely contributes no toxic effects to the cell. In fact, GABA at high concentrations (25-200 rnM) stabilizes and protects isolated thylakoids against freezing darnage in the presence of salt, exceeding the cryoprotective properties of proline (Heber et al., 1971). Evidence that GABA possesses hydroxyl radical scavenging activity, exceeding that of proline and glycine betaine at the sarne concentrations (16 mM), has also been demonstrated in vitro (Srninioff and Cumbes,
1989). Trossat et al. (1997) showed that GABA can be synthesized from 4-
-84- aminobutyraldehyde (a product of the polyamine catabolic pathway) by the chloroplast-
localized BADH that is involved in glycine betaine synthesis. Whether GABA has a
specific role (i-e. osmolyte/osmoprotectant) during water deficit conditions, or is
meiabolized (e.g. to support the production of known osmolytes such as proline) may be resolved with transgenic plants that overproduce GABA (Baum et al., 1996).
Altematively, specific targeting of these genes to roots or organellar membranes (e.g. chloroplasts) may be worthy and result in improved resistence to various plant stresses.
For exarnple, the availability of choline transport into the chloroplasts is likely the lirniting factor of glycine betaine synthesis (Nuccio et al., 1998). More pressing perhaps, is the need to generate anti-sense plants or find knock-out mutants of these transporters, which will provide a better understanding of their role. Only through approaches such as these can the true physiological function of GABA in plants be detennined. Aebi H (1984) Catdase in Vitro. In L Packer, ed, Methods in Enzymology, Vol. 105.
Academic Press Inc., London, pp 121-126
André B (1995) An overview of membrane transport proteins in Saccharomyces
cerevisiae. Yeast 11: 1575-16 11
Anonymous (1972) Dissociation constants of organic bases in aqueous solution. In RC
Weast, ed, Handbook of Chernisuy and Physics, CRC Press, Cleveland, Ohio, pp
Dl 17-Dl 18
Arazi T, Baum G, Snedden WA, Shelp BJ, Fromm H (1995) MolecuIar and
biochernical analysis of calmodulin interactions with the calmodulin-binding domain
of plant glutamate decarboxylase. Plant Physiol 108: 55 1-56 1
Atkins CA, Canvin DT (1994) Photosynthesis and CO2evolution by leaf discs: gas
exchange, extraction and ion-exchange fractionation of lJC-labeled photosynthetic
products. Can f Bot 49: 1225-1234
Aurisano N, Bertani A, Reggiani R (199%) Anaerobic accumulation of 4-
aminobutyrate in rice seedlings: causes and significance. Phytochernistry 38: 1147-
1150
Aurisano N, Bertani A, Reggiani R (1995b) Involvement of calcium and calmodulin in
protein and amino acid metabolism in nce roots under anoxia. Plant Ce11 Physiol. 36:
1525-2529
Aurisano N, Bertani A, Reggiani R (1996) Evidence for the involvement of GTP-
binding proteins in the process on anaerobic y-aminobutyrate accumulation in rice
-86- roots. J Plant Physiol149: 5 17-519
Ballantyne JS, Chamberlin ME (1994) Regulaüon of cellular amino acid levels. In K
Shnge, ed, Cellular and Molecular Physiology of CeIl Volume Regulation, CRC
Press, Boca Raton, Florida, pp 11 1 - 122
Baum G, Chen Y, Arazi T, Takatsuji H, Fromm H (1993) A plant glutamate
decarboxylase containing a calmodulin binding domain. J Bi01 Chem 268: 19610-
19617
Baum G, Lev-Yadun S, Fridmann Y, Arazi T, Katsnelson H, Zik M, Fromm H
(1996) Calmodulin binding to glutamate decarboxylase is required for regulation of
glutamate and GABA metabolism and normal developments in plants. EMBO J 15:
2988-2996
Bennett MJ, Marchant A, Green HG, May ST, Ward SP, Miliner PA, Walker AR,
SchuIz B, Feldmann KA (1996) Arabidopsis AUXI gene: a permease-like regulator
of root gravitropism. Science 273: 948-950
Bergman A, Gardestrom P, Ericson I (1980) Method to obtain a chlorophyll-free
preparation of intact rnitochondria from spinach leaves. Plant Physiol66: 442-445
Binzel ML, Hasegawa PM, Rhodes D, Handa S, Handa AK, Bressan RA (1987)
Solute accumulation in tobacco cells adapted to NaCl. Plant Physiol84: 1408-1415
Boia~MC, Santa-CNZ A, Cayuela E, Pérez-Alfocea F (1995) Short-term solute
changes in leaves and roots of cultivated and wild tomato seedlings under salinity. J
Plant Physiol 147: 463-468
Boorer KJ, Fischer WN (1997) Specificity and stoichiometry of the Arabidopsis
H+/amino acid transporter AAPS. J Bi01 Chem 272: 13040-13046 -87- Borden LA, Smith KE, Hartig PR, Branchek TA, Weinshank RL (1992) Molecular
heterogeneiry of the y-aminobutyric acid (GABA) transport system. J Bi01 Chem 267:
21098-21 104
Borden LA, Dhar TG, Smith KE, Branchek TA, Gluchowski C, Weinshank RL
(1994) Cloning of the human homologue of the GABA transporter GAT-3 and
identification of a novel inhibitor with selectivity for dus site. Receptors Channels 2:
207-2 13
Borden LA, Smith KE, Gustafson EL, Branchek TA, Weinshank RL (1995) Cloning
and expression of a betaine/GABA transporter from human brain. J Neurochem 64:
977-984
Bown AW, Chung 1, Snedden W, Shelp B (1989) Specific glutamate cotransport into
mesophyll cells and efflux of the major metabolite 4-arninobutyric acid. In J. Dainty,
ed, Plant Membrane Transport, EIsevier Science Publishers, City, State, pp 329-334
Bown AW, Shelp BJ (1989) The metabolism and physiological roles of 4-aminobutyric
acid. Biochem (Life Sci Adv) 8: 2 1-25
Bown AW, Shelp BJ (1997) The metabolism and functions of y-arninobutync acid. Plant
Physiol 115: 1-5
Bradford M (1976) A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein dye-binding. Anal Biochem 72:
248
Brechtel CE, Hu L, King SC (1996) Substrate specificity of the Escherichia coli 4-
aminobutyrate carrier encoded by gabP. J Bi01 Chem 271: 783-788
Breitkreuz KE, Shelp BJ (1995) Subcellular compartmentation of the 4-aminobutyrate -88- shunt in protoplasts früm deveIoping soybean cotyledons. Plant Physiol 108: 99-103
Bush DR (1993) Proton-coupled sugar and amino acid transporters in plants. Annu Rev
Plant Physiol Plant Mol Bi01 44: 513-542
Bush DR, Langston-Unkefer PJ (1988) Arnino acid transport into memebrane vesicles
isolated from zucchini: evidence of a proton-amino acid symport in the plasmalemma.
Plant Physiol88: 487-490
Canvin DT, Salon C (1997) Photorespiration and CO,-concentrating mechanisrns. In DT
Dennis, DH Turpin, DD Lefebvre, DB Layzell, eds, Plant Metabolism, Addison
Wesley Ltd, Harlow, England, pp 3 14-340
Carme10 V, Santos H, Sa-Correia 1 (1 997) Effec t of extrace!Iular acidification on the
activity of plasma membrane ATPase and on the cytosolic and vacuolar pH of
Saccharomyces cerevisiae. Biochim B iophys Acta 1325: 63-70
Carroll AD, Fox GG, Laurie S, Phillips R, Ratcïiffe RG,Stewart GR (1994)
Ammonium assimilation and the role of y-aminobutync acid in pH homeostasis in
carrot ce11 suspensions. Plant Physiol 106: 5 13-520
Chang HC, Bush DR (1997) Topology of NAT;?, a prototypical example of a new family
of arnino acid transporters. J Bi01 Chem 272: 30552-30557
Chen Y, Baum G,Fromm H (1994) The 58-kD cdmodulin-binding glutamate
decarboxylase is a ubiquitous protein in petunia organs and its expression is
developrnentally regulated. Plant Physiol 106: 1381 - 1387
Chen L, Bush DR (1997) LHT1, a lysine and histidine specific amino acid transporter in
Arabidopsis. Plant Physiol 115: 1127- 1134
Cholewa E, Cholewinski AJ,Shelp BJ, Snedden WS, Bown AW (1997) Cold shock- -89- stimulated y-arninobutyric acid synthesis is mediated by an increase in cytosolic Ca",
not by an increase in cytosolic H'. Can J Bot 75: 375:382
Christensen HN, Greene AA, Kakuda DK, MacLeod CL (1 994) Specia. transport and
neurological significance of two amino acids in a configuration conventionally
designated as D. J Exp Bi01 196: 297-305
Chung 1, Bown AW, Shelp BJ (1992) The production and efflux of 4-aminobutyrate in
isolated rnesophyll celis. Plant Physio199: 659-664
Clark JA, Deutch AY, Gallipoli PZ, Amara SG (1992) Functional expression and CNS
distribution of a B-alanine-sensitive neuronal GABA transporter. Neuron 9: 337-348
Cooper P, Seiman MT (1974) An analysis of the effects of tobacco mosaic virus on
growth and the changes in the fi-ee amino compounds in young tomato plants. Ann
Bot 38: 625-638
Crawford LA, Bown AW, Breitkreuz KE, Guinel FC (1994) The synthesis of y-
arninobutyric acid in response to treatments reducing cytosolic pH. Plant Physiol 104:
865-87 1
Delauney AJ,Verma DPS ( 1990) A soybean Al-pyrroline-5-carboxylate reductase gene
was isolated by functional complementation in Escherichia coli and is found to be
osmoregulated. Mol Gen Genet 221: 299-305
Delauney AJ, Verma DPS (1993) Proline biosynthesis and osmoregulation in plants.
Plant J 4: 2 15-223
Dennis DT, Huang Y, Negm FB (1997) Glycolysis, the pentose phosphate pathway and
anaerobic respiration. In DT Dennis, DH Turpin, DD Lefebvre, DB Layzell, eds,
Plant Metabolism, Addison Wesley Ltd, Harlow, England, pp 105- 123 -90- Dixon ROD, Fowden L (1961) y-Aminobutyric acid metabolism in plants. Ann Bot 25:
513-530
Dohlmen RJ, Strasser AWM, Honer CB, Hollenberg CP (1991) An efficient
transformation procedure enabling long-term storage of competent cells of various
yeast genera. Yeast 7: 69 1-692
Douce R, Bourguignon J, Brouquisse R, Neuburger M (1987) Isolation of plant
mitochondria: general principles and cnteria of integrity. In L Packer, R Douce, eds,
Methods in Enzymology, Vol. 148. Acadernic Press, Inc., London, pp 403-415
Ferson AE, Wray LV, Fisher SH (1996) Expression of the Bacillus subtilis gubP gene
is regulated independently in response to nitrogen and amino acid availability. Mol
Microbiol22: 693-70 1
Fischer WN, Kwart M, Hummel S, Frommer WB (1995) Substrate specificity and
expression profile of amino acid transporters (AAPs) in Arabidopsis. J Bi01 Chem
270: 16315-16320
Fischer WN, André B, Rentsch D, Krolkiewicz S, Tegeder M, Breitkreuz KE,
Frornmer WB (1998) Amino acid transport in plants. Trends Plant Sci 3: 188-195
Flores HE, Protacio CM, Signs MW (1989) Primary and secondary metabolism of
polyamines in plants. In JE Poulton, JT Romeo, EE Conn, eds, Plant Nitrogen
Metabolisrn, Rec. Adv. Phytochem, Vol 23, Plenum Press, New York, pp 329-393
Ford Y-Y, Ratcliffe RG, Robins RJ (1996) Phytohormone-induced GAI3 A production
in transformed root cultures of Datura stramonium: an in vivo NI5 NMR study. J Exp
Bot 47: 81 1-818
Forlani G, Scaineliî D, Nielsen E (1997) Al-pyrrotine-5-carboxylatedehydrogenase -91- from cultured cells of potato. Plant Physiol 113: 1413- 1418
Fougère F, Le Rudulier D, Streeter JG (199 1) Effects of salt stress on amino acid,
organic acid, and carbohydrate composition of roots, baccteriods, and cytosol of alfalfa
(Medicago sativa L.). Plant PhysioI. 96: 1228-1236
Frommer WB, Hummel S, Riesmeier JW (1993) Expression cloning in yeast of a
cDNA encoding a broad specificity amino acid pennease from Arnbidopsis thaliana.
Proc Natl Acad Sci USA 90: 5944-5948
Frommer WB, Hummel S, Unseid M, Nimemann O (1995) Seed and vascular
expression of a high-affinity transporter for cationic arnino acids in Arabidopsis. Proc
Natl Acad Sci USA 92: 12036- 12040
Frommer WB, Kwart M, Hirner B, Fischer WN, Hummel S, Ninnemann O ( 1994)
Transporters for nitrogenous compounds in plants. Plant Mol Bi01 26: 1651 - 1670.
Fromrner WB, Ninnemann O (1995) Heterologous expression of genes in bacterial,
fungal, animal. and plant cells. Annu Rev Plant Physiol Plant Mol Bi01 46: 419-444
Gaiiego PP, Whotton L, Picton S, Grierson D, Gray JE (1995) A role for glutamate
decarboxylase during tomato ripening: the characterisation of a cDNA encoding a
putative glutamate decarboxylase with a calmodulin-binding site. Plant Mol Bi01 27:
1143-1 151
Good AG, Muench DG (1994) Purification and characterization of an anaerobically
induced alanine aminotransferase frorn barley roots. Physiol Plant 99: 1520-1525
Grenson M (1992) Amino acid transporters in yeast: structure, function and regulation.
In J De Pont ed, Molecular Aspects of Transport Proteins, Elsevier Science Pub, pp Grenson M, Muyldermans F, Broman K, Vissers S (1987) 4-Aminobutyric acid
(GABA) uptake in Baker's yeast Saccharomyces cerevisiae is mediated by the general
arnino acid permease the proline permease and a GABA-specific permease integrated
into the GABA-catabolic pathway. Biochem (Life Sci Adv) 6: 35-39
Guastella J, Nelson N, Nelson H, Czyzyk L, Keynan S, Miedel MC, Davidson N,
Lester HA, Kanner BI (1990) Cloning and expression of a ntbrain GABA
transporter. Science 249: 1303- 1306
Guimbal C, Klostermann A, Kilimam MW (1995) Phylogenetic conservation of 4-
arninobutyric acid (GABA) transporter isoforms. Eur J Biochem 234: 794-800
Handa S, Bressan RA, Handa AK, Carpita NC, Hasegawa PM (1983) Solutes
contributing to osmotic adjustment in cultured plant cells adapted to water stress.
Plant Physiol73: 834-843
Hanson AD, Rathinasabapathi B, Rivoal J, Burnet M, Dillon MO, Gage DA (1994)
Osmoprotective compounds in the Plumbaginaceae: a natural experiment in metabolic
engineering of stress tolerance. Proc Natl Acad Sci USA 91: 306-3 10
Hare PD, Cress WA (1997) Metabolic implications of stress-induced proline
accumulation in plants. Plant Growth Reg 21: 79-102
Harrington HM, Henke RR (198 1) Amino acid transport into cultured tobacco cells. 1.
Lysine transport. Plant Physiol. 67: 373-378
Hearl WG, Churchich JE (1984) Interactions between Carninobutyrate
aminotransferase and succinic sernialdehyde dehydrogenase, two rnitochondrial
enzymes. J Bi01 Chem 259: 11459-1 1463
Heber U, Tyankova L, Santarius KA (197 1) Stabilization and inactivation of biological -93 - membranes during freezing in the presence of amino acids. Biochim Biophys Acta
241: 578-582
Hinier B, Fischer WN,Rentsch D, Kwart M, Frommer WB (1998) Developmental
control of H+/aminoacid permease gene expression during seed development of
Arabidopsis. Plant J 14: 535-544
Hoagland DR, Amon DI (1938) Water-culture method for growing plants without soil.
Cal AgExp Sta Cir 347
Hohberg N, Bülow L (1998) Improving stress tolerance in plants by gene transfer.
Trends Plant Sci 3: 61-66 Y
Hu LA, King SC (1998) Functional sensitivity of polar surfaces on transmembrane helix
8 and cytoplasmic loop 8-9 of the Escherichia coli GABA (Caminobutyrate)
transporter encoded by gabP: mutagenic analysis of a consensus arnphipathic region
found in transporters from bacteria to mamals. Biochem J 330: 77 1-776
Jackson C, Dench JE, Hail DO, Moore AL (1979) Separation of mitochondna from
contamination subcelluiar structures utilizing silica sol gradient centrifugation. Plant
Physiol. 64: 150-153
Jauniaux JC, Vandenbol M, Vissers S, Brornan K, Grenson M (1987) Nitrogen
catabolite regulation of proline permease in Saccharomyces cerevisiae. Cloning of the
PUT4 gene and study of PUT4 RNA levels in wild-type and mutant strains. Eur J
Biochem 164: 601-606
John RA, Charteris A, Fowler LJ (1978) The reaction of amino-oxyacetate wi%
pyridoxal phosphate-dependent enzymes. Biochem J 171: 77 1-779
Jones MM, Osmond CB, Turner NC (1980) Accumulation of solutes in ieaves of -94- sorghum and sunflower in response to water deficits. Aust J Plant Physiol7: 193-205
Journet EP, Borner WD, Douce R (1982) Glutamate metabolism triggered by
oxaloacetate in intact plant mitochondna. Arch Biochem Biophys 214: 366-375
Kanner BI (1994) Sodium-coupled neurotransmitter transport: structure, function and
regulation. J Exp Bi01 196: 237-249
Kathiresan A, Tung P, Chinnappa CC, Reid DM (1997) y-Aminobutync acid
stimulates ethylene biosynthesis in sunflower. Plant Physiol 115: 129-135
Kavi Kishor PB, Hong Z, Mao G-H, Hu C-AA, Verma DPS (1995) Overexpression of
A'-pyrroline-5-carboxylate synthetase increases proline production and confers
osmotolerance in transgenic plants. Plant Physiol. 108: 1387-1394
Kim JW, CIoss EI, Aibritton LM, Cunningham JM (199 1) Transport of cationic
amino acids by the mouse ecotropic retrovirus receptor. Nature 352: 725-728
KMmerer TW (1987) Alcohol dehydrogenase and pyruvate decarboxylase activity in
leaves and roots of Eastern cottonwood (Pop~ilusdeltoides Bartr.) and soybean
(Glycine max L.). Plant Physiol84: 1210- 1213
King SC, Fleming SR, Brechtel CE (1995) Ligand recognition properties of the
Escherichia coli Caminobutyrate transporter encoded by gabP. J Bi01 Chem 270:
19893-19897
Kinraide TB, Etherton B (1980) Electrical evidence for different mechanisms of uptake
for basic, neutral, and acidic amino acids in oat coleoptiles. Plant PhysioL65: 1085-
1089
Kiyosue T, Yoshiba Y, Suchi-Sin P, Yamaguchi-Shinozaki K, Shinozaki K (1996) A
nuclear gene, encoding mitochondrial proline dehydrogenase, an enzyme involved in -95- proline metabolism, is upregulated by proline but downregulated by dehydration in
Arabidopsis. Plant Ce11 8: 1323- 1335
Kühn C, Frommer WB (1995) A novel zinc finger protein encoded by a couch potato
homologue from Solmurn tiiberosum enables a sucrose transport-deficient yeast strain
to grow on sucrose. Mol Gen Genet 247: 759-763
Lansac AR, Sullivan CY, Johnson BE (1996)Accumulation of fiee proline in sorghum
(Sorghum bicoïor) pollen Can J Bot 74: 40-45
Leigh RA, Ahmad N, Wyn Jones RG (198 1) Assessrnent of glycinebetaine and proline
compartmentation by analysis of isolated beet vacuoles. Planta l53: 34-41
Li 2-C, Bush DR (1990) ApH-dependent amino acid transport into plasma membrane
vesicles isolated from sugar beet leaves. Plant Physiol94: 268-277
Li 2-C, Bush DR (1991) ApH-Dependent arnino acid transport into plasma membrane
vesicles isolated from sugar beet (Beta vidgaris L.) leaves. Plant Physiol96: 1338-
1344
Li 2-C, Bush DR (1992) Structural determinants in substrate recognition by proton-
arnino acid symports in plasma membrane vesicles isolated fkom sugar beet leaves.
Arch Biochem Biophys 294: 5 19-526
Ling V, Snedden WA, Shelp BJ, Assrnan SM (1994) Analysis of a soluble calmodulin
binding protein from fava bean roots: Identification of glutamate decarboxylase as a
calmodulin activated enzyme. Plant Ce11 6: 1135- 1 143
Liron 2, Wong E, Roberts E (1988) Studies on uptake of y-aminobutyric acid by mouse
brain particles; toward the development of a model. Brain Res 444: 119-132
Liu Q-R, L6pez-Corcuera B, Mandiyan S, Nelson H, Nelson N (1993) Molecular -96- characterization of four pharmacologically distinct y-aminobutyric acid transporters in
mouse brain. J Bi01 Chern 268: 2106-21 12
Liu Q-R, Mandiyan S, Nelson H, Nelson N (1992) A family of genes encoding
neurotransrnitter transporters. Proc Natl Acad Sci USA 89: 6639-6643
L6pez-Corcuera B, Liu Q-R, Mandiyan S, Nelson H, Nelson N (1992) Expression of a
mouse brain cDNA encoding novel y-arninobutyric acid transporter. J Bi01 Chem
267: 17491-17493
Matsuda N, Nakano A (1998) RMAI, an Arabidopsis thaliana gene whose cDNA
suppresses the yeast sec15 mutation, encodes a novel protein with a RLNG finger
motif and a membrane anchor. Plant Ce11 Physiol39: 545-554.
Mayer R, Cherry J, Rhodes D (1990) Effect of heat shock on arnino acid metabolism of
cowpea cells. Plant Physiol94: 796-8 10
McCourt P, Keith K (1997) Molecular biology of development In DT Dennis, DH
Turpin, DD Lefebvre, DB Layzell, eds, Plant Metabolism, Addison Wesley Ltd.
Harlow, England, pp 37-49
McCutcheon SL, Ciccarelii BW, Chung 1, Shelp BJ, Bown AW (1988) L-Glutamate-
dependent medium alkalinization by Asparagus mesophyll cells. Plant Physiol88:
1024- 1047
McIntire SL, Reimer RI, Scbrrske K, Edwards RH, Jorgensen EM (1997)
Identification and characterization of the vesicular GABA transporter. Nature 389:
870-876
McKelvey J, Rai R, Cooper TG (1990) GABA transport in Saccharomyces cerevisiae.
Yeast 6: 263-270 -97- Minet M, Dufour ME, Lacoute F (1992) Complementation of Saccharomyces
cerevisiae auxotrophic mutants by Arabidopsis thaliana cDNAs. Plant J 2: 4 17-422
Moreau F, Romani R (1982) Preparation of avocado mitochondna using self-generated
Percoll density gradients and changes in buoyant density during ripening. Plant
Physiol70: 1380- 1384
Nelson H, Mandiyan S, Nelson N (1990) Cloning of the human brain GABA
transporter. FEBS Lett 269: 18 1- 184
Neuburger M (1985) Preparation of plant mitochondria, criteria for assessrnent of
mitochondrkd integrity and purity, survivd in vitro. In R Douce and DA Day, eds,
Encyclopedia of Plant Physiology, New Series, Vol. 18. Springer-Verlag, Berlin, pp
7-24
Niegemam E, Schulz A, Bartsch K ( 1993) Molecular organization of the Escherichia
coli gab cluster: nucleotide sequence of the structural genes gabD and gabP and
expression of the GABA pennease gene. Arch Microbiol160: 454-460
Nikawa J, Hosaka K, Tsukagoshi Y, Yarnashita S (1990) Prirnary structure of the yeast
choline transport gene and regulation of its expression. I Bi01 Chem 265: 15996-
16003
Nishimura M, Douce R, Akazawa T (1982) Isolation and characterization of
metabolically competent mitochondria from spinach leaf protoplasts. Plant Physiol
69: 9 16-920
Nuccio M, Nolte KD, Russell BL, Rathinasabapathi B, Gage DA, Hanson AD (1998)
Engineering glycine betaine synthesis using choline monooxygenase. Amencan
Society of Plant Physiologists Annual Meeting, Abstract No. 291 -98- Peng 2,Lu Q, Verma DP (1996) Reciprocal regulation of A'-pyrroline-5-carboxylate
synthetase and proline dehydrogenase genes control proline levels during and after
osmotic stress in plants. Mol Gen Genet 253: 334-341
Rampuîh AL, Bown AW (1996) Rapid y-arninobutyric acid synthesis and the inhibition
of the growth and development of oblique-banded leaf-roller larvae. Plant Physiol
111: 1349-1352
Ratcliffe RG (1995) Metabolic aspects of the anoxie response in plant tissue. In N
Smimoff, ed, Environment and Plant Metabolisrn: Flexibility and Acclimaüon, Bios
Scientific, Oxford, pp l 11 - 127
Rathinasabapathi B, Burnet M, Russell BL, Gage DA, Liao PC, Nye GJ, Scott P,
Golbeck JH, Hanson AD (1997) Choline rnonooxygenase, an unusual iron-sulfur
enzyme catalyzing the fist step of glycine betaine synthesis in plants: prosthetic
group chxacterization and cDNA cloning. Proc Nat1 Acad Sci USA 94: 3454-3458
Rayapati PJ, Stewart CR, Hack E (1989) Pyrroline-5-carboxylate reductase is in pea
(Pisum sativum L.) leaf chloroplasts. Plant Physiol 91: 58 1-586
Raymond P, Gidrol X, Salon C, Pradet A (1987) Control involïing adenine and
pyridine nucleotides. In PK Stumpf, EE Conn, eds, The Biochemistry of Plants: A
Comprehensive Treatise, Vol 1 1: Biochemistry of Metabolism, Academic Press, San
Diego, California, pp. 129- 176
Reggiani R, Cantu CA, Brimballa 1, Bertani A (1988) Accumulation and
interconversion of arnino acids in rice roots under anoxia. Plant Ce11 Physiol29: 981-
987
Rentsch D, Boorer KJ, Frommer WB (1998) Structure and hnction of plasma -99- membrane amino acid oligopeptide and sucrose transporters from higher plants. J
Membrane Bi01 162: 177- 190
Rentsch D, Himer B, Schmelzer E, Frommer WB (1996) Salt stress-induced proline
transporten identified by suppression of an arnino acid transport targeting mutant.
Plant Ce11 8: 1437-1446
Rentsch D, Laloi M, Rouhara 1, Schmelzer E, Delrot S, Frommer WB (1995) NTR1
encodes a high afinity oligopeptide transporter in Arabidopsis. FEBS Lett 370:
264-268
Rhodes D (1987) Metabolic responses to stress. In "The Biochemistry of Plants" (DD
Davies ed), Vol. 12, Academic Press, New York, pp. 201-24 1.
Rhodes D, Handa S, Bressan RA (1986) Metabolic changes associated with adaptation
to plant cells to water stress. Plant Physiol82: 890-903
Rhodes D, Hanson AD (1993) Quaternary ammonium and tertiary sulfonium compounds
in higher plants. Annu Rev Plant Physiol Plant Mol Bi01 44: 357-384
Roberts JK, Callis J, Wemmer D, Walbot V, Jardetzky O (1984) Mechanisms of
cytoplasmic pH regulation in hypoxic maize root tips and its role in survival under
hypoxia. Proc Nat1 Acad Sci USA 81: 3379-3383
Robert. JKM, Hooks MA, Miaullis AP, Edwards S, Webster C (1992). Contribution
of malate and amino acid rnetabolism to cytoplasmic pH regulation in hypoxic rnaize
root tips studied using magnetic resonance spectroscopy. Plant Physiol98: 480-487
Robinson SP, Jones GP (1986) Accumulation of glycinebetaine in chloroplasts provides
osmotic adjustment during salt stress. Aust J Plant Physiol 13: 659-668
Rose MD, Winston F, Hieter P (1990) Methods in Yeast Genetics, Cold Spnng Harbor -100- Laboratory Press, New York, pp 1-198
Satya Narayan V, Nair PM (1985) Purification and characterization of glutamate
decarboxylase from Solanurn tuberostcm. Eur J Biochem 150: 53-60
Satya Narayan V, Nair PM (1986) The Caminobutyrate shunt in Solanrrm tuberosum.
Phytochemistry 25: 997- 1001
Satya Narayan V, Nair PM (1989) Potato tuber succinic semialdehyde dehydrogenase:
purification and charactenzation. Arch. Biochem. Biophys. 275: 469-477
Satya Narayan V, Nair PM (1990) Metabolisrn, enzyrnology and possible roles of 4-
aminobutyrate in higher plants. Phytochemistry 29: 367-375
Schmidt R, Wiiimitzer L (1988) High efficiency Agrobacterium tumefaciens-mediated
transformation of Arabidopsis thaliana leaf and cotyledon explants. Plant Ce11
Reports 7583-586
Schnarrenberger C, Oeser A, Tolbert NE (1971). Development of microbodies in
sunflower cotyledons and castor endosperm during germination. Plant Physiol48:
566-574
Schousboe A, Wu J-Y, Roberts E (1973) Purification and characterization of the 4-
aminobutyrate-2-ketoglutaratetransaminase from mouse brain. Biochem 12: 2868-
Scott EM, Jakoby WB (1959) Soluble y-aminobutyric glutamic transaminase from
Psezrdomonasfluorescens. J Bi01 Chem 234: 932-937
Secor J, Schrader LE (1985) Amino acid eMux from cells and leaf discs. In R Schibles,
ed, World Soybean Conference III - Proceedings, Westview Press, Boulder, Colorado, Serraj R, Shelp BJ, Sinclair TR (1998)Accumulation of y-arninobutyric acid in
nodulated soybean in response to drought stress. Physiol Plantarum 102: 79-86
Shantz LM, Viswanath R, Pegg AE (1994) Role of the 5'-untranslated region of mRNA
in the synthesis of S-adenosylmethionine decarboxylase and its regulation by
spermine. Biochem J 302: 765-772
Shelp BJ (1987)The composition of phloem exudate and xylem Sap from broccoli
(Brassica oleracea var. itczlica) supplied with NH,+ NO, or m4N03.J Exp Bot 38:
1619-1636
Shelp BJ, Walton CS, Snedden WA, Tuin LG, Oresnik 1J, Layzeil DB (1995) GABA
shunt in developing soybean seeds is associated with hypoxia. Physiol Plant 94: 219-
228
Snedden WA, Arazi T, Fromm H, Shelp BJ (1995) Calciurn/calmodulin activation of
soy bean glu tamate decarboxylase. Plant Physiol 108: 543-549
Snedden WA, Chung 1, Pauls RH, Bown AW (1992) Protod-glutamate symport and
the regulation of intracellular pH in isolated mesophyll cells. Plant Physiol99: 665-
67 1
Snedden WA, Koutsia N, Baum G, Fromm H (1996) Activation of a recombinant
petunia glutamate decarboxylase by calciurn/calrnoduIin or by a monoclonal antibody
which recognizes the calmodulin binding domain. J Bi01 Chem 271: 4 148-4 153
Stambuk BU, De Araujo PS, Panek AD, Serrano R (1996) Kinetics and energetics of
trehalose transport in Saccharonzyces cerevisine. Eur J Biochem 237: 876-88 1
Steward FC, Dunan DJ (1965) Metabolism of nitrogenous compounds. In FC Steward.
ed, Plant Physiology: A Treatise Vol IV A. Academic Press, Toronto, pp 379-386 -102- Streeter JG, Thompson JF (1972a) In vivo and in vitro studies on y-arninobutytic acid
metabolism with the radish plant (Raphanus sativus L.). Plant Physiol49: 579-584
Streeter JG, Thompson JF (1972b) Anaerobic accumulation of y-aminobutyric acid
and alanine in radish leaves (Raphanus sativus L.). Plant Physiol49: 572-578
Swanson GT, Umbach JA, Gundersen CB (1994) Glia of the cholinergie electromotor
nucleus nf Torpedo are the source of the cDNA encoding a GAT- 1-1ike GABA
transporter. J Neurochem 53: 1- 12
Talibi D, Grenson M, André B (1995) Cis- and rrans-acting elements determining
induction of the genes of the y-aminobutyrate (GABA) utiIization pathway in
Saccharomyces cerevisiae. Nucleic Acids Res 23: 550-557
Tanaka K, Akutsu H, Ozaki Y, Kyogoku Y, Tomita K (1978) Molecular
confirmations of y-arninobutync acid and y-amino-P-hydroxybutyricacid in aqueous
solutions. Bull Chern Soc Jpn 51: 2654-2658
Taylor DC, Shelp BJ, Nelson LM, Grodzinski B ( 1988) Carbon and nitrogen
partitioning in young nodulated pea (wild type and nitrate reductase-deficient mutant)
plants exposed to NH,NO,. Physiol Plant 74: 593-601
Thomas JC, De Armond RL, Bohnert HJ (1992) Influence of NaCl on growth, proline,
and phosphoenolpyruvate carboxylase levels in Mesembryanthernum crystallinum
suspension cultures- Plant Physiol98: 626-63 1
Tokunaga M, Nakano Y, Kitaoka S (1976) The GABA shunt in the callus cells denved
from soybean cotyledon. Agr Biol Chern 40: 115-120
Tramontano W, Jouve D (1.997) TrigoneIline accumulation in salt-stressed legumes and
the role of other osmoregulators as ce11 cycle control agent. Phytochem 44: 1037- 1040 -103- Trossat C, Rathinasabapathi B, Hanson AD (1997) Transgenically expressed betaine
aldehyde dehydrogenase efficiently catalyzes oxidation of
dimethylsulfonioproprionaldehyde and a-aminoaldehydes. Plant Physiol 113: 1457-
1461
Tsushida T and Murai T (1987) Conversion of glutamic acid to y-arninobutyric acid in
tea leaves under anaerobic conditions. Agric Biol Chem 51: 2805-287 1
Tuin LG, Shelp BJ (1994) In situ ["Cl-glutamate metabolism by developing soybean
cotyledons. 1. Metabolic routes. J Plant Physiol 143: 1-7
Tuin LG, Shelp BJ (1996) In situ ["CI-glutamate metabolism by developing soybean
cotyledons. II. The importance of glutamate decarboxylation. J. Plant Physiol 147:
7 14-720
Turano FJ, Fang TK (1998) Charactenzation of two glutamate decarboxylase cDNA
clones from Arabidopsis. Plant Physiol 117: 14 11- 142 1
Van Cauwenberghe OR (1998) Partial Purification and Characterization of a Gamma-
Arninobutyrate Transaminase from Nicotiana tabaccim. University of Guelph, Ph. D.
Thesis
Vance C (1997) The molecular biology of N rnetabolism. In DT Dennis, DH Turpin. DD
Lefebvre, DI3 Layzell, eds, Plant Metabolism, Addison Wesley Ltd, Harlow, England,
pp 449-477
Verbruggen N, Viliarroel R, Van Montagu M (1993) Osmoregulation of a
pyrroline-5-carboxylate reductase gene in Arabidopsis thhaliana. Plant Physiol 103:
77 1-78 1
Vissers S, André B, Muylderrnans F, Grenson M (1989) Positive and negative -104- regulatory elements control the expression of the UGA4 gene coding for the inducible
4-aminobutyric-acid-specific pexmease in Saccharomyces cerevisiae. Eur J Biochem
181: 357-36 1
Vissers S, André B, Muyldermans F, Grenson M (1990) Induction of the
4-aminobutyrate and urea-catabolic pathways in Saccharomyces cerevisiae. Specific
and cornrnon transcriptional regulators. Eur J Biochem 187: 6 11-6 16
Voetberg GS, Sharp RE (1991) Growth of the maize pnmary root tip at low water
potentials. III. Role of increased prorine deposition in osmotic adjustment. Plant
Physiol96: 1 125-1 130
Wagner GJ (1979) Content and vacuole/extravacuole distribution of neutral sugars, fiee
amino acids, and anthocyanin in protoplasts. Plant Physiol. 64: 88-93
Wallace W, Secor J, Schrader L (1984) Rapid accumulation of y-aminobutyric acid and
alanine in soybean leaves in response to an abrupt transfer to Iower temperature,
darkness, or mechanical manipulation. Plant Physiol75: 170- 175
Walton CS (1993) Glutamate metabolism in developing soybean seed: significance of
the GABA shunt. University of Guelph, Masters Thesis
Weretilnyk EA, Hanson AD (1988) Betaine aldehyde dehydrogenase polymorphism in
spinach: genetic and biochemicai charactenzation. Biochem Genet 26: 143- 15 1
Weretilnyk EA, Hanson AD (1990) Molecular cloning of a plant betaine-aldehyde
dehydrogenase, an enzyme implicated in adaptation to salinity and drought. Proc Nad
Acad Sci USA 87: 2745-2749
Williamson CL, Slocum RD (1992) Molecular cloning and evidence for osmoreguIation
of the A'-pyrroline-5-carboxylate reductase @rd)gene in pea (Pisurn sativum L.). -105- Plant Physiol 100: 1464-1470
Yamaki S (1982) Distribution of sorbitol, neutrai sugars, free amino acids, malic acid
and hydrolytic enzymes in vacuoles of apple cotyledons. Plant Cell Physiol23: 88 1-
889
Yamauchi A, Uchida S, Kwon HM,Preston AS, Robey RB, Garcia-Perez A, Burg
MB, Handler JS (1992) Cloning of a Na'- and CLdependent betaine transporter that
is regulated by hypertonicity. J Bi01 Chem 267: 649-652
Yamaura 1, Matsumoto T, Funatsu M, Shinohara T (1988) Purification and some
properties of succinic semiaidehyde dehydrogenase from barley seeds. Agric Bi01
Chem 52: 2929-2930
Yancey PH (1994) Compatible and counteracting solutes. In K Strange, ed, Cellular and
Molecular Physiology of Ce11 Volume Regulation, CRC Press, Boca Raton, Florida,
pp 81-109
Yoshiba Y, Kiyosue T, Nakashima K, Yamaguchi-Shinozaki K, Shinozaki K (1997)
Regulation of levels of proline as an osmolyte in plants under water stress. Plant Ce11
Physiol38: 1095- 1102
Yun SJ, Oh SH (1998) Cloning and charactenzation of a tobacco cDNA encoding
calcium/calmodulin-dependent glutamate decarboxylase. Mol Cells 8: L 25- 129
Zik M, AraY T, Snedden WA, Fromm H (1998) Two isoforms of glutamate
decarboxylase in Arabidopsis are regulated by calcium/calmoàulin and differ in organ
distribution. Plant Mol Bi01 37: 967-975
Zou J, Taylor DC (1997) Cloning and molecular charactenzation of an Arabidopsis
thaliana RING zinc finger gene expressed preferentially during seed development. -106- Gene 196: 29 1-295