The Plant Cell, Vol. 7, 899-906, July 1995 O 1995 American Society of Plant Physiologists

Regulation of Lysine and Threonine Synthesis

Gad Galili Department of Plant Genetics, The Weizmann lnstitute of Science, Rehovot 76100, Israel

INTRODUCTION

Human and monogastric animals cannot synthesize 10 out of the 20 amino acids and therefore need to obtain these from their diet. Among the essential amino acids, lysine and threo- nine are considered to be exceedingly important in that they are the most limiting essential amino acids in cereal grains, which represent the largest source of food worldwide (Bright and Shewry, 1983). Because of the nutritional importance of lysine and threonine, the regulation of their metabolism has been studied extensively at the biochemical, genetic, and, more recently, molecular levels. Like many bacterial species, higher plants synthesize lysine and threonine from aspartate using two different branches of the aspartate family pathway, as shown schematically in Figure 1. The enzymes involved in ly- sine and threonine synthesis have been reviewed in detail (Bryan, 1980). In this review, I discuss the complex biochemi- cal, cellular, developmental, physiological, and environmental controls of the synthesis of lysine and threonine. I also focus on specific enzymes that play major regulatory roles in the synthesis of these amino acids.

BIOCHEMICAL REGULATION OF LYSINE AND THREONINE SY NTHESIS

Analyses of crude and purified enzyme preparations have shown that the aspartate family pathway in higher plants is regulated by several end product feedback inhibition loops (Figure 1, curved arrows). Similar feedback inhibition loops also operate in many bacterial species (Cohen and Saint- Girons, 1987). The activity of aspartate kinase (AK), the first enzyme in the pathway, is feedback inhibited by both lysine and threonine. In addition, lysine inhibits the activity of dihy- drodipicolinate synthase (DHPS), whereas threonine inhibits the activity of homoserine dehydrogenase (HSD). Therefore, in addition to feedback inhibiting the first enzyme in the path- way, lysine and threonine also inhibit the activities of the first enzymes after the branch point, that is, enzymes that are spe- cific for their own synthesis (Figure 1).

Main Regulatory Enzymes of the Aspartate Family Pathway

Aspartate Kinase

AK, the first enzyme of the aspartate family pathway, is neces- sary for the synthesis of all of its end product amino acids (Figure 1). AK catalyzes the phosphorylation of aspartate to form 3-aspartyl phosphate, with the accompanying hydroly- sis of ATP. Severa1 lines of evidence have shown that plants generally possess at least two or three AK isozymes. In most plant species, AK activity is additively inhibited by lysine and threonine. In addition, S-adenosyl methionine, which is derived from methionine, may act synergistically with lysine to inhibit the lysine-sensitive AK activity (Rognes et al., 1980). In bar- ley, AK activity can be separated into three isozymal peaks (Rognes et al., 1983). One isozyme, AK-I, is inhibited by threo- nine, whereas the other two, AK-II and AK-III, are inhibited by lysine. Similarly, one of two carrot AK isozymes is sensitive to lysine, whereas the other is sensitive to threonine (Sanako and Komamine, 1978). Two or three AK isozymes have also been reported in maize (Doston et al., 1989; Azevedo et al., 1992).

At least some AK isozymes in plants appear to be bifunc- tional proteins that contain both AK and HSD activities (Wilson et al., 1991; Azevedo et al., 1992). This observation was not unanticipated, because two of the Escherichia coli AK iso- zymes, AK-I and AK-II, both of which are lysine insensitive, exist as similar bifunctional proteins (Cohen and Saint-Girons, 1987). Amino acid sequence analysis of several peptides from a carrot bifunctional enzyme showed that it is most closely related to the E. coli AK-IIHSD-I isozyme (Wilson et al., 1991). The presence of a bifunctional AKIHSD enzyme in plants is of particular interest in terms of the evolution of these enzymes, because not all bacterial species possess such a bifunctional enzyme (Cohen and Saint-Girons, 1987).

The existence of bifunctional AKIHSD isozymes in plants has been confirmed by cloning and analysis of cDNAs and genes encoding such enzymes from several plant species (Weisemann and Matthews, 1993; Ghislain et al., 1994; Muehlbauer et al., 1995). The derived sequences of the plant enzymes are similar to the deduced amino acid sequences of the E. coli AK-IIHSD-I and AK-IIIHSD-I1 enzymes. So far, no

900 The Plant Cell

i noserine 2-3 dih

phpsphofjomoserjne

Figure 1. The Aspartate Family Biosynthetic Pathway.

The major regulatory enzymes and their products are indicated on a transmission electron micrograph of a plastid. The pathway from threonineto isoleucine is not shown. Yellow curved arrows represent feedback inhibition by the end product amino acids. AK, aspartate kinase; DHPS,dihydrodipicolinate synthase; HSD, homoserine dehydrogenase. The branches of lysine and threonine syntheses appear to be localized in theplastid, whereas methionine appears to be synthesized in the cytosol (for review, see Bryan, 1990).

lysine-sensitive AK isozyme that is similar to the £ coli lysine-sensitive AK-III has been cloned from plants, and it is thus stillnot known whether plants also contain an AK isozyme that lacksHSD activity, as is the case with £. coli AK-III (Cohen and Saint-Girons, 1987).

Dihydrodipicolinate Synthase

DHPS catalyzes the first reaction that is unique to lysine bio-synthesis, the condensation of 3-aspartic semialdehyde withpyruvate to form 2,3-dihydrodipicolinate. Among the majorregulatory enzymes of the aspartate family pathway in plants,DHPS is the most sensitive to feedback inhibition by lysine(las between 10 and 50 nM). DHPS is ~10-fold more sensi-tive to lysine inhibition than are plant lysine-sensitive AKs (I0.sbetween 100 and 700 u,M) and ~100-fold more sensitive to ly-sine inhibition than is £ coli DHPS (I0.5 of ~1000 nM). DHPShas been purified to near homogeneity from wheat (Kumpaisalet al., 1987), tobacco (Ghislain et al., 1990), maize (Frisch etal., 1991), and pea (Dereppe et al., 1992), and DHPS cDNAsand genes have been cloned from several plant species(Kaneko et al., 1990; Frisch et al., 1991; Vauterin and Jacobs,1994). Results of DNA gel blot analysis suggest that maizemay contain several DHPS genes (Frisch et al., 1991). Simi-larly, two different DHPS cDNAs have been cloned fromcommon wheat (Kaneko et al., 1990).

Homoserine Dehydrogenase

HSD catalyzes the first reaction that is uniquely associatedwith threonine, methionine, and isoleucine biosynthesis, theconversion of 3-aspartic semialdehyde into homoserine. Higherplants generally possess at least two forms of HSD—athreonine-sensitive form and an insensitive form (Bryan, 1980;Leaetal., 1985). HSD has been purified to near homogeneityfrom a variety of plant species. Carrot HSD can be reversiblyinterconverted in vitro from a threonine-sensitive trimeric formto a threonine-insensitive dimeric form, depending on whetherthe medium contains threonine (Matthews et al., 1989). It isstill not known whether this interconversion occurs in vivo and,if so, whether it helps regulate threonine synthesis. Charac-terization of purified HSD (Wilson et al., 1991; Azevedo et al.,1992) and cDNA clones (Weisemann and Matthews, 1993;Ghislain et al., 1994; Muehlbauer et al., 1995) has confirmedthat the HSD isozymes are linked to AK on single proteins.

Intracellular Localization of the Aspartate FamilyPathway

Many of the enzymes involved in lysine and threonine synthe-sis have been localized to plastids (Bryan, 1990). Analysesof cloned DNA sequences confirm that these enzymes are

Lysine and Threonine Synthesis 901

synthesized with transit peptides that direct them into the plastid (Kaneko et ai., 1990; Frisch et al., 1991; Weisemann and Matthews, 1993; Ghislain et ai., 1994; Muehlbauer et ai., 1995). Operation of the aspartate family pathway inside the plastid makes biological sense because some of the biochemical reac- tions require energy that is largely produced within this organelle. Nevertheless, some of the aspartate family enzymes, particularly those involved in the synthesis of methionine and S-adenosyl methionine, have been localized to the cytosol (Bryan, 1990). Severa1 functional and regulatory aspects of the intracellular localization of the aspartate family enzymes are discussed in later sections.

Functional Studies of the Aspartate Family Pathway

Although in vitro studies of enzyme activities have pinpointed the major regulatory enzymes of the aspartate family pathway, they have not provided sufficient evidence regarding the ex- act functional roles of each of these enzymes. Nor have these studies shown how these major regulatory enzymes operate in concert to regulate this complex pathway. Functional evi- dente of this regulation has been obtained by analyzing plants possessing feedback-insensitive forms of the major regulatory enzymes. These plants include both mutant strains and trans- genic plants that overexpress bacterial genes encoding several major regulatory enzymes of the aspartate family pathway. The results of these studies have also highlighted important regula- tory differences between the aspartate family pathways of higher plants and bacteria.

Analyses of Plant Mutants

In E. coli, the rate-limiting enzyme for lysine biosynthesis is AK, with DHPS playing only a secondary role. This was shown by the analysis of E. coli mutants selected for their resistance to inhibition by toxic lysine analogs such as thiosine or S-amino- ethyl I-cysteine (AEC). These analogs compete with lysine for incorporation into proteins, and only mutants that overproduce lysine survive such treatment (Bright and Shewry, 1983; Jacobs et ai., 1987). The E. colilysine-overproducing mutants selected on this basis possess a modified, lysine-insensitive form of the AK-111 isozyme (Boy and Patte, 1972). This suggests that feedback inhibition of AK by lysine is the main point of regula- tion for lysine biosynthesis in €. coli(Cohen and Saint-Girons, 1987). Because the bacterial mutants overproduce lysine but not threonine, it is likely that inhibition of HSD by threonine is the main point of regulation for threonine biosynthesis (Fig- ure 1; Cohen and Saint-Girons, 1987). Any physiological role of DHPS inhibition by lysine in E. coliis questionable, because mutants with feedback-insensitive DHPS have not been found, despite the use of strong selective agents (Cohen and Saint- Girons, 1987). This is in accord with the fact that the E. coli

AK-I11 is considerably more sensitive to lysine inhibition than is E. coli DHPS

Similar selections for AEC-resistant, lysine-overproducing plant mutants have been successful for several species, in- cluding carrot, tobacco, rice, and bulrush millet (for review, see Galili et ai., 1994, and references therein). The AEC resistance in tobacco cells resulted from the expression of a modified, less lysine-sensitive DHPS (Negrutiu et ai., 1984). This indi- cates that in plants, in contrast with E. coli, DHPS is the major rate-limiting enzyme in lysine biosynthesis and AK plays a sec- ondary role. Indeed, plant DHPSs are generally 20 to 100 times more sensitive to lysine inhibition than is E. coli DHPS.

Plants containing feedback-insensitive AK isozymes have, however, been identified by selecting for resistance to growth inhibition by high concentrations of lysine plus threonine. These two amino acids together completely inhibit the activity of ali AK isozymes, resulting in methionine starvation (Figure 1; Green and Phillips, 1974; Jacobs et ai., 1987). Mutants pos- sessing feedback-insensitive AK isozymes have been selected in a number of plant species by their resistance to growth on lysine plus threonine (for review, see Galili et ai., 1994). These mutants overproduce free threonine and, to a lesser extent, isoleucine, methionine, and, in some cases, lysine. These results show that in plants, in contrast to bacteria, AK limits mainly threonine but not lysine synthesis. Indeed, HSD activ- ity could play only a secondary role in limiting threonine synthesis in plants, because in contrast to bacteria, plants con- tain at least one HSD isozyme that is insensitive to feedback inhibition.

= 0.2 mM and 1 mM, respectively).

Analyses of Transgenic Plants

Although plant mutants have contributed to our understand- ing of the regulation of lysine and threonine synthesis, such mutants have several major limitations because of the selec- tion procedures by which they are obtained. Because the sensitivities of AK and DHPS to feedback inhibition are the major limiting factors for lysine and threonine synthesis, mu- tants isolated for resistance to lysine plus threonine or to AEC possess structurally modified enzymes rather than enzymes with altered expression or compartmentation. Such mutants are therefore not useful for understanding regulation by the levels, intracellular localization, or tissue-specific expression of the enzymes. These limitations have been overcome by ex- pressing bacterial genes encoding feedback-insensitive AK and DHPS enzymes in transgenic plants (Glassman et al., 1992; Perl et al., 1992; Shaul and Galili, 1992a, 1992b).

To create such transgenic plants, the coding sequences of the bacterial AK and DHPS genes have been fused to the con- stitutive 35s promoter from cauliflower mosaic virus. Some chimeric constructs have also included DNA encoding a plastid transit peptide to direct the bacterial enzymes into the organelle. Expression of the bacterial AK results in significant overproduc- tion of free threonine, whereas expression of the bacterial

902 The Plant Cell

DHPS causes significant overproduction of lysine (Glassman, 1992; Perl et al., 1992; Shaul and Galili, 1992a, 1992b), in agree- ment with studies of plant mutants.

In addition, the transgenic plants have led to important ad- ditional discoveries. First, they have shown that localization of DHPS in the plastid is crucial for lysine synthesis, whereas localization of AK within the organelle improves threonine production but is not essential (Shaul and Galili, 1992a, 1992b). The biological significance of this observation is not clear be- cause it is still unknown whether all plant AK isozymes are localized in plastids. It is possible that some metabolic inter- mediates of the aspartate family pathway can shuttle between the plastids and cytosol. Studies of transgenic plants have also demonstrated not only that AK and DHPS are regulated by feedback inhibition but also that the levels of these enzymes limit the rate of production of threonine and lysine. When the entire green tissue from 1-month-old plantlets derived from transgenic plants possessing various levels of the bacterial enzymes was analyzed, significant positive correlations were detected between the levels of the bacterial AK and DHPS en- zymes and the levels of free threonine and lysine, respectively (Shaul and Galili, 1992a, 1992b). Thus, the biochemical regu- lation of threonine and lysine synthesis apparently operates on both the K, and Vmax of AK and DHPS. Finally, the lysine and threonine levels of the transgenic plants vary in different tissues and at different stages of development, suggesting the presence of complex developmental regulatory signals lead- ing to differential expression of aspartate family genes.

Concerted Regulation of Lysine and Threonine Synt hesis

Transgenic plants expressing both a feedback-insensitive AK and a feedback-insensitive DHPS (Frankard et al., 1992; Shaul and Galili, 1993) contain free lysine levels that far exceed those in plants expressing only the insensitive DHPS or AK. This lysine increase is also accompanied by a significant reduction in threonine accumulation as compared with plants express- ing the insensitive AK only. These results reveal two major aspects of the regulation of the aspartate family pathway. First, the balance between lysine and threonine synthesis appears to be determined by a competition between DHPS and HSD for limiting amounts of their common substrate, 3-aspartic semi- aldehyde, whose level is apparently determined primarily by the activity of AK (see Figure 1). This agrees with the fact that 3-aspartic semialdehyde is a very unstable compound in vitro and that the level of aspartate, the precursor of the aspartate family pathway, has never been found to be limiting (Frankard et al., 1991, 1992; Perl et al., 1992; Shaul and Galili, 1992a, 19924 1993). Second, among the aspartate family amino acids in plants, lysine synthesis appears to be under the most stringent regulation because DHPS is so sensitive to lysine inhibition. When DHPS becomes deregulated, the branch lead- ing to lysine synthesis competes strongly with the other branch of the pathway, and a considerable amount of 3-aspartic

semialdehyde is thus converted into lysine at the expense of threonine.

DEVELOPMENTAL AND ENVIRONMENTAL REGULATION OF LYSINE AND THREONINE SYNTHESIS

Are Multiple AK lsozymes Redundant or Essential?

Although the aspartate family pathway has been studied mostly at the biochemical level, several lines of evidence suggest that this pathway is also subject to complex developmental and environmental regulation. For instance, AK, the first enzyme of the aspartate family pathway, apparently has major regula- tory importance for the synthesis of all of the aspartate family amino acids. Close examination of the aspartate family path- way from a biochemical viewpoint suggests that despite the major regulatory importance of AK, feedback inhibition of the lysine-sensitive AK isozymes by lysine itself is unlikely to play a major regulatory role. DHPS is much more sensitive to ly- sine inhibition than is AK, and lysine levels should therefore never accumulate enough to inhibit AK activity. If the level of lysine-sensitive AK activity is not limiting, then the threonine- sensitive AK isozyme may be biochemically redundant, as is the case in E. co/i(Cohen and Saint-Girons, 1987). On the other hand, if different AK isozymes are expressed in cell- and tissue- specific manners, then the different AK isozymes may not be redundant.

Indeed, several lines of evidence suggest that there is no absolute redundancy of AK activities in plants and that most if not all of the AK isozymes are functional and essential. First, although joint addition of lysine and threonine strongly inhibits plant growth, each of these amino acids by itself can inhibit growth when added alone at higher concentrations (Bryan, 1980). Moreover, the sensitivity of plant growth to either lysine or threonine varies markedly at different stages of develop- ment (H. Karchi, L. Yang, R. Amir, and G. Galili, unpublished results). These findings imply that both the lysine- and threonine-sensitive AKs are essential for plant growth. Sec- ond, the relative activities of lysine- and threonine-sensitive AKs change dramatically during plant development, with the lysine-sensitive activity dominating in rapidly growing cells (Bryan, 1980, 1990). In addition, the levels of AK and DHPS mRNAs are higher in young leaves than in elongating leaves (X.Z. Zhu and G. Galili, unpublished observations). Third, lysine-sensitive AK should be only weakly affected by lysine because of the extreme sensitivity of DHPS to lysine inhibi- tion (see Figure 1). This is not the case in vivo; mutations rendering plant lysine-sensitive AKs insensitive to lysine cause overproduction of threonine (see Galili et al., 1994, and refer- entes therein).

Moreover, although the lysine-sensitive AKs appear to be subjected to regulation in vivo, they still do not limit lysine syn- thesis. In fact, the mechanism of AK regulation is even more complex. So far, all of the threonine-overproducing plant

Lysine and Threonine Synthesis 903

mutants have been shown to possess a modified, lysine- insensitive form of the lysine-sensitive AK but not of the threonine-sensitive AKlHSD enzymes (see Galili et al., 1994, and references therein). The reason for this is still unknown, but it may result from different mechanisms of allosteric inhi- bition of the lysine- and threonine-sensitive AKs. Indeed, a recent report (Muehlbauer et al., 1995) suggested that the al- losteric regulation of the threonine-sensitive AKlHSD enzymes of maize (and perhaps of other plants as well) may occur by a novel, two-component mechanism in which the AK domain acts as a sensor of the threonine concentration by autophos- phorylation of a histidine residue in this domain. The phosphate is then transferred from the sensor to a specific aspartic acid residue that may be located in the HSD domain to initiate a response regulatory program. This hypothesis is based on amino acid similarities of the maize AKlHSD enzymes to other proteins regulated by an analogous two-component mechanism.

Lysine Synthesis during Plant Development 1s Not Dependent Solely on DHPS Activity

Despite the key regulatory role of DHPS, lysine production also depends on additional enzymes. Expression of a feedback- insensitive DHPS should elevate lysine synthesis only in cells in which the entire lysine biosynthetic pathway is active. The levels of free lysine in young tobacco leaves are significantly higher than those in older leaves, even though younger leaves have lower expression of the bacterial DHPS (O. Shaul and G. Galili, unpublished results). This observation is in accord with the findings that the aspartate family pathway, as a whole, is more active in rapidly growing young leaves than in old leaves (X.Z. Zhu and G. Galili, unpublished data).

Regulation of Lysine and Threonine Synthesis during Seed Development

The major amino acids transported into seeds are asparagine andlor glutamine. Seeds are thus expected to possess the en- tire enzymatic machinery needed to convert asparagine and glutamine into the other amino acids. To test whether seeds have the potential to synthesize lysine and threonine autono- mously, transgenic tobacco plants were created that express bacterial feedback-insensitive AK and DHPS under the con- trol of the seed-specific promoter from the bean phaseolin storage protein gene (Karchi et al., 1993,1994). Seed-specific expression of the bacterial feedback-insensitive AK resulted in a significant increase in free threonine in mature seeds. Simi- larly, plants expressing the bacterial feedback-insensitive DHPS in a seed-specific manner synthesized higher than nor- mal levels of free lysine during seed development. However, the leve1 of free lysine was significantly reduced in mature seeds, most likely because it is efficiently catabolized (Karchi et al., 1994). These results indicate that the aspartate family pathway is functional in seeds.

The timing of expression of the lysine and threonine bio- synthesis genes as well as the metabolism of asparagine into aspartate and aspartate family amino acids seems to occur relatively late during seed development, similar to the timing of expression of storage protein genes (Karchi et al., 1994). These observations suggest that genes encoding storage pro- teins as well as genes encoding enzymes involved in amino acid metabolism are regulated coordinately during seed de- velopment. If there is indeed a coordinated regulation of genes encoding storage proteins and metabolic enzymes, one might expect to find common promoter elements in them. Although detailed studies of the promoters of genes encoding amino acid biosynthesis enzymes are still lacking, it appears that they may contain at least some promoter elements that are similar to promoter elements of storage protein genes (see the follow- ing section).

Environmental Control of Lysine and Threonine Synthesis

The production of lysine and many other amino acids in yeast is coordinately regulated by nutritional conditions. This regu- lation is termed the general control (GCN) system. The GCN4 transcription factor, which binds to specific short conserved regions in the promoters of many genes encoding enzymes of amino acid metabolism, is a major determinant of this regu- lation (Hinnebusch, 1988). Many plant genes contain GCN4-like boxes in their promoters (Müller and Knudsen, 1993). These genes encode not only enzymes of amino acid metabolism but also storage proteins and a few enzymes that are not in- volved in amino acid metabolism. This further suggests that amino acid metabolism and storage protein synthesis are coor- dinately regulated during seed development. Interestingly, the Opaque-2 transcription factor of maize, which is an important regulator of storage protein synthesis, binds to promoters of GCN4-activated yeast genes and can also complement yeast mutants lacking GCN4 activity (Mauri et al., 1993). The func- tional significance of the GCNClike boxes in regulating gene expression in seeds and nonseed tissues is still not under- stood. It has, however, been demonstrated that the GCN4- like box might be a key element in regulating the response of a barley storage protein promoter to nitrogen (Müller and Knudsen, 1993).

LYSINE CATABOLISM

Lysine catabolism has been studied extensively in microor- ganisms, fungi, and animals. In plants, lysine catabolism has been studied mostly in seeds. In cereal and legume seeds, lysine is catabolized via saccharopine and a-aminoadipate semialdehyde, the same pathway as in mammalian cells (Fig- ure 2; Mazelis, 1980). The first reaction in this pathway is the conversion of lysine plus a-ketoglutarate into saccharopine,

904 The Plant Cell

L-1 ysine

1-LKR

Saccharopine

L-a-aminoadipate J - semialdehyde

kASD L-Pipecolic acid

L-a-aminoadipate

Figure 2. Enzymatic Catabolism of Lysine in Seed.

ASD, aminoadipic acid semialdehyde dehydrogenase; LKR, lysine ketoglutarate reductase (also known as saccharopine dehydrogenase, NADP+-lysine forming); SDH, saccharopine dehydrogenase, NAD+- glutamate forming. Reproduced from Galili et al. (1994).

which is catalyzed by the enzyme lysine ketoglutarate reduc- tase (Arruda and Da silva, 1983).

The activity of lysine ketoglutarate reductase may be sub- ject to complex developmental and environmental regulation, similar to that of seed storage proteins and enzymes of amino acid biosynthesis. Lysine ketoglutarate reductase activity in developing maize grains and tobacco seeds appears coordin- ately with the onset of storage protein synthesis (Brochetto- Braga et al., 1992; Karchi et al., 1994), and this activity is noticeably lower in developing grains of opaque-2 maize rnu- tants than in wild-type grains (Brochetto-Braga et al., 1992). lncreasing the concentration of free lysine in developing tobacco seeds, either by exogenous administration or by endogenous overproduction (via expression of the bacterial feedback-insensitive DHPS), causes an -10-fold increase in the activity of lysine ketoglutarate reductase (Karchi et al., 1994). Thus, free lysine accumulation, and possibly its incorporation into seed proteins, is apparently a highly regulated process.

FUTURE PROSPECTS

Molecular and Cellular Dissection

Cloned genes for aspartate family enzymes help in elucidat- ing the cellular and developmental regulation of expression of these genes as well as the in situ localization of their mRNAs and proteins. Several of these genes have been cloned from different plant species, among them genes that appear to encode threonine-sensitive AKlHSD enzymes and others that encode DHPS. Presumably, additional genes encoding aspar- tate family enzymes will soon be cloned. Cloning of the lysine-sensitive AK is of particular importance, because it will

provide a critical tool to address the question of the lack of redundancy of AK isozymes in plants.

It will also be important to perform functional analyses of promoters driving the expression of aspartate family genes and to identify transcription factors and other proteins that bind to these promoters. These types of studies may help in elucidat- ing general control mechanisms as well as in identifying mechanisms involved in the regulation of lysine and threonine synthesis. The available physiological data indicate that the promoters of the aspartate family genes are likely to be com- plex. These promoters are likely to contain severa1 regulatory sequences that are also present in storage protein genes, such as the Opaque-2 box. If the aspartate family enzymes in seeds and other vegetative tissues are encoded by common genes, these genes may possess additional controlling elements that are different from those of storage protein genes and that regu- late their expression in tissues other than seeds. Several putative regulatory elements have already been identified in the promoters of aspartate family enzymes by sequence ho- mology (Ghislain et al., 1994), but their functional significance has still to be determined.

The intracellular localization of the aspartate family enzymes and its regulatory consequence also need to be examined. Many of the aspartate family enzymes have been localized to plastids, but it is still not clear whether this is true for all enzymes. This question is particularly relevant for the differ- ent AK isozymes, because AK has been observed to function in both the plastid and the cytosol (Shaul and Galili, 1992b). The cloning of the entire set of AK isozymes as well as the production of isozyme-specific antisera based on amino acid sequence variations between the sequences will be required to address fully the question of AK localization.

Genetic Dissection

Genetic identification and subsequent cloning of new struc- tural and regulatory genes in the aspartate family pathway are also called for. Two genetic approaches can be used: (1) the production of auxotrophs that cannot synthesize lysine and threonine; and (2) the identification of regulatory mutants in which the synthesis and final accumulation of lysine and threo- nine are enhanced or suppressed. Auxotrophs may be obtained by employing negative complementation (Last, 1993) using cloned genes or by selecting plants that require aspartate family amino acids for growth. Selections for regulatory mutants can be performed either by classical mutagenesis or by more re- cently developed approaches such as insertion mutagenesis, transposon mutagenesis, or activation of genes by transform- ing plants with strong enhancers (Hayashi et al., 1992). However, selection for regulatory mutants of the aspartate fam- ily pathway is not simple. Because lysine and threonine synthesis is regulated primarily by feedback inhibition of DHPS and AK, mutants selected for resistance to toxic lysine and threonine analogs generally result in mutations in the coding sequences of these enzymes that render them feedback

Lysine and Threonine Synthesis 905

insensitive rather than in mutations in regulatory genes. Nevertheless, genetic dissections may be aided significantly because overproduction of lysine or threonine causes severely abnormal plant phenotypes (Frankard et al., 1992; Shaul and Galili, 1993) and lysine overproduction rendem plants extremely sensitive to growth on exogenous threonine (Perl et al., 1992; Shaul and Galili, 1992a). The latter effect of lysineoverproduc- tion results because the endogenously overproduced lysine and the exogenously added threonine significantly inhibit AK, causing starvation for methionine (Figure 1). Mutagenized plants could thus be screened for the enhancement or sup- pression of the abnormal phenotypes as well as for suppression of the threonine sensitivity. Indeed, we have recently obtained putative mutants with suppressed threonine sensitivity (H. Karchi, L. Yang, and G. Galili, unpublished data). These puta- tive mutants may result from any of severa1 defects, such as defective lysine synthesis, increased lysine catabolism, in- creased accumulation of soluble methionine, or altered lysine sequestration.

Environmental and Physiological Controls

Lysine and threonine synthesis requires carbon, nitrogen, and energy. Little is known about the interrelationships among these factors, and it will be important to study how lysine and threonine synthesis is regulated by carbon and nitrogen flows as well as by energy obtained from photosynthesis. Again, the abnormal phenotypes associated with lysine and threo- nine overproduction will be invaluable in unraveling these questions. We have recently found that transgenic lysine- overproducing tobacco plants, when grown under relatively low light intensity (50 pE m-2 sec-l), lose much of their mor- phological abnormalities and accumulate much lower lysine levels than plants grown at 200 pE m-2 sec-I. Whether this is dueto reduced energy available from photosynthesis or to light-regulated expression of aspartate family genes remains to be determined. A similar approach could be used to test how lysine and threonine synthesis is regulated by photorespi- ration and by various plant hormones.

Practical lmplications and Applications

The developments described previously have the potential to lead to the production of more nutritious transgenic crop plants. Targeting overproduction of free lysine and threonine to spe- cific tissues, such as seed, can eliminate problems associated with abnormal phenotypes and reduced yields. Moreover, clon- ing of cDNAs encoding lysine-ketoglutarate reductase or other lysine catabolic enzymes may enable the elimination of this activity in seeds, using such procedures as antisense suppres- sion. Another essential requirement will be to incorporate the increased free lysine and threonine into seed proteins; this may ensure the stable accumulation of both amino acids and minimize negative effects of high levels of these free amino acids on seed development and germination. lncorporation

of free lysine and threonine into proteins may be enhanced by transforming plants with genes encoding nove1 or modi- fied seed proteins that are relatively rich in these amino acids (Wallace et al., 1988).

ACKNOWLEDGMENTS

I thank Prof. Jonathan Gressel, Dr. Eliot Herman, and Mr. lgal Avivi for critical reading of the review and for helpful comments as well as my students Xiao-Zhu Zhu and lnbal Ben-Tzvi Tzchori for helpful dis- cussions. The work from my laboratory was supported in part by a grant from the National Council for Research and Development, Israel; the Gesellschaft Fuer Biotechnologische Forschung MBH, Braun- schweig, Germany; and the Leo and Julia Forchheimer Center for Molecular Genetics. The author is an incumbent of the Bronfman Chair of Plant Sciences.

REFERENCES

Arruda, P., and Da silva, W.J. (1983). Lysine-ketoglutarate reductase activity in developing maize endosperm. Phytochemistry 22,

Azevedo, R.A., Smith, R.J., and Lea, P.J. (1992). Aspartate kinase regulation in maize: Evidence for cepurification of threonine-sensitive aspartate kinase and homoserine dehydrogenase. Phytochemistry

Boy, E., and Patte, J.C. (1972). Multivalent repression of aspartic semi- aldehyde dehydrogenase in Escherichia coli K-12. J. Bacteriol. 112,

Bright, S.W.J., and Shewry, P.R. (1983). lmprovement of protein quality in cereals. CRC Crit. Rev. Plant Sci. 1, 49-93.

Brochetto-Braga, M., Leite, A., and Arruda, P. (1992). Partia1 purifi- cation and characterization of lysine-ketoglutarate reductase in normal and opaque-2 maize endosperms. Plant Physiol. 98,

Bryan, J.K. (1980). Synthesis of the aspartate family and branched- chain amino acids. In The Biochemistryof Plants, Vol. 5, B.J. Miflin, ed (New York: Academic Press), pp. 403-452.

Bryan, J.K. (1990). Advances in the biochemistry of amino acid bio- synthesis. In The Biochemistry of Plants, Vol. 16, B.J. Miflin and P.J. Lea, eds (New York: Academic Press), pp. 161-196.

Cohen, G.N., and Salnt-Glrons, I. (1987). Biosynthesis of threonine, lysine and methionine. In Escherichia coli and Salmonella typhimu- rium: Cellular and Molecular Biology, F.C. Neidhardt, ed (Washington, DC: American Society for Microbiology), pp. 429-444.

Dereppe, C., Bold, G., Ghisalba, O., Ebert, E., and Schar, HrP. (1992). Purification and characterization of dihydrodipicolinate synthase from pea. Plant Physiol. 98, 813-821.

Doston, S.B., Somers, D.A., and Gengenbach, B.G. (1989). Purifi- cation and characterization of lysine-sensitive aspartate kinase from maize. Plant Physiol. 91, 1602-1608.

Frankard, V., Ghislain, M., Negrutiu, I., and Jacobs, M. (1991). High threonine producer mutant of Nicotiana sylvestris (Spegazzini and Comes). Theor. Appl. Genet. 82, 273-282.

2687-2689.

31, 3371-3374.

84-92.

1139-1147.

906 The Plant Cell

Frankard, V., Ghlslain, M., and Jacobs, M. (1992). Two feedback- insensitive enzymes of the aspartate pathway in Nicotiana sylves- tris. Plant Physiol. 99, 1285-1293.

Frisch, D.A., Tommey, A.M., Gengenbach, B.G., andSomem, D.A. (1991). Direct genetic selection of a maize cDNA for dihydrodipicoli- nate synthase in an Escherichia coli da@ auxotroph. MOI. Gen. Genet. 228, 287-293.

Galili, G., Shaul, O., Perl, A., and Karchi, H. (1994). Synthesis and accumulation of the essential amino acids lysine and threonine in seeds. In Seed Development and Germination, H. Kigel and G. Galili, eds (New York: Marcel Dekker), pp. 811-831.

Ghislain, M., Frankard, V., and Jacobs, M. (1990). Dihydrodipicolinate synthase of Nicotiana sylvestris, a chloroplast-localized enzyme of the lysine pathway. Planta 180, 480486.

Ghlslain, M., Frankard, V., Vandenbossche, D., Matthews, B.F., and Jacobs, B. (1994). Molecular analysis of the aspartate kinase- homoserine dehydrogenase gene from Arabidopsis thaliana. Plant MOI. Biol. 24, 835-851.

Glassman, K.F. (1992). A molecular approach to elevating free lysine in plants. In Biosynthesis and Molecular Regulation of Amino Acids in Plants, B.K. Singh, H.E. Flores, and J.C. Shannon, eds (Rock- ville, MD: American Society of Plant Physiologists), pp. 217-228.

Green, C.E., and Philllps, R.L. (1974). Potential selection system for mutants with increased lysine, threonine and methionine in cereal crops. Crop Sci. 14, 827-830.

Hayashl, H., haja, I., Lubenaw, H., Schell, J., and Walden, R. (1992). Activation of a plant gene by T-DNA tagging: Auxin-independent growth in vitro, Science 258, 1350-1353.

Hinnebusch, A.G. (1988). Mechanisms of gene regulation in the general control of amino acid biosynthesis in Saccharomyces cerevisiae. Microbiol. Rev. 52, 248-273.

Jacobs, M., Negrutiu, I., Dirks, R., and Cammaerts, D. (1987). Se- lection programs for isolation and analysis of mutants in plant cell cultures. In Plant Biology, Vol. 3, C.E. Green, D.A. Somers, W.P. Hackett, and D.D. Biesboer, eds (New York: Alan R. Liss), pp.

Kaneko, T., Hashimoto, T., Kumpaisal, R., and Yamada, Y. (1990). Molecular cloning of wheat dihydrodipicolinate synthase. J. Biol. Chem. 265, 17451-17455.

Karchi, H., Shaul, O., and Galili, G. (1993). Seed specific expression of a bacterial desensitized aspartate kinase increases the produc- tion of seed threonine and methionine in transgenic tobacco. Plant

Karchi, H., Shaul, O., and Galili, G. (1994). Lysine synthesis and catabolism are coordinately regulated during tobacco seed devel- opment. Proc. Natl. Acad. Sci. USA 91, 2577-2581.

Kumpalsal, R., Hashimoto, T., and Yamada, Y. (1987). Purification and characterization of dihydrodipicolinate synthase from wheat sus- pension cultures. Plant Physiol. 85, 145-151.

Last, R.L. (1993). The genetics of nitrogen assimilation and amino acid biosynthesis in plants: Progress and prospects. Int. Rev. Cytol.

Lea, P.J., Wallsgrove, R.M., and Miflin, B.J. (1985). The biosynthe- sis of amino acids in plants. In Chemistry and Biochemistry of the

243-264.

J. 3, 721-727.

143, 297-330.

Matthews, B.F., Farrar, M.J., and Gray, A.C. (1989). Purification and interconversion of homoserine dehydrogenase from Daucus carota cell suspension cultures. Plant Physiol. 91, 1569-1574.

Mauri, I., Masslmo, M., Lohmer, S., Motto, M., Salamini, F., Thompson, R., and Martegani, E. (1993). Functional expression of the transcriptional activator Opaque-2 of Zea mays in transformed yeast. MOI. Gen. Genet. 241, 319-326.

Mazells, M. (1980). Amino acid metabolism. In The Biochemistry of Plants, Vol. 5, B.J. Miflin, ed (New York: Academic Press), pp. 541-567.

Muehlbauer, G.J., Somem, D.A., Matthews, B.F., and Gengenbach, B.G. (1995). Molecular genetics of the maize (Zea mays L.) aspar- tate kinase-homoserine dehydrogenase gene family. Plant Physiol.

Miiller, M., and Knudsen, S. (1993). The nitrogen response of a bar- ley C-hordein promoter is controlled by positive and negative regulation of the GCN4 and endosperm box. Plant J. 4, 343-355.

Negrutiu, I., Cattolr-Reynearts, A., Verbruggen, I., and Jacobs, M. (1984). Lysine werproducer mutants with an altered dihydropicolinate synthase from protoplast culture of Nicotiana sylvestris (Spegazzini and Comes). Theor. Appl. Genet. 68, 11-20.

Perl, A., Shaul, O., and Galili, G. (1992). Regulation of lysine synthesis in transgenic potato plants expressing a bacterial dihydrodipicolinate synthase in their chloroplasts. Plant MOI. Biol. 19, 815-823.

Rognes, S.E., Lea, P.J., and Miflin, B.J. (1980). S-Adenosyl methionine-A nove1 regulator of aspartate kinase. Nature 287, 357-359.

Rognes, SE., Bright, S.W.J., and Miflin, B.J. (1983). Feedback- insensitive aspartate kinase isoenzymes in barley mutants resis- tant to lysine plus threonine. Planta 157, 32-38.

Sanab, K., and Komamlne, A. (1978). Change in the proportion of two aspartokinases in carrot root tissue in response to in vitro cul- ture. Plant Physiol. 61, 115-118.

Shaul, O., and Galili, G. (1992a). lncreased lysine synthesis in trans- genic tobacco plants expressing a bacterial dihydrodipicolinate synthase in their chloroplasts. Plant J. 2, 203-209.

Shaul, O., and Gallli, G. (1992b). Threonine overproduction in trans- genic tobacco plants expressing a mutant desensitized aspartate kinase from Escherichia coli. Plant Physiol. 100, 1157-1163.

Shaul, O., and Galili, G. (1993). Concerted regulation of lysine and threonine synthesis in tobacco plants expressing bacterial feedback- insensitive aspartate kinase and dihydrodipicolinate synthase. Plant MOI. Biol. 23, 759-768.

Vauterin, M., and Jacobs, M. (1994). lsolation of a poplar and Arabidop sis thaliana dihydrodipicolinate synthase cDNA clone. Plant MOI. Biol. 25, 545-550.

Wallace, J.C., Galili, G., Kawata, E.E., Cuellar, R.E., Shotwell, M.A., and Larkins, B.A. (1988). Aggregation of lysine-containing zeins into protein bodies in Xenopus oocytes. Science 240, 662-664.

Weisemann, J.M., and Matthews, B.F. (1993). ldentification and ex- pression of a cDNA from Daucus carota encoding a bifunctional aspartokinase-homoserine dehydrogenase. Plant MOI. Biol. 22, 301-312.

Wilson, B.J., Gray, A.A., and Matthews, B.F. (1991). Bifunctional

106, 1303-1312.

Amino Acids, G.C. Barrett, ed (London: Chapman and Hall), pp. 197-226.

protein in carrot contains both aspartokinase and homoserine dehy- drogenase activities. Plant Physiol. 97, 1323-1328.

DOI 10.1105/tpc.7.7.899 1995;7;899-906Plant Cell

G. GaliliRegulation of Lysine and Threonine Synthesis.

 This information is current as of December 5, 2018

 

Permissions 8X

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw153229

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists