High free-methionine and decreased lignin content result from a mutation in the Arabidopsis S-adenosyl-L-methionine synthetase 3 gene


For correspondence (fax: +1 515 254 2619; e-mail: Mitchel.Tarczynski@pioneer.com).


As an approach to understand the regulation of methionine (Met) metabolism, Arabidopsis Met over-accumulating mutants were isolated based on their resistance to selection by ethionine. One mutant, mto3, accumulated remarkably high levels of free Met – more than 200-fold that observed for wild type – yet showed little or no difference in the concentrations of other protein amino-acids, such as aspartate, threonine and lysine. Mutant plants did not show any visible growth differences compared with wild type, except a slight delay in germination. Genetic analysis indicated that the mto3 phenotype was caused by a single, recessive mutation. Positional cloning of this gene revealed that it was a novel S-adenosylmethionine synthetase, SAMS3. A point mutation resulting in a single amino-acid change in the ATP binding domain of SAMS3 was determined to be responsible for the mto3 phenotype. SAMS3 gene expression and total SAMS protein were not changed in mto3; however, both total SAMS activity and S-adenosylmethionine (SAM) concentration were decreased in mto3 compared with wild type. Lignin, a major metabolic sink for SAM, was decreased by 22% in mto3 compared with wild type, presumably due to the reduced supply of SAM. These results suggest that SAMS3 has a different function(s) in one carbon metabolism relative to the other members of the SAMS gene family.


The sulfur amino-acid methionine (Met) is required in the diets of non-ruminant animals. To meet their nutritional requirements for Met, non-ruminant animals must obtain this essential amino-acid from plants either directly or indirectly. In some diets, particularly those used for agricultural livestock, Met is a limiting component for maintaining optimal animal growth. Increasing the accumulation of Met in plants could increase the nutritional quality of plants for use in animal diets. In addition to its nutritional value, Met has an essential role in cellular metabolism, as it is necessary for protein biosynthesis, and it is the precursor for the biosynthesis of S-adenosylmethionine (SAM), polyamines and the phytohormone ethylene. SAM, a direct product of Met catabolism, is a substrate in numerous transmethylation reactions, including several reactions that occur in the biosynthesis of lignin (Campbell and Sederoff, 1996).

In plants, the essential amino-acids Met, lysine, threonine and isoleucine are synthesized from a common precursor, aspartate. Met biosynthesis can be distinguished from the other aspartate-derived amino acids by a process termed transsulfuration, which involves the transfer of sulfur from cysteine to homocysteine via the intermediate cystathionine (Giovanelli et al., 1978). Cystathionine is formed from the condensation of O-phosphohomoserine and cysteine, catalyzed by cystathionine gamma synthase (CgS). CgS, which is the first enzyme unique to Met biosynthesis, is positioned at a metabolic branchpoint between Met and threonine biosynthesis. Cystathionine beta lyase (CbL) and methionine synthase (MS), respectively, catalyze the remaining two steps unique to Met biosynthesis.

Although relatively little is known about the regulation of Met metabolism, its cellular concentration is tightly controlled (Datko et al., 1978). Lysine and threonine biosynthesis have been shown to be co-ordinately regulated (Shaul and Galili, 1993). Similar regulation can be inferred for Met, as the biosynthesis of this sulfur amino-acid is intrinsically linked to the other aspartate-derived amino-acids, and multiple pathways converge to provide precursors for its synthesis. Several studies have indicated that Met may have a role in regulating its own biosynthesis. For example, Thompson et al. (1982) found that CgS activity in the alga Lemna was decreased to 15% of control levels by the addition of Met to the culture medium. Later, Giovanelli et al. (1985) concluded that Met feedback regulates its own de novo biosynthesis at the step of cystathionine synthesis. More recent studies have indicated that CgS is regulated by Met or one of its metabolites at the level of mRNA stability. Chiba et al. (1999) isolated mutants of Arabidopsis and showed that discrete point mutations in exon 1 of CgS resulted in increased stability of CgS mRNA, which resulted in increased free Met accumulation.

To identify additional regulatory steps that control Met metabolism, we used a genetic approach to isolate Met over-accumulation mutants (mto) based on selection by ethionine tolerance. Ethionine is a toxic analog of Met and may compete with this sulfur amino-acid for protein and SAM synthesis. Mutants that are resistant to ethionine have been isolated from yeast and higher plants, and several have been reported to accumulate high free-Met (Bartlem et al., 2000; Cherest et al., 1973; Gonzales et al., 1984; Inaba et al., 1994; Madison and Thompson, 1988). Two Arabidopsis ethionine-resistant mutants, mto1 and mto2 have been previously isolated and characterized. Mutations in the genes that code for CgS and threonine synthase (TS) were determined to be responsible for the mto1 and mto2 phenotypes, respectively (Bartlem et al., 2000; Chiba et al., 1999). While the exact mechanism(s) of ethionine tolerance is not fully understood (Alix, 1982), an increase in the free Met concentration within cells may lead to more dilute ethionine concentrations, and reduce the possibility of ethionine being used in essential cellular processes.

In this report, we describe the isolation and characterization of a third Arabidopsis ethionine-resistant mutant, mto3. This mutant dramatically over-accumulates free Met; however, growth and development of mto3 appear to be normal. The gene responsible for the high free-Met content of mto3 was isolated by map-based cloning and determined to code for a novel S-adenosylmethionine synthetase.


Isolation and phenotypic characterization of mto3 plants

Ethionine is a toxic analog of Met, and it likely impedes plant growth by its incorporation into protein and SAM in place of Met (Alix, 1982). Thus, cells or tissues exposed to ethionine that overaccumulate free Met may dilute their endogenous ethionine concentration, allowing for increased resistance to the analog. We screened approx. 100 000 EMS-mutagenized M2 seeds and isolated several ethionine-resistant mutants. One mutant, mto3, was selfed and propagated to obtain a homozygous line, which was further characterized. At strong, growth-inhibiting concentrations of ethionine (100 µm), wild-type seed could germinate and partially expand cotyledons, but root and leaf growth were completely inhibited. Seeds from the mto3 mutant, however, were able to germinate and develop normally in the presence of 100 µm ethionine, producing new root and shoot growth (Figure 1a). Culture of mto3 plants in soil showed no consistent, visible growth differences when compared with wild type (Figure 1b), except a slight delay in germination.

Figure 1.

Ethionine selection and growth of mto3.

(a) Germination and seedling growth of mto3 and wild type on medium containing 100 µm ethionine.

(b) mto3 and wild-type plants following germination and 4-week culture in soil.

The concentration of free Met in young leaves of mto3 plants was increased more than 200-fold, whereas the related amino acids aspartate/asparagine, threonine, and lysine, showed less than a two-fold change (Table 1). All other protein amino-acids were relatively unchanged (i.e. two-fold or less; data not shown), indicating that the mto3 mutation is relatively specific for Met. In addition to leaves, analyzes of root and stem material harvested from young mto3 plants showed that these tissues also accumulated high concentrations of free Met (data not shown). Interestingly, the free Met concentration in stem tissue was more than five-fold higher than that for leaves. Cystathionine, a precursor amino acid for Met biosynthesis, was increased by approximately 10-fold (Table 1).

Table 1.  Concentration of free amino-acids in the aerial portion of mto3 plants harvested prior to bolting
Amino acidConcentration nmol/g (FWT)Fold
Wild typemto3
  1. Three independent experiments were conducted containing the aerial portions of more than 100 plants each. Values are averages ± SEs.

Asp/Asn308 ± 44307 ± 411.0
Glu/Gln2422 ± 2702056 ± 1940.9
Lys9 ± 210 ± 31.1
Thr434 ± 66607 ± 161.4
Met15 ± 83655 ± 720244
Ile54 ± 1195 ± 81.8
Pro325 ± 40669 ± 302.1
Cystathionine0.9 ± 0.08.8 ± 2.29.8

Genetic analysis and map-based cloning of the mto3 locus

Reciprocal crosses were made between mto3 and Col-0. All F1 plants were sensitive to ethionine and contained wild-type levels of free Met. F2 plants showed a 3 : 1 segregation of ethionine-sensitive-to-ethionine-resistant plants, indicating that mto3 is a single-gene, recessive mutation. To confirm that ethionine resistance and high free-Met are caused by the same genetic mutation, free Met was determined in eight ethionine-resistant F2 plants. All ethionine-resistant F2 plants contained high free-Met. The mto3 mutation was then mapped to chromosome 3, approximately 5 cm from marker nga162 and 4 cm from marker atDMC1 (Figure 2a). Because the mto1 and mto2 genes were mapped previously to the upper part of chromosome 3 and chromosome 4, respectively (Bartlem et al., 2000; Inaba et al., 1994), and because the mto3 gene did not show co-segregation with the two previously identified SAMS genes, SAMS1 and SAMS2 (Peleman et al., 1989a; Peleman et al., 1989b), we concluded that mto3 was a novel mutant. Thus, we proceeded to clone the gene responsible for the mto3 phenotype by map-based cloning.

Figure 2.

Map-based cloning of mto3.

(a) Physical mapping of the mto3 mutation. The mto3 mutation was mapped to a 80-kb region covered by three Bac clones, MGD8, MTO12, and MKP6, which were sequenced by the Kazusa sequencing group. This 80 kb region contains six open reading frames including SAMS3 as indicated (arrow).

(b) Sequence alignment of region near the mto3 mutation. Various SAMS genes were aligned by using the Clustal program; the region near the ATP binding domain is shown. Amino-acid residues in boldface type are the consensus residues for the ATP binding domain. An arrow indicates the amino-acid change in the ATP binding domain resulting from a point mutation in mto3. Protein sequences were obtained from GenBank and their accession numbers are: humanSAMS, human SAMSCAA48726; yeastSAMS, yeast SAMS1 p10659; E. coliSAMS, E. coli SAMSAAA24164; tomatoSAMS, tomato SAMS1Z24741; Arabidopsis SAMS, aSAMS1p M33217; aSAMS2p M55077; aSAMS3p AB022216. MaizeSAMS, maize SAMS1p, maize SAMS2p, maize SAMS3p, and maize SAMS4p are four SAMS genes identified from a maize EST database (Pioneer Hi-Bred International, Inc., Johnston, IA, USA).

Using 750 ethionine-resistant F2 plants, we mapped the mto3 gene to an 80-kb region on chromosome 3, which contains six open reading frames encoding an unknown protein, a kinesin-like protein, a new S-adenosylmethionine synthetase, a pto kinase interactor, a DNA binding protein, and a ribosomal protein (Figure 2a). The new SAMS gene, termed SAMS3, is 91% and 89% identical in deduced amino-acid sequence to the previously identified SAMS1 and SAMS2 genes, respectively. We focused on further characterization of SAMS3, assuming a role in Met metabolism. SAMS3 showed complete co-segregation with mto3 in more than 1400 meiosis events, indicating strong genetic linkage between SAMS3 and the mto3 phenotype. To determine if SAMS3 contained a lesion that could be responsible for the mto3 phenotype, we compared SAMS3 sequences from wild type and mto3 (Figure 2b). The sams3 gene from mto3 showed a single point mutation that resulted in an amino-acid change from alanine to threonine. This alanine residue is a highly conserved residue in the ATP binding domains of all SAMS. To confirm that this single mutation in sams3 leads to the mto3 high-Met phenotype, we isolated a SAMS3 genomic fragment, including promoter and terminator elements, by PCR from Col-0, and transformed mto3 with this fragment. Free Met was then measured in four herbicide-resistant transformants. All four transformed, T1 mto3 plants showed wild-type levels of free Met, demonstrating that wild-type SAMS3 can completely complement the mto3 mutation (Table 2). In the selfed, T2 generation, segregation of ethionine-resistant plants to ethionine-sensitive plants was approx. 1 : 3. In addition, we verified that wild-type SAMS3 was introduced in the transformed mto3 lines by using a CAPS marker developed from SAMS3 (data not shown). We concluded that this single point mutation in the SAMS3 gene is responsible for the high free-Met phenotype observed for mto3.

Table 2.  Complementation of the mto3 phenotype by wild-type SAMS3 gene
 Free Met
content in leaf
(nmol/g FWT)
Segregation of Ethionine
Resistance in T2 generation
Plant lineResistant
  1. Met was measured in transgenic T1 plants. Ethionine resistance was determined in transgenic T2 plants. T2 seeds were plated on MS medium in the presence of 100 µm ethionine.

Wild type
F1 plant
Transgenic #1
Transgenic #2
Transgenic #3
Transgenic #4

SAMS activity, SAM concentration, and SAMS3 gene expression

SAMS is encoded by a small gene family. For Arabidopsis, SAMS1 and SAMS2 have been characterized previously, and SAMS3 has been identified in this research. An additional SAMS gene, SAMS4 (GenBank accession number AC006922), has been recently identified by sequence homology following the completion of the Arabidopsis genome sequence. To determine if the identified lesion in sams3 can affect total SAMS activity and SAM concentration, we measured total SAMS protein (Figure 4a), activity and SAM (Figure 3) from the aerial portions of more than 100 pooled plants each for wild type and mto3. Total SAMS protein concentration was not altered in mto3 compared with wild type. However, total SAMS activity and SAM content was decreased by approx. 30% and 35%, respectively, in this mutant. These data indicate that the point mutation resulting in a single amino-acid change from alanine to threonine in SAMS3 negatively affects SAMS3 catalysis, and that this alteration results in a decrease in the concentration of SAM.

Figure 4.

SAMS3 gene expression.

(a) Western analysis of total SAMS protein. Ten micrograms of protein each from wild-type and mto3 vegetative tissue was separated by SDS-PAGE. A maize SAMS2 polyclonal antibody was used to detect Arabidopsis total SAMS protein in leaf tissue.

(b) Total SAMS gene expression. Ten micrograms of total RNA for each tissue was separated by formamide gel. The signal was detected following membrane hybridization with 32P-labled SAMS3 coding sequence. SAMS3 coding sequence has at least 81% identity to the coding sequences of SAMS1, SAMS2, and SAMS4.

(c) 32P-labeled SAMS3 3′ untranslated region was used for SAMS3 gene expression. Sequence alignments showed that the selected probe has less than 50% identity with comparable regions of the other Arabidopsis SAMS genes.

(d) 18S rRNA probe used as loading control.

Figure 3.

SAMS activity and SAM concentration in wild type and mto3.

(a) SAMS activity. Wild-type plants and mto3 plants were cultured in flats containing soil under identical conditions. The aerial portion of more than 100 plants for each genotype was sampled and pooled prior to bolting. SAMS was assayed by monitoring the incorporation of 35S label into SAM. For each genotype, values are the average of three replicates + SE. SAMS activity is expressed as nanomoles SAM produced per milligram protein per minute.

(b) SAM concentration. Plant material prepared for determining SAMS activity was also used for SAM determinations. SAM was quantified by HPLC. For each genotype, values are the average of two replicates + SE.

To understand why SAMS1 and SAMS2 do not compensate for the loss of SAMS3 function, we determined SAMS3 gene expression in different tissues (Figure 4). Expression of SAMS3 was strong in stem and leaf tissues. SAMS3 gene expression patterns were not substantially different than those for all SAMS genes, and their expression patterns did not clearly differ from those determined for wild type.

Lignin content

SAM is a methyl donor for several reactions in lignin biosynthesis. Highly lignified tissues such as stem tissue might be expected to have increased levels of SAMS. mRNA analyzes showed that the various SAMS genes are expressed strongly in stem tissue (Figure 4). To determine if decreased SAM concentration affects lignin accumulation, Klason lignin was measured in two individual experiments for wild-type plants and mto3 plants. For each experiment, the aerial portions of more than 100 wild-type plants and 100 mto3 plants were collected 2 weeks after flowering and pooled separately. The Klason lignin content for mto3 plants was decreased compared with wild type in both experiments (wild type, 8.44% DWT and 7.04% DWT; mto3, 6.15% DWT and 5.97% DWT), with an average loss equal to 21.7% (Figure 5).

Figure 5.

Lignin content in wild type and mto3.

Wild-type plants and mto3 plants were cultured in flats containing soil under identical conditions. The aerial portion of more than 100 plants for each genotype was sampled and pooled 2 weeks after flowering. For each genotype, values are the average of two independent experiments + SE. Klason lignin was determined as described in Experimental procedures.


SAMS catalyzes the transfer of the adenosyl moiety of ATP to the sulfur atom of Met to form SAM, requiring the complete dephosphorylation of ATP. SAM, the major methyl group donor for numerous transmethylation reactions in all living cells, is a precursor for the biosynthesis of polyamines and the phytohormone ethylene, and is involved in the regulation of metabolism. SAMS genomic and cDNA clones have been isolated from a wide variety of species (Espartero et al., 1994; Izhaki et al., 1996; Kim et al., 1995; Markham et al., 1984; Peleman et al., 1989a; Schroder et al., 1997; Thomas and Surdin-Kerjan, 1987). SAMS genes comprise a small gene-family, and, at least within a species, appear to be highly conserved. However, the specific function(s) of the individual SAMS genes within a gene family is not well understood.

In Arabidopsis, four SAMS genes have been identified by analysis of the sequenced genome. Two SAMS genes, SAMS1 and SAMS2, have been previously characterized, and they are 97% identical at the amino-acid level. RNA profiling of these genes showed that they were similarly expressed relative to one another, but that the magnitude of expression was tissue-dependent (Peleman et al., 1989a; Peleman et al., 1989b). For example, both SAMS genes were highly expressed in stem tissues but relatively weakly expressed in pod tissues. In addition, experiments with the reporter gene β-glucuronidase showed that expression of SAMS1 5′ sequences conferred a similar expression pattern, with highest expression in vascular tissues (Peleman et al., 1989a; Peleman et al., 1989b). In transgenic experiments, simultaneous overexpression and co-suppression of the SAMS1 gene from Arabidopsis in tobacco resulted in complex developmental phenotypes (Boerjan et al., 1994). In tissues where co-suppression occured, SAMS activity was strongly reduced, free Met accumulated to high levels, and the plants showed abnormal growth patterns, such as stunting and dark green sectors that were associated with veins.

In this work, we identified a novel SAMS gene, SAMS3. A single amino-acid change in the ATP binding domain of SAMS3 resulted in a strong reduction of SAMS activity and SAM concentration and more than a 200-fold increase in free Met. However, in contrast to the results observed for SAMS co-suppression, the mto3 plants showed normal growth and development, indicating that SAMS3 has a different function in vivo than the other SAMS genes. In addition, from unpublished experiments with maize, we have isolated insertional inactivation lines of two individual SAMS genes by transposable Mutator-element- tagging. Characterization of plants from these lines showed that, despite gene inactivation, Met did not accumulate and growth appeared relatively unaffected. However, similar to the co-suppression results of tobacco, SAMS co-suppression experiments with maize did result in high levels of free Met accumulation in vegetative tissues, and the plants were severely abnormal in their growth and development (C. Li et al., unpublished).

Further evidence for different roles of SAMS genes can be obtained from disruption experiments with yeast. In this organism, two SAMS genes have been identified, and, like SAMS genes from other species, they are highly conserved (92% similar in amino-acid sequence) (Thomas and Surdin-Kerjan, 1987; Thomas et al., 1988). Analyzes of mutant strains with individual disruptions in SAMS1 and SAMS2 showed distinct phenotypic differences; for example, a strain disrupted in SAMS1 was resistant to ethionine, whereas a strain disrupted in SAMS2 was sensitive to this Met analog (Thomas and Surdin-Kerjan, 1987; Thomas and Surdin-Kerjan, 1991).

Arabidopsis SAMS3 protein has approx. 90% identity with SAMS1, SAMS2 and SAMS4 (GenBank accession number AC006922). This raises the question of why endogenous SAMS isoforms can not compensate for the loss of SAMS3function. Several explanations are possible. First, SAMS3 could be differentially expressed in tissues or cells relative to the other SAMS genes. However, at the level of our RNA analyzes, we did not find evidence to support this possibility, as SAMS3 gene expression showed no clear tissue specificity relative to total SAMS expression. The spatial expression patterns observed for total SAMS in our studies are generally consistent with those reported for the individual Arabidopsis SAMS1 and SAMS2 genes (Peleman et al., 1989a; Peleman et al., 1989b). Second, SAMS3 could be differentially expressed during plant growth and development. While our work does not address this possibility, differential expression of individual SAMS genes during growth and development has been reported (Espartero et al., 1994; Gomez-Gomez and Carrasco, 1998; Schroder et al., 1997). For example, in Pisum sativum, the SAMS2 gene was weakly expressed in nearly all tissues examined, but showed its highest level of expression at the apex. In addition, following pollination of this same species, SAMS1 was specifically up-regulated in ovaries, whereas SAMS2 was expressed constitutively (Gomez-Gomez and Carrasco, 1998). A critical analysis of the expression patterns for the individual SAMS genes from various tissues during growth and development of mto3 together with SAMS activity determinations should provide further insight regarding this possibility. Third, SAMS3 could have different physicochemical properties from those of other SAMS. Schroder et al. (1997) determined various properties of three SAMS isoenzymes from Catharanthus roseus following recombinant expression in bacteria. No significant differences were detected in (i) optima for temperature and pH (ii) Km for substrates (iii) inhibition by reaction products or ethionine, and (iv) native protein size. These authors suggested the possibility that the different SAMS isoforms reflect specificities in the association with enzymes that use SAM. Thus, SAMS3 may not be differentially regulated at transcriptional or kinetic levels, but, rather, it may form an association with another enzyme(s) to have a specific effect on metabolism. While speculative, one possibility consistent with our data is that SAMS3 may associate with a methyltransferase involved in lignin biosynthesis. Several methylation reactions occur during the biosynthesis of lignin, and highly lignified tissues such as stem must consume large amounts of SAM. Metabolite channeling could provide a more efficient mechanism for the methylation of lignin monomers. Peleman et al. (1989a) suggested a role for the Arabidopsis SAMS1 gene in lignin biosynthesis based on its strong expression in vascular tissues. SAMS3 is also highly expressed in vascular tissue, and, in mto3, the lesion in SAMS3 resulted in a reduction of the lignin content in these plants, demonstrating a role for this enzyme in lignin biosynthesis. As an initial test of this possibility, we expressed the Arabidopsis SAMS2 gene under the control of a constitutive promoter (modified CaMV 35S promoter) in mto3. Unlike complementation of the mto3 phenotype by wild-type SAMS3 and its regulatory sequences (Table 2), analyzes of 10 transgenic plants containing the SAMS2 construct showed no complementation of the mto3 phenotype (unpublished data). While supportive, additional experiments addressing metabolite channeling specifically are needed to evaluate this possibility.

The accumulation of free Met observed in mto3 is likely the result of a decrease in flux from Met to SAM due to the alteration in SAMS3. However, the regulation of Met biosynthesis in higher plants is not well understood, and other possible regulatory mechanisms may contribute to Met accumulation in mto3 (Leustek et al., 2000 and Ravanel et al., 1998). Early work in Lemna indicated that CgS is a potential regulatory enzyme for Met synthesis, and that Met concentrations may have a role in regulating this enzyme (Giovanelli et al., 1985; Thompson et al., 1982). More recently, CgS in Arabidopsis was shown to be regulated at the level of mRNA stability, possibly by Met or one of its metabolites (Chiba et al., 1999). In yeast, the regulation of sulfur amino-acid biosynthesis has been studied extensively. In this microorganism, SAM has been identified as a key metabolite that at elevated levels negatively affects the transcription of several Met-related pathway genes. For example, in strains with disrupted SAMS1 and SAMS2 genes, exogenously applied Met can not repress the transcription of Met-related pathway genes, indicating that elevated SAM concentration is essential for repression of Met biosynthesis (Thomas and Surdin-Kerjan, 1997). Met synthesis in enteric bacteria has also been shown to be negatively regulated by SAM (Saint-Girons et al., 1988). However, whether similar regulation occurs in higher plants is not clear (Chiba et al., 1999; Giovanelli et al., 1985; Ravanel et al., 1998; Thompson et al., 1982). In the SAMS co-suppression experiments of Boerjan et al. (1994) and in mto3, the free Met concentrations were markedly increased. Thus, Met itself can not be a key regulator of its own synthesis. Northern analyzes of steady-state RNA for various genes associated with Met metabolism in mto3 showed no clear negative regulation of these genes in response to elevated Met concentrations (data not shown). However, several genes encoding proteins involved in the synthesis of Met including cysteine synthase1, CgS and MS were modestly up-regulated (approx. two-fold) in mto3. While speculative, given the decreased level of SAM in this mutant, these results are consistent with the possibility that SAM is a key regulator of Met synthesis, perhaps affecting the expression of several genes. The accumulation of very high concentrations of free Met in mto3 may result in part from the effect of decreased SAM concentration on genes involved in Met synthesis; for example, by increasing the stability of CgS mRNA (Chiba et al., 1999).

Plant TS, in contrast to bacterial threonine synthase, is strongly stimulated by SAM in vitro, and it has been implicated in regulating the flux between Met and threonine synthesis (Curien et al., 1998; Madison and Thompson, 1976). However, in mto1 where the CgS gene was affected, free Met increased 40-fold and SAM increased three-fold, whereas the free threonine level was similar to wild type (Bartlem et al., 2000; Inaba et al., 1994). In mto3, where free Met increased more than 200-fold and SAM decreased by 35%, free threonine was, again, not affected. These results are intriguing, as they imply that SAM may not lead to activation of TS in vivo. Clearly, however, other factors must be considered, for example, the spatial separation of cytoplasmically synthesized SAM and plastid-localized TS. Additional experiments with mto3 should continue to provide new insights regarding the regulation of methionine metabolism.

Experimental procedures

Plant materials and culture

Arabidopsis thaliana (L.) Heynh. ecotype Wassilewskija (Ws) was used as wild type. Columbia (Col-0) was used in mapping experiments and, in some experiments, as an additional control. Ethyl methanesulfonate (EMS)-mutagenized M2 seeds of Ws harboring a gl1 mutation were purchased from Lehle Seeds (Round Rock, TX, USA). Seeds were germinated either on Gamborg's B5 medium or soil (GIBCO BRL, Grand Island, NY, USA). Following germination, plants transferred to or germinated directly on soil were housed in a Conviron growth chamber at 25°C/20°C (day/night) with a photoperiod of 16 h light/8 h darkness. All plants were analyzed at the 3- to-5-leaf stage of development prior to bolting unless otherwise stated.

Mutant isolation

Mutants were isolated essentially as described by Inaba et al. (1994). Briefly, surface-sterilized seeds were cultured aseptically on Gamborg's B5 medium containing 0.8% agar and 100 µm DL-ethionine. Seedlings that produced strong root growth and expanded four leaves in the presence of ethionine were collected and transferred to soil. These plants were grown to maturity and self-pollinated to produce M3 seeds. M3 seeds were then similarly re-screened for ethionine resistance.

Genetic analysis, mapping, and cloning of the mto3 gene

Homozygous mto3 plants (ecotype Ws) were crossed to Col-0. F2 seeds were obtained by self-pollination of the F1 plants. F2 seeds were grown on Gamborg's medium containing 100 µm ethionine. Ethionine-resistant F2 plants were transferred to soil and used for mapping. DNA was prepared from parental, F1, and 750 ethionine-resistant F2 plants. The mto3 gene was mapped to chromosome 3 by using amplified simple sequence length polymorphisms (SSLPs; Bell and Ecker, 1994) and cleaved amplified polymorphic sequences (CAPS; Konieczny and Ausubel, 1993) as markers. To refine further the map position of the mto3 gene, CAPS markers were developed by sequencing PCR-cloned genomic DNA from Ws near the mto3 locus. We developed several CAPS markers on BAC clones, MGD8, MKP6, and MEB5, based on sequence differences between the two ecotypes that resulted in restriction site changes. A SAMS3 gene-specific CAPS marker used oligo primers phn33707 5′ GCTCAGGAGTCTCATCAGTAG and phn33708 5′CTGTATTAGGTTTCTTTCGTG with restriction enzyme Bsa BI.

Complementation of mto3

A 3.8-kb SAMS3 genomic fragment including promoter and terminator was isolated by PCR from wild-type plants (Col-0). Oligo primers used for PCR were phn35281 5′ CACCGCAAACTCAATCACATAC and phn41924 5′CAGCCGTCTATTTTGGTTTTTC. The genomic fragment was then cloned into a binary vector (PHP12943; Pioneer Hi-Bred International, Inc., Johnston, IA, USA) containing a CaMV 35S promoter and Bialaphos resistance gene (BAR) as a selectable marker. The construct was introduced into Agrobacterial strain LBA4404 by electroporation and was used to transform mto3 by vacuum infiltration (Clough and Bent, 1998). Free Met was determined on leaf samples collected from herbicide resistant transformants.

Free amino-acid and lignin determinations

Wild-type plants and mto3 plants were cultured in flats containing soil under identical conditions. For free amino-acid analyses, more than 100 plants each were sampled for various tissues prior to bolting. Three independent experiments were conducted. The collected samples were immediately ground in liquid nitrogen and stored at −80°C until assayed. Free amino-acids were extracted from tissues with 80% (v/v) ethanol at 42°C for 10 min. Samples were centrifuged at 10 000 g for 10 min at RT. Precipitates were twice re-extracted as before. The supernatants were pooled and vacuum dried. Free amino-acids were oxidized in 200 µl performic acid (nine parts 88% formic acid, one part 30% hydrogen peroxide), vortexed, and incubated at RT for 60 min. The samples were then vacuum dried at RT, and the oxidized amino-acids were resuspended in 200 µl of amino-acid sample dilution buffer (Beckman Instruments, Inc., Fullerton, CA, USA; Cat no. 727410). Free amino-acids were analyzed using a Beckman 6300 high-performance ion-exchange analyzer. Separation was performed using a 4-mm-diameter × 120 mm hydrolysate column. Amino acids were eluted by Na+ gradient and were detected with ninhydrin. For lignin analyzes, plants were cultured and pooled as for amino-acid analyzes, except that only the aerial portion was harvested 2 weeks after flowering. Two independent experiments were conducted. Plant material was freeze-dried, ground to pass a 1-mm sieve, and extracted with boiling neutral detergent (Van Soest et al., 1991) using filter bags in a batch fiber analyzer (ANKOM, Fairport, NY, USA). The residual neutral detergent fiber (NDF), a pectin-free, cell-wall preparation, was oven dried (55°C) and used for quantification of Klason lignin according to Kaar et al. (1991), modified to accommodate analysis of 100 mg NDF.

SAMS activity and SAM determinations

The aerial portion of more than 100 plants cultured as described for amino-acid analyzes was pooled and used for SAMS activity and SAM measurements. For SAMS activity determinations, soluble protein was extracted in triplicate by homogenizing approx. 250 mg each of frozen tissue in 0.5 ml extraction buffer (100 mm Tris (pH 7.5), 2 mm EDTA, 20% glycerol, 20 mmβ-mercaptoethanol, 1 mm DTT) at 4°C. After centrifugation at 10 000 g for 10 min, the protein concentration of the supernatant was determined using a reagent supplied by Pierce Chemical Co., (Rockford, IL, USA) (Cat no. 1856210). Total SAMS activity was assayed essentially as described by Boerjan et al. (1994). Briefly, approx. 60 µg of extracted protein was incubated in 0.25 ml of a reaction mixture containing 100 mm Tris (pH 8.0), 30 mm MgSO4, 10 mm KCl, 20 mm ATP, and 5 mm35S-Met (15 µCi). Control reactions contained all reagents except for ATP. Reactions were incubated for 1 h at 25°C and were terminated by placement on a mixture of ice and water. Thirty microliters of the reaction mixture was then spotted on a phosphocellulose filter (Whatman International, Clifton, NJ, UAS) in triplicate. The filter was air-dried, and washed three times with ice-cold water for 5 min each at RT. The washed filter was then transferred to scintillation vials containing 1 ml of 1.5 m ammonium hydroxide. After 5 min, scintillation liquid was added, and the sample was counted by scintillation spectrometry. Plant material used for SAMS activity determinations was also used to determine SAM content. SAM was determined in duplicate essentially as described by Wise and Fullerton (1995), with the following modifications to the extraction procedure. Approx. 200 mg of plant material was homogenized in 0.4 ml of 0.1 m NaOAc (pH 6.0) at 4°C. Protein was then precipitated by the addition of 0.4 ml of 30% (w/v) TCA. Samples were mixed vigorously and incubated on ice for 30 min. Precipitated protein was pelleted by centrifugation at 10 000 g for 10 min. Supernatants were collected and immediately frozen in liquid nitrogen and stored at −80°C until analysis.

Western and Northern blots

Plants were cultured and tissues were harvested as described for amino-acid analyzes. For Western analyzes, protein was extracted with SDS sample buffer (50 mm TrisCl (pH 7.5), 1% SDS, 50 mm DTT) and quantified using a reagent supplied by Pierce Chemical Co. (Cat no. 1856210). Protein samples were separated by 10% SDS-PAGE, Tris-glycine gel and transferred to a membrane according to standard protocols (Sambrook et al., 1989). A maize SAMS polyclonal antibody was used to detect total SAMS protein. Signals detection was by ECL Western Blotting Detection Reagent (Amersham Pharmacia Biotech, Piscataway, NJ, USA; Code No. RPN2106). For Northern analyzes, total RNA was isolated from leaf and stem tissues using Trizol Reagent according to the manufacturer's instructions (GibcoBRL, Gaithersburg, MD, USA; Cat no. 15596–018). Leaf tissue was harvested prior to bolting, and stem tissue was harvested and pooled 2 weeks after flowering. All samples were immediately frozen in liquid nitrogen and stored at −80°C until analyzes. RNA was isolated for each tissue from 1 g of frozen sample. RNA electrophoresis, membrane transfer, and hybridization were performed according to standard protocols (Sambrook et al., 1989). For analyzes of total SAMS gene expression, SAMS3 coding sequence was amplified using primers (GAAACTAAAGAGGCAGAAAGAG and GCTCTGTATTAGGTTTCTTTCG) and used as a probe. SAMS3 coding sequence showed at least 81% identity to the coding sequences of SAMS1, SAMS2, and SAMS4. Membranes were washed with 2XSSC and 0.1% SDS at 65°C. For SAMS3 gene-specific expression, a SAMS3 gene-specific probe was amplified from genomic DNA by PCR using primers (CTGCTGCCTATGGTCACTTTGG and CAAACAAAGGAGACTAATGTCG). Sequence alignments showed that this probe had less than 50% identity with comparable regions of the other Arabidopsis SAMS genes. Membranes were washed with 0.1XSSC and 0.1% SDS at 65°C. As a loading control, the membranes were also hybridized with an 18S rRNA probe (Nairn and Ferl, 1988). The 32P-labeled probes were prepared by random primer labeling using a rediPrimeII kit (Amersham Pharmacia Biotech).


We thank Vincent Sewalt for analysis of Klason lignin content, Carolyn Wise for analysis of SAM concentration, Elena Rus for amino-acid analysis, and Thomas Leustek for manuscript comments.