* This work was supported by Grants MCB-9982895 and IBN-0211421 from the National Science Foundation and CRIS no. 5335-21430-005-00D from the USDA/ARS. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § Current address: Protein Chemistry and Engineering, Faculty of Agriculture, Kyushu University, Fukuoka 812-8581, Japan.
1-Amino-cyclopropane-1-carboxylate synthase (ACS, EC 184.108.40.206) is the key enzyme in the ethylene biosynthetic pathway in plants. The completion of the Arabidopsis genome sequence revealed the presence of twelve putative ACS genes, ACS1–12, dispersed among five chromosomes. ACS1–5 have been previously characterized. However, ACS1 is enzymatically inactive whereas ACS3 is a pseudogene. Complementation analysis with the Escherichia coli aminotransferase mutant DL39 shows that ACS10 and 12 encode aminotransferases. The remaining eight genes are authentic ACS genes and together with ACS1 constitute the Arabidopsis ACS gene family. All genes, except ACS3, are transcriptionally active and differentially expressed during Arabidopsis growth and development. IAA induces all ACS genes, except ACS7 and ACS9; CHX enhances the expression of all functional ACS genes. The ACS genes were expressed in E. coli, purified to homogeneity by affinity chromatography, and biochemically characterized. The quality of the recombinant proteins was verified by N-terminal amino acid sequence and MALDI-TOF mass spectrometry. The analysis shows that all ACS isozymes function as dimers and have an optimum pH, ranging between 7.3 and 8.2. Their Km values for AdoMet range from 8.3 to 45 μm, whereas their kcat values vary from 0.19 to 4.82 s–1 per monomer. Their Ki values for AVG and sinefungin vary from 0.019 to 0.80 μm and 0.15 to 12 μm, respectively. The results indicate that the Arabidopsis ACS isozymes are biochemically distinct. It is proposed that biochemically diverse ACS isozymes function in unique cellular environments for the biosynthesis of C2H4, permitting the signaling molecule to exert its unique effects in a tissue- or cell-specific fashion.
The gas ethylene has been known since the beginning of the century to be used by plants as a signaling molecule for regulating a variety of developmental processes and stress responses (
). These include seed germination, leaf and flower senescence, fruit ripening, cell elongation, nodulation, wounding and pathogen responses. Ethylene production is induced by a variety of external factors, including wounding, viral infection, elicitors, auxin treatment, and Li+ ions (
). The structures show that the enzyme is a homodimer, and its overall fold and active sites are similar to those of aminotransferases even though the two enzymes have completely different catalytic activities. The tertiary structures together with available biochemical data explain the catalytic roles of the conserved and non-conserved active site residues (
). The question immediately arises as to why there are so many ACS isozymes for synthesizing ethylene in Arabidopsis. It has been postulated that the presence of ACS isozymes may reflect tissue-specific expression that satisfies the biochemical environment of the cells or tissues in which each isozyme is expressed (
). For example, a group of cells or tissues with low concentration of the ACS substrate, AdoMet, would express a high affinity (low Km) ACS isozyme. Accordingly, the tissue-specific expression of distinct ACS isozymes, defined by their unique biochemical properties, would allow C2H4 to be made in a tissue-specific manner to mediate various biological responses. Herein, we report the biochemical characterization of the Arabidopsis ACS gene family members as a first step toward understanding the role of multigene families in plants in general and of ACS in particular.
pET22, pET28, pET32 vectors, and thrombin protease (restriction grade) were purchased from Novagen (Darmstadt, Germany). pBAD-HisA vector was from Invitrogen (Carlsbad, CA), and Escherichia coli BL21-CodonPlus™(DE3)RIL strain was from Stratagene (La Jolla, CA). Superdex™-200 was from Amersham Biosciences. pQE80 vector and Ni-NTA agarose were from Qiagen (Valencia, CA). Complete™-EDTA-free was purchased from Roche Applied Science (Indianapolis, IN). The AdoMet used in the enzyme assay (contains 59% of (S,S)diastereomer form) was purchased from Roche Applied Science. Sinefungin and AVG were purchased from Sigma. All other chemicals used for biochemical analysis were of analytical grade. Oligonucleotides were purchased from Operon Technologies, Inc. (Alameda, CA).
Standard protocols were followed for DNA manipulations (
). DNA sequencing was used to confirm that no spurious mutations were introduced during mutagenesis.
Expression of ACS
(a) Plant Material—A. thaliana ecotype Columbia seeds were surfaced sterilized for 8 min in 5% sodium hypochlorite, 0.15% Tween 20, excessively rinsed in distilled water, and plated on Petri dishes onto sterilized filter paper discs on top of 0.8% agar (Select Agar, Invitrogen) containing 0.5× Murashige-Skoog salts (Invitrogen), 0.5 mm MES, pH 5.7, 1% sucrose, 1× vitamin B5. The plates were incubated in the dark at 25 °C for 7 days after cold treatment at 4 °C for 2 days. Intact etiolated seedlings were removed and placed in Petri dishes containing 0.5× Murashige-Skoog salts solution buffered at pH 5.7 with 0.5 mm MES and supplemented with 20 μm IAA or 50 μm CHX. The seedlings were incubated for 2 h in the dark at room temperature with shaking (100 rev/min). Mock controls were incubated with an equal amount of solvent (ethanol) used to prepare the stock solution of the chemicals. Five g-fresh weight of seedlings were removed, briefly blotted dry, frozen in liquid nitrogen, and stored at –80 °C. Three-week-old light-grown plants were grown as described in the Arabidospsis Biological Resources Center (ABRC) manual. Roots, leaves, stems, flowers, and siliques were collected and quick-frozen in liquid nitrogen and stored at –80 °C.
RT-PCR—Total RNA was isolated from 7-day old etiolated seedlings, treated with or without 20 μm IAA or 50 μm CHX, and from various parts of light grown plants, using the RNeasy kit (Qiagen) or with the TRIzol reagent (Invitrogen). Genomic DNA was removed by treating 1–2 mg of total RNA with 10 units of RNase-free DNase I (Roche Applied Science) by incubating for 15 min at 37 °C in 10 μlof1× DNase I buffer (50 mm Tris-HCl, pH 7.5, 10 mm MgCl2) containing 40 units of RNase inhibitor. 5 μg of DNase treated RNA primed with oligo (dT)24 primer were used for first strand synthesis with SuperScriptII reverse transcriptase (Invitrogen) according to the manufacturer's instructions.
PCR was performed in the GeneAmp PCR System 9700 (Applied Biosystems). The total reaction volume was 25 μl containing 2.5 μl of 1st strand cDNA, 1 μl of primer solution (10 μm), 2.5 μl of 10× PCR buffer, 2 μl dNTP mix, and 0.2 μl ExTaq enzyme (Takara). The reaction mixtures were subjected to the following PCR conditions: 94 °C for 2 min, one cycle; 94 °C for 15 s, 62 °C for 15 s, and 72 °C for 3 min for 35 cycles; followed by one cycle of incubation at 72 °C for 3 min. The primers used are: ACS1 F, 5′-TGTCTCAGGGTGCATGTGAGAATCAACTT-3′ and R, 5′-AGCTCGAAGCAATGGTGAATGAGGAGACA-3′; ACS2 F, 5′-GCGACTAACAATCAACACGGAG-3′ and R, 5′-ACATTATCCCTGGAGACGAGAGAC-3′; ACS4 F, 5′-CCAAGTCTCTTCGTATTTCCTT-3′ and R, 5′-TAGTCGGAAAACCCAGTTAGAGAC-3′; ACS5 F, 5′-CCAGCTATGTTTCGATCTAATCGAGTCATGGTTAAC-3′ and R, 5′-TCCATGAAACCCGGAAAACCCAGTTAGAGACTGTC-3′; ACS6 F, 5′-TGACGGTCACGGCGAGAATTCCTCTTATT-3′ and R, 5′-CCTGAGGTTACTCTGCCAACACTTCTTCT-3′; ACS7 F, 5′-AACAACAACAACGTCGAGCTTTCTCGAGT-3′ and R, 5′-AGATCCCGGAGATATATTCAGGTTCAGCT-3′; ACS8 F, 5′-CGATCTCATTGAGTCATGGCTTGCTAAGA-3′ and R, 5′-ACGGTCCATCAACGAACCTCTTCAATCTA-3′; ACS9 F, 5′-GGATGGGAAGAATACGAGAAGAACCC-3′ and R, 5′-ATCACTCTTCTACTATCTGTTGACTC-3′; ACS10 F, 5′-AGGGTTATTGTTCCGTTACAAGGTGTGGT-3′ and R, 5′-TACGGAACCATCCTGGTTCGATACAGTGA-3′; ACS11 F, 5′-AACCTCGGCTAACGAGACTCTAATGTTCT-3′ and R, 5′-ATGACACGATGAGCCTGGAGAGATGTTAA-3′; ACS12 F, 5′-TTCGTCGGCTCTCTCATTCCTTGTTGTCT-3′ and R, 5′-ACCACCACTCTCAGCACATGGTATTCCTA-3′.
For ACT8, the primers used were the same as described in Ref.
Full-length open reading frame clones (ORFs) for all the putative ACS genes were constructed by: 1) PCR from preexisting full-length cDNAs (ACS1 U26543; ACS2 M95595; ACS4 U23481; ACS5 L29261), using the following gene-specific primers: ACS1: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGTCTCAGGGTGCATGTGAG-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTTAAGCTCGAAGCAATGGTGA; ACS2: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGGGTCTTCCGGGAAAAAAT-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTCATGCTCGGAGAAGAGGTGA-3′); ACS4: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGGTTCAATTGTCAAGAAAA-3′) and R,(5′-AGTCGACGGCCCATGAGGCCCTATCGTTCCTCAGCCTCACG-3′); ACS5: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGAAACAGCTTTCGACAAAA-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTCATCGTTCATCAGGTACACG-3′).
2) RT-PCR poly(A)+ RNA from CHX-treated etiolated seedlings was purified by oligo(dT) cellulose (Qiagen) following the manufacturer's instructions. First-strand cDNA was synthesized from 5 μg of poly(A)+ RNA with oligo dT18 as a primer and 1000 units of SuperScript II RT enzyme (Invitrogen) as previously described (
). First-strand cDNA was re-suspended into 50 μlofdH2O after phenol/chloroform extraction followed by EtOH precipitation. Using gene-specific primers, 0.5 μl of 1st-strand cDNA and 3.75 units of Expand High Fidelity DNA polymerase (Roche Applied Science), cDNAs were amplified using the following conditions. After initial denaturation for 2 min at 94 °C, touch-down PCR (
) was performed, consisting of 12 cycles of 15 s denaturation at 94 °C, 30 s annealing at 63 °C with negative ramp 6 °C at 0.5 °C per cycles and 3 min at 68 °C, followed by 18 cycles of 15 s at 94 °C,30sat57 °C and 3 min at 68 °C. A final elongation was undertaken at 68 °C for 7 min. Amplified cDNAs were purified with the Qiagen PCR purification kit.
The gene-specific primers used were: ACS6: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGGTGGCTTTTGCAACAGAG-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTTAAGTCTGTGCACGGACTAG-3′); ACS7: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGGGTCTTCCTCTAATGATG-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTCAAAACCTCCTTCGTCGGTC-3′); ACS8: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGGGTCTCTTGTCAAAGAAA-3′) and R, (5′-AGTCGACGGCCCATGAGGCCCTATCGTTCCTCGGGTTCACG-3′); ACS9: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGAAACAACTGTCGAGAAAA-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTCATCGTTCATCAGGTACACG-3′); ACS10: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGACCCGTACCGAACCAAAC-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTCAATTTTGAGATTTACATGT-3′); ACS11: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGTTGTCAAGCAAAGTTGTT-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTCAACGTTCTGATTCACAAGT-3′); ACS12: F, (5′-GAATTCGGCCGTCAAGGCCAGAAGGAGATATAACCATGAGGTTGATAGTACCTCTC-3′) and R, (5′-AGTCGACGGCCCATGAGGCCTCAAGATCTAAATGATTCAGC-3′).
The ORFs were cloned into the pUNI51 vector and sequenced. pUNI51 was derived from pUNI50 (
) by introducing a new polylinker containing two SfiI sites (A and B) that allow cloning of a full-length ORF unidirectionally by having the SfiI A and B sites at the 5′- and 3′-ends of the ORF, respectively. The sequences of all eleven putative ACS ORFs are error-free and can be transferred into any desirable expression vector by in vitro cre/lox recombination (
). Fig. 1A shows the overall strategy for constructing the expression vector, pBAD-Trx-His for achieving this goal. The NdeI site present in front of the thioredoxin ORF was changed to the NcoI site by site-directed mutagenesis using the primer P1, 5′-CTTTAAGAAGGAGATATACCATGGGAAGCGATAAAATTATTCACC-3′ (underlined shows the NcoI restriction site) giving rise to pET321. The NdeI site present in front of the His tag was deleted and replaced with the MscI-BglII linker, 5′-CCATCATCATCATCATCACTCTTCTGGA-3′ (underlined shows the His6 tag sequence) giving rise to pET322. A thrombin protease site and a NdeI site were introduced in front of the multicloning site (MCS) of pET32 with a KpnI-BamHI linker, 5′-CGGTGGTGGCTCCGGTCTGGTGCCACGCGGT AGTCATATGGATATCG-3′ (bold and underlined show thrombin and NdeI sites, respectively), replacing the KpnI/BamHI sites of pET322, giving rise to pET323.
The pBAD-HisA vector was modified as follows: the NdeI site present downstream of the origin of replication was deleted by site-directed mutagenesis using the primer P2, 5′-CGGTATTTCACACCGCATATCGTGCACTCTCAGTACAATC-3′ (bold shows mutated base). The other NdeI site present in the MCS of pBAD-HisA was deleted by BglII/HindIII digestion and filling in with T4 DNA polymerase. An XbaI site was introduced between the pBAD promoter and the ribosome-binding site (rbs) by site-directed mutagenesis using the primer P3, 5′-CATACCCGTTTTTTTGGTCTAGAAGGAGGAATTAACCATG-3′ (underlined shows the XbaI site). Finally, the XbaI-XhoI fragment of pET323 containing the rbs, Trx tag, His tag, thrombin site, and MCS was subcloned into the XbaI/XhoI sites of the modified pBAD-HisA giving rise to the pBAD-Trx-His vector used in this study.
Expression in E. coli—The ACS ORFs were subcloned into pBAD-Trx-His and pET22 vectors as NdeI/SacI fragments (ACS2, -4, -5, -6, -8, -9, and -12) or as NdeI/NotI fragments (ACS7, -10, and -11) (
) giving rise to pBAD-ACS2–12 and pET22-ACS2–12 plasmids, respectively. The pBAD-Trx-His expression system produces His-tagged proteins that can be purified by affinity chromatography. The pET22 expression system (
) produces native proteins that can be purified by conventional purification procedures (non-affinity chromatography). We decided to use the pBAD expression system because it gave the same levels of enzyme activities as the pET22 system (see Table I) while allowing protein purification by affinity chromatography. Attempts to purify the ACS isozymes using the IMPACT I (Intein mediated purification with an affinity chitin-binding tag; New England Biolabs) expression system were unsuccessful.
Table IActivity of the Arabidopsis ACS isozymes in E. coli
E. coli transformants harboring the expression vectors pET22-ACS2–12 and pBAD-ACS2–12 were cultured in LB and RM media (1× M9 salts-2% Casamino acids-1 mm MgCl2) plus 0.2% glucose, respectively, with 150 μg/ml ampicillin at 37 °C in a buffered flask with constant shaking at 300 rpm, until the cell cultures reached an OD600 of 0.8. Isopropyl-1-thio-β-d-galactopyranoside (pET plasmid) or l-(+)-arabinose (pBAD plasmid) was added to each culture to a final concentration of 1 mm or 0.2% (w/v), respectively, and the cell cultures were then allowed to grow for an additional 4 h at 30 °C. Cells were pelleted by centrifugation at 3,000 × g for 15 min, washed once in half of the original volume with 10 mm Tris-HCl buffer, pH 8.0, containing 0.1 m NaCl, and stored at –70 °C. For large scale protein preparation, a single colony containing pBAD-ACS plasmid was first inoculated into 100 ml of RM plus glucose medium, incubated until OD600 reached 1.0, then transferred into 3 to 6 liters of the same medium containing 150 μg/ml ampicillin, and induced under the same conditions. Cells were harvested by centrifugation at 1,500 × g for 20 min at 4 °C.
Assay for ACS Activity
ACS activity was assayed in a reaction mixture (total volume, 0.4 ml) containing 100 mm HEPES buffer, pH 8.0–10 μm PLP-120 μm AdoMet and an appropriate enzyme fraction. The mixture was incubated at 30 °C for 20 min, and the reaction was terminated with 50 μl of 10 mm HgCl2. The amount of ACC formed was determined according to (
). One unit of enzyme activity converts 1 μmol of AdoMet to ACC per hour at 30 °C.
The overall protein purification scheme is shown in Fig. 1B. Frozen cell pellets were resuspended in 1:20 of the original volume in extraction buffer containing 50 mm potassium phosphate, pH 8.0, 10 μm PLP, 0.1 m NaCl, 1 mm PMSF, 2 mm EDTA, 2 mm NDSB201–100 μl/ml lysozyme, and proteinase inhibitor Complete™ (1 tablet per 50 ml of buffer). The cells were lysed by sonication. An aliquot of the cell extract was analyzed on SDS-PAGE. After centrifugation at 20,000 × g for 30 min, the supernatant was used as crude ACS solution. Twenty percent of the streptomycin sulfate solution was added to the crude ACS solution to the final concentration of 1.5% to precipitate genome DNA. The precipitate was removed by centrifugation. The supernatant was precipitated with ammonium sulfate (100% saturation), and the resulting precipitate was collected by centrifugation in a small volume of 50 mm potassium phosphate buffer, pH 8.0, 10 μm PLP, 1 mm PMSF, 0.5 m NaCl, 2 mm NDSB201-Complete™ and desalted on a Sephadex-25 column (2.5 × 30 cm) equilibrated with the same buffer. Imidazole solution, 1 m, pH 8.0, was added to the fraction with ACS activity, to a final concentration of 10 mm, and applied to Ni-NTA agarose column (1 × 3 cm) equilibrated with the same buffer. Weakly bound proteins were washed from the resin with 50 mm potassium phosphate buffer, pH 8.0, 10 μm PLP, 1 mm PMSF, 0.5 m NaCl, 20 mm imidazole, 2 mm NDSB201-Complete™. The His6-tagged protein was eluted with 50 mm potassium phosphate buffer, pH 8.0, 10 μm PLP, 1 mm PMSF, 0.5 m NaCl, 250 mm imidazole, 2 mm NDSB201-Complete™. Eluted fractions with ACS activity were buffer-exchanged on a Sephadex-25 m in 20 mm Tris-HCl buffer, pH 8.0, 10 μm PLP, 150 mm NaCl, 2 mm NDSB201–2.5 mm CaCl2. Subsequently, thrombin protease (1 unit/mg ACS) was added to the thioredoxin-His6-tag-ACS solution and digested at 20 °C for 16 h. The digest was precipitated with ammonium sulfate and the resulting precipitate was dissolved in a small volume of 50 mm potassium phosphate buffer, pH 8.0, containing 10 μm PLP, 0.1 mm PMSF, 0.3 m NaCl, 2 mm NDSB201, and fractionated on a Superdex™-200 column (1 × 30 cm) equilibrated with the same buffer at a flow rate of 0.4 ml/min using an Amersham Biosciences FPLC system.
Protein Determination and N-terminal Sequence Analysis
Protein concentration was determined with the bicinchoninic acid method (
) using marker proteins: β-galactosidase (116 kDa), phosphorylase b (97 kDa), fructose-6-phosphate kinase (84 kDa), bovine serum albumin (66 kDa), glutamic dehydrogenase (55 kDa), ovalbumin (45 kDa), and glyceraldehyde-3-phosphate dehydrogenase (36 kDa). Superdex™-200 (1 × 30 cm) column was calibrated with following marker proteins (Sigma): apoferritin (443 kDa), α-amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (67 kDa), and carbonic anhydrase (29 kDa). MALDI-TOF mass analysis was carried out on a Bruker Reflecton II MALDI-TOF mass spectrometer. Samples were prepared by mixing one part of protein sample with nine parts of matrix (saturated sinapinic acid or cyano-4-hydroxycinnamic acid in 0.1% trifluoroacetic acid, 30% acetonitrile), and 1 μl of this solution was applied onto the sample probe. External calibration was carried out using two ion peaks from bovine serum albumin: [M + 2H]2+, 33,216; [M + H]+, 66,432.
Determination of Kinetic Parameters for ACS
ACC synthase activity was measured at different AdoMet concentrations ranging from 3 to 200 μm, in a reaction mixture containing 100 mm buffer-10 μm PLP-1 mm dithiothreitol. The buffers used were: pH 5.0–5.5, sodium acetate buffer; pH 6–7, MES buffer; pH 7–8.5, HEPES buffer; pH 8.5–9.0, EPPS buffer; pH 10–11, glycine-NaOH buffer. The amount of protein used for each reaction was ACS2, 77 ng; ACS4, 80 ng; ACS5, 21 ng; ACS6, 9.8 ng; ACS7, 210 ng; ACS8, 10 ng; ACS9, 9.8 ng; and ACS11, 87 ng, respectively. Km and Ki measurements were carried out at the optimum pH of each isoenzyme. The Km value for AdoMet was calculated from Lineweaver-Burk plots. Each ACS was incubated with various concentrations of AVG (
), and the rate of ACC formation was determined. The Ki values were determined using Dixon plots.
The ACS ORFs were subcloned downstream of the T5 promoter in the His tag-containing pQE80 vector. ACS1,2,4,5,6,8,9,12 were subcloned as NdeIblunt/SacI fragments into BamHIblunt/SacI-digested vectors. ACS10 was subcloned as a NdeIblunt/NotIblunt fragment into BamHIblunt/SalIblunt-digested pQE80. ACS11 was subcloned as NdeIblunt/SspIblunt into BamHIblunt/SacIblunt pQE80 vector. Each expressed ACS protein contains a His tag at its N terminus. The AATase cDNA, pAAT37 (
). E. coli DL39 was transformed with the pQE-ACS plasmids and the transformants were plated on M9 medium plus various amino acids and incubated at 30 °C for 2 days, and photographed. The transformants were also grown in liquid M9 medium in the presence of various amino acids at 30 °C for 24 h, and the absorbance at 600 nm was determined.
). Phylogenetic analysis reveals that the genes fall into three branches (Fig. 2B). ACS10 and 12 are phylogenetically related to alanine and aspartate ATases, respectively (Fig. 2B).
The polypeptides of the ACS gene family are quite similar in size, ranging between ∼50.9 (ACS7) to ∼61.2 kDa (ACS10) (Table I). ACS3 is not included in the analysis because it is a truncated polypeptide (
). In addition, the eleven conserved residues between ACS and ATases are also present in the Arabidopsis ACS gene family members, except for the tyrosine (Y) residue in box 2, which is part of the PLP-binding site (
), and the glycine (G) residue in box 3, which has been replaced by serine (S) and phenylalanine (F), respectively, in ACS10 and ACS12 (Fig. 3, compare residues with open circles). The conserved glutamate (E) residue in box 1, which is responsible for substrate specificity (
), is present in all the members of the ACS gene family, except ACS10 where it has been replaced by glutamine (Q) (Fig. 3). This indicates that this E residue is not the sole determinant for ACS substrate specificity because ACS12 has the E47 residue but is lacking ACS activity. Furthermore, the serine (S) residue in the hypervariable C terminus, which is a phosphorylation site (
), is present in all ACS isozymes except ACS7, ACS10, and ACS12, because all three have a truncated C terminus compared with the other ACS isozymes (Fig. 3). ACS10 and 12 have longer N termini than all the other ACS isozymes (Fig. 3). The amino acid and nucleotide sequence identity among the various members of the ACS gene family varies from 32 to 91% and 34 to 84%, respectively, indicating that the ACS gene family is quite divergent (Table II).
Table IIPercent amino acid and nucleotide sequence identity of the Arabidopsis ACS gene family members
The authenticity of the various ACS isozymes was verified by expression experiments in E. coli. Full-length ORFs were subcloned into pET22 and pBAD-Trx-His vectors, and ACS activity was determined as described under “Experimental Procedures.” Table I shows that all ACS isozymes are enzymatically active, except for ACS1, ACS10, and ACS12. The inactivity of ACS1 is due to the absence of the tripeptide PSN in the conserved box 4 (Fig. 3 and Ref.
). The inactivity of ACS10 and ACS12 was quite puzzling because both proteins have most of the hallmarks of the authentic ACS isozymes. We entertained two possibilities: (a) ACS10 and 12 are pseudogenes like ACS1 or (b) they encode another enzymatic activity, preferably aminotransferases, a close relative to ACS (
). The first possibility was tested by constructing hybrid or modified ACS10 and 12 proteins to determine whether they become enzymatically active. We tested whether the presence of their N-terminal extension or the absence of their C-terminal peptides are responsible for their inactivity. We constructed a N-terminal-truncated ACS12 (N-ACS12; Arg66–Ser495, 430 amino acid residues) and N-ACS12 plus the C-terminal peptide of ACS8 (C8; Thr434–Arg469), N-ACS12+C8 (466 amino acid residues). Both hybrid proteins (N-ACS12 and N-ACS12+C8ACS8) were expressed in E. coli, and their ACS activity was nil. We also constructed and tested mutants of ACS12, N-ACS12, and N-ACS12+C8 by altering their F92Y and box 2 sequence from YKPFEGL to YQDYHGL (underlined amino acids show the amino acids changed by site-directed mutagenesis to the same residues as in other ACS isozymes). The activity of all six mutants was nil. A clue to the possible function of ACS10 and 12 was provided by the phylogenetic tree (Fig. 3). Their phylogenetic resemblance with the alanine and aspartate ATases raised the prospect they may encode ATases (Fig. 1B). This possibility was tested by complementation experiments of the E. coli ATase mutant DL39 (
). The results of Fig. 4 show that ACS10 and ACS12 are ATases with broad specificity for aspartate and aromatic amino acids such as tyrosine and phenylalanine. Their activity for branched chain amino acids is nil. All functional ACS isozymes do not have ATase activity (Fig. 4). The data of Fig. 4 have also been confirmed by growth of each transformant in liquid media supplemented with various amino acids (data not shown). Accordingly, the Arabidopsis ACS gene family consists of eight functional (ACS2, 4–9, and 11) and one non-functional (ACS1) members.
Expression of the ACS Genes—Expression studies using RT-PCR show that IAA induces six ACS genes in 7-day-old etiolated seedlings (compare lanes 1 and 2 with lane 3 in Fig. 5A). It has been previously reported that ACS5 is not auxin inducible (
). The possibility exists, however, that the induction of ACS5 by auxin is transient and detectable only by short auxin treatment. CHX treatment of etiolated seedlings strongly enhances mRNA accumulation of all ACS genes (compare lanes 1 and 2 with lane 4 in Fig. 5A). The expression of ACS1 is nil in the presence of auxin or CHX (compare lanes 1 and 2 with lanes 3 or 4 in Fig. 5A). RT-PCR analysis of mRNAs from various parts of light grown seedlings reveals differential expression among the various ACS genes during plant growth and development (Fig. 5). Specifically, all ACS genes except ACS1 and ACS4 are expressed in the roots (lane 1, Fig. 5B). ACS1 and ACS5 are expressed in leaves, ACS6 in flowers, and ACS1, -2, -5, -8, -9, and -11 in siliques (Fig. 5B).
Expression and Purification of the ACS Isozymes—A number of E. coli expression systems were tested prior to deciding on the use of the pBAD-Trx-His expression vector (see “Experimental Procedures”). The expressed proteins are N-terminal fusions of ACS with thioredoxin and His6. The 16-kDa thioredoxin tag enhances the solubility of the fusion protein and allows easy identification of its removal after thrombin digestion by SDS-PAGE (size reduction of the digested fusion polypeptide). Fig. 6A shows a Coomassie-stained gel of total protein extracts from E. coli strains expressing various ACS proteins. All ACS isozymes are efficiently expressed and their molecular sizes range from 60–65 kDa indicating that they are indeed fusions of thioredoxin and His6. (Fig. 6A; compare lane V E. coli extract to those expressing the various ACS isozymes). The expression level of each fusion protein is approximately the same (Fig. 6A; compare the intensity of the bands marked with a solid circle). The majority of the fusion proteins were localized in the insoluble fraction of the total E. coli extracts. Among the various ACS isozymes, some are more soluble than others. The order of decreased solubility is: ACS7 > 6 > 2 > 4 > 8 > 9 > 5 > 11. The various ACS isozymes were purified from the soluble fraction by Ni-affinity chromatography. Subsequently, the Trx-His-ACS were digested with thrombin protease and purified by filtration chromatography on Superdex™-200. Fig. 7 shows the elution profiles of the various ACS isozymes. ACS2, -6, -7, and -9 are eluted as a single peak whereas ACS4, -8, -5, and -11 have a broad elution profile (two or more active peaks). We collected the last active peak with a retention time of 33–35 min for further analysis. ACS activity eluted at this retention time corresponds to a molecular size of ∼100 kDa (dimer) as determined using known sized proteins. Fig. 6B shows that the purified ACS isozymes consist of a single band on a SDS-PAGE and stained by Coomassie Blue. The yield of the various ACS isozymes varies from 0.03 to 17 mg per 3 liters of E. coli culture (Table IV). Table III summarizes the results of the ACS isozyme purification. The purification scheme developed allows purification of the ACS isozymes to homogeneity with 10–600-fold purification depending on the ACS isozyme. Thrombin digestion yields a polypeptide with three additional amino acids, GSH, at the N terminus.
Table IVEnzymatic properties of the Arabidopsis ACS isozymes
Molecular Mass and Subunit Structure—The molecular size of the eight functional ACS isozymes, estimated by SDS-PAGE, ranges from 45 to 50 kDa, whereas their sizes determined by gel filtration on a Superdex™-200 column ranges from 93 to 105 kDa (Table IV), indicating these functional ACS proteins are dimers. Since the size of the purified proteins determined by SDS-PAGE is smaller than the predicted size (see Table IV), the prospect is raised that the purified ACS isozymes have undergone partial proteolysis during purification. This may be due to nonspecific proteolytic degradation by thrombin or during purification. The authenticity of the N termini was determined by sequencing the N terminus of the purified polypeptides. The results presented in Table IV indicate that all ACS isozymes have the correct and expected N-terminal sequence. The possibility of a proteolytic degradation at the C termini was determined by mass spectrometry using a MALDI-TOF mass spectrometer as described under “Experimental Procedures.” The molecular size of each ACS determined by this method is shown in Table IV. The data show that the size of ACS4 and ACS7 is the same as that predicted. However, the size of ACS2, -5, -6, -8, -9, and -11 was smaller by 2–5 kDa than the predicted size for each protein, suggesting a cleavage at the C terminus. Fig. 8 shows the putative cleavage sites at the C termini that occurred during purification. For example, the experimentally determined size of 51,695 Da for ACS2 corresponds to an ACS2 polypeptide that arises from GSH-(Met1–Lys460). The cleavage of the Lys460–Lys461 bond in the intact ACS2 protein results in a 51,642 Da protein very similar in size to the experimentally determined value of 51,695 Da. Using similar calculations, we determined that purified ACS5, -6, -8, -9, and -11 arise from the cleavages of Lys450–Lys451, Glu441–Glu442, Arg436–Ser437, Arg444–Ile445, and Arg439–Arg440 bonds, respectively (Fig. 8). These results suggest that the purified ACS isozymes, except ACS4 and -7, are cleaved at the C-terminal region of the molecule during purification despite the presence of the protease inhibitors, Complete™ and PMSF, during the purification procedure.
Enzymatic Properties of ACS Isozymes—Table IV shows the enzymatic properties of the Arabidopsis ACS isozymes. The pH optima vary from 7.3 to 8.2. These values are smaller than those previously reported for various ACS isozymes purified from E. coli or plant tissues (compare pH values in Table IV with Table V). The Km values range from 8.3 to 45 μm whereas the Vmax ranges from 13 to 324 μmol of ACC/mg of protein/h at 30 °C. The inhibitor constants for 2 known inhibitors, AVG and sinefungin, vary from 0.019–0.80 μm and 0.15–12 μm, respectively. AVG is a more effective inhibitor of ACS activity than sinefungin.
Table VEnzymatic properties of ACSes from various plant species: a comparison
). The biological significance of multigene families in general and of the ACS gene family in particular is unknown. While the various ACS isozymes catalyze the same biochemical reaction, it is not known whether their biological function(s) are distinct or overlapping. Genetic evidence (
) support the view that each member of the ACS gene family may have a distinct biological function. It has been postulated that tissue specific expression of a particular ACS isozyme satisfies the biochemical environment of the cells and tissues in which each isozyme is expressed (
). Accordingly, the distinct biological function of each isozyme is defined by its biochemical properties. Such a concept enhances the physiological fine-tuning of the cell and demands that the enzymatic properties of each isozyme be distinct (
The availability of the Arabidopsis genome sequence provided the opportunity to experimentally test this proposition with a complete ACS family. The Arabidopsis ACS family consists of nine isozymes. Eight of them are enzymatically active and one, ACS1, is enzymatically inactive (
The genome also contains three annotated genes, ACS3, ACS10, and ACS12, with great resemblance to the various members of the ACS gene family. ACS3 is a pseudogene representing a truncated version of ACS1 (
). ACS10 and ACS12 encode aminotransferases (Table I).
The biochemical characterization of the ACS isozymes requires an expression system that provides high levels of active enzymes. We carried out a number of preliminary experiments for determining the best E. coli expression system to be used. We expressed the ACS8 protein using the pET28 (His6-tagged ACS8) and pET32 (Trx-His6-tagged ACS8) expression vectors, both containing the T7 promoter (
). The enzyme activities of the crude extracts were 56 and 40 nmol/100 μl/h at 30 °C for pET28-ACS8 and pET32-ACS8, respectively. These activities were one-fourth to one-fifth of those obtained with the pET22 (native ACS8) and the pBAD (Trx-His6-tagged ACS8) expression vectors containing the T7 and pBAD promoter, respectively (Table I). Accordingly, we used the pBAD instead of the T7 promoter containing expression vector because it allows protein purification by affinity chromatography. The enzyme activities recovered with the eight ACS isozymes varied greatly ranging from 0.75% (ACS4) to 26.2% (ACS6) (Table III). This was due to the solubility differences among the various isozymes and not due to differential stability of the isozymes. The yield of ACS6, the most soluble member, was 35 times higher than that of ACS4 (Table III). This was quite evident after digestion of Trx-His6-tagged ACS4, -5, and -11 with thrombin protease, which led to a large amount of insoluble proteins. In addition gel filtration of the digested ACS4, -5, and -11 gave many peaks with enzyme activity on a Superdex™-200 column (Fig. 7). Thrombin digestion resulted in the precise removal of the N-terminal polypeptide (thioredoxin) used for affinity purification, as determined by amino acid sequence of the purified enzymes (Table IV). In addition mass spectrometry was used to determine whether the purified proteins were intact. All purified ACS isozymes were truncated at the C terminus except for ACS4 and -7 (Fig. 8). The putative proteolytic cleavage sites were in the hypervariable C-terminal region defined after the conserved arginine (R) residue among all ACS isozymes in box 7 (Fig. 3). The C termini contain many amino acids with positive (Arg and Lys) and negative (Asp and Glu) charges. These amino acids are potential targets of endogenous trypsin- and acid protease-like proteases despite the presence of protease inhibitors, Complete™ and PMSF, during the purification procedure. Also the possibility exists that the excess amount of thrombin protease used resulted in secondary cleavage of arginine residues at the C terminus. ACS4 and ACS7 were purified as intact polypeptides because ACS4 does not have proteolytic susceptible sites like the remaining ACS isozymes and ACS7 is missing the C-terminal hypervariable region (Fig. 3).
The variable region of the C terminus is the domain responsible for phosphorylation (
). This suggests that the enzyme properties determined in this study with C-terminal-truncated proteins may represent the values of the intact proteins. Phosphorylation of the C terminus regulates the protein stability of the enzyme mediated by the ETO1 gene product. Phosphorylated ACS is more stable than non-phosphorylated ACS because it prevents protein-protein interaction between ACS and ETO1 (
). Furthermore, mutations in the C terminus such as eto2 and eto3 enhance the stability of ACS5 and ACS9, respectively, by preventing the interaction of ETO1 with the ACS protein. It remains to be determined whether ACS7 is also post-transcriptionally regulated because the regulatory C-domain responsible for protein stability is missing.
The enzymatic properties of the ACS isozymes support Rottman's proposition regarding the multiplicity of ACS isozymes (
). They are distinct when compared within the ACS gene family and within those from other plant species (Tables IV and V). The ACS isozymes are not only biochemically divergent but also their genes have a divergent pattern of expression and response to the plant hormone auxin (Fig. 5). Promoter GUS fusions with the members of the ACS gene family reveal a spectacular and highly divergent pattern of expression throughout the life cycle of Arabidopsis.2 Yeast microarray data show that expression divergence is the major reason for maintaining duplicated genes in a genome. More importantly, a large number of duplicated genes have diverged quickly in expression, and the vast majority of gene pairs eventually become divergent in expression (
). The remainder are unable to provide ACC in the ripening fruit because they are expressed elsewhere, restricted by their tissue specific expression. Determination of the AdoMet concentration in single cells or group of cells using laser capture microdissection (
) put forward a similar concept regarding multilocus enzymes in man: “Although in general the different isozymes which make up these various multilocus sets are very similar to one another in their catalytic functions, differences in their kinetics, their inhibition characteristics and other properties such as stability have been noted in quite a number of cases. It is difficult, in viewing the striking tissue differentiations that occur, not to conclude that the detailed enzymic properties of the different isozymes of the set have been tailored in the course of evolution so they are appropriate for the specific metabolic roles they subserve in the particular intracellular environment of the tissues in which they are found. However, in most cases the exact nature of such presumed functional differences has not been clearly defined and this is a major task for the future.”
We thank Barbara J. Bachman (Yale University) for providing the DL39 E. coli strain and Steve Elledge (Baylor College of Medicine) for the pUNI50 vector. We also thank Alan Smith and Richard C. Winant of the Stanford University Protein and Nucleic Acid Facility for determining the N-terminal sequence of the ACS isozymes. The ATase cDNA pAAT37 was provided by Stephen Gantt (University of Minnesota).
Saltveit Jr., M.E.
Ethylene in Plant Biology. Academic Press, Inc.,