purification and characterization of pea thioredoxin f expressed in escherichia coli

Plant Molecular Biology 26: 225-234, 1994. © 1994 Kluwer Academic Publishers. Printed in Belgium. 225

Purification and characterization of pea thioredoxin f expressed in Escherichia coli

Michael Hodges, Myroslawa Miginiac-Maslow, Paulette Decottignies, Jean-Pierre Jacquot, Mariana Stein, Loic Lepiniec, Claude Cr6tin and Pierre Gadal Laboratoire de Physiologie V@dtale Mol~culaire, Centre de Recherches sur les Plantes, Bt~timent 630, URA CNRS 1128, Universitd Paris-Sud, 91405 Orsay Cedex, France

Received 21 January 1994; accepted in revised form 9 June 1994

Key words: chloroplast, Escherichia coli, fructose 1,6-bisphosphatase, light activation, NADP-dependent malate dehydrogenase, thioredoxin f

Abstract

The recently cloned cDNA for pea chloroplast thioredoxin f was used to produce, by PCR, a fragment coding for a protein lacking the transit peptide. This cDNA fragment was subcloned into a pET expression vector and used to transform E. coli cells. After induction with IPTG the transformed cells produce the protein, mainly in the soluble fraction of the broken cells. The recombinant thioredoxin f has been purified and used to raise antibodies and analysed for activity. The antibodies appear to be specific towards thioredoxin f and do not recognize other types of thioredoxin. The recombinant protein could activate two chloroplastic enzymes, namely NADP-dependent malate dehydrogenase (NADP- MDH) and fructose 1,6-bisphosphatase (FBPase), both using dithiothreitol as a chemical reductant and in a light-reconstituted/thylakoid assay. Recombinant pea thioredoxin f turned out to be an excellent catalyst for NADP-MDH activation, being the more efficient than a recombinant m-type thioredoxin of Chlamydomonas reinhardtii and the thioredoxin of E. coli. At the concentrations of thioredoxin used in the target enzyme activation assays only the recombinant thioredoxin f activated the FBPase.

Introduction

Thioredoxins are ubiquitous, low-molecular- weight (12 kDa) proteins which are characterized by an active site, Cys-Gly-Pro-Cys, where the two cysteine residues in close proximity can form an intramolecular disulfide bridge [4, 5]. Thioredox- ins participate in numerous redox reactions via a reversible disulfide/dithiol reduction reaction in- volving these cysteines [6]. In photosynthetic cells they play an important role in the light-dependent ferredoxin/thioredoxin regulatory system [9] which modifies the activities of certain chloroplastic enzymes [6].

In most non-photosynthetic cells only one type of thioredoxin is present. However, photosynthetic cells contain several types of thioredoxins which can be distinguished by their primary amino acid sequence, subcellular localization, reducing system and function. In the chloroplasts, two types of thioredoxin are present. Those capable of activating chloroplast fructose 1,6-bisphosphatase (FBPase) [6, 24] and certain other key enzymes of CO: assimilation (sedoheptulose 1,7- biphosphatase [19] and phosphoribulokinase [28]) are called f-type thioredoxins, while those preferentially activating NADP-dependent malate dehydrogenase (NADP-MDH) [15] are

226

termed m-type thioredoxins. A third type of thioredoxin which is found in the cytoplasm and is specifically reduced by an NADP/thioredoxin system is called an h-type thioredoxin [10].

It appears that thioredoxin f shows a rather high specificity for interaction with target enzymes in the reductive pentose phosphate pathway, as the above mentioned enzymes cannot be signifi- cantly activated by other thioredoxins [6]. To study thioredoxin specificity at the molecular level it is necessary to have sufficient pure protein to carry out extensive structural and biochemical studies. However, thioredoxin f is difficult to purify [12] in quantities sufficient for such studies. Until now, only one f-type thioredoxin has been extensively studied, including full amino-acid sequencing, cDNA cloning and expression in E. coli [1, 16]. However, the recombinant protein was obtained as insoluble inclusion bodies which required a harsh denaturation/renaturation treatment to be solubilized [ 1 ]. In order to get a more general view of the properties and specificity of thioredoxin f, the coding sequence of a cDNA of thioredoxin f from pea has been cloned into a pET expression vector to overexpress this protein in E. co6. The recombinant protein, which in our case was found to be soluble, has been purified, and characterized and antibodies specific to thioredoxin f have been produced. The success of this procedure opens up the possibility to inves- tigate thioredoxin f specificity by site-directed mutagenesis.

Materials and methods

Bacterial strains and plasmids

Escherichia coli strain DH5~ F' obtained from Gibco BRL was used to produce high yields of plasmids and E. coli strain BL21 (DE3) was used for production of thioredoxin f encoded by recombinant pET vectors (gift from Dr L. Thelander). Bacteria were grown at 37 °C on Luria-Bertani (LB) medium, supplemented with ampicillin (50 #g/ml) when the bacteria carried plasmids conferring resistance to this antibiotic.

Isolation of plasmid DNA, preparation of DNA fragments, ligation and transformation of E. coli cells were carried out as described elsewhere [21].

Amplification by PCR of cDNA corresponding to a mature pea thioredoxin f-coding sequence

Two oligonucleotides were constructed based on the previously described nucleotide sequence of pea thioredoxin f [18]. The upstream 20-mer oligonucleotide was derived from the nucleotide sequence homologous to the non-coding strand and corresponding to the amino acid sequence Val-Gly-Lys-Val-Thr taken as the N-terminus of the mature thioredoxin f protein. It included an Nco I restriction site (underlined) at the 5' end, giving the following sequence: 5-CC ATG GTA G G G AAA GTA ACC-3.

The construction of the upstream primer re- sulted in the addition of a methionine to the recombinant protein in place of a threonine. The N-terminus was chosen based on an amino acid sequence comparison between the thioredoxin f of pea [18] and spinach [16], choosing the first strictly homologous sequence close to the proposed N-terminus [1, 16] (see Fig. 1).

The downstream 20-mer oligonucleotide was derived from the nucleotide sequence of pea thioredoxin f complementary to the non-coding strand and corresponding to the amino acid sequence Val-Arg-Ser-Ser. It included the stop codon and a Bam HI restriction site (underlined), giving the following sequence: 5-GGA TCC TAA CTA GAC CGA AC-3.

The PCR reaction was initiated directly with an aliquot of the plasmid containing the full-length cDNA of pea thioredoxin f [ 18]. The plasmid was denatured by heating for 3 min at 97 ° C. The sample was then subjected to PCR in the presence of 400 nM of each oligonucleotide, 200 #M dNTPs, 1.3 mM MgC12 and 2.5 units Taq poly- merase (total volume 100 #1). 35 cycles (1 min 95 °C, 2min 55 °C, 3min 72 °C) were per- formed, followed by an elongation of 6 rain at 72 °C.

Cloning of the PCR product

The double-stranded PCR product having the correct size was purified and rendered blunt- ended by treatment with the Klenow fragment, and ligated into pBluescript SK + plasmid prepared previously by restriction with Eco RV. Fol- lowing transformation of E. coli strain DH5e F ' , the nucleotide sequence of the insert was verified by the dideoxy sequencing method using a Pharmacia T7 sequencing kit with SK and KS primers.

Expression of pea thioredoxin f cDNA in E. coli

An Nco I/Bam HI restriction fragment was obtained by digestion of the sequencing vector, purified and ligated into the pET 8c expression vector and introduced into the DH5c~ F' E. coli strain. Plasmid DNA was prepared from the ampicillin-resistant transformants and tested by restriction using Nco I and Bam HI to see if the inserted cDNA was indeed present. The resulting construct, named pETf2, was introduced into the BL21 (DE3) E. coli strain to test its ability to direct the synthesis of recombinant pea thioredoxin f.

A colony of BL21 (DE3) E. coli containing pETf2 was used to start an overnight 30 ml LB culture, supplemented with ampicillin, which was then used to inoculate the final 1.6 1 LB culture. When the culture reached an absorbance of 0.6 at 600 nm, 100/zM IPTG was added and the bacteria were grown for a further 4 h. Cells were collected by centrifugation, resuspended in 50 ml of buffer containing 30 mM Tris-HC1 pH 7.9, 1 mM EDTA, 0.5 mM benzamidine, 500/~M PMSF and 1.4 mM 2-mercaptoethanol.

Extraction and purification of the recombinant protein

The bacterial cell suspension was freeze/thawed twice (liquid N2/37 o C) before disruption by three passages through a precooled French pressure cell (Amicon), at 60 MPa and centrifugation for

227

30min at 20000 ×g. The volume of the supernatant was adjusted to 150 ml with the resuspen- sion buffer and this gave the crude extract for recombinant thioredoxin f purification. After a heat shock of 5 min at 70 °C the denatured proteins were removed by centrifugation for 30 min at 20000 x g and the resulting supematant was fractionated by addition of solid ammonium sulfate. The proteins precipitating between 35-90% saturation were collected by centrifugation at 20000 x g for 30 min and resuspended in a small volume (15 ml) of buffer A (30mM Tris-HC1 pH 7.9) containing 200 mM NaC1. After the addition of 7.5 mM DTT, this sample was loaded onto a Sephadex G-50 column equilibrated with 30mM Tris-HC1 pH 7.9 containing 200mM NaC1 and proteins were eluted in 5 ml fractions by gravity flow. Fractions containing thioredoxin were detected by their ability to activate NADP- MDH in the presence of DTT (see below), pooled and dialysed against buffer A on an Amicon cell equipped with a YM30 membrane. The proteins were loaded onto a DEAE-Sephacel column equilibrated with buffer A. The thioredoxin f activity was found in the passed-through fraction which was concentrated on an Amicon cell as above. This procedure yielded ca. 6 mg of recombinant thioredoxin f per litre of bacteria culture. The quantity of purified protein was calculated by the absorbance at 280 nm using a coefficient of 0.85 mg/ml.

HPLC chromatography and N-terminal amino acid sequence determination

A fraction of the recombinant protein was subjected to HPLC by reverse-phase liquid chromatography using a 0.46 cm x 15 cm Vydac C4 column where the protein was eluted with a 35-70% acetonitile gradient in 0.1% trifluoro- acetic acid, at a flow rate of 1 ml/min. This sample was then used to determine the N-terminal amino acid sequence by automated Edman degradation, using an Applied Biosystems 476 A sequencer with on-line HPLC detection ofphenylthiohydan- toin amino acids.

228

Antibody production

Antibodies raised against the recombinant pea thioredoxin f were prepared as described [26]. Rabbits were injected with 0.7 mg purified thioredoxin f emulsified with Freund's complete adju- vant followed by three intraveinous booster in- jections of 0.7 mg ofthioredoxin f. After bleeding, the antibodies in the serum were precipitated with 33 ~o ammonium sulfate, resuspended in buffer A and purified by affinity chromatography on protein A-sepharose. After washing the column with buffer A containing 150 mM NaC1, the IgG fraction was eluted with 0.2 M sodium citrate (pH 2.8) and reprecipitated with 33 ~o ammonium sulfate. For western blot analyses, the antibodies were further 'cleaned' on nitrocellulose filters saturated with a crude DH5e bacterial protein extract.

phate, 0.9 mM NADP, 14 mM 2-mercaptoethanol, 0.7 U glucose 6-phosphate dehydrogenase and 1.8 U phosphoglucoisomerase. The activity was determined as the increase in absorbancy at 340 nm.

Light activation of maize leaf NADP-MDH [9] or spinach leaf FBPase [23]

The target enzyme was activated in a 30 #1 mixture containing 100mM Tris-HCl pH 7.9, washed pea thylakoids (30/~g chlorophyll), 10 #M ferredoxin, 2/zM ferredoxin/thioredoxin reductase (purified from spinach), variable concentrations of thioredoxins and either NADP- MDH or FBPase. The activation was carried out under air at 25 ° C with 400/~mol quanta m - 2 s - of white light for various times. The respective enzymatic activities were then measured as described above using 25 #1 aliquots.

Enzyme assays

DTT-dependent activation of maize leaf NADP- MDH [14]

The enzyme was activated in 25/~1 medium containing 40 mM Tris-HC1 pH 7.9, 4 mM DTT, 3 #M NADP-MDH and variable (0 -40~M) amounts of thioredoxin. After incubation at 20 ° C, an aliquot of 20/~1 was used to determine the activity of NADP-MDH at 30 °C in a reaction medium (1 ml) containing 100 mM Tris-HCl pH 7.9, 150/~M NADPH and 750 ~M oxaloace- tic acid. The amount of NADPH oxidized by the enzyme was monitored as a decrease in absorbancy at 340 nm.

DTT-dependent activation of spinach leaf FBPase [231

The enzyme was activated in 100/,1 medium containing 100 mM Tris-HC1 pH 7.9, 5 mM DTT, 0.6 #M FBPase and thioredoxin as indi- cated. After incubation at 20 °C, an aliquot of 30/~1 was taken to determine the activity of FB- Pase in a reaction mixture (1 ml) containing 100mM Tris-HCl pH7.9, 0.1raM EDTA, 1.5 mM MgSO4, 2.5 mM fructose, 1,6-bisphos-

Polyacrylamide gel electrophoresis and western blot

Proteins were separated by SDS-PAGE following a modified method [22] using a 6~o stacking gel and a 16.5 ~o separating gel containing 0.133 ~/o glycerol in 1 M Tris-HC1 pH 8.45 and 0.1 ~o SDS. The anode running buffer was 0,2 M Tris-HC1 pH 8.9 and the cathode running buffer was 0.1 M Tris, 0.1 M Tricine and 0.1 ~o SDS giving a pH of 8.3. Proteins were visualized by Coomassie Brilliant Blue staining.

For western blot analyses, the proteins were separated as above and transferred to nitrocellulose at 60 V, 4 h. After blocking the sites with TSMB buffer (30 mM Tris-HC1 pH 7.9, 150 mM NaC1 supplemented with 4~o (w/v) powdered milk) the filters were incubated overnight at 4 °C in TSMB containing antibodies raised against the recombinant thioredoxin f. After three washes with TSMB buffer and an hour incubation with sheep anti-rabbit antibodies conjugated to horse- radish peroxidase activity, the thioredoxin f was visualized with 4-chloro-l-naphthol as substrate in the presence of H20 >

229

Results

The availability of a full-length cDNA, LTfP1, coding for a thioredoxin f of pea [ 18] enabled us to amplify, by PCR, a fragment of this clone for the production in E. coli of a mature form of the protein. A partial comparison of the pea and spinach thioredoxin f primary amino acid sequences (Fig. 1) showed that the two proteins had a high degree of similarity beginning from the valine at position 74 based on the complete pea sequence. Since it is not certain from the literature where the spinach thioredoxin f begins [ 1 ], val-74 was chosen as the first amino acid for the recombinant pea thioredoxin f. A methionine was added immediately upstream of the chosen valine by the ATG codon, present in the Nco I restriction site. The Nco I and Bam HI sites included in the PCR product were used to clone the fragment into the expression vector pET 8c. The resulting construction was used to transform E. coli strain BL21 (DE3).

Small scale production (100 ml of culture) was carried out to see if the recombinant protein was located in the soluble or insoluble fraction of the E. coli cells. The insoluble fraction was resolubi- lized in 4 M urea and 10 mM CHAPS and incubated at 37 °C for an hour. After centrifugation, the supernatant was dialysed overnight against buffer A. Western blots and activity studies showed that higher amounts of recombinant thioredoxin f were present in the soluble fraction than in the solubilized insoluble fraction (as with recombinant NADP-MDH produced using the

same vector [13 ]). Therefore, the large-scale production and purification of the pea thioredoxin f was carried out using only the soluble fraction of broken E. coli cells.

It has been mentioned that the purification of large amounts of thioredoxin f from higher plants [24] and green algae [12] is difficult. The recombinant pea thioredoxin f was purified from E. coli

rather easily, giving a sufficiently large quantity of highly pure protein. The purity of the recombinant thioredoxin f preparation was confirmed by SDS-PAGE (Fig. 2A, lane3) and by HPLC chromatography. When chromatographed on a reverse-phase column, a single symmetric peak

Reccmbinant

Spinach 57 _ _ 19WPIV * o* o o ~o o* * **********W****

Pea 50 ~ - ISVSVRSSLEIg~ IV

Reccn~binant

Fig. 1. Partial amino acid sequence alignment of thioredoxin f from pea [18] and spinach [1, 16] showing the strategy for the choice of the N-terminus. The underlined methionines are proposed N-terminal amino acids of the processed thioredoxin ffrom spinach [1, 16]. The numbers correspond to the amino acid numbering of the complete protein with transit peptide. * indicate identical and ° corresponds to conserved amino acids between the two sequences.

Fig. 2. A comparison of the three different thioredoxin prepa- rations used in this work. A. SDS/PAGE of the purified recombinant proteins of thioredoxin of E. coli (lane 1), the thioredoxin m from Chlamydomonas reinhardtii (lane 2), thioredoxin f of pea (lane 3) and the crude protein extract used to purify the thioredoxin f (lane 4). B. A western blot using the antibodies raised against the recombinant thioredoxin f of pea. Each lane contains equal amounts of either recombinant thioredoxin from E. coli (lane 1), C. reinhardtii m-type (lane 2) and h-type (lane 3) thioredoxins or pea thioredoxin f (lane 4). Lane M contains molecular mass markers at 27, 36, 48, 58, 80 and 110 kDa.

230

was observed. Fig. 2A also shows the recombinant E. coli thioredoxin (lane 1) and the Chlamy- domonas reinhardtii m-type thioredoxin (lane 2) used in this work along with the crude protein extract used in the thioredoxin f purification (lane 4). An N-terminus amino acid sequence analysis of the recombinant protein was per- formed, no heterogeneity could be detected and the result was in full agreement with the one predicted from the cDNA sequence. This shows that proteases present in E. coli do not degrade the N-terminus of the recombinant protein, except for the removal of the added methionine which is indeed predicted from the work of Hirel et al. [11].

Polyclonal antibodies were raised against the purified recombinant pea thioredoxin f. When western blot analyses were carried out using equal amounts ofthioredoxins from E. coli, m-type and h-type thioredoxins from C. reinhardtii, and purified recombinant pea thioredoxin f, the antibodies only recognized the thioredoxin f protein (Fig. 2B). It was also seen that the thioredoxin f antibodies did not cross-react with thioredoxin m from spinach leaves nor with a thioredoxin from Dictyostelium discoideum (data not shown). The inability of the antibodies to cross-react with the E. coli thioredoxin shows that the purified protein is free from the native thioredoxin of E. colt This is important when investigating the specificity and kinetic characteristics of the purified thioredoxin f as E. coli thioredoxin activates NADP-MDH in a similar manner to m-type thioredoxins [8].

The ability of the recombinant pea thioredoxin f to activate chloroplastic NADP-MDH and FB- Pase either in the presence of DTT or in a reconstituted light activation assay was investigated. The activation by thioredoxin f was compared to the ability of a recombinant m-type thioredoxin from C. reinhardtii (Ch2) and E. coli thioredoxin to activate these enzymes under the same experimental conditions. It has already been shown that recombinant E. coli thioredoxin has the same efficiency as spinach leaf thioredoxin m in activating NADP-MDH [7], and that native Ch2 is almost as efficient [7].

Figure3 shows the time course of DTT-

~E

480 .~_ E

"~ 320 ._N

160 E3 <: Z "6 Oi E 0

,5., c

i

lO

Minutes of activation

20

Fig. 3. Time courses of dithiothreitol-dependent activation of sorghum N A D P - M D H . Thioredoxin (15/2M) was incubated in the presence of 5 m M DTT and 3 ~ M N A D P - M D H and at different times an aliquot was taken and used to determine the activity of N A D P - M D H , as described in the Materials and methods. Recombinant proteins of thioredoxin f from pea (O) and thioredoxin m from Chlamydomonas reinhardtii (0) .

dependent activation of NADP-MDH in the presence of recombinant thioredoxin f and Ch2. Surprisingly, the f-type thioredoxin appeared to give higher activation rates of NADP-MDH although after 15 min a similar NADP-MDH activity was attained (see Table 1). The ability of the three different thioredoxins to activate NADP-MDH in the presence of DTT was investigated as a function of thioredoxin concentration using an activation time of 2 min where the activation rate is linear. It can be seen from Fig. 4A that thioredoxin f led to a better activation of NADP-MDH (as expected from the data in

Table 1. Thioredoxin mediated activation of chloroplast enzymes either in the presence of DTT (+ DTT) or in a reconstituted light-activation/thylakoid system (+Light) . Experimental conditions as described in the Materials and methods. N D stands for not determinable.

Thioredoxin Condition Activation Activation of NADP-MDH of FBPase

Vmax S~/2 Vmax S~/2 (nmol/min) (~M) (nmol/min) (/~M)

f-type + DDT 330 2.2 74.1 0.25 + Light 411 5.8 58.0 3.30

m-type + DTT 320 6.2 ND ND + Light 398 10.0 ND ND

Vma x was measured after an activation incubation time of 15 min.

Fig. 3) than the two other thioredoxins investigated which exhibited similar efficiencies. The calculated $1/2 values for thioredoxin are summa- rized in Table 1 and show values similar to those already published for E. coli and native Ch2 thioredoxins [7] of 5.6/~M while recombinant thioredoxin fgave a slightly lower value of 2.2 #M. These data show that thioredoxin f is capable of activating NADP-MDH, and when compared to the other recombinant thioredoxins tested, it appears to be more efficient. This ability of thioredoxin f to activate NADP-MDH is confirmed in Fig. 4B which shows the light-dependent activation of NADP-MDH in a reconstituted chloroplast system. Again, recombinant f-type thioredoxin seemed to be more efficient, with respect to the kinetics of activation of NADP-MDH, compared to the recombinant m-type thioredoxin (data not shown). This is reflected by the some- what lower $1/2 value for thioredoxin f. These results are important when taking into account the discrepancies which exist in the literature concerning the ability of thioredoxin f to activate NADP-MDH [12, 20, 24, 25].

480

d: 320

ID

• - 160

" O~ o 0 < z 480 0 E c

320 L

160 - r

A . o

i i i i

10 20 30 40

B i

I - "

i i i I

10 20 30 40 50

Thioredoxin concn., gM

Fig. 4. Thioredoxin-dependent activation rates of sorghum N A D P - M D H either reduced by (A) DTT or (B) in a light- reconstituted/thylakoid system as described in the Materials and methods. Recombinant proteins of thioredoxin f from pea (O), thioredoxin m from Chlamydomonas reinhardtii ( 0 ) and thioredoxin from E. coli (A) (only in A).

231

" o

c~

Z

0

E ,_~ ._

> 30

60

0 5 10

60

30

, 1 0 20 30 40

Thioredoxin eoncn., gM

Fig. 5. Thioredoxin-dependent activation rates of spinach chloroplastic FBPase either reduced by (A) DTT or (B) in a light-reconstituted/thylakoid system as described in the Ma- terials and methods. Recombinant proteins of thioredoxin f from pea (©), thioredoxin m from Chlamydomonas reinhardtii ( 0 ) and thioredoxin from E. coli (A ) (only in A).

The different types of thioredoxin were also tested for their ability to activate chloroplastic FBPase. Figure 5A shows the DTT-dependent activation of FBPase in the presence of different concentrations of recombinant f-type, Ch2 or E. coli thioredoxins after an activation time of 1 min. It can be seen that only the recombinant pea thioredoxin f could activate the FB Pase under the given experimental conditions, thus confirming the assignment of the recombinant protein to an f-type thioredoxin. The calculated S1/2 value for pea thioredoxin f is shown in Table 1 and the value of 0.25/~M is similar to that already published for both the recombinant and native spinach thioredoxin f [ 1 ]. Figure 5B shows the light- dependent activation of FBPase in the presence of recombinant thioredoxin f or Ch2. Again, the data show that in the reconstituted chloroplast system only the recombinant thioredoxin f activated FBPase, however in the presence of atmo- spheric oxygen, higher concentrations of thioredoxin f are required to bring about FBPase activation (see Table 1). This result is in contrast

232

to the observations made using a recombinant spinach thioredoxin f in a light-activation assay carried out under argon in a reconstituted system where the S m for thioredoxin f was similar to the DTT-dependent assay [ 1].

Discussion

An active recombinant thioredoxin f from pea has been successfully produced and purified from E. colicells. The majority of the protein was found in the soluble fraction which facilitated the protein purification using standard procedures. Com- pared to the spinach thioredoxin f construct used by Aguilar et al. [1 ], the expression of our pea thioredoxin f did not require any harsh denaturation/renaturation steps. The protein appeared to be stable in E. coli, which only removed the methionine added to the thioredoxin primary amino acid sequence. It seems that the N- terminus is not important in the activation properties of the thioredoxin f as a previously described recombinant thioredoxin f from spinach (see [ 1 ] and Fig. 1) having a longer and different N- terminus was also fully active and similar to the native protein in its ability to activate chloroplastic FBPase.

The observed specificity of the antibodies raised against the recombinant pea thioredoxin f is not surprising. Numerous thioredoxin amino acid sequences are found in the literature (e.g. [17]) and their comparison shows that thioredoxin f is not very similar to other thioredoxins (30 ~o similarity), except for the conserved active site sequence. This is shown indirectly by the fact that the thioredoxin f antibodies do not recognise any of the other thioredoxins tested. These specific antibodies will be useful in purification work [2] and screening other cDNA libraries.

A certain diversity of results concerning the ability of thioredoxin f to activate NADP-MDH is present in the literature. Early data showed that the thioredoxin f from spinach could activate NADP-MDH but that the m-type thioredoxin was more efficient [25]. However, a thioredoxin f from Anabaena sp. 7119 was shown not to ac-

tivate NADP-MDH [27]. Huppe et al. [12] reported that C. reinhardtii thioredoxin f did not activate NADP-MDH and that antibodies raised against the spinach thioredoxin f reacted only weakly with the purified C. reinhardtii protein. These are similar observations to those of Prado et al. [20] who reported the purification of a pea thioredoxin f which does not activate NADP- MDH and which shows no immunological cross- reactions between pea and spinach thioredoxin f proteins with the corresponding antibodies. In this work, however, it is shown that the recombinant pea thioredoxin f is capable of activating NADP-MDH, both in the presence of DTT and in a reconstituted light/thylakoid system. Indeed, the results presented above indicate that thioredoxin f appears to be more efficient in activating NADP-MDH than both a recombinant m-type thioredoxin from C. reinhardtii and the thioredoxin from E. coli. The $1/2 values obtained for the activation of FBPase in the presence of DTT are similar to those already reported for the native and recombinant spinach thioredoxin f [ 1, 8], indicating that the recombinant protein is fully active. The increase in the S~/2 value for FBPase activation in the reconstituted system by the thioredoxin f appears to be due to the oxidizing effect of air.

The contradictory reports in the literature concerning the specificity of thioredoxin f and its ability to activate NADP-MDH are not easy to ex- plain. It could be that other thioredoxins exist in the plant which can equally activate FBPase but have different abilities to activate NADP-MDH. Two forms of thioredoxin f have already been purified using FBPase affinity chromatography [3]. However, they both activated NADP-MDH but differed in their ability to activate chloroplastic sedoheptulose bisphosphatase. Indeed, the exact physiological roles of the numerous different thioredoxins are not known. Hopefully, the availability of fulMength cDNAs coding for these thioredoxins will enable, with the production of transformed plants containing either mutated or antisense constructions, a study of the physiological function of each type of thioredoxin.

Acknowledgements

This work was supported by an ECC grant EEC ERB CI 1CT920070.

References

1. Aguilar F, Brunner B, Gardet-Salvi L, Stutz E, Sch~irmann P: Biosynthesis of active spinach-chloroplast thioredoxin f in transformed E. coli. Plant Mol Biol 20: 301-306 (1992).

2. Bakrim N, Echevarria C, Cr6tin C, Arrio-Dupont M, Pierre J-N, Vidal J, Chollet R, Gadal P: Regulatory phos- phorylation of Sorghum leaf phosphoenolpyruvate carboxylase. Identification of the protein-serine kinase and some elements of the signal-transduction cascade. Eur J Biochem 204:821-830 (1992).

3. Buc J, Rivi6re M, Gontero B, Sauve P, Meunier J-C, Ricard J: Affinity chromatography, on fructose-bisphosphatase-sepharose of two chloroplastic thioredoxin f. Pu- rification and comparative molecular properties. Eur J Biochem 140:199-202 (1984).

4. Buchanan BB: Role of light in the regulation of chloroplast enzymes. Annu Rev Plant Physiol 31:341-374 (1980).

5. Buchanan BB: Regulation of CO2 assimilation in oxy- genic photosynthesis: the ferredoxin/thioredoxin system. Arch Biochem Biophys 288:1-9 (1991).

6. Cs6ke C, Buchanan BB: Regulation of the formation and utilization of photosynthate in leaves. Biochim Biophys Acta 853:43-63 (1986).

7. Decottignies P, Schmitter J-M, Jacquot J-P, Dutka S, Pi- caudA, GadalP: Purification, characterization, and complete amino acid sequence of a thioredoxin from a green algae, Chlamydomonas reinhardtii. Arch Biochem Biophys 280:112-121 (1990).

8. De Lamotte-Guery F, Miginiac-Maslow M, Decot- tignies P, Stein M, Minard P, Jacquot J-P: Mutation of a negatively charged amino acid in thioredoxin modifies its reactivity with chloroplastic enzymes. Eur J Biochem 196: 287-294 (1991).

9. Droux M, Miginiac-Maslow M, Jacquot J-P, Gadal P, Crawford NA, Kosower NS, Buchanan BB: Ferredoxin- thioredoxin reductase: a catalytically active dithiol group links photoreduced ferredoxin to thioredoxin functional in photosynthetic enzyme regulation. Arch Biochem Bio- phys 256:372-380 (1987).

10. Florencio FJ, Yee BC, Johnson TC, Buchanan B: An NADP/thioredoxin system in leaves: purification and characterization of NADP-thioredoxin reductase and thioredoxin h from spinach. Arch Biochem Biophys 266: 496-507 (1988).

11. Hirel Ph H, Schmitter J-M, Dessen P, Fayat G, Blan- quet S: Extent of N-terminal methionine excision from

233

E. coli proteins is governed by the side-chain length of the penultimate amino acid. Proc Natl Acad Sci (USA) 86: 8247-8251 (1989).

12. Huppe HC, de Lamotte-Gu6ry F, Jacquot J-P, Buchanan BB: The ferredoxin-thioredoxin system of a green algae, Chlamydomonas reinhardtit. Planta 180:341-351 (1990).

13. Issakidis E, Miginiac-Maslow M, Decottiginies P, Jac- quot J-P, Cr6tin C, Gadal P: (1992) Site-directed mutagenesis reveals the involvement of an additional thioredoxin-dependent regulatory site in the activation of recombinant sorghum leaf NADP-malate dehydrogenase. J Biol Chem 267:21577-21583 (1992).

14. Jacquot J-P, Buchanan BB, Martin F, Vidal J: Enzyme regulation in C4 photosynthesis. Plant Physiol 68: 300- 304 (1981).

15. Jacquot J-P, Vidal J, Gadal P, Sch~irmann P: Evidence for the existence of several enzyme specific thioredoxins in plants. FEBS Lett 96:243-246 (1978).

16. Kamo M, Tsugita A, Wiessner C, Wedel N, Bartling D, Herrmann RG, Aguilar F, Gardet-Salvi L, Sch~lrmann P: Primary structure of spinach-chloroplast thioredoxin f. Protein sequencing and analysis of complete cDNA clones for spinach-chloroplast thioredoxin f. Eur J Bio- chem 182:315-322 (1989).

17. Le Mar6chal P, Hoang BMC, Schmitter J-M, Van Dors- selaer A, Decottignies P: Purification, properties and primary structure of thioredoxin from Aspergillus nidulans. Eur J Biochem 210:421-429 (1992).

18. Lepiniec L, Hodges M, Gadal P, Cr6tin C: Isolation, characterization and nucleotide sequence of a full-length pea cDNA encoding thioredoxin-f. Plant Mol Biol 18: 1023-1025 (1992).

19. Nishizawa AN, Buchanan BB: Enzyme regulation in C4 photosynthesis. Purification and properties of thioredoxin-linked fructose bisphosphatase and sedoheptulose bisphosphatase from corn leaves. J Biol Chem 256:6119- 6126 (1981).

20. Prado FE, Lazaro J J, Hermoso R, Chueca A, Lopez- Gorg6 J: Purification and properties of pea (Pisum sati- rum L.) thioredoxin f, a plant thioredoxin with unique features in the activation of chloroplast fructose-l,6- bisphosphatase. Planta 188:345-353 (1992).

21. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY (1989).

22. Sch~gger H, von Jagow G: Tricine-sodium dodecyl sulfate polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 166:368-379 (1987).

23. Sch~lrmann P, Jacquot J-P: Improved in vitro light activation and assay systems for two spinach chloroplast enzymes. Biochim Biophys Acta 569:309-312 (1979).

24. Schtlrmann P, Maeda K, Tsugita A: Isomers in thioredoxins of spinach chloroplasts. Eur J Biochem 116:37-45 (1981).

25. Tsugita A, Maeda K, Schttrmann P: Spinach chloroplast

234

thioredoxins in evolutionary drift. Biochem Biophys Res Comm 115:1-7 (1983).

26. Vidal J, Godbillon G, Gadal P: Recovery of active highly purified phosphoenolpyruvate carboxylase from specific immunoabsorbent column. FEB S Lett 118:31-34 (1980).

27. Whittaker MM, Gleason FK: Isolation and characteriza-

tion of thioredoxin f from the filamentous cyanobacte- rium, Anabaena sp. 7119. J Biol Chem 259:14088-14093 (1984).

28. Wolosiuk RA, Crawford NA, Yee BC, Buchanan BB: Isolation of three thioredoxins from spinach leaves. J Biol Chem 254:1627-1632 (1979).

purification and characterization of pea thioredoxin f expressed in escherichia coli

Documents