Biosci. Biotechnol. Biochem.

, 74 (3), 627635, 2010

Assimilation of Formaldehyde in Transgenic Plants Due to the Introduction of the Bacterial Ribulose Monophosphate Pathway Genes
Li-mei C HEN,1;2 Hiroya Y URIMOTO,3 Kun-zhi L I,1;2 Izumi O RITA,3 Motomu A KITA,4 Nobuo K ATO,3;5 Yasuyoshi SAKAI,3 and Katsura I ZUI1;4; y
Graduate School of Biostudies, Kyoto University, Oiwake-cho, Kitashirakawa, Sakyo-ku, Kyoto 606-8502, Japan Biotechnology Research Center, Kunming University of Science and Technology, Bailong Campus, Bailongsi, Bailong Road, Kunming 650224, China 3 Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Oiwake-cho, Kitashirakawa, Sakyo-ku, Kyoto 606-8502, Japan 4 School of Biology-Oriented Science and Technology, Kinki University, 930 Nishimitani, Kinokawa, Wakayama 649-6493, Japan 5 Faculty of Bioenvironmental Science, Kyoto Gakuen University, Sogabe-cho, Kameoka, Kyoto 621-8555, Japan
2 1

Received November 16, 2009; Accepted December 8, 2009; Online Publication, March 7, 2010 [doi:10.1271/bbb.90847]

Formaldehyde (HCHO) is an air pollutant suspected of being carcinogenic and a cause of sick-house syndrome. Microorganisms called methylotrophs, which can utilize reduced C1 compounds such as methane and methanol, x and assimilate HCHO, whereas most plants are unable to assimilate HCHO directly. We found that a bacterial formaldehyde-xing pathway (ribulose monophosphate pathway) can be integrated as a bypass to the Calvin-Benson cycle in transgenic Arabidopsis thaliana and tobacco by genetic engineering. These plants showed enhanced tolerance to HCHO and enhanced capacity to eliminate gaseous HCHO by xing it as a sugar phosphate. Our results provide a novel strategy for phytoremediation of HCHO pollution, and also represent the rst step toward the production of plants that can assimilate natural gas-derived C1 compounds. Key words: Arabidopsis thaliana; Calvin-Benson cycle; formaldehyde; ribulose monophosphate pathway; transgenic plant

Formaldehyde (HCHO), industrially produced from natural gas, is used in many products, including adhesives, bonding agents, and solvents. HCHO reacts as an electrophile with the side-chains of arginine and lysine and the amino groups of RNA and DNA, and causes protein-protein, protein-DNA, and DNA-DNA cross-links. Therefore, it is suspected to be carcinogenic and a cause of sick-house syndrome.13) HCHO is also known as one of the major volatile organic compounds (VOCs) of air pollution and the World Health Organization has established an air quality guideline of 0.1 mg m3 . Utilization of houseplants for the removal
y

of VOCs was rst proposed by Wolverton et al.4) Although some houseplants were found to have relatively high ability to remove HCHO from the air,5) later studies revealed that the main organisms involved were not the plants themselves, but rather soil microorganisms symbiotically living with the plants.6) A trial to enhance the capacity for HCHO detoxication in plants was made by overexpressing an enzyme for HCHO oxidation, glutathione-dependent formaldehyde dehydrogenase.7) It is known that plants generally have a very limited capacity for assimilating gaseous HCHO,8) though HCHO is present as a natural metabolite in plants, mainly in forms bound with cofactors such as tetrahydrofolate and glutathione.9) On the other hand, some microorganisms are known to x HCHO into their cell constituents when they produce it endogenously as a metabolite and when they are exposed to exogenous HCHO. The ribulose monophosphate (RuMP) pathway is one of the HCHO-xation pathways found in microorganisms called methylotrophs, which utilize onecarbon compounds, such as methane and methanol, as the sole carbon source.1012) The key enzymes of this pathway are 3-hexulose-6-phosphate synthase (HPS), which xes HCHO to D-ribulose 5-phosphate (Ru5P) to produce D-arabino-3-hexulose 6-phosphate (Hu6P), and 6-phospho-3-hexuloisomerase (PHI), which converts Hu6P to fructose 6-phosphate (F6P). The bacterial RuMP pathway and the plant CalvinBenson cycle share common metabolic features: (i) both pathways x a one-carbon unit to ribulose phosphate; (ii) the xation reaction eventually yields F6P; and (iii) Ru5P is regenerated from F6P through rearrangement reactions. Based on these facts, we intended to introduce

To whom correspondence should be addressed. School of Biology-Oriented Science and Technology, Kinki University, 930 Nishimitani, Kinokawa, Wakayama 649-6493, Japan; Fax: +81-736-77-4754; E-mail: izui@waka.kindai.ac.jp Abbreviations: BA, N 6 -benzyladenine; F6P, fructose 6-phosphate; G6P, glucose 6-phosphate; Gen, gentamycin; HCHO, formaldehyde; HPS, 3-hexulose-6-phosphate synthase; Hu6P, D-arabino-3-hexulose-6-phosphate; Km, kanamycin; MS, Murashige and Skoog; NAA, naphthalene acetic acid; NMR, nuclear magnetic resonance; Nos, nopaline synthase; 3-PGA, 3-phosphoglycerate; PHI, 6-phospho-3-hexuloisomerase; rbcS, the gene for the small subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase; Ri5P, D-ribose 5-phosphate; rmpA, the gene for 3-hexulose 6-phosphate synthase; rmpB, the gene for 6-phospho-3-hexuloisomerase; Ru5P, D-ribulose 5-phosphate; RuBP, D-ribulose 1,5-bisphosphate; RuMP, ribulose monophosphate; Spe, spectinomycin; Su7P, sedoheptulose 7-phosphate; TP, transit peptide; TCA, trichloroacetic acid; TLC, thin layer chromatography; VOC, volatile organic compound; Xu5P, D-xylulose 5-phosphate

628

L. C HEN et al.

RuMP pathway HCHO fixation (Bacteria) HCHO HPS Ru5P /RuBP regeneration (Common)
CH 2 OH CO HCOH HCOH CH 2 O- P CH 2 OH HOCH CO HCOH HCOH CH 2 O- P

PHI
C5
CH 2 OH CO HOCH HCOH HCOH CH 2 O- P

Hu6P

C5 C5

C7 C3 C3

C4 C3

Ru5P

F6P

RuBP

C3

FBP

CO2 fixation & reduction (Plants)

CO2

3-PGA

C3

Calvin-Benson cycle
Fig. 1. Schematic Representation of the Bacterial RuMP Pathway and the Plant Calvin-Benson Cycle. HPS and PHI denote 3-hexulose-6-phosphate synthase and 6-phospho-3-hexuloisomerase respectively. The abbreviations for several sugar phosphates are as follows: Ru5P, ribulose 5-phosphate; Hu6P, hexulose 6-phosphate; F6P, fructose 6-phosphate; FBP, fructose 1,6-bisphosphate; RuBP, ribulose 1,5-bisphosphate; 3-PGA, 3-phosphoglycerate. The other metabolites in the pathway are symbolized merely by their carbon numbers for simplicity.

a HCHO-xing reaction catalyzed by HPS and PHI into plants and to couple it with the RuMP regeneration reaction of the Calvin-Benson cycle. In the pathway designed, the CO2 -xation and reduction steps in the cycle are bypassed by the HPS- and PHI-catalyzed reactions (Fig. 1). In the present study, we produced transgenic Arabidopsis and tobacco plants that expressed the HPS- and PHI-encoding genes, rmpA and rmpB respectively, from a methylotrophic bacterium, Mycobacterium gastri MB19,13) and targeted their products to the chloroplasts with articially-added transit peptides.14) The resulting plants showed an enhanced capacity for HCHO assimilation via the installed RuMP pathway. Although HCHO is strongly toxic to all organisms, the transgenic plants showed strong tolerance to this compound. The immediate use of this novel strategy might be in phytoremediation of HCHO pollution. More importantly, the present work might be the rst step toward the metabolic engineering of plants to confer ability to assimilate various C1 compounds.

was removed. Similarly, the SphI-XbaI fragment (about 600 bp) from pTA-rmpB was ligated into vector, pUC-rbcS-3C to generate a plasmid named pUC-rbcS-rmpB (about 5.2 kb). The pUC-rbcS-rmpA plasmid was then digested with HindIII and BamHI, and a 2.4-kb fragment containing the rbcS-3C promoter and the rmpA-coding regions was obtained. This fragment was subcloned into binary vector pPZP21116) at the HindIII and BamHI sites to generate the plant expression vector pPZP211-rmpA, which carries a chimeric gene consisting of the rbcS3C promoter with its transit peptide, the rmpA-coding region, and the nopaline synthase (Nos) terminator. Similarly, the HindIII-XbaI fragment (about 2.4 kb) from pUC-rbcS-rmpB was prepared and then subcloned into another binary vector, pPZP221,16) at the HindIII and XbaI sites, generating another plant expression vector, pPZP221-rmpB, which carries a chimeric gene, consisting of rbcS-3C promoter with its transit peptide, the rmpB-coding region and Nos terminator. The two plant expression vectors were transferred into Agrobacterium tumefaciens C58C1 (pMP90) by electroporation using a Gene Pulser (BioRad, Hercules, CA) with parameters 200 ohms and 2.5 kV/0.2 cm. A. tumefaciens colonies were selected on LB-agar plates containing 100 mg/ml of spectinomycin (Spe). Plant transformation. Arabidopsis transformation was carried out essentially as described by Clough and Bent,17) with minor modications. Arabidopsis thaliana plants were grown in pots in soil, covered with a window screen, under constant light at 22  C in a growth chamber. The A. tumefaciens C58C1(pMp90) strain harboring the binary vector was grown in LB medium with Spe (100 mg/ml) and gentamycin (Gen, 25 mg/ml), until the culture reached an OD600 reading of 0.81.0. Arabidopsis plants at the ower budding stage were inltrated under vacuum with A. tumefaciens cells that had been centrifuged and resuspended in 5% sucrose plus Silwet L-77 (0.005%) solution. Seeds were collected from the inltrated plants and sown on agar medium containing kanamycin (Km, 50 mg/ml) or Gen (80 mg/ml) for transgenic rmpA and rmpB plant selection respectively. To generate double transformants, rmpA and rmpB transformants were reciprocally re-transformed with the rmpB and rmpA constructs. The double transformants were selected on agar medium containing both Km and Gen. Tobacco (Nicotiana tabacum cv. Xanthi) was transformed essentially by following the leaf disk co-cultivation protocol of Horsch et al.18) Co-cultivation was initiated by dipping leaf disks in an Agrobacterium suspension, blotting them on sterile tissue paper, and incubating them for 2 d on Murashige and Skoog (MS) salts mixture medium containing naphthalene acetic acid (NAA, 0.21 mg/ml), N 6 benzyladenine (BA, 0.02 mg/ml), and 3% sucrose for callus induction. The leaf disks were then transferred onto a medium containing antibiotics for transgenic plant selection, and NAA (0.105 mg/ml), BA (0.5 mg/ml), and 3% sucrose for shoot induction. Carbenicillin was included in the medium (500 mg/ml) to inhibit Agrobacterium growth.

Materials and Methods

Plant materials. Arabidopsis thaliana ecotype Columbia and Nicotiana tabacum cv Xanthi were used in this study. Transgenic plants of T2 generation were used in the experiments. Plasmid construction. The coding regions of the rmpA (624 bp) and rmpB (600 bp) genes were amplied by PCR from pUC118-rmpAB.13) For amplication of the rmpA coding region, sense primer, 50 -GCATGCAGCTCCAAGTCGCCATC-30 (containing an SphI site) and antisense primer 50 -GGATCCTCAGGCGATCGCGGCGCGCAG-30 (containing a BamHI site) were used. For amplication of the rmpB coding region, sense primer 50 -GCATGCCGCAAGCCGCAGAAGCC30 (containing a SphI site) and antisense primer 50 -TCTAGATCACTCGAGGTTGGCGTGGCG-30 (containing an XbaI site) were used. The amplied coding regions were subcloned into a TA cloning vector (Takara Bio., Ohtsu, Japan) to generate two plasmids, which were designated pTA-rmpA and pTA-rmpB respectively. The sequences of the rmpA and rmpB coding regions in the two plasmids were conrmed. The SphI-BamHI fragment (about 624 bp) was isolated from pTA-rmpA and ligated into vector pUC-rbcS-3C14,15) to generate a new plasmid named pUC-rbcS-rmpA (about 5.4 kb). In the resulting plasmid, the coding region of the rmpA gene was inserted downstream of the transit peptide of the rbcS gene and the core coding region of the rbcS gene

Formaldehyde Assimilation in Transgenic Plants As for Arabidopsis, tobacco double transformants were generated by re-transformation. Tobacco seedlings were tested for antibiotic resistance by germinating them on MS medium containing 1% sucrose and antibiotics. Growth and maintenance of transgenic plants. For aseptic growth of Arabidopsis plants, seeds were surface sterilized, cold treated at 46  C for 12 d, and germinated on a solidied medium containing agar (1%), MS/G salts (1/2 Murashige and Skoog salts, 1/2 Gamborgs B5 salts, 0.05% MES (pH 5.7), and 1% sucrose, with and without antibiotics. The seedlings were grown in a growth chamber at 22  C under constant light. Resistant seedlings were transplanted to Japanese Green Soil (Sanwa Press, Tsu, Mie, Japan) for propagation, and were grown under constant light at 22  C in a greenhouse. For aseptic growth of tobacco, seeds were surface sterilized and germinated on a solidied medium containing agar (0.8%), MS salts (pH 5.7), and sucrose (3%), with and without antibiotics, in a growth chamber at 25  C under constant light. For propagation of tobacco, the selected seedlings were transplanted to soil and grown under constant light at 25  C in a greenhouse. Northern analysis. Transgenic plants were rst selected on agar medium with antibiotics and then grown for 2 weeks on medium without antibiotics to obtain vigorous growth. Total RNA was isolated from the seedlings using TRIzol Reagent (Invitrogen, Carlsbad, CA) as described by the manufacturer. The coding regions of the rmpA and rmpB genes were used as probes. In Northern blot analysis, total RNA was fractionated by electrophoresis on 1.2% agarose gels containing 6% v/v formaldehyde, blotted onto nylon membranes, and hybridized with [32 P]-labelled DNA probes. Western analysis and measurement of enzyme activity. Arabidopsis and tobacco seedlings (12 g), grown on agar medium or soil, were harvested and homogenized in 10 volumes w/v of protein extraction buer (50 mM phosphate buer pH 7.5, 10% glycerol, 5 mM MgCl2 , 10 mM 2-mercaptoethanol, 1 mM phenylmethylsulfenyl uoride, 2 mM EDTA, and 10% w/v insoluble polyvinylpolypyrollidone). Each extract was centrifuged at 15;000 g for 20 min at 4  C. The supernatant was collected and the protein concentration was determined using a Bio-Rad protein assay kit with bovine serum albumin as the standard. Intact chloroplasts were isolated from Arabidopsis leaves as described by Triboush et al.,19) with modications. The chloroplasts were disrupted and centrifuged to obtain the soluble protein fraction. For Western blot analysis, the proteins were separated in SDSPAGE gel (15%), and then transferred to polyvinylidene diuoride (PVDF) membranes by electroblotting. The membranes were rst treated with rabbit antibodies raised against HPS or PHI from Methylomonas aminofaciens 77a,20) and then with a goat antibody raised against rabbit IgG which had been conjugated with peroxidase. Enzyme activity was determined spectrophotometrically by monitoring the HCHO-dependent production of NADPH at 340 nm, as described by Kato.21) The enzyme reaction was performed at 30  C in an assay mixture containing 50 mM potassium phosphate buer pH 7.5, 2.5 mM MgCl2 , 2.5 mM Ribose 5-phosphate (Ri5P), 2.5 mM NADP , 10 U phosphoriboisomerase, 10 U phosphoglucoisomerase, and 10 U glucose phosphate dehydrogenase with 10 U PHI or HPS, 2.5 mM HCHO and the protein extract in a total volume of 1 ml. The reaction was initiated by adding 50 ml of 50 mM HCHO. One unit of HPS or PHI activity was dened to correspond to the production of 1 mmol of NADPH per min at 30  C. HCHO-Tolerance experiments and measurements of gaseous HCHO uptake. Arabidopsis seeds were sown on agar medium with Km (50 mg/ml) and/or Gen (70 mg/ml). Resistant seedlings were selected, and transferred to fresh agar medium without antibiotics to allow them to recover their growth for one week. Then they were transferred to agar medium containing various concentrations of HCHO to test their resistance for 1 week. To test the transgenic tobacco plants, shoots were generated from leaf discs and further selected on agar medium with Km (100 mg/ml) and/or Gen (100 mg/ml). Resistant shoots were cut and transferred to agar medium containing HCHO and grown for 2 weeks. For measurements of gaseous HCHO resistance and uptake, plant seedlings were selected on agar medium with antibiotics. The seedlings

629

were then transferred into plant boxes (370 ml) and grown for 34 weeks. A microcentrifuge tube (500 ml) containing 50 ml HCHO (37%) was then placed in the corner of each plant box. Three to 4 weeks later, the cover of each plant box was quickly replaced with a new one with a hole in the center, and a HCHO detector (Gastec Passive Dositube, 91D, Ayase, Kanagawa, Japan) was set in the hole to detect gaseous HCHO. Three and a half hours later, the measurement was stopped and the results were photographed. The concentration of HCHO remaining in the box was estimated following the instructions of the supplier. Measurement of aqueous HCHO-uptake. For measurements of aqueous-HCHO uptake, plant seedlings were selected as described above and then grown on MS agar medium without antibiotics for 2 weeks. Plant seedlings (0.3 g fresh weight) were collected and submerged in 10 ml of water containing various concentrations of HCHO in a 50 ml-ask for 30 h at 20  C under light. Samples of the HCHO solutions were then taken for measurement of the HCHO concentration. A mixture of 50 ml of sample solution, 950 ml of H2 O and 1 ml of freshly prepared Nash reagent (15.0 g of ammonium acetate dissolved in 99.5 ml of H2 O, 0.3 ml of acetic acid, and 0.2 ml of acetone) was incubated at 30  C for 30 min.22) The HCHO concentrations were measured at 410 nm with pure HCHO as standard. Tracer experiments. For long-term tracer experiments, Arabidopsis seedlings were selected and grown on MS medium for 2 weeks. The plant seedlings (0.25 g) were collected and submerged in a solution containing 5 mM KHCO3 , 2 mM HCHO, 0.1% MES (pH 5.7), and [14 C]HCHO (Sigma, St. Louis, MO, 2.8 mCi) and incubated under constant light at 25  C overnight. After incubation, the plants were washed with cooled sterile water 45 times to remove free [14 C]HCHO from the surfaces. They were homogenized in liquid-nitrogen and extracted with 10% trichloroacetic acid (TCA). The insoluble fraction was obtained by centrifugation and washed once with 5% TCA, twice with 50% ethanol, and once with 100% ethanol by ltration. The radioactivity in the insoluble fraction was measured using a Clear-sol scintillation cocktail in a liquid scintillation counter. In short-term tracer experiments, seedlings (1 g of fresh weight) were immersed in a 20-ml solution containing 5 mM KHCO3 , 1 mM HCHO, 0.1% MES (pH 5.7), and [14 C]HCHO (Sigma, 2.0 mCi) and incubated under constant light at 25  C for 60 min. After incubation, the plants were washed with cooled sterile water 45 times to remove free [14 C]HCHO from the surfaces. The soluble metabolites in the plants were extracted with 80% v/v hot ethanol, then dried and redissolved in methanol (100 ml). A portion of the sample was subjected to radioactivity measurement, and the rest of the sample was analyzed by thin-layer chromatography (TLC). One-dimensional TLC was performed on thin-layer cellulose plates (Merck no. 1.05716) using butanol-propionic acid-water (equal parts of butanol-water, 370:25 v/v, and propionic acid-water, 18:22 v/v) as the solvent system as described by Pedersen et al.23) In [13 C]HCHO labeling experiments, the plants (1.5 g) were submerged in a solution containing 5 mM KHCO3 , 2 mM [13 C]HCHO (Cambridge Isotopes Laboratories, Andover, MA) and 0.1% MES (pH 5.7), and incubated under constant light at 25  C for 5 or 10 min. After incubation, the plants were washed with cooled sterile water 45 times to remove free [13 C]HCHO from the surfaces. They were homogenized in liquid nitrogen, then 2 ml of 100 mM potassium phosphate buer (pH 7.4) was added and the mixture was boiled for 3 min. Cellular debris was removed by centrifugation. The supernatant was frozen, lyophilized, and resuspended in 0.5 ml of the phosphate buer containing 5% D2 O, and placed in a 5-mm NMR tube. NMR analysis. [13 C]NMR analysis was performed (at Yunnan University, China) on a Bruker DRX 500-MHz instrument (Karlsruhe, Germany). Broad-band decoupling experiments were carried out for qualitative analysis of NMR spectra for extracts from the plants labeled for 5 min with [13 C]HCHO. The acquisition parameters included a 6.8-ms (30 ) pulse with broad-band proton decoupling, a spectral width of 32,679 Hz, an acquisition time of 0.5 s, and a decay time of 1.2 s. The sample temperature was maintained at 25  C, and 32,000 data points were acquired for each sample. Scans (n 1;200) were acquired for each sample, and a line broadening of 4 Hz was used in the processing of free induction decay.

630

L. C HEN et al.

A
NPTII

PrbcS TP PrbcS aacC1 TP

rmpA rmpB

NosT NosT

CK

B1

A1

AB1

AB2

rmpA RNA 720 bp rRNA

CK

A1

B1

AB1

AB2

rmpB RNA 700 bp rRNA

D
Mga A1 B1 AB1 AB2 CK HPS protein Mga B1 A1 AB1 AB2 CK PHI protein
Fig. 2. Generation of Transgenic Arabidopsis Plants Overexpressing HPS and PHI. A, DNA constructs used in the expression of the rmpA and rmpB genes and in targeting the HPS and PHI proteins to the chloroplasts. NPTII and aacC1 are the Km- and Gen-resistance genes respectively. PrbcS, the tomato rbcS-3C promoter; TP, the coding region for the transit peptide of tomato rbcS-3C; NosT, the nopaline synthase (Nos) terminator. B and C, Northern blot analysis of rmpA and rmpB gene expression in transgenic Arabidopsis plants respectively. Total RNA samples (5 mg each) from seedlings were analyzed. The full-length coding regions of the rmpA and rmpB genes were used as probes. The lower panels show ethidium bromide-stained rRNA in the agarose gels as loading controls. A1, a line of rmpA-transformed line; B1, a rmpB-transformed line; AB1 and AB2, two lines transformed with both rmpA and rmpB; CK, a negative control line transformed with the empty vector (pPZP211). D, Accumulation of HPS (upper panel) and PHI (lower panel) in chloroplasts of the transgenic Arabidopsis plants, as shown by Western blot analysis using rabbit antisera. Five mg of soluble protein extracted from the chloroplasts was applied to each lane. Experiments were repeated 3 times, and representative data are shown. Mga denotes crude cell extract (1 mg protein) from M. gastri MB19, which was used as a positive control. All lines were the same as used in Northern blot analysis as above. Quantitative analysis of [13 C]-enriched F6P and G6P in extracts from the plants labeled for 10 min with [13 C]HCHO were accomplished through inverse-gated decoupled experiments with a long decay. The parameters included a 2.5-ms (45 ) pulse, a spectral width of 31, 250 Hz, an acquisition time of 0.5 s, and a delay time of 20 s. Decoupling was achieved using the composite pulse Waltz-16 only during the acquisition period. The sample temperature was maintained at 25  C, and 32,000 data points were acquired for each sample. Twelve hundred scans were acquired for each sample, and a line broadening of 4 Hz was used in the processing of free induction decay. Resonances were assigned based on the chemical shifts of authentic samples and by comparing the observed chemical shifts with previously published values.24) Broad-band decoupling experiments were used to cross-check the quantities estimated from the inversedgated experiments. The peaks were calibrated relative to [13 C]HCHO sealed in a capillary and inserted in the sample tube.

Results
Expression of bacterial HCHO-assimilation enzymes in Arabidopsis chloroplasts Arabidopsis thaliana was transformed with two kinds of DNA constructs containing the rmpA and rmpB genes (Fig. 2A). Both of them were placed under the control of the tomato rbcS-3C promoter with its transit peptide sequence following Miyagawa et al.14) Transgenic lines (at least 20 of each) containing rmpA (named A1A20), rmpB (B1B20), or both rmpA and rmpB (AB1AB20)

were obtained, as described in Materials and Methods. Plants transformed with one of the two empty vectors, pPZP211 or pPZP221,16) were produced as controls. One line of the former transformants, named CK, was used as a control in this work. All the transformants appeared to be indistinguishable from wild-type plants in the vegetative and reproductive stages when grown either on an antibiotic-free agar medium or on soil. Seeds of each plant line were collected, and T2 generation plants were used in further experiments. Expression of the rmpA and rmpB genes in the transgenic plants was conrmed by Northern blot analysis (Fig. 2B, C), and by Western blot analysis using rabbit antibodies against the HPS and PHI proteins (Fig. 2D). The HPS and PHI proteins were found only in the extracts of chloroplasts puried by subcellular fractionation from transformants carrying the rmpA and/or rmpB genes respectively. Correct removal of the transit peptide in the chloroplasts was indicated by the molecular sizes of HPS (21 kDa) and PHI (25 kDa), which were identical to those of the original proteins. These results indicate that the HPS and PHI proteins were successfully expressed and targeted to the chloroplasts. A coupled spectrophotometric method21) was used to assay HPS and PHI activities in the leaf extracts,

Formaldehyde Assimilation in Transgenic Plants

631

A
A1

4 mM AB1

B
A1

8 mM AB1

C
A1

10 mM AB1

CK

B1

CK

B1

CK

B1

Fig. 3. HCHO Tolerances of the Transgenic Arabidopsis Plants. A, B, and C show the growth of transgenic Arabidopsis plants on agar medium containing 4, 8, and 10 mM HCHO respectively. Experiments were repeated many times, and representative data are shown. The transgenic lines are described in the legend to Fig. 2.

and it revealed easily detectable levels of enzyme activity in the transgenic plants. The HPS specic activities in the extracts from CK, A1, B1, AB1, and AB2 lines were 0, 0.24, 0, 0.28, and 0.20 mmol/min/mg protein respectively, and the PHI-specic activities from the CK, A1, B1, AB1, and AB2 lines were 0, 0, 0.20, 0.16, and 0.18 mmol/min/mg protein respectively. Enhanced tolerance of transgenic A. thaliana planted on agar media containing HCHO Since HCHO is highly toxic to plants, we speculated that transgenic plants engineered with the HCHOxation pathway would become tolerant to higher concentrations of HCHO than control plants. The HCHO tolerance of transgenic A. thaliana was investigated on agar medium containing HCHO. After germination, Km- and/or Gen-resistant seedlings of the A1, B1, AB1, and the CK transgenic plants were grown aseptically on agar medium without antibiotics for 7 d and then transferred to agar media containing various concentrations of HCHO. At 4 mM HCHO, no dierence was observed in growth among the seedlings (Fig. 3A). At 8 mM HCHO, the CK and B1 plants did not grow, and they showed chlorosis and wilting. On the other hand, about 50% of the A1 plants survived. This indicates that the production of HPS in chloroplasts enhanced HCHO tolerance to some extent. No growth inhibition was observed in the AB1 transgenic plants (Fig. 3B). In the presence of 10 mM HCHO, only the AB1 transgenic plants survived (Fig. 3C). Thus co-expression of HPS and PHI endowed the plants with the ability to detoxify HCHO more eectively than individual expression of HPS or PHI. Enhanced tolerance and uptake capacity of transgenic A. thaliana to gaseous HCHO To determine whether the AB transgenic plants would also show higher tolerance and uptake capacity as to gaseous HCHO, the AB1, AB2, and CK lines, which had been grown separately in sealed boxes, were exposed to HCHO evaporated from a micro test tube (0.5 ml) containing HCHO solution (37%, 50 ml) for about 3 weeks (Fig. 4). Within 1 week after the start of exposure, chlorosis and wilting occurred in the CK plants, but not in the AB1 or AB2 plants, indicating enhanced tolerance of the double transformants to gaseous HCHO. After 3 weeks, the covers of the plant boxes were removed and quickly replaced with covers

AB1

AB2

CK

> 20 ppm About 2 ppm About 2 ppm

Fig. 4. Gaseous HCHO Tolerance and Uptake Capacity of Transgenic Arabidopsis Plants. Details of the experiment are given in Materials and Methods. At the termination of the experiment, the remaining gaseous HCHO levels were roughly estimated using HCHO detector tubes (Gastec Passive Dositube, 91D). The two double-transgenic lines, AB1 and AB2, were used.

equipped with HCHO dose-monitoring tubes in order to determine roughly the gaseous HCHO levels remaining in the boxes. As shown in the lower part of Fig. 4, the concentrations were found to be less than 2 ppm for the AB1 and AB2 lines, but more than 20 ppm for the CK plants. These results indicate that the HCHO assimilation pathway strongly enhanced not only the tolerance of the transgenic AB plants to exogenous HCHO, but also their ability to take up and eliminate gaseous HCHO. Removal of exogenous HCHO by transgenic A. thaliana To test the ability of the transgenic plants to remove exogenous HCHO, experiments were performed as described by Achkor et al.7) The seedlings of AB1 and CK plants were suspended in a solution (0.2% MES,

632

L. C HEN et al.

Radioactivity of 14C in insoluble fraction (dpm), x 10-3

B
Radioactivity of 14C in soluble fraction (dpm), x 10-3 120
50 40 30 20

80

Spot 3 Spot 2

40

10

Spot 1

0 AB1 CK

AB1

CK AB1 CK

Fig. 5. Fixation of Radioisotope Labelled [14 C]HCHO by Transgenic Arabidopsis Plants. Seedlings were submerged in water containing [14 C]HCHO and additives, as described in Materials and Methods. A, Incorporation of [14 C] from [14 C]HCHO into the trichloroacetic acid (TCA)-insoluble fraction of plant materials after long-term incubation. B, Radioactivity incorporated into the soluble fraction after short-term incubation. The soluble fraction consisted of extracts with 80% hot ethanol. C, Autoradiogram of cellulose TLC for the soluble fraction of B. Experiments were repeated 3 times, and representative data are shown. The transgenic lines of AB1 and CK are described in the legend to Fig. 2.

5 mM KHCO3 , pH 5.7) containing various concentrations (10, 8, and 5 mM) of HCHO, and shaken for 30 h under light. The concentrations of HCHO decreased to 67% and 80% of the initial concentration of 10 mM, for AB1 and CK1 respectively. In the same way, decreases to 30% (AB1) and 69% (CK1) of the initial concentration of 8 mM, and the decreases to 9% (AB1) and 46% (CK1) of the initial concentration of 5 mM were observed. These results indicate that at lower initial concentrations, 5 and 8 mM, the amounts of HCHO eliminated by AB1 and CK1 were about 50 mmol and 25 mmol respectively. Thus the apparent capacity for HCHO removal of AB1 was about 2-fold higher than CK1. However, it must be kept in mind that the elimination of HCHO measured here might have been due not only to cellular metabolism (degradation and assimilation) but also to HCHO-adduct formation with various cell constituents.9) Augmented HCHO assimilation in transgenic A. thaliana To demonstrate the assimilation of HCHO through the engineered HCHO-xation reaction in the AB1 transgenic plants, a tracer experiment with [14 C]HCHO was conducted under the conditions described for the above experiment, except that the initial HCHO concentration was 2 mM. As a precaution, non-radioactive KHCO3 was included in the solution to dilute out the [14 C]CO2 that can be produced from [14 C]HCHO during incubation. After 18 h of incubation under light, the seedlings were washed and then ground in the presence of TCA. The radioactivity of the TCA-insoluble fraction was about 10-fold higher in the AB1 transgenic plants than in the CK control plants (Fig. 5A). Since HCHO-adducts are supposed to be decomposed by TCA-treatment, the radioactivity in this fraction can be ascribed to highmolecular-weight cell constituents. When short-term incubation (1 h) was carried out and the soluble fraction was prepared by hot-ethanol extraction, the radioactivity in this fraction was about 2-fold higher for the AB1 plants than for the CK controls (Fig. 5B). Cellulose TLC analysis of this fraction indicated that the radioactivity was incorporated into a discrete number of metabolite

groups (Fig. 5C). Among the three major spots in Fig. 5C, spot 1 had mobility similar to those of sugar monophosphates. Although no characterization of the spots was carried out, these results indicate that HCHO assimilation capacity was augmented in the transgenic plants as compared with the control plants. Functioning of the engineered HCHO xation pathway To conrm that the HCHO xation pathway was indeed established as a bypass of the Calvin-Benson cycle by overexpression of HPS and PHI in the transgenic plants, the [13 C]nuclear magnetic resonance (NMR) technique was used to detect the labeled intermediates in the Calvin-Benson cycle after the A1, B1, AB1, and CK plants were incubated with [13 C]HCHO for 5 min as described above. Labeled metabolites were extracted from the plants with phosphate buer and then subjected to [13 C]NMR analysis. The [13 C]NMR spectra are shown in Fig. 6. An extract of wild-type (WT) Arabidopsis plants without [13 C]HCHO treatment was used to monitor the background [13 C]NMR signal levels of the intermediates which is shown in Fig. 6E. The resonance signal peaks corresponding to the initial HCHO xation product, [1-13 C]F6P, with chemical shifts at 63.77, 64.05, and 64.76 ppm were observed in the AB1 spectra (Fig. 6A), but no such signals were detected in the spectra of the CK (Fig. 6B), A1 (Fig. 6C), or B1 (Fig. 6D) extracts. Signal peaks corresponding to the other intermediates, including [1-13 C]Ru5P, [1-13 C]ribose 5-phosphate (Ri5P), [3-13 C]sedoheptulose 7-phosphate (Su7P), [113 C]ribulose 1,5-bisphosphate (RuBP), [3-13 C]3-phosphoglycerate (3-PGA), and [1-13 C]xylulose 5-phosphate (Xu5P) with chemical shifts at 171.53, 98.77, 73.56, 68.46, 66.77, and 63.36 ppm respectively, also appeared in the AB1 spectra (Fig. 6A), no such signals were detected at the corresponding positions in the spectra of the CK (Fig. 6B), A1 (Fig. 6C), or B1 (Fig. 6D) extracts. Two specic resonance signal peaks with chemical shifts at 72.18 and 72.3 ppm, which appeared only in the spectra of AB1 (Fig. 6A) and A1 (Fig. 6C) extracts, might correspond to [1-13 C]Hu6P. In addition, part of the [1-13 C]F6P might have been converted to

Formaldehyde Assimilation in Transgenic Plants

Hu6P(?) 3-PGA F6P Hu6P(?) Ru5P Ru5P G6P Ref 3-PGA F6P F6P Xu5P Su7P RuBP Ri5P G6P G6P Ref

633

B
Hu6P(?)

Chemical shift (ppm)

Ri5P

Chemical shift (ppm)

Fig. 6. [13 C]NMR Analysis of the Metabolites of HCHO Assimilation by Transgenic Arabidopsis Plants. Complete [13 C]NMR spectra (left) and expanded regions of interest (right) of leaf extracts of AB1 (A), CK (B), A1 (C), and B1 (D) treated with [13 C]HCHO for 5 min, as well as leaf extract of WT without [13 C]HCHO treatment (E). Experiments were carried out in the manner described in Fig. 5, except for the use of [13 C]HCHO instead of [14 C]HCHO. For details, see Materials and Methods. The transgenic lines are described in the legend to Fig. 2. Peak assignments are as follows: Ref, reference; Ru5P, [1-13 C]Ru5P; Ri5P, [1-13 C]Ri5P; Su7P, [3-13 C] Su7P; RuBP, [1-13 C] RuBP; 3-PGA, [3-13 C]3-PGA; F6P, [1-13 C] F6P; Xu5P, [1-13 C] Xu5P; Hu6P, [1-13 C] Hu6P; G6P, [1-13 C] G6P.

[1-13 C]glucose 6-phosphate (G6P), which appeared as two strong resonance signal peaks, at 92.83 and 96.66 ppm in the AB1 spectra. G6P and F6P are known to be readily inter-convertible by the action of G6P isomerase. There were some background signals of [1-13 C]F6P and [1-13 C]G6P in the spectra of WT extract (Fig. 6E), presumably due to the natural abundance of 13 C in atmospheric CO2 . Similar levels of signal were also detected in the extracts of CK (Fig. 6B), A1 (Fig. 6C), and B1 (Fig. 6D). Hence, quantitative [13 C]NMR analysis was performed with extracts from the AB1 and CK plants incubated with [13 C]HCHO for 10 min to determine whether the [1-13 C]F6P and [113 C]G6P contents in the AB1 extract truly increased.24) The results indicated that the [1-13 C]F6P and [113 C]G6P levels in the AB1 extract were almost 3-fold higher than those in the CK extract (Fig. 7). This indicates that [13 C]HCHO was rapidly xed to F6P owing to ecient functioning of the installed HCHOxation pathway, and subsequently was incorporated into the expected positions of the intermediates in the Calvin-Benson cycle. The data indicats that the [13 C]HCHO was indeed assimilated in the engineered HCHO xation pathway, which functioned as a bypass of the cycle. Enhancement of HCHO tolerance in transgenic tobacco by co-expression of HPS and PHI To determine whether the strategy would work generally in other plants, the rmpA and rmpB construct was introduced into tobacco. Ten lines of rmpA transformants (A1A10) and 15 lines of rmpB transformants (B1B10), and ten lines of double-gene transformants (AB1AB10) were obtained. Control lines (CK) were

F6P Carboh 13 (g/g fresh wt)

2.5 2 1.5 1 0.5 0

G6P

Carboh 13 (g/g fresh wt)

6 5 4 3 2 1 0 AB1 CK

AB1 CK

Fig. 7. Quantitative Analysis of [1-13 C] F6P and [1-13 C] G6P in Cell Extracts from the AB1 and CK Seedlings after Incubation with [13 C] HCHO (2 mM) for 10 min. Experiments were carried out as described in Materials and Methods. The transgenic lines are described in the legend to Fig. 2.

transformed with the empty vector pPZP211. Expression of the designated genes in the tobacco transformants was conrmed by Northern blot analysis, Western blot analysis, and enzyme assay in the same manner as described above for Arabidopsis (data not shown). These transgenic tobacco plants were treated with HCHO, and the results for representative lines (A1, B7, AB7, and CK) are shown in Fig. 8. Shoots of these lines were cut and transferred to agar medium containing 10 mM HCHO and grown for 2 weeks. Control and B7 plants (Fig. 8A and B) were severely damaged by HCHO, and root growth was almost completely inhibited. The A1 shoots were damaged by HCHO (Fig. 8C), but not as severely as the control and B7 shoots. A1 root growth

634

L. C HEN et al.

+ HCHO A1

- HCHO

B7

D
A1 X B7

Fig. 8. Tolerance of Transgenic Tobacco Plants to HCHO in Agar Medium. Two-week-old shoots of tobacco transformants were transplanted to agar medium containing 10 mM HCHO, and 2 weeks later photos were taken. A, B, C, and D denote lines CK, B7, A1, and AB7 respectively.

Fig. 9. Enhanced HCHO-Tolerance of a Cross-Progeny (A1 B7) between the A1 (rmpA, male) and B7 (rmpB, female) Transgenic Lines. Seedlings of each line were grown on agar medium in a plant box for 3 weeks, and then a 500 ml-microtube containing 20 ml of 37% w/v HCHO was placed in each box. Transgenic plants grown without exposure to HCHO are also shown as controls. Photographs were taken after 2 weeks of exposure.

was inhibited, but not completely, by HCHO. The AB7 shoots were not signicantly damaged by HCHO (Fig. 8D), and they produced strong root systems and thus exhibited vigorous growth. These results were similar to those observed with transgenic Arabidopsis, indicating that the RuMP pathway functioned as expected in the transgenic tobacco. In order to demonstrate a synergistic eect of the co-expression of HPS and PHI on HCHO tolerance, it is preferable to prepare lines of double gene transformants by cross pollination between the single gene transformants rmpA (A1) and rmpB (B7). In the resulting progeny, the dosage of each gene cannot exceed the dosage of either parent. Line B7 was pollinated with line A1, and cross progenies A B were obtained. A line named A1 B7, which must have had both genes as determined by its tolerance to the two antibiotics for selection, was used in the experiment. Seeds of line A1, line B7, and line A1 B7 were sown on agar medium and grown for 3 weeks, and then a microtube (0.5 ml) containing 20 ml of 37% HCHO was placed in each sealed plant box and HCHO was allowed to evaporate. After 2 weeks, line A1 and line B7 showed a marked tendency to chlorosis and growth inhibition, while line A1 B7 showed only slight chlorosis and better growth (Fig. 9). Radioactive tracer experiments were also performed for transgenic tobacco plants, and similar data to those for transgenic Arabidopsis were obtained (data not shown). These results strongly indicate that the functioning of the RuMP pathway plays a key role in conferring HCHO tolerance on plants.

Discussion
The present study demonstrates, for the rst time, the integration of a bacterial HCHO xation pathway into the Calvin-Benson cycle in transgenic plants. It is quite

feasible that this pathway functions in plant leaves, because: (i) Gaseous HCHO is assumed to diuse into the leaves through the stomata eciently, as is the case for NO2 ,25) (ii) the large metabolic ow of the CalvinBenson cycle provides for a sucient supply of the acceptor, RuMP and prompt removal of the product, F6P, and (iii) no additional energy is required. Therefore the engineered HCHO xation pathway can be considered a bypass of the CO2 -xing reaction in photosynthetic plants (see Fig. 1). Overexpression of a glutathione-dependent HCHO dehydrogenase (FALDH) has been reported to enhance HCHO elimination in Arabidopsis.7) We produced the same transgenic Arabidopsis plants reported by that group and compared their tolerance to HCHO. In contrast to our transgenic plants expressing HPS and PHI, the FALDH overexpressing plants did not grow on an agar medium containing 6 mM HCHO (data not shown), indicating that no or very weak tolerance to HCHO can be conferred by the introduction of this gene. More recently, an adverse eect of FALDH overexpression on plant growth was reported by the same research group (ref. 7).26) This FALDH system requires, in addition to the cofactor glutathione, regeneration systems for the reduced glutathione and NAD , whereas the RuMP pathway for HCHO xation described here does not require any additional energy (it is exergonic) or cofactors,12) and the growth of the transgenic plants was not inhibited in any way. Furthermore, the regeneration pathway of Ru5P is inherent in plants. Hence we believe that the RuMP pathway is a more suitable HCHO elimination system in plants. Our preliminary experiments with larger tobacco (about 90 cm in height) in 300-liter chamber indicated that the rate of HCHO absorbance was about 20% faster

Formaldehyde Assimilation in Transgenic Plants

635

for the transgenic tobacco (AB) than the wild-type plants.27) If this strategy can be applied to indoor houseplants, the resulting genetically modied plants can be used to maintain a healthy household environment, and may be more favorably accepted by the public than genetically engineered foods. It is also noteworthy that there are no direct routes to produce biomass from reduced C1 compounds such as methane (a greenhouse gas), methanol, or HCHO in plants. Therefore, the RuMP-plant system may provide a novel photosynthetic platform for the assimilation of these natural gas-derived C1 -compounds and biogas emitted from various plants.

3)

4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17) 18) 19) 20) 21) 22) 23) 24) 25) 26) 27)

Acknowledgments
This work was supported in part by grants from the program Research for Promoting Technical Seeds of the Japan Science and Technology Agency (to K.I.), and from Project Research in the School of Biology-Oriented Science and Technology of Kinki University (to K.I.), by a grant from the Japan Society for the Promotion of Science under the FY2002 JSPS Postdoctoral Fellowship for Foreign Researchers (to L-M.C.), by the foundation (2004PY01-5) of Yunnan province and Kunming University of Science and Technology for training adult and young leaders of Science and Technology in Yunnan and by the foundation (30670163) of the National Natural Science Foundation of China (to L-M.C.), and by Grants-in-Aid for Scientic Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. Y.S. was also supported by the Asahi Glass Foundation. We are grateful for the technical assistance of Mr. Minori Kira of Kinki University in part of this study. Thanks to Dr. S. Shigeoka, to Dr. M. Tamoi, Kinki University, to Dr. F. Sato and Dr. K. Ifuku, Kyoto University, and to Dr. M. Sugita, Nagoya University for valuable suggestions.

References
1) 2) Teng S, Beard K, Pourahmad J, Moridani M, Easson D, Poon R, and OBrien PJ, Chem. Biol. Interact., 130, 285296 (2001). Naya M and Nakanishi J, Toxicol. Pharmacol., 43, 232248 (2005).

Im H, Oh E, Mun J, Khim J, Lee E, Kang H, Kim E, Kim H, Won N, Kim Y, Jung W, and Sul D, J. Proteome Res., 5, 1354 1366 (2006). Wolverton BC, McDonald RC, and Warkins EAJr, Econ. Bot., 38, 224228 (1987). Wolverton BC, "How to Grow Fresh Air, Penguin Books, New York, pp. 2029 (1996). Wolverton BC and Wolverton JD, J. Miss. Acad. Sci., 38, 1115 (1993). Achkor H, Daz M, Fernandez MR, Biosca JA, Pares X, and Martnez MC, Plant Physiol., 132, 22482255 (2003). Giese M, Bauer-Doranth U, Langebartels C, and Sandermann HJr, Plant Physiol., 104, 13011309 (1994). Hanson AD, Gage DA, and Shachar-Hill Y, Trends Plant Sci., 5, 206213 (2000). Yurimoto H, Kato N, and Sakai Y, Chem. Rec., 5, 367375 (2005). Kato N, Yurimoto H, and Thauer RK, Biosci. Biotechnol. Biochem., 70, 1021 (2006). Yurimoto H, Kato N, and Sakai Y, Appl. Microbiol. Biotechnol., 84, 407416 (2009). Mitsui R, Sakai Y, Yasueda H, and Kato N, J. Bacteriol., 182, 944948 (2000). Miyagawa Y, Tamoi M, and Shigeoka S, Nat. Biotechnol., 19, 965969 (2001). Sugita M, Manzara T, Pichersky E, Cashmore A, and Gruissem W, Mol. Gen. Genet., 209, 247256 (1987). Hajdukiewicz P, Svab Z, and Maliga P, Plant Mol. Biol., 25, 989994 (1994). Clough SJ and Bent AF, Plant J., 16, 729734 (1998). Horsch RB, Fry JE, Homann NL, Eichholts D, Rogers SG, and Fraley RT, Science, 227, 12291231 (1985). Triboush SO, Danilenko NG, and Davydenko OG, Plant Mol. Biol., 16, 183189 (1998). Sakai Y, Mitsui R, Katayama Y, Yanase H, and Kato N, FEMS Microbiol. Lett., 176, 125130 (1999). Kato N, Methods Enzymol., 188, 397401 (1990). Nash T, Biochem. J., 55, 416421 (1953). Pedersen TA, Kirk M, and Bassham JA, Biochim. Biophys. Acta, 112, 189203 (1966). Yasueda H, Kawahara Y, and Sugimoto S, J. Bacteriol., 181, 71547160 (1999). Hanson PJ, Rott K, Taylor GEJ, Gunderson CA, Lindberg SE, and Ross-Todd BM, Atmos. Environ., 23, 17831794 (1989). Espunya MC, Diaz M, Moreno-Romero J, and Martnez MC, Plant Cell Environ., 29, 10021011 (2006). Sawada A, Oyabu T, Chen LM, Li KZ, Hirai N, Yurimoto H, Orita I, Sakai Y, Kato N, and Izui K, Int. J. Phytoremediation, 9, 487496 (2007).