• Open Access

Production of cyanophycin, a suitable source for the biodegradable polymer polyaspartate, in transgenic plants


  • Katrin Neumann,

    1. Agrobiotechnologie, Universität Rostock, Justus-von-Liebig-Weg 8, 18059 Rostock, Germany
    Search for more papers by this author
    • These authors contributed equally to this project.

  • Dirk Paul Stephan,

    1. Biologie VIII: Molekulare Zellphysiologie, Universität Bielefeld, Universitätsstraße 25, 33615 Bielefeld, Germany
    Search for more papers by this author
    • These authors contributed equally to this project.

  • Karl Ziegler,

    1. Biochemie der Pflanzen, Humboldt-Universität zu Berlin, Chausseestraße 117, 10115 Berlin, Germany
    Search for more papers by this author
    • These authors contributed equally to this project.

  • Maja Hühns,

    1. Agrobiotechnologie, Universität Rostock, Justus-von-Liebig-Weg 8, 18059 Rostock, Germany
    Search for more papers by this author
  • Inge Broer,

    Corresponding author
    1. Agrobiotechnologie, Universität Rostock, Justus-von-Liebig-Weg 8, 18059 Rostock, Germany
      Correspondence (fax 0049 381 4984082; e-mail: inge.broer@uni-rostock.de)
    Search for more papers by this author
  • Wolfgang Lockau,

    1. Biochemie der Pflanzen, Humboldt-Universität zu Berlin, Chausseestraße 117, 10115 Berlin, Germany
    Search for more papers by this author
  • Elfriede K. Pistorius

    1. Biologie VIII: Molekulare Zellphysiologie, Universität Bielefeld, Universitätsstraße 25, 33615 Bielefeld, Germany
    Search for more papers by this author

Correspondence (fax 0049 381 4984082; e-mail: inge.broer@uni-rostock.de)


The production of biodegradable polymers in transgenic plants in order to replace petrochemical compounds is an important challenge for plant biotechnology. Polyaspartate, a biodegradable substitute for polycarboxylates, is the backbone of the cyanobacterial storage material cyanophycin. Cyanophycin, a copolymer of l-aspartic acid and l-arginine, is produced via non-ribosomal polypeptide biosynthesis by the enzyme cyanophycin synthetase. A gene from Thermosynechococcus elongatus BP-1 encoding cyanophycin synthetase has been expressed constitutively in tobacco and potato. The presence of the transgene-encoded messenger RNA (mRNA) correlated with changes in leaf morphology and decelerated growth. Such transgenic plants were found to produce up to 1.1% dry weight of a polymer with cyanophycin-like properties. Aggregated material, able to bind a specific cyanophycin antibody, was detected in the cytoplasm and the nucleus of the transgenic plants.


Transgenic plants can be utilized to produce renewable resources. This approach is a CO2-neutral, environmentally acceptable and competitive way to supply raw materials for industrial purposes. The production of renewable resources in transgenic plants can be achieved by altering the ratios of their natural ingredients, such as amylose and amylopectin in potato tubers in order to produce large amounts of the industrially interesting amylopectin (Kull et al., 1995). Furthermore, plants can be modified to produce new materials normally isolated from animals or bacteria, such as the spider silk protein (Scheller et al., 2001), or to produce oral vaccines (Richter et al., 2000; Kim and Langridge, 2003).

An important aim is the production of biodegradable polymers in genetically modified plants. A well-known example is the synthesis of polyhydroxybutyrate (PHB), a thermoplastic with properties similar to polyethylene and polypropylene. The genes necessary for the synthesis of PHB have been isolated from Ralstonia eutropha and successfully transferred to a variety of plants, such as Arabidopsis thaliana, Beta vulgaris, Brassica napus, Nicotiana tabacum and Solanum tuberosum (Poirier et al., 1992; Steinbüchel et al., 1992; Nawrath et al., 1994; Houmiel et al., 1999; Bohmert et al., 2000, 2002). In order to obtain significant amounts of PHB, the three enzymes of the entire pathway must be targeted to the plastids of A. thaliana and expressed in a coordinated manner (Bohmert et al., 2000; Menzel et al., 2003). However, in transgenic plants producing high levels of the polymer (up to 40% dry weight), severe negative effects on growth and development have been observed (Bohmert et al., 2000, 2002).

Poly-amino acids, such as poly-γ-glutamate, polyaspartate and poly-ɛ-lysine, are used as dispersants, thickeners or additions in hydrogels (Oppermann-Sanio et al., 1999; Oppermann-Sanio and Steinbüchel, 2002). Poly-γ-glutamate and poly-ɛ-lysine are synthesized by bacteria and can even be isolated from culture medium. Polyaspartate is a soluble, non-toxic, biodegradable polycarboxylate (Tabata et al., 2000) used in a number of industrial, agricultural and medical applications (Schwamborn, 1996; Oppermann-Sanio et al., 1999; Joentgen et al., 2001; Zotz et al., 2001; Oppermann-Sanio and Steinbüchel, 2002), complementing the application range of PHB. No organisms have been identified that produce polyaspartate. Hence, it has to be produced by chemical synthesis (Schwamborn, 1996). However, homo- and copolymers of polyaspartate can be obtained from the natural storage compound cyanophycin (multi-l-arginyl-poly-l-aspartic acid; Figure 1a) (Simon, 1971, 1987; Allen, 1988) by hydrolysis under mild conditions (Joentgen et al., 2001). Cyanophycin is synthesized by many cyanobacteria and some non-cyanobacterial eubacteria (Krehenbrink et al., 2002; Ziegler et al., 2002) via non-ribosomal polypeptide synthesis. The building block (‘monomer’) of this polymer is β-aspartylarginine (Figure 1a). Only one enzyme, the cyanophycin synthetase, that is encoded by a gene named cphA, is necessary for cyanophycin synthesis. This enzyme catalyses the formation of the peptide bonds linking the aspartate residues of the backbone, as well as the formation of the isopeptide bonds that connect arginine to the β-carboxy groups of the aspartate residues. Cyanobacterial cyanophycin is polydisperse (25–125 kDa), water-insoluble and stored in granules without membrane. However, a highly water-soluble polymer with properties very similar to cyanobacterial cyanophycin has been produced in Escherichia coli expressing a cyanophycin synthetase gene of Desulfitobacterium hafniense (Ziegler et al., 2002).

Figure 1.

(a) Schematic structure of cyanophycin. The branched polypeptide consists of a poly-α-aspartate backbone to which arginine residues are connected via isopeptide bonds. (b–f) Electron micrographs of cyanophycin granules in Thermosynechococcus elongatus BP-1 cells harvested in the stationary growth phase (b–d) and of cyanophycin granules in Escherichia coli cells transformed with a construct containing the cyanophycin synthetase gene of Thermosynechococcus elongatus BP-1 (e, f). Ultrastructural investigation after glutaraldehyde/osmium tetroxide fixation (b, e) and immunocytochemical detection of cyanophycin by the anti-cyanophycin antiserum and gold-coupled anti-rabbit immunoglobulin G (IgG) (c, d, f). Bar = 0.5 µm; Arg, arginine; Asp, aspartic acid.

Here, we show that the production of water-soluble cyanophycin is possible in transgenic tobacco and potato plants expressing the coding region of the cyanophycin synthetase gene from Thermosynechococcus elongatus BP-1 under the control of the cauliflower mosaic virus (CaMV) 35S promoter.


Cloning of the cyanophycin synthetase gene

A chimeric gene construct containing the cyanophycin synthetase coding region (cphA) from T. elongatus BP-1 (Berg et al., 2000) fused to the CaMV 35S RNA promoter and transcription termination signal (Baulcombe et al., 1986; Eckes et al., 1989) was introduced into a plant transformation vector to generate p35ScphA. In Figure 1, electron micrographs of cyanophycin granules in T. elongatus are shown. Transgenic E. coli harbouring p35ScphA were shown to produce granules of a dense material that was identified by its reaction with an antibody specific for cyanophycin (Figure 1) and by biochemical analyses (data not shown), thus demonstrating the correct function of the chimeric gene.

Generation of transgenic tobacco plants

The plants were produced by leaf disc transformation with Agrobacterium tumefaciens LBA4404 harbouring p35ScphA. The transformation experiments were repeated twice and the rates of regeneration were compared with those of the plant transformation vector containing only the nptII gene (Herrera-Estrella et al., 1983). The mean rate of regeneration was 86% relative to that of the control. The transgenic plants showed a significantly reduced growth rate. Most of the transgenic plants grown for approximately 8 weeks started to develop a modified phenotype consisting of variegated leaves that were thicker than the leaves of control plants. Moreover, the chloroplasts in the cyanophycin-producing plant cells differed morphologically from the wild-type chloroplasts. They contained both fewer and smaller grana stacks, indicating that cyanophycin production might cause stress of a yet unknown nature for the plant (Figure 2).

Figure 2.

Changes in leaf morphology (left) and in the appearance of chloroplasts (right) observed for most of the transgenic tobacco lines expressing the cphA gene from Thermosynechococcus elongatus BP-1 and producing soluble cyanophycin. (a, b) Non-transgenic control. (c, d) Transgenic lines expressing the cphA gene from T. elongatus constitutively in the cytoplasm. Bar = 0.5 µm.

The presence of the transgene-specific messenger RNA (mRNA) was demonstrated by reverse transcriptase-polymerase chain reaction (RT-PCR) in the majority of the regenerated plant lines (74% of 46 lines). The transgene-specific mRNA was detectable in all transgenic lines showing the modified phenotype. According to Northern analyses, truncated forms of the cyanophycin synthetase mRNA dominated in most of the transgenic plants tested. Full-length transcripts were detected only in a few transgenic lines (data not shown). Hence, the transcript originating from a bacterial gene might be incorrectly processed in the plant, leading to reduced amounts of full-length mRNA.

Analysis of the cyanophycin content of transgenic tobacco plants

Most of the transgenic plants contained both a water-soluble and a water-insoluble form of cyanophycin. This material was isolated from aqueous extracts of lyophilized leaves by a published procedure (Ziegler et al., 2002), and analysed by sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by staining with Coomassie Brilliant Blue (Figure 3). The isolated polymer was polydisperse. Its maximum size of about 35 kDa corresponds to a degree of polymerization of approximately 125, taking β-Asp-Arg as the principal monomeric unit of the polymer (Figure 1a). Several defined smaller polypeptide bands were visible in such electropherograms (Figure 3). When incubated with the cyanophycin-specific, hydrolytic enzyme cyanophycinase (Richter et al., 1999), the polymer was rapidly degraded (Figure 3). Amino acid analyses showed that the polypeptide contained aspartate : arginine : lysine in a molar ratio of 1 : 1.05 : 0.1 (mean values of two determinations). The assay developed for the quantification of the polypeptide produced by the plants (see ‘Experimental procedures’) used two recombinant enzymes, cyanophycinase (Richter et al., 1999) for depolymerization and a plant-type asparaginase with isoaspartyl peptidase activity (Hejazi et al., 2002) to hydrolyse the depolymerization product, β-aspartylarginine. The latter enzyme has been shown to be specific for isoaspartyl peptides, being unable to hydrolyse α-aspartylarginine (Hejazi et al., 2002). This confirms that the arginine residues of the plant-derived material were connected to the polyaspartate backbone via isopeptide bonds, as in authentic cyanophycin. Thus, the size, composition and structure of the plant-derived polymer were very similar to those of cyanophycin produced by transgenic E. coli expressing active cyanophycin synthetase (Ziegler et al., 1998; Aboulmagd et al., 2000).

Figure 3.

Sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the polymer isolated from transgenic tobacco (a) and transgenic potato (b) plants constitutively expressing the cphA gene from Thermosynechococcus elongatus in the cytoplasm. +, Extract treated with cyanophycinase (0.5% by weight), a cyanobacterial enzyme degrading cyanophycin; –, extract not treated with cyanophycinase; 2, 14, 19, 21, different transgenic tobacco and potato lines; control, plant line harbouring only the nptII gene.

The polymer was detected in 15 of the 19 transgenic tobacco lines harbouring the transgene-specific RNA. All positive lines but one, which produced only very small amounts of polymer, showed the modified phenotype. The content of the polymer, quantified in eight samples from four transgenic lines (lines 21, 22, 24, 29), ranged from 21 to 83 mg/g soluble protein (2.1%−8.3%), the mean being 48.5 mg/g (4.85%). For eight transgenic lines, the polymer content per gram dry weight was determined (Table 1); the mean was 3.15 mg/g dry weight (0.315%). The maximum amount of cyanophycin was found in line 2 with a total amount of 1.14% dry weight (0.37% of the soluble and 0.77% of the insoluble form). This line showed a strongly reduced growth rate. Both forms of cyanophycin exhibited the same degree of polymerization.

Table 1.  Cyanophycin content of various transgenic tobacco and potato lines (n.t., not tested). The number of + signs indicates the degree of phenotypical changes
 Transgenic lineTotal content of cyanophycin (mg/g dry weight)Content of insoluble cyanophycin (mg/g dry weight)Content of soluble cyanophycin (mg/g dry weight)Phenotypical changes
Tobacco 211.447.733.71+++
14 0.7650.1350.63+
21 1.2635n.t.n.t.+
22 1.939n.t.n.t.
23 2.7255n.t.n.t.++
24 3.4875n.t.n.t.+
29 0.959n.t.n.t.++
39 2.661.431.23++
Potato19 2.3471.860.487+++
21 1.16831.16830+

All tobacco lines tested were fertile and produced seeds. The F1 generation of three transgenic lines was analysed further. Cyanophycin-producing seedlings were easily identified by the occurrence of variegated leaves and a slightly reduced growth rate. The cyanophycin content was measured in five descendants of each line. In line 14, small amounts of the water-soluble form were detected (one-sixth of that in the parental line). The mean cyanophycin content was 0.09 ± 0.05 mg/g dry weight. The descendants of line 2, which contained the highest amount of cyanophycin, lost their kanamycin resistance completely. From the 60 seedlings grown on medium without kanamycin, none exhibited the morphological changes associated with cyanophycin production. In this line, cyanophycin production seems to be lost in the next generation. In contrast, line 39 produced about half the amount of the water-soluble form relative to the parental line; however, the insoluble form was increased (1.2-fold). The mean cyanophycin content was 1.65 ± 0.9 mg/g dry weight.

Electron microscopic and immunocytochemical analysis

In electron micrographs, loosely packed aggregates with a size up to 1.5 µm were visible in the cytoplasm of transgenic tobacco plants (Figure 4). They consisted of strands differing in thickness and length that were lying on top of each other. The aggregates reacted with an antibody specific for cyanophycin. After staining plant leaf tissues for protein with the azo dye Naphthol Blue Black, the aggregates could be observed under a light microscope (Figure 4).

Figure 4.

Microscopic detection of cyanophycin in transgenic tobacco leaves. (a, b) Electron micrographs of the ultrastructure of cyanophycin (overview and magnification). (c) Light micrograph of a Naphthol Blue Black-stained cross-section of a tobacco leaf containing cyanophycin (marked by arrows). (d) Immunocytochemical detection of cyanophycin by the anti-cyanophycin antiserum and gold-coupled anti-rabbit immunoglobulin G (IgG). Filled bars = 0.5 µm; open bar = 10 µm; C, cytoplasm; Ch, chloroplast; Cy, cyanophycin; E, epidermis; PM, palisade mesophyll; SM, spongy mesophyll.

Production of cyanophycin in transgenic potato

Our goal is the production of cyanophycin in transgenic potato tubers, so as to allow the harvesting of this polymer from the residues of starch isolation. In order to ascertain the feasibility of cyanophycin production in potato and to compare the effects caused by this production with those observed in tobacco, constitutive expression of the foreign gene in the cytoplasm was achieved by leaf disc transformation using Agrobacterium tumefaciens LBA4404 harbouring p35ScphA. The transformation experiments were repeated twice. The mean rate of regeneration was 70% relative to that of the control plants transformed with vector containing only the nptII gene. As described for tobacco, a modified phenotype consisting of variegated leaves and a decelerated growth rate was observed in plants producing cyanophycin. The presence of the transgene-specific mRNA was demonstrated by RT-PCR in 77% of the 13 potato lines tested. As observed for tobacco, a soluble polymeric substance of up to 35 kDa, which was degradable by cyanophycinase, could be extracted from the leaves of four of the transgenic lines (Figure 3). The amount of small-sized fragments was reduced compared with that in tobacco. The polymer content was about 30 mg/g soluble protein (3%). This corresponds to 1.99 mg/g dry weight on average (0.199%; measured for lines 19 and 21; Table 1). Both the soluble and insoluble forms of cyanophycin were detected. The maximum amount of cyanophycin found in line 19 was 0.049% of the soluble and 0.19% of the insoluble form (dry weight), giving a total amount of 0.24%. By transmission electron microscopy, aggregates resembling those observed in tobacco, but with a size of up to 4 µm in diameter, were detected in both the leaves and tubers of transgenic potato plants. In some plants, these aggregates were present in both the cytoplasm and the nucleus (Figure 5), probably as a result of inclusion of part of the aggregates by the newly formed nuclear membrane after cell division. The aggregates located in both compartments could be stained with Naphthol Blue Black.

Figure 5.

Electron micrographs of cyanophycin in tubers of transgenic potato plants. (a) Detection of cyanophycin in the cytoplasm and in the nucleus. (b) Magnification of cyanophycin in the nucleus. (c) Magnification of cyanophycin in the cytoplasm. Filled bars = 0.5 µm; open bar = 2 µm; C, cytoplasm; Cy, cyanophycin; N, nucleus; Nc, nucleolus; S, starch.

Tubers were produced from three of the analysed plant lines. Those obtained from the line producing the most cyanophycin (line 19) did not display eyes and could not be propagated further. For the other two lines, plants were germinated from the tubers. In both plants, the content of cyanophycin was increased compared with that in the original transformant (up to 8.4-fold). These R1 plants produced normal amounts of tubers.

The presence of cyanophycin in the tuber could only be demonstrated by electron microscopy, which showed the occurrence of granules (Figure 5).


Only two transgene-encoded biodegradable polymers of technical use have so far been produced in plants (Poirier et al., 1992; Steinbüchel et al., 1992; Nawrath et al., 1994; Zhang et al., 1996; Houmiel et al., 1999; Bohmert et al., 2000, 2002). We have achieved the production of cyanophycin in transgenic tobacco and potato plants. From transgenic plants producing adequate amounts of cyanophycin, the polymer can be extracted, purified and converted to polyaspartate, for which technical applications exist (Schwamborn, 1996; Joentgen et al., 2001; Zotz et al., 2001). The constitutive expression of the cyanophycin synthetase gene originating from T. elongatus BP-1 resulted in the production of the polypeptide with a maximum degree of polymerization of approximately 125. Without optimization of the gene construct, up to 1.1% of the dry weight consisted of cyanophycin. In contrast with cyanobacteria, where only an insoluble form is present, in plants we identified both an insoluble and water-soluble form of cyanophycin. The molecular basis for the different solubility properties is not known (Ziegler et al., 2002).

The small amount of full-length transcripts and the presence of several short mRNAs seem to indicate that incorrect processing of the mRNA takes place. This has occasionally been observed when transgenes of bacterial origin are expressed in plants (Van Aarssen et al., 1995; Diehn et al., 1998; Haffani et al., 2000). Modifications of the transgene are likely to enhance mRNA stability and may thereby further increase the polymer content. Often, this strategy proves to be successful (Van Aarssen et al., 1995; Strizhov et al., 1996; De Rocher et al., 1998). The use of artificial genes encoding polyaspartate may be another possibility. Synthetic genes have been successfully employed for the production of the bioelastomer (GVGVP)121 in transformed E. coli (Guda et al., 1995), Aspergillus nidulans (Herzog et al., 1997) and transgenic tobacco plants, via the chloroplast (Guda et al., 2000) or nuclear genome (Zhang et al., 1996). However, whilst the bioelastomer is composed of a five-amino-acid repeat consisting of three different amino acids, polyaspartate contains only the one amino acid, for which just two codons exist. Consequently, an artificial gene coding for polyaspartate would be highly repetitive, rendering the possibility of correct expression in transgenic plants rather unlikely. Furthermore, polyaspartate is a polyanion, whereas cyanophycin is a zwitterionic molecule in which each arginine residue (see Figure 1a) carries two charges, a positive charge at its guanidino group and a negative charge at its α-carboxy group. There is one known case in which a polyanion with charge properties similar to those of polyaspartate, poly-γ-glutamate, has been produced by an organism without being secreted, namely in the nematocysts of Cnidaria. In these explosive organelles, poly-γ-glutamate, together with positively charged counterions, is used to build up a very high osmotic pressure (Szczepanek et al., 2002). For osmotic reasons, and because polyaspartate is a cation exchanger, it is considered to be unlikely that this polyanion could be produced and stored inside plant cells in large amounts.

A slight decrease in the regeneration rate in transformation experiments, a decelerated growth rate, variegated leaves and changes in chloroplast morphology appear to be consequences of cyanophycin production in plants. Such a negative effect of polymer accumulation on transgenic plants has also been described, and even to a stronger degree, for transgenic plants producing PHB (Bohmert et al., 2000, 2002). Hence, transformed cells producing large amounts of cyanophycin in the cytoplasm may not be able to grow, thus leading to the observed reduced rate of regeneration. The reasons for these side-effects could be: (i) the production of cyanophycin could deplete the amino acid resources of the plant; (ii) the presence of cyanophycin aggregates in the cytoplasm of the plant might interfere with the normal function of this compartment. The observed side-effects cannot be ascribed to the presence of cyanophycin in the nucleus, as this has only been demonstrated for potato tubers, and equal damage was observed in tobacco and potato leaves.

The amount of cyanophycin detected in the next generation compared with the T0 plant varied from line to line. Cyanophycin production was lost, reduced or even enhanced after meiosis, enabling the establishment of stably producing lines. The loss of cyanophycin production in line 2 was probably due to an inactivation of the transgene. Interestingly, the activity of the nptII gene present on the same T-DNA was lost as well. The deviation was relatively high between the descendants analysed for each line.

In contrast, the amount of cyanophycin in transgenic potato lines seemed to be increased in plants grown from tubers of the original transformants. This difference between tobacco and potato could be due to changes in the plant genome brought about by meiosis.

The production of PHB in transgenic plants led to similar problems that were solved, in part, by refinement of the expression conditions (Nawrath et al., 1994). Preliminary results have indicated that direction of the cyanophycin synthetase to the chloroplast via a transit peptide increases the polymer content significantly. Furthermore, these plants exhibit no obvious phenotypical changes. Taking this and the possible optimizations mentioned above into account, we hope to produce sufficient amounts of cyanophycin in transgenic plants in the future. Although, in recombinant E. coli strains, up to 24% cyanophycin per gram of cellular dry matter can be produced on a technical scale (Frey et al., 2002), the need for cost-intensive bioreactors could reduce the cost-effectiveness of the production procedure. In the case of transgenic potato tubers, the production as a by-product of starch isolation could be cost-effective despite the fact that, even after refinement of the expression conditions, the plants may never be able to produce such a large amount of cyanophycin. It will be necessary to determine the optimal potato cultivar to enable both high starch production and the synthesis of the polypeptide cyanophycin. A possible solution could be the expression of an additional tuber-specific amino acid biosynthesis gene, supplementing the amino acid pool in the tubers.

If the improvement in cyanophycin production on a DNA, RNA and protein level is successful enough to produce sufficient amounts of the polymer, the isolation of cyanophycin from the vestiges of starch extraction may then be a cost-effective way to obtain the raw material for the production of polyaspartate homo- and copolymers, i.e. biodegradable substitutes for polyacrylates (Schwamborn, 1996).

Experimental procedures

Growth conditions

The growth of T. elongatus BP-1 (formerly Synechococcus elongatus) was conducted as described in Michel et al. (1998). E. coli and Luria Broth Agrobacterium tumefaciens LBA4404 were grown on LB medium under standard conditions. For electron microscopic analysis, E. coli cells were grown three times for 24 h each on EMM E. coli minimal medium plates supplemented with 5 mm l-proline. Tobacco (N. tabacum) var. Petit Havanna SRI and potato (S. tuberosum) var. Désirée were grown on Murashige and Skoog (MS) medium (Murashige and Skoog, 1962) containing 30 g/L sucrose (tobacco) or 20 g/L sucrose (potato). The plants were cultivated in growth chambers at 24 °C/18 °C (tobacco) or 22 °C/18 °C (potato) with a 16 h/8 h light/dark cycle.

Vector construction, plant material and transformation

The 2.7 kb coding region of the cyanophycin synthetase (cphA) gene from T. elongatus BP-1 was isolated from pSEK#1 (Berg et al., 2000) after restriction with NdeI/SalI. The NdeI site was filled in to generate a blunt end. Together with a SalI/KpnI fragment containing the transcription termination signal of the CaMV 35S RNA gene, the fragment was inserted into the binary vector pLH9000 (Hausmann and Töpfer, 1999) digested with StuI and KpnI. The resulting vector pcphA was digested with SmaI, and the CaMV 35S RNA promoter (digested with EcoRI and filled in to generate a blunt end) was inserted to generate p35ScphA. The binary vector was introduced into Agrobacterium tumefaciens LBA4404 (Hoekema et al., 1983). N. tabacum cv. Petit Havanna SRI and S. tuberosum cv. Désirée were transformed using Agrobacterium tumefaciens-mediated gene transfer, as described in Wohlleben et al. (1988) and Düring et al. (1993).

Nucleic acid isolation and analysis

Genomic DNA and total DNA were extracted from young leaves using the Qiagen DNeasy Plant Mini-Kit and the Qiagen RNeasy Plant Mini-Kit, respectively, according to the manufacturer's instructions (Qiagen, Hilden, Germany). The presence of the transgene in the plant genome was established by PCR using primers amplifying an 856 bp fragment of the coding region (cyel1-fw, TTACTGGAGCATTCGGCGTCATA; cyel1-rv, TTCCACAATGACGGATTTGGAGA). RT-PCR, using an oligo-dT-primer to synthesize the cDNA and the primers described above, was performed to demonstrate the presence of the transgene-specific RNA in transformed plants. Northern analyses were carried out as described previously (Neumann et al., 1997), with the exception that the DIG-labelled DNA probes were constructed by PCR using the primers described above.


SDS-PAGE was carried out on slab gels as described previously (Ziegler et al., 1998). For analysis of the water-soluble form of cyanophycin, the separation gels contained 15% (w/v) acrylamide and 6 m urea.

Analysis of water-soluble and water-insoluble forms of the polymer

Lyophilized plant material, ground in a mortar in the presence of glass beads (0.25–0.5 mm in diameter), was extracted three times with 50 mm Tris/HCl (pH 8.1) by vortexing and centrifugation (10 min, 16 100 g). The total protein content of such extracts was determined as described by Ziegler et al. (1998). The combined pellets were analysed for insoluble cyanophycin as described by Simon (1973). The soluble form of the polymer was purified from the supernatants by heat treatment, proteinase K digestion and precipitation with ethanol (Ziegler et al., 2002). For amino acid analysis, the polymer was purified further. The dried ethanol precipitate was dissolved in a minimum volume of 25 mm ammonium acetate (pH 4.0) and loaded on to an EconoPac S cartridge (5 mL, Bio-Rad, Munich, Germany) equilibrated with this buffer. After washing with increasing concentrations of ammonium acetate (pH 4.0), the polymer was eluted with 0.5 m ammonium acetate (pH 4.0) and lyophilized. Hydrolysis with HCl and quantitative amino acid analysis were carried out as described previously (Ziegler et al., 2002). For routine quantification of the soluble polymer, the ethanol-precipitated material from 25–50 mg of lyophilized plant material was dissolved in 0.25 mL potassium phosphate (pH 7.4) and completely hydrolysed enzymatically with a mixture of recombinant cyanophycinase and isoaspartyl dipeptidase (Richter et al., 1999; Hejazi et al., 2002). The mass of cyanophycin was calculated from the aspartic acid content of the hydrolysates, determined enzymatically as in Hejazi et al. (2002).

Light and electron microscopic analysis

Electron microscopy and immunocytochemical procedures were performed as described previously (Michel et al., 1998; Stephan et al., 2000). Cells were fixed with either glutaraldehyde and osmium tetroxide (ultrastructural investigations) or glutaraldehyde alone (immunocytochemical investigations). The antiserum used was raised against the water-insoluble cyanophycin isolated from the cyanobacterium Synechocystis PCC 6803 (Stephan et al., 2000). The antiserum dilution was 1 : 100. As the second antiserum, a gold-coupled anti-rabbit immunoglobulin G (IgG) was used (dilution, 1 : 30). For light microscopic analysis, the samples were treated as for immunocytochemistry, but were embedded in Technovit 7100 resin (Heraeus Kulzer, Hanau, Germany). Slices of 1 µm were placed on microscope slides, dried and stained for 1–2 min with either 0.05% Naphthol Blue Black (Sigma-Aldrich, Taufkirchen, Germany) in 7% acetic acid or 0.1% Coomassie Serva Blue in 7% acetic acid, and processed overnight in 7% acetic acid. Preparations were examined on a Zeiss light microscope (Zeiss, Oberkochen, Germany).


We are indebted to Norika AG (Groß Lüsewitz, Germany) for supplying plant material. We thank Professor W. Wohlleben (University of Tübingen), Dr U. Kahmann (University of Bielefeld) and Dr H. Junghans (Norika AG) for many helpful discussions. This work was supported by grants from Bayer AG (Leverkusen, Germany) to W. Lockau and K. Ziegler and from the Bundesministerium für Verbraucherschutz, Ernährung und Landwirtschaft represented by Fachagentur für Nachwachsende Rohstoffe (Gülzow, Germany).