Sugarcane plants were developed that produce the sucrose isomers trehalulose and isomaltulose through expression of a vacuole-targeted trehalulose synthase modified from the gene in ‘Pseudomonas mesoacidophila MX-45’ and controlled by the maize ubiquitin (Ubi-1) promoter. Trehalulose concentration in juice increased with internode maturity, reaching about 600 mm, with near-complete conversion of sucrose in the most mature internodes. Plants remained vigorous, and trehalulose production in selected lines was retained over multiple vegetative generations under glasshouse and field conditions.
Sugarcanes (Saccharum hybrids) are attractive candidates for metabolic engineering aimed at sustainable production of value-added biomaterials and feedstocks, particularly those derived from sucrose (α-d-glucopyranosyl-1,2-d-fructofuranose), the major storage product in sugarcane (Birch, 2007). Sucrose is unique as a non-reducing glucose–fructose disaccharide. It is the world’s most abundant sugar with a vital role as the transport carbohydrate in plants. Enjoyment of its sweet taste has been innate in humans for tens of thousands of years, possibly arising much earlier as an indicator of safety and nutritional value in foods (van der Poel et al., 1998).
Some microbes convert sucrose into structural isomers such as isomaltulose (palatinose, α-d-glucopyranosyl-1,6-d-fructofuranose) and trehalulose (α-d-glucopyranosyl-1,1-d-fructose), possibly to sequester the sugar in a form that confers an advantage against competing species (Börnke et al., 2001; Wu and Birch, 2005; Watzlawick and Mattes, 2009). These sucrose isomers have greater acid stability than sucrose, and they are naturally present in various sucrose-containing foods (Lina et al., 2002). For humans, they are nutritional sugars that combine the sweet taste profile of sucrose with potential benefits for consumers including acariogenicity and slower digestion into monosaccharides (Takazoe, 1989; Ooshima et al., 1991; Lina et al., 2002). Trehalulose is interesting because of very high solubility which can be an advantage in foods such as jams and for industrial purposes that call for highly concentrated sugar solutions.
Use of sucrose isomers is currently limited by the expense of microbial or enzymatic conversion from more abundant, plant-derived sucrose (Schiweck et al., 1991). Sucrose isomerase (SI) enzymes can convert dissolved sucrose to isomers without any need for cofactors, and the conversion can be remarkably complete because of the multi-step mechanism and the lower free energy of the isomers (Tewari and Goldberg, 1991; Veronese and Perlot, 1999; Zhang et al., 2003; Rhimi et al., 2008; Watzlawick and Mattes, 2009). The known SI enzymes from various microbes are all in the same structural family (TIM-barrel family 13 glycosyltransferases), but they vary in kinetic efficiency and in product ratios, ranging from fairly specific isomaltulose (IM) or trehalulose (TH) synthases to mixed-isomer producers and hydrolases producing mostly hexoses from sucrose under certain conditions. At present, the enzymes with highest reported product specificity are from Pantoea dispersa UQ68J (91% IM and 3% TH yield from sucrose at 30–35 °C; Wu and Birch, 2005) and ‘Pseudomonas mesoacidophila’ strain MX-45 (91% TH and 8% IM yield from sucrose at 20 °C; Nagai et al., 1994). The sequence and structure of the trehalulose synthase from the whitefly Bemisia argentifolii are unknown (Salvucci, 2003).
Cloning of the genes encoding microbial SIs opened the possibility for more efficient production of isomers by direct conversion from sucrose in engineered plants. However, plants do not readily metabolize the isomers, so efficient conversion of sucrose from the active metabolic pool in the cytosol of growing tissues impairs growth (Börnke et al., 2002a; Wu and Birch, 2007).
In potato tubers, low-level IM production (∼15 μmol/g fresh weight) was achieved by apoplastic targeting of SI (Börnke et al., 2002b), albeit with severe adverse effects on starch metabolism during storage (Hajirezaei et al., 2003). In sugarcane, high-level production of IM (up to 440 μmol/g fresh weight) without evident adverse effects was achieved by vacuolar targeting of the UQ68J SI with high IM product specificity (Wu and Birch, 2007). Therefore, we were interested to determine the potential for TH production in sugarcane engineered to express a vacuole-targeted MX-45 SI with high TH product specificity.
Results and discussion
High-level trehalulose accumulation in mature internodes under glasshouse conditions
Young plants screened after 2–3 weeks in pots in their first vegetative generation (VG1) showed variable sugar concentrations in juice from leaves, extracted by freeze–thaw treatment and centrifugation. UbiKN controls yielded up to 30 mm sucrose, 5 mm hexoses and no TH. UbiTS lines yielded up to 35 mm TH without evident reduction in endogenous sugars. Growth and development of the TH-producing lines were not visibly different from UbiKN controls at any stage during callus culture, plant regeneration and growth of the transgenic plants through VG1.
No TH was detectable in juice from mature internodes of sugarcane background genotype Q117 or in 35 tested UbiKN control lines harvested after 12 months of growth in the glasshouse. TH was detected at this time, at concentrations up to 542 mm, in 76% of 54 independent transgenic lines that had initially been selected for geneticin resistance after bombardment using coprecipitated pUbiKN and pUbiTS, which was designed for expression of vacuole-targeted Trehalulose synthase (TS). Total sugar concentrations varied widely from 320 to 688 mm among both control and UbiTS populations, with no discernable relationship to TH level.
Lines selected for highest TH levels again showed high levels (up to 616 mm TH with near-complete conversion from sucrose) in mature internodes of a subsequent vegetative generation harvested after 12–18 months in the glasshouse. IM was present at about 12% of total isomer levels, and there was substantial variability between (single stalk) replicates in residual sucrose and total sugar levels (Figure 1).
Growth and trehalulose production in the field
There was substantial variability in germination under field conditions of buds from glasshouse-grown stalks in their first vegetative generations after regeneration from callus. Transgenic lines with no detectable TH ranged from 50% to 80% germination, while lines with >300 mm TH in mature internodes ranged from 55% to 93% germination. These lines ratooned normally (formed shoots from basal nodes after removal of above-ground stalks) in both glasshouse and field conditions. This is not surprising as wild Saccharum species germinate well with very little stored sugar (Bull and Glasziou, 1963), and it is unlikely that sugarcanes selected by humans for sweetness use more than a tiny fraction of stored sugar for germination.
Eight out of nine tested UbiTS transgenic lines that accumulated >100 mm TH in mature-stalk juice in the glasshouse gave similar results at 11 months in their first field propagation. The exception gave no detectable TH in the field, possibly because of transgene silencing. Although the Ubi promoter is relatively resistant to the full effects of silencing in sugarcane, progressive and stochastic silencing effects have been observed for Ubi-reporter constructs (Birch et al., 2010). In field-grown UbiTS plants, TH concentration increased progressively down the stem profile (Figure 2). As in the case of IM-synthase transformants (Wu and Birch, 2007), this pattern most likely arises from (i) continuous delivery of TS expressed from the constitutive Ubi promoter (Zhang et al., 1999), (ii) high catalytic efficiency (Watzlawick and Mattes, 2009) allowing occasional TH production before TS inactivation in the hostile vacuolar environment (Gnanasambandam and Birch, 2004) and (iii) probable low capacity for TH metabolism in plants (Loreti et al., 2000). For higher yield and purity of this valued sugar, it will be useful to achieve patterns of developmental expression, compartmentation and TS enzyme stability yielding high TH content further up the harvested stalk profile.
In UbiTS transformants, IM comprised about 14% of the sucrose isomer content at all locations along the stalk, which is equivalent to the ratio from action of purified TS on sucrose at ambient temperatures of 25–30 °C (Nagai et al., 1994), indicating similar stability of IM and TH in sugarcane stalks. The total sugar accumulation pattern showed a rapid increase to about internode 10, which is typical of sugarcane, and reached impressive levels (around 800 mm in juice) by normal commercial standards (Figure 2). Among the eight lines with >100 mm TH, none showed under these field conditions the effect previously observed under glasshouse conditions in some IM-accumulating lines wherein substantial isomer concentrations were stacked on apparently undiminished sucrose concentrations (Wu and Birch, 2007).
In whole-cane juice, isomers reached up to 347 mm, comprising more than 40% of total sugars, with no clear relationship between isomer concentration and total sugar concentration in this first field propagation (Figure 3). Noting the variability between interspersed plots of the clonal Q117 controls, we do not think that this trial stage is necessarily a good indicator of total sugar accumulation potential. Initial growth of sugarcane in the field from thin glasshouse-derived planting material was weaker than that from stronger field-sourced setts. Experience from conventional breeding in sugarcane shows that competition effects between adjacent lines interfere more severely with biomass than with sugar concentrations (Stringer and Cullis, 2002). However, the quantitative effects of competition between adjacent plots and the rate of recovery from likely epigenetic tissue culture and glasshouse effects over initial field generations are not yet well understood.
Four lines with various levels of TH and total sugar in the first field trial were replanted and tested in a second field propagation. Those with lower total sugar concentration in the FP1 appeared to recover towards typical commercial levels, and TH concentration in whole-cane juice was maintained in the FP2 with the exception of line 2131 in which it apparently decreased (Figure 4). The transgenic lines were vigorous in appearance, but larger field plots in subsequent trial stages are required for reliable biomass yield estimates in sugarcane (Jackson and McRae, 2001).
This is the first account of high-level accumulation of a value-added biomaterial under field conditions following metabolic engineering of sugarcane. Expression of a single transgene for a vacuole-targeted sucrose isomerase resulted in accumulation of the high-value sugar trehalulose up to 600 mm in juice from mature internodes and 300 mm in whole-cane juice, in vigorous sugarcane plants grown over multiple vegetative generations under commercial field conditions. This result supports the potential for development of sugarcane as a sustainable platform for production of sugar-derived biomaterials, with clear opportunity for advances through tailored expression patterns, compartmentation and stability of associated enzymes.
Constructs and gene transfer
‘Pseudomonas mesoacidophila’ strain MX-45 (Nagai et al., 1994) was obtained from the International Patent Organism Depository, Tsukuba, Japan. The mutB gene (see GenBank Accession DQ304536) was PCR amplified using primers (ATATGGATCCACTAGTAATGGAGGAGGCCGTAAAGCCGGGCGCGCCATGGTG and AATAGAGCTCAGTGGTGGTGGTGGTGGTGCTTCACCTTGTAGATGCCG) to eliminate a 69-bp sequence encoding the presumed periplasmic leader peptide and to introduce a 6 × His C-terminal extension for potential use in immunochemical localization studies. Trehalulose synthase function was confirmed by expression from bacterial plasmid pET24b (Novagen, Madison, WI, USA). The TS-coding region was coupled to the N-terminal leader and vacuole compartmentation signal (ERsNTPP) from sweet potato sporamin and arranged for expression from the maize ubiquitin (Ubi-1) promoter in a construct (pUbiTS) analogous to that previously described for isomaltulose synthase (Wu and Birch, 2007). There are vacuoles of diverse function in various sugarcane tissues, as indicated by differences in uptake and/or hydrolysis of reporter compounds for pH and protease activity (Rae et al., 2009). ERsNTPP-reporter proteins do not accumulate in any non-lytic sugarcane vacuole type (Gnanasambandam and Birch, 2004). The isomaltulose levels reported previously are only feasible from targeting to the large sugar-storage vacuole in mature sugarcane stalk parenchyma (Wu and Birch, 2007).
The trehalulose synthase construct and selectable marker construct pUbiKN were coprecipitated onto tungsten microprojectiles and introduced into sugarcane (cultivar Q117) embryogenic callus, followed by selection for geneticin resistance and regeneration of transgenic plants, essentially as described previously (Bower et al., 1996).
Sugarcane growth and sugar analysis
Sugarcane cultivars are vegetatively propagated for both commercial and experimental purposes. In the case of transgenic lines, there may be epigenetic changes in both endogenous and transgene expression during the initial vegetative generations after plant regeneration from the callus tissue required for the gene transfer process (Lourens and Martin, 1987; Burner and Grisham, 1995; Hansom et al., 1999; Vickers et al., 2005). Therefore, we number the vegetative generations commencing with VG1 for the primary regenerated plants. It is also known that performance at harvest in plant and ratoon crops in field trials is affected by the quality of the planting material and as well as the genotype (Srivastava et al., 2006). Plants grown under glasshouse conditions commonly produce thinner stalks (and hence setts for replanting) which result in visibly weaker initial growth in field plantings relative to field-sourced planting material. It can take one or more cycles of field propagation to obtain stalks of normal commercial size. Therefore, we number the cycles of field propagation commencing with FP1 for the first field planting of each line. In the work described here, there was a single glasshouse generation, so FP1 corresponds with VG2.
For glasshouse experiments, VG1 plants were grown as a single stalk in 4-L pots of commercial potting mix at 28 °C, under natural light conditions with daily watering and with 5 g/pot of the controlled release fertilizer Osmocote® (Scotts, Pty Ltd, Baulkham Hills, NSW, Australia) applied every second month. Plants were grown to maturity, normally developing between 20 and 35 internodes by 12 months.
Transverse sections about 5 mm in height were collected from the middle of the second-lowest internode. Juice was extracted by crushing the whole section in a hydraulic arbour press (5 ton) then heated at 100 °C for 10 min to inactivate enzymes and separated from cell debris by two centrifugation steps of 15 min each at 18000 g.
Selected lines were vegetatively propagated for analyses in the following generations, in the glasshouse as described earlier and in some cases under field conditions at Kalamia in North Queensland. In the field, each plant was allowed to form a stool with multiple stalks. Leaves were discarded at harvest, and the most mature stalk from each of four stools was pooled for extraction of whole-cane juice in a small roller mill. In some cases, selected internodes were also sampled to examine sugar profiles down the stem, and these samples were processed as described earlier for glasshouse-grown canes. The internode below the node bearing the leaf with the top visible dewlap (TVD) was numbered one, with higher numbers for older internodes (Van Dillewijn, 1952).
The sugar composition of extracted juice was determined by isocratic HPLC at high pH with pulsed electrochemical detection as previously described (Wu and Birch, 2007).
The authors acknowledge the excellent technical assistance of Ms Wendy Chan throughout this project. We thank Dr J. Stiller, Ms S. Lee-Goh, Ms K. Pirlo and Mrs E. Keresztné-Határvölgyi for their contributions in the preparation of the TS construct and its transfer into sugarcane. We thank Dr S. Basnayake (UQ), Mr T. Morgan (CSR) and members of the CSR Technical Field Department for assistance with plot design, field planting and sampling. This research was supported under the ARC Linkage Scheme in a collaboration between CSR Sugar Limited and The University of Queensland.