CRC SUGAR Industry Innovation through Biotechnology Research Project Final Report ALTERNATIVE SUGARS: NEW OPTIONS FOR THE SUGAR INDUSTRY Report prepared by Anne Rae and Jason Hodoniczky. Project 2b8 Project participants: Jason Hodoniczky, Gregory Robinson, Graham Bonnett, John Manners, Ulrike Kappler, Victoria Haritos, Anne Rae Contact: Anne Rae Senior Research Scientist CSIRO Plant Industry 306 Carmody Road, St. Lucia, Qld. 4067 Telephone: 07 3214 2379 Facsimile: 07 3214 2920 Email: Anne.Rae@csiro.au CONTENTS Page No 1.0  SUMMARY ........................................................................................................... 3  2.0  BACKGROUND .................................................................................................. 4  3.0  OBJECTIVES ...................................................................................................... 4  4.0  METHODOLOGY ............................................................................................... 5  4.1  Identification of candidate sugars............................................................. 5  4.2  Functionality of candidate sugars............................................................. 6  4.2.1  Assay for sweetness ........................................................................ 7  4.2.2  Assay for cariogenicity................................................................... 8  4.2.3  Assays for digestibility ................................................................... 8  4.3  Enzymatic synthesis of sugars................................................................... 8  5.0  RESULTS .............................................................................................................. 9  5.1  Selection of candidate sugars .................................................................... 9  5.2  Strategies for synthesis of candidate sugars .......................................... 11  5.2.1  Gentiobiitol ................................................................................... 11  5.2.2  Glucosyl sucrose ........................................................................... 11  5.2.3  Ketosucrose ................................................................................... 12  5.3  Physical and sensory properties of candidate sugars............................ 12  5.3.1  Sweetness....................................................................................... 12  5.3.2  Cariogenicity................................................................................. 13  5.3.3  Digestibility and probiotic activity.............................................. 14  5.3.4  Summary....................................................................................... 16  6.0  OUTPUTS........................................................................................................... 16  7.0  INTELLECTUAL PROPERTY:.................................................................... 17  7.1  Project IP .................................................................................................. 17  7.2  Sub-contracts ............................................................................................ 17  8.0  ENVIRONMENTAL AND SOCIAL IMPACTS: ................................... 17  9.0  EXPECTED OUTCOMES .............................................................................. 18  CRC Sugar Industry Innovation through Biotechnology Date (May/2010) 2 10.0  FUTURE NEEDS AND RECOMMENDATIONS .................................. 18  11.0  PUBLICATIONS ARISING FROM THE PROJECT .......................... 18  12.0  ACKNOWLEDGMENTS ................................................................................ 18  13.0  REFERENCES ................................................................................................... 19  14.0  APPENDIX 1 .................................................................................................... 19  15.0  APPENDIX 2 .................................................................................................... 19  16.0  APPENDIX 3 .................................................................................................... 19  17.0  APPENDIX 4 .................................................................................................... 19  18.0  APPENDIX 5 .................................................................................................... 19  19.0  APPENDIX 6 .................................................................................................... 19  20.0  APPENDIX 7 .................................................................................................... 19  21.0  APPENDIX 8 .................................................................................................... 19  22.0  APPENDIX 9 .................................................................................................... 20  CRC Sugar Industry Innovation through Biotechnology Page 2 Date (May/2010) 1.0 SUMMARY Sugarcane has a highly effective carbohydrate biosynthetic and storage metabolism that has facilitated its use for the production of sucrose. Sugars are increasingly seen as low-cost, renewable organic resources which can be modified to produce food ingredients and industrial raw materials. For the sugar industry, alternative sugars offer a means of diversification in an area close to the existing core business. However a major restriction to development of alternative products has been ownership of enabling intellectual property by third parties. This project aimed to identify alternative sugars with desirable commercial properties and capture the IP to enable their production. The initial phase of the project was a scoping study to identify novel, naturally- occurring sugars and enzyme systems that may be involved in their manufacture by collating information from the literature and patent databases. Sugars that occur naturally in sugarcane and closely related species were also examined for potential as higher value products. Preferred candidates were simple modifications of sucrose where the gene sequences encoding the enzymes were available and no prior IP existed. Four sugars with potential applications as alternative sweeteners or chemical feedstocks were identified. Two of the candidate sugars could be either purchased directly or made by chemical synthesis from a purchased precursor. The remaining two candidate sugars were not available commercially and could not be synthesised easily. We proposed to make these sugars by cloning and expressing the genes that encode the enzymes from their native sources and then using the enzymes to synthesise the novel sugars in vitro. Two enzymes were expressed and characterised. Although neither of these enzymes carried out the predicted reactions, both enzymes were new; one is a dehydrogenase and the other is a glucosidase acting on gluco-oligosaccharides. The potential value of any novel sugar depends on its physical and sensory properties. For application as an alternative sweetener, a novel sugar ideally needs to be as sweet as sucrose but offer health benefits, particularly low cariogenicity (tooth decay) and low calorie-yield. We developed methods that can be used in the laboratory to test industry-relevant properties of sugars, specifically sweetness, cariogenicity and digestibility. A set of commercially available sugars, including several alternative sweeteners, was used to test the assays and provide a comparison with the novel sugars. Sweetness relative to sucrose and glucose was determined by a two-way preference ingestion assay with Drosophila melanogaster (fruit flies). Production of acid by the oral bacterium Streptococcus mutans was used as an assay to detect potentially cariogenic sugars. Calorie yield of sugars was measured by assays for digestibility by yeast invertase and rat α-glucosidase/sucrase. We also tested whether the sugars were able to inhibit the digestion of sucrose, and whether the sugars could promote the growth of ‘healthy’ bacteria in the gut. The results showed that two alternative sugars derived from sucrose have the properties required for an alternative sweetener. We also identified a disaccharide which is sweet-tasting and able to inhibit the digestion of sucrose. Further research will be required to develop an economic production system for these candidate sugars. The tests developed in this project also identified some interesting relationships between sugar structure and sensory or nutritive properties. Further analysis of these relationships may allow design of new sweeteners with optimal properties. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 3 The outcome of this work is an improved ability to develop new sugar derivatives as alternative sweeteners. The information and tools developed by the project will assist future efforts to exploit new options for diversification in the sugar industry. 2.0 BACKGROUND Naturally occurring sugars are important as food ingredients, as feedstocks for fermentation processes such as ethanol generation, and as industrial raw materials. In the food industry, sugars add not only sweetness, but also colour, texture and preservation qualities. Sugars may be modified chemically to produce surfactants, emulsifiers and preservative coatings. With the projected decline in petrochemical- based materials, sugars are increasingly seen as low-cost, renewable organic resources for the chemical industry. Sugarcane has a highly effective carbohydrate biosynthetic and storage metabolism that has facilitated its use for the production of sucrose. The price of sucrose has fluctuated greatly in recent years and there is a desire in the industry to buffer these effects through diversification. Alternative sugars offer a means of diversification in an area close to the existing industry’s core business. For the Australian industry to compete in the alternative sugar product area we propose that it will be necessary to: (a) identify sugars with potential functionality that matches a commercial opportunity in the market (b) establish an IP position that will allow the protection and development of a process for the economical production of the sugar in either enzymatic, microbial or plant-derived production systems. The aim of this project was to identify novel sugars with desirable commercial properties and capture the IP to enable production. The project is well-aligned with the Program 2 aim of developing technologies for delivery of high-value materials from sugarcane. One of the major restrictions to development of sugarcane as a biofactory has been ownership of enabling intellectual property by third parties. This position will create new options for the sugar industry for diversification in the area of novel sugars. 3.0 OBJECTIVES The aim of the project was to identify and capture market opportunities for production of novel sugars. The specific aims of the project were: (i) Identify sugars of potential commercial interest AND (ii) Identify a novel source of the enzymes and clone the genes Achieved. The first two objectives were achieved by a scoping study in the initial part of the project. Novel naturally-occurring sugars and enzyme systems that may be involved in their manufacture were sought by collating information from the literature and patent databases. Preferred candidates were simple modifications of sucrose where the gene sequences encoding the enzymes were available and no prior IP existed. Four candidate sugars that met these criteria were identified: three sugars with potential applications as alternative sweeteners and one sugar as a potential chemical feedstock. Genes that were predicted to synthesise two of the sugars were cloned. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 4 (iii) Demonstrate production of the sugar Not achieved. One of the candidate sugars was purchased and a second was made under contract by a chemical synthesis company. The other two candidate sugars were not available commercially, and therefore we attempted to make them enzymatically. Two enzymes for synthesis were cloned and expressed in E.coli. Although the enzymes were active, they did not make the predicted sugars in vitro. (iv) Test the properties of novel sugars relevant to potential applications Achieved. The potential value of a novel sweetener depends on its physical and sensory properties, principally sweetness, digestibility and cariogenicity. Tests for these properties were developed and used to determine which sugars matched the characteristics of commercial sweeteners. (v) Protect the IP for exploitation Achieved. The IP produced by the project was examined carefully against the criteria of novelty and potential market value. Although some of the candidate sugars had the properties of a sweetener, no economic production system could be identified, making it unlikely that these new sugars would be competitive in the marketplace. Therefore, patent protection was not sought. (vi) Prepare a plan for further research and commercialisation Achieved. Some outcomes from this work have been approved for publication. One paper is already published, one submitted to a journal and one is still being prepared. The information on candidate sugars from the initial scoping study has not been disclosed and may become the subject of further research if opportunities arise. 4.0 METHODOLOGY 4.1 Identification of candidate sugars The initial phase of the project was a desktop exercise to identify novel sugars and enzyme systems that may be involved in their manufacture. Preferred candidates were simple modifications of sucrose where the gene sequences encoding the enzymes were available and no prior IP existed. We examined naturally occurring sugars in the categories of sucrose isomers, sucrose-based oligosaccharides, sugar alcohols and rare sugars. The criteria for selecting candidate sugars were: i. novel or rare sugar, not currently in commercial production ii. potential applications as low calorie sweeteners or as stereo-specific starting material for chemical synthesis of high value products iii. experimental tools such as gene sequences and a source of DNA are available One additional candidate came from a project in the first phase of the CRC where sugarcane plants engineered to make sorbitol were also found to contain a novel sugar identified as gentiobiitol (Fong Chong et al., 2007, 2009). This sugar was assessed for its potential value as a sweetener. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 5 For each candidate sugar, a detailed assessment was prepared, covering background, properties, feasibility, IP position and market potential. 4.2 Functionality of candidate sugars An industry partner, who requested confidentiality, provided information on market trends and on functional requirements for new sweetener products. Novel sugar products ideally need to be as sweet as sucrose but offer health benefits, particularly low cariogenicity and low calorie-yield. We developed methods that can be used in the laboratory to test industry-relevant properties of sugars, specifically sweetness, digestibility and cariogenicity. New collaborations with Professor Carol Morris at SCU and with Dr Elizabeth McGraw at UQ were important in developing these methods. The details of these methods are included in two draft publications (see Section 11). The methods are described briefly below. To test the assays and provide a comparison with the novel sugars, we have used a set of commercially available sugars, listed in Table 1. A series of sugar alcohols, sucrose isomers and other di- and trisaccharides were sourced for comparison to the candidate sugars. The set included several commercial sweeteners. Table 1 Summary of sugar structures used for comparison MW sugar name structural information sugar alcohols 152.15 xylitol reduced xylose 182.17 sorbitol reduced glucose 344.31 maltitol reduced maltose 344.32 gentiobiitol reduced gentiobiose 506.45 maltotriitol reduced maltotriose sucrose isomers 342.3 sucrose (glc-1,2-fru) 342.3 turanose (glc-1,3-fru) 342.3 leucrose (glc-1,5-fru) 342.3 palatinose (glc-1,6-fru) disaccharides 378.33 trehalose.2H20 (glc-1,1-glc) 342.2 kojibiose (glc-1,2-glc) 360.31 maltose.H20 (glc-1,4-glc) 342.3 gentiobiose (glc-β1,6-glc) 342.3 melibiose (gal-1,6-glc) 312.3 isoprimeverose (xyl-1,6-glc) trisaccharides 504.44 melezitose (glc-1,2-fruc-1,3-glc) 504.44 panose (glc-1,6-glc-1,4-glc) 504.44 maltotriose (glc-1,4-glc-1,4-glc) 504.44 erlose (glc-1,4-glc-1,2-fru) 504.44 1-kestose (glc-1,2 -fru-β1,2 -fru) 594.51 raffinose.5H20 (gal-1,6-glc-1,2-fru) candidate oligosaccharide all structures alpha linked unless indicated CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 6 4.2.1 Assay for sweetness There is no laboratory assay for sweetness and the taste profiles of new food products are determined by trained panels of human testers. Since the novel sugars, by definition, did not have food safety approval, human taste tests could not be used. However, tests in animal model systems can be a good substitute and are also useful when applying for approval from the regulatory authority, Food Standards Australia and New Zealand (FSANZ). It is important that the model animal used has similar preferences for sweet compounds to the human taste profile. Drosophila melanogaster (fruit fly) was identified as a species which has sweet taste preferences closer to humans than many other animals, including other mammals (Gordesky-Gold et al. 2008). We have established a two-way preference ingestion assay which enabled sweetness, relative to sucrose, to be determined. In developing this assay we collaborated with Dr Elizabeth McGraw at the University of Queensland, to access materials and methods for handling Drosophila. Initially a lab strain of Drosophila melanogaster was used. A wild strain was also captured locally and bred in case the lab strain had lower discriminating ability. In the assay (Figure 1), 96-well plates covered with Parafilm were prepared with droplets of 0.5% agarose containing a test sugar and a coloured dye (red or blue food dye). The droplets were presented in an alternating pattern and assays were prepared in duplicate with dye colours reversed to rule out possible colour preferences. Flies were kept on a 12 h light-dark cycle and were starved for 24 h prior to the assay. Approximately 50 flies per assay were then allowed to feed on the sugar solutions in darkness for 2 h. The flies were then killed by placing the dishes at -20oC for 48 h and the colour of each fly’s abdomen (Figure 1) was scored as red, blue, purple (feeding on both sugars) or clear (indicates no feeding). The results were expressed as a preference index (PI) which equals the number of flies in red/blue + 0.5 times the number of purple flies divided by the total number of flies feeding. Each assay was performed several times with fresh flies. Figure 1 Set-up and scoring of bioassay for sweetness based on two-way preference by fruit flies (Drosophila melanogaster). On the left is a 96 well plate covered with parafilm and prepared with alternating sugar solutions in 0.5 % agarose containing either a red or blue dye. After feeding for two hours, flies were scored by the colour of their abdomen; shown are magnified images of flies with clear, red and blue abdomens. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 7 4.2.2 Assay for cariogenicity Cariogenicity (tooth decay) is caused by a combination of microbial activities in the mouth. Early and late-colonising bacterial species use sucrose obtained from food for growth and produce dextrans, which form plaque, and acids, which attack tooth enamel. An oral isolate of Streptococcus mutans was obtained from the UQ Culture Collection and maintained at 37oC in brain-heart infusion (BHI) medium. Cells were resuspended in a medium that replicates saliva conditions and incubated with a sugar solution for 90 min at 37oC. The change in pH of the medium was then measured. A fall in pH indicates that S. mutans is able to use the sugar as a substrate for growth and produce acid. 4.2.3 Assays for digestibility Digestibility of sugars by oral and intestinal enzymes is an indicator of whether they are calorigenic. In developing these assays, collaboration with Professor Carol Morris and researchers in project 2b3 (“Bioactive Natural Products from Sugarcane”) was very valuable in defining the most appropriate enzymes and methods. Assays were developed with yeast invertase and rat α-glucosidase/sucrase as models for human salivary invertase and human intestinal glucosidase/sucrase respectively, dominant enzymes that metabolise sucrose. The assays were tested on the panel of reference sugars described above as well as on the two candidate sugars, turanose and gentiobiitol. Products of invertase digestions were monitored by HPLC. Intestinal glucosidase reactions were performed at 37oC and stopped by the addition of 1 M Tris. An aliquot was then used for the determination of glucose release using a hexokinase assay monitored by absorbance at 340 nm and calculated from a standard curve. As a control to demonstrate that glucose release was due to glucosidase activity, the specific inhibitor acarbose was included in the reaction (0.4 mg/mL). In addition to testing digestibility by glucosidase, we tested whether the sugars were able to inhibit the digestion of sucrose or isomaltose. The assay was performed as above, but with the addition of 15 mM isomaltose or 55 mM sucrose, both of which are digested rapidly by glucosidase. Acarbose was used in control reactions, as above. Percentage inhibition was calculated by comparison to digestion of either isomaltose or sucrose alone. Sugars that are not digested by intestinal enzymes may support the growth of beneficial (“probiotic”) bacteria in the gut. Plant & Food Research (New Zealand) have been contracted to assay the effect of a number of sugars on growth of Bifidobacterium lactis HN019 (DR10™) & Lactobacillus rhamnosus HN001 (DR20™). These assays are in progress and results should be known by the end of June 2010. 4.3 Enzymatic synthesis of sugars Gene sequences were synthesised by GeneArt Ltd. then subcloned into a protein expression vector. The initial vector system chosen utilised the T7 promoter with a 6 x His tag to produce high level expression of the protein of interest in E. coli. The tag allows purification on a nickel column and identification of the enzyme by western blots using a Ni-NTA-alkaline phosphatase detection system. A vector that includes the Trx sequence, which is known to increase solubility of cloned proteins, was also tested. The vectors were initially expressed in E. coli strain BL21 (DE3) and strains were either auto-induced or induced by the addition of IPTG (0.3 mM) for 1-3 h at room temperature. Proteins in both the soluble and insoluble fractions were CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 8 recovered, then separated on SDS-polyacrylamide gels, transferred to PDVF membrane and detected with the Ni-conjugate. Insoluble inclusion bodies were purified and refolding conditions were tested using a kit from Athena Enzyme Systems (QuickFold™). The system provides 15 combinations of refolding reagents in a factorial matrix design to identify key buffer components resulting in soluble protein. The identity and viability of novel glucosidase enzymes were tested in an assay with the substrate, nitrophenol-glucoside. In this assay, cleavage of the glucoside unit releases nitrophenol which is monitored by an increase in absorbance at 595 nm. Purchased yeast glucosidase was used as a positive control. Affinity purification using the Ni tag, ‘Talon’ affinity column was tested and the fractions analysed on SDS-polyacrylamide gels as above. A large-scale purification was carried out by the UQ Protein Expression Facility (PEF). 5.0 RESULTS 5.1 Selection of candidate sugars Candidate sugars were selected according to the criteria described above (Section 4.1). The three candidates identified by literature searches are shown in Table 2 and described below. A fourth candidate, gentiobiitol, was identified in another CRC project (see Section 4.1). Additional sugars that fulfill some of the criteria were also identified. In some cases, lack of tools such as sequences of the genes that synthesise the sugars prevented further work, or the IP position may have been weaker. These sugars could become viable candidates in the future if further information becomes publicly available. Sugars that occur naturally in sugarcane and closely related species were also examined for potential as higher value products. There was little information in the literature on the occurrence of rare sugars in sugarcane. In order to address this question, a CRC-funded vacation scholar, Louise Ryan, worked with the project for 6 weeks to survey the sugars present in 7 species. Donna Glassop was also involved in the analysis and identification of sugars, using her experience from the metabolomics work in CRC project 1ai “Genes for enhanced sucrose accumulation” (Glassop et al. 2007). The results showed that many soluble sugars are present in sugarcane and closely-related species but at such low concentrations that their extraction would not be commercially viable. The results were published: Glassop D., Ryan L.P., Bonnett G.D. and Rae A.L. (2010) The complement of soluble sugars in the Saccharum complex. Tropical Plant Biol. 3:110–122. (Appendix 1). CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 9 Table 2 Brief description of the three candidates selected from literature search Sugar Enzyme Source Application 1 α-1,2-glucosyl glucosyl cyanobacteria Non-digestible sucrose hydrolase/transfer (Anabaena/Nostoc) sweetener -ase 2 Turanose α-Glucosidase Honeybee (Apis Low calorie (sucrose mellifera) sweetener isomer) 3 3-ketosucrose Glucoside 3- Agrobacterium Chemical dehydrogenase tumefaciens feedstock 1. α-1,2-Glucosyl sucrose Glucans containing α-1,2-linkages are considered highly desirable as sweeteners because they are indigestible and acariogenic. Most commercially available gluco- oligosaccharides are produced by the dextransucrase enzyme from Leuconostoc and contain a mixture of α-1,2- and α-1,6-linkages linkages. Enzymes that catalyse solely α-1,2-linkages have not previously been described. However, the trisaccharide, formed by the addition of glucose to sucrose was recently identified as a product of the cyanobacterium Nostoc (syn. Anabaena). The novel sugar is probably synthesised by the activity of a glucosyl hydrolase/transferase enzyme. The availability of the full genome sequence of Nostoc allowed us to identify and clone the genes that may encode this enzyme. 2. Turanose Honey contains a mixture of sweet-tasting, low calorie or low cariogenic compounds derived from sucrose, including the sucrose isomer, turanose and oligosaccharides such as erlose and theanderose. These are thought to be synthesised by an α- glucosidase/transferase enzyme in the honeybee crop. The sequence of the enzyme is available and may provide a novel means of making these sugars. 3. 3-Ketosucrose This sugar was originally identified in extracts of Agrobacterium. The reports generated great interest because the introduction of a polymerisable double bond on the sugar molecule enables stereoselective addition of new functional groups. For example, derivatisation with vinyl side groups can be used to produce biodegradable polymers and latexes. Although the Agrobacterium genome sequence has been completed, it was not possible to identify the gene encoding the dehydrogenase enzyme due to lack of homologues for comparison. Recently, genes encoding glucoside-3-dehydrogenases in other species have been identified, enabling comparisons with Agrobacterium. This enzyme may open the market for new chemical products derived from sucrose. 4. Gentiobiitol Gentiobiitol is a disaccharide alcohol which is probably synthesised by transfer of glucose onto sorbitol by a β-glucosidase/transferase enzyme. The sugar is not in commercial production and the work by Fong Chong et al. (2010) is the first report of its biological synthesis. Disaccharide alcohols such as maltitol and isomaltitol are commercially produced for applications as low calorie sweeteners. Gentiobiitol would be expected to share some properties with these sugars, except that it may not be sweet, as gentiobiose is known to taste bitter. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 10 5.2 Strategies for synthesis of candidate sugars The production strategies used for the selected sugars are shown in Table 3. Amongst the candidate sugars, one could be purchased commercially (turanose) and one could be synthesised from a commercially-available sugar (gentiobiitol produced by the reduction of gentiobiose). The other two candidate sugars, glucosylsucrose and ketosucrose were not commercially available and we proposed to synthesise these by cloning and expressing the genes that encode the enzymes from their native sources. Table 3. Summary of strategies for producing selected novel sugars Target Sugar Potential Native source Production Strategy application Turanose Sweetener Honey bees Purchase Gentiobiitol Sweetener “Sorbitolcane” Chemical synthesis from produced by Fong gentiobiose Chong et al. in a previous CRC project Glucosyl sucrose Sweetener Anabaena/Nostoc Enzymatic synthesis cyanobacteria using gene cloned from Nostoc Ketosucrose Chemical Agrobacterium Enzymatic synthesis feedstock tumefaciens bacteria using gene cloned from Agrobacterium 5.2.1 Gentiobiitol The chemical synthesis company, Epichem Ltd. was contracted to synthesise gentiobiitol from 5 g of gentiobiose as starting material. The product was treated to remove impurities and analysed by 1H NMR spectroscopy. The product was delivered on 2 February 2009. Interestingly, the Epichem chemists found that they were unable to crystallise the gentiobiitol as the sugar was highly hygroscopic. This property would limit the applications of gentiobiitol as a sweetener. For example, gentiobiitol could not be used by the spoonful to replace sucrose, but could still be useful in manufactured products such as bottled drinks and baked goods, where its hygroscopic nature might be an advantage. 5.2.2 Glucosyl sucrose The trisaccharide, glucosyl sucrose was detected in the cyanobacterium, Nostoc, and is thought to be synthesised from sucrose by the action of a glucosidase enzyme. Two genes encoding putative glucosidases (aG1 and aG2) were identified, synthesised, and cloned into E.coli. Although the majority of the protein was recovered as insoluble inclusion bodies, significant amounts were soluble. Partially purified aG1 and aG2 enzymes were obtained by affinity chromatography. Assays with the artificial substrate, nitrophenol-glucoside, confirmed that the enzymes had glucosidase activity. Production of enzyme aG2 was scaled up by recloning into a high-expression vector, and a large-scale production and purification was carried out by the UQ Protein Expression Facility (PEF). The fractions were tested for activity against a range of substrates. The results suggested that while the enzyme is an active glucosidase, its CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 11 specificity is for longer chain glucosides such as malto-oligosaccharides and that it has little activity against sucrose in the present form. The detailed experimental results are shown in Appendix 2. This enzyme may be active against sucrose under conditions that were not tested, or alternatively, a different enzyme may be responsible for synthesis of glucosyl sucrose in Nostoc. Since no activity against sucrose was detected and the enzyme described here was not stable, no further purification and analysis was performed. 5.2.3 Ketosucrose Ketosucrose has previously been detected in the bacterium Agrobacterium tumefaciens and is thought to be synthesised from sucrose by the enzyme, glucoside- 3-dehydrogenase (G3DH). When the sequence of the Agrobacterium genome was published it was not possible to identify the gene encoding G3DH due to a lack of well-defined homologues in other species. Since then, the G3DH gene has been identified in a number of bacterial species including Halomonas and Gramella. We used these sequences to identify the homologous gene in Agrobacterium tumefaciens and Stenotrophomonas maltophilia. The A. tumefaciens and S. maltophilia G3DH genes were synthesised and subcloned into a protein expression vector and expressed in E.coli. A number of induction, expression and refolding strategies were tested. Soluble enzyme was recovered and assays with a model substrate suggested that the enzyme retained activity. However, after incubation of the enzyme with sucrose, we were not able to detect ketosucrose in the reaction products. The detailed experimental results are shown in Appendix 3. In Agrobacterium, this enzyme is secreted into the periplasmic space which may indicate that it interacts with other proteins to achieve the conversion of sucrose to ketosucrose. Future work may be able to resolve this process. 5.3 Physical and sensory properties of candidate sugars Assays were developed for sweetness, cariogenicity (tooth decay), digestibility and probiotic activity. These assays were used to test two of the candidate sugars as well as a panel of commercially available sugars. A summary of the results is presented here. The full methods and results will be described in two journal publications; the drafts of these papers are attached to this report as Appendix 4 and Appendix 5. 5.3.1 Sweetness Sweetness relative to sucrose and glucose was determined by a two-way preference ingestion assay with Drosophila melanogaster (fruit flies). Two of the candidate sugars were tested as well as a panel of commercially available sugars. Initially the accuracy of the assay was confirmed by comparing known sugars. Flies showed a strong and reliable preference for 5 mM or 2 mM sucrose solutions over water. The next series of assays confirmed that the flies could pick a sweeter solution reliably (10 mM sucrose compared to 2 mM sucrose). When the flies were presented with 10 mM fructose compared to 10 mM glucose, fructose was preferred. This was the expected result, as human taste tests rate fructose to be approximately two times as sweet as glucose. These results confirmed that the assay has good discriminating power and that differences were statistically significant. The sweetness of two candidate sugars, gentiobiitol and turanose has been tested against both glucose and sucrose. These results indicated that gentiobiitol is not as CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 12 sweet as sucrose but has the same sweetness as glucose, on a molar basis. Turanose was found to have a similar sweetness to sucrose. The relative sweetness of the two candidate sugars is shown diagrammatically in Figure 2. Analysis of the sweetness of a large panel of commercially available sugars highlighted some interesting relationships between structure and sweetness. The results showed that α-linked sugars were generally more palatable than β-linked sugars. Conversion of a β-linked sugar to a sugar alcohol appeared to improve its palatability. Amongst the isomers of sucrose, turanose had a similar sweetness preference, while leucrose and palatinose were judged less sweet. Gentiobiitol Turanose Aspartame Sucrose Acesulfame ‐K Maltitol 0,1 1 10 100 Glucose Saccharin Glucose Syrup Fructose Cyclamate Sucralose Xylitol Stevia Figure 2 Sweetness of gentiobiitol and turanose relative to a range of commercial sweeteners including sucrose. 5.3.2 Cariogenicity Dietary sugars are used by oral bacteria and produce acid that contributes to tooth decay. Production of acid by the oral bacterium Streptococcus mutans was used as an assay to detect potentially cariogenic sugars. The results (Figure 3) showed that S. mutans produced significant amounts of acid when incubated with sucrose, as expected. The starch-derived glucobioses and all of the β-linked glucobioses except for gentiobiose were also used by the bacteria. However isoprimeverose (a disaccharide derived from xyloglucans) and the isomers of sucrose were not used by the bacteria. The candidate sugars turanose and gentiobiitol were not used by S.mutans indicating that that are likely to be non-cariogenic. These two sugars behaved in a similar way to the commercial sweeteners palatinose (= isomaltulose) and maltitol. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 13 kojibiose nigerose maltose isomaltose sophorose laminaribiose cellobiose gentiobiose sucrose turanose leucrose palatinose maltitol gentiobiitol isoprimeverose -10 0 10 20 30 40 Figure 3 Change in pH following growth of Streptococcus mutans on various sugar substrates. Results are expressed as % change in pH compared to the pH at time zero. No change in pH indicates that the sugar would not contribute to tooth decay. Error bars represent standard error. 5.3.3 Digestibility and probiotic activity Assays were developed with yeast invertase and rat α-glucosidase/sucrase as models for human oral and intestinal digestion respectively. The assays were tested on the panel of reference sugars described above as well as on the two candidate sugars, turanose and gentiobiitol. Products of invertase digestions were monitored by HPLC and the results showed that only erlose, raffinose and 1-kestose were digested. These sugars showed complete breakdown within 30 min, similar to the breakdown of sucrose as the positive control (data not shown). The assay with rat intestinal enzymes determined the digestibility of the candidate and reference sugars (Figure 4). The enzyme showed different activities on the three sucrose isomers tested. Leucrose was digested at a similar rate to sucrose, while turanose was only partially digested and palatinose was digested minimally. Most of the α-glucobioses were digested. Gentiobiitol and the related disaccharide alcohol, maltitol were not digested. The results suggest that gentiobiitol would be non- calorigenic while turanose would release calories more slowly than sucrose. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 14 % change pH sucrose isomers α‐glucobioses 0.03 0.08 maltose 0.07 0.06 0.02 0.05 nigerose sucrose 0.04 kojibiose leucrose 0.030.01 turanose 0.02 isomaltose palatinose 0.01 0 0 0 10 20 30 40 0 10 20 30 40 Concentration (mM) Figure 4 Activity of rat glucosidase on various sugar substrates In addition to testing digestibility by glucosidase, we tested whether the sugars were able to inhibit the digestion of sucrose or isomaltose. Products such as acarbose which inhibit sucrose digestion have applications as pharmaceutical products to help control diabetes. Neither of the candidate sugars showed significant inhibition of sucrose digestion. However, several of the reference sugars showed an unexpected inhibitory activity. Isomaltose digestion was inhibited by xylitol, while sucrose digestion was inhibited by xylose and by isoprimeverose, a disaccharide (Xyl-Glc) derived from xyloglucan (Figure 5A). A dose-response curve for isoprimeverose was obtained (Figure 5B). Isoprimeverose was purchased for use as one of the reference sugars and it has not been well characterised. The activity of isoprimeverose in inhibiting sucrose digestion has not previously been described. The results suggest that isoprimeverose and other xyloglucan derivatives may be worth investigating as a dietary additive to reduce sugar uptake. A B 60 min incubation with 10 mM sugar (Acarbose control 0.6 mM) 40 100.00 35 30 75.00 25 50.00 20 15 25.00 10 5 0.00 0 0 5 10 15 20 -5 -25.00 xylose xylitol isoprimeverose acarbose concentration (mM)-10 isomaltose inhibition (%) sucrose inhibition (%) Figure 5 (A) Inhibition of digestion of sucrose and maltose by alternative sugars. (B) Dose-response curve for the activity of isoprimeverose on inhibition of sucrose digestion. Error bars represent standard deviation. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 15 Inhibition α-glucosidase (%) Glucose released (umol / min) inhibition sucrose digestion (%) In the human gut, sugars that are not digested may offer additional health benefits by supporting the growth of probiotic bacteria which produce short-chain fatty acids. As the two candidate sugars were shown to be poorly digested by mammalian enzymes, assays for growth of probiotic bacteria using these sugars as substrates are being carried out. The results of these assays should be known by the end of June 2010. 5.3.4 Summary A summary of the results from analysis of the candidate sugars is shown in Table 4. Gentiobiitol was sweet tasting. It did not support the growth of oral bacteria. It was not digested by invertase or by glucosidase. Turanose was also sweet tasting and was also not utilised by oral bacteria. Turanose was not digested by invertase and was only slowly digested by glucosidase. These properties match the activities of the commercial sweeteners maltitol and palatinose (isomaltulose). One of the reference sugars, isoprimeverose, also showed some interesting properties. In addition to being undigested, isoprimeverose was able to inhibit the digestion of sucrose to some extent. As the sweetness assay indicated that isoprimeverose is palatable, this sugar may have potential as a food supplement to improve glycaemic index. Table 4 Summary of the results from analysis of sugar properties. Sugar name Sweet taste Cariogenic Invertase Glucosidase Inhibition Sucrose Yes Yes Digestible Digestible ‐ Xylitol Yes No Not digestible Not digestible Yes Isomaltulose Yes No Not digestible Not digestible No + 15 other reference  sugars Turanose Yes No Not digestible Not digestible No Gentiobiitol Yes No Not digestible Not digestible No Isoprimeverose Yes No Not digestible Not digestible Yes 6.0 OUTPUTS 1. An assessment of opportunities for developing alternative sugars from sucrose based on technical feasibility, IP and market opportunities. 2. Expression and characterisation of two enzymes that were predicted to synthesise novel sugars from sucrose. Although neither of these enzymes carried out the predicted reactions, both enzymes were new; one is a dehydrogenase and the other is a glucosidase acting on gluco-oligosaccharides. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 16 3. A bioassay for estimating the relative sweetness of novel sugars based on a behavioural assay with fruit flies. 4. Methods for analysing the oral and intestinal digestibility of novel sugars. 5. Demonstration that two alternative sugars derived from sucrose (gentiobiitol and turanose) have the properties required for an alternative sweetener. 6. Identification of a disaccharide (isoprimeverose) which is sweet-tasting and able to inhibit the digestion of sucrose. 7.0 INTELLECTUAL PROPERTY: 7.1 Project IP (i) The information in the initial scoping study on candidate sugars and the processes underlying their production represents IP of potential value to the CRC and includes confidential information from the commercial partner. This information has been protected as a trade secret. (ii) Two potential sweeteners have been identified. This project IP was examined carefully against the criteria of novelty and potential market value. Although the candidate sugars had the properties of a sweetener, no economic production system could be identified, making it unlikely that these new sugars would be competitive in the marketplace. Therefore, patent protection was not sought and release of the information as journal papers has been approved. (iii) A potential inhibitor of glucose release has been identified. However, patent protection was not pursued because the level of inhibition was below that of the commercial inhibitor, acarbose. This would probably require further investigation before a competitive product could be developed. 7.2 Sub-contracts Four subcontracts were entered into during the course of the project. All contracts were discussed and agreed with the Commercialisation Manager before signing. (i) The chemical synthesis company, Epichem Ltd. was contracted to synthesise gentiobiitol (Appendix 6). On the advice of the CRC lawyers, the contract was modified to ensure that (i) the CRC retained all of the compound synthesised, and (ii) the CRC was granted first rights to any new synthetic technologies developed in the process. (ii) The UQ Protein Expression Facility was contracted to purify two enzymes (Appendix 7 and 8). (iii) Plant & Food Research (New Zealand) have been contracted to perform probiotic assays (Appendix 9). 8.0 ENVIRONMENTAL AND SOCIAL IMPACTS: There were no environmental or social impacts from conducting the project. Although production of alternative sugars in transgenic sugarcane has been suggested, future implementation of this technology is more likely to involve in vitro production CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 17 systems such as microbial bioreactors, as these would be cheaper and faster to implement than production in a transgenic plant. 9.0 EXPECTED OUTCOMES The outcome of this work is an improved ability to exploit new options for diversification in the sugar industry. Although the project has not produced a new commercial product, the information and tools developed by the project will assist future efforts to develop new sugar derivatives as alternative sweeteners. 10.0 FUTURE NEEDS AND RECOMMENDATIONS The research described here is at a very early stage with respect to producing a commercial sweetener. The scoping study showed that candidate sugars and enzymes of synthesis can be identified. However the cloning and expression of those enzymes showed that reproducing native enzyme activities in vitro can be very difficult. Techniques for regulating enzyme action and modifying enzyme specificity and kinetics have been described and could be applied to these enzymes if the product was valuable enough. Turanose was identified as a potential alternative sweetener. In further work, the enzymes that are predicted to synthesise turanose from sucrose in honey bees should be cloned and tested. During the scoping study, several other sugars with potential as alternative sugars were identified, but the sequences of enzymes that may synthesise these sugars were not available at that time. As more complete genome sequences and comparative analyses become available, these sugars may become more attractive prospects for in vitro production. The tests developed in this project identified some interesting relationships between sugar structure and sensory or nutritive properties. Further analysis of these relationships may allow design of new sweeteners with optimal properties. 11.0 PUBLICATIONS ARISING FROM THE PROJECT 1. Glassop D., Ryan L.P., Bonnett G.D. and Rae A.L. (2010) The complement of soluble sugars in the Saccharum complex. Tropical Plant Biol. 3:110–122. DOI 10.1007/s12042-010-9049-y (Appendix 1) 2. Hodoniczky J., Robinson G.J., McGraw E.A. and Rae A.L. Fruit fly bioassay to distinguish ‘sweet’ sugar structures. Submitted to J. Agric. Food Chem. (Appendix 4) 3. Hodoniczky J., Clayton, D., Morris, C. and Rae A.L. Oral and intestinal digestion of carbohydrates in structurally relevant terms. Manuscript in preparation. (Abstract included as Appendix 5) 12.0 ACKNOWLEDGMENTS We wish to thank Dr Donna Glassop (CSIRO Plant Industry) and Ms Louise Ryan (CRC Vacation Scholar) for a survey of the range of sugars present in sugarcane and related species. We thank Dr Elizabeth McGraw (University of Queensland) for advice on the experiments with fruit flies and for access to specialist equipment for growing and handling the flies. We also thank Professor Carol Morris and members of her research group (Southern Cross University) for advice and assistance in the assays for digestibility. CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 18 13.0 REFERENCES Fong Chong B, Bonnett GD, Glassop D, O’Shea MG and Brumbley SM. (2007) Growth and metabolism in sugarcane are altered by the creation of a new hexose-phosphate sink. Plant Biotechnology Journal 5, 240–253 doi: 10.1111/j.1467- 7652.2006.00235.x Fong Chong B, Abeydeera WP, Glassop D, Bonnett GD, O'Shea MG and Brumbley SM. (2010) Co-ordinated synthesis of gentiobiitol and sorbitol, evidence of sorbitol glycosylation in transgenic sugarcane. Phytochemistry. 71(7):736-41. Glassop D., Roessner U., Bacic A., and Bonnett G.D. (2007) Changes in the sugarcane metabolome with stem development. Are they related to sucrose accumulation? Plant Cell Physiol. 48(4): 573–584. doi:10.1093/pcp/pcm027 Gordesky-Gold, B., Rivers, N., Ahmed, O.M. and Breslin, P.A.S. (2008) Drosophila melanogaster prefers compounds perceived sweet by humans. Chemical Senses 33, 301-309. 14.0 APPENDIX 1 Glassop D., Ryan L.P., Bonnett G.D. and Rae A.L. (2010) The complement of soluble sugars in the Saccharum complex. Tropical Plant Biol. 3:110–122. DOI 10.1007/s12042-010-9049-y 15.0 APPENDIX 2 Enzymatic synthesis of glucosyl sucrose 16.0 APPENDIX 3 Enzymatic synthesis of ketosucrose 17.0 APPENDIX 4 Hodoniczky J., Robinson G.J., McGraw E.A. and Rae A.L. Fruit fly bioassay to distinguish ‘sweet’ sugar structures. Manuscript submitted to Journal of Agricultural and Food Chemistry. 18.0 APPENDIX 5 Hodoniczky J., Clayton, D., Morris, C. and Rae A.L. Oral and intestinal digestion of carbohydrates in structurally relevant terms. Manuscript in preparation. 19.0 APPENDIX 6 Contract with Epichem Ltd. 20.0 APPENDIX 7 Contract #1 with UQ Protein Expression Facility 21.0 APPENDIX 8 Contract #2 with UQ Protein Expression Facility CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 19 22.0 APPENDIX 9 Contract with Plant and Food Research (New Zealand) CRC Sugar Industry Innovation through Biotechnology Date (May/2010) Page 20 Tropical Plant Biol. (2010) 3:110–122 DOI 10.1007/s12042-010-9049-y The Complement of Soluble Sugars in the Saccharum Complex Donna Glassop & Louise P. Ryan & Graham D. Bonnett & Anne L. Rae Received: 1 December 2009 /Accepted: 25 February 2010 /Published online: 30 March 2010 # Springer Science+Business Media, LLC 2010 Abstract The use of sugarcane as a biofactory and source manipulated. Since species from the Saccharum complex of renewable biomass is being investigated increasingly due can be interbred, any genes leading to the production of sugars to its vigorous growth and ability to fix a large amount of of interest could be introgressed into commercial Saccharum carbon dioxide compared to other crops. The high biomass species or manipulated through genetic engineering. resulting from sugarcane production (up to 80 t/ha) makes it a candidate for genetic manipulation to increase the Keywords GC-MS .Metabolite analysis . Soluble sugar . production of other sugars found in this research that are Sugarcane of commercial interest. Sucrose is the major sugar mea- sured in sugarcane with hexoses glucose and fructose Abbreviations present in lower concentrations; sucrose can make up to DM dry mass 60% of the total dry weight of the culm. Species related to FM fresh mass modern sugarcane cultivars were examined for the presence GC gas chromatography of sugars other than glucose, fructose and sucrose with the MS mass spectrometry potential of this crop as a biofactory in mind. The species examined form part of the Saccharum complex, a closely- related interbreeding group. Extracts of the immature and mature internodes of six different species and a hybrid were Introduction analysed with gas chromatography mass spectrometry to identify mono-, di- and tri-saccharides, as well as sugar Sugarcane (Saccharum hybrid) has a specialised metabo- acids and sugar alcohols. Thirty two sugars were detected, lism that efficiently synthesises and stores sucrose at higher 16 of which have previously not been identified in concentrations than most plants. Carbon is partitioned sugarcane. Apart from glucose, fructose and sucrose the between sinks at the meristematic regions and the storage abundance of sugars in all plants was low but the research tissue in the stalk depending on developmental age and demonstrated the presence of sugar pathways that could be environmental influences. Whilst the major storage com- pound is the disaccharide sucrose, in immature tissues, the constituent monosaccharides, glucose and fructose are Communicated by: Ray Ming present at higher levels than sucrose (Hoepfner and Botha D. Glassop (*) : L. P. Ryan :G. D. Bonnett :A. L. Rae 2003). A variety of other soluble sugars has also been CSIRO Plant Industry, Queensland Bioscience Precinct, detected in analyses of stem tissue from sugarcane (Glassop 306 Carmody Road, St. Lucia Qld. 4067, Australia et al. 2007). As the price of sucrose on world markets is e-mail: donna.glassop@csiro.au volatile, there is increasing interest in the extraction of : : higher value products from sugarcane (Edye et al. 2006).D. Glassop L. P. Ryan G. D. Bonnett :A. L. Rae The presence and therefore potential exploitation of soluble Co-operative Research Centre for Sugar Industry Innovation through Biotechnology, University of Queensland, sugars other than sucrose in sugarcane and related species St. Lucia Qld. 4067, Australia has not been thoroughly explored. Tropical Plant Biol. (2010) 3:110–122 111 Saccharum is a genus in the grass family Poaceae, tribe Table 1 Percent moisture content of sugarcane internodes (± standard Andropogoneae (Daniels and Roach 1987). Modern sugar- deviation). Letters represent least significant difference (P≤0.05) between all samples (species and internode position), samples with cane cultivars are derived predominantly from interspecific the same letter are not significantly different crosses between S. officinarum L. and S. spontaneum L.. Together with Saccharum, the genera Erianthus, Miscanthus, Species Internode position Narenga and Sclerostachya form the “Saccharum complex”, a closely-related interbreeding group (Mukherjee 1957). Immature Mature Genera within the Saccharum complex can be forced to interbreed with Saccharum and thus if high value sugars are S. spontaneum 71.74±7.05bc 58.62±6.06a present within any of these species the genes responsible for E. arundinaceous 75.31±1.13cd 68.43±2.79b their synthesis could be introgressed into agronomically S. robustum 79.17±2.94de 61.45±2.25a superior Saccharum hybrids. M. sinesis 81.56±2.22e 70.80±4.73bc An increased role has been ascribed to sugars in the S. edule 88.93±0.69f 73.07±0.03bc regulation of metabolites. Trehalose is postulated to modu- S. officinarum 89.02±1.21f 73.03±0.69bc late hexokinase activity which is implicated in sugar sensing comm.hybrid Q117 89.57±0.19f 70.34±4.06bc and plant development (Bosch 2005; Rolland et al. 2006; Zhang et al. 2006). There are also reported links between sugar levels and gene expression through complex signal transduction networks (Smeekens 2000; Koch 2004; and S. edule had the highest and S. spontaneum the lowest Rolland and Sheen 2005; Felix et al. 2009). Other sugars water content in mature internodes (Table 1). (palatinose, turanose, cellobiose, gentiobiose, lactulose and leucrose) have been implicated in repressing gibberellin Sucrose, Glucose and Fructose Content signalling in barley embryos (Loreti et al. 2000). Conse- quently identifying the range of sugars present in sugarcane Significant differences were observed for sucrose, glucose may give an additional benefit through the study of gene- and fructose content between species and between inter- expression and metabolic regulation in sugarcane. nodes within a species (Fig. 1). Glucose and fructose In this survey, we analysed the soluble sugars from a concentrations were similar to each other within each variety of species belonging to the Saccharum complex sample, irrespective of species or internode maturity because identification of these less abundant sugars may (Fig. 1a). The concentrations of these hexoses were higher indicate the existence of pathways of sugar biosynthesis in immature internodes than in mature internodes. The other than those involving sucrose, fructose and glucose. sucrose content in the mature internode was generally The other motive of this work was to search for sugars that higher than in the immature internode for all species may be of higher economic value than sucrose. Together except E. arundinaceous and M. sinesis, where the sucrose with our understanding of sugar metabolism, knowledge content did not increase with maturity (Fig. 1b). S. of alternative sugars may be useful in developing sugar- robustum and S. spontaneum displayed a smaller increase cane as a biofactory and may illustrate areas for future in sucrose content between immature and mature internodes manipulation. compared to the other species. The lower levels of sucrose in S. robustum and higher levels in S. officinarum are used as a taxonomic indicator of these two species (Whalen Results and Discussion 1991; Irvine 1999). Amongst the three main constituents of the sugarcane Water Content stem (fibre, water and sucrose) an association has been demonstrated between the amount of sucrose accumulated The three main components of sugarcane are soluble and the water content. Bonnett et al. (2006) and Keating et sugars, fibre and water. Significant differences (P≤0.05) al. (1999) observed in commercial cultivars, that as sucrose in water content were observed between species and stages concentration increases above 100 mg g−1 fresh mass (FM), of development. Water content in the immature internodes the water content decreases at a proportional rate of 1:1. In ranged between 72 and 90% with the highest water contents the current experiment, this observation was upheld for the detected in commercial hybrid Q117 and S. officinarum commercial hybrid Q117, and was also seen in S. officinarum (Table 1). The lowest water content in immature internodes and S. edule (Fig. 2). S. spontaneum, S. robustum, M. sinesis was found in S. spontaneum. In the mature internodes, and E. arundinaceous do not fit the above mentioned model there was a rearrangement of the order of species with because they have sucrose concentrations lower than moisture contents ranging from 58 to 73%; S. officinarum 100 mg g−1 FM (Fig. 2). 112 Tropical Plant Biol. (2010) 3:110–122 Fig. 1 Mean sugar concentra- 180 tions per g dry weight of A. immature and mature internodes 160 of genotypes from the Saccharum complex. a. 140 Fructose—dark shading and Glucose—light shading; 120 b. Sucrose. Error bars represent standard errors (n=3). Letters 100 represent least significant difference (P≤0.05) between all 80 samples (species and internode position), samples with the 60 same letter are not significantly different 40 20 0 immature internodes mature internodes 600 B. 500 400 300 200 100 0 immature internodes mature internodes Other Sugars present but could not be identified. A total of 32 sugars were identified and their relative abundance measured. In addition to the more abundant sugars associated with Only glucose, fructose and sucrose were quantified as they sugarcane, a number of sugars present in lower concen- were present at a sufficient concentration to be detected by trations were detected (Table 2, relative abundance values HPLC. The other sugars were present in such low con- are presented) and identified by combined gas chromatog- centrations that they could not be quantified or detected by raphy mass spectrometry. Identification of sugars was HPLC. Bosch et al. (2003) measured five sugars in sugar- limited by the availability of sugar mass spectras in the cane (inositol, raffinose, maltose, xylose and trehalose) at libraries, and it is possible that a wider range of sugars were concentrations ranging from 0.0005 to 0.5 µmole g−1 mg Sucrose / g dry weight mg sugar / g dw S. officinarum e g S. officinarum g E. arundinaceous abc de E. arundinaceous e S. spontaneum abc cd S. spontaneum c M. sinesis f abc M. sinesis abc S. robustum a e S. robustum e S. edule bcd f S. edule f Commercial hybrid ab h Commercial hybrid i S. officinarum g abc S. officinarum bc E. arundinaceous a a E. arundinaceous a S. spontaneum d a S. spontaneum a M. sinesis d ab M. sinesis abc S. robustum cd a S. robustum a S. edule g abc S. edule abc Commercial hybrid g abc Commercial hybrid abc Tropical Plant Biol. (2010) 3:110–122 113 180 spread in plants and is a component of β-D-glucans from E. arundinaceous 160 M. sinesis related grasses including sorghum, maize, barley, rye and S. edule 140 S. officinarum wheat (Woolard et al. 1976). Mixed-linked glucans were S. robustum visualised in the cell walls of rind, parenchyma, phloem and 120 S. spontaneum comm.hybrid Q117 vascular parenchyma cells of immature Q117 internodes, 100 though this particular glucan was not observed in xylem 80 or bundle sheath cells (Fig. 3a,b). Callose was present in the sieve plates and wall deposits of phloem in Q117 60 mature internodes (Fig. 3c). Relative abundance values of 40 laminaribiose were significantly different between species 20 (P≤0.01) and internodes (P≤0.01); with higher values observed in S. robustum and M. sinesis for both internodes 0 500 600 700 800 900 and mature S. spontaneum internodes than in the other -1 species examined in this study.water content mg g fm Fig. 2 Water content of individual replicate internode samples plotted Cell Wall Hemicellulosic Polysaccharides against sucrose concentration on a fresh mass basis. The line represents the 1:1 trend of some sugarcane species to replace water at a rate of 1 g sucrose g−1 water when sucrose concentrations above In addition to β-glucans, cell wall polysaccharides from 100 mg g−1 FM are achieved grasses typically include hemicelluloses and small quanti- ties of pectins (Carpita 1996). It is possible that the xylose and arabinose detected are intermediaries in the metabolism FM. Xylitol, trehalose, arabinose, galactose and maltose of the hemicellulose polysaccharide, arabinoxylan, which is were also detected via high performance anion exchange an important component of grass cell walls (Cobbett et al. chromatography in progeny obtained from a cross between 1992). Arabinose levels were significantly different between Brazilian cultivars SP80-180 and SP80-4966 (Felix et al. species (P≤0.05) and internodes (P≤0.01), while xylose 2009). levels were only significantly different between internodes Sixteen new sugars were identified in addition to (P≤0.01). The relative abundance values for both xylose those already reported in sugarcane and its relatives by and arabinose were higher in immature internodes than Mutalshaikhov and Ismailov (1976), Bosch et al. (2003), mature internodes for all species examined. This may reflect Felix et al. (2009) and Glassop et al. (2007) (Table 2). In higher demand for precursors of cell wall polysaccharides in the samples of the species examined in this research young tissues. neither theanderose nor 1-kestose were found, although they have previously been observed in fresh cane juice by Cell Wall Pectic Polysaccharides Eggleston et al. (2004). Arabinose is an important constituent of pectic cell wall Components of Cell Wall Polysaccharides polysaccharides such as arabinogalactans and rhamnogalac- turonans, and is found in hydroxyproline-rich cell wall A number of the sugars found can be attributed to glycoproteins (Cobbett et al. 1992; Ralet et al. 1994; Burget intermediates in core metabolism, particularly cell wall et al. 2003). Together with the sugar acids, galactonic (C1 synthesis and breakdown. For example, cellobiose, the carboxylic acid), galacturonic (C6 carboxylic acid) and disaccharide consisting of two glucose molecules joined by glucuronic acids, the galactose-arabinose disaccharide, a β(1 → 4) linkage, is likely to be a component of the cell which was identified by comparison to a private library of wall polysaccharide, cellulose. Cellulose is ubiquitous in mass spectra (pers. comm., U. Roessner, University of higher plant cell walls and its presence has been confirmed Melbourne) is also linked to pectic polysaccharides (Nichols amongst β-D-glucans of sorghum, maize, barley, rye and 1974). Within the species of the Saccharum complex wheat (Woolard et al. 1976). In the present study, cellobiose examined here, the relative abundance values of the sugar levels were not significantly different (P>0.05) between the acids were low, with significant differences between species species and internodes examined. (galacturonic acid, P≤0.01) and internodes (galacturonic Similarly, laminaribiose, the β(1→ 3) linked disaccharide and glucuronic acids, P≤0.01 for both), again suggesting of glucose, is a component of the structural and wound- that turnover of pectic cell wall precursors is higher in induced glucan, callose, or the mixed-link glucans which younger tissues. Galactonic acid had significant differences have been described in grasses (Carpita 1996). Although not between species (P≤0.01) and internodes (P≤0.05) with previously identified in sugarcane, laminaribiose is wide- S. officinarum, E. arundinaceous and M. sinesis containing Sucrose mg g- 1 fm 114 Tropical Plant Biol. (2010) 3:110–122 Table 2 Relative abundance of sugars in members of the Saccharum complex from GC-MS analysis. Values are relative responses of the individual sugars compared to the internal standard, ribitol, and adjusted to be expressed on an equivalent fresh mass basis. Statistical significance between species (S) and internodes (I), species × internodes (SI). Mean values are shown with standard errors in brackets (n=3) Statistical Stalk maturity significance Immature internode Saccharides S I SI Saccharum Erianthus Saccharum Miscanthus sinesis Saccharum Saccharum edule Saccharum officinarum aruundinaceous spontaneum robustum hybridQ117 Cell wall cellobiose 0.73 (0.18) 0.20 (0.12) 0.06 (0.04) 0.36 (0.10) 1.71 (0.58) 1.16 (0.30) 0.15 (0.03) laminarabiose a a 1.12 (0.37) 0.57 (0.08) 2.61 (1.36) 10.75 (2.84) 12.6 (7.44) 1.93 (0.54) 3.45 (0.57) Cell wall hemicellulosic arabinose b a 10.46 (0.64) 18.60 (2.11) 26.49 (3.48) 20.49 (2.28) 24.9 (1.72) 17.3 (0.81) 14.4 (0.27) xylose a b 8.76 (0.21) 17.92 (1.64) 27.87 (4.27) 9.17 (1.60) 15.6 (2.26) 8.58 (0.38) 7.49 (0.26) Cell wall pectic galactonic acid a b 16.72 (0.72) 8.90 (1.35) 38.71 (9.38) 40.76 (9.38) 53.2 (8.40) 28.2 (2.30) 32.9 (4.50) galacturonic acid a a 6.20 (0.28) 4.64 (0.26) 4.16 (0.39) 6.08 (1.26) 8.87 (1.11) 7.71 (0.24) 26.5 (8.02) gal-ara 2.85 (0.62) 2.04 (0.31) 2.90 (0.80) 4.82 (0.76) 3.60 (0.92) 1.81 (0.27) 1.82 (0.20) galactose b 59.15 (1.89) 293 (147) 437 (106) 425 (62) 80 (26) 51 (1.44) 489 (125) glucuronic acid a 18.78 (3.72) 67.61 (1.39) 12.35 (0.99) 39.07 (5.90) 22.3 (4.02) 12.8 (3.76) 27.3 (1.72) Metabolism gentiobiose a 10.71 (0.63) 22.94 (8.93) 78.31 (15.9) 60.26 (11.4) 12.8 (3.15) 13.0 (2.05) 18.8 (1.44) ribose a b 6.13 (1.28) 5.03 (0.97) 8.92 (2.57) 7.87 (1.04) 6.05 (0.69) 7.79 (0.97) 10.4 (0.16) fructose 6 P a a 14.72 (1.53) 9.21 (2.87) 85.36 (11.6) 62.74 (12.9) 22.5 (6.49) 15.6 (1.01) 29.2 (1.14) glucose 6 P a a 17.12 (1.81) 13.55 (3.32) 74.63 (4.26) 70.58 (19.0) 36.2 (11.8) 21.1 (0.72) 35.4 (3.94) Physiological stress maltose a b 0.26 (0.02) 0.20 (0.09) 0.70 (0.54) 0.07 (0.02) 1.16 (0.11) 0.33 (0.09) 0.17 (0.09) maltotriose a a 1.85 (0.69) 1.70 (0.23) 4.47 (1.84) 7.85 (1.68) 0.77 (0.16) 2.14 (0.57) 4.25 (0.49) trehalose a 11.42 (0.78) 67.95 (41.2) 69.60 (17.9) 233 (83) 20.6 (1.08) 10.6 (2.41) 173 (5.16) raffinose a a 3.92 (1.19) 7.40 (1.38) 5.23 (1.69) 14.46 (4.18) 3.04 (0.53) 3.87 (0.88) 11.4 (1.21) Alcohols/Polyols erythritol a b 0.06c 0.05c 0.61 (0.40) 0.08 (0.05) 1.01 (0.42) 0.07 (0.02) galactinol a a a 11.78 (1.21) 27.05 (4.86) 8.98 (4.24) 14.81 (1.27) 20.3 (3.40) 12.0 (1.74) 2.53 (0.85) inositol 1P b b 2.69 (0.16) 1.68 (0.72) 5.00 (0.91) 7.70 (1.19) 3.24 (0.54) 2.40 (0.11) 4.42 (0.83) mannitol a a b 49.02 (8.01) 111.5 (43.3) 72.43 (20.3) 26.54 (10.3) 40.4 (6.09) 50.5 (11.2) 400 (49) myo-inositol a a a 81.09 (8.27) 15.24 (1.57) 184 (22) 171 (44) 144 (13) 135 (10) 90 (2) pinitol a a 17.46 (1.51) 49.93 (22.9) 22.70 (3.55) 133 (14) 14.7 (4.02) 13.5 (1.53) 24.9 (1.92) sorbitol a a a 2.98 (2.06) 6.81 (2.53) 6.46 (1.02) 14.99 (2.34) 5.52 (0.91) 5.11 (1.21) xylitol a a 0.66 (0.10) 0.63 (0.09) 0.27 (0.09) 0.56 (0.11) 1.18 (0.17) 0.64 (0.04) 0.24 (0.06) Rare plant saccharides melezitose a a 1.14 (0.26) 2.21 (0.31) 5.90 (0.85) 8.54 (1.39) 2.20 (0.68) 1.44 (0.10) 0.85 (0.08) turanose 0.49 (0.07) 3.28 (0.65) 16.81 (0.33) 7.01 (0.38) 1.82 (0.31) 0.41 (0.20) 5.38 (1.29) asignificantly different at P0.01; bsignificantly different at P0.05 conly one rep detected Sthtaislkmmeattaubriotylite; therefore no standard error can be provided Mature internode Saccharides Saccharum Erianthus Saccharum Miscanthus sinesis Saccharum Saccharum edule Saccharum officinarum aruundinaceous spontaneum robustum hybridQ117 Tropical Plant Biol. (2010) 3:110–122 115 Table 2 (continued) Stalk maturity Mature internode Saccharides Saccharum Erianthus Saccharum Miscanthus sinesis Saccharum Saccharum edule Saccharum officinarum aruundinaceous spontaneum robustum hybridQ117 Cell wall cellobiose 0.04 (0.01) 0.05 (0.02) 0.08 (0.05) 0.04 (0.001) 0.23 (0.09) 0.02 (0.001) 0.10 (0.03) laminarabiose 4.76 (1.09) 2.42 (1.29) 20.30 (12.4) 18.38 (4.52) 35.9 (2.70) 3.76 (0.78) 3.82 (1.50) Cell wall hemicellulosic arabinose 6.86 (0.40) 6.05 (0.58) 10.75 (3.55) 14.23 (2.95) 8.18 (1.35) 9.83 (0.53) 6.15 (4.23) xylose 5.18 (1.39) 1.37 (0.11) 3.88 (2.14) 3.69 (1.18) 2.13 (0.99) 3.72 (0.44) 2.75 (2.29) Cell wall pectic galactonic acid 18.90 (4.43) 7.82 (0.22) 29.22 (2.09) 29.61 (1.06) 28.4 (11.6) 27.0 (0.69) 20.0 (10.4) galacturonic acid 4.51 (0.51) 1.06 (0.18) 1.98 (0.61) 3.14 (1.81) 5.58 (2.37) 6.03 (2.71) 6.38 (1.82) gal-ara 2.67 (0.37) 0.76 (0.22) 4.37 (0.94) 1.95 (0.24) 4.95 (0.32) 2.79 (0.53) 3.94 (0.91) galactose 71.97 (4.50) 263 (62.4) 77.7 (30.9) 156 (13) 51 (26) 61.7 (1.19) 157c glucuronic acid 4.52 (2.96) 2.62 (0.41) 2.69 (1.26) 4.68 (2.21) 0.90 (0.16) 2.18 (1.28) 2.35 (1.99) Metabolism gentiobiose 9.58 (0.46) 26.14 (4.57) 84.13 (42.3) 42.55 (18.9) 14.9 (4.39) 8.54 (1.76) 14.9 (1.23) ribose 3.51 (0.27) 1.88 (0.25) 2.44 (0.77) 6.15 (0.20) 1.96 (0.48) 2.72 (0.16) 2.16 (1.01) fructose 6 P 5.29 (0.70) 14.03 (0.38) 83.35 (67.9) 22.46 (6.40) 12.0 (2.65) 11.4 (1.99) 7.55 (5.38) glucose 6 P 10.79 (1.78) 15.25 (0.67) 25.94 (7.78) 32.68 (11.0) 20.5 (5.00) 20.5 (4.14) 11.0 (7.90) Physiological stress maltose 0.13 (0.04) 0.17 (0.03) 0.07 (0.02) 0.07 (0.03) 0.87 (0.49) 0.29 (0.07) 0.13 (0.03) maltotriose 19.15 (3.47) 4.54 (0.56) 90.70 (44.4) 61.31 (50.1) 18.6 (7.42) 34.1 (5.72) 72.2 (25.0) trehalose 9.48 (1.14) 39.67 (19.3) 76.99 (22.3) 276.7 (73.5) 11.3 (1.16) 6.56 (1.30) 110 (55) raffinose 7.33 (0.15) 10.20 (1.58) 3.37 (0.60) 6.59 (1.91) 1.40 (0.25) 21.3 (2.89) 13.4 (3.28) Alcohols/Polyols erythritol 0.1c 0.06 (0.03) 1.05 (0.43) 1.46 (0.64) galactinol 0.82 (0.28) 17.27 (3.65) 0.24 (0.17) 10.82 (7.74) 0.38 (0.04) 0.11 (0.03) 0.37 (0.23) inositol 1P 2.25 (0.10) 3.45 (0.31) 2.98 (0.60) 4.97 (2.20) 4.85 (1.52) 2.82 (0.24) 1.69 (0.46) mannitol 24.5 (10.0) 8.02 (1.57) 3.18 (1.77) 2.81 (1.03) 1.90 (0.76) 20.6 (7.83) 58.1 (55.4) myo-inositol 37 (4) 18.89 (1.60) 41.73 (11.9) 58.62 (17.1) 23.0 (3.80) 84.5 (4.46) 24.0 (15.9) pinitol 7.69 (0.26) 13.23 (1.87) 15.21 (1.16) 31.19 (9.26) 6.71 (1.17) 6.06 (0.59) 12.9 (3.79) sorbitol 3.75 (1.18) 6.34 (0.88) 6.05 (0.79) 12.03 (4.29) 6.32 (1.17) 3.90 (0.05) 6.74 (3.52) xylitol 0.19 (0.05) 0.16 (0.01) 0.06 (0.03) 0.33 (0.01) 0.25 (0.07) 0.20 (0.05) 0.10 (0.05) Rare plant saccharides melezitose 1.71 (0.05) 3.46 (0.50) 11.95 (2.56) 10.80 (0.42) 7.39 (2.10) 1.57 (0.45) 2.10 (0.36) turanose 1.81 (0.45) 5.82 (2.14) 15.02 (5.12) 4.63 (0.68) 5.70 (0.94) 1.27 (0.17) 12.6 (2.46) 116 Tropical Plant Biol. (2010) 3:110–122 Fig. 3 Detection of polysaccharides in stem tissue of commercial in sieve plates (solid arrows) and wall deposits (open arrows) in the hybrid Q117 by histochemical staining and immunolabelling. a, b show phloem in a longitudinal section of internode 10 stained with the sections of fixed tissue embedded in paraffin; c–f show free hand aniline blue fluorochrome. Blue autofluorescence of unlabelled cell sections. a. In a section of internode 1 labelled with antibody to the walls is also visible. d. In an unstained section, only blue (1–3),(1–4)-β-glucan and an Alexa-Fluor 488 conjugated secondary autofluorescence is visible. e. In a transverse section of the tenth antibody, green fluorescence was observed in cell walls of rind, internode stained with I2-KI, starch is visualised as black staining in parenchyma, phloem and vascular parenchyma cells, but not in the the chlorenchyma cells near the rind (arrows). f. In an unstained xylem or bundle sheath cells. b. In control experiments where the section, there is no dark staining in the chlorenchyma. Tissues present primary antibody was omitted, only weak background fluorescence include epidermis (e), metaxylem vessels (x), phloem (p), and was observed. c. Yellow fluorescence indicates the presence of callose parenchyma (par). Bars: a, b 100 μm; c–f 200 μm higher levels in mature than immature tissues. The relative difference between species (P>0.05) was observed there abundance of galactose-arabinose disaccharide was not was a species by internode interaction (P≤0.05). A small significantly different between species and internodes (P> amount of free ribose is not surprising since it is a main 0.05 for both), though S. officinarum had a higher relative component of RNA, ATP and other metabolites, and the abundance level in mature than immature tissues which was presence of ribose has been detected in other plant tissues the opposite of the remaining species examined. (Seymour et al. 1989). Gentiobiose (β-1,6-glucosyl-glucose) has been identified Sugars With Other Roles in Metabolism in microorganisms, but it occurs rarely as a free sugar in plants, although there is evidence of β-glucosidases with The relative abundance values of ribose in the samples were transferase activity which could synthesise this sugar (Zhifang low, but higher in immature internodes than mature inter- and Loescher 2003). Gentiobiose has been detected in plant nodes (significantly different, P≤0.01). While no significant tissue by GC separation methods (Seymour et al. 1989) and Tropical Plant Biol. (2010) 3:110–122 117 it is found as an esterified side-chain in secondary 0.01). In E. arundinaceous and commercial hybrid Q117 metabolites such as crocin. Free gentiobiose was recently the differences in relative abundance values between found in tomato fruit, where it was implicated in signalling immature and mature internodes were small but there was during fruit ripening (Dumville and Fry 2003). Gentiobiose more trehalose in the immature internodes, while there was relative abundance values were significantly different no large difference between relative abundance values of between species (P≤0.01) with higher levels observed in immature and mature internodes for S. officinarum, S. immature than mature tissue of S. officinarum, M. sinesis, robustum, S. edule and S. spontaneum. M. sinesis had a S. edule and commercial hybrid Q117, similar to the higher relative abundance of trehalose in the mature difference seen between immature and mature tomato fruit. internode than immature internode. In our study, the highest levels of trehalose were found in M. sinensis which had Sugars Related to Adaptation to Stress relative abundance values in the mature tissues that were double the levels in other samples. Maltose and maltotriose are likely to be derived from the Raffinose (α -1,6-galactosyl-sucrose) and its homologous α(1 → 4)-glucan, starch (Weise et al. 2004; Smith et al. series of galactosyl sucroses are synthesised from sucrose by 2005) which was shown to be present in small amounts in the transferase activity of various α-galactosidases, using stems of the commercial variety Q117 (Fig. 3e). Maltose galactinol as the galactosyl donor (Kandler and Hopf 1980). was detected at both stages of development in all species Relative abundance levels for galactinol in immature inter- examined, with significant differences between species nodes were higher than mature internodes for all species (P≤0.01) and internodes (P≤0.05). The presence of examined; with significant differences between species, maltose was associated with tissue maturation in tomatoes internodes and species × internodes interactions (P≤0.01 (Roesnner-Tunali et al. 2003). The trisaccharide maltotriose for all parameters). It has been suggested that raffinose was present in both stages of internode development, with and the raffinose-family of oligosaccharides are involved all genotypes except E. arundinaceous exhibiting a higher in stress tolerance and they are commonly found in seeds relative abundance level in the mature internodes than the as a dessication tolerance agent (Claeyssen and Rivoal immature internodes; significant difference is noted between 2007). In the present study M. sinesis, S. robustum and S. species and internodes (P≤0.01 for both) with no significant spontaneum had higher relative abundance values of interaction (P>0.05). raffinose in the immature than mature internodes, the In some plant species including several pteridophytes, opposite of what was observed in the remaining species; members of the Apiaceae and the xerophyte Myrothamnus with significant differences between species and a signifi- flabellifolius, trehalose is the primary carbohydrate for cant species × internode interaction (P≤0.01 for both). translocation and storage, and it has also been implicated in Sugar alcohols or polyols are thought to play an important sugar sensing and responses to abiotic stress (Goddijn et al. role as osmoregulators, associated with osmotic stress 1997; Goddijn and van Dun 1999; Avonce et al. 2005). In caused by temperature, drought, salinity or high sugar con- previous studies, trehalose was positively correlated with centrations (Bieleski 1982; Pommerrenig et al. 2007), sucrose in the sugarcane hybrid Q117 (Glassop et al. 2007), possibly by scavenging hydroxyl radicals and preventing but negatively correlated in the hybrid cultivars N19 and oxidative damage (Smirnoff and Cumbes 1989; Loescher US6656-15 (Bosch et al. 2003). Trehalose and trehalose-6- and Everard 1996; Nishizawa et al. 2008). In some species, phosphate are thought to mediate carbon metabolism and polyols, like mannitol (Apiaceae, Oleaceae, Rubiaceae partitioning through modulation of hexokinase activity in a and Scrophulariaceae), and sorbitol (woody Rosaceae, similar manner to that observed in yeast (Goddijn and Spiridaeoideae, Pyroideae and Prunoideae), can be syn- Smeekens 1998; Müller et al. 1999; Eastmond et al. 2003; thesised and translocated from leaves in addition to Schluepmann et al. 2004; Avonce et al. 2005). McCormick sucrose (Bieleski 1982; Nadwodnik and Lohaus 2008). et al. (2008) observed a decreased ratio of trehalose-6- These sugar alcohols may represent as much as 30% of the phosphate synthase : trehalose-6-phosphate phosphatase carbon fixed by polyol translocating plants (Bieleski 1982; (TPS:TPP) in sugarcane leaves with reduced photosynthesis Loescher and Everard 1996; Nadwodnik and Lohaus due to cold-girdling. Transgenic sugarcane overexpressing 2008). TPS from the fungus Grifola fondosa accumulated trehalose In our study, the relative abundance of mannitol in all which conferred increased drought tolerance, though no genotypes, except the commercial cultivar Q117, were effects on carbon partitioning/accumulation were reported slightly higher in immature than mature internodes; in Q117 (Zhang et al. 2006). relative abundance levels in the immature internodes were Amongst the samples examined in this study, there was more than 3 times higher than in mature internodes; with no clear association between sucrose content and trehalose. significant differences between species (P≤0.01), inter- There were significant differences between species (P≤ nodes (P≤0.01) and a significant species × internode 118 Tropical Plant Biol. (2010) 3:110–122 interaction (P≤0.05). Mannitol has been detected in more internodes. The remaining species had only a slightly than 100 vascular plants (Nadwodnik and Lohaus 2008) higher abundance of xylitol in immature than mature and it is a major translocatable sugar in some species internodes. Xylitol detection in plants as a product of plant (Claeyssen and Rivoal 2007; Trip et al. 1965) where it may metabolism has not been convincing due to the method of play a similar role to sucrose in transferring photosynthate analysis and the presence of xylitol has often been to various sinks (Bieleski 1982). Studies in transgenic attributed to fungal pathogens degrading xylans (Bieleski plants have also suggested that mannitol may have a role in 1982). The propagation of the Saccharum species in field resistance to salt stress (Abebe et al. 2003; Zhifang and conditions cannot guarantee that the cane was not colonised Loescher 2003). by fungi, though no infestation was observed. The presence Within the commercial cultivar Q117, sorbitol was only of xylitol in sugarcane needs to be further examined to detected in mature internodes. While low relative abundance ensure that it is a sugarcane metabolite and not a fungal levels were observed in both internodes for the remaining metabolite. species there are significant differences between species, Sugar alcohols, including sorbitol, mannitol, erythritol internodes and species × internode interactions (P≤0.01 for and xylitol are commercially produced for applications as all parameters). The presence of sorbitol in the woody low calorie sweeteners. The results suggest that sugar Rosaceae family is well documented, with concentrations alcohols naturally occur within species of the Saccharum ranging from 15 to 80% of the soluble carbohydrate content complex and although they are in small quantities there depending on tissue/organ (Bieleski 1982; Nadwodnik and may be potential to increase the yields of these compounds. Lohaus 2008). Plantago coronopus plants undergoing salt The effect of accumulating an alternative sugar product in stress had increased sorbitol concentrations that may play a the tissue would need to be examined carefully. Transgenic role as an osmoregulator (Gorham et al. 1981). sugarcane plants expressing the sorbitol-6-phosphate dehy- The cyclic polyol, myo-inositol, and its monomethylated drogenase gene, from Malus domestica, were able to derivative, pinitol, were present in all the Saccharum accumulate sorbitol in the leaf lamina (∼120 mg g−1 DM) complex species. Differences in myo-inositol relative and stalk pith (∼10 mg g−1 DM) and although this was not abundance values are significant for species, internodes lethal there was a negative effect on growth and metabolism and species × internode interaction (P≤0.01 for all (Fong Chong et al. 2007). parameters). The abundance of myo-inositol is higher in immature tissue than mature tissue for all species except E. Sugars Rarely Found in Plants arundinaceous where the values are similar. Relative abundance values for pinitol were significantly different The detection of turanose and melezitose in the samples for species and internodes (P≤0.01 for both) with slightly may indicate that there was some insect exudate present on higher values in immature than mature internodes for all the material. Turanose is an isomer of sucrose that is species. Pinitol was found to be present in a small number reportedly not synthesised, cleaved or transported by plant of Proteaceae species; when present normally in concen- enzymes (Sinha et al. 2002), while melezitose has only trations higher than inositol (Bieleski and Briggs 2005). In rarely been reported in plants (Farrant et al. 2009). There alfalfa, pinitol levels increased under salt stress suggesting was no significant difference (P>0.05) for turanose, but a that pinitol was acting as a compatible solute (Fougere et al. higher abundance of turanose in immature internodes of 1991). Pinitol has also been linked to drought or salt S. spontaneum and M. sinesis than in mature internodes, tolerance in Sesbania aculeate (Ashraf and Harris 2004), which was the opposite of that observed for the remaining Mesembryanthemum crystallinum (Paul and Cockburn species. The amounts of the trisaccharide melezitose was 1989) and soybean (Streeter et al. 2001). significantly different between species and internodes (P≤ The four-carbon polyol, erythritol was not detected in S. 0.01 for both), with higher abundance in the immature than edule or in the mature internodes of commercial hybrid mature tissue for all species except S. edule that showed no Q117 and M. sinesis, but was found in all of the other difference between internodes. Both turanose and melezitose samples; with significant difference between species have been detected in honey and are thought to be formed (P≤0.01) and a significant species × internode interaction by the action of honeybee or aphid glucosidases which also (P≤0.05). Erythritol has been identified in a range of have some transferase activity, particularly at high concen- plants including some grasses and Primula (Stacey 1974; trations of sucrose (Da Costa Leite et al. 2000; Hogervorst Bieleski 1982). et al. 2007). It is possible that small amounts of turanose and Differences in relevant abundance values of xylitol were melezitose are formed in plants such as sugarcane, as a side significant for species and internodes (P≤0.01 for both) reaction similar to the 1-kestose production from sucrose by with greater differences observed in immature internodes of invertase from some organisms at high sucrose concen- S. edule, S. officinarum and S. robustum than mature trations (Pollock and Cairns 1991). Tropical Plant Biol. (2010) 3:110–122 119 Conclusion Moisture Content Measurements Apart from glucose, fructose and sucrose, the quantities of Moisture content was determined after weighing samples of soluble sugars detected were low, but it is clear that a wider each internode (fresh mass—FM) and then drying to range of sugars is made by species in the Saccharum complex constant weight (dry mass—DM). than has previously been described. Some of the sugar alcohols found in these species would have value as Glucose, Fructose and Sucrose (GFS) Extraction alternative products if the concentrations could be increased. It may be possible to selectively breed or genetically modify Stem samples (<2 g) were weighed into 15 mL tubes. particular species to increase the production of valuable Distilled water (9.9 mL) was added to each tube (Hamilton sugars for harvest. The ability to use sugarcane as a bio- diluter). Samples were incubated in a waterbath overnight factory for alternative sugar production has been demon- at 70°C; the solution was decanted and stored in a 50 mL strated by Wu and Birch (2007) with the production of tube. Another 9.9 mL water was added to the original transgenic sugarcane synthesising isomaltulose. The positive sample, and the samples were incubated overnight at 70°C results of Wu and Birch (2007) demonstrate that total sugar again. The second solution was mixed with the first, and an content of sugarcane may be increased and introduction of aliquot was taken for HPLC analysis. HPLC analysis was other sugars could be a way of increasing value of performed as described in Glassop et al. (2007). sugarcane whilst maintaining sucrose content. The roles of the reported complement of sugars in Metabolite Extraction and GC-MS sugarcane need to be further researched. Carbon partitioning and accumulation can be affected by a large number of Aqueous extractions were performed to prepare samples for molecules with a signalling role, including sugars. Sugar analysis of sugars by combined gas-chromatography/mass sensing and signalling can modulate plant growth, develop- spectrometry (GC-MS). Aliquots of 60 µL were dried for ment and response to stress. Understanding these pathways GC-MS analysis. Sugars were identified by comparison to a will be important in designing strategies to optimise the library of standards using methods developed by Glassop et production of both sucrose and alternative products with al. (2007). The following changes were made to the GC- higher value from sugarcane. MS protocol, in order to increase the separation of sugars to avoid co-elution. The oven temperature initiated at 70°C, increased to 160°C at 6°C/min, a second temperature Materials and Methods increase to 226°C at 2°C/min, a final temperature increase to 330°C at 6°C/min and hold for 10 min. The MS scanned Plant Tissue from mass 70 to 600 m/z every 0.4 s (with an interscan time of 0.05 s), starting at 7.5 min and ending at 73 min. Samples of stalk internode were collected from an immature internode (internode 4) and a mature internode Peak Identification and Determination of Relative (internode 7–15) from plants grown in soil at the CSIRO Abundance Method Development Davies Laboratory in Townsville, Australia (Lat. 19° 15′S, Long. 146°46′E) in the late afternoon 18th October 2006. One chromatogram of each cultivar and each internode was Plants were regularly watered and grown as single isolated examined in detail, using characteristic mass to charge plants in a germplasm garden in the open under the same ratio’s (m/z), to identify which sugars were present. These conditions. Internodes were numbered from the top of the sugars were then incorporated into an automated method stem towards the base in accordance with Moore (1987); that identified their presence by confirming m/z, retention internode 1 is immediately below the node to which the time and comparison with mass spectra from publically and first fully expanded leaf subtends. Triplicate samples were privately available libraries specific for each sugar. The collected from a multistem plant of the following species: automated method was then used to identify and measure Erianthus arundinaceous (IJ76-394), Miscanthus sinensis the peak areas from the chromatograms of all samples. This (NC-100), Saccharum spontaneum (Mandalay), S. edule procedure resulted in the measurement of 28 sugars via GC- (NG57-78), S. robustum (NG-289), S. officinarum (Goru) MS. Peak area is equivalent to metabolite concentration; and Saccharum hybrid (cv. Q117). Internode samples were since calibration curves were not performed for each quickly frozen in liquid nitrogen after harvesting and stored metabolite the peak area is used as a relative abundance at −80°C. Frozen tissue was ground using liquid nitrogen value. Differences in relative abundance values are indica- and a mortar and pestle. Ground tissue was used for tive of differences in concentration. However the peak area measuring variables as detailed below. of one metabolite cannot be compared to another metabolite 120 Tropical Plant Biol. (2010) 3:110–122 due to different chemistries which affect sensitivity. (Goodacre et al. 2007). As zeros are present in the data Glucose, fructose and sucrose were not measured from set a set value is added to all values. The set value was the GC-MS chromatograms but by HPLC because the determined from the smallest peak area greater than zero levels of abundance of these three sugars were outside the and dividing this value by 1,000. Log 10 transformation detection range of the GC-MS chromatograms, which was performed on these values. The transformed values were designed to ensure detection of sugars present at low were analysed with ANOVA, Genstat. Genstat was also concentrations. HPLC techniques are well established for used for the ANOVA and least significant difference GFS measurements. analysis of moisture content and GFS data. Normalisation of Results Acknowledgments Louise Ryan was supported by a vacation student scholarship from the Cooperative Research Centre for Sugar Industry Innovation through Biotechnology. The authors would also Peak areas obtained from the quantification method were like to thank both CSIRO internal reviewers and the anonymous divided by peak area of ribitol (internal standard) and fresh journal reviewers for their suggested improvements to the manuscript. or dry weights of samples extracted to get relative abundance value /g fresh or dry mass respectively, in accordance with methods presented in Glassop et al. (2007). References Histochemistry and Immunolabelling Abebe T, Guenzi AC, Martin B et al (2003) Tolerance of mannitol- accumulating transgenic wheat to water stress and salinity. Plant Physiol 131:1748–1755 For histochemical staining, thin sections of sugarcane stalk Ashraf M, Harris PJC (2004) Potential biochemical indicators of tissue were cut by hand with a razor blade. Callose ((1-3)- salinity tolerance in plants. Plant Sci 166:3–16 β-glucan) was detected by fluorescence of sections under Avonce N, Leyman B, Thevelein J et al (2005) Trehalose metabolism and glucose sensing in plants. Biochem Soc Trans 33:276–279 UV illumination after staining with the aniline blue Bieleski R (1982) Sugar alcohols. In: Loewus FA, Tanner W (eds) fluorochrome (Evans et al. 1984; Biosupplies Australia Plant carbohydrates vol 13A. Springer-Verlag, New York, p 158 Pty. Ltd., Parkville Vic. 3052, Australia) according to the Bieleski RL, Briggs BG (2005) Taxonomic patterns in the distribution manufacturer’s protocol. Starch was detected by staining of polyols within the Proteaceae. Aust J Bot 53:205–217 Bonnett GD, Hewitt ML, Glassop D (2006) Effects of high with I2-KI and visualised as black deposits by bright field temperature on the growth and composition of sugarcane microscopy. In control experiments, similar tissues sections internodes. Aust J Agric Res 57:1087–1095 that had not been stained with histochemical reagents were Bosch S (2005) Trehalose and carbon partitioning in sugarcane. viewed under UV or bright field illumination. Dissertation,University of Stellenbosch Bosch S, Rohwer JM, Botha FC (2003) The sugarcane metabolome. For immunolabelling, stalk tissue was fixed in 77th South African Sugarcane Technologists Association Con- paraformaldehyde-glutaraldehyde, embedded in paraffin gress, 19-22 August. Durbin. 7:129–133 and sectioned as described in Rae et al. (2005). A mouse Burget EG, Verma R, Molhoj M et al (2003) The biosynthesis of L- monoclonal antibody with specificity for (1–3),(1–4)- β- arabinose in plants: Molecular cloning and characterization of a golgi-localized UDP-D-xylose 4-epimerase encoded by the glucan (Meikle et al. 1994) was purchased from Biosupplies MUR4 gene of Arabidopsis. Plant Cell 15:523–531 Australia Pty. Ltd. Sections were labelled with 10 µg/ml Carpita NC (1996) Structure and biogenesis of the cell walls of antibody following the method of Rae et al. (2005) except grasses. Ann Rev Plant Physiol Plant Mol Biol 47:445–476 that the secondary antibody was chicken anti-mouse IgG Claeyssen É, Rivoal J (2007) Isozymes of plant hexokinase: occur- rence, properties and functions. Phytochemistry 68:709–731 conjugated to Alexa Fluor 488 used at a final concentration Cobbett CS, Medd JM, Dolezal O (1992) Suppressors of an arabinose- of 2 µg/ml. In control sections, the primary antibody was sensitive mutant of Arabidopsis thaliana. Aust J Plant Physiol omitted. 19:367–375 All sections were examined under a Zeiss Axioskop Da Costa Leite JM, Trugo LC, Costa LSM et al (2000) Determination of oligosaccharides in Brazilian honeys of different botanical microscope and images recorded with an Olympus DP-70 origin. Food Chem 70:93–98 camera. Daniels J, Roach BT (1987) Taxonomy and evolution. In: Heinz DJ (ed) Sugarcane improvement through breeding. Elsevier, Amsterdam, p 7 Statistical Analysis Dumville JC, Fry SC (2003) Gentiobiose: a novel oligosaccharide in ripening tomato fruit. Planta 216:484–495 Eastmond PJ, Li Y, Graham IA (2003) Is trehalose-6-phosphate a Metabolomic experiments yield values that vary from regulator of sugar metabolism in plants? J Exp Bot 54:533–537 single digits to numbers in the hundreds and thousands; Edye LA, Doherty WOS, Blinco JA (2006) The sugarcane biorefinery: statistical analysis of these data sets with this variation can energy crops and processes for the production of liquid fuels and renewable commodity chemicals. Int Sugar J 108:19–27 lead to invalid conclusions. Transformation of the data is Eggleston G, Legendre B, Tew T (2004) Indicators of freeze-damaged common; log 10 transformation was used in this experi- sugarcane varieties which can predict processing problems. Food ment, in order to overcome the degree of variation Chem 87:119–133 Tropical Plant Biol. (2010) 3:110–122 121 Evans NA, Hoyne PA, Stone BA (1984) Characteristics and Müller J, Wiemken A, Aeschbacher R (1999) Trehalose metabolism in specificity of the interaction of a fluorochrome from aniline blue sugar sensing and plant development. Plant Sci 147:37–47 (Sirofluor) with polysaccharides. Carbohydr Polym 4:215–230 Mutalshaikhov G, Ismailov ZF (1976) Investigation of the carbohy- Farrant JM, Lehner A, Cooper K et al (2009) Desiccation tolerance in drates of Saccharum spontaneum and Arundo donax. Chem Nat the vegetative tissues of the fern Mohria caffrorum is seasonally Compd 10:655–656 regulated. Plant J 57:65–79 Nadwodnik J, Lohaus G (2008) Subcellular concentrations of sugar Felix JdM, Papini-Terzi FS, Rocha FR et al (2009) Expression profile alcohols and sugars in relation to phloem translocation in of signal transduction components in sugarcane population Plantago major, Plantago maritima , Prunus persica and Apium segregating for sugar content. Trop Plant Biol 2:98–109 graveolens. Planta 227:1079–1089 Fong Chong B, Bonnett GD, Glassop D et al (2007) Growth and Nichols BW (1974) The structure and function of plant glycolipids. In metabolism in sugarcane are altered by the creation of a new Plant Charbohydrate Biochemistry, Pridham JB (ed) Academic hexose-phosphate sink. Plant Biotechnol J 5:240–253 Press, London, pp 97–108 Fougere F, Le Rudulier D, Streeter JG (1991) Effects of salt stress on Nishizawa A, Yabuta Y, Shigeoka S (2008) Galactinol and raffinose amino acid, organic acid, and carbohydrate composition of roots, constitute a novel function to protect plants from oxidative bacteroids, and cytosol of alfalfa (Medicago sativa L.). Plant damage. Plant Physiol 147:1251–1263 Physiol 96:1228–1236 Paul M, Cockburn W (1989) Pinitol, a compatible solute in Glassop D, Bacic A, Bonnett GD et al (2007) Changes in the Mesembryanthemum crystallinum L.? J Exp Bot 40:1093–1098 sugarcane metabolome with stem development. Are they related Pollock CJ, Cairns AJ (1991) Fructan metabolism in grasses and to sucrose accumulation? Plant Cell Physiol 48:573–584 cereals. Annu Rev Plant Physiol 42:77–101 Goddijn OJM, Smeekens S (1998) Sensing trehalose biosynthesis in Pommerrenig B, Papini-Terzi FS, Sauer N (2007) Differential regula- plants. Plant J 14:143–146 tion of sorbitol and sucrose loading into the phloem of Plantago Goddijn OJM, van Dun K (1999) Trehalose metabolism in plants. major in response to salt stress. Plant Physiol 144:1029–1038 Trends Plant Sci 4:315–319 Rae AL, Perroux J, Grof CPL (2005) Sucrose partitioning between Goddijn OJM, Verwoerd TC, Voogd E et al (1997) Inhibition of vascular bundles and storage parenchyma in the sugarcane stem: trehalase activity enhances trehalose accumulation in transgenic a potential role for the ShSUT1 sucrose transporter. Planta plants. Plant Physiol 113:181–190 220:817–825 Goodacre R, Broadhurst D, Smilde AK et al (2007) Proposed Ralet MC, Thibault JF, Faulds CB et al (1994) Isolation and minimum reporting standards for data analysis in metabolomics. purification of feruloylated oligosaccharides from cell walls of Metabolomics 3:231–241 sugar-beet pulp. Carbohydr Res 263:227–241 Gorham J, Hughes L, Wyn Jones RG (1981) Low-molecular-weight Roesnner-Tunali U, Hegemann B, Lytovchenko A et al. (2003) carbohydrates in some salt-stressed plants. Physiol Plant 53:27–33 Metabolic profiling of transgenic tomato plants overexpressing Hogervorst PAM, Wackers FL, Romeis J (2007) Detecting nutritional hexokinase reveals that the influence of hexose phosphorylation state and food source use in field-collected insects that synthesize diminishes during fruit development. Plant Physiol 133:84–99 honeydew oligosaccharides. Funct Ecol 21:936–946 Rolland F, Baena-Gonzalez E, Sheen J (2006) Sugar sensing and Hoepfner SW, Botha FC (2003) Expression of fructokinase isoforms signalling in plants: conserved and novel mechanisms. Annu Rev in the sugarcane culm. Plant Physiol Biochem 41:741–747 Plant Biol 57:675–709 Irvine JE (1999) Saccharum species as horticultural classes. Theor Rolland F, Sheen J (2005) Sugar sensing and signalling networks in Appl Genet 98:186–194 plants. Biochem Soc Trans 33:269–271 Kandler O, Hopf H (1980) Occurrence, metabolism, and function of Schluepmann H, van Dijken A, Aghdasi M et al (2004) Trehalose oligosaccharides. In: Preiss J (ed) The biochemistry of plants: a mediated growth inhibition of Arabidopsis seedlings is due to comprehensive treatise, vol 3 carbohydrates: structure and trehalose-6-phosphate accumulation. Plant Physiol 135:879–890 function. Academic, New York, pp 221–270 Seymour FR, Unruh SL, Nehlich DA (1989) Quantitation of free Keating BA, Robertson MJ, Muchow RC et al (1999) Modelling sugars in plant tissue by G.l.c. of their peracetylated aldononitrile sugarcane production systems I. Development and performance and ketoxime derivatives. Carbohydr Res 191:175–189 of the sugarcane module. Field Crop Res 61:253–271 Sinha AK, Hofmann MG, Römer U et al (2002) Metabolizable and Koch K (2004) Sucrose metabolism: regulatory mechanisms and non-metabolizable sugars activate different signal transduction pivotal roles in sugar sensing and plant development. Curr Opin pathways in tomato. Plant Physiol 128:1480–1489 Plant Biol 7:235–246 Smeekens S (2000) Sugar-induced signal transduction in plants. Annu Loescher WH, Everard JD (1996) Sugar alcohol metabolism in sinks Rev Plant Physiol 51:49–81 and sources. In: Zamski E, Schaffer AA (eds) Photoassimilate Smirnoff N, Cumbes QJ (1989) Hydroxyl radical scavenging activity distrubtion in plants and crops. CRC, New York, pp 185–207 of compatible solutes. Phytochem 28:1057–1060 Loreti E, Alpi A, Perata P (2000) Glucose and disaccharide—sensing Smith AM, Zeeman SC, Smith SM (2005) Starch degradation. Annu mechanisms modulate the expression of alpha-amylase in barley Rev Plant Biol 56:73–98 embryos. Plant Physiol 123:939–948 Stacey BE (1974) Plant polyols. In: Pridham JB (ed) Plant McCormick AJ, Cramer MD, Watt DA (2008) Differential expression carbohydrate biochemistry. Academic, London, pp 47–59 of genes in the leaves of sugarcane in response to sugar Streeter JG, Lohnes DG, Fioritto RJ (2001) Patterns of pinitol accumulation. Trop Plant Biol 1:142–158 accumulation in soybean plants and relationships to drought Meikle PJ, Hoogenraad NJ, Bonig I et al (1994) A (1→3, 1→4)—β- tolerance. Plant Cell Environ 24:429–438 glucan-specific monoclonal antibody and its use in the quantitation Trip P, Nelson CD, Krotkov G (1965) Selective and preferential and immunocytochemical location of (1→3, 1→4)—β—glucans. translocation of C14-labeled sugars in white ash and lilac. Plant Plant J 5:1–9 Physiol 40:740–747 Moore P (1987) Anatomy and morphology. In: Heinz DJ (ed) Weise S, Weber AM, Sharkey T (2004) Maltose is the major form of Developments in crop science 11—sugarcane improvement carbon exported from the chloroplast at night. Planta 218:474– through breeding. Elsevier, New York 482 Mukherjee SK (1957) Origin and distribution of Saccharum. Bot Gaz Whalen MD (1991) Taxonomy of saccharum (Poaceae). Baileya 119:55–61 23:109–125 122 Tropical Plant Biol. (2010) 3:110–122 Woolard GR, Rathbone EB, Novellie L (1976) A hemicellulosic beta- tolerance in sugarcane (Saccharum Officinarum L.). J Integr D-glucan from the endosperm of sorghum grain. Carbohydr Res Plant Biol 48:453–459 51:249–252 Zhifang G, Loescher WH (2003) Expression of a celery Wu L, Birch RG (2007) Doubled sugar content in sugarcane plants mannose-6-phosphate reductase in Arabidopsis thaliana modified to produce a sucrose isomer. Plant Biotechnol J 5:109–117 enhances salt tolerance and induces biosynthesis of both Zhang SZ, Yang BP, Feng CL et al (2006) Expression of the Grifola mannitol and a glucosyl-mannitol dimer. Plant Cell Environ Frondosa trehalose synthase gene and improvement of drought- 26:275–283 APPENDIX 2 – ENZYMATIC SYNTHESIS OF GLUCOSYL SUCROSE The trisaccharide, glucosyl sucrose was detected in the cyanobacterium, Nostoc, and is thought to be synthesised from sucrose by the action of a glucosidase enzyme. Two genes encoding putative glucosidases (aG1 and aG2) were identified, synthesised, and cloned into E.coli. The vector system chosen utilises the T7 promoter with a 6 x His tag to produce high level expression of the protein of interest in E. coli. The tag allows purification on a nickel column and identification of the enzyme by western blots using a Ni-NTA-alkaline phosphatase detection system. Initially, the genes aG1 and aG2 were expressed by autoinduction in BL21 (DE3) cells. Although the majority of the protein was recovered as insoluble inclusion bodies, significant amounts were soluble (Figure a1-1). insoluble soluble auto 0 hr auto1 2 1 2 1 2 80kDa70 aG1 aG2 50 1. aG1 2. aG2 Figure A1-1 Analysis of enzyme production in E. coli cultures that express the Nostoc genes aG1 or aG2. After overnight autoinduction, some soluble protein at the correct molecular weight (arrows) was detected on a protein blot developed with Ni-NTA alkaline phosphatase. Methods for purifying the enzymes were then tested. Following affinity-purification using the Ni tag, partially purified aG1 and aG2 enzymes were recovered (Figure A1-2A). The identity and viability of these enzymes were tested in an assay with the substrate, nitrophenol-glucoside. In this assay, cleavage of the glucoside unit releases nitrophenol which is monitored by an increase in absorbance at 595 nm. Purchased yeast glucosidase was used as a positive control. Figure A1-2B shows that the fractions recovered from the affinity purification contained glucosidase activity. These experiments confirmed that the gene predictions were correct and that the enzymes had the desired activity after expression and purification. A aG1 F1 F2 F3 F4 F5 aG2 F1 F2 F3 F4 F5 115kDa 80 70 aG1: 93kDa aG2: 98kDa 50 B 0.12 1 0.80.1 0.6 0.08 0.4 0.2 0.06 0 0.04 0 0.05 0.1 Yeast glucosidase (U) 0.02 0 1 2 3 4 aG1 fractions aG2 fractions Figure A1-2 (A) Protein blot showing partially purified aG1 and aG2 enzymes in the eluted fractions F1 to F5 following separation on a Talon affinity column. (B) Assay for glucosidase activity in fractions from the affinity column. Results are expressed as absorbance at 595 nm, equivalent to release of nitrophenol from the substrate. Maximum activity was obtained in fraction 2, corresponding to the presence of the bands shown in (A). Inset shows a linear relationship between absorbance and the amount of the control enzyme, yeast glucosidase. Further experiments were then carried out to increase the ratio of soluble to insoluble enzyme produced by this system. For enzyme aG2, this was achieved by re-cloning into a vector that includes the Trx sequence, which is known to increase solubility of cloned proteins (Figure A1-3). This approach was not successful with aG1, which appeared to be degraded. insoluble soluble aG2 aG1 aG2 aG1 3. 2. 1. 0hr 3. 2. 1. 3. 2. 1. 0hr 3. 2. 1. Induction time 115kDa 80 65 50 aG1-Trx: 110kDa aG2-Trx: 116kDa AAAAbbbbssssoooorrrrbbbbaaaannnncccceeee 555599995555nnnnmmmm AAAAbbbbssss.... 555599995555nnnnmmmm Figure A1-3 Protein blot showing production of aG1 and aG2 enzymes following recloning with the Trx tag and induction with IPTG (0.3 mM) for 1- 3 h at room temperature. Increased amounts of soluble aG2 were obtained however aG1 appeared to be degraded into smaller fragments. Larger scale expression of enzyme aG2 in the Trx vector was then tested in two different cell lines (Figure A1-4) to determine the optimum conditions. Prolonged expression resulted in degradation of the protein, suggesting that protease inhibitors would be required. A large-scale production and purification was carried out by the UQ Protein Expression Facility (PEF) using vectors and conditions described here. A B C D 140kDa 115 A: soluble T=0 80 B: soluble T=3 Rosetta C: soluble T=o/n Rosetta D: soluble T=3 BL21 Figure A1-4 Western blot using Trx monoclonal antibody on soluble (left) and insoluble (right) fractions following expression of vector aG2 pET32a induced with 0.3 mM IPTG for 3 h at room temperature. The protein has a predicted molecular weight of 115.7 kDa and predicted pI of 5.77. Fractions from the large scale purification peformed by PEF were separated on a SDS- polyacrylamide gel transferred to a PVDF membrane and labelled with the Trx antibody. The results (Figure A1-5A) showed that the aG2-Trx fusion protein was recovered in fractions #4-11, however, a large amount of low molecular weight degradation product was also detected. The activity of fractions #6-11 was tested with the artificial substrate, nitrophenyl-6Glc, which contains a short chain of 6 glucose units. Fractions #8-10 were active against this substrate (Figure A1-5B). The activity of fraction #9 was shown to increase in the presence of divalent cations (Figure A1-5C), which is characteristic of glucosidases. Yeast glucosidase was used as a control in these experiments. Unfortunately the enzyme appeared to be very unstable, as activity was abolished when trying to remove imidazole by dialysis. Incubation of the enzyme with sucrose and maltose suggested that there was no activity with these substrates, as no glucosyl-sucrose could be detected by HPLC. However, the enzyme was shown to be active against maltotriose. The results suggest that this enzyme is indeed a glucosidase, but that it has specificity for longer chain glucosides such as malto-oligosaccharides and has little activity against sucrose in its present form. As the enzyme was not active on sucrose and was unstable no further expression & purification was carried out. f4 f11 A B A405 0.008 0.067 0.091 0.09 0.084 0.074 (blank substracted) fraction 6 7 8 9 10 11 room temperature 37 C incubation C yeast glc f9+DTT f9+Mn2+ f9+Mg2+ yeast glc f9+DTT f9+Mn2+ f9+Mg2+ 1.614 0.02 0.028 0.039 1.595 0.044 0.086 0.064 Figure A1-5 Purification and assay of glucosidase. (A) Western blot using Trx monoclonal antibody on fractions from affinity purification of aG2- Trx protein showed recovery of enzyme in fractions 4-11. (B) Activity of fractions against the substrate nitrophenyl-6xGlc, showed highest activity in fractions 8-10. (C) Activity of fraction #9 increased in the presence of the divalent cations Mn+ and Mg+. APPENDIX 3 - ENZYMATIC SYNTHESIS OF KETOSUCROSE Ketosucrose has previously been detected in the bacterium Agrobacterium tumefaciens and is thought to be synthesised from sucrose by the enzyme, glucoside-3- dehydrogenase (G3DH). When the sequence of the Agrobacterium genome was published it was not possible to identify the gene encoding G3DH due to a lack of well- defined homologues in other species. Since then, the G3DH gene has been identified in a number of bacterial species including Halomonas and Gramella. We used these sequences to identify the homologous gene in Agrobacterium tumefaciens and Stenotrophomonas maltophilia. The A. tumefaciens and S. maltophilia G3DH genes were synthesised and subcloned into a protein expression vector. The vector system chosen utilises the T7 promoter with a 6 x His tag to produce high level expression of the protein of interest in E. coli. The tag allows purification on a nickel column and identification of the enzyme by western blots using a Ni-NTA-alkaline phosphatase detection system. The gene was initially expressed in E. coli strain BL21(DE3). Analysis of the proteins in both the soluble and insoluble fractions on SDS-polyacrylamide gels indicated that the protein was being produced but not folded correctly, resulting in the production of insoluble inclusion bodies. This is a common problem in protein expression studies and a variety of strategies have been used by other researchers to overcome the problem. Growing the cultures under different temperatures or with different conditions for induction was tested but did not improve solubility of the protein. We also tested expression of the plasmid in a range of host strains containing various chaperones (plasmid set from Takara Ltd.) to assist with folding but this was not successful. Advice from Dr Ulrike Kappler suggested that folding may be assisted by secretion of the protein into the periplasmic space, so the G3DH genes were re-cloned into an expression vector containing a periplasmic export signal (plasmid pET22b). This strategy resulted in the best recovery of soluble protein, however significant amounts remained in inclusion bodies in the insoluble fraction (Figure A2-1). insoluble soluble G E G E E (A. tumefaciens): 66kDa G (S. maltophilia): 65kDa Figure A2-1 Analysis of glucoside-3-dehydrogenase (G3DH) enzyme produced in E.coli cultures that express the gene from A. tumefaciens (E) or S.maltophilia (G). Expression was induced by adding IPTG (0.3 mM) for 2 h at room temperature. Bands at the correct molecular weight are circled. Some soluble protein was observed following recloning of the G3DH genes into an expression vector containing a periplasmic export signal, however large amounts were present in the insoluble fraction. In order to obtain larger amounts of soluble protein, inclusion bodies were purified and refolding conditions were tested using a kit from Athena Enzyme Systems (QuickFold™). The system provides 15 combinations of refolding reagents in a factorial matrix design to identify key buffer components resulting in soluble protein. The FAD cofactor (20 μM) was also included in all buffers. The amount of soluble protein recovered from each buffer combination was analysed by SDS-PAGE, transferred to PDVF membrane followed by detection with the Ni-conjugate. The results showed that several of the buffers allowed the protein to re-fold and remain in the soluble fraction (Figure A2-2). 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 80kDa 65kDa 50kDa Figure A2-2 Protein blot of soluble enzyme recovered by dialysis after refolding tests in 15 buffer combinations. In each lane, a blue band at 65 kDa molecular weight indicates that refolding was successful in this buffer. A further test was carried out on the refolded enzyme samples to test whether the enzymes retained dehydrogenase activity. This assay used an electron acceptor substrate (DCPIP) as an analogue for the native sugar substrate. Activity was detected as a decrease in absorbance at 595 nm, corresponding to electron transfer activity of the recombinant protein (Figure A2-3). This experiment demonstrated that our predictions of the enzyme action were correct and that the enzyme had been successfully cloned and expressed. buffer 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 ΔAbs(595) 0.10 0.17 0.21 0.17 0.21 0.20 0.06 0.02 0.00 0.00 0.03 0.06 0.04 0.02 0.00 Figure A2-3 Results of an activity assay for enzyme recovered from each of 15 refolding buffer combinations. Dehydrogenase activity is expressed as change in absorbance at 595 nm. Buffers 2 to 6 produced soluble enzyme which retained dehydrogenase activity. Refolding using buffers #2 to #6 was then tested with a larger scale sample of protein. After dialysis and concentration of the refolded sample, the protein was successfully recovered from the soluble fraction (Figure A2-4). Electron acceptor assays confirmed that the protein retained the ability to transfer electrons. Although it was not possible to obtain a sample of ketosucrose as a standard, methods were available to assay for ketosucrose by HPLC, thin layer chromatography and colorimetric determination in the presence of NaOH. However, when the enzyme was incubated with sucrose under a variety of conditions, no ketosucrose could be detected in the reaction products. Since the native enzyme is found in the periplasm of Agrobacterium, it is possible that it requires additional proteins for activity against sucrose. Future work may be able to identify the accessory proteins and complete the synthesis of ketosucrose. Resolubilsed Refolded G Insoluble fraction Inclusion body After dialysis E. coli after o/n MW E & G & concentration incubation 65 kDa soluble fraction Refolded G E. coli after o/n T = 0 controls After dialysis incubation Figure A2-4 Protein blot of soluble enzyme recovered after refolding APPENDIX 4 Fruit fly bioassay to distinguish ‘sweet’ sugar structures Jason Hodoniczky1,2, Gregory J. Robinson1,2, Elizabeth A. McGraw3 & Anne L. Rae1,2 1 CSIRO Plant Industry, 306 Carmody Rd, St Lucia, Queensland 4067, Australia. 2 Cooperative Research Centre for Sugar Industry Innovation through Biotechnology, University of Queensland, St Lucia, Queensland 4072, Australia. 3 School of Biological Sciences, University of Queensland, St Lucia, Queensland 4072, Australia. Correspondence: Dr Anne Rae (anne.rae@csiro.au) Keywords: carbohydrate, behavior, Drosophila, glucobiose, sucrose, glucose Abstract Palatable response to dietary sugars plays a significant role in influencing metabolic health. New structures are being explored with beneficial health properties, although consumer acceptance relies heavily on desirable sensory properties. Despite the importance of behavioral responses, the ability to elucidate structure-preference relationships of sugars is lacking. Using a wild population of Drosophila melanogaster as a model, we performed pair-wise comparisons across structural groups to characterize a fruit fly bioassay to determine sugar palatability. Preference was successfully described in structurally relevant terms, particularly through the ability to test sugars of related structures directly, in addition to standard sucrose comparisons. The fruit fly bioassay also provided the first report on the palatability of gentiobiitol and in making reference to known human preferences also raises opportunities for greater understanding of behavioral response to sugars generally. Page 1 of 18 Introduction With a foundation dating back centuries, and once reserved for the privileged, sweet tasting carbohydrates (predominantly sucrose) are extensively added to many modern food products. As concerns grow about their health implications current efforts are focussed on developing alternative sugars with the proviso that suitable properties, including desirable taste, are maintained. Due to the complexities of performing human sensory trials, published data on structure-taste relationships is fragmentary, not well supported by experimental evidence or quoted in product information without reference. With a host of new carbohydrate products being developed and entering the market in recent years, an ability to assess define structures with favourable response is of interest. The utility of the fruit fly as a model for research in chemosensory behavior is well accepted, with a degree of similarity between flies and mammals in the way palatable responses are invoked {Scott, 2005 #370}. A recent report has also demonstrated similarity in the response of Drosophila melanogaster (fruit fly) to high intensity sweeteners recognised by humans {Gordesky-Gold, 2008 #195}, extending on earlier work using Phormia regina (blow fly){Ahamed, 2001 #206}. This is particularly intriguing considering the variation in response displayed for these compounds in more closely related mammals, such as new world monkeys {Glaser, 2002 #368; Laska, 1998 #343}. With this is mind, the current study characterized structural groups of sugars in a fruit fly bioassay by comparing preferences to known ‘sweet’ sugars. The format also proved useful in targeting linkages responsible for heightened palatability through multiple comparisons with di- and trisaccharides containing related glucosidic linkages. Page 2 of 18 Materials & Methods Materials The di- and trisaccharides, kojibiose, nigerose, isomaltose, sophorose, gentiobiose, leucrose, and panose were sourced from Carbosynth (Berkshire, UK), while erlose, maltotriose and maltotriitol were supplied by Hayashibara Biochemical Laboratories (Okayama, Japan). Laminaribiose was purchased from Megazyme (Co. Wicklow, Ireland) and all remaining carbohydrates were purchased from Sigma-Aldrich (St. Louis, MO). A 200 mM stock solution of each carbohydrate was prepared in distilled water (dH20) and stored in aliquots at -20 oC. Synthesis of gentiobiitol Synthesis of gentiobiitol via the reduction of gentiobiose was performed by Epichem Ltd (Murdoch, Australia) and was based on a modification to the method of Abdhel-Akher et al.{Abdel-Akher, 1951 #367} Briefly, 4.8 g of gentiobiose was dissolved in 50 mL of dH20 and combined with 1.0 g of sodium borohydride in 25 mL dH20. The reaction was allowed to proceed at room temperature for 4 h and quenched with acetic acid after confirming the reaction to be complete by thin layer chromatography. The solution was then concentrated and the product precipitated using methanol. Further purification, deacetylation and concentration were carried out to yield a final solution of 0.70 M gentiobiitol in dH20 (4.05 g, 99 % purity). Collection and maintenance of Drosophila melanogaster populations A new population of wild-type flies was established from 10 female Drosophila melanogaster captured on the University of Queensland campus between 27th February 2009 and 6th March 2009. Traps consisted of empty 1 mL pipette tip boxes baited with mashed banana and sprinkled with live yeast {Loeschcke, 2007 #358}. Standard 1 mL pipette tips Page 3 of 18 with the ends cut off were inserted to create a one-way entrance to the trap. Traps were deployed in the field for 24 h surrounded by Tanglefoot® insect barrier (The Tanglefoot Co. Grand Rapids, MI) to prevent ants and other crawling insects from entering. Traps were inspected for flies and females were transferred to separate vials containing standard corn meal nutrient medium. The individual females were monitored until their offspring eclosed and their sons could be identified. Identification of Drosophila melanogaster males was carried out by examining the sex combs, as D. melanogaster show distinct differences to other commonly found Drosophila species. Laboratory stocks of Oregon-R were used in pilot experiments. All flies were reared in 250 mL bottles at 25 °C on a 12:12 h light:dark regime on standard corn meal medium. Two-choice behavioral assay Assays were carried out using a 96-well plate with a layer of Parafilm® stretched over it to keep the food and flies on the surface and out of the wells. Test sugars were mixed with either brilliant blue (25 mg mL-1) or erythrosine (90 mg mL-1) (New Directions; Marrickville, Australia) in 0.5 % agarose. Initially, plates were replicated with the colours inverted for each sugar to test for colour bias, but after 12 trials this approach was not continued as it was determined that colour had no affect on preference as had been reported previously {Thorne, 2004 #359}. A minimum of 50, < 5 day-old flies were starved for 24 h on filter paper soaked in dH20 for each experiment {Thorne, 2004 #359},{Al-Anzi, 2006 #208},{Dahanukar, 2001 #360},{Dahanukar, 2007 #283},{Gordesky-Gold, 2008 #195},{Ueno, 2008 #361}. Both males and females were used as it has been reported that sex does not affect feeding preference {Dahanukar, 2001 #360}. All feeding experiments were carried out in the morning as circadian rhythm can affect feeding behavior {Meunier, 2007 #362},{Xu, 2008 Page 4 of 18 #363}. Flies were allowed to feed for 2 h in the dark before being frozen for 48 h. Scoring of abdomen color was carried out visually with a dissecting microscope (Olympus SZ51, Center Valley, PA/ Zeiss Stemi 2000, Thornwood, NY) and the flies were grouped into the following categories: red (R), blue (B), purple (P) and none. Red and blue flies were distinct whereas the purple flies varied in shade depending on the quantity of sugar eaten. Preference Index (PI) was calculated as the number of red or blue flies + ½ number of purple flies divided by the total number of flies that fed; PI = R or B + ½ P / (R + B + P). Tests were only included in PI calculations if more than 20 % of the flies were feeding {Dahanukar, 2001 #360},{Gordesky-Gold, 2008 #195}. When using the wild-type population, feeding rate < 20% were not observed and rates were commonly > 50 %. Each paired comparison was replicated at least 3 times with separate plates and flies. Results & Discussion D. melanogaster were exposed to traditional sugars, including glucose and sucrose, and a range of potential or current alternative sweeteners. Fruit fly sugar preferences were determined using two-choice behavioral assays based on the method of Tanimura et al.{Tanimura, 1982 #344} Initial optimization was performed to confirm that food dyes did not influence preference and to establish reproducible conditions for fruit fly behavior. In addition to confirming earlier studies, including the importance of ageing {Nestel, 1985 #375; Nigg, 1995 #374}, we found that a consistent time of day for starving and feeding improved reproducibility of assays. We suggest this improvement is likely due to the effect of circadian rhythms on feeding behaviour {Meunier, 2007 #362},{Xu, 2008 #363}. A fruit fly line newly established from a wild population was employed for data on carbohydrate structures as it was generally observed to be more consistent in its behavioural response, displaying higher feeding rates throughout the experiments than inbred laboratory Page 5 of 18 lines. Preference index (PI) was calculated to measure the palatability of one carbohydrate in relation to another, determined from the feeding experiments following scoring of abdomen colour. PI for each carbohydrate choice was a proportion of 1.0, with a value of 0.50 equating to an equal preference of the two sugars being tested. This approach enabled gustatory responses to sugars to be reported in a defined and sensitive manner. The carbohydrate structures used in the study (Table 1) were selected to represent broad structural groups and target specific monosaccharide linkages through selection of related di- and trisaccharides. Tests were performed with equimolar (4 mM) solutions, including standard comparisons paired with either sucrose or glucose. The first comparisons focussed on defining the preferences of the fruit fly, relative to commonly held views of human carbohydrate preference. A comparison of four α- glucobioses along with their corresponding β-glucobiose were initially all tested against equimolar sucrose (Figure 1a), confirming the general preference for alpha structures over beta known in humans{Pangborn, 1966 #345}. While the α-glucobiose samples were more preferred than their corresponding β-glucobiose in each case, significant consumption of the beta sugars did occur. This was particularly unexpected with the β-1,6 glucobiose (gentiobiose) which has been reported to have a ‘bitter’ taste in humans{Birch, 1970 #347},{Cote, 2009 #348},{Pangborn, 1961 #346}. A similar comparison was therefore conducted by directly pairing the alpha and beta carbohydrates as the two opposing choices (Figure 1b). Differences were more pronounced using a direct approach, with the preference for α- over β-glucobioses confirmed and all statistically significant. The greatest difference in PI occurred with the α-1,4 and β-1,4 samples (maltose and cellobiose respectively), suggesting a possible strong palatable response by Drosophila towards α-1,4 carbohydrates in general. Interestingly, the difference between the α-1,6 (isomaltose) and β-1,6 (gentiobiose) sugars remained small. This limited difference between gentiobiose and the Page 6 of 18 related α-linked structure may suggest a somewhat elevated preference for gentiobiose relative to other β-linked structures. The plausibility of this may be strengthened by the presence of β-1,6 glucans in yeast cell walls, an extract of which is included in the artificial diet of the fruit fly and found as part of their natural diet. Similarly, with such a pairwise test this may also indicate a reduced preference for isomaltose, or more specifically α-1,6 linked glucose. To characterise our model system further, we calculated preferences for a second structural group of sugars, the sucrose isomers. Sucrose and three isomers were compared to either sucrose or glucose (again equimolar). In these assays (Figure 1c) turanose was a more strongly preferred isomer than leucrose or palatinose. A previous study detecting responses of sugar-sensing neurons in fruit fly has reported similar findings, with sucrose and turanose displaying greater responses over leucrose and palatinose {Dahanukar, 2007 #283}. These preferences correlate with those suggested for humans {Godshall, 2007 #356},{Shibuya, 2004 #357}, although turanose is poorly studies in humans and suggested to be approximately half as ‘sweet’ as sucrose. D. melanogaster could be seen to have an elevated response to turanose, which may be a consequence of the greater natural abundance over the two other isomers tested, since turanose is present at higher levels in nectar and honey{Burgin, 1997 #349},{de la Fuente, 2007 #350}. In broad terms, these experiments reinforced the utility of the fruit fly behavioral assay relative to human preferences, as glucose comparisons resulted in consistently greater PI for the opposing carbohydrate than sucrose for all samples. Similarly glucose is generally accepted as being approximately 75 % as ‘sweet’ as sucrose in humans. Furthermore, the control comparing sucrose to itself in Figure 1C resulted in a PI close to 0.5 as expected. Page 7 of 18 Having characterized a range of linkages with generally accepted information on palatability, we then applied the assay to a compound with no previous data. The ability to increase preference for α-linked disaccharides following conversion to a sugar alcohol is known; with disaccharide alcohols such as maltitol now commonly used as alternative sweeteners with desirable dental health properties. The effect of reduction on β-linked structures is less studied. Conversion of the β-glucobiose, cellobiose, into a sugar alcohol has been shown to decrease ‘sweetness’ in one study with human subjects{Kearsley, 1980 #351}. In the present study gentiobiitol, the reduced product of the β-1,6 linked glucobiose (gentiobiose), was assessed to determine any changes to preference. Gentiobiitol has recently been identified in transgenic sugarcane plants engineered to produce sorbitol {Chong, #365} and the potential applications of this novel sugar as an alternative sweetener is of interest. A comparison to equimolar sucrose was first performed (Figure 2a), and gentiobiitol was shown to be somewhat palatable with a PI of approximately 0.30, though not significantly more than gentiobiose compared to sucrose in the previous data (Figure 1a). Further comparisons were conducted with glucose and maltitol, which revealed preferences close to 0.50 or similar palatability to these known sweeteners. Glucose and maltitol have similar levels of ‘sweet’ taste in humans as seen with the fruit fly data and further testing of the β-linked sugar alcohol gentiobiitol is of interest. The comparison of gentiobiitol with gentiobiose was also conducted as it has been shown that comparing similar structures directly, resolves preferences between structures more clearly than indirect comparisons of both structures to sucrose for example. In this instance, although not statistically significant (at p < 0.05), the final comparison in Figure 2a does suggest preference for the sugar alcohol over the related β-glucobiose. A similar comparison of the sugar alcohol sweetener malitol was conducted along with the gentiobiitol data (Figure 2b), pairing sucrose, glucose, maltotriitol or maltose in feeding assays. Relative to sucrose, maltitol was reported to have a PI of 0.44 (± 0.02), and with glucose a PI of (0.62 ± 0.03). So while these comparisons are somewhat higher relative Page 8 of 18 to those of gentiobiitol and may seem to contradict the result between maltitol and gentiobiitol directly in Figure 2a, we are attempting to differentiate changes in PI of ~ 0.1 which does seem to be at the limits of the bioassay. We do again see that results between the sucrose and glucose comparisons are greater with the less ‘sweet’ compound glucose, further indicative of consistency between assays. Final behavioral experiments in Figure 3 were performed firstly to investigate the effect of chain length through analysis of trisaccharide structures. Trisaccharides displayed elevated preferences when compared to sucrose, which is where variation to the accepted carbohydrate preferences of humans can be seen. Stronger preference towards erlose and melezitose (Figure 3a) over sucrose would not be expected in humans, as it is generally accepted that increased chain length decreases palatability, although further opportunity exists here for more detailed human studies also. For the fruit fly, however, this is not unexpected because of the greater likelihood of being encountered in their natural diet. Melezitose and erlose are present in honeydew for example, and have been described as a food source for other fruit fly species{Christenson, 1960 #364},{Hendrichs, 1993 #373},{Hogervorst, 2007 #352}. Additionally, the strong preference for these trisaccharides has been reported in other insects {Tinti, 2001 #353},{Glaser, 2002 #368}. In Figure 3b maltotriose exhibited a discerning preference over maltose, with increased chain length of an additional α-1,4 glucose invoking an elevated response in D. melanogaster. This provides a further example of the strength of the bioassay in comparing related structures to define preferences structurally, as separate comparisons of maltotriose or maltose to sucrose (Figure 3b and 1a respectively) showing minimal difference. Preference for starch related linkages i.e. those containing α-1,4 or α-1,6 glucose, were able to be explored in the trisaccharide comparisons also, and highlighted because of their dietary Page 9 of 18 significance. Panose, which contains both an α-1,4 and α-1,6 linkage, showed higher preference over glucose although a markedly low PI when compared to maltose (Figure 3a). This was consistent with the suggestion of a reduced preference for isomaltose (α-1,6 glucobiose) mentioned earlier. A separate assessment of maltotriose preference revealed a high PI (0.90 ± 0.02) over panose (Figure 3b), and further evidence for the preference of α- 1,4 glucose over α-1,6 by D. melanogaster. This preference for α-1,4 glucose also dominates response towards sugar alcohols, with maltotriose showing high PI over both maltitol and maltotriitol (Figure 3b). Similarly, in an earlier comparison maltitol demonstrated a significantly lower PI when paired with maltose (Figure 2a). So while this deviates from the increased response of humans to maltitol over maltose, it is consistent with the strong response invoked by α-1,4 glucose with fruit fly. In terms of human preference for starch- related structures, we are not aware of any extensive studies to define the relative preference of α-1,4 glucose over α-1,6, though it has been reported that soluble starch can enhance sucrose ‘sweetness’ in humans {Kanemaru, 2002 #355}. So whilst human taste may be influenced by starch related carbohydrates, they directly invoke strong appetitive behavior in Drosophila with an ability to discriminate α-1,4 over α-1,6 glucose. In summary, we have demonstrated the strength of a behavioral assay using the model organism Drosophila melanogaster that while seemingly simple in its execution, is reproducible, sensitive and can reveal specific structure-preference relationships of carbohydrates not easily obtained with humans. We extend work on the similarities between the fruit fly and human responses (summarised in Table 2) and note the strength in the similarity between species for disaccharides, with deviation reported in the response towards trisaccharides (particularly those containing α-1,4 glucose). Opportunities for improved understanding of the response of specific carbohydrate structures in humans have been highlighted, including the sucrose isomer turanose, and starch related glucosidic linkages. Page 10 of 18 The latter may be particularly of interest because of the health benefits of altered starch structures for example, and sensory evaluation is being performed after incorporation into common food products {Baixauli, 2008 #371; Sanz, 2009 #372}. Finally, results suggest a sugar alcohol produced from a β-glucobiose may prove palatable and consider it of interest to examine the human response to gentiobiitol, particularly alongside gentiobiose. Acknowledgements We acknowledge the support of the Cooperative Research Centre for Sugar Industry Innovation through Biotechnology for project funding and thank Dr Rosanne Casu and Dr Graham Bonnett for their comments on the manuscript. References Page 11 of 18 Table 1 : Structural groups of carbohydrates used in preference assays. Rationalization of carbohydrate groups aided structure-preference analyses. All structures are alpha-linked except the group of beta-glucobioses as shown, and the reduced structure derived from the β- 1,6 member of this group known as gentiobiitol. Monosaccharides are abbreviated to glc = glucose, and fru = fructose. α ‐glucobioses kojibiose glc‐1,2‐glc nigerose glc‐1,3‐glc  maltose glc‐1,4‐glc isomaltose glc‐1,6‐glc β‐glucobioses sophorose glc‐1,2‐glc laminaribiose glc‐1,3‐glc  cellobiose glc‐1,4‐glc gentiobiose glc‐1,6‐glc sucrose isomers sucrose glc‐1,2‐fru turanose glc‐1,3‐fru leucrose glc‐1,5‐fru palatinose glc‐1,6‐fru sugar alcohols maltitol reducedmaltose gentiobiitol reduced gentiobiose maltotriitol reduced maltotriose trisaccharides melezitose glc‐1,2‐fru‐1,3‐glc erlose glc‐1,4‐glc‐1,2‐fru panose glc‐1,6‐glc‐1,4‐glc maltotriose glc‐1,4‐glc‐1,4‐glc Page 12 of 18 1.00 a 0.90 0.80 0.70 0.60 0.50 0.40 ** ** * 0.30 0.20 0.10 α β α β α β α β 0.00 glc‐1,2‐glc glc‐1,3‐glc glc‐1,4‐glc glc‐1,6‐glc 1.00 *b 0.90 0.80 ** ** 0.70 * 0.60 0.50 0.40 0.30 0.20 0.10 α β α β α β α β 0.00 glc‐1,2‐glc glc‐1,3‐glc glc‐1,4‐glc glc‐1,6‐glc 1.00 c 0.90 ** ** 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 sucrose turanose leucrose palatinose Figure 1: Characterizing sugar preferences of Drosophila melanogaster. Pairwise preference assays were performed with < 5 day old fruit flies by introducing them to a 96- well plate containing two equimolar sugar samples with either a red or blue food dye. After feeding, scoring of the abdomen color enabled calculation of the preference index (PI). A PI > 0.5 indicates elevated preference towards the indicated test sugar. Preference of α- over β- gluocobioses was examined by comparing all 8 samples to sucrose (a) as well as by directly comparing the alpha and beta structures to one another (b). The plots in b, which show PI for Page 13 of 18 Preference index (PI) both sugars, more clearly reveals the preference for α-glucobiose carbohydrates than when separately compared to sucrose. Further characterization of preference for a structural group was performed by examining sucrose isomers (c). PI was calculated relative to both sucrose (black bars) and glucose (white bars). A similar response towards the isomer turanose as sucrose was seen, while leucrose and palatinose were less preferred. The general observation of an increased response to all test sugars when compared to glucose than sucrose, confirms the elevated response of D. melanogaster towards sucrose over glucose that is consistent with humans. The control of sucrose with itself shows a PI close to 0.5 as expected also. Errors shown are ± SEM, (n = 3 – 6). Paired t-test performed with * corresponding to p < 0.05 and ** p < 0.01. Page 14 of 18 1.00 a 0.90 0.80 0.70 0.60 0.50 0.40 * 0.30 0.20 0.10 0.00 sucrose glucose maltitol gentiobiose 1.00 b 0.90 0.80 0.70 0.60 0.50 0.40 0.30 * 0.20 0.10 0.00 sucrose glucose maltotriitol maltose Figure 2 : Preference of novel compound using fruit fly bioassay. (a) Preference of the sugar alcohol gentiobiitol was determined relative to known ‘sweet’ compounds, including sucrose, glucose and maltitol, as well as gentiobiose from which is was prepared. While less preferred than sucrose, a similar PI to equimolar glucose and maltitol was observed. An increased preference relative to gentiobiose is suggestive of an improved palatable response to -linked structures through conversion to a sugar alcohol, though further testing is required. (b) Further assays with the known sugar alcohol maltitol enabled further, albeit indirect, comparisons for gentiobiitol. The sucrose and glucose comparisons revealed only slightly elevated PI for maltitol than the same result for gentiobiitol, though discerning a difference in PI of approximately 0.1 seems at the limits of the assay and may be seen to confirm the similarity of gentiobiitol and maltitol. Unlike gentiobiitol, maltitol is less preferred than the structure from which it is derived, with maltitol expected to be more ‘sweet’ in humans. This can be seen as consistent for the fruit fly, however, because of their high palatable response towards -1,4 glucose than seen in humans. Interestingly, the effect Page 15 of 18 Preference index (PI) of an additional -1,4 glucose seems minimal with the presence of a sugar alcohol, though, as PI for maltitol did not vary significantly when compared to maltotriitol. Errors shown are ± SEM, (n = 3 – 5). Paired t-test with * corresponding to p < 0.05 and ** p < 0.01. Page 16 of 18 1.00 a 0.90 melezitose erlose panose 0.80 * * 0.70 ** * 0.60 0.50 0.40 0.30 0.20 ** 0.10 0.00 sucrose glucose sucrose glucose maltose glucose maltose 1.00 b ** * ** ** 0.90 0.80 ** 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 sucrose glucose maltose maltitol panose maltotriitol Figure 3 : Preference trisaccharides with an emphasis on starch-related structures. (a) The PI was determined for the trisaccharide indicated at the top of the bars, compared to glucose as well as sucrose and / or maltose. In general trisaccharides showed higher preference over glucose and consitituent disaccharides which is consistent with other insect studies, though increased chain length is suggested to reduce human response. A distinct reduction in PI with panose when paired with maltose provides further evidence of lower palatability towards α-1,6 over α-1,4 glucose structures. (b) Maltotriose shows elevated PI when paired with a variety of structures, including sugar alcohols. Again the strong response to α-1,4 over α-1,6 glucose when assessed against panose is observed. Additionally, increased chain length is favourable in the case of the high PI for maltotriose when compared to maltose. Indirect comparisons looking across comparisons with sucrose is unable to discern such a clear difference as this direct comparison. Errors shown are ± SEM (n = 3 - 4). Paired t-test with * corresponding to p < 0.05 and ** p < 0.01. Page 17 of 18 Preference index (PI) Table 2 : Summary of carbohydrate preferences by fruit fly relative to human responses. Results from two-choice behavioral assays performed with Drosophila melanogaster are summarised with respect to structure-preference relationships. Comparisons to human preferences avoid the complexities associated with quoting relative values, rather providing a summary of accepted views and highlighting gaps in knowledge. Fruit fly Human α‐glucobiose > β‐glucobiose α‐glucobiose > β‐glucobiose  sucrose > glucose sucrose > glucose  sucrose, turanose > leucrose, palatinose  sucrose > turanose, leucrose, palatinose maltose > maltitol maltitol> maltose gentiobiitol≅maltitol ≅ glucose maltitol≅ glucose, gentiobiitolunknown α‐1,4 glucose > α‐1,6 glucose any difference unknown Page 18 of 18 APPENDIX 5 Oral and intestinal digestion of carbohydrates in structurally relevant terms Jason Hodoniczky1,3, Dionne Clayton2, 3, Carol Morris2, 3 & Anne L. Rae1,3* 1CSIRO Plant Industry, 306 Carmody Rd., St Lucia, Queensland 4067, Australia 2 Centre for Phytochemistry and Pharmacology, Southern Cross University, Military Road, Lismore, NSW 2480, Australia 3Cooperative Research Centre for Sugar Industry Innovation through Biotechnology, University of Queensland, St Lucia, Queensland 4072, Australia Correspondence: *Email: anne.rae@csiro.au Telephone: +61 7 32142379 Keywords: Sugar, glucoside, glucobiose, Streptococcus, -glucosidase Abstract The impacts of dietary carbohydrates on human health are well recognized, representing both the need to minimize adverse effects (through high caloric and cariogenic sugars) and opportunity for improvements in wellbeing (with prebiotics for example). With the increasing market presence of alternative carbohydrate products aiming to address a range of modern health issues, a comparative study of oral and intestinal digestion across structural groups was conducted. Use of in vitro oral and intestinal -glucosidase assays provided comparison of various glucosidic linkages and chain length. Fermentation of carbohydrates by Streptococcus mutans highlighted the diversity of structures utilized by the oral bacterium, though -1,2, -1,3 and -1,6 glucobioses were additional structures to sugar alcohols and sucrose isomers found not to promote formation of organic acids. A mammalian glucosidase assay defined the relative digestibility of sucrose isomers, and the effect of variable linkages among -glucobiose substrates, with -1,4 > -1,3 > -1,2 > -1,6. Investigation of starch-related trisaccharides by anion-exchange chromatography revealed the reduction in glucose release observed for maltotriose, erlose and maltotriitol resulted from feedback inhibition of digestion products. Further comparisons of biological interactions among varying carbohydrate structures are likely to be informative in the design of functional food products. Epichem Pty Ltd Murdoch University South Street MURDOCH WA 6150 Phone +61 (8) 9360 7696 Facsimile +61 (8) 9360 7699 ABN 80 106 769 902 Dr Jason Hodoniczky CSIRO Molecular and Health Technologies Gate 3, Normanby Rd, Clayton, VIC, 3168 QUOTATION N°. 081030-CSIRO Item Structure Quantity Time Cost (ex GST) OH O All product derived $4,000 HO O from 1g Gentiobiose 1 HO OH 1-2 weeksOH HO All product derived HO OH from 5g Gentiobiose $6,000 OH Notes: 1. The product to be supplied >95% purity. 2. Gentiobiose starting material supplied by client. Robert Gauci Head of Laboratory (WA) 30 October 2008 General Conditions of Sale for Epichem Pty Ltd 1. Compounds are characterised by MP, TLC, GC/MS, 1H NMR as appropriate to their physical properties. Other analyses such as HPLC, CHN, 13C NMR (including 2D experiments), etc, are also available but may attract a surcharge unless specified in the Variations on General Conditions above. 2. Compounds supplied are free from significant impurities by 1H NMR (300MHz) and TLC. Any special requirements for purity can usually be met but may attract a surcharge unless specified in the Variations on General Conditions above. 3. Compounds are supplied for laboratory purposes only. The toxicological properties of these compounds may not have been established. It is the buyer’s responsibility to ensure the compounds are handled in an appropriate manner by suitably qualified personnel. 4. Epichem Pty Ltd shall not in any event be liable for any incidental, consequential or special damages of any kind resulting from any use or failure of any compound supplied. 5. Epichem Pty Ltd shall not be held liable for any loss, damage or penalty as a result of any delay in or failure to manufacture or deliver any compounds. 6. Any taxes, duty, custom, inspection or testing fee, or any other tax, fee or charge whatsoever imposed by any governmental authority shall be met by the buyer. 7. Unless specified otherwise, the compounds are supplied on a non-exclusive basis to the buyer and no intellectual property rights are conferred to them. Epichem retains any intellectual property rights which may result from any novel synthetic chemistry it develops in the preparation of any compounds. 8. Unless specified otherwise in the Variations on General Conditions above, Epichem reserves the right to supply the compounds or any intermediates or by-products prepared during the synthesis to other parties. 9. Unless otherwise stated, quotations are in Australian dollars and valid for 30 days.