Fructan Synthesis

The model developed by Edelman and Jefford (13) for synthesis of the inulin type fructans in Helianthus tuberousus has been applied to both dicots and monocots although the latter species are often assumed to contain additional enzymes that synthesize the more complex fructans known to occur in situ. Briefly, the model is that suscose:sucrose fructosyl transferase (1-SST; E.C. synthesizes the trisaccharide 1-kestose (isokestose) from sucrose, and that trimer elongation occurs via the reaction of fructan:fructan fructosyl transferase (FFT; The continued widespread application of this model was challenged several years ago largely due to concerns about the purity of the enzymes studied and because the oligomeric fructans produced in vitro frequently do not reflect those found in vivo (10, 11). These concerns are valid and will be discussed; nonetheless, application of this model to the synthesis of grass fructans still serves well as a foundation upon which to examine the recent literature on fructan synthesis in grasses.

2.1. Trisaccharide formation

Sucrose: sucrose fructosyl transferase activities in grasses have been documented in many studies, yet few of the enzymes have been rigorously purified prior to characterization and in most studies, the trisaccharide allegedly produced by SST was not identified. Characterization of the trisaccharide produced by putative SST containing preparations is critical as a complete model for synthesis of fructans in some grass species (e.g. barley, wheat) must account for the synthesis of both 1-kestose and 6-kestose. Studies of impure preparations have led to concerns over whether the invertase activity present in SST preparations is a function of the SST enzyme itself or results from contaminating invertases. As Pollock and Cairns (10) have pointed out, invertases can cleave not only sucrose to its component monomers but can also transfer fructosyl residues from one to another sucrose resulting in the production of 6-kestose, 1-kestose or neokestose depending upon the reaction conditions. Hence, invertase contamination of SST preparations could result in the erroneous conclusion that trisaccharide synthesis resulted solely from the in vitro SST activities measured. Most of the literature that suggests involvement of SST in synthesis of trimeric fructans in grasses is not definitive due to the study of impure preparations.

Recently, a series of studies have established the nature of two enzymes from barley that synthesize fructan trisaccharides (14,15,16,17). Simmen et al. (14) identified two isoforms of SST that were induced by light and separated them from the predominant invertases, which in the absence of light induction did not produce fructan trimers. Using sucrose as the substrate, one of the light-induced, partially purified SST isoforms produced 1-kestose as the sole trisaccharide, hence this isoform is referred to as 1-SST. Glucose and a small amount of fructose were also produced by this 1-SST. The fructose production was attributed to a low level of contamination by invertase. The second SST isoform incubated with sucrose produced glucose and 6-kestose, hence this isoform is subsequently referred to as a 6-SST. This 6-SST preparation produced considerably more fructose than was produced by the 1-SST preparation. The authors attribute the production of 6-kestose and fructose to the 6-SST because the production of both of these was only slightly reduced, and reduced to the same extent, when the 6-SST preparation was incubated with sucrose in the presence of pyridoxal. Pyridoxal is an inhibitor of invertases from many species including those from barley that are not induced by light (14,18). Interestingly, this partially purified 6-SST was also shown to produce bifurcóse when incubated in the presence of both sucrose and 1-kestose. Although the purity of the 6-SST and the 1-SST was not established in this study, the selective induction by light and minimal inhibition of activity by pyridoxal certainly suggests that barley leaves contain light inducible 1-SST and 6-SST, each with some ability to function as an invertase.

2.1.1. Sucrose:sucrose 1-fructosyl transferase (1-SST)

The most highly purified 1-SSTs isolated in grasses have been isolated from barley and tall fescue. Luscher et al. (19) have recently purified a 1-SST activity from barley leaves to apparent homogeneity. Analysis by isoelectric focusing revealed the presence of two isoforms with pis of 4.93 and 4.99. Examination of these isoforms by SDS-PAGE showed that both had a 50 kD subunit and a 22 kD subunit. A thorough study by Luscher and Nelson (20) of SST extensively purified from tall fescue showed the presence of three isoforms with pis of 4.13, 4.23 and 4.36. Examination of the most purified preparations of these three isoforms by SDS-PAGE showed 3 to 5 proteins. All three preparations contained proteins of approximately 80, 58 and 25 kD. While the latter study included proteolytic inhibitors to eliminate proteolytic degradation as an explanation of the multiple bands found in extensively purified preparations of SST, neither of these studies determined if the native molecular weight of the active enzymes could account for two subunits of approximately 22-25 kD and 50 - 58 kD. Such experiments were done with a 1-SST purified from roots of Cichorium intybus and showed that the native molecular weight was 69 kD as determined by gel filtration under native conditions. Only two bands were detected by SDS-PAGE and the molecular masses were 24 and 49 kD. The purity of the enzyme was well established as there was only one band on a silver-stained native electrophoresis gel and the preparation yielded only one N-terminal sequence. Additionally, MALDI-TOF MS of this same fraction detected molecular ions at m/z 19896 and at 49092 and the heterodimer at 68792 (21); thus, the heterodimeric structure of this 1-SST is clear. Hence, the 1-SST isolated from barley likely was purified to homogeneity and those from tall fescue were highly purified, but not to homogenity.

Characterization of the reactions catalyzed by the 1-SSTs from barley and tall fescue showed considerable variation. Both isoforms of the 1-SST from barley produced 1-kestose, glucose and fructose when incubated with sucrose. The production of fructose was saturated at 30 mM sucrose. The production of 1-kestose and glucose, which paralleled each other, was not saturated even at 500 mM sucrose. Both isoforms released fructose when incubated with 1-kestose. Consequently, it appears that both of these isoforms from barley contain SST, invertase and fructan exohydrolase activities. The three isoforms isolated from Festuca produced 1-kestose from sucrose making it appropriate to refer to these as 1-SSTs, but there was also some invertase function (contaminant or inherent) present. Furthermore, each isoform also produced neokestose and at least one isoform produced nystose making these isoforms quite broad in their catalytic capabilities. Indeed, this is perhaps too broad to be considered similar to the 1-SSTs in the original Edelman and Jefford (13) model. When a preparation containing a mixture of all three isoforms of 1-SST from tall fescue was incubated with sucrose, in addition to the aforementioned products, a complex series of oligofructans ranging from DP 4-8 was produced and appear identical to the fructans isolated from the growing leaf. Both of these studies provide useful information, but do not yet allow a definitive conclusion that the same protein is responsible for all of the activities.

2.1.2. Sucrose:sucrose 6-fructosyl transferase (6-SST)

Subsequent studies of the 6-SST from barley leaves included further purification and characterization (15), cloning and functional expression of the gene in Nicotiana plumbaginifolia (16), a plant lacking fructans, and in Pichia pastoris (17), a methyltropic yeast lacking in fructans and in sucrose metabolizing enzymes (22). These studies have provided the first thorough description, both physically and kinetically, of such a fructan synthesizing enzyme from a grass species. The 6-SST purified from barley leaves had a molecular mass of 67 kD as determined by size exclusion chromatography under nondenaturing conditions, and was composed of two subunits, one -50 kD and the other -20 kD, as determined by electrophoresis under denaturing conditions (15,16). The deduced amino acid sequence of the longest cDNA obtained was shown to contain the sequence of both subunits (16). Isoelectric focusing analysis of the 6-SST purified from barley revealed an isoform with a pi of 4.9 and another isoform with a pi of 5.1. These isoforms were shown to have almost completely identical peptide fragment patterns upon digestion with trypsin (16) and indistinguishable catalytic properties and, therefore, may represent the same gene product with minor variations in post-translational modifications (e.g. the extent of glycosylation).

The substrate specificity and the identity of reaction products of the 6-SST purified from barley leaves were thoroughly examined (15,16). Both isoforms used sucrose to produce 6-kestose as the predominant fructan along with traces of 1-kestose, bifurcóse and 6b-kestin. Additionally, both purified isoforms hydrolyzed sucrose to glucose and fructose. The ratio of the invertase function to the transferase functions of the purified enzymes was shown to be regulated by various fructans or related carbohydrates. For example, when sucrose was the sole substrate the transfer of fructosyl units to water (invertase function) was the major activity and the production of fructans was a minor activity. However, of particular importance was the observation that purified enzyme incubated with both sucrose and 1-kestose synthesized bifurcóse at rates greater than those of either sucrose hydrolysis or 6-kestose production in the presence of only sucrose (15). This observation prompted the name of the enzyme to be changed to sucrose:fructan 6-fructosyltransferase (6-SFT) to better reflect its catalytic capabilities. The ability of this enzyme to synthesize fructans larger than trisaccharides will be discussed further in the following section on trimer elongation. Expression of the 6-SFT cDNA in protoplasts of Nicotiana plumbaginifolia resulted in a functional enzyme with activities much the same as those of the enzymes purified from barley leaves (16). Likewise, expression of the 6-SFT cDNA in Pichia pastoris verified that the enzyme functions as an invertase, a 6-SST and a 6-SFT (17). Interestingly, the expression of the enzyme in Pichia also resulted in 1-SST activity suggesting that under some circumstances the protein is capable of forming both 1-kestose and 6-kestose. However, because this activity was not present when the enzyme was purified from barley or when the cDNA was expressed in Nicotiana, its relevance is unclear. Nonetheless, it is clear that the sucrose:fructan 6-fructosyl transferase present in leaves of barley has several catalytic functions in addition to that of trisaccharide synthesis.

2.2. Chain elongation

The formation of fructans greater than DP 3 was attributed to the action of fructan: fructan fructosyl transferase (FFT) in the classic model of Edelman and Jefford (13). This enzyme reversibly transfers fructosyl residues between oligomeric and polymeric fructans with 1-kestose being the preferred donor; sucrose is not a fructosyl donor but it is an acceptor (13, 23). In this model, chain elongation is by transfer of fructosyl units from fructans and does not occur directly from sucrose. The terminology of Waterhouse and Chatterton (24) for fructan structures greater than DP 3 is used in the following sections.

2.2.1. Fructan:fructan 1-fructosyl transferase (1-FFT)

To facilitate purification of FFT from grasses containing fructans more complex than inulins and whose substrate trisaccharides were not readily commercially available, Jeong and Housley (25) devised an assay for measurement of FFT activity that is independent of SST activity for the generation of FFT substrate. This proved crucial to characterization of FFT from wheat. This assay relies upon the transfer of the terminal fructose from nonradiolabeled 1-kestose to [14C] sucrose and subsequent detection of the production of radioactive 1-kestose. Using this assay at pH 7, where many fructan metabolizing enzymes have little or no activity, these authors chromatographically separated FFT activity of wheat leaves from fructan exohydrolase (FEH) and SST. The absence of SST in the FFT preparation was established as no radiolabeled trisaccharide was synthesized when sucrose, both radiolabeled and unlabeled, were provided as substrate. The absence of invertase was concluded as the synthesis of trisaccharides from sucrose by invertase typically requires concentrations of sucrose much higher than the 20 mM used in this study and typically results in the production of multiple trisaccharides. Although the absence of various contaminating activities was addressed in this study, the actual purity of the enzyme preparation was not established.

The FFT isolated from wheat used 1-kestose to synthesize 1,1-kestotetraose (1,1-nystose) and 1,1,1-kestopentaose (1,1,1-logose). It was concluded that this enzyme is a 1-FFT with the ability to synthesize inulin-type fructans up to DP 5. Although inulin-type fructans up to DP 6 are found in wheat, they do not appear to be abundant (6). The enzyme was unable to synthesize fructans with the (5-2,6-linkage but could use 6-kestose as an acceptor of a p-2,1-fructosyl transfer from 1-kestose to form bifurcóse and 6,1-kestoteraose (6,1-nystose), a fructan that is rarely found in wheat. Since wheat contains highly branched fructans with mixtures of (3-2,1 and (32,6- fructosyl units and a terminal glucose (6,7), other activities are clearly required to synthesize the full spectrum of fructans present in wheat. This could be satisfied by a 6-FFT such as that found in Asparagus officinalis (26). Such an FFT has not been isolated from a grass. It should be noted that this FFT was demonstrated to transfer the terminal fructose from 1-kestose to sucrose synthesizing yet another 1-kestose. (3-2,1-linked oligofructans of DP 4 and 5 were also able to donate fructosyl units to sucrose, but at much lower rates than that when 1-kestose was the donor. Whether or not this ability to transfer fructosyl units to sucrose is an inherent property of the FFT or due to the presence of another enzyme was not fully established.

2.2.2. Sucrose:fructan 6-fructosyl transferase (6-SFT)

Another enzyme capable of elongating fructan trisaccharides is the 6-SFT purified from barley leaves that was discussed in section 2.1 for its ability to synthesize 6-kestose. As stated, a particularly important observation was that this enzyme can synthesize bifurcóse (1 and 6-kestotetraose) from sucrose and 1-kestose and that it does so at rates greater than those at which it synthesizes 6-kestose from sucrose. Duchateau et al. (15) thoroughly examined the substrate specificity and the reaction products of this 6-SFT. To do this, they purified a variety of potential acceptor fructans from several sources and tested each fructan in the presence of sucrose for the ability of the purified 6-SFT to transfer a fructosyl unit to the fructan. The reaction products were identified via co-chromatography with known standards on two different HPLC separation matrices. Incubation of the purified 6-SFT with sucrose plus each of eight different oligofructans resulted in the production of a single new fructan containing the addition of one (3-2,6- fructosyl moiety linked either to the terminal glucose or to a (32,6- fructosyl chain. This study documented the ability of this enzyme to synthesize four different pentaoses and one kestohexaose. Thus, this purified 6-SFT synthesizes the formation of the trisaccharide 6-kestose, initiates a new (3-2,6-linked chain and extends an existing [3-2,6-fructan chain.

The relative rates of the different fructosyl transferase activities of 6-SFT were also examined (15). Sucrose as the sole substrate resulted in 78% of the total fructosyl transfer being to water (invertase function), ~ 20% to sucrose (6-SST function) and negligible 6-SFT activity was observed. Trisaccharide and tetrasaccharide fructans as substrates in the absence of sucrose resulted in a severe reduction of the total fructosyl transferase activity. Either tri- or tetrasaccharide fructans and sucrose together as substrates resulted in total fructosyl transferase activities comparable to those observed with sucrose alone; however, the ratio of the transferase functions was significantly changed. Of the total transferase activity observed when sucrose and 1-kestose were both present, 78% was to 1-kestose (6-SFT function), <20% was to water, and <5% was to sucrose. Cloning and expression of functional 6-SFT has verified that these activities result from expression of the 6-SFT cDNA. It is, therefore, clear that this enzyme synthesizes a trisaccharide and a variety of DP 4, 5 and 6 fructans.

2.3. Reconsidering the model

There are differences between enzymology of fructan synthesis described above and that described in the model Edelman and Jefford developed based on studies of Helianthus tuberosus (13). The key difference is the role of sucrose. In grasses, at least in barley, the elongation of fructan chains greater than DP 3 occurred by the direct transfer of a fructosyl unit from sucrose to an elongating fructan chain. In contrast, in Helianthus the only fructan synthesized directly from sucrose was the trisaccharide, 1-kestose. As pointed out by Wiemken et al. (27), this difference is not trivial because the transfer of fructosyl units from sucrose to a fructan is exergonic and irreversible. These authors elaborated upon the scheme presented by Duchateau et al (15) and presented a new model for fructan synthesis in grasses. In this model, 1-SST is still the enzyme responsible for synthesis of 1-kestose, which is the preferred acceptor for elongation by 6-SFT. 6-SFT

uses sucrose as the fructosyl donor and attaches it to, first, 1-kestose to form bifurcóse. Then, 6-SFT uses sucrose as a source of a fructosyl group to attach to bifurcóse. Additional cycles of 6-SFT use of sucrose would result in the synthesis of graminans. The data certainly support this model for synthesis of graminan type fructans.

0 0

Post a comment