Fueled by the current interest in the link between diet and health (Ma et al., 2008; Estruch et al., 2009; Lomax and Calder, 2009; Schiffrin et al., 2010), there is now a surge of research by various groups into the biological activities and potential beneficial effects of dietary saccharide biopolymers (Schepetkin and Quinn, 2006; Cumashi et al., 2007; Vos et al., 2007; Rideout et al., 2008; Weickert and Pfeiffer, 2008; Bruzzese et al., 2009; Chan et al., 2009; Courtois, 2009; Graff et al., 2009; Jacobs et al., 2009; Liu et al., 2009). These biopolymers represent a broad range of structurally diverse (Schepetkin and Quinn, 2006), non- or low-digestible, dietary soluble fibers that have been derived from different species of flora including higher plants, fungi, lichens and algae, and have been demonstrated (in-vitro or in-vivo) to affect beneficially one or more target cellular or body functions (Cumashi et al., 2007; Ho et al., 2007; Hua et al., 2007; Vos et al., 2007; Kim and Joo, 2008; Ma et al., 2008; Bruzzese et al., 2009; Chan et al., 2009; Courtois, 2009; Graff et al., 2009; Lomax and Calder, 2009; van den Heuvel et al., 2009).

The physiological effects attributed to these ‘functional’ polysaccharides can be divided into six main categories: (1) beneficial effects on colonic microflora (Jacobs et al., 2009) and gastrointestinal physiology; that is functioning as a prebiotic (Macfarlane et al., 2008; de Vrese and Schrezenmeir, 2008; Lomax and Calder, 2009); (2) immunomodulatory effects (Leung et al., 2006; Schepetkin and Quinn, 2006; Cumashi et al., 2007; Hua et al., 2007; Vos et al., 2007; Benyacoub et al., 2008; Kim and Joo, 2008; Bruzzese et al., 2009; Burrows et al., 2009; Graff et al., 2009; Mizuno et al., 2009); (3) anti-angiogenetic (Cumashi et al., 2007) and anti-tumor effects (Chan et al., 2009; Liu et al., 2009); (4) altered lipid metabolism (Rideout et al., 2008; Chen and Huang, 2009); (5) improved bioavailability of essential minerals (Bo et al., 2006; van den Heuvel et al., 2009); and (6) other beneficial health effects such as enhanced production of growth factors involved in re-epithelization and wound healing (Jettanacheawchankit et al., 2009).

In this respect, many of the dietary saccharide polymers studied (isolated from >35 species) (Schepetkin and Quinn, 2006), exhibiting multifunctional physicochemical and physiological characteristics, are now being redefined as secondary metabolites (Vos et al., 2007; Macfarlane et al., 2008; Weickert and Pfeiffer, 2008; Courtois, 2009), or biological response modifiers (BRMs) (Leung et al., 2006).

These BRM polysaccharides (Table 1) have diverse structural complexity, and belong to a wide range of low or high molecular weights, soluble or insoluble, acidic or neutral, homo- or hetero-polymers, which may or may not undergo varying degrees of degradation in the digestive tract either by bacterial enzymes or by specific conditions associated with the digestive process (Sinnott et al., 2007; Macfarlane et al., 2008; van den Broek et al., 2008; Arasaradnam et al., 2009; Courtois, 2009; Jacobs et al., 2009; Schiffrin et al., 2010). Recent evidence suggests that the potential for these polysaccharides to undergo some degree of degradation is highly dependent on the metabolic potential encoded by the combined genomes of the gut microbiota (Arasaradnam et al., 2009; Tuohy et al., 2009).

Table 1 Examples of dietary polysaccharides that have been studied in relation to their biological activities

Current studies into the link between the bioconversion of these dietary polysaccharides, their bioavailability and their downstream effects on the host metabolism and physiology are using metabolomics and metagenomics approaches (Jacobs et al., 2009). These and other innovative approaches in the field of colonic fermentation (Chassard et al., 2007; Arasaradnam et al., 2009) are providing novel insights into gut microbial–human mutualism (Possemiers et al., 2009), its impact on regulating human health and disease, and the importance of dietary modulation (Leung et al., 2006; Macfarlane et al., 2008; Bruzzese et al., 2009; Liu et al., 2009; Lomax and Calder, 2009; Tuohy et al., 2009; Schiffrin et al., 2010).

Although many of the physiological traits attributed to dietary polysaccharides may be secondary, that is, related to their effects on the gut microflora and its biochemical activities (Schiffrin et al., 2010), evidence from various pharmacodynamic and pharmacokinetic studies also indicate microflora-independent immunomodulatory effects (Schepetkin and Quinn, 2006; Ho et al., 2007; Hua et al., 2007).

This ‘direct’ immunomodulatory effect is instigated by the binding of certain BRM polysaccharides (for example, β-glucans) (Dalmo and Bogwald, 2008) to specific receptors on immune cells in the gut-associated lymphoid tissues, and results in the intracellular activation of signal transducers and transcription factors that are associated with various effector functions of the immune response (Vetvicka et al., 1996; Schepetkin and Quinn, 2006; Ho et al., 2007; Hua et al., 2007; Vos et al., 2007; Paur et al., 2008; Burrows et al., 2009; Graff et al., 2009; Mizuno et al., 2009).

Although the potential mechanisms of the interaction and the subsequent signal transduction pathways are not fully understood, the evidence supports the direct binding of BRM polysaccharides (or their fragments) to pattern recognition receptors. These receptors are key players in the immune response (receptors for recognition of microbial polysaccharides) and include toll-like receptors (Leung et al., 2006; Hua et al., 2007; Dalmo and Bogwald, 2008; Graff et al., 2009), non-toll pattern recognition receptors (for example, β-glucan receptor or dectin-1), complement receptor type 3 and certain transmembrane lectins (Vetvicka et al., 1996; Vos et al., 2007; Gunning et al., 2009). Recognition by these pattern recognition receptors can result in intracellular signaling cascades for example, activation of protein kinases or nuclear transcription factor kappa B. This can in turn result in subsequent activation or inactivation of a wide spectrum of target genes involved in the regulation of a variety of cellular responses, such as expression of various cell-surface receptors (Kim and Joo, 2008) and cytokine production (Leung et al., 2006; Schepetkin and Quinn, 2006; Cumashi et al., 2007; Dalmo and Bogwald, 2008; Paur et al., 2008; Jettanacheawchankit et al., 2009).

One of the primary immunomodulatory effects of BRM polysaccharides is to promote or alter various leukocyte activities, in particular those of macrophages (Leung et al., 2006; Schepetkin and Quinn, 2006) and immune-regulatory Gamma delta T cells (a sub-population of intraepithelial lymphocytes in the gastrointestinal tract) (Graff et al., 2009), via changes in cytokine expression (Yoshino et al., 2000; Hua et al., 2007; Benyacoub et al., 2008; Kim and Joo, 2008; Burrows et al., 2009; Schiffrin et al., 2010). This in turn can impact both the innate and adaptive arms of the immune response and thus result in activation or dampening of these responses (Schepetkin and Quinn, 2006; Cumashi et al., 2007; Ho et al., 2007; Vos et al., 2007; Graff et al., 2009; Mizuno et al., 2009).

Dietary polysaccharide induced immunomodulatory activities of note include: increased macrophage cytotoxic and phagocytic activities, altered pro- and anti-inflammatory (Leung et al., 2006; Schepetkin and Quinn, 2006; Ho et al., 2007; Benyacoub et al., 2008; Kim and Joo, 2008; Mizuno et al., 2009; Schiffrin et al., 2010) and Th1-Th2 balance (Yoshino et al., 2000; Vos et al., 2007; Burrows et al., 2009), as well as altered expression of certain adhesion molecules (Yoshino et al., 2000; Schepetkin and Quinn, 2006; Cumashi et al., 2007; Ho et al., 2007; Graff et al., 2009). All of which supports the notion that certain dietary plant polysaccharides may have significant immunomodulatory potential (Schepetkin and Quinn, 2006; Cumashi et al., 2007; Ho et al., 2007; Benyacoub et al., 2008; Burrows et al., 2009; Trinchero et al., 2009; Tuohy et al., 2009).

In view of this and given the important link between changes in the immune response and the glycosylation of various effector glycoproteins (Alavi and Axford, 2008), the aim of this study was to examine the possible affects, if any, that oral administration of dietary plant polysaccharides may have on the glycosylation status of serum glycoproteins in a cohort of healthy subjects. We chose a commercial, mixed saccharide dietary supplement (AA), the constituents of which have previously been shown to exhibit prebiotic (Sinnott et al., 2007), as well as possible immunomodulatory activities (Lefkowitz et al., 2000; Koetzner et al., 2010), and for the purposes of this study have no known reported toxicity or negative side effects associated with their use in human subjects (Stancil and Hicks, 2009).

Using AA we carried out an open-label dosing study to first evaluate safety of increased doses, and second to monitor the affect on the N-glycosylation status of serum glycoproteins in healthy subjects.

Materials and methods

Study cohort

Healthy volunteers were recruited from the staff at St George's University of London. The study design was approved by the Ethics Committee and volunteers gave written informed consent.


The supplement was Advanced Ambrotose (AA) powder (provided by Mannatech, Inc., Dallas, TX, USA). AA is a plant-derived dietary supplement that contains a standardized mixture of partially purified polydisperse saccharide biopolymers: aloe vera gel powder (containing glucomannan-based polysaccharides), arabinogalactan (high-M.wt soluble arabinose and galactose-containing dietary fiber), gum ghatti and gum tragacanth (mainly soluble mixture of complex galactose polymers; with other saccharide constituents including arabinose, rhamnose, fucose, glucose, mannose, xylose and uronic acids), glucosamine (plant-derived supplemental glucosamine HCl), Undaria pinnatifida fucoidans (sulfated fucose polymer that contains other monosaccharides, galactose and mannose, derived from edible brown seaweed) and rice starch.

Study design

An open-label dosing study was set up. The study was carried out in two phases.

Phase 1

Pilot study was carried out to establish safety and dose necessary to impact serum N-glycosylation profiles. Six individuals (age range 28–55; 3 females), ingested increasing doses of AA supplement (increased by 1.3 g/week) for 7 weeks (starting with 1.3 g/day for week 1, and reaching 9.1 g/day at week 7). From week 7 to week 11, the group was divided into two arms; three stopped their supplement while the other three volunteers continued with supplementation at increasing doses (reaching 14.3 g/day by week 11).

Phase 2

A larger study was performed to evaluate glycosylation changes pre- and during-supplementation. A total of 12 volunteers (age range 22–40; 6 females), were followed for 16 weeks. The first 7 weeks, designated as the pre-supplementation period (week −7 to 0), was to establish baseline serum N-glycosylation patterns and normal fluctuations. Following this period, the individuals ingested increasing doses of AA supplement (week 1–4 at 1.3 g/day; week 5–6 at 2.6 g/day and week 7–8 at 5.2 g/day). This was followed by a final week with no supplement intake (week 9).

Sample collection

Serum was extracted from blood samples and stored at −40 °C. Samples were tested in a blind manner. The codes were not revealed until the study was completed.

Glycosylation analysis

Serum (40 μl) glycoproteins were deglycosylated using a Peptide N-Glycosidase-F (New England BioLabs Inc., Hitchin, Herts, UK) according to the manufacturer's instructions. Released N-glycans were isolated by C18 (Supelco, Poole, Dorset, UK), and graphitized carbon columns, freeze-dried, reconstituted in 100 μl of deionised water, ready for Glycan profiling by Matrix assisted laser desorption ionisation time-of-flight mass spectrometry (MS; Kratos Analytical, Ltd., Manchester, UK), set in positive linear mode and calibrated with a dextran ladder. Dihydroxy-benzoic acid was used as the matrix.

Statistical analysis

Where necessary the data were log transformed. To allow for time trends analysis of covariance with time as a continuous predictor and subject as a categorical factor was used. For each glycan structure, two regression models were fitted. The first ignored AA and fitted a straight-line relationship with time. The second allowed for different regression lines in the pre-supplement and supplement periods. These two models were compared using an F test. The difference between the sums of squares was calculated. The ratio of this to the residual mean square for the larger, two period model=F ratio; used to test effect of AA.

Correction for multiple comparisons

The Bonferroni method was applied and significance was set at P0.006.


Only the data following the start of AA were used to analyse regression of each glycan on dose and subject as a categorical factor. Time was included as a second covariate (to allow for possible time trends); such that if there was a significant dose trend even after adjustment for time, the effect of dose was fairly secure.


Safety and compliance

Phase 1

All six subjects completed the trial. AA was well tolerated even at high doses, with no adverse reactions.

Phase 2

All six male and five female volunteers completed the trial. However, because of palatability issues at high doses, four of five female volunteers stopped supplementation earlier than planned; two at week 6 and two at week 7. One female volunteer withdrew at week 1 (because of palatability issues), and had data for pre-AA phase only, and was therefore omitted from the analysis.

Serum N-glycan analysis

Serum N-glycan profiles for each individual at weekly intervals were examined. Eight main biantennary serum glycoprotein-derived N-glycan structures corresponding to different m/z values were identified (Figure 1). Peak intensities were analyzed.

Figure 1
figure 1

Schematic diagram of the serum protein-derived biantennary N-linked glycan structures that were analyzed by MS. Eight main biantennary serum glycoprotein-derived N-glycan structures were identified. These comprised six neutral and two acidic (sialylated) glycans. The glycans are designated as G0, G1 and G2, according to the number of terminal galactose residues, and monosialylated and or A2, according to the number of terminal sialic acid residues that decorate the core pentasaccharide (GlcNAc2, Man3). Three of the glycans are F; giving rise to G0F, G1F and G2F structures. Monosialylation and monogalactosylation may occur on either the α1–3 or α 1–6 arm of the biantennary structures.

Phase 1

Analysis of the data from the pilot study demonstrated changes in the serum N-glycan profile in response to increasing doses of AA supplementation. Example of the observed changes for each of the six volunteers, at four time points, for four of the glycans, agalactosylated (G0), monogalactosylated (G1), G1fucosylated (G1F) and disialylated (A2), are shown (Figure 2). The time points chosen were time=0 (representing baseline values), time=1 (demonstrating that changes in serum protein glycosylation were evident within 1 week of supplementation at 1.3 g/day;), time=7 (9.1 g/day; time point at which the group were subdivided into two arms; three individuals continued with their oral supplementation at increasing doses, whereas the remaining three discontinued their supplementation) and time=11 (the end of the study demonstrating possible differences between the serum glycosylation profile of the individuals who continued compared with those that discontinued with their AA supplementation).

Figure 2
figure 2

Results from the pilot study; examining possible changes in serum protein N-glycosylation in response to different doses of AA. The MS percentage intensity for G0, G1, G1F and A2, for each of the six volunteers, at four time points is shown. Week 0 (baseline), week1 (AA at 1.3 g/day), week 7 (AA at 9.1 g/day) and week 11 (n=3 continued to take increasing doses of AA, reaching a final dose of 14.3 g/day, represented by the solid line; and n=3 discontinued their AA supplementation, represented by dashed line, respectively).

Week 0–7: all apart from the monosialylated glycans (not shown), showed some degree of change in response to AA supplementation.

Week 7–11: comparisons of the glycan profiles in the three individuals that discontinued with those that continued AA supplementation showed differences. These differences were confined mainly to the neutral glycans (G0F, G1, G1F, digalactosylated (G2) and G2F) and were particularly noticeable for G1F, the levels of which reverted (increased) towards baseline (week=0) when AA was discontinued (Figure 2).

Phase 2

The 7-week pre-supplementation monitoring of serum protein glycan profiles allowed us to evaluate and monitor normal glycosylation fluctuations and to establish statistically meaningful baseline levels. Comparison of the glycan profiles during the supplementation period with the time trends observed during the pre-supplementation period detected distinct changes in the relative intensities of seven of eight glycans studied. These changes were particularly striking within the first 1–2 weeks of supplementation (Figure 3). G0 was the exception, showing major fluctuations in the pre- as well as supplementation period.

Figure 3
figure 3

Serum protein N-glycan profiles before and after oral AA supplementation. Plots of MS observations for all subjects for whom there were measurements made before and during AA supplementation, demonstrating the changes in the eight serum protein-derived biantennary N-glycans in response to oral AA supplementation (zero time is the last pre-AA measurement). The changes were significant (F test for the effect of AA; P<0.001) for all the serum protein N-glycans except G0.

Analysis of covariance using the two regression model was used to assess the strength of evidence that serum glycosylation changes followed AA supplementation. A graphical example of the model as applied to the analysis of the data for G2 glycans is given (Figure 4). Using this analysis, AA supplementation was found to result in significant changes in the levels of all the serum protein-derived N-linked glycans (G0F, G1, G1F, G2, G2F, monosialylated and A2; P<0.001) except G0 (P=0.2).

Figure 4
figure 4

Example of the two regression model used for the analysis of covariance showing the change in serum protein G2 glycan levels following AA supplementation (P<0.001). For each glycan structure, two regression models were fitted. The first ignored AA and fitted a straight-line (solid blue line) relationship with time. The second allowed for different regression lines in the pre-supplement and supplement phase (solid black lines; coefficient −0.501 and 0.018, respectively). This was achieved by creating two time variables; the first is time up to AA, but is equal to zero afterwards. The second time variable is time after AA, but is equal to zero before AA.

Dose effect

Regression analysis (with time as a second covariate) following the start of AA was used to assess the strength of evidence that change did occur following supplementation with AA and to examine the effect of dose. The results (Table 2) show dose effects for G1 and G2; the levels of which decreased significantly with increasing AA dose (coefficient factor by which mean glycan levels are multiplied/week −0.130 and −0.690; P=0.02 and 0.005, respectively), and for A2; the levels of which increased significantly with increasing AA dose (Coefficient factor =1.083; P=0.005).

Table 2 Effect of increasing dose of AA supplement on serum protein N-glycosylation


Our study reports novel findings showing that the ingestion of a standardized mixture of plant-derived polysaccharides can induce significant changes in the N-glycosylation status of serum glycoproteins in normal healthy individuals. AA supplementation caused an overall, dose dependent, shift towards increased sialylation, resulting in significantly lower levels of neutral glycans (G1 and G2: −0.130 and −0.690 per week, respectively), and increased levels of fully processed, sialylated, glycans (A2: by a factor of 1.083 per week).

Given that glycosylation is a key post-translational modification that can significantly affect the overall biophysical and biochemical functions of proteins (Alavi and Axford, 2008; Biol-N′garagba and Louisot, 2003), the observed shift in the N-glycosylation status of serum glycoproteins in response to AA supplementation may be significant and could reflect changes in one or more physiological parameters.

Interpretation of the biological significance of the observed shift towards more silaylated serum glycoproteins is complex (Alavi and Axford, 2008) and requires further more detailed structure-function examination of given serum glycoproteins (for example, IgG). However, what is known is that increased levels of sialic acid can affect absorption, serum half-life and clearance from the serum, as well as the physical, chemical and immunogenic properties of various key serum glycoproteins (Alavi and Axford, 2008; Bork et al., 2009; Schauer, 2009).

Indeed, sialylation changes can have a key role in many aspects of the immune response, in particular in relation to the regulation of inflammatory processes, as in the case of IgG, where sialylation has been shown to function as a ‘switch’. The presence of sialylation is associated with the steady state, normal, anti-inflammatory function of IgG, whereas its absence is associated with the pro-inflammatory effector function (Alavi and Axford, 2008; Anthony et al., 2008).

However, with this in mind, it is important to note that the elucidation of a specific physiological role for a given glycan modification, such as sialylation, poses formidable challenge. The presence or absence of a particular sugar residue, at a particular glycosylation site on a given protein, may have different roles in different biological settings. In the case of sialic acid, this is further complicated by the fact that sialylation is linkage-specific, and that the presence of the different forms of sialic acid may give rise to very different biological scenarios (Alavi and Axford, 2008; Schauer, 2009). Further research is therefore needed to more clearly elucidate specific effects and their consequent biological significance.

Although the question of whether these serum glycosylation changes are directly linked to the constituents of AA per se, or the affect of these on gut microflora (Sinnott et al., 2007) requires further investigation, the novel finding that small amounts of a polydisperse mixture of dietary polysaccharides can alter the glycosylation of serum proteins is unexpected and intriguing. This is especially true as it provides a potential mechanism through which dietary polysaccharides may alter the effector functions of circulating molecules. This notion is particularly relevant when one considers the fact that, time and again, cross-sectional (Bo et al., 2006) and longitudinal (Estruch et al., 2009) studies indicate an association between dietary-fiber intake and levels of inflammatory serum markers (for example, an inverse correlation with interleukin-6 and tumor necrosis factor-α receptor-2) (Ma et al., 2008). Continued efficacy studies using randomized placebo-controlled trials examining AA and its specific components are warranted to evaluate possible effect(s).

In conclusion, we have demonstrated that supplementation with dietary plant polysaccharides in AA can result in significant serum protein glycomodifications in normal healthy individuals. These finding are interesting and merit further research, where it is hoped that an integrative approach using glycomics together with other—omics such as metabolomics and metagenomics, focusing on the gut associated microbe–host mutualism and its impact on physiology, can help unravel the exogenous and endogenous effects of dietary components and hence, generate novel hypotheses for innovative dietary interventions that may impact health.