Stable isotope ( δ 13 C ) profiling of xylitol and sugar in South Africa

FUNDING: None Xylitol is an alternative sweetener to sucrose, glucose and fructose, and is available under a number of brands in South Africa. Carbon stable isotope values (δ13C) of a selection of commercially available xylitol products (n=28) were analysed and compared with sugar samples (n=29). Sugarcane (C4) and beet sugar (C3) derived sugar samples aligned with published values of source, although two samples that indicated a sugarcane origin suggested a beet sugar origin. Control corn-derived samples defined a stepwise xylose to xylitol discrimination of +0.7‰. The distinction between C3and C4-derived xylitol was less clear with three samples difficult to define (range = -14.8 to -17.1‰). The values for a suite of xylitol samples (-22.3‰ to -19.7‰; n=8) that aligned closely with a suspected C3-derived xylose, were ~8‰ more positive than known birch isotope values. Some xylitol samples may thus represent (1) a mixture of C3and C4-derived products, (2) derivation from a CAM species source or (3) different processing techniques in which the discrimination values of xylose from corn, and xylose from birch, may differ because of the respective chemical processing techniques. No samples that claimed a birch bark origin were within the range of samples suggested to be corn derived (i.e. -13.0‰ to -9.7‰, n=16). We suggest that the threshold values provided are relatively robust for defining the origins of xylitol and sugar, and can be used in determining the authenticity and claims of suppliers and producers.


Introduction
2][3] Added sugars in processed foods and beverages are linked to the development of diabetes, obesity and metabolic syndrome. 6,7][10] The intake of alternative sweeteners (artificial and non-nutritive) may reduce these health risks 9,11 and consequently the alternative sweetener market has increased in recent times 5,12 .
Xylitol (D-erythro-pentitol), a reduced-calorie sweetener, is a five-carbon sugar alcohol or polyol (C 5 H 12 O 5 ), also known as wood or birch sugar 10,13 , that occurs naturally in small quantities in a variety of plants 9,[14][15][16] , fruits and vegetables 4,[17][18][19] and is even produced in the human body 10 .It has been used as a sweetening agent in food since the 1960s 20,21 , in chewing gums, mints, sweets and toothpaste and also as a sugar substitute in confectionery and drinks 4,18 .Evidence suggests that it may prevent dental caries 15 , ear and upper respiratory tract infections 4,18 , the development of obesity 22 , and can be safely consumed by sufferers of diabetes type I and II, as it has the capacity to stabilise blood sugar levels 4,22 .In comparison with other alternative sweeteners, xylitol is similar in sweetness to sucrose with a lower calorie content (2.4 vs 4.0 cal/g). 16,22 an industrial scale, xylitol is currently produced by chemical reduction of xylose (D-xylose), traditionally derived from the bark of birch trees (Betula pendula) and other hardwoods. 4,23,24More recently, corncobs (Zea mays) 25 , sugarcane bagasse, and wheat, sorghum and rice straw [26][27][28] have been used as sources of xylose; in China, xylitol production from corncob reached 50 000 tons in 2008 25,29 .Production from corncob is favoured as it is less expensive than production from birch. 30Alternative methodologies to produce xylitol utilise yeasts (e.g.Candida sp.), bacteria and fungi. 4,17,31,32The annual demand for xylitol is over 100 000 tons worldwide, with a selling price of USD4-5/kg and an economic value of up to USD537 million per year. 4,25,33Asia produces 50% of the world's total xylitol, with the balance produced in Europe, the USA and Australia. 25The largest producer of xylitol in the world has production plants in Finland, the USA and China. 25,346][37][38] This has the overall negative effect of raising health concerns, reducing consumer confidence, and decreasing the sale of authentic products. 36Stable isotope analysis is one method to identify food adulteration, where the botanical origin, geographical origin, and specific farming regime or production system of a product can potentially be determined. 35,36,38Stable isotope analysis has been successfully utilised to detect the adulteration of honey with cheaper sugars 37,[39][40][41] and olive oil with pomace oil 42 , and CO 2 in apple cider (C 3 ) with a C 4 carbon isotope value may suggest the incorporation of C 4 sugars 43 .
5][46][47][48] The majority of plants utilise the Calvin cycle (C 3 ) and have tissues with a mean δ 13 13 C values ranging from -27‰ to -12‰. 49The 13 C/ 12 C isotope ratio can thus be used to identify a product's botanical origin. 38 South Africa, xylitol is available from local retailers, supermarkets and health stores, with a number of brands available to the public.Xylitol is one of the more recent alternatives to sugar (sucrose, fructose, glucose) and claims on the origins of products vary, from those indicating a birch bark or corn origin, to those for which the source is unknown or not disclosed.The aim of this study was to isotopically profile xylitol products available in South Africa, in an attempt to define their likely origin: from C 3 (e.g.birch bark) or C 4 (e.g.corn) plants.For comparison, we analysed sugar samples that were expected to reflect distinct origins from either sugarcane Saccharum spp.(C 4 ) or sugar beet Beta vulgaris (C 3 ).By doing so we attempt to present a critical assessment of these commercially available sweeteners that adds to consumer transparency and the integrity of food products currently available on the South African market.

Samples
Samples of xylitol products (n=28; including duplicates for two brands) available in South Africa were purchased from various retailers (January 2015 -November 2016), mostly in the Gauteng Province.Information was collected from each product regarding (1) country of origin and (2) claimed source (if disclosed).Of the 28 samples, 10 disclosed the source on the packaging.Anonymity of the selected brands is assured through assigned control numbers to each sample.In all instances we assumed products to be 100% xylitol, as indicated by the packaging.
To understand the relationship between xylose and xylitol we sourced two laboratory xylose samples, a sample of xylose from a Chinese supplier and derived xylitol (from the supplier of one of the brands for which we sourced repeat samples).To understand the likely source species we compared them to published isotope values of corn and birch -the two most common sources of xylitol.In addition, we measured the carbon isotope value of a birch bark and a birch leaf sample collected on the University of the Witwatersrand east campus (in Johannesburg) in December 2016.
We obtained additional sugar (fructose, glucose and sucrose) samples as controls (n=29) because South African table sugar (sucrose) is mainly derived from sugarcane (C 4 ) and European sugar from sugar beet (C 3 ).Some of these samples also included laboratory samples (n=8), of which three samples were indicated to be derived from Agave (a CAM photosynthesiser) and two samples were indicated to be derived from coconut, Cocos nucifera.
To assess the likely mixing of xylitol from different sources, in which the different sources may contain different size xylitol crystals, we selected three samples (X1, X3, X7) and separated them into >500 μm or <500 μm constituents (Labotec™, Star Screens, Booysens, Johannesburg, South Africa, sieve conforms to SABS197-ISO9002 specifications) for stable isotope analysis.A 28-33 g amount of each sample was separated into the different constituent size classes and prepared for isotope analysis.

Sample preparation for δ 13 C analyses
Duplicate subsamples (0.4-0.5 mg) of all xylitol, xylose and sugar samples were weighed into tin cups (pre-cleaned in toluene) for analysis.
For every four samples we weighed a gelatine (Merck) laboratory working standard (0.2, 0.4, 0.6 mg).Standards were of variable mass in order to determine if there was any sample size effect.The samples were combusted at 1020 °C on an EA1112 Elemental Analyser coupled to a DeltaV Plus stable light isotope mass spectrometer by a Conflo IV interface (all equipment supplied by Thermo Scientific, Bremen, Germany).These analyses were performed in the Mammal Research Institute stable light isotope laboratory at the University of Pretoria.Precision on the standard analyses was 0.11‰ and no sample size effect was noted.The stable isotopic values are expressed in delta (δ) notation in parts per thousand (per mille, ‰), relative to the international standard Vienna PeeDee belemnite (VPDB).
The sugar samples were clearly separated into two clusters: those C 4 derived (range: -13.1‰ to -10.0‰, n=18; cf.mean literature cane value of -12.7‰ 41,48 ), and those C 3 derived (range: -27.0‰ to -23.5‰, n=6; cf.mean sugar beet value of -25.6‰ 41,59 ).Three of the four C 3 commercially available sugar samples were sourced from Europe (one from the UK and two from Finland), i.e. they were of sugar beet origin.Two of the C 3 samples were indicated to be sugarcane derived (laboratory sucrose sample S24=-23.5‰ and fructose sample S15=-25.3‰).

Discussion
The hexose sugars sampled in this study suggest that stable isotope analysis is able to clearly define the origin of C However, two forms of evidence suggest that these samples are birch authentic; first, three of the samples (within 0.3‰ of each other) are from Finland and we suggest that we are less likely to observe any adulteration in the country in which this process was developed, and second, two xylose values (albeit laboratory samples) that are similar to these samples lend support to the fact that they may be pure C 3 derived (in the same way that we recognise a close isotopic link between C 4 xylose and the derived xylitol in the factory control).The different discrimination values between xylose (and xylitol) and the plants from which they are thought to be derived, i.e.C 3 or C 4 , may be a consequence of different chemical processes, and this remains to be investigated further.If not, then an entire suite of available C 3 xylitol products may be consistently adulterated with C 4 xylitol.
Only one sample (X14) was within 2.0‰ of the published values for birch.This sample warranted further investigation and for this and a number of other products, contact with the supplier was made to obtain further insight on their products.It was ascertained that xylitol in the product X5 was supplied by the packager of X15.Also, X15's supplier supplied the xylitol in X14, but further communication revealed that X14 had been withdrawn from the market because of suspected mistrust with X15.Because X14 was 'suspicious', one of us (C.S.) tasted it.The lack of a subtle cooling effect in the mouth (a characteristic of xylitol) 5 and an aftertaste reminiscent of other sweeteners (e.g.stevia) added basis to the idea that we were dealing with a likely adulterated food product.
Analysing the contents of all samples is beyond the scope of this study, but given this outcome, it is an investigation that should be undertaken.
Two brands are represented more than once (Figure 1), and in each case the within-product isotope values are different; which may be explained by a change in supplier.
To test the possibility that sample contents may comprise mixed origins we analysed samples separated by size (< or >500 μm; n=3 samples).For each (although the proportions of each size class were different between products), we found no difference in the isotope values of the two fractions.This analysis included one sample for which the source (C3 or C4) was not clear.This finding suggests, at least at this scale, that the sample was not derived from separate sources.
In 36% (n=10) of xylitol samples, the country of origin was not disclosed on the packaging, and information regarding the origin and source was sometimes vague, e.g.'fibrous parts of plant', or lacking.The South African Foodstuffs, Cosmetics and Disinfectants Act (Act no.54 of 1972), which includes regulations on the labelling and advertising of foods (Government Gazette No 37695, May 2014), prescribes numerous requirements that do not appear on a number of the products.These inconsistencies, in what appear to be standard labelling requirements, may result in consumers being mis-or uninformed regarding the products they are purchasing.Xylitol available in South Africa is not produced locally but, to the best of our knowledge, imported.In China, the waste biomass available after preparing corncobs (C 4 ) for the food industry provides a substrate for xylitol production 25,29 , yet negative perceptions regarding the source (and country) of certain products may persist for consumers.In the USA, hydrosylate from birch trees (C 3 ) is utilised as a substrate for xylitol production 25 and it is also potentially produced from other hardwood (C 3 ) species 26 .Even though xylitol produced from birch may appeal to a discerning market, although it may be more expensive, there is no other major difference between birch-sourced and cornsourced xylitol.Therefore, both information that is provided and a lack of information, can be misleading for the consumer.
Xylitol is traded all over the world.South Africa is simply one example of an importer vulnerable to adulteration and/or misinformation originating at either the source or locally; making this work of both national and international relevance.The carbon isotope data prima facie suggest that most xylitol samples represent origins suggesting either a C 3 or a C 4 source.For four samples, we question a single C 3 or C 4 source (intermediate isotope values; X8, X15, X5), or authenticity as xylitol Because marketers are not obliged to disclose the origins or methods of processing, the intention is thus not to challenge the authenticity of those marketing these products, but rather to challenge the information (or lack thereof) in the marketing process.Stable isotope analyses may be a suitable method to distinguish the origin of commercially available xylitol, especially when this is not clearly stated on the packaging or label.This study thus provides information that may be used to profile C 3 -or C 4 -derived sugar and xylitol products, and to lay a foundation for further investigations regarding these products in the food market.However, a priori information may be required regarding chemical processes that present source-product variation in isotope signatures, before conclusive origins (i.e.C 3 or C 4 ) can be defined, and before any isotope interpretation can be applied in a forensic assessment.

Table 1 :
24 one of the manufacturing processes, xylose in hemicellulosic hydrolysate fractions is converted to xylitol by a chemical catalytic reaction,24but other processes may be used and it is unclear if different chemical processes are the reason for the range of values observed.If the processing method is discounted, then the variation in isotope signatures in commercially available xylitol in South Africa may indicate many different sources.δ 13 C values of different xylitol granule size components (>500 μm and <500 μm), with proportions (%) of each size category indicated 13and C 4 -derived sugars, with δ13C values ranging from 27.0‰ to -23.5‰ and -13.1‰ to -10.0‰, respectively.These values in turn align, respectively, with the published δ13C values for sugarcane and beet sugar.Sugars derived from Agave (a CAM photosynthesiser) yielded expected intermediate isotope values, but this finding does not affect the analysis, as CAM plants are not routinely used for xylitol production.The xylitol isotope values are less clear.Because the chemical processes involved in the production of xylitol may vary, and are more complex than those for sugar production, we cannot be sure that stable isotopes unambiguously link source and product.4plant).In support of this conclusion is a mean corn value of -12.4‰ reported in the literature.If this δ13C discrimination from xylose to xylitol is minimal for C 4 -derived xylitol, it is difficult to align the group of eight xylitol samples (-22.3‰ to -19.7‰) to a C 3 origin if the mean published birch (-28.5‰) and the local birch (mean=-26.3‰)values are significantly more negative than the xylitol samples claimed to be derived from birch.http://www.sajs.co.za Volume 113 | Number 5/6 May/June 2017 41,5955][56][57][58]4‰,includesseeds and cob values48,[50][51][52][53], mean birch Betula pendula δ 13 C=-28.5‰,includesleavesandwood[54][55][56][57][58]and(2)sugar;meansugarcaneSacharrum spp.δ 13 C=-12.7‰41,48,meansugarbeetBetavulgaris δ 13 C=-25.6‰41,59.The δ 13 C values of 29 xylitol, 3 xylose and 29 sugar samples.See Supplementary tables 1 and 2 for isotope values of samples presented in the figure, and values and references for calculated mean endpoint values for xylitol and sugar.However,other reasons may explain these findings, including that: (1) the samples represent both C 3 and C 4 origins (mixture of different products; although this was not apparent in the three samples for which we measured the isotope value of different size granules), (2) different processing techniques result in fractionation processes that result in isotope values of the product not aligned with the source (we suggest that this is unlikely because the wide range of isotope values obtained would then suggest many processing techniques), (3) the isotope values of the samples do align with the source, but it is rather the source that (for whatever reason) does not have a clearly distinguishable C 3 or C 4 signature, (4) these intermediate products represent xylitol with a CAM origin (although in no sample was this disclosed, and we are not aware of any xylitol derived from a CAM species), or (5) adulteration with non-xylitol additives (e.g.flow agents, other artificial sweeteners like stevia or erythritol) affected the final δ 13 C of the marketed product.
*indicates laboratory samples.For xylitol samples: Xy indicates xylose samples (grey markers), similar numbers indicate same brand name.For sugar samples: sucrose unless indicated; F, fructose; G, glucose.Grey sugar markers indicate Agave (CAM photosynthesiser) derived sugars; similar numbers indicate same brand name; open sugar markers indicate sugar samples derived from coconut blossom.Figure 1: http://www.sajs.co.za