In order to keep abreast of current trends in gluten-free technology from the consumer perspective, the author recently purchased the gluten-free cookbook put out by the writers at
America’s Test Kitchen. The
company’s so-called The How Can It Be
Gluten Free Cookbook (2014) is something of a mixed bag; many of the
recipes are retooled versions of their particular recipes, such as their
five-banana banana bread, for particular gluten-free reformulations available
to the lay cook. Some of the recipes are nothing more than a bit of legerdemain
to fill space and were never contaminated by gluten to begin with, however, and
the book does lose some points for this bait-and-switch maneuver. For the
recipes that are present, however,
the cogent and usually decently accurate depictions of the chemistry
accompanying the processes and the storytelling of how each variation of a
recipe was tested are useful guides to the home cook.
This post is not intended as a review of the book, although I consider it a worthy enough text. Rather, I want to dedicate the present discussion to a curious observation made within its pages on the superiority of psyllium husk in yeast-dough baked goods, such as pizza crust and sandwich breads. The Editors at
Test kitchen attribute this to the fact that it “binds more effectively with
water… [as] a result, psyllium does a better job of strengthening the protein
network so it is capable of holding in lots of gas and steam during baking” (p
16). Later, the Editors write that “its chemical composition is similar to that
of xanthan gum, but it has a higher viscosity, so it is able to bind water even
more effectively… psyllium interacts strongly with the proteins in gluten-free
flours… providing a strong enough structure to support highly leavened bread
once the bread cools” (p 21). The Editors note that this advantage is only
observed in yeasted products; chemically-leavened doughs and batters were
generally drier and had an objectionable texture when psyllium was used (p 16).
These statements are, to my eye, somewhat conflicting. They raise several key questions that will be explored here:
1.) Is the mechanism as to why psyllium works in the way that it does strictly related to its viscosity?
2.) Is the chemical structure of psyllium truly comparable to that of xanthan gum?
3.) What is the reason for psyllium’s superior performance in yeasted products in particular?
4.) What is the reason for the comparatively poorer performance of psyllium fiber in chemically-leavened breads.
In answering these inquiries, perhaps the most sensible place to begin would be at the most basic level: that of the chemical structure. We will here consider three types of gums that see common use in gluten-free batters as a replacement for gluten: xanthan gum, guar and/or locust bean gum, and psyllium husk fiber. We must also consider the role of gluten in normal bread and just why so many gluten-free breads resemble nothing so much as a baker’s nightmare encounter with Ceres’ “uncanny valley”. It will suffice to say here that gluten is a long, viscoelastic polymer comprised of partially denatured and cross-linked large glutenin proteins and smaller, labile gliadin subunits. The addition of a hydrocolloid to “replace” gluten’s functionality is something of a fool’s errand, for the addition of a carbohydrate cannot alone replace the intrinsic elastic stretch-pull of true wheat gluten that can allow bread dough to take immense abuse and turn into a warm, glossy, living thing to the touch. Less poetically, a hydrocolloid alone cannot imitate this very critical aspect of the gluten network developed in true wheat dough, and thus the reason that hydrocolloids are added to gluten-free batters is to approximate the viscosity of wheat dough in early development. There are some additional issues with this approach due to the nature of the starch granules in non-wheat doughs and batters, but this problem will need to be treated in a future article in order to consider the problem of hydrocolloids in isolation.
Guar and locust bean gum are both β(1 à 4) D-mannopyranosyl units on main chain. They differ in relative amounts of substitution of α-D-galactopyranosyl units via the C-6-O. 0.05 – 0.25 wt% in foods is a common level of incorporation, depending on the exact empirical properties of the gum used. Both can also interact with xanthan to form a gel of 3D structure and show a synergistic gestalt effect on viscosity as a result. (2:4 xanthan-guar is one measured maximum) (Rocks 1971; Casas and García-Ochoa 1999).
Xanthan is a rather different beast. The hydrocolloid that we call xanthan gum is actually a fermentation product of Xanthomonas campestris, which is a native infection of cruciferous vegetables. A pyruvated trisaccharide side-chain consisting of mannose, glucose, and mannose, respectively is linked by definition to every other β(1 à 4)-linked D-glucopyranosyl unit of the cellulose chain. The viscosity increases with increasing degree of substitution of pyruvate. As with all non-starch hydrocolloids, increasing addition of xanthan to a model rice-based gluten-free dough resulted in increased water binding as found by thermogravimetric analysis, in turn resulting in greater viscosity of the batter (Crockett, Ie, and Vodovotz 2011). Thus, the Editors of
America’s Test Kitchen are correct
on this basic point. However, increasing substitution of xanthan (3 and 5%) in
the same model system depressed the final loaf volume and increased loaf
hardness over the minimum (2%). Thus, viscosity, and by extension water
binding, appear to be inadequate to explain improved gas retention and crumb quality
when considered alone.
Both of these gums share a characteristic of essential linearity, similar in some respects to the linearity of amylose. Psyllium fiber is also similar in this respect, but there are some key differences to consider. The “fiber” of psyllium husk is actually the extracted mucilage of Plantago ovata and consists of a mixture of gel-forming and non-gel-forming fractions (Fischer, Yu, Gray, Ralph, Anderson, and Marlett 2004). The gel-forming fraction was reported as consisting of about 57 – 58% of the original husk weight (Fischer, Yu, Gray, Ralph, Anderson, and Marlett 2004). The non-gel-forming fraction appears to consist of polyuronide compounds, rhamnose, and other, minor sugars, whereas the gel-forming fraction has most conclusively been shown to consist of a highly branched arabinoxylan in about a 3:1 ratio of xylose to arabinose moieties (Fischer, Yu, Gray, Ralph, Anderson, and Marlett 2004). The main chain of this gel-forming polysaccharide consists of β-(1 à 4) D-xylopyranosyl units. Contrasted with the regularity of xanthan gum’s substituted chain, single xylanopyranosyl units are linked through C-2, while trisaccharide L-arabinofuranose-α-(1 à 3)-D-xylanopyranose-β-(1 à 3)-L-arabinofuranose chains are linked through C-3 (Fischer, Yu, Gray, Ralph, Anderson, and Marlett 2004).
The question of whether this structure can be considered truly analogous to that of xanthan is tricky. The two bear similar features in the linear, β-(1 à 4) linked polysaccharide base chain and the trisaccharide substituents common to both. The identity of the component monomer sugars and the pyruvate-containing moiety of xanthan not seen in psyllium mucilage are also key differences between the two. On its own, xanthan is non-gelling, whereas psyllium mucilage is when sufficiently pure. Indeed, it is this latter characteristic that is implicated in its hypocholesterolaemic character. Thus, the chemical structure of xanthan and psyllium husk fiber, to say nothing of guar gum, are not sufficiently similar in this author’s estimation to support the assertion of the Editors at America’s Test Kitchen on this point.
What reason, then, may we suspect for the superior performance of psyllium fiber in yeast doughs? The answer may lie in that very qualifier: in speaking of yeasted doughs, we speak of a peculiar method of leavening a dough over time. Cappa (2013) investigated the effect of psyllium husk fiber along with sugar beet fiber in 5:1 and 1:1 combinations in gluten-free breads at 1.5% of the total weight. These authors found that the most “liquid-like” of their doughs, containing more water, that also contained psyllium rose most readily throughout the course of a 60 min proof time after yeast had been added. In contrast, the more solid doughs rose only modestly up to the first 20 – 30 min, and no further. Doughs containing the 1:1 mixture of psyllium and sugar beet fiber were notably poorer in terms of rise than their counterparts that contained the greater proportion of psyllium. These facts suggest two things:
- Firstly: That the presence of ample, even excess water is necessary for proper rising of psyllium-supplemented yeast doughs. This likewise implies that the purported water-binding ability of psyllium is as important as the Editors make it out to be.
- Secondly: That psyllium fiber exhibits an influence on yeast dough that encourages gas retention above that of other fibers.
The second suggestion is interesting, for it again concerns the chemical structure of the gel-forming psyllium hydrocolloid. The gel-forming ability suggests that a film is formed during hydration to encourage retention of the CO2 produced by yeast fermentation (Cappa, Lucisano, and Mariotti 2013). Indeed, a film of hydrocolloid was observed in unbaked gluten-free doughs containing psyllium when examined using confocal microscopy (Mariotti, Lucisano, Panani, and Ng 2009). The gel formed by psyllium mucilage is weak at normal pH used in baked food systems and melts about 40 °C, well above the room temperature or typical “warm place” conditions used for bread-proofing in the home kitchen (Farahnaky, Askari, Majzoobi and Mesbahi 2010). The translation of a weak gel formed upon hydration in a gluten-free dough into a functional film at the edges of gas bubbles generated during the production of CO2 by yeast is fairly easy to visualize. In contrast, the inability of xanthan to form gels without the presence of guar or a related hydrocolloid is a most likely reason for the poorer performance of xanthan compared to psyllium in yeast doughs (Renou, Petibon, Malhiac, and Grisel 2013).
The author was not able to find any papers directly comparing the performance of psyllium fiber with other gums in chemically-leavened quick breads, such as muffins or baking powder biscuits. Thus, we move into the less certain realm of speculation at this point. It is plausible that the melting temperature of the weak psyllium gel (40 °C) is low enough that the structure of the gel enabling gas retention during proofing of yeast doughs is destroyed during the early stages of baking, well below the gelatinization temperature of the potato, rice, and tapioca flours and starches that provide the commonest basis of gluten-free goods. Thus, if no gel is present to retain gases produced by the reaction of chemical leavening agents, gas retention is likely to be poor, and the texture and volume of baked products is likely to suffer. There is, to my thinking, no need to invoke an uncertain interaction of psyllium with the proteins of native flours to explain the phenomenon at play.
The next question that arises naturally from this line of thought, and with which we might test our hypothesis, is whether another sort of gel that remains intact above the melting temperature of the psyllium gel might exhibit better performance in chemically-leavened products. Felicitously, the xanthan-guar gel is just such a creature. Xanthan and guar interact most likely interact via interactions between highly-ordered helical segments of the xanthan chain with unsubstituted segments of the randomly-substituted guar polymannose main chain (Renou, Petibon, Malhiac, and Grisel 2013). The formation of hydrophilic “cross-linking” interactions between these domains accounts well for the increase in viscosity over either gum used alone at low temperatures, and indeed the gel exhibits some viscoelastic properties after heating, which may well account for improved gas retention in chemically-leavened goods (Renou, Petibon, Malhiac, and Grisel 2013). In the absence of emulsifiers and dough conditioners, a xanthan-guar mixture yielded a yeast-leavened rice-based loaf of superior volume and tenderness compared with either hydrocolloid alone (Demirkesen, Mert, Sumnu, and Sahin 2010). Surprisingly, however, few authors have investigated the matter in quick-bread systems; the closest case that the author could find was that of rice cakes made with baking powder utilizing different hydrocolloids and hydrocolloid blends in their formulations. The highest number and area of pores was indeed obtained with the xanthan-guar blend above that of guar or xanthan alone, thus supporting the notion that a viscoelastic gel interaction between xanthan and guar contributes to the superior texture and gas retention of chemically-leavened baked goods above those made with psylllium husk fiber (Turabi, E.; Sumnu, G.; Sahin, S. 2010).
Thus, it appears that we can answer the questions raised by the statements of the Editors as follows:
1.) No; differences in viscosity alone are not adequate to explain the observed differences in performance in gluten-free bread systems.
2.) Yes and no; the component monomers of psyllium and xanthan polysaccharides differ greatly, which lends a weak gelation ability to the former only, but a certain surface similarity is present in terms of the structure of the backbone and the trisaccharide substituents on the main chain.
3.) The reason for psyllium fiber’s superiority as a gluten replacement in yeasted products appears to have more to do with formation of a film to retain gases during the proofing step than contribution to viscosity or purported interaction with cereal proteins
4.) Speculatively, the early melting temperature of the psyllium fiber gel during baking is possibly the reason for the observed poor quality of chemically-leavened goods made with psyllium.
Cappa, C.; Lucisano, M.; Mariotti, M. Influence of Psyllium, sugar beet fibre and water on gluten-free dough properties and bread quality. Carbohydrate Polymers. 2013, 98, 1657 – 1666
Casas, J. A.; García-Ochoa, F. Viscosity of solutions of xanthan/locust bean gum mixtures. Journal of the Science of Food and Agriculture. 1999, 79, 25 – 31
Crockett, R.; Ie, P.; Vodovotz, Y. How Do Xanthan and Hydroxypropyl Methylcellulose Individually Affect the Physicochemical Properties in a Model Gluten-Free Dough. Journal of Food Science. 2011, 76, 6, 2011
I.; Mert, B.;
Sumnu, G.; Sahin, S. Rheological properties of gluten-free bread formulations. Jouranl of Food Engineering. 2010, 96, 295 – 303
Farahnaky, A.; Askari, H.; Majzoobi, M.; Mesbahi, Gh. The impact of concentration, temperature and pH on dynamic rheology of psyllium gels. Journal of Food Engineering. 2010, 100, 294 – 301
Fischer, M. H.; Yu, N.; Gray, G. R.; Ralph, J.; Anderson, L.; Marlett, J. A. The gel-forming polysaccharide of psyllium husk (Plantago ovata Forsk). Carbohydrate Research. 2004, 339, 2009 – 2017
Kool, M. M.; Schols, H. A.; Delahaije, R. J. B. M.; Sworn, G.; Wierenga, P. A.; Gruppen, H. The influence of the primary and secondary xanthan structure on the enzymatic hydrolysis of the xanthan backbone. Carbohydrate Polymers. 2013, 97, 368 – 375
Mariotti, M.; Lucisano, M.; Pagani, M. A.; Ng, P. K. W. The role of corn starch, amaranth flour, pea isolate, and Psyllium flour on the rheological properties and the ultrastructure of gluten-free doughs. Food Research International. 2009, 42, 8, 963 – 975
Quantitative analysis of macro and micro-structure of gluten-free rice cakes containing different types of gums baked in different ovens. Food Hydrocolloids. 2010, 24, 755 – 762
Renou, F.; Petibon, O.; Malhiac, C.; Grisel, M. Effect of xanthan structure on its interaction with locust bean gum: Toward predication of rheological properties. Food Hydrocolloids. 2013, 32, 331 – 340
Rocks, J. K. Xanthan gum. Food Technology. 1971, 25, 5, 22
Wang, F.; Wang, Y.-J.; Sun, Z. Conformational Role of Xanthan in its Interaction with Locust Bean Gum. Journal of Food Science. 2002, 67, 7, 2002