wheat germ agglutinin, WGA). Indeed, phytic acid is known to infer the availability of minerals, but also to display antioxidant and anticancer properties and possibly other health effects (Kumar et al. 2010). WGA is generally considered to have adverse health effects principally due to its binding and damaging of the gut epithelial cell layer. However, a recent review on lectins points out the possible difference in effects according to the WGA amount and the lack of toxic evidence when consumed in transformed products after cooking, baking or extrusion (van Buul and Brouns 2014). Enzymes with adverse hydrolytic and oxidative activities, mainly lipases and lipoxygenases, are also unwanted activities responsible for storage damaging of germ and derived products (Sjovall et al. 2000). As other fractions, germ also contains undesirable compounds from grain contaminated with fungi. Recent studies analyzing the distribution of mycotoxins (Gimenez et al. 2013) reveal that wheat germ and derived oil contain deoxynivalenol (DON), but do not represent a significant source considering the tolerable daily intake. Analyses of mycotoxins in maize give low level of fumonisins and DON in refined oil and margarine even if present in the germ fraction. A significant amount of ZEA is detected, on the contrary, probably due to its lipophilic nature (Escobar et al. 2013). However, these authors estimate the dietary exposure to be less than 0.2% tolerable daily intake.
Differences in the distribution of some chemical compounds are observed between different cereal grains. For example, phytic acid is mainly found in the embryo in maize (87% of the total), whereas it is mainly located in the aleurone layer in wheat (Evers et al. 1999). Therefore, precise allocation of a specific compound in a cereal germ sample, as well as its quantification, must pay attention to the isolation method and the considered tissue as the embryonic axis and scutellum do not have identical composition (Pomeranz 1988) due to a distinct difference in physiological role in grain. Hand dissection of tissues and identification of biochemical markers specific for each tissue (Hemery et al. 2009) allow improvement in the characterization of cereal fraction.
Degerming prior to milling is more common and easier in maize than for other grains due to the larger size of the grain and germ and the importance of the germ oil extraction for further use. It is realized either using dry or wet processes. In the dry processing, the embryonic axis can be dislodged from the endosperm without grinding by mechanically shearing or impacting the grain. Mehra and Eckhoff (1997) have used a degerminator (Beall, Decatur, Illinois, U.S) to recover maize embryo, but other methods are also described to isolate a germ fraction (Gerschwiler et al. 2006; Hemery et al. 2007). A wheat germ by‐product, mainly corresponding to germ axis, is possibly recovered from a germ separator or traditional roller mills but with a low efficiency (15–20% of the total germ [Hemery et al. 2007]). Posner and Li (1991) have developed a two‐step tempering and processing procedure that allows the recovering of both the embryonic axis and the scutellum reaching 50–100% of the total germ. Isolation of a germ enriched fraction is also effective from hull‐less barley using a Fitzpatrick comminuting mill followed by sieving (Moreau and Hicks 2013).
In the wet‐milling process, a germ fraction is separated from a protein, starch or fibre‐enriched fraction. This method classically uses denaturing solutions for proteins as it generally proceeds in presence of sulphur dioxide and lactic acid at a low pH (Vignaux et al. 2006). However, Singh and Eckhoff (1996) have combined a dry and wet procedure to quicker recover the germ with an equivalent yield but avoiding the use of sulphur dioxide. Singh et al. (2001) compare different processing methods and have described significant differences in the corn germ oil yield and biochemical composition.
The efficiency of germ isolation is generally not evaluated due to difficulties to give a precise quantification of each tissue in recovered milling fractions. Different approaches have therefore been developed in order to monitor germ isolation during fractionation. Using transgenic maize lines expressing the green fluorescent protein (GFP) either in the endosperm or germ, Shepherd et al. (2008) compared the efficiency of different processing strategies through the analysis of this fluorescence marker. Specific biochemical markers such as tocopherols and a wheat germ agglutinin (WGA) also allow the monitoring of the germ composing oil (Morrison et al. 1982) or the embryonic axis (Hemery et al. 2009). Barnes and Jorgensen (1987) and Barnes and Tester (1987) developed spectral methods to quantify wheat germ in milling products based on the detection of the fluorescence emission of glycoflavones, which are the phenolic compounds known to be present at higher concentration in germ than in other tissues. Hsi‐Mei (1999) has compared the relative germ content in different millstream flour samples using this method.
Germ isolation aims first to reduce the lipid content of flour and thus avoid possible oxidation during storage. It is also used for oil recovery that is achieved by mechanical pressing (Al‐Obaidi et al. 2013) or solvent extraction (Piras et al. 2009) removing, respectively, 50% or 90% of the total grain lipids. Alternative approaches have been developed where supercritical carbon dioxide is used for extraction instead of toxic solvents (Zacchi et al. 2006; Piras et al. 2009; Ko et al. 2012; Ozcan et al. 2013). This method is found very efficient to extracts lipids as well as fat‐soluble compounds like tocopherols. Gelmez et al. (2009) defined the optimal pressure, temperature and time of CO2 extraction based not only on yield, but also on the efficiency of phenolic compounds and tocopherols as well as antioxidant activities retained throughout the extraction. More recently, Fang and Moreau (2014) have developed an aqueous enzymatic method to extract oil from wheat germ with high yield prior (72%) and after demulsification (around 60%).
Since germ lipids are mainly unsaturated, they are prone to oxidation and degradation under the conditions used for oil extraction and storage. Autoxidation of unsaturated lipids in germ or derived oil occurs rapidly, which leads to possible off‐flavors and formation of even toxic products. The most common classical method to prevent germ spoilage is dry heating or steaming (Srivastava et al. 2007), but other efficient methods have also been developed such as continuous microwave (Sjovall et al. 2000; Xu et al. 2013) or gamma irradiation (Jha et al. 2013). Krings et al. (2000) demonstrated the potential of roasting germ to delay the increase of peroxide value and conjugated diene hydro‐peroxide concentration as well as decomposition of tocopherol. Germ oil oxidation is potentially reduced by vacuum packaging, packing under an inert gas and refrigeration/freezing (Naz et al. 2005). Other methods like addition of antioxidant molecules from other plants, as those found in rosemary methanolic extracts, also prevent oxidation (Yesil‐Celiktas et al. 2009).
Extraction of oil releases a germ fraction enriched in protein (>30%) with well‐balanced amino acid content, notably including lysine. This remaining fraction is also rich in flavonoids, minerals (potassium, magnesium and calcium mainly) and phenolic acids (Zhu et al. 2006; Zhu et al. 2011). Therefore, defatted wheat germ can be used as an interesting nutritional food additive or as a raw matter for extraction of molecules with health functionalities.
As germ can be easily isolated, it has even been used as a bioreactor for recombinant protein production allowing an easy recovery of the produced heterogeneous protein (Zhang et al. 2009; Paraman et al. 2010). The only disadvantage is that the target recombinant protein is isolated with a number of proteins from germ and a high level of lipids especially in corn. Choice of the dry or wet fractionation process in this case has to pay attention to the recombinant protein sensitivity towards the generally harsh conditions used in the latter. Recently, it has also been shown possible to increase the germ size and therefore seed oil content by introducing wheat puroindolines into corn (Zhang et al. 2010).
3.5 Bran fractions – a source of micronutrients to exploit?
Bran fractions are considered by millers as by‐products and are therefore not standardized. They correspond to classical fractions obtained by debranning or milling from grains concomitantly with flours and semolina or hominies. They are mainly constituted of outer layers of grains ranging from the aleurone layer to the outer pericarp, without the eventual hulls, with a minor proportion of remaining starchy endosperm. Bran fractions differ greatly in tissue composition and thus biochemical composition and functional properties, depending on grains (oat vs. wheat), process (debranning vs. milling) and extraction rate used. Oat bran, but no other bran, is defined by AACC (http://www.aaccnet.org) based on its chemical composition and extraction rate, but not from the anatomical point