**1. Introduction**

Mineral micronutrient sufficiency—a prerequisite for human well-being—can be ensured by diet diversification or consumption of mineral-dense produce [1]. In human diets, mineral micronutrients are predominantly acquired from plant-based sources, in particular staple grain [2,3]. However, most mineral micronutrients (manganese (Mn), iron (Fe), copper (Cu) and zinc (Zn)) in grain are tightly bound in phytate (myo-inositol hexakisphosphate), a phosphorus (P)-rich salt, which cannot be digested by mammals. This makes phytate-bound mineral elements poorly bioavailable and ineffectively exploited for normal body functions [4]. Furthermore, mineral density of the cereal grain has been for a long time regarded as of minor importance compared to the crop yield [5] resulting in prevalent micronutrient deficiencies in humans [6].

Efforts to increase bio-available concentrations of mineral elements in staple crops to remedy mineral micronutrient deficiencies—particularly in marginal populations—have been invested recently, and are referred to as biofortification [7,8]. Of the seven mineral elements often lacking in our diets, Fe deficiency is most widespread, affecting up to 60% of the global population [7]. However, increasing bio-available Fe concentration through the agronomic and genetic approaches in crops is challenging [9] for several reasons: (i) poor Fe availability in the soils limits uptake into plant roots [10], (ii) strict metabolic control over Fe accumulation and sequestration in plants tissues (sufficiency ranging between 50 and 150 mg Fe·kg−<sup>1</sup> dry weight in leaves of crop plants [11]), (iii) removal of Fe-rich layers during the processing of staple grain [12], and (iv) poor Fe bioavailability from phytate-rich produce such as cereal grain [13].

A large degree of variation in the accumulation of Fe in grain and seeds has been found in different crops, which is not a result from just the environmental factors. For example, in bread wheat (*Triticum aestivum* L.) grain, the variation in total Fe concentration, i.e., a ratio between the minimum and maximum total Fe concentration in grain, was up to 1.76 [14], in barley (*Hordeum vulgare* L.) the variation was up to 4.5 [15], in rice (*Oryza sativa* L.) up to 10.7 in flooded conditions and up to 288 in unflooded conditions [16], in pearl millet (*Pennisetum glaucum* (L.) R. Br.) up to 4.4 [17], in chickpea (*Cicer arietinum* L.) up to 3.2, and in pea (*Pisum sativum* L.) up to 3.5 [18]. Following predominantly classic breeding strategies, the existing natural variation in Fe density has been exploited for the development of biofortified varieties within the HarvestPlus programme [8], which demonstrated, for different crops and in different populations, that consumption of Fe-biofortified crops provides significantly more bioavailable Fe.

Despite its obvious importance for human nutrition, the filling of the staple grain with Fe, and understanding tissue-specific partitioning of Fe and Fe ligands remains a poorly understood subject [19]. Most grain filling processes take place through phloem tissues, which deliver Fe remobilised from the leaves. The presence of awns (bristle such as structures extending from lemmas), exhibiting photosynthetic activity accompanied by transpiration activity in wheat [20], may therefore play a role in the grain filling. This connection has not been investigated so far. It is, however, well-accepted that the level of phloem-mobility of a mineral element significantly affects its concentration and location in the grain, with elements such as calcium (Ca), exhibiting poor phloem mobility, not easily reaching the filial tissues of the grain and mostly remaining in the pericarp (maternal) tissues of the grain [12,21,22]. Iron has intermediate phloem mobility [23,24], so relatively large concentrations (exceeding those in leaves) of Fe have been found in some filial tissues of different staple grain, particularly the aleurone and embryo with values in the range from 200 to 400 mg Fe·kg−<sup>1</sup> and from 100 to 200 mg Fe·kg−1, respectively [12,15,25–28]. In these grain tissues, Fe was found to strongly co-localise with P [12,28–30]. Since approximately 80% of total P in the grain is in the form of phytate stored mainly in the aleurone cells [31], it has been inferred that the majority of Fe is bound to phytate in these tissues. Co-localisation analyses can, however, only predict potential ligands, not unambiguously determine the Fe binding environment, so conclusions must be drawn carefully when P is being used as a proxy for phytate. Using X-ray absorption near edge structure (XANES), which enables simultaneous analysis of Fe chemical form (speciation) and the type of complexing agents, it has been shown that around 80% of Fe in the whole grain of different wheat cultivars is bound to phytate, 15% to 24% as Fe2<sup>+</sup> (ferrous) and 57% to 85% as Fe3<sup>+</sup> (ferric) species [25]. Obtaining Fe K-edge XANES spectra of sufficiently high signal-to-noise ratio is a challenging task, with Fe concentrations typically found in grains and particularly in endosperm (<20 mg Fe·kg−<sup>1</sup> in wheat [12,32], barley [15], and Tartary buckwheat (*Fagopyrum tataricum* Gaertn.) [21]), thus such studies remain scarce. One way to circumvent these technical challenges is to combine reliable Fe distribution mapping, which identifies tissues or cell-types with the largest Fe concentrations, with micro-XANES analysis. One such study, conducted on cotyledons (containing on average 187 mg Fe·kg−<sup>1</sup> [21]) of Tartary buckwheat grain showed that 47% of Fe was bound to phytate, 22%, of that as Fe2<sup>+</sup> and 25% as Fe3+, while the remaining Fe3<sup>+</sup> was bound to citrate [33]. Furthermore, in wheat aleurone, modified aleurone (surrounding the crease) and in nucellar projection the micro-XANES analysis indicated that Fe was bound to phytate/citrate, phytate, and Fe-nicotianamine/Fe oxide-hydroxide, respectively [28]. However, tissue-specific Fe speciation was not resolved [28] and the XANES analysis in pericarp and embryo have not been acquired so far.

Therefore, the aim of the study was to compare mineral element composition of the awned and awnletted (those that have short or no awns) cultivars and to determine tissue-specific Fe speciation and Fe ligands in the contrasting cultivars to test the following hypotheses: (i) the presence of awns affects the mineral element composition of the wheat grain, (ii) majority of Fe is bound to phytate in di fferent tissues of wheat grain, (iii) Fe speciation and Fe ligands across di fferent wheat cultivars are stable, and (iv) Fe ligand profile depends on local Fe concentration.
