Detecting changes in the nutritional value and elemental composition of transgenic sorghum grain

Nuclear Instruments and Methods in Physics Research B 363 (2015) 183–187

Contents lists available at ScienceDirect

Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb

Detecting changes in the nutritional value and elemental composition of transgenic sorghum grain R. Ndimba a,c,⇑, A.W. Grootboom b, L. Mehlo b, N.L. Mkhonza b, J. Kossmann c, A.D. Barnabas a, C. Mtshali a, C. Pineda-Vargas a,d a

iThemba LABS, National Research Foundation, South Africa Council for Scientific and Industrial Research (CSIR) Biosciences, Pretoria, South Africa Institute for Plant Biotechnology, University of Stellenbosch, Matieland, South Africa d Faculty of Health and Wellness Sciences, CPUT, Bellville, South Africa b c

a r t i c l e

i n f o

Article history: Received 7 April 2015 Received in revised form 9 September 2015 Accepted 16 September 2015 Available online 30 September 2015 Keywords: Sorghum Transgenic Micro-PIXE Essential amino acids Minerals

a b s t r a c t We have previously demonstrated that poor digestibility in sorghum can be addressed by using RNA interference (RNAi) to suppress kafirin synthesis. The approach resulted in a twofold improvement in overall protein digestibility levels. In the present study, the effect of this targeted kafirin suppression on other grain quality parameters was investigated. Several significant changes in the proximate composition, amino acid profile and the bulk mineral content were detected. Importantly, the most limiting amino acid, lysine, was significantly increased in the transgenic grains by up to 39%; whilst mineral elements in the bulk, such as sulphur (S) and zinc (Zn) were reduced by up to 15.8% and 21% respectively. Elemental mapping of the grain tissue, using micro-PIXE, demonstrated a significant decrease in Zn (>75%), which was localised to the outer endosperm region, whilst TEM revealed important changes to the protein body morphology of the transgenic grains. Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction Sorghum (Sorghum bicolor (L.) Moench) is Africa’s contribution to the elite cereal crops of the world and, as such, it is ranked as the fifth most important human staple after wheat, rice, maize and potatoes [2]. Although sorghum is important for food security, it is generally viewed as nutritionally inferior to other major cereals, because of its dominant proteins, the kafirins, which are difficult to digest, and are furthermore deficient in the essential amino acids lysine, methionine and tryptophan [3]. To improve the nutritive value of this crop, a recent study utilised RNA interference (RNAi) technology to suppress kafirin synthesis in the public Sorghum line P898012 [1]. Several of the resultant transgenic lines demonstrated a significant increase in overall in vitro protein digestibility (of up to 53%) [1]. Although this was a welcome improvement to the nutritional value of sorghum, it was not clearly established if the genetic alteration had any unintended effects on other important grain quality characteristics. To address this concern the present study was initiated to evaluate ⇑ Corresponding author at: Materials Research Dept., NRF: iThemba LABS, P.O. Box 7129, Somerset West, South Africa. Tel.: +27 21 843 1165; fax: +27 21 843 3543. E-mail address: [email protected] (R. Ndimba). http://dx.doi.org/10.1016/j.nimb.2015.09.056 0168-583X/Ó 2015 Elsevier B.V. All rights reserved.

potential differences in the grain of two independent transgenic lines (featuring kafirin suppression) and their non-transgenic parental counterpart at the level of the proximates, the bulk mineral content and the total amino acid profile. Transmission electron microscopy (TEM) was also utilised to compare differences in the morphology of grain protein bodies, whilst micro-proton-induced X-ray emission (micro-PIXE) spectroscopy was used to resolve spatial differences in mineral concentrations within individual grains. The results of this study will serve to enhance our present understanding of the effects of kafirin suppression on sorghum grain quality, and will further highlight if there are any inadvertent changes in the transgenic grain material that may warrant closer investigation.

2. Materials and methods 2.1. Plant material Plants of two independent transgenic T5 sorghum lines, featuring the pABS044 construct, for the targeted suppression of select gamma- and alpha-kafirins (full details of the original transformation reported in [1]) and their non-transgenic parental counterpart, P898012, were grown under controlled conditions in a

184

R. Ndimba et al. / Nuclear Instruments and Methods in Physics Research B 363 (2015) 183–187

containment glasshouse located at the Biosciences Division of the Council for Scientific and Industrial Research (CSIR, South Africa). Plants were moved randomly at each watering to minimise positional effects, and before anthesis, individual panicles were bagged to prevent outcrossing. At full maturity, grains from the transgenic (designated herein as TG2 and TG3) and non-transgenic (designated herein as wild-type WT) plants were hand-harvested, cleaned and milled to fine flour; or kept as whole grains as required for the intended analysis. The two independent transgenic lines TG2 and TG3 were selected because of the results of Western Blot analyses which indicated that there was complete suppression of the targeted kafirins, namely, gamma-1 (25 kDa), gamma-2 (50 kDa) and alpha-kafirin A1 (25 kDa) [1]. 2.2. Compositional analyses 2.2.1. Proximate analysis The proximate analyses of the samples for moisture, crude protein, crude fat and total ash were carried out in triplicate according to standard protocols. In brief, the weight difference method was used to determine the moisture content after drying the samples at 100 ± 5 °C for 24 h; and the ash content after sample ignition at 500 °C [4]. Crude protein (N  6.25) was determined by the Dumas combustion method [5] and crude fat by means of ether Soxhlet extraction [4]. 2.2.2. Amino acid analysis The protein-bound amino acid content of the samples was analysed in triplicate according to [6], using reverse phase-high pressure liquid chromatography (RP-HPLC). For cysteine and methionine determination, a separate hydrolysis step involving performic acid oxidation was performed. The amino acid analysis was carried out at the accredited facility of the South African Grain Laboratory (SAGL) in Pretoria, South Africa. 2.2.3. Bulk mineral content Approximately 5 g of the ground sorghum samples was freezedried to a constant dry weight over a period of four days. A half gram of each sample was then digested with 10 ml of HNO3:HCl, 4:1, for destruction of organic matter using a microwave digester. The digested samples were then resuspended in 50 ml distilled water and thoroughly mixed before being analysed by inductively coupled plasma mass spectrometry (ICP MS) at the ICP Laboratory, Central Analytical Facility, Stellenbosch University. The instrument was calibrated using certified mixed standard reference materials from the National Institute of Standards and Technology (NIST). The results reported here were limited to the following main mineral elements: phosphorus (P), potassium (K), magnesium (Mg), sulphur (S), calcium (Ca), iron (Fe) and zinc (Zn). 2.3. Statistical analysis For the compositional data, at least three independent determinations were made for each parameter investigated, and the results expressed as the mean ± standard deviation. To evaluate differences between the means at the 5% significance level, one way analysis of variance (ANOVA) and the Tukey mean separation test was performed using Statistica for Windows Version 12.6 (Statsoft Inc., USA). 2.4. Transmission electron microscopy (TEM) To evaluate differences in the ultrastructure of the protein bodies in the transgenic and wild-type grain, TEM analysis was performed. In brief, small segments (1–2 mm3) of the peripheral endosperm were fixed in 3% (v/v) glutaraldehyde buffered with

0.1 M sodium cacodylate (pH 7.2) at room temperature for 24 h, followed by post-fixation in 2% (v/v) osmium tetraoxide at 4 °C for 24 h. The samples were then dehydrated in a graded ethanol series, before being infiltrated and polymerised in Agar Low Viscosity resin at 70 °C for 16 h. Ultrathin sections were prepared using a ultramicrotome fitted with a diamond knife, stained with 2% (w/v) uranyl acetate and Reynold’s lead citrate, and examined with a Tecnai G2 transmission electron microscope (situated at the Department of Physics, University of Cape Town). All of the TEM images reported depict protein bodies from the subaleurone layer of the grain endosperm. 2.5. Micro-PIXE analysis Dry mature sorghum grains were selected for micro-PIXE analysis, and due to their low moisture content, no elaborate fixation treatment was deemed necessary [7]. Selected grains were embedded in EpoFixTM (Struers) commercial resin and longitudinally sectioned through the median using a rotating diamond-coated blade, operated at a low speed (100 rpm), which cleanly cut the sample into half. Photomicrographs were then made of each halfgrain sample using a Nikon SMZ1500 stereomicroscope fitted with a digital camera. A minimum of three different half-grain samples for each genotype was then coated with a thin layer of carbon before mounting for micro-PIXE analysis. Micro-PIXE on the halfgrain samples was carried out using a proton beam of 3.0 MeV energy and a current of 100 pA, at the Materials Research Department, iThemba LABS. The proton beam was focused to a 3  3 lm2 spot and raster scanned over a sample area of 2 mm2, using square scan patterns and a data matrix of up to 128  128 pixels. Both micro-PIXE and proton backscattering (BS) spectra were collected simultaneously in event-by-event mode. Following data collection, quantitative elemental maps were generated, using the Dynamic Analysis method, as part of the GeoPIXE II software package [8]. Additionally, micro-PIXE spectra were extracted from a defined outer and inner region of the endosperm of each sample, to obtain average concentration values for particular mineral elements of interest. For data processing, each half-grain sample was treated as ‘infinitely thick’, and the main constituent of the biological matrix was assumed to be cellulose, following the similar approach of [9,10]. 3. Results and discussion The main nutritional components analysed in the transgenic and wild-type sorghum grain samples are shown in Table 1. Proximate analysis is an important tool for evaluating the quality of foodstuffs and is often used as the basis for establishing the overall nutritional value. The proximate components determined in this study included values for the moisture, crude protein, crude fat and the total ash content. The mean values of all measured parameters were found to be statistically equal (at the 5% level) for TG3 and WT grain. Grain from TG2 however, was found to be statistically different from the WT, in terms of its moisture, crude fat and total ash content. Higher levels of moisture (25% increase) and crude fat (16% increase) were recorded for TG2 grain; whilst total ash was significantly reduced (by 21%), in comparison to the WT. According to the consensus document of the Organization for Economic Co-operation and Development (OECD), the following mean range values for grain sorghum are noted: moisture 9.2–12.5%; crude fat 0.8–4.3%; and total ash 1.5–3.3% [11]. The mean values for moisture (11.16%), crude fat (2.75%) and total ash (1.61%) in the TG2 sample therefore fall within the bounds of normal variation for grain sorghum, and as such, these differences are not regarded as biologically significant. The increased moisture

185

R. Ndimba et al. / Nuclear Instruments and Methods in Physics Research B 363 (2015) 183–187

Table 1 Parameters investigated to compare the composition/nutritional profile of transgenic sorghum (TG2 and TG3), versus their wild-type (WT) non-transgenic parental counterpart. Means with same superscript within the row are not significantly different at p < 0.05. Parameter analysed Proximate analysis (g/100 g wholegrain flour)

Protein bound amino acids (mmol/100 g wholegrain flour)

Mineral content (mg/kg wholegrain flour) by ICP MS analysis

Moisture (fw) Crude protein (dw) Crude fat (dw) Ash (dw) Isoleucine* Leucine* Methionine* Cysteine Phenylalanine* Threonine* Tyrosine Valine* Lysine* Tryptophan* Aspartic acid Glutamic acid Serine Glycine Histidine* Arginine* Alanine Proline P K Mg S Fe Zn

WT

TG2

TG3

8.95 ± 0.08a 12.82 ± 0.69a 2.37 ± 0.04a 2.03 ± 0.03a 5.65 ± 0.29a 16.43 ± 1.13a 1.73 ± 0.13a 2.26 ± 0.50a 5.59 ± 0.15a 5.22 ± 0.36a 3.49 ± 0.17a 8.25 ± 0.41a 2.38 ± 0.24a 0.84 ± 0.06a 9.25 ± 0.19a 25.50 ± 0.71a 7.95 ± 0.51a 7.63 ± 1.00a 3.25 ± 0.09a 5.09 ± 0.35a 16.15 ± 0.68a 12.39 ± 0.36a 4426 ± 138a 4686 ± 272a 1899 ± 48a 1328 ± 52a 51.9 ± 1.9a 42.6 ± 1.9a

11.16 ± 0.20b 11.90 ± 0.95a 2.75 ± 0.13b 1.61 ± 0.04b 5.28 ± 0.20b 14.21 ± 1.49a 1.75 ± 0.15a 1.85 ± 0.37a 4.88 ± 0.27a 5.43 ± 0.17a 2.99 ± 0.37a 7.96 ± 0.25a 3.15 ± 0.19b 0.88 ± 0.03a 9.37 ± 0.27a 22.05 ± 1.54b 7.03 ± 0.69a 7.89 ± 0.78a 2.97 ± 0.26a 5.66 b ± 0.44 14.14 ± 1.23a 10.07 ± 0.99b 4215 ± 439a 4635 ± 101a 1882 ± 259a 1118 ± 50b 45.0 ± 5.1a 34.5 ± 4.5b

9.06 ± 0.15a 12.33 ± 0.56a 2.46 ± 0.06a 1.98 ± 0.22a 4.92 ± 0.18b 13.94 ± 1.50a 1.68 ± 0.30a 1.70 ± 0.20a 4.79 ± 0.18a 5.05 ± 0.20a 2.93 ± 0.17a 8.05 ± 0.84a 3.31 ± 0.39b 0.87 ± 0.04a 9.38 ± 0.76a 22.03 ± 1.01b 6.90 ± 0.60a 7.18 ± 0.63a 2.84 ± 0.25a 5.14 ± 0.42ab 14.30 ± 1.33a 9.75 ± 0.73b 4106 ± 104a 4544 ± 138a 1787 ± 78a 1248 ± 7a 49.9 ± 4.0a 33.6 ± 2.1b

fw = fresh weight; dw = dry weight. * Essential amino acids.

content for TG2 may however be a source of concern, as high moisture levels are known to facilitate insect infestation and microbial growth, which in turn contributes to grain spoilage and unnecessary post-harvest losses [12]. In terms of the amino acid profile, the concentration levels of 18 different amino acids were determined and for the purpose of a meaningful comparison were expressed in terms of the molar amount per 100 g of the wholegrain flour sample. According to a statistical analysis of the results, the transgenic grain samples consistently differed from the WT, mainly in terms of 5 amino acids, namely, glutamic acid, proline, arginine, isoleucine and lysine. In the case of glutamic acid and proline, a significant reduction (of up to 13.6% and 21.3% respectively) in the transgenic lines was evident, which is likely attributable to the targeted suppressed kafirins, which are well-known to feature high levels of these two specific amino acids [13]. Interestingly, in the case of arginine, which shares a common biosynthesis pathway with proline and glutamic acid [14], a significant increase of 11% was found for TG2, but the <1% increase for TG3 was not significantly different from the WT. It is therefore not clear if there is a consistent and tangible impact on arginine levels in the transgenic grain. However, in the case of the essential amino acids, isoleucine and lysine, the pattern appears more obvious. In comparison to WT, isoleucine was reduced by 6.5% in TG2, and 12.9% in TG3; whilst lysine was significantly increased in these lines, by 32% and 39%, respectively. Given that kafirins are virtually lysine-free, the enhanced lysine levels in the transgenic lines are likely attributable to an increased accumulation of lysine-rich non-kafirin grain proteins, such as the albumins, globulins and glutelins [15]. Although the increase in the lysine content of the transgenic lines is a welcome improvement to the nutritional value of the grain, there appears to be a concomitant decrease in another essential amino acid, isoleucine. This reduction is probably attributable to the targeted suppression of the alpha-kafirins, which are characterised by

relatively high levels of isoleucine [16]. Interestingly, it is also noted that lysine and isoleucine are linked via their biosynthesis pathways to the metabolic precursor pyruvate. Therefore it may be speculated that the increased demand for lysine production in the transgenic lines, may have caused a diversion of metabolic resources away from other pyruvate-derived amino acids, such as alanine, valine, leucine and isoleucine [17]. A survey of the average concentrations of these pyruvate-derived amino acids (Table 1), reveal that they are consistently lower in value for the transgenic lines in comparison to the WT, but a statistically significant decrease was only found in the case of isoleucine. However, according to the normal range of isoleucine levels found in sorghum, as specified by the OECD [11], the reduced isoleucine levels observed in the transgenic lines of the present study, do not fall outside the accepted known range of variation, and therefore this reduction would not qualify as a major cause for concern. In terms of the bulk mineral content, a general decrease in the concentration of minerals was apparent in TG2 and TG3 in comparison to the WT. However, according to the statistical test, only zinc was identified as significantly reduced in TG2 and TG3 (by 19% and 21% respectively); whilst sulphur was also highlighted as significantly reduced, but only in TG2 (by 15.8%). According to the OECD reference range (which is cited to be 16.9–47.1 mg/kg for zinc) the lowered zinc content of the transgenic lines (TG2 = 34.5 mg/kg; TG3 = 33.6 mg/kg) is not below the known level of variation found in sorghum. This is similarly the case for sulphur. Although the average bulk S content in TG2 (1118 mg/kg) is significantly reduced in comparison to the WT (1328 mg/kg), this reduced level is still within the OECD reference range of 900–1700 mg/kg [11]. In order to gain spatial comprehension of the observed mineral losses in the transgenic grains, as revealed by the bulk technique of ICP analysis, micro-PIXE was performed on independent half-grain samples of each genotype (n = 3), with a primary focus on the basal portion of the grain, which includes the endosperm region, and the

186

R. Ndimba et al. / Nuclear Instruments and Methods in Physics Research B 363 (2015) 183–187

Fig. 1. Randomly selected grains (n = 3), from the transgenic lines (TG2 and TG3) and the wild-type (WT) were halved as shown in (A) and prepared for micro-PIXE as described in the text. The proton beam was scanned over an area of approximately 2.3  2.3 mm2 as demarcated by the red box in (A) and elemental maps were produced using the Dynamic Analysis method in GeoPIXE. (B) shows a typical S distribution map of the scanned area of the grain, which highlighted two main areas of interest: an outer endosperm region featuring relatively high levels of S, and an inner endosperm region with lower concentration levels. These regions of interest were encircled, as depicted by the green demarcation in (C) and (D), and GeoPIXE region selection tools were utilised to obtain the mean concentration values and the minimum detection limit for S and Zn in these specific parts of the grain (see Table 2). Gross morphological features of the grain are pointed out in (A), P = pericarp; En = endosperm; and Em = embryo. It is noted that the outer endosperm region as presented in this study includes the aleurone layer. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 2 Average concentration of S and Zn (mg kg 1 dw) in the selected regions of the outer and inner endosperm of wild-type (WT) and transgenic sorghum (TG2 and TG3) grain. Results are presented as the mean ± SE of three replicates, with the minimum detection limit (mg kg 1 dw) indicated in brackets. Means with a different superscript within the column indicate a significant difference at p < 0.05. Genotype

Outer endosperm S

WT TG2 TG3

Inner endosperm Zn

a

3521 ± 555 (18) 2470a ± 321 (16) 4319a ± 907 (19)

S a

114 ± 12 (1) 27.3b ± 15 (1) 28.3b ± 7 (1)

Zn a

954 ± 152 (7) 577a ± 85 (9.7) 743a ± 51 (10)

3.3b ± 2 (0.6) 2.4b ± 0.4 (0.7) 4.2b ± 2.8 (0.8)

Fig. 2. TEM of protein bodies in the subaleurone layer of the wild-type (WT) and transgenic sorghum lines (TG2, TG3). Black arrow points to internal concentric ring structure of WT protein body. White arrows point out invaginations in the lobed protein bodies of the transgenic grains. Dense protein matrix of the transgenic grains is pointed out by white dotted arrows.

surrounding pericarp, but excludes the embryo (see Fig. 1A). Interest was directed to this portion of the grain because of the central importance of the endosperm for nutrient storage, and because of its relatively simple morphology, which facilitates an easy comparison of regions of interest, between different sample elemental maps. In view of sulphur’s role as a key indicator of the presence of protein [18], quantitative elemental maps depicting the distribution of S in the transgenic and WT were surveyed, which revealed two distinctive regions of interest: an outer endosperm and an inner endosperm region (as depicted in the example shown in Fig. 1B–D). Using GeoPIXE region selection analysis, elemental concentration data specific to the outer and inner endosperm regions for the WT and transgenic grains were extracted and evaluated. The results of this analysis, as it pertained to S and Zn, are shown in Table 2. Overall, the findings showed that the concentrations of both S and Zn, were not significantly different amongst the WT and transgenic samples, in the inner endosperm. However, in the outer endosperm, a significant difference was evident for Zn, but not for S. In the transgenic

lines, the average Zn concentration in the outer endosperm was 27.3 mg/kg for TG2 and 28.3 mg/kg for TG3, which represents >75% decrease, in comparison to WT value of 114 mg/kg. Given that kafirins are known to be highly concentrated in the outer endosperm [19], it seems that the genetic suppression of the targeted kafirins in TG2 and TG3, is linked to an unintended result of decreased overall Zn concentration levels, within this specific area of the grain. A possible explanation for this finding may be due to the fact that gamma-kafirins contain relatively high levels of cysteine [20], which is one particular amino acid moiety that has been shown to exhibit a strong affinity for Zn [21,22]. Interestingly, a reduction in cysteine content was noted for TG2 and TG3 (Table 1), however the result was not revealed as statistically significant at the 5% level. Given that the most prevalent forms of human malnutrition, involve low levels of dietary zinc [23], the finding that there are significant losses of this mineral element within the transgenic grains, reveals an important unintended difference, that may compromise the overall nutritional benefit derived from the other improvements in the protein quality of the grain.

R. Ndimba et al. / Nuclear Instruments and Methods in Physics Research B 363 (2015) 183–187

Because kafirins are stored in distinctive protein body structures that are particularly prevalent in the grain sub-aleurone layer [19], an interest in examining morphological changes to these structures was pursued using TEM. As shown in Fig. 2, contrasting morphological features were evident in both TG2 and TG3 in comparison to the WT. Protein bodies in the WT, were of a well-defined polygonal shape, ranging in size from 1–2 lm in diameter, with dark-staining inclusions that were at times organised in a clear concentric ring structure. In comparison, the protein bodies of TG2 and TG3, were highly amorphous, with deep invaginations that yielded a lobed overall protein body structure. No internal concentric ring structures or dark staining inclusions were found within the protein bodies of TG2 and TG3. However, the prominence of a dark protein matrix surrounding the protein bodies of the transgenic lines was more evident in comparison to the WT. The observed changes in protein body morphology are consistent with findings reported elsewhere for kafirin-suppressed sorghum and is linked to enhanced protein digestibility, due to an increased surface area that is more accessible and susceptible to proteolytic attack [1,15]. The noticeable increase in the dense protein matrix surrounding the lobed protein bodies, has also been reported, and is likely composed of the more lysine-rich sorghum grain proteins [15].

4. Conclusion In this study, a range of different analytical techniques were used to detect compositional changes in transgenic sorghum with targeted kafirin suppression, in comparison to the wild-type. Consistent differences of significance were found in the amino acid profile, grain Zn content and in the general morphology of the grain protein bodies. Although much of the deviation from the wild-type could be explained in terms of the intended kafirin suppression, or, was within the acceptable range of variation documented for grain sorghum, one important unintended result that was revealed from the micro-PIXE analysis was that a significant decrease in Zn concentration was localised to the outer endosperm. This study therefore highlights the need to carry out further investigations of the changes to the transgenic grain that may be localised only to the outer endosperm, and furthermore, may only be perhaps evident at the molecular level.

187

Acknowledgements Financial and technical support for this study was provided by NRF: iThemba LABS, CSIR Biosciences and the Institute for Plant Biotechnology at the University of Stellenbosch. Significant assistance with TEM imaging was provided by M. Jaffer at the University of Cape Town, South Africa. References [1] A.W. Grootboom, N.L. Mkhonza, Z. Mbambo, M.M. O’Kennedy, L.S. da Silva, J. Taylor, J.R.N. Taylor, R. Chikwamba, L. Mehlo, Plant Cell Rep. 33 (2014) 521–537. [2] J. Lindsay, INTSORMIL Scientific Publications; , 2010 (accessed 21 February 2015). [3] P.R. Shewry, J. Cereal Sci. 46 (2007) 239–250. [4] AOAC International, Official Methods of Analysis, 17th ed., Association of Official Analytical Chemists Inc, Arlington, Virginia, USA, 2002. [5] AACC International, Approved Methods of the AACC, tenth ed., American Association of Cereal Chemists Inc, St Paul, Minnesota, USA, 2000. [6] L. Mehlo, Z. Mbambo, S. Bado, J. Lin, S.M. Moagi, S. Buthelezi, S. Stoychev, R. Chikwamba, Mutat. Res. 749 (2013) 66–72. [7] K. Vogel-Mikuš, P. Pelicon, Vavpeticˇ, I. Kreft, M. Regvar, Nucl. Instr. Meth. B 267 (2009) 2884–2889. [8] C.G. Ryan, J. Imaging Syst. Technol. 11 (2000) 219–230. [9] E.T.F. Witkowski, I.M. Weiersbye-Witkowski, W.J. Przybylowicz, J. MesjaszPrzybylowicz, Nucl. Instr. Meth. B 130 (1997) 381–387. [10] W.J. Przybylowicz, C.A. Pineda, A.D. Barnabas, J. Mesjasz-Przybylowicz, Nucl. Instr. Meth. B 150 (1999) 282–290. [11] Organization for Economic Co-operation and Development, Consensus Document on compositional consideration for new varieties of grain sorghum. Series on the Safety of Novel Foods and Feeds No. 19, , 2010 (accessed 21 February 2015). [12] B. Dubale, A. Solomon, B. Geremew, G. Sethumadhava Rao, S. Waktole, Afr. J. Food Agric. Nutr. Dev. 14 (2) (2014) 8676–8694. [13] M. Lange, E. Vincze, H. Wieser, J.K. Schjoerring, P.B. Holm, J. Agric. Food Chem. 55 (2007) 6074–6081. [14] B.G. Forde, P.J. Lea, J. Exp. Bot. 58 (9) (2007) 2339–2358. [15] L. Da Silva, R. Jung, Z. Zhao, K. Glassman, J. Taylor, J.R.N. Taylor, J. Cereal Sci. 54 (2011) 160–167. [16] J.S. Wall, J.W. Paulis, Advances in Cereal Science and Technology, vol. II, American Association of Cereal Chemists, St Paul, 1977, pp. 132–219. [17] D. Schulze-Siebert, D. Heineke, H. Scharf, G. Schultz, Plant Physiol. 76 (1984) 465–471. [18] E.F. Garman, G.W. Grime, Prog. Biophys. Mol. Biol. 89 (2005) 173–205. [19] J.M. Shull, J.J. Watterson, A.W. Kirleis, Protoplasma 171 (1992) 64–74. [20] I.N. El Nour, A.D.B. Peruffo, A. Curioni, J. Cereal Sci. 28 (1998) 197–207. [21] B. Trzaskowski, L. Adamowicz, P.A. Deymier, J. Biol. Chem. 13 (2008) 133–137. [22] C. Mendoza, Int. J. Food Sci. Technol. 37 (2002) 759–767. [23] L. Lu, S. Tian, H. Liao, J. Zhang, X. Yang, J.M. Labavitch, W. Chen, PLoS ONE 8 (2) (2013) e57360.