Introduction

Globally, 2 billion people suffer from severe micronutrient deficiencies due to the non-availability of nutritious food (Mayer et al. 2008; White and Broadly 2009). Micronutrients constitute minerals and vitamins that are derived from the diet, as they could not be synthesized in the body (Bouis and Welch 2010). The daily requirement of each micronutrient varies according to the stage of development. An average healthy adult requires 15 mg each of iron (Fe) and zinc (Zn) and 600 µg of vitamin A; however, this requirement is not met through the food consumed by the majority of the population. The present-day food constitutes cereals in significant proportion, and these cereals have high calorific value. Thus, the normal staple diets supply only 2–3 mg of Fe, 7–8 mg of Zn, and traces of vitamin A, which is not sufficient to sustain the human body (https://www.who.int/vmnis/database/en/). These values are projected to decrease due to the poor availability of micronutrients in the frontline crops like rice and wheat. Furthermore, while micronutrients play a prime role in regulating the metabolic activities in the cells and tissues, their deficiency results in irreparable consequences. Approximately 43% of the children below the age of five are prone to anemia due to Fe deficiency (Stevens et al. 2013). In the case of zinc, 17.3% of the population faces inadequacy in dietary zinc (Wessells and Brown 2012). Similarly, 190 million children at their pre-school age face vitamin A deficiency (Imdad et al. 2017; Visser et al. 2017). Previously, White and Broadly (2005) showed that 340 million children suffer from malnutrition, leading to stunting, wasting, or overweight, which results in disorders like xeropthalmia, skin disorders, and cancers. This has necessitated the biofortification of major cereals through breeding and transgene-based approaches (Garg et al. 2018). Further, studies on the mechanism of micronutrient uptake, the effect of soil fertility, processing techniques, targeted genes for nutrient mobilization, and anti-nutritional factors that hinder the bioavailability of micronutrients were performed (Welch and Graham 2004; Palmgren et al. 2008; Cakmak 2009; de Valenca et al. 2017; Das et al. 2019; Hossain et al. 2019).

Advancements in genetic dissection of micronutrient biosynthesis and accumulation in major cereals had pinpointed the genes, alleles, and QTLs underlying these complex traits. This information was further used in molecular breeding for improving the micronutrient content of the grains. Recently, recombinant DNA and genome editing approaches were also being deployed in major cereals for biofortification (Majumdar et al. 2018). These studies also widened the understanding of the research gaps and limiting factors in attaining the target mineral bioavailability in cereals. Though significant progress has been made in major graminaceous species (like rice, wheat, maize, and sorghum), a class of minor cereals, including small millets, remain underutilized and neglected (Muthamilarasan and Prasad 2021). Small millets constitute eleven millet species that are rich in minerals, vitamins, essential amino acids, and antioxidants (Saleh et al. 2013; Muthamilarasan et al. 2016; Vetriventhan et al. 2020). The micronutrient content of millet grains in comparison with major cereals is provided in Table 1. Despite their nutritional superiority over the major cereals, deciphering the genetic determinants of micronutrient contents in minor millets and exploiting them for the genetic improvement in cultivated varieties with enriched minerals and vitamins remain elusive. However, studies on major cereals provide a roadmap for identifying the genes, alleles, and QTLs underlying micronutrient traits in other graminaceous species. Further deployment of omics tools will enable the manipulation of target genes for improving the nutritional content of minor species per se. Also, it will facilitate the transfer of candidate genes to other major cereals through transgene-based approaches. Thus, identifying the genetic determinants holds the key for such biofortification programs, and given their importance, the present review enumerates the knowledge generated so far in understanding the genetic determinants of micronutrient traits in graminaceous crops using different approaches. It also provides the roadmap for further studies to enhance the nutritional potential of small millets and other graminaceous crops to provide a long-term and sustainable solution to micronutrient deficiency prevalent worldwide.

Table 1 Micro- and macronutrient content of millets and cereals
Full size table

Current understanding of the mechanism of micronutrient uptake and their manipulation to enhance mineral absorption

The solubilization and mobilization of minerals from the soil to the grains are intricate processes involving different biochemical and molecular components. As the mineral uptake is proportional to the concentration or richness in the soil, agronomic biofortification provided limited success in improving the nutritional status of major cereals (Garg et al. 2018). However, an exogenous supply of minerals resulted in the accumulation of unused elements in the soil, which could be toxic to plant growth at higher concentrations. On the other hand, growing the crops in micronutrient deficient soils results in decreased accumulation, affecting the grain nutritional content (Shukla et al. 2016). For example, the uptake of zinc has been severely affected in crops grown in calcareous soils (Gupta et al.2016). Thus, gaining mechanistic insights into the nutrient uptake will provide insights into understanding the nutritional quality of these crops. Nutrient absorption from the soil could be facilitated by the secretion of redox enzymes, chelating compounds, and the association of microbes in the soil. Graminaceous crops absorb the micronutrients by chelation, while non-graminaceous species imbibe the micronutrients by reduction reactions occurring in the rhizosphere. Cereals eventually capture these micronutrients from their rhizosphere, where the nutrients enter the roots by acidification of their plasma membrane (Morrissey and Guerinot 2009). Due to this humification and change in the Eh (redox potential) of soils, the tightly held cations in the soil surface are released toward the roots by a cation exchange reaction (Gaxiola et al. 2007). To stimulate this absorption, plants also release organic acids and phytosiderophores to facilitate the uptake of essential nutrients like Fe and Zn. One such phytosiderophores is mugienic acid that has a higher affinity for absorbing Fe and Zn by chelation. Cereals like rice, wheat, and maize release these compounds into their rhizosphere using a transporter-like TOM 1 (Ishimaru et al. 2006). Rice secretes 2-deoxymugineic acid (2-DMA) while barley releases 3-epihydroxymugineic acid. These compounds effectively form complexes with the available Fe and Zn from the rhizosphere and effects absorption by chelation (Schaaf et al. 2004).

After absorption from the soil, further studies on the ascent of micronutrients under normal and controlled conditions in rice described the role of roots for a continuous uptake under control conditions. Typically, the roots absorb the required nutrients for the growth and development up to the plant establishment phase. After the ripening stage, the sinks depend entirely on their leaves for nutrient translocation (Cakmak and Kutman 2018). Hence, the roots initially transport the nutrients from the soil toward the stem and leaves which are stored as a reserve for distribution to other parts. Thus, identifying the genes involved in this uptake mechanism and exploiting them for enriching the cereal grains using transgenic-based approaches have derived importance (Fig. 1). Biofortification of the mainstream cereals by manipulating such genes involved in the uptake process has increased the Fe uptake up to 3.7-fold in rice (Goto et al. 1999). Improvising the Fe uptake in rice by overexpressing the chelating gene NAS1 (nicotianamine synthase), iron influx gene YSL2 (yellow stripe-like2), iron uptake gene IDS3 (iron-deficiency-specific clone 3), and manipulation of Fe uptake regulating translocators [vacuolar iron transporters VIT1 and VIT2, and hemerythrin motif-containing really interesting new gene (RING)-and zinc-finger protein HRZ1] has significantly increased the Fe concentration in the grains (Suzuki et al. 2008; Lee et al. 2012; Ishimaru et al. 2010). The uptake of Fe by reducing Fe3+ to Fe2+ to improve the solubility enhances the Fe uptake in plants (Kim and Guerinot 2007). A major proportion of the Fe that gets absorbed is accumulated as ferritin in the grains' aleurone layer, which is facilitated by the ferroxidases. These ferroxidases can conserve up to 4500 atoms of iron in their complexes. Manipulating the ferritin genes has proved to be successful in improving grain Fe content in cereals. For example, manipulating soybean ferritin genes with rice globulin promoters in the rice cultivar, Swarna, showed a 3.7-fold increase in Fe content (Vasconcelos et al. 2003; Paul et al. 2012, 2016). This was considered a significant leap in achieving Fe-rich rice to circumvent anemia and related issues in the human population.

Fig. 1
figure 1

Genes and transporters underlying uptake, translocation and storage of iron and zinc in graminaceous crops. ZIP: zinc-regulated transporter; ZIP2: zinc-regulated transporter 2; YS: yellow stripe; YSL: yellow stripe-like transporter; MTP: metal transporter protein; ENA 1: efflux transporter of nicotianamine 1; VIT: vacuolar iron transporter; IRT1: iron-regulated transporter 1; NRAMP 1: natural resistance-associated macrophage protein 1; FRDL 1: ferric reductase defective-like 1 transporter; TOM1: transporter of mugineic acid family phytosiderophores 1; FER: ferritin; NAS: nicotianamine synthase. Figure generated using Biorender

Full size image

Biofortification for Fe content in rice has been successfully achieved by Masuda et al. (2013) through seven transgenic approaches. In the first approach, endosperm-specific expression of Fe storage protein, ferritin (SoyferH1, SoyferH2, Pvferritin), was performed to achieve a twofold increase in Fe content. In the second approach, nicotianamine synthase genes (OsNAS1, 2, 3; HvNAS1) were overexpressed to produce the metal chelator, nicotinamine. The transgenic plants showed a threefold increase in grain Fe content. Enhanced expression of a Fe(II)-nicotianamine transporter gene (OsYSL2), in the third approach, increased the Fe content by fourfold in the grains. Approach four had introduced a mugineic acid synthesis gene of barley (HvIDS3) in rice to enhance the Fe content in grains by 1.4-fold. The fifth approach was to overexpress the Fe transporters, OsIRT1 and OsYSL15, the sixth approach involved the overexpression of Fe homeostasis-related transcription factor (Iron-related bHLH transcription factor, OsIRO2), and the seventh was to knock down the vacuolar Fe transporters (OsVIT1 and OsVIT2), which resulted in increased Fe accumulation in the grains. Combining these approaches had also proven successful in improving the Fe content. For example, approaches one, two, and three were combined to produce Fe-fortified rice, which showed 4—to sixfold high Fe content in the seeds (Masuda et al. 2013).

Agronomic biofortification by incorporating biofertilizers in the soils eventually plays a crucial role in solubilizing the micronutrients for uptake by the plants. Microbiome in the rhizosphere, including terrestrial fungi and bacteria, establishes the essential association with the plants to solubilize the soil nutrients for enabling the plants to take up through symbiotic relationships. This underlines the importance of other external factors like symbionts and fertilizer application in improving nutrient availability. An increase in the acquisition of Fe and Zn was successfully achieved by inoculation of arbuscular mycorrhizae in the rhizosphere (Coccina et al. 2019), ferti-fortification by incorporating increased external soil application (Clemente et al. 2007), foliar application of micronutrients (Cakmak et al. 2010), iron solubilizers like azotobacter and azospirillum (Hussain et al. 2018), and iron- and zinc-coated fertilizers (Kutman et al. 2010). Understanding the mechanism underlying these accessory uptakes and identifying the genes having roles in the processes could also extrude a more significant drift in enhancing the absorption of micronutrients into the plants from the soil.

Conventional plant breeding approaches for enhancing micronutrient contents

Conventional breeding for biofortification has been successful in delivering improved lines to the farmers. However, the success was limited to the crops with excellent genetic diversity, whereas transgene-based approaches were used in the species that had limited genetic diversity, reduced heritability, and linkage drag. Crossing the parent lines with high micronutrient content with the recipient lines with better agronomic traits for several generations develops elite lines with both enhanced micronutrient content and desired agronomic traits. This has been demonstrated in several varieties, including IR6844 (IR 8 × Taichung Native 1) and IR68144-3B-2–2-3 (IR72 X Zawa Bonday), which retained 80% of the Fe content in the polished seeds (Gregorio et al. 2000; Virmani and Ilyas-Ahmed 2008; Has et al. 2005). The Consultative Group for International Agricultural Research (CGIAR), in collaboration with CIAT (International Center for Tropical Agriculture) and IFPRI (International Food Policy Research Institute), established the HarvestPlus initiative, which focuses on breeding for biofortified crop species. The initiative was successful in enhancing Fe, Zn, and vitamin-A in several cultivated species, including rice, wheat, and maize (Bouis and Welch 2010). In addition to deploying breeding to improve the nutritional levels, the approach has also been instrumental in reducing the antinutritional content in grains. For example, marker-assisted backcross breeding developed low phytic acid grains in rice, maize, and wheat, ensuring better bioavailability in the human body after consumption (Virmani and Ilyas-Ahmed 2008; Velu et al. 2014; Jeng et al. 2012). In India, the Indian Council of Agricultural Research (ICAR) and International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) had released several biofortified varieties for cultivation by farmers.

Dissecting micronutrient traits using genetics and genomics approaches

The agronomic and conventional breeding approaches operate without the knowledge of the genes, alleles, and QTLs underlying the desired traits, whereas genetics and genomics had proved successful in identifying the precise molecular determinants, thereby enabling the researchers to manipulate them for trait improvement. In this direction, a considerable amount of work has been done in the previous decades to identify the genetic determinants underlying grain micronutrient contents (Table 2). Providing direct access to the genes, alleles, and QTLs regulating these traits overcome the limitations in agronomic biofortification and conventional breeding.

Table 2 Genetic determinants identified in graminaceous crops having roles in micronutrient traits
Full size table

Rice

Rice being a major staple cereal necessitates the enhancement of its grain micronutrient content. Lu et al. (2008) initially identified micronutrient QTLs in 241 RILs obtained from a cross between Minghui 63 and Zhenshan 97. The QTLMapper1.0 detected 10 QTLs for micronutrients, of which two major QTLs for Fe (qFE-1, qFE-9) and three major QTLs for Ca [qCA-5 (chromosome 5), qCA-9 (chromosome 9), and qCA-4 (chromosome 4)], one minor QTL for Mn (qMN-1, on chromosome 1) and Cu (qCU-2 on chromosome 1), three minor QTLs for Zn (qZN-5, qZN-7, and qZN-11) were detected. The QTLs, qFE-1, qFE-9, qZn-5, qZn-11, and qMN-1 with digenic interactions for micronutrients were also located. This presented the involvement of more genes in enhancing the micronutrient status. Similarly, five QTLs in RILs between Madhurkar and Swarna for Fe and Zn concentrations on chromosomes 1, 3, 5, 7, and 12 were detected by Anuradha et al. (2012). Interestingly, the study showed a co-localization of Fe and Zn QTLs on chromosome 7 which underlined their interrelationship. Further, putative genes for Fe (metal tolerance protein, OsMTP1; and OsYSL1), zinc (OsNAS1-2; acireductone dioxygenase, OsARD2; and OsIRT1), as well as both Zn and Fe (adenine phosphoribosyltransferase, OsAPRT; and OsNAS3), were identified from this study, and this could be implied in biofortifying rice for micronutrients. Parallel co-localization in double haploids (DH) was seen in QTLs of Mg/Mn, Mg/P, and Mn/Zn present on the 8th and 9th chromosomes. This describes the desirable selection for combined improvements for micronutrients in cereals. Double haploids were also utilized by Swamy et al. (2018a, b) to identify 59 QTLs and six gene families such as endogenous ferritin (OsFER), OsNAS, natural resistance-associated macrophage protein (OsNRAMP), OsYSL, and zinc-induced facilitator-like (OsZIFL) for Zn content in rice.

Later, Calayugan et al. (2020) assessed rice DH derived from IR05F102 X IR69428 for three different seasons by best linear unbiased estimates (BLUEs) from SNPs. These data were further used for inclusive composite interval mapping to identify 23 QTLs distributed on all the chromosomes except 4, 8, and 11. The dissected QTLs for micronutrients were also found to be associated with eight agronomically significant traits. Among all, two QTLs for Fe (qFe9.1 and qFe12.1) on chromosomes 9 and 12, and 4 QTLs for Zn (qZn1.1, qZn5.1, qZn9.1, and qZn12.1) on chromosomes 1, 5, 9, and 12, respectively, were notable. Interestingly, the SNP interval 9,809,545–9,819,278 showed co-localization of qFe9.1 and qZn9.1 on chromosome 9. This corresponds to the previous studies in understanding the genic linkages and confirms their interrelationship. Further, candidate genes within the QTLs like, LysM receptor-like kinase 10 (OsLysM-RLK10) and Receptor-like Cytoplasmic Kinase 276 (OsRLCK276) (within qFe9.1), SWEET13 (sugars will eventually be exported transporter), and OsSWEET13 (within qFe12.1), ARGOS-like (OsARL1e), OsGATA8, Sar1b and OsGATA14 (within qZn1.1), and Os09g0511500 (within qZn9.1) were identified. Among these genes, the position of rice Zrt and Irt-like protein 6 (OsZIP6) was toward the right of the QTL qZn5.1, which was an Fe transporter, also stratifies the association between iron and zinc uptake in crops (Calayugan et al. 2020).

Although several QTLs have been identified, they differ with respect to the confidence level and ambiguous localization in the chromosomes (Fig. 2). Hence, meta-QTL analysis was more preferable to project the consensus QTLs with their exact location. This robust genome-based technique pools and analyzes data that have already been reported to show concordance level of the traits/QTLs. Several research groups have performed meta-QTL analysis, and one such study by Dixit et al. (2019) identified meta-QTLs for Fe and Zn, on chromosomes 3 and 2. The in silico analysis of these QTLs also found few candidate genes, namely OsFDR3, Auxin-responsive Aux/IAA gene (IAA5), Proton-dependent Oligopeptide Transporter (OsPOT), and OsZIP4. Another meta-QTL analysis linked to Fe and Zn also detected 48 QTLs on chromosome 12 and identified 663 candidate genes, which could be widely used for marker-assisted breeding and biofortification in rice (Raza et al. 2019). These genes and QTLs identified through the meta-QTL analysis could be further employed for downstream studies to characterize and deploy them in trait improvement.

Fig. 2
figure 2

Physical map of rice showing the locations of QTLs identified for micronutrient traits by different studies. The vertical bars represent chromosomes and the QTL for each micronutrient is mapped in colored blocks. Also, the reference to the respective study has been provided to the right of each box

Full size image

Maize

In maize, Brunson and Quackenbush (1962) showed the variability in the germplasm to facilitate the breeding for provitamin A. Studies on the structural differences between α- and β- carotene revealed the role of lycopene epsilon cyclase (lcyE) locus (Harjes et al. 2008), and also, methodologies to improve carotenoid contents by incorporating beta-carotene hydroxylase 1 (crtRB1) and lcyE in maize were suggested (Yan et al. 2010). Later, in CIMMYT, GWAS for carotenoid content with 380 inbred maize lines identified 476,000 SNPs and their validation identified genes like crtRB1 and nonheme di-iron β-ring hydroxylase (HYD5 and HYD1), which had a crucial role in improving Vitamin A in maize. Moreover, genes like carotenoid cleavage dioxygenases (CCD1), Geranylgeranyl pyrophosphate synthase 2 (GGPS2), 1-deoxy-d-xylulose 5-phosphate synthase (DXS1), and Zeaxanthin epoxidase (ZEP1) were found to have a putative role in enhancing the total carotenoid content (Suwarno et al. 2015). Earlier, Aluru et al. (2008) had demonstrated that overexpressing bacterial phytoene synthase gene, crtB, and crtl with super γ-zein promotor increased the carotenoid (especially β-carotenoid) accumulation by 34-fold, for four consecutive generations, and could help develop genetic stocks. Recently, Natesan et al. (2020) have also developed introgression lines with 7.3-fold higher β-carotene using crtRB1 3'TE as a genic marker from HP467-15, UMI 1200, and UMI 1230. Interestingly, Gautam et al. (2010) had also explained that β-carotene has a significant role in increasing the bio-accessibility of other minerals. Thus, improving β-carotene could consequently enrich maize with other essential minerals.

QTLs for different micronutrients in maize revealed markers for Zn and Fe. Forty-eight genes associated with Fe and Zn in maize were classified into families, namely ferritin (1), cation efflux (1), ZIP (13), natural resistance-associated macrophage protein (NRAMP, 16 genes), and yellow stripe (17) gene family. This also identified 34 SSRs from 28 putative genes for Fe/Zn transporters, which was amplified in 124 inbred lines. Several SNPs were seated in the exons of candidate genes like ZmZIP11 and ZmZIP7. One SNP in ZmZIP11 exhibited polymorphism among the inbreds that could be used as a genic marker in molecular breeding programs (Sharma and Chauhan 2008). Later, Lungaho et al. (2011) had identified 3 small QTLs for grain Fe content using linear unbiased predictors (BLUPs) among RILs of B73 and Mo17. Molecular breeding using these small QTLs would prove to be a burdensome task. Later, Qin et al. (2012) had published data of a linkage map for Fe and Zn in Mo17, Mu6, and SDM inbred lines that progressed the QTL studies in maize for micronutrients. Here, the QTL identification and analysis were performed using inclusive composite interval mapping and identified 31 QTLs for Zn and Fe. Also, Jin et al. (2013) had performed both QTL analysis (178 × P53) and meta-analysis for QTLs (previous QTL research) associated with Fe and Zn and identified ten mQTLs and 5 QTLs using BioMercator2.1. Out of these ten mQTLs, eight were associated with both Fe and Zn, which could be used to explain their correlation in simultaneous improvements.

In addition to QTLs, GWAS was performed by Hindu et al. (2017) to identify the genomic regions associated with Fe and Zn in 923 inbred lines. This revealed 347,765 SNPs across varying environmental conditions. Out of all the SNPs, 46 SNPs were called for Fe and Zinc, of which 26 were for Fe and 20 for Zn. These associated SNPs could be used to assist the molecular approaches in identifying target donors from germplasm. Alternatively, transgenic approaches to enhance the Fe and Zn uptake by overexpressing the phytase gene (Drakakaki et al. 2005), and RNAi of ZmZIP5 for the uptake of zinc (Li et al. 2019) are other alternative methods to reduce the antinutritional factors like phytic acid in maize for enhancing the micronutrient bioavailability.

Barley

Genetic determinants for micronutrients and vitamin A in barley were also identified using different QTL, GWAS, and GBS approaches. Biparental mapping (Clipper and Sahara 3771) was used extensively to identify QTLs for Zn in barley by Sadeghzadeh et al. (2010, 2015) wherein they had identified QTLs for Zn on chromosome 2HS. Subsequently in 2015, they detected extra QTLs on the chromosomes 2HL, 3H, and 4H. Like rice and maize, the double haploid population in barley (Clipper and Sahara 3771) was used by Lonergan et al. (2009) to execute the association study using Map Manager QTX and QGene 4.0. They located 5 QTLs for grain Zn content on three chromosomes, namely 1H, 2H, and 5H (Lonergan et al. 2009). Following this, Mamo et al. (2014) experimented with a marker-linked association in 298 landraces of barley and identified a single QTL, Zn-qtl-6H_SCRI_RS_10655, on 6H chromosome. Later, critical QTLs such as IL157 on 1H and IL146 on 4H were dissected from introgression libraries designed for Fe and Zn (Reuscher et al. 2016).

Broader studies with 45 QTLs for 13 micronutrients [Ba (2), Ca (2), Cu (4), Fe (11), K (2), Mg (3), Mn (6), Na (4), S (3), Si (5), and Zn (3)] by GWAS (TASSEL) projected three significant QTLs for Zn viz, Zn-1H-21.97, Zn-2H-87.34, and E-5H-44.99 (Gyawali et al. 2017). Besides, population structure analysis across three regions with DArT™ markers and BLUEs presented 43 MTAs for micronutrients, viz. 15 for Cu, 6 for Fe, 9 for Mn, and 13 for Zn. The MTAs associated with Zn were present on the chromosome 2H (2H|bPb9754) near the 2 YSL (yellow stripe-like) genes (AtYSL2 and YSL9). MTAs associated with Cu had the highest association with the marker 2H|bPb4040. The pleiotropic effects among these micronutrient QTLs were between Zn, Cu, and Mn, which was linked to the marker bPb4909 (1H), Zn and Cu were linked to markers bPb9754 and bPb4040 (2H), and Zn and Fe were linked to bPb8836 (6H). This pleiotropism has to be further phenotypically evaluated to describe the morphological effects in enhancing the Fe and Zn content in barley. Finally, the validation of these QTLs established MTP5 as one of the candidate genes for Zn accumulation in barley (Detterbeck et al. 2019).

Sorghum

One of the significant problems in sorghum is the low bioavailability of micronutrients and vitamin A. The genetic variation in the level of carotenoids in sorghum grains was used for QTL and GWAS mapping. Fernandez et al. (2008) initially developed a RIL population with 85 SSRs to associate QTLs for carotenoids relating to their endosperm color. Nine putative genes for endosperm color, with Phytoene synthase 3 (Psy3) as a major contributor, were detected with a co-localization of a carotenoid Ccd1a gene. CIM among the QTLs revealed significant co-localization of QTLs for endosperm color with carotenoids. Phytoene desaturase (Pds) and Psy3 genes showed association with β- carotene and zeaxanthin, while six genes, Crtrb1 (5%), Ccd1a, Vp14, Lcye, Psy1, and Psy 3 (1%), showed significant association with lutein content. The elemental genotypic and phenotypic variation studied with the GBS data across three locations with Sorghum Association Panel was available from Morris et al. (2013), and these data exhibited a moderate heritability for Cu (45%), which was higher than the heritability of Ca, Co, S, P, Se, Mo, Mn, Fe, Zn, and K (< 30%). GWAS in these accessions identified several potential candidates, namely Sobic.007G064900 (for Zn content) was homologous to AtZIP5, Sobic.003G349200 (for Mn content) was homologous to AtMTP11, Sobic.001G213400 locus (for Fe content), and Sobic.001G443900 locus (for Mg content) which has been annotated as Peptide transporter (PTR2). Ionomics approaches in these genes were later suggested for dissecting the genomics of these genes (Shakoor et al. 2016). Later, Phuke et al. (2017), with 336 RILs, projected the negative association of Fe and Zn with yield, and their positive association with 100 seed weight suggested a selection pressure for bolder seeds in biofortification of sorghum. They also observed a higher environmental influence for Fe than Zn.

Another GWAS for carotenoids in sorghum revealed 14 SNPs for β-carotene from which only three were located in the proximity of major genes, viz. cytochrome P450 97A (CYP97A) (chromosome 2), PDS (chromosome 6), and geranylgeranyl diphosphate synthase (GGPPS; chromosome 2). For zeaxanthin, overall, 38 SNPs were identified with 12 association regions, out of which 4 showed presence near the priori gene candidates, and they were MDS (chromosome 4), ZDS (chromosome 2), DXR (chromosome 3), and ZEP (chromosome 6). For lutein, no significant SNPs were detected. This study indicated a profound association between the SNPs and the Zeaxanthin content as zeaxanthin epoxidase (ZEP) and 2-C-methyl-D-erythritol-2,4-cyclodiphosphate synthase (MDS) genes were associated with zeaxanthin. They proposed that ZEP could have a pertinent role in controlling carotenoid content in sorghum and could be further utilized in increasing the provitamin A content in sorghum. Development of RILs and ILs is in progress to validate the markers (Cruet-Burgo et al. 2020). Probably, marker assisted selection using the identified markers could aid in fast forwarding the bio-fortification process in sorghum.

Genetics and genomics of micronutrient traits in small millets

Minor millets are hidden reserves for several traits, and they are highly nutritious than mainstream cereals. They possess the desirable glycemic index and are gluten-free in nature. Each millet has its distinctive features, which are yet to be explored (Muthamilarasan and Prasad, 2021). Among them, little millet exhibit abundance in micronutrients with a considerably higher iron content. Studies portrayed ferric chelate reductase (FRO2) in metal uptake of plants, which was characterized in little millet using next-generation sequencing with the rice ortholog, OsFRO2. The amplified FRO2 (2.7 kb) gene from little millet (RLM-37 genotype) showed a significant similarity with the sequence of OsFRO2. The protein sequence of FRO2 incorporated the domains like NOX_Duox_Like_FAD_NADP and ferric reductase and showed fascinatingly higher similarity in 3D structure of both little millet FRO2 and OsFRO2. This study recommended that desirable alleles and micronutrient associated gene orthologues in millets could be identified using the already reported genes in staple crops (Chandel et al. 2017). Previously, Kumar et al. (2015) had evaluated 113 genotypes of finger millet which is known for its Ca content which is of about 450 mg per 100 g. Molecular screening of this germplasm with 23 anchored-SSR primers designed based on Ca transporters and sensors revealed polymorphism for 14 markers, and they contributed to almost 83 alleles. Dendrogram analysis divided the accessions into 7 clusters, and the association studies with TASSEL and STRUCTURE identified nine markers linked with the Ca content, which could be validated in the near future (Kumar et al. 2015). The studies in small millets are still in its budding state, and several efforts are ongoing to bring these in the mainstream research.

Transcriptomic approaches to understand micronutrient traits in major cereals

At present, the NCBI SRA database contains easily accessible RNA-seq data of more than 35,250 samples of rice, and also, Gene Expression Omnibus includes a massive amount of transcriptome profile data achieved from microarrays (GEO: https://www.ncbi.nlm.nih.gov/geo/). Several transcriptomic studies have been reported in rice and mostly based on stress. There are fewer reports in micronutrient content. In one of the studies by Zheng et al. (2009), microarray and transcriptomic analysis were performed to understand the antagonistic interaction of Phosphorous (P) and Iron (Fe). To decipher the complex nature of the rice transcriptome, Lu et al. (2010) used the rice cv. Indica and Japonica to develop the whole genome transcriptome profiles. Successively, Dong et al. (2018) grew the rice with and without minerals, and used root samples to generate the entire transcriptome RNA seq data. They explained the role of Ser/Arg proteins in the regulation of mineral nutrient homeostasis. Similarly, Sperotto et al. (2012) performed the expression profiling of 25 genes associated with metals and their homologs, out of which nine (OsNAC5, OsFRO1, OsNRAMP1,7 and 8, OsYSL6,8 and 4, and OsNAS1) were related to Fe and Zn uptake. Recently, Ren et al. (2019) analyzed large-scale RNA-seq data and identified 1584 novel peptides, which further improvised the annotation of the rice genome.

Messias et al. (2014) performed the expression analysis for the carotenoid-related genes among 22 landraces of corn (kernel) by RT-qPCR in the hybrid 30F53 at different developmental stages and amplified the genes ZmCCD1, CYP97C, HYD3, and PSY1. The results showed that the PSY1 expression was high at 16 to 22 days after pollination (DAP). Whereas both CYP97C and HYD3 exhibited a peculiar expression pattern wherein an increase in expression was observed at 10 DAP, it decreased at 13–16 DAP, followed by a sharp increase at 19 DAP. Inclusive to all these findings, there was a correlation in the expression of HYD3 with PSY1 and CYP97C. However, ZmCCD1 showed a negative correlation with the content of carotenoids. Also, it explained that the conversion to lutein and zeaxanthin from β-carotene could be determined by the expression analysis of CYP97C and PSY1 (Messias et al. 2014). Elucidation of elemental accumulation and variation in grains can be made easier with transcriptomic studies. For instance, in barley (cv Golden promise), Tauris et al. (2009) isolated the cells embryo, aleurone, endosperm, and the transfer cell tissues using Laser-assisted microdissection followed by an RNA extraction. Affymetrix-based microarray (22 K Barley GeneChip) was used to achieve tissue gene expression profile. This explained the abundance of 25 genes related to metal homeostasis, which were further categorized into several gene families like HMA (heavy metal ATpase), ZIP (Zrt-, Irt-like proteins), NRAMP (natural resistance-associated macrophage proteins), CAX (cation exchanger), VIT1 (vacuolar iron transporter), CDF (cation diffusion facilitator), ZIF1 (zinc-induced facilitator1), NAAT (nicotianamine aminotransferase) metallothionein, NAS (nicotianamine synthase), and YSL (yellow stripe-like). Using this gene expression data, a zinc trafficking path in barley grains to the maturing seed from the phloem was framed. HvYSL9 was expressed in all other tissues except for the embryo and is expected to have a metal transport role in cells (Tauris et al. 2009). Similarly, Detterbeck et al. (2019) studied the comparative transcriptome of barley lines with low and high Zn content. They reported 26 differentially expressed genes (DEGs), among which 19 could not be annotated. Further, data search revealed the presence of homeostasis genes, which were earlier reported by Tauris et al. (2009), and they were PCS (phytochelatin synthase), ZIFL (zinc-induced facilitator-like), and PME (pectin methylesterase). Concurrently, they also proposed that MTP and YSL transporters could have a significant role in Zn content in barley grains (Detterbeck et al. 2019).

Transcriptomic studies in small millets to understand micronutrient traits

The sequence information for all the species of small millets is unavailable, and they are yet to be disseminated for utilization, except for foxtail millet. Hence, the identified transcriptomes in small millets serve as a desirable approach in targeting the key genes in millets and cereals. Finger millet is rich in Ca content and has a variable concentration of Ca in plant tissues due to its differential gene expression. For the first time, Singh et al. (2014) conducted a transcriptome-wide investigation for Ca sensor genes in GPHCPB-45(GP-45) and GPHCPB-1 (GP-1) genotypes of the finger millet showing a considerable difference in their Ca content (about 100 mg). The tissues were collected for RNA isolation from four different spike developmental stages like spike emergence (S1), pollination (S2), dough (S3), and maturation (S4) and were subjected to transcriptome sequencing using Illumina HiSeq 2000 platform. De novo assembly of the transcriptome data against CDS of calcium sensor genes from rice was performed. They identified 82 distinctive calcium sensor genes, which were categorized into eight families [calmodulin (CaM) and calmodulin like (CaML), calcineurin B-like protein (CBLs), calcium-dependent and CaM-independent protein kinases (CDPKs), CaM-dependent protein kinases (CaMK), phosphoenolpyruvate carboxylase kinase-related kinases (PEPRKs), CDPK-related protein kinases (CRKs), Ca2+-/CaM-dependent protein kinases (CCaMK), and CBL-interacting protein kinases (CIPKs)]. Out of these, 24 genes in GP-45 (high-Ca content) and 11 genes in GP-1 (low-Ca content) showed upregulated expression (Singh et al. 2014), and these genes could be further validated and used for Ca biofortification.

Similarly, in another study, the investigators identified Ca-related genes using rice data from MPSS and Affymetrix gene expression. The identified genes (CAX1, TPC1, CaMK2, CaMK2, calmodulin, tubulin, and 14–3-3) were amplified and cloned with the vector pGEM-T Easy (Promega) and sequenced by Applied Biosystems 370. The Ca content determination revealed a considerable variation during the reproductive and vegetative stages. It was found that the flag leaf tissues accumulated more Ca with respect to the spike. The accession GPHCPB-45 had higher Ca than GPHCPB-1 with a difference of almost 100 mg. Also, the expression of the two-pore channel (TPC1) and CAX1 was higher in the spike of GPHCPB-45 could be the probable reason of higher Ca content. Even CAM showed a higher expression pattern similar to Ca2+ ATPases in GPHCPB-45. Interestingly, a contrasting Ca accumulation pattern was observed in both the genotypes wherein GPHCPB-45 had a higher accumulation of Ca in seed, whereas GPHCPB-1 showed a higher accumulation of Ca in flag leaf. Hence, this differential expression could be validated to understand the uptake of Ca. The study also proposed that the CAX1 gene could be a putative gene in enhancing the seed calcium content. Moreover, a construct containing both CAX and CaM gene would help in the biofortification of grains and cereals (Mirza et al. 2014). Recently, Kokane et al. (2018) investigated the role of the protein CIPK with respect to CAXs and Ca2+ translocation during grain filling in cv. GP-1 and GP-45. Their differential gene expression revealed that CaM was highly expressed in the GP-45 seeds, suggesting that it might have a potential role in Ca2+ movement and interacts with Ca2+ ATPase. Moreover, CAX1 showed higher abundance in vegetative tissues and in developing spikes, whereas CAX3 was highly expressed in spikes. Hence, they suggested that Ca2+ translocation and accumulation might be controlled by tripartite interactions (Kokane et al. 2018). Despite these studies, the transcriptomes of several other millets are yet to be studied to understand the differentially expressed genes during grain filling and maturation stages that have role in biosynthesis and accumulation of micronutrients in the grains.

Proteomics and metabolomics studies in major cereals and small millets

The proteome profiling of cereals for their micronutrients began from Gayen et al. (2016), who compared the nutritional content of transgenic rice with Xa21 (bacterial blight resistant) against the wild type using 2D gel electrophoresis and observed no significant variation in their nutrient profile. Most of the proteomics study in rice is predominantly analyzed for its stress response rather than its nutritional content. Metabolomics of cooked grains from 10 rice varieties using UPLC-MS for around 3097 metabolites revealed a key variation based on race (aus, indica and japonica). They also explained that the genes associated with the biosynthesis pathway showed variation in SNP. For tocopherol and phenolics, the gene γ-TMT was associated with vitamin E content in rice (Heuberger et al. 2010). Calingacion et al. (2012) studied three waxy rice cultivars (TSN1, HNN, and KNL) by integrating genome-wide genotyping with metabolomics (H-NMR, derivatized GC–MS, ICP-MS, and headspace GC–MS). SNP association and metabolic profile revealed that the metabolic profile for each rice variety was unique and showed relevance to nutritional content as the geographical origin of crop plays a significant role in plant metabolite profile. Parallelly, Hu et al. (2015) performed metabolite profiling of the rice cv. japonica and indica to dissect 121 seed metabolites that showed a correlation of phenotype with the metabolites which were dependent on the geographical location of rice origin.

Being natural and common pigment of grains, carotenoids are widely studied in metabolomics studies. CYMMIT and HarvestPlus have come together to fortify corn with provitamin A using breeding strategies. Worldwide, carotenoid characterization has been performed using targeted metabolomics. For instance, Kuhnen et al. (2011) assessed the carotenoid content in 26 landraces of maize (white, variegated, orange, purple, and yellow) using HPLC–UV-VIS to characterize the carotenoids, from which they found a significant amount of lutein and zeaxanthin, with traces of α-, β- carotene, and β-cryptoxanthin. Uarrota et al. (2014) characterized the carotene content in eight landraces, viz. MPA, Roxo, Roxa, and Palha, which had a higher accumulation of carotenoids like α-carotene, cis and trans-β-carotene and also β-cryptoxanthin. They detected the presence of non-provitamin A like lutein and zeaxanthin. Later, Messias et al. (2014) studied the carotenoid biosynthesis pathway and its catabolism across 22 landraces of maize (yellow, white, and orange). They used HPLC–DAD to characterize the carotenoids to differentiate α- and β-carotene and α- and β-cryptoxanthin. Non-provitamin A like, lutein and zeaxanthin were also characterized to be higher in MC3 and MC14. Similarly, in another study, varieties of sweetcorn, viz. Jingtian 3 and 5, and varieties of waxy corn, viz. Jingtianzihuanuo 2, Jingnuo 8 and Suyunuo 11, were collected and found that α-cryptoxanthin (provitamin A), lutein, and zeaxanthin were the more significant carotenoids present in corn varieties (Song et al. 2015).

Beyond these approaches, thirty-five orange and twenty-six white maize landrace procured from different areas of Malawi revealed lutein to be the most abundant carotenoid followed by zeaxanthin, β-carotene, and β-cryptoxanthin in the orange landraces. Their analysis suggested that orange maize could be a provitamin A natural source (Hwang et al. 2016). Succeeding them, four maize landraces were evaluated for their carotenoid content and compared to commercial lines. The local landraces (Nano di Verni) contained higher amounts of β-carotene and β- cryptoxanthin than lutein (Capocchi et al. 2017). There are many reports on characterizing carotenoids in maize germplasms to increase provitamin A content by breeding or genome editing strategies. Zn could be bound to high molecular weight protein in the grains, and this was studied by Dionisio (2018). They conducted a proteomics study on Zn binding proteins in the grains of barley using three different techniques viz. SDS-PAGE (protein separation), PVDF (blotting), and DTZ stain. Later, they identified the Zn binding proteins like 7S globulins, β and γ hordeins, dehydrins and LEA, and prolamin family members, which would help in biofortification programs. They also suggested that these proteins might have a role in binding other cationic elements, which could be validated with further investigations (Dionisio 2018). Expanding the proteomic and metabolomic studies to other underutilized graminaceous crop species will widen the understanding of micronutrient biosynthesis and accumulation in the grains, which could further be exploited using multi-omics approaches for enhancing the contents.

Phenomics studies to understand micronutrient traits in graminaceous crops

There are various high throughput tools available for phenotyping in rice from its leaves (Micol et al. 2009), rice panicle (Ikeda et al. 2010), the morphology of roots (Zhu et al. 2011), biomass of shoot (Golzarian et al. 2011), grain yield (Duan et al. 2015) and photosynthetic integrity (Bauriegel et al. 2011). Techniques like visible-light imaging, near-infrared imaging, hyperspectral imaging, fluorescence imaging, digital X-ray radiography, and X-ray computed tomography are the major photonic-based strategies used in phenomics (Yang et al. 2013). Unfortunately, there is no technique yet that could help in phenotyping the micronutrient traits. Several phenotyping facilities are present to determine the morphological phenotypes of maize, but these do not help understand the micronutrient content. Hence, most of the phenomics studies in maize are based on metabolite analysis, as discussed earlier. In barley, Long et al. (2017) developed RNAi lines of HvIRT1 that showed a 5% decrease in the gene expression under Mn supplementation. The plants were grown in alkaline soil depriving the plant for Mn, and their phenotype was monitored. The RNAi lines showed less biomass in the shoot with chlorosis symptoms, whereas the wild type grew properly with increased biomass. Upon exposure to radiolabeled Mn, the RNAi lines showed reduced Mn uptake than the wild type. Interestingly, the RNAi lines showed a twofold higher accumulation of Mn in the root apoplast, which might be the probable reason that Mn could not be loaded to stele, xylem, and pericycle cells. Hence, this proposed the role of the IRT1 gene in barley for higher Mn accumulation (Long et al. 2017). To study the decorticated carotenoid content in sorghum grains, Fernandez et al. (2008) analyzed the grain endosperms by color using Konica Minolta colorimeter CR-300, and out of the three (a, b, and c) output spectrums, b was selected, which infers the role of the intensity of the yellow endosperm color. These studies necessitate the importance of phenomics for studying the molecular determinants of grain micronutrients.

Integrated omics approaches and strategies for genetic improvement in micronutrient traits

Genetic engineering has been an indispensable tool for manipulating the genome to increase the nutrient content (Table 3). Agrobacterium-mediated transformation was used to transfer the ferritin gene from soybean to rice to enhance Fe content (Goto et al. 1999; Sivaprakash et al. 2006). Later, Ye et al. (2000) incorporated the biosynthetic pathway for β- carotene in rice endosperm to enhance the provitamin A. Similarly, Lee et al. (2009) used nicotianamine synthase gene activation to biofortify rice with Fe. Interestingly, Singh et al. (2017) developed transgenic rice lines with enhanced Fe, Zn, and β-carotene content by incorporating three genes in one locus, namely nicotianamine synthase 1 (AtNAS1) from Arabidopsis, bacterial carotene desaturase (CRTI), ferritin (PvFerritin from bean), and phytoene synthase (ZmPSY) from maize (Singh et al. 2017). Earlier, Colmsee et al. (2012) had developed the OPTIMAS—Data Warehouse (DW) for maize, and it compiles data generated from all the omics techniques, including experimental and analytical data from various research. The import tool of the OPTIMAS-DW was based on Javascript that aided in uploading the data from the experiments performed. For easy accessibility of the data, the tool supports various file formats that can further analyze and visualize datasets (Colmsee et al. 2012). In order to study the reproducibility of different profiling techniques, Zeng et al. (2014) selected three tissues to perform metabolomics and microarray studies from 11 traditional varieties of maize. They compared the data from both the studies and suggested that microarray data could be hindered by the variant genomic sequences with unknown functions (Zeng et al. 2014).

Table 3 Transgenic crops developed with enhanced micronutrient traits using Agrobacterium-mediated or biolistic approach
Full size table

Several transgenic maize varieties have been produced for disease resistance, but only a few are made to enhance micronutrient trait. One such study was performed by overexpression of bacterial phytoene synthase gene, crtB, and crtl in maize using super γ-zein promotor that increased the carotenoid (especially β-carotenoid) accumulation by about 34-fold, and the trait was reproduced for four generations. They have used a biolistic approach for transforming maize zygotic embryos with two plasmid constructs, viz. pBAR184 + pRBS + pRIS and pBAR184 + pRB + pRI, wherein pRB contained crtB gene and pRI contained crtI gene. The transgenic lines showed varying carotenoid content in each generation and individual kernels from a particular year. This could be the result of epigenetic factors or germplasm that were selected for transformation. They also concluded that the enhanced β-carotene content shows a higher correlation with the increased abundance of lycopene β-cyclase gene (Aluru et al. 2008).

Tiong et al. (2014) developed transgenic barley lines overexpressing HvZIP7 gene and observed that it was highly induced in leaves and roots with higher accumulation of Zn during deficiency of Zn with no alteration in the Fe, Mn, Cu, and Cd content. However, the study does not explain the accumulation of Zn in the grain (Tiong et al. 2014). Earlier, Podar et al. (2012) have observed that the HvMTP1 helps in transporting Zn to the vacuole when expressed in yeast and proposed a hypothesis where they have mentioned that overexpressing HvMTP1 with D-hordein (endosperm specific promotor) could enhance the endosperm Zn content. Later, they expressed this HvMTP1 gene along with D-hordein in the Golden Promise cultivar of barley and studied the gene expression. RT-qPCR aided in the selection of lines with transformed homozygous plants (2-homozygous out of 3). To research and compare the gene abundance, they used Golden rice as control, GFP (with CaMV35S promoter) expressing homozygous line and the two homozygous transformed lines. The RT-qPCR revealed that both the transformed lines had a higher expression level of HvMTP1 transcript than the GFP-expressing line and control, post 21 DAP. The transformed plants were cultivated until maturity in Zinc devoid soil. The grains were harvested at maturity, and the Zn content was measured using ICP-OES for all three transformed lines. Higher accumulation of Zn was observed in the grains of transformed homozygous plants than the transformed heterozygous, control, and GFP expression plant. The homozygous transformed lines also showed a higher accumulation of Cu, Fe, and Mn. They again performed the same experiment by adding Zn in soil. Interestingly, all the plants showed an average increment of 103% Zn accumulation compared to the plants grown in Zn devoid soil. Surprisingly, there was no fluctuation in the Fe content in both conditions. They also confirmed that the HvMTP1 is expressed in the grain endosperm using three different techniques of DTZ (Zn-specific stain), ICP-OES, and fluorescence analysis with Synchrotron X-ray (Menguer et al. 2017). In barley, loading of Mn in the grain could be very fast during early developmental stages of grain due to the increased expression of HvIRT1. Surprisingly, Long et al. (2017) also observed an increase in expression of HvIRT1 (35-fold) in the root. They developed HvIRT1 RNAi lines of barley and observed that the RNAi lines showed diminished photosynthesis function due to low Mn content. In rice and Arabidopsis thaliana, it has already been reported that IRT1 functions in the regulation of Fe, but in this study, no such role for Fe was observed (Long et al. 2017). Hence, this study could help understand the evolution of the IRT1 gene.

Sorghum, although a staple food, lacks essential micronutrients like Fe, Zn, and β-carotene. Lipkie et al. (2013) developed transgenic lines to increase the bioavailability β-Carotene in sorghum. Transgenic sorghum lines were generated using the Agrobacterium-mediated transformation strategy in the immature embryo of sorghum TX430. The transgenic were developed using three construct, viz. (i) ABS168 contained golden rice genes (Paine et al. 2005), but the sorghum promoters of α-Kafirin with ZmPST-1 and β-kafirin with CRT-1 gene from Erwinia uredovora with selectable marker PMI gene, (ii) ABS188 contained all genes from the previous vector along with LPA-1 gene (Low phytic acid) (Increased Zn and Fe bioavailability) and (iii) ABS203 contained and extra AtDXS gene (for increased isoprenoid) + γ-zein promoter, HvHGGT gene + α-Kafirin along with the genes in the vector ABS168. The homozygous transgenic plants were achieved with multiple self-pollination. They have harvested transgenic sorghum grains followed by a porridge preparation, which was used for carotenoid determination. The carotenoid determination using HPLC–DAD plus YMC C30 column showed the presence of several carotenoids like zeaxanthin, lutein, α- and β-cryptoxanthin, all-trans-β-carotene and cis-β-carotene isomers, and provitamin A. One of the transgenic sorghums, Homo188-A, showed the highest bioavailability of provitamin A. On the other hand, no major effect was observed in the transgenic lines containing the LPA-1, HGGT, and DXS (Lipkie et al. 2013).

Similarly, in another study, three vector constructs were made using combinations of genes like Pantoea ananatis CRTI, ZmPSY1, HvHGGT, E. coli PMT AtDXS that were cloned in pSBI super-binary vector. The genotype TX430 was grown in greenhouse, and the immature embryo was harvested for Agrobacterium-mediated transformation. The selected transgenic lines were selfed to achieve 3:1 transgenic to non-transgenic seeds, which was further confirmed with qPCR. Finally, to achieve 100% homozygous seeds, the plants were self-pollinated. The carotene content of the seeds was analyzed using HPLC. The PSY1 protein accumulation was determined using LC–MS/MS, which showed a higher accumulation of all-trans β-carotene in the transgenic seeds (visibly orange in color). Unfortunately, at room temperature, during seed storage, the β-carotene is degraded due to oxidation. Therefore, they also co-expressed the HGGT (homogentisate geranylgeranyl transferase) gene, positively affecting tocopherol and tocotrienol biosynthesis. Vitamin E is a known antioxidant and hence reduced the oxidative degradation of β-carotene significantly. Thus, Che et al. (2016) successfully developed transgenic sorghum lines with increased Vitamin A content. Despite these reports and success stories in major cereals, no attempt has been made till now in genetic engineering or genome editing of small millet species to understand or improve the grain micronutrient content.

Roadmap to improve micronutrient content and availability in graminaceous species

Micronutrient availability in plants is a crucial content in ensuring food security to the people. Owing to these, several genes and QTLs associated with micronutrient traits have been identified and discussed in the previous sections. These findings had a significant impact on increasing the bioavailability of micronutrients in cereals (Fig. 3). The QTLs, alleles and genes could be effectively introgressed for improving the micronutrient status of the ruling crop varieties and the parents of the elite hybrids. This would satisfy the final utility of the consumers. Commercializing these cultivars at an affordable cost to the underprivileged society would further reduce the nutritional hindrances existing across the world.

Fig. 3
figure 3

Schematic representation of current and future strategies for micronutrient trait improvement in graminaceous species

Full size image

In addition, there have been several reports where genome editing has benefitted grass family members like in maize; zinc finger nucleases (ZFNs) were used to insert PAT gene cassettes into the endogenous maize ZmIPK1 gene along with alterations in inositol phosphate profile for the development of maize seeds (Shukla et al. 2009). Similarly, in rice, ZFNs were used for identification of safe domains for introduction of genes in rice plants to aid in gene stacking and gene integration (Cantos et al. 2014). Modern techniques of genome editing like TALENs (Transcription Activator- Like Effector Nucleases) and CRISPR are also exploited in improving crops like maize, rice, wheat, etc. TALEN-based mutagenesis was employed in sabotaging bacterial blight susceptibility gene, SWEET14, to develop blight-resistant rice (Li et al. 2012). Identical studies have reported that powdery mildew resistance in wheat was developed by knocking out three homologs of MLO gene (Jung and Altpeter 2016). TALENs were also employed for improving agronomic traits, as in case of maize, the GLOSSY phenotype was enhanced by removing GL2 genes (Kannan et al. 2017). However, nowadays CRISPR Cas technique is most prevalent for approaches like site-directed mutagenesis using embryo bombardment in plants like maize and wheat (Svitashev et al. 2015; Liang et al. 2017). In fact, the CRISPR Cas technique has been optimized in monocots by deploying hexaploid wheat (Triticum aestivum) as a template (Liang et al. 2017). Several transgenic approaches are in progress to attain higher nutritional levels while inculcating advanced genome editing techniques to develop superior cultivars in a short breeding cycle. The prospectus of biofortification in cereals should also be widened in the future by identifying desirable genomic regions from small millets as they are closely related to the major cereals. In small millets, nutritional content could be improved by exploiting the knowledge of micronutrient associated biochemical pathways and their regulation using comparative genomics approaches and genome editing. This substantiates and calls for the proportional focus on exploring the nutritionally rich small millets in the near future.

Conclusions

The importance of micronutrients in the grains of cereal crops has been well recognized, and several focused studies were performed to identify the genes, alleles, and QTLs underlying this complex trait. In addition, breeding and biotechnological approaches are already being deployed for the genetic improvement in major cereals, as evident from the literature; however, a portion of graminaceous crops labeled as 'small millets' remains understudied. Though these crops are well known for their superior grain quality traits, including elemental composition, they are not much explored. The development of genomic resources is the first step to proceed with any study to delineate the genes or genomic regions that regulate the complex traits. To achieve this, next-generation sequencing technology should be used to sequencing the genomes and transcriptomes for identifying the genes and pathways and develop large-scale genome-wide molecular markers that are useful for genotyping purposes. The availability of germplasm resources in global repositories should be exploited for this purpose for extensive phenotyping of micronutrient traits. Advanced approaches like GWAS could be effectively used to pinpoint the marker-trait associations as the QTLs underlying micronutrient traits in these crops. Once identified, the information could be further used to functionally characterize the genes and corresponding gene families in millet species and to establish their role in regulating micronutrient content in the grains. These data can be further used to genetically enhance the micronutrient traits in millets per se, and also, comparative genomics could facilitate the implementation of this information for genetic improvement in other graminaceous species.