Accelerated Generation of Elite Inbreds in Maize Using Doubled Haploid Technology

Abstract

The creation of homozygous parental lines for hybrid development is one of the key components of commercial maize breeding programs. It usually takes up to 6 to 7 generations of selfing to obtain homozygous inbreds from the initial cross using the conventional pedigree method. Using doubled haploid (DH) method, concurrent fixation of all the genes covering entire chromosomes is possible within a single generation. For generation of DH lines, haploids are generated first by several means such as in-vitro method using tissue culture technique and in-vivo method using the haploid inducer (HI) lines. Of which, tissue culture-based methods have shown little promise for large-scale DH production as it needs good infrastructures and technical requirements. In contrast, inducer-based method provides more optimistic solutions for large-scale DH lines production. Due to its rapidity, DH technology is now being adopted in many countries including India for reducing the breeding cycle.

1. Introduction

Maize breeding strategies rely heavily on the creation of homozygous parental lines for hybrid breeding. Using the traditional pedigree approach, it might take up to 7 generations of selfing to achieve homozygous inbreds from the first cross (Figure 1). In this context, because of its economic and logistical practicality, the creation of doubled haploid (DH) has received a lot of attention for varietal development in the last two decades [1]. The DH approach allows for simultaneous fixation of all genes across complete chromosomes in a single generation [2]. Haploids are initially created by a variety of methods, including an in-vitro way utilizing tissue culture techniques and an in-vivo method employing haploid inducer (HI) lines. On the other hand, Tissue culture-based approaches have not shown much promise for large-scale DH production since they necessitate adequate infrastructure and technical requirements. In contrast, because of its viability for large-scale DH line production, the inducer-based technique is deemed more hopeful and cost-effective [3]. According to the parent from which the haploids are being formed, inducer-based haploids may be divided into maternal haploid and paternal haploid [1]. With the exception of their cytoplasm, the genetic constitutions of both forms of haploids are identical. Once haploids are created, their genetic constitutions can be duplicated in a single step to produce homozygous DH lines. DH technique has recently been implemented in commercial maize breeding programs in various countries because of its speed in decreasing the breeding cycle.

Figure 1.

Conventional breeding method for obtaining inbred lines.

2. Genetic factors involved in maize haploid induction

The inheritance of haploid induction rate (HIR) has been extensively investigated during the last two decades. In-vivo haploid induction in maize is accomplished using three approaches (Figure 2). The first method uses a mutation originating from ‘Stock 6’ to induce maternal haploid induction, and it is commonly employed in commercial maize breeding programs [3]. On chromosome 1, a pollen-specific phospholipase-A gene known as Matrilineal (MTL) was discovered with a 4-bp insertion in the final exon of the gene, causing a flame shift mutation and premature stop codon [4, 5, 6]. When coupled with the MTL gene in homozygous recessive form (mtl/mtl), a second mutant gene called ZmDMP (encoding DUF679 domain membrane protein) on chromosome 9 increases the haploid induction rate [7]. As a result, mutations in the MTL and ZmDMP genes are required for maize maternal haploid induction. The second strategy, on the other hand, entails a mutation in the indeterminate gametophyte1 (ig1) gene, which was identified in Wisconsin-23 (W23) inbred for paternal haploid induction [8]. It was previously recognized that ig1 on chromosome 3 encodes a lateral organ boundary (LOB)-domain protein, which is part of a wide family of transcription factors important for plant lateral organ development [9]. Because the underlying determinants for both procedures (maternal and paternal) were different, the mechanisms of haploid induction in both ways differed greatly in the commercial breeding programs. Aside from these two key changes, induction of haploids in maize is influenced by several variables, including the maternal genetic background of the donor germplasms and the environment in which the induction crosses are made [1, 10, 11]. Another approach for generating both maternal and paternal haploids from a single inducer is CENH3-based haploidization [12]. CENH3 gene codes a histone H3 protein variant found in centromeric nucleosomes that have two primary domains. In CENH3 protein, the N terminal tail domain has little in common with regular histone H3, but the C terminal histone fold region has a lot in common with normal histones [13]. At the homozygous stage, a mutation in CENH3 is fatal because chromosomes fail to segregate to poles during cell division due to the lack of functional centromeres [14].

Figure 2.

Production of haploids using in-vivo induction system. A: Maternal haploid induction using mtl mutant; B: Paternal haploid induction using ig1 mutant; C: Both maternal and paternal haploid induction using cenh3 mutant.

3. Haploids identification after induction crosses

During an induction cross, haploids appear at a frequency of ~10% depending on the HIR, while the remaining 90% of seeds are diploid with no utility [3]. As a result, distinguishing haploids from diploid offspring at the seed, seedling, or mature plant stage is critical. Reducing the number of progenies would be helpful since it would lower the cost of developing DH lines. Different morphological and molecular indicators can be included in the inducer genotypes utilized in the DH development process [15, 16]. The dominant genetic marker produced in the seed or seedling stage can be included in mother haploid inducers, allowing haploids formed from induction crosses to be differentiated [1]. In most maize breeding programs across the world, haploid inducer with R1-nj (Navajo) allele is commonly utilized for haploid identification [1, 15]. The purple color of the scutellum of the embryo and the aleurone layer of the endosperm in diploid seeds influence the Navajo phenotype. Anthocyanin, on the other hand, is exclusively found in the endosperm of haploid seeds, not in the embryo. R1-nj marker makes it easier to distinguish haploids from diploids based on visual inspection due to the different colors of the embryo [17, 18]. The existence of a dominant inhibitor allele C1-I in tropical elite inbred lines inhibits anthocyanin production on seeds, which is a fundamental restriction of the R1-nj marker-based method [16]. When identifying haploids exclusively based on the R1-nj marker is challenging, the Pl1 gene is employed as an alternative [19]. Pl1 gene induces light-independent anthocyanin synthesis in seedling roots, allowing haploids and diploids to be identified that were previously misclassified by the R1-nj marker. The Pl1 gene, on the other hand, can frequently lead to misinterpretation due to the formation of red roots in seedlings after exposure to sunlight [19]. Furthermore, in the adult stage, recessive phenotypic mutations such as liguleless can be utilized to identify haploids [1].

Because of the numerous drawbacks of phenotypic morphological markers, multiple attempts have been made to use genetic markers based on the xenia effect of high oil content for haploid identification [20]. The use of a haploid inducer with a high oil content would be advantageous since the high oil marker is not genotype-dependent, allowing it to be applied to practically all genotypes, including landraces and wild cousins like teosinte [1]. As a result, the genes that cause high oil content may be targeted in order to create inducers with high oil content. The effectiveness of the oil-based identification technique, on the other hand, is dependent on a large difference in oil content between source germplasm and inducer, since a little difference would result in a higher number of false positives and false negatives [21, 22]. Automating the process of haploid identification would be a cost-effective and practical solution since it would considerably cut the cost of wages for those participating in the haploid identification process [23]. Several mechanical approaches have been altered based on R1-nj marker expression on embryo and endosperm employing multispectral, hyperspectral, and fluorescence imaging techniques (Figure 3). In this case, an imagining-based automated approach powered by machine learning and deep learning understanding might be a feasible option since it decreases the time and effort required to identify haploids [23, 24]. As a result, numerous approaches for haploid identification employing in-vivo HI are available. The main task is to create an HI with a mix of appropriate markers that can solve all of the problems associated with the identifying procedure. No marker method can give a universal strategy that is applicable to all germplasm [1]. As a result, developing an inducer with a proper marker system for the desired breeding program is critical before starting any DH programs. As a result, it is often recommended to create a set of HI with distinct marker systems appropriate for different types of source germplasms.

Figure 3.

Development of automation system based on hyperspectral spectroradiometer using a machine learning algorithm.

4. Chromosome doubling of identified haploids

The next objective is to create DH lines from the haploids after a successful induction cross using an appropriate inducer [3]. Haploids are normally infertile since they only have one copy of each chromosome, thus they must be chromosomally duplicated. Chemicals that prevent haploid seedlings from mitotically duplicating are used to achieve artificial chromosomal duplication (Figure 4). Colchicine is the chemical of choice in DH pipelines for artificial chromosomal doubling [1, 3]. Initially, haploid seeds are recognized using any of the markers and then germinated on paper towels until the coleoptiles reach a length of 2 cm. Before submersion in colchicine, the coleoptile tip is cut off, and the seedlings are rinsed out under tap water. The seedlings were then placed in trays filled with peat moss and kept at room temperature until they reached the three-leaf stage. Viable seedlings were then transplanted to a DH nursery field with suitable row-to-row and plant-to-plant spacing [3].

Figure 4.

Pipeline for doubled haploid production following colchicine treatment.

5. Application of doubled haploid technology

Visual selection based on classic pedigree breeding methods within segregating populations for numerous generations is the most common strategy for inbred development in maize [25]. Recurrent selection is another method for improving the breeding population mean by recombining superior progeny after selection [26]. However, using these approaches takes longer to obtain the appropriate amount of homozygosity. In comparison to other approaches, the DH method may achieve homozygosity in a single generation (Figure 5). As a result, DH production would be a feasible alternative to conventional methods for rapidly generating homozygous lines [3]. Because they are 100 percent homozygous, DH lines meet all of the DUS (Distinctness, Uniformity, and Stability) requirements for varietal development [1]. The DH population may also be used to gain knowledge on the genetic architecture of complex characteristics through breeding. Because the DH population is made up entirely of additive genetic variation due to homozygosity at all loci, the selection response is substantially higher than in other segregating populations [27, 28]. Additionally, DH breeding can be used with a marker-assisted backcrossing program to transfer the favorable allele of the concerned trait through either phenotypic or marker-assisted procedures, or a mix of both, by omitting the self-pollination stages at the end of the program [29]. Many commercial breeding programs have recently combined DH technology with genomic selection to improve genetic gain, especially for characteristics governed by a large number of QTL with low heritability [30]. Individual haploid plants are genotyped to find superior haploids, followed by self-pollination to establish homozygous lines using genomic selection [1]. Recently, after genetic alterations of three essential genes involved in meiotic recombination (REC8, PAIR1, and OSD1 genes) and a single gene involved in haploid induction (MTL), parthenogenesis was designed to fix heterosis [31, 32, 33, 34]. Therefore, there are plenty of opportunities to combine these strategies in order to boost breeding yield.

Figure 5.

Chromosome constitution of various stages during homozygous lines development in a single generation.

6. Conclusion

In summary, DH technology has revolutionized commercial maize breeding programs by offering economic viability. This technology can also be integrated with the other major crops of economic importance to fast-track their breeding program. Integration of recent biotechnological approaches in the DH program further enhances the output of the breeding cycle per unit of time.

Acknowledgments

The participant is also thankful to the Human Resource Development Group (HRDG) division of the Council of Scientific & Industrial Research (CSIR), New Delhi, India for the Junior Research Fellowship (File No.: 09/083(0383)/2019-EMR-I) to pursue his Ph.D. program. The participant also shows his sincere gratitude to Vignesh Muthusamy, Rajkumar U Zunjare, and Firoz Hossain for correcting the manuscript.

Conflict of interest

The authors declare no conflict of interest.