Haploid differentiation in maize kernels based on fluorescence imaging

Abstract

A new fluorescence-based method for inbred haploid differentiation in maize kernels was developed by utilizing the R1-nj colour marker in combination with fluorescence microspectroscopy and imaging. Seven inbred lines with varying R1-nj expression were used in this study. The fluorescence response of the diploid kernels at the embryonic dye spot was shown to simultaneously exhibit lower intensity and occur at a higher wavelength than the fluorescence of the dye-lacking haploid embryos. Intensity and area thresholds were applied to fluorescence images to sort the haploids from mixed sample populations, and sorting efficiencies of greater than 80% were achieved in all seven inbred lines (with values greater than 90% for five lines). The potential for highthroughput sorting when fluorescence imaging is combined with existing technologies for seed handling as well as high sorting efficiency may make fluorescence a viable and promising alternative to current sorting methods for some inbred lines.Key words: maize -haploid -diploid -fluorescencesortingThe development of new hybrids of maize (Zea mays L.) with desirable traits (e.g. high yield, pest tolerance) is of considerable importance to the global economy, with the 2014 worldwide production estimated at 1.02 billion metric tons (FAO 2014). In modern commercial maize breeding programmes, new hybrids are generated by crossing homozygous inbred lines that have been produced via doubled haploid (DH) line development (Murigneux et al. 1993, R€ ober et al. 2005, Prasanna 2012). Traditional inbred line development via manual self-pollination requires 6-8 generations to produce a highly homozygous line, whereas DH line development only requires 2-3 generations for a 100% homozygous line. This reduction in line development time allows breeders not only to produce more inbred lines in less time, but also to react to new selection targets quickly. In the DH process, the haploid kernels need to be collected from the mixed population as only the haploids are suitable for the next generation of inbreeding.Most methods for haploid kernel selection rely on pioneering work by Nanda and Chase, who used purple embryo marker (PEM) stock as the male parent to produce the anthocyanin marker R1-Navajo (R1-nj) in the progeny kernels (Nanda and Chase 1966). The dominantly inherited R1-nj marker causes a xenia effect resulting in pigmentation of the embryo and on the cap of the aleurone of the seed. As is the case in many angiosperms, sexual reproduction occurs through the double fertilization process where fusion of one sperm and the egg constitutes the union of genetic information that produces the diploid embryo, while union of the second sperm and the central cells develops into triploid endosperm. The DH system takes advantage of the phenomenon of haploid induction. In this case, it is known as maternal haploid induction because haploid embryos contain the cytoplasm of the female (donor) parent and the inducer line is the male in the cross. On average, 10% of double fertilizati...

A Diallel Analysis of a Maize Donor Population Response to In Vivo Maternal Haploid Induction: I. Inducibility

“…The effective and efficient use of the maize in vivo maternal DH system relies on the ability to efficiently produce, sort, and effectively identify haploid progeny. There are multiple avenues to increase the efficiency of this system: (i) development and use of improved inducer lines (Coe, 1959;Rober et al, 2005;Geiger, 2009;Prigge et al, 2011), (ii) improved methods of haploid selection through new traits and/or automation (Jones et al, 2012;Melchinger et al, 2015;Smelser et al, 2015;Boote et al, 2016;De La Fuente et al, 2017), and (iii) the method considered in this and other studies (Kebede et al, 2011;Prigge et al, 2011), namely improvement of rates of induction in donor materials.…”

“…Prigge et al, 2011), (ii) improved methods of haploid selection through new traits and/or automation(Jones et al, 2012;Melchinger et al, 2015;Smelser et al, 2015;Boote et al, 2016;De La Fuente et al, 2017), and (iii) the method considered in this and other studies(Kebede et al, 2011;Prigge et al, 2011), namely improvement of rates of induction in donor materials.…”

Discrimination of haploid and diploid maize kernels via multispectral imaging

“…The method described by Boote et al. () provides the most similar method, using fluorescence imaging. However, this method only considers one kernel at a time, and while discussion of automation was provided, no current method exists for that system, while for the VideometerLab 3 system a method of automated seed feeding exists (shown here: ftp://videometerlab:multispec@www.videometer.com/Videos/2014_July_VideometerLab_AutoFeeder.AVI).…”

“…Although the method in Jones et al (2012) provided a discriminative sort, there was no validation of the true identity of the haploid/hybrid kernels. The method described by Boote et al (2016) provides the most similar method, using fluorescence imaging. However, this method only considers one kernel at a time, and while discussion of automation was provided, no current method exists for that system, while for the VideometerLab 3 system a method of automated seed feeding exists (shown here: ftp://videometerlab:multispec@www.videometer.…”

“…Although no fully implemented system is being commercially used, several other pilot studies have been published using other markers to discriminate between the haploid and diploid fractions using instrumentation. Traits such as the difference between the embryo weights of the haploid and diploid seed (Smelser et al 2015), spectral differences using NIRS and SIMCA (Jones et al 2012), fluorescence imaging (Boote et al 2016) and oil content (Melchinger et al 2015) are all being tested for their utility as automated selection criteria. Each method has its strengths and drawbacks.…”