Mutations in the dwarf3 gene confer height stability in sorghum

Dwarfism is a useful trait in many crop plants because it contributes to improved lodging resistance and harvest index. The mutant allele dw3‐ref (dwarf3‐reference) of sorghum [Sorghum bicolor (L.) Moench] is characterized by an 882 bp tandem duplication in the fifth exon of the gene that is unstable and reverts to wild‐type at a frequency greater than 0.001 in many genetic backgrounds. The goal of this research was to identify stable alleles of dw3 (dwarf3) that could be backcrossed into elite parent lines to improve height stability of the crop. To discover new alleles of dw3, a panel consisting mostly of sorghum conversion lines (SC‐lines) was screened by polymerase chain reaction for the 882 bp tandem duplication in the fifth exon of dw3‐ref. Sanger sequencing was used to characterize the DNA sequence of this fragment in genotypes that did not contain the 882 bp tandem duplication. Sequence analysis identified three indel mutations, including an 82 bp deletion, a 6 bp duplication, and a 15 bp deletion in this region of the gene. Field trials of the donor genotypes with these new alleles indicated no wild‐type revertants of dw3‐sd3 (dwarf3‐stable dwarf), dw3‐sd4, and dw3‐sd5. These alleles were backcrossed into Tx430. Field trials of backcross progeny (BC2F4) with the dw3‐sd3, dw3‐sd4, and dw3‐sd5 alleles indicated no revertants. The plant height and flowering time characteristics of the backcross progeny were similar or slightly shorter and earlier than the recurrent parent. These findings demonstrate that dw3‐sd3, dw3‐sd4, and dw3‐sd5 alleles will be useful in breeding for the stable dwarf trait.

height stability of the crop.To discover new alleles of dw 3 , a sorghum diversity panel was screened by DNA sequencing to discover genotypes with different mutations in this gene.One insertion and two deletion mutations were identified and characterized.Field trials of genotypes with these new alleles indicated stable dwarf phenotypes.These alleles were backcrossed into Tx430 and derived backcross progeny (BC 2 F 4 ) were shown to be dwarf stable.The plant height and flowering time characteristics of the backcross progeny were similar to the recurrent parent.

INTRODUCTION
Dwarfism is a useful trait in sorghum [Sorghum bicolor (L. ) Moench] because it enables the use of mechanical harvesting, improves lodging resistance, and enables increased fertilizer-use efficiency (Hedden, 2003;Khush, 2001).The dw 3 (dwarf3) gene of sorghum and br2 (brachytic2) gene of maize (Zea mays L.) are orthologous.Previous studies have shown that the dwarf phenotypes associated with the recessive dw 3 and br2 alleles are caused by mutations in a P-glycoprotein (PGP), which functions as a polar auxin transporter (Multani et al., 2003).PGPs are encoded by multidrug-resistant genes and are classified as adenosine triphosphate-binding cassette (ABC) transporters.These proteins transport hydrophobic molecules out of the cell (Noh et al., 2001).Overexpression experiments in Arabidopsis thaliana AtPGP1 resulted in longer hypocotyls in plants with gene overexpression and shorter hypocotyls when antisense DNA was over-expressed (Noh et al., 2001;Sidler et al., 1998).These dwarf phenotypes were characterized by shorter internode lengths than wild-type plants, caused by the lack of auxin stimulus required for cell elongation (Karper & Quinby, 1947;Multani et al., 2003).The mutation in dw 3 -ref (dwarf 3 -reference) of sorghum is an 882 bp tandem duplication in exon 5, which causes the protein to be nonfunctional (Multani et al., 2003).This allele is known to be unstable with progeny of a homozygous dw 3 -ref plant reverting to Dw 3 tall, wild-type at a rate of around 1/600 plants, depending on the genetic background of the accession (Karper, 1932).Karper (1932) hypothesized that the reversion phenotype was caused by a mutation during gametogenesis, which Multani et al. (2003) confirmed as an unequal crossing-over event during meiosis.In dw 3ref sorghum, the gamete with the gene losing the duplication becomes Dw 3 wild type and exhibits a tall phenotype since the encoded auxin transporter can properly function (Multani et al., 2003).
The presence of height mutants in sorghum seed production is often managed through rouging, which increases labor and production costs and is not always reliable because environmental conditions such as drought can reduce plant height and make it difficult to differentiate height mutants from dwarf genotypes.Stable dwarfism ensures homogenous plant height and simplified mechanized harvesting.The development of stable height sorghum varieties or hybrids represents a solution to the problem.
A stable dwarf allele, dw 3 -sd2 (dwarf 3 -stable dwarf), was previously identified in the sorghum line Tx2737, whose stable dwarf phenotype is due to a 6 bp deletion, which caused a two amino acid deletion in the translated protein (Barrero Farfan et al., 2012).This discovery pointed to the potential existence of different dwarf mutations in the dw 3 gene, some of which could be additional stable sources of the trait.In this study, a panel of dwarf-stature sorghum accessions representing the genetic diversity of the crop was screened for novel dw 3 mutations.New dw 3 alleles with small insertions or deletions (indels) were identified and characterized for stability and impacts on plant height and maturity.

Sorghum diversity panel
A sorghum diversity panel consisting of 300 sorghum breeding lines and sorghum conversion lines (SC-lines) described by Casa et al. (2008) was screened by polymerase chain reaction (PCR) to discover novel dw 3 alleles.Fresh tissue samples were collected from each accession in the sorghum diversity panel by sampling a 5 cm segment of young leaf tissue in a 1.5 mL tube.The samples were processed by adding liquid nitrogen to each tube and grinding the leaf tissue with a small pestle to obtain a fine powder.DNA was extracted from each sample using a high throughput cetyltrimethylammonium bromide (CTAB) DNA extraction protocol (Dietrich et al., 2002).The CTAB buffer was made of 100 mM Tris pH 7.5, 0.7 M NaCl, 10 mM ethylenediaminetetraacetic acid (EDTA), 1% CTAB, and 0.17% beta-mercaptoethanol (BME).First, 500 μL pre-warmed buffer at 65˚C was added to each sample tube, mixed, and incubated at 65˚C for 30 min in a shaking water bath.The samples were thoroughly mixed during the incubation period, extracted with 500 μL of 24:1 chloroform/octanol mixture, and centrifuged at 14,000 × g for 10 min.A volume of 350 μL of the aqueous layer was transferred to a new tube, and 350 μL of cold (−20˚C) isopropyl alcohol was added.After mixing gently to precipitate DNA, the samples were spun for 10 min at 14,000 × g to pellet the DNA.The pellets were washed with 500 μL 70% ethanol.After discarding the ethanol, the tubes were briefly centrifuged, and the residual ethanol was removed using a pipette.Pellets were rehydrated in 100 μL double-distilled water and incubated on a heat block at 75˚C for about 10 min to destroy residual DNase activity.
To extract DNA of the backcross progeny, a modified protocol combining some steps of the CTAB (Allen et al., 2006) and the mixed alkyl trimethyl ammonium bromide (MATAB) (Risterucci et al., 2000) procedure was used.Leaf samples were ground to a fine powder using a Qiagen tissue lyser at 30 m/s, to which 750 μL of extraction buffer solution (1.2 M NaCl, 100 mM Tris pH 8, 20 mM EDTA pH 8, 2% CTAB, and 0.1% BME) was added, followed by 30-min incubation at 65˚C in a water bath with gentle vortexing every 5 min.To each tube was added 750 μL of 24:1 chloroform:isoamyl alcohol.After mixing by continuous hand shaking for 1 min, the product was centrifuged at 13,000 × g for 20 min, then 600 μL of the supernatant was transferred to a new 1.5 mL Eppendorf tube and 600 μL of cold (−20˚C) isopropyl alcohol was added to each tube.Tubes were gently shaken for 1 min, stored at −20˚C for 2 h, centrifuged at 13,000 × g for 20 min, and the supernatant removed.The pellet was washed once in 500 μL of cold (−20˚C) 70% ethanol and air dried.DNA was resuspended in HPLC grade water.
DNA concentration and quality were checked on a Thermo-Scientific Nanodrop 1000 Spectrophotometer, and samples with A 260/280 above 1.8 were considered pure and diluted to 50 ng/μL prior to performing the PCR.

PCR screening for novel dw 3 alleles
The panel of 300 sorghum accessions was screened using PCR to amplify the 882 bp tandem duplication in dw 3 .The sorghum reference genome sequence obtained from Phytozome (Goodstein et al., 2012) was used to design the primers.The forward and reverse primers were labeled 4F (5′-CGTCCTGCAGAAGATGTTCATGAAGG-3′) and 5R (5′-GTGCGCCACCACGATGGTGGTGC-3′), respectively (Barrero Farfan et al., 2012).PCR was performed in a 25 μL reaction containing 1X MangoMix (Bioline USA Inc.), 0.4 μM of each primer, 0.125 μg of DNA, 1.25 μL dimethyl sulfoxide, and double-distilled water.A PCR protocol with a touchdown step (Don et al., 1991) was used to amplify the locus, including an initial denaturation cycle at 95˚C for 3 min, followed by 40 cycles of denaturation at 95˚C for 45 s, annealing for 55 s, and one cycle of extension at 72˚C for 2.5 min.

Core Ideas
• Dwarfism is a useful trait in many crop plants including sorghum.• Sorghum dwarf stability is linked to naturally occurring mutations in the dwarf3 gene.• The new dwarf3 alleles dw 3 -sd3, dw 3 -sd4, and dw 3 -sd5 confer height stability to sorghum plants.
The annealing temperature was set at 68˚C for 10 cycles, stepping down 1˚C every five cycles until reaching 65˚C for the final 35 cycles.The PCR reactions were performed on a Bio-Rad C1000 Thermal Cycler (Bio-Rad Laboratories, Inc.), and the results were visualized on agarose gel made with 1% TAE buffer.

DNA sequencing and sequence analysis
The amplified DNA samples from genotypes that did not include the 882 bp tandem duplication were purified using the ExoSAP-IT Express PCR product cleanup reagent using the manufacturer's instructions (ThermoFisher.com).Sanger sequencing was performed at Azenta Life Sciences using the 4F primer (5′-CGTCCTGCAGAAGATGTTCATGAAGG-3′) (Barrero Farfan et al., 2012).Multiple sequence alignment of the samples was done with the dw 3 reference sequence of BTx623, Sobic.007G163800.1,via ClustalW multiple sequence alignment tool of BioEdit (Hall, 1999).

Dwarf allele introgression
Tx430 carries the unstable dw 3 -ref allele that can revert to wild-type.A wild-type revertant of Tx430 (Dw 3 ), genetically similar to Tx430 dw 3 -ref except for the mutation, was identified and used as a male parent in developing the backcross populations (Figure 1).Tx430 Dw 3 was crossed as a male to sorghum lines carrying unique dw 3 alleles.The F1 progeny from each cross were self-pollinated, and the F 2 seeds were planted in plots to verify 3:1 segregation for the Dw 3 allele.The short plants from each row carrying the dw 3 alleles were backcrossed to Tx430 Dw 3 , and the resulting BC 1 F 1 plants were self-pollinated to obtain BC 1 F 2 populations.Short plants in each BC 1 F 2 family were backcrossed to Tx430 Dw 3 and the resulting BC 2 F 1 plants were selfpollinated to obtain BC 2 F 2 families.The dw 3 plants in each family were self-pollinated for two generations to generate an array of BC 2 F 4 progeny for each of the new dw 3 alleles (Figure 1).

Stability performance of the new dwarf3 alleles
The stability of each new dw 3 allele was assessed in field trials at the Purdue Agronomy Center for Research and Education (ACRE) in 2020.Each of the donor lines and the recurrent parent (Tx430, SC124, SC134, and SC991) were planted in an experiment using a randomized complete block design with four replicates.Each plot was planted as a 91.4-m single-row plot with 0.76 m between plots.The number of plants for each plot was recorded and the frequency of Dw3 plants was calculated as the ratio of tall plants over the total number of plants recorded for the genotype.Dw3 plants are visually significantly taller than the dwarf dw3 entries, hence they are easily identified in the field.
The new dw 3 alleles from SC124, SC134, and SC991 were backcrossed into Tx430.Nine backcross progeny were advanced to BC 2 F 4 for each new allele and assessed in field trials at the ACRE in 2020 using a randomized complete block design with three replicates.Each plot was planted as a 91.4m single-row plot.The number of plants for each plot was recorded, and the frequency of Dw3 plants was calculated as the ratio of tall plants over the total number of plants recorded for the genotype.Based on the overall performance of the backcross progeny in 2020, the 10 best BC 2 F 5 lines were selected and planted at ACRE in 2022 along with their parents.Each BC 2 F 5 line was planted in two replications with each replication planted as a 3.04-m single-row plot with 0.76 m between plots.Five random plants from each plot were randomly tagged to determine the average flowering time based on days from planting to 50% flowering and plant height measured as apex height.

Allele screening and description of the new dwarf alleles
PCR analyses were used to discover new dw 3 alleles in the sorghum diversity panel.The PCR product was 2006 bp long when the exon contained the tandem duplication (dw 3 -ref) and 1124 bp long when the product contained no duplication (Figure 2).PCR products examined on 1% agarose gel revealed that 181 of the 300 accessions had a 2 kb band, which were assumed to be genotypes containing the dw 3 -ref allele with the 882 bp duplication (Figure 2).All other lines exhibited a roughly 1 kb band that could indicate either wild-type Dw 3 plants or novel dw 3 alleles.Therefore, all the samples with the 1 kb band were Sanger sequenced.
Sanger sequencing and sequence alignment of the accessions that had a 1 kb PCR fragment confirmed that those genotypes did not have the 882 bp duplication in the fifth exon; however, many accessions carried additional mutations (Figure 3).Unique indels were identified in the population, including one insertion and two deletions (Table 1).The newly identified alleles are dw 3 -sd3 (82 bp deletion), dw 3 -sd4 (6 bp duplication), and dw 3 -sd5 (15 bp deletion) (Table 1).

Stability performance
Field trials conducted to examine the stability of the three new dw 3 alleles revealed that dw 3 -sd3 (82 bp deletion), dw 3 -sd4 (6 bp duplication), and dw 3 -sd5 (15 bp deletion) were stable (Figure 4), while dw 3 -ref exhibited a reversion rate for plant height under field conditions on average one in 300 plants (Table 2; Table S1).The backcross conversions of dw 3 -sd3, dw 3 -sd4, and dw 3 -sd5 in the Tx430 genetic background were also shown to be stable with no mutants observed in 14,601, 16,685, and 14,822 plants, respectively (Table 3; Table S2).
The average flowering times of selected BC 2 F 5 progeny were between 69.6 and 71.4 days with individual progeny being similar or slightly earlier than Tx430 (Table 4; Table S3).
The average plant heights of BC 2 F 5 progeny were between 102 and 112 cm with individual progeny being similar to or slightly shorter than Tx430 (Table 4).

Dwarfism is a useful trait in sorghum
Selective breeding for advantageous plant traits can increase their value by improving the efficiency of how they are grown and harvested.In the case of sorghum, dwarf plants are useful because they generally exhibit improved resistance to lodging and are easier to mechanically harvest (Karper & Quinby, 1947).Due to the potential value of the dwarfing trait, it is useful to understand the natural diversity of the dw 3 locus.Identifying multiple sources of a desirable allele could be useful for introgression projects since the linkage drag from different genetic backgrounds may improve heterosis when derived lines are used in hybrid seed production.A survey of the diversity of a trait at a given locus can offer insights into past breeding methods and may reveal important gene regions, as indicated by high sequence conservation.

4.2
Sorghum dwarf height stability is linked to naturally occurring mutations Multani et al. (2003) noted that new dwarfing alleles could be used to address the instability of the tandem duplication of dw3-ref in sorghum.In this research, three new, stable dwarfing mutations were discovered in the dw 3 gene.Deletions and insertions have good potential for use in developing new stable dwarf phenotypes due to their likely disruption of protein function.The goal of this research was to discover new stable dw 3 alleles in diverse genetic backgrounds to improve the diversity of this region of Chromosome 7 in sorghum.
Novel dw 3 alleles containing insertions/deletions were identified in the sorghum diversity panel.The newly identified indels included an 82 bp deletion discovered in SC124, a 6 bp insertion in SC134, and a 15 bp deletion in SC991 for which the alleles were named dw 3 -sd3, dw 3 -sd4, and dw 3 -sd5, respectively.The 82 bp deletion (dw 3 -sd3) caused a frame shift in the DNA code, which led to the introduction of an early stop codon at position 963 in the peptide chain of SC124 resulting in the rest of the fifth exon not being translated.The 6 bp insertion (dw 3 -sd4) led to the amino acids A1036 and L1037 added to the peptide chain without affecting the reading frame.The 15 bp deletion in dw 3 -sd5 led to five amino acids being deleted from the peptide chain with no additional disruptions observed in the amino acid sequence.

Dwarf stability
Field trials demonstrated that dw 3 -sd3, dw 3 -sd4, and dw 3 -sd5 alleles were dwarf-stable.A study of backcross progeny with more than 14,000 plants representing each of these new alleles demonstrated that no revertant plants were found in any backcross lines that contained the dw 3 -sd3, dw 3 -sd4, and  (2012).
The phenological characteristics of the backcross progeny indicated that many of the lines were similar or slightly earlier in flowering time and similar or shorter in plant height compared to the recurrent parent Tx430.Given the importance of early maturity and short plant stature in sorghum production, this is a positive outcome that may add value to these lines as trait-donors (Karper & Quinby, 1947).
Utilizing dwarf-stable parent lines in seed production is desirable because of the reduced need for rouging in seed production fields.Rouging describes the process of manually removing unwanted height mutants and other off-types from fields prior to pollen shed.When unfavorable weather or staffing challenges cause delays in rouging, any tall mutant plants can shed wild-type pollen and substantially increase the frequency of off-types in the next generation.If the frequency of height mutants is high enough, these seed lots may be discarded, thereby increasing the cost of these products.The commercialization of dwarf-stable parent lines and hybrids greatly reduces this risk.Farmers also prefer the appearance of dwarf stable sorghum hybrids because these products are more uniform in production fields.

Concluding remarks
These studies suggest that dw 3 -sd3, dw 3 -sd4, and dw 3 -sd5 will be useful in breeding for a stable dwarf phenotype without negatively altering other agronomic characteristics.

D A T A AVA I L A B I L I T Y S T A T E M E N T
Field trial data for plant height stability and flowering time of donor and recurrent parents and backcross progenies are made available in Tables S1-S3.

F
I G U R E 2 PCR products from a screen of Sobic.007G163800(dwarf3) in the sorghum diversity panel.The PCR fragments were visualized in a 1% agarose gel stained with ethidium bromide.The genotypes carrying the dw 3 reference allele produced PCR products (2006 bp) that are nearly two times longer than the WT (1124 bp) or new dw 3 alleles without the tandem duplication (1042, 1109, and 1130 bp).

F
I G U R E 3 (A) Gene scheme of Sobic.007G163800showing the insertions and deletions found in the sorghum diversity panel.The indels are located on exon 5 from Sorghum bicolor v3.1.1.(B) Sequence alignment of the region of exon 5 of Sobic.007G163800amplified from SC124, SC134, and SC991 compared to the reference genome, BTx623.The positions of the dwarf3 indels are shown in red for the dw 3 -sd3 allele (82 bp deletion), dw 3 -sd4 allele (6 bp duplication), and dw 3 -sd5 allele (15 bp deletion).Positions do not include the untranslated regions (UTRs).
dw 3 -sd5 alleles, whereas the check with the reference allele (Tx430 dw3-ref) had a revertant rate of about one in 270 plants, similar to what was reported by Barrero Farfan et al.
Description of new dwarf3 alleles discovered in the sorghum diversity panel.Mutation previously discovered.All sequence positions are derived from the genomic sequence of Sobic.007G163800 from version 3.1.1 of Sorghum bicolor which carries the 882 bp tandem duplication.Plant height stability of indel donor lines.Plant height stability of backcross progeny.
T A B L E 1 a T A B L E 2 Variation in flowering time and plant height among backcross progeny with dw 3 -sd3, dw 3 -sd4, and dw 3 -sd5.