Physical mapping of the wheat genes in low-recombination regions: radiation hybrid mapping of the C-locus

This work reports the physical mapping of an important gene affecting spike compactness located in a low-recombination region of hexaploid wheat. This work paves the way for the eventual isolation and characterization of the factor involved but also opens up possibilities to use this approach to precisely map other wheat genes located on proximal parts of wheat chromosomes that show highly reduced recombination. Mapping wheat genes, in the centromeric and pericentromeric regions (~ 2/3rd of a given chromosome), poses a formidable challenge due to highly suppressed recombination. Using an example of compact spike locus (C-locus), this study provides an approach to precisely map wheat genes in the pericentromeric and centromeric regions that house ~ 30% of wheat genes. In club-wheat, spike compactness is controlled by the dominant C-locus, but previous efforts have failed to localize it, on a particular arm of chromosome 2D. We integrated radiation hybrid (RH) and high-resolution genetic mapping to locate C-locus on the short arm of chromosome 2D. Flanking markers of the C-locus span a physical distance of 11.0 Mb (231.0–242 Mb interval) and contain only 11 high-confidence annotated genes. This work demonstrates the value of this integrated strategy in mapping dominant genes in the low-recombination regions of the wheat genome. A comparison of the mapping resolutions of the RH and genetic maps using common anchored markers indicated that the RH map provides ~ 9 times better resolution that the genetic map even with much smaller population size. This study provides a broadly applicable approach to fine map wheat genes in regions of suppressed recombination.


Introduction
Hexaploid bread wheat is one of the most important crops that feed the growing human population (Shiferaw et al. 2013;Shewry and Hey 2015). It is one of the most cultivated crop plants and is ranked number one in annual consumption worldwide. However, changing climatic conditions and increasing human populations are posing an inevitable challenge for global wheat production (Miraglia et al. 2009;Godfray et al. 2010;Asseng et al. 2011;Enghiad et al. 2017;Ficke et al. 2017). To enhance the worldwide output of wheat, it is crucial to uphold the sustainability of agricultural production methods (Godfray et al. 2010;Hawkesford et al. 2013).
Numerous measures and technologies have been adopted and are being implemented to promote the sustainability of agricultural production systems, particularly for crop plants (Akpınar et al. 2013;Whitford et al. 2013;Dwivedi et al. 2017;Chaudhary et al. 2018;Borrill et al. 2019;Rasheed Communicated by Peter Langridge. 159 Page 2 of 14 and Xia 2019; Tadesse et al. 2019;Wu et al. 2020). Examples of such initiatives include the integration of novel genes and alleles with agronomic significance through genetic improvement, as well as the incorporation of useful traits related to yield and stress tolerance against biotic and abiotic factors into cultivars (Mujeeb-Kazi and Rajaram 2002;Alonso and Ecker 2006;Gill et al. 2011;Akpınar et al. 2013;Longin and Reif 2014;Tadesse et al. 2019;Bailey-Serres et al. 2019;Wu et al. 2020).
Traditional breeding and marker-assisted selection-based approaches are commonly used in crop breeding programs worldwide to transfer genes, with genetic recombination playing a vital role in the process of genetic transfer and mapping studies (Tanksley and McCouch 1997;Varshney et al. 2005;Rawat et al. 2009;Gupta et al. 2010;Wulff and Dhugga 2018;Rasheed and Xia 2019;Tadesse et al. 2019). By combining donor and recipient chromosomes, researchers can identify the relative position of new genes or QTL and their alleles and use linked or flanking markers to facilitate their deployment in crop plants like wheat.
Hexaploid bread wheat has a complex genome structure in terms of its size (> 14 Gb), genome organization, content, and polyploid nature (Consortium (IWGSC) et al. 2018;Ramírez-González et al. 2018). Wheat chromosomes show much higher gene content along their distal parts (toward the telomere) than the proximal parts (centromeric and pericentromeric regions). Furthermore, the distribution of the recombination frequency on wheat chromosomes is highly uneven (Saintenac et al. 2009;Consortium (IWGSC) et al. 2018;Ramírez-González et al. 2018); around 30% of the wheat genes are housed in low-recombination regions of the chromosomes which are about 2/3rd part of a given wheat chromosome (Saintenac et al. 2009;Consortium (IWGSC) et al. 2018). Due to very limited recombination events in the proximal part of the wheat chromosomes, precise mapping, fine mapping, and map-based cloning studies of agronomically important genes are primarily restricted to the telomeric or sub-telomeric regions of the wheat chromosomes (Krattinger et al. 2009).
Mapping wheat genes in the pericentromeric or centromeric region based on genetic mapping poses a big challenge (Saintenac et al. 2009;Philippe et al. 2013) and requires alternative approaches (Mahlandt et al. 2021). A recombination-independent approach known as radiation hybrid (RH) mapping has been developed in human genetics and was explored in both animal and plant genetics Harris 1975, 1977;Riera-Lizarazu et al. 2000Wardrop et al. 2002). This method uses gamma irradiationinduced chromosomal breaks to order marker scaffolds in a given RH panel Harris 1975, 1977;Riera-Lizarazu et al. 2000Wardrop et al. 2002). In plants, RH mapping was first explored to map maize chromosome 9 (Riera-Lizarazu et al. 2000) and was subsequently applied in barley (Wardrop et al. 2002) and wheat (Kalavacharla et al. 2006;Riera-Lizarazu et al. 2010;Tiwari et al. 2012aTiwari et al. , 2016. In our previous studies, we successfully implemented RH mapping to generate D-genome and then whole genome RH maps for the wheat Consortium (IWGSC) et al. 2018). The RH mapping panels developed in wheat can also provide an alternative for gene mapping studies, particularly where genetic mapping is challenging due to lower rates of recombination. Since RH mapping relies on radiation-induced breaks and on their deletion and retention patterns, a marker scaffold can be ordered as well as used to locate a dominant phenotype (Bassi et al. 2013;Mahlandt et al. 2021). Pollen irradiation-based RH panels in wheat can be used to map dominant loci in wheat irrespective of their location in any part of the wheat chromosome (Tiwari et al. 2012a;Mahlandt et al. 2021).
The wheat inflorescence is represented by a spike with several spikelets on an axis with each spikelet bearing multiple florets (flowers) (Koppolu and Schnurbusch 2019). In cereal crops, floral architecture and spike morphology contribute to yield and bear grains, the most important part of the plant for human consumption (Ma et al. 2018;Koppolu and Schnurbusch 2019). Spike length, number of spikelets per spike, number of grains per spike, and grain weight per spike are all traits included in the yield component definitions in the cereals crop plants (Guo et al. 2018). There are several genes and factors which affect the spike morphology and related traits in cereals (Koppolu and Schnurbusch 2019), but only a few of them have been characterized at the molecular level and these examples include the squarehead, Q, and sphaerococcum, S, loci (Faris et al. 2003;Cheng et al. 2020).
A dominant locus mapped on chromosome 2D of wheat is responsible for the compact spike (C-locus) that regulates spike length and compactness in hexaploid wheat (Johnson et al. 2008;Yu et al. 2022;Wen et al. 2022) and defines a subspecies (Triticum aestivum ssp. compactum) of hexaploid wheat (Mac Key 1954) also called club wheat (Nilsson Ehle 1911; Rao et al. 1972;Johnson et al. 2008). It remains uncharacterized at the molecular level, mainly due to its location in a low-recombination region. The C-locus has a pleiotropic effect on many traits as it affects spike morphology, grain size, shape, number, and perhaps other aspects of plant development and directly or indirectly affects the agronomic performance of wheat such as height (Johnson et al. 2008;Wen et al. 2022). There is a keen interest in understanding this trait and its relationship with other spikecompacting genes in wheat and related germplasm.
A limited number of mapping studies were conducted on this trait and there is no consensus about the placement of the C-locus (Johnson et al. 2008;Yu et al. 2022;Wen et al. 2022). Johnson et al. 2008 used two genetic mapping populations and placed the C-locus close to the centromeric region of chromosome 2D. However, due to suppressed recombination events in this region, the study was not able to place the C-locus concerning the centromere. This locus is a good candidate to test the efficiency of mapping dominant genes using RH mapping in chromosomal regions with very limited recombination. This study presents an integrated approach by combining genetic mapping and RH mapping to precisely map C-locus in hexaploid wheat and its localization on the physical map chromosome 2D of the wheat reference genome.

Plant material
The hexaploid wheat variety Corrigin (Triticum aestivum ssp. compactum; 2n = 6x = 42; AABBDD) with a standard club phenotype was used as the male parent for the development of an RH panel for the C-locus (Rosielle et al. 1991). A tetraploid cultivar 'Langdon' (Triticum turgidum; 2n = 28; AABB) with a laxed spike was used as the female parent for the RH crosses.
For developing the genetic mapping population, the club cultivar, Corrigin was used as a female parent, and synthetic hexaploid wheat TA8051 as the male parent (Mahlandt et al. 2021).

Development of club-wheat radiation hybrid (club-wheat RH) mapping panel
To generate an irradiated pollen panel for the C-locus, the hexaploid wheat club cultivar 'Corrigin' was grown to flowering, and dehiscent wheat spikes were excised from the plants with stems kept in water.
The spikes were irradiated with γ-rays and pollen from the irradiated spikes was immediately used to pollinate already emasculated spikes of tetraploid wheat Langdon (Riera-Lizarazu et al. 2010;Tiwari et al. 2012b). All radiation experiments used a Gamma Cell 220 irradiator at the Radiation Center of Oregon State University, USA. A radiation dose curve using treatments of 0-, 10-, 15-, and 20-Gy irradiated pollen samples was performed (Fig. 1a). We selected 15 Gy dosage treatment to develop the club-RH panel (Fig. 1a). Method described by Mahlandt et al (2021) was used to develop club-RH panel using optimized 15 Gy treatment (Fig. 1b).  Schematic presentation of the club-wheat RH mapping panel development-(i) In the first step, a continuous supply of wheat spikes was obtained by growing club-wheat seeds in the greenhouse. (ii) In the next step, freshly dehiscing pollen grains were irradiated with optimized dosages. (iii) In the third step, irradiated pollen was used to pollinate already emasculated tetraploid wheat spikes, and viable seeds were harvested 20 days after pollination. (iv) Greenhouse planting of the club-wheat RH 1 seeds for genotyping and phenotyping of the club-wheat RH 1 plants. (v) club-RH panel combining genotyping and phenotyping data sets to precisely map C-locus

Development of a genetic mapping population
The club cultivar Corrigin was used as a female parent and synthetic hexaploid wheat (SHW) line TA8051 [Prelude tetraploid/Aegilops tauschii (TA1604)] was used as the male line for the initial cross ( Fig. 2d and e) to generate F 1 plants. Corrigin × Synthetic F 1 s were selfed to generate an F 2 population of 1000 plants (2000 gametes). A set of 702 F 2 individuals (1404 gametes) were used to genetically map the C-locus. Leaf tissue from this F 2 population was collected for DNA extraction, and the lines were phenotyped visually for compact or not-compact phenotypes.

DNA extraction, marker development, and genotyping assays
DNA extraction from 15-Gy pollen plant lines and genetic mapping population, parental lines, and genetic stocks followed the method described by Mahlandt et al. 2021. For genome-specific PCR-based marker development, the Chinese Spring reference genome of hexaploid bread wheat (IWGSC 2018) was used to obtain gene sequences for the A, B, and D sub-genome homologs.
Homologous sequences were then used to develop genome-specific primers using the online GSP tool (Wang et al. 2016). Genome specificity of the developed primers was confirmed by parental testing of all markers on nullisomic-tetrasomic lines as well as wheat deletion bin lines lacking the chromosome 2D, and specific chromosomal regions on chromosome 2D, respectively (Sears 1966;Endo and Gill 1996). All the PCRs and genotyping work followed the protocols described by Mahlandt et al. 2021. A list of all the developed and mapped 2D-specific markers and their sequences are provided in supplementary table 1 (ST. 1).

Initial characterization of club-wheat RH panel
For the initial characterization of the club-RH panel, DNA was extracted from a set of 282 15-Gy club-RH 1 s. DNA was quantified using a NanoDrop ND-1000 UV-Vis Spectrophotometer (Thermo Scientific), and subsets of samples was diluted to 20 ng/μl. Initial assessment of the 15-Gy club-RH panel for genome-wide radiation-induced deletions was done as described in Mahlandt et al. 2021 using a set of 14 SSR markers from all d-genome chromosomes (two per chromosome).
Polymerase chain reaction assays and the markers used were as described by Tiwari et al. (2012a;. Wheat genomic SSR markers previously mapped on chromosome 2D and in previous club mapping studies (Johnson et al. 2008) were first tested on a set of parental lines, wheat deletion stocks for chromosome 2D, and a set of non-irradiated (0-Gy) control F 1 hybrids (without any deletions). The primer sequences of these SSR markers were obtained from the GrainGenes database (http:// wheat. pw. usda. gov).

Phenotypic evaluation and statistical analysis
Corrigin, LDN, non-irradiated (0-Gy) control F 1 hybrids (derived from a cross between Corrigin and Langdon) and club-RH 1 panel were grown in a greenhouse with one plant per pot at a temperature of 25 °C and a 16/8-h light/dark photoperiod. Since each seed of the RH panel exhibits a unique event, data were recorded per individual plant and at least 5 tillers per plant were used for phenotyping. As Fig. 2 Phenotypic analysis of normal wheat and club wheat with locus. a Spike morphology of a normal hexaploid wheat (wt) and club wheat (CW) used in this study. b a close look at the spikelet arrangement in wheat (WT) and club wheat (CW) lines. c Compari-son of rachis after removing spikelets from the wheat (WT) and club wheat (CW) lines. Comparison of spikes of the synthetic hexaploid wheat TA8051 (d) and Corrigin line (e) used for genetic mapping of C-locus in this study explained in Fig. 2, visible differences in Corrigin and Langdon spikes are quite distinct. We used the spikes of Corrigin, Langdon, and synthetic wheat line TA8051 and observed a significant difference in the spike compactness between Corrigin and Langdon as well as between Corrigin and synthetic wheat TA8051 (Fig. 2a-e).
The RH panel along with parental lines (Corrigin, Langdon, and F 1 -0 Gy-control) were harvested at maturity, and at least three spikes per plant were used for qualitative data recording. For the club-wheat RH population, we had two groups of phenotypes such as club type (like CW wheat) or Langdon type (lax type spikes). Since club-phenotype is controlled by a dominant locus, as expected, all the heterozygous plants showed a high degree of compactness.

RH mapping
Previous mapping (Johnson et al. 2008) established that the C-locus was located on chromosome 2D; hence, we designed chromosome 2D-specific markers covering various deletion bin stocks of chromosome 2D (Endo and Gill 1996), and 75 PCR-based chromosome 2D-specific markers were selected for RH mapping. Initial grouping and map construction was based on RH marker data using the software package Carthagene (de Givry et al. 2005).
RH groups were determined using minimum two-point LOD scores of 10.0 and a threshold distance of 0.3. RH groups were assigned to the physical map of the 2D chromosome by their sequence alignment to the reference genome by BLAST-based searches. After this step, the resulting groups were used for constructing maps for individual subgroups. The marker order of initial RH maps were improved using the Carthagene commands 'greedy,' 'annealing,' 'flip,' and 'polish' using default settings (Mahlandt et al. 2021).
Sequences of the mapped gene markers, ESTs, and common SSR markers anchored on Ae. tauschii and wheat reference genomes were used to make synthetic comparisons on the collinearity between RH and physical maps of the C-locus region (Luo et al. 2009(Luo et al. , 2017Appels et al. 2018). The number of obligate breaks in the RH groups was counted using the 'mapocb' command for a given map (de Givry et al. 2005). To analyze overall marker retention patterns across the different RH groups covering different parts of the wheat chromosome arms, we merged the RH groups using a threshold distance of 0.5 and LOD of 5 and the combined map was collinear with wheat reference chromosome based on anchored markers.

Initial genetic analysis and genetic mapping
To assess the segregation of the club phenotype in the mapping populations, the χ 2 Chi-squared test statistic was calculated and summed for the club, heterozygous, and parental phenotypes with an expected 1:2:1 ratio for the club × Synthetic F 2 population. The P value was calculated using the Chi-squared distribution with two degrees of freedom for F 2 s. To assess the segregation of genotypes for the selected markers and to construct a linkage map with C-locus, phenotypic and genotypic datasets, we used our earlier approach optimized as described (Mahlandt et al. 2021).
The linkage mapping was performed using the R package 'R/qtl' (Broman et al. 2003) as described in Mahlandt et al. (2021). Briefly, a greedy algorithm was used to establish initial marker order using a sliding window of 5 permuted markers in the calculation, and then, the Kosambi map function was used for final map estimation. The 'checkAlleles' command was used to check allelic information, and LOD scores and recombination fractions were estimated as per default settings. The R package 'R/LinkageMapView' was used to draw a linkage map (Ouellette et al. 2018).

Sanger sequencing of PCR amplicons
For the amplicon sequencing of the targeted PCR products, using the Sanger sequencing method, first, PCR products were run on a 1.5% agarose gel stained with ethidium bromide to visualize amplicons and to confirm single bands of high concentration. Next, PCR products were mixed with ExoSAP-IT enzyme (Thermo Fisher Scientific) at a 5:1 PCR product-to-enzyme ratio to remove dNTPs and purify amplicons. About 2 ul of purified PCR products was next added to a BigDye Terminator version 3.1 master mix (Applied Biosystems) containing 0.5 ul BigDye terminator, 1.75 μl 5 × sequencing buffer, 1 μl of 4 μM forward or reverse primer, and 4.75 μl molecular biology grade water. A thermal cycler was used to run the sequencing reaction as follows: 94 °C for 15 s, 50 °CC for 4 s, 60 °CC for 2 min, for 25 × cycles. Sequencing products were purified using a sodium acetate/EDTA/ethanol precipitation method and run on a 3730xl DNA analyzer (Applied Biosystems) for sequencing.

Physical localization of the C-locus region
Four different mapping panels: F 2 population (Corrigin × CS 2D); Corrigin and synthetic wheat F2 population; club-RH panel; and Ae. tauschii genetic map (Luo et al. 2009) did not show any major rearrangement or major difference in the order of the common markers (Luo et al. 2009), and it suggested that flanking markers can be used for the chromosome landing on the reference genome physical maps of hexaploid bread wheat and its diploid D-genome donor Ae. Tauschii to delineate the C-locus physical region. PCR-based sequences of the flanking gene markers were used to identify their physical location of the reference genome using BLAST-based similarity searches.

Map comparisons and mapping resolutions
For orienting the RH map and to compare the map resolution and mapping potential of the RH panel concerning the genetic map, we used an integrated genetic map (with BAC contigs) of chromosome 2D of Ae. tauschii, generated by genotyping of 572 individuals of an F 2 population (Luo et al. 2009) as well as the genetic map generated in this study using an F 2 population derived from a cross between Corrigin and SHW lines. Common anchored markers were used for the comparative analysis. Mapping resolution of a given RH map is a function of the number of lines used and a total number of chromosomal breaks induced (Tiwari et al. 2012b).

Phenotypic and genetic analysis
The Compactum locus or C-locus results in an extremely compact spike, and this pronounced phenotype has established a unique wheat subspecies T. aestivum ssp. compactum (club wheat). In this study, we used a club wheat cultivar Corrigin, for mapping and genetic characterization studies (Fig. 2). It was selfed for at least 2 generations in the greenhouse using single seed descent to ensure purity, penetrance, and phenotypic expression. Next, we characterized the parental lines and the mapping populations to confirm the nature and penetrance of the phenotype. From here on, we will use club wheat for Corrigin (containing C-locus) and C-locus for the locus/loci controlling the club or compact spike phenotype (Fig. 2).
In club wheat, the length of the spike was highly reduced as compared to the normal bread wheat cultivars (compared with bread wheat cultivars including Jagger, Winsome, and synthetic hexaploid wheat; based on visual observations). However, the number of spikelets per spike was found to be higher than standard hexaploid wheat cultivars (Fig. 2). Our genetic analysis confirmed that C-locus has very high penetrance and heritability and there is no segregation of the club phenotype in the parental line suggesting homozygosity. Other spike components (number of anthers, ovaries, and seed per spikelet) were similar to a standard bread wheat. Following fertilization, club-wheat spikes produced fertile grains with grains similar to normal wheat controls.
To confirm the dominant expression of C-locus and to analyze Mendelian inheritance, we analyzed the F 1 plants that originated between the crosses of club and non-club lines and derived respective RH and genetic populations. All the F 1 plants between Corrigin x Langdon (a tetraploid wheat with non-club phenotype) and Corrigin × synthetic wheat (TA8051) showed club phenotype without any exception, confirming the dominant nature of the phenotype.
Further, the F 2 population raised by selfing F 1 plants (Corrigin × synthetic wheat TA8051) segregated 3 club:1lax, indicating that C-locus is controlled by a dominant locus. We observed that ~ 25% of the F 2 population showed Corrigin-type spikes, about 50% of the lines in the F 2 population were recorded as heterozygous and about 25% of the lines showed spikes similar to synthetic wheat (relaxed/ non-club); these results are consistent with the published studies (Table 1.). To rule out any segregation distortion in the F 2 progeny, we tested the segregation ratios of two flanking markers (club-wheat × synthetic wheat TA8051). These markers followed an expected 1:2:1 ratio, suggesting no influence of segregation distortion in downstream linkage analysis (Table 1).
Since previous studies have indicated that C-locus is localized in the low-recombination region of wheat chromosome 2D, to perform precise mapping of the C-locus, a two-step mapping analysis was performed. In the first step, a recombination-independent RH mapping approach was used to localize C-locus. In the second step, cross-validation of the mapped C-locus interval was determined using a large F 2 population (1404 gametes; derived from a cross between Corrigin and synthetic wheat TA8051).

Development of club-wheat radiation hybrid (club-wheat RH) panel
Since C-locus is controlled by a dominant locus, deletion of the chromosomal region where the C-locus is located should show a non-club phenotype. Using this hypothesis, we can assay the RH panel made by crossing the gamma-irradiated pollen samples of the club line Corrigin (AABBDD) by pollinating Langdon (2n = 4× = 28; AABB; a tetraploid wheat cultivar with lax spike structure). Each seed set on the emasculated 'Langdon' spike after pollination with irradiated club-wheat pollen represents a unique mutation or chromosomal deletion event for the d-genome chromosomes (Fig. 1b).
Our work on club-wheat RH panel development (Fig. 1b) yielded a total of 900 RH 1 seeds (from about 112 emasculated 'Langdon' spike), only 350 germinated

Characterization of club-wheat RH panel using molecular markers
For developing the club-wheat RH panel, we used irradiated pollens from club-wheat (n = 3×; ABD) to pollinate emasculated spikes of tetraploid wheat (n = 2×x = AB). Thus, the d-genome chromosomes in this panel are expected to be in the hemizygous condition and the d-genome chromosomes/2D markers can directly be applied using the club-wheat RH panel.
To perform characterization of the club-wheat RH panel, we assayed a set of 279 RH lines using a panel of 14 SSR markers (2 markers per d-genome chromosome). The marker retention frequency is a key determinant of the mapping potential of the RH lines, and this can be defined as the fraction of the markers retained in a given RH panel. For example, if a certain 2D-specific marker is present in 85 out of a total of 100 lines of a given RH panel then its retention frequency can be estimated as 85% (deletion frequency can be 15%). Using a set of 14 SSR markers specific to all seven d-genome chromosomes, it was estimated as 85%.

RH mapping of the C-locus
Since we started this study with the knowledge that C-locus is located on chromosome 2D, we assayed the mapping potential of this panel across the whole chromosome 2D using chromosome-wide markers. The marker retention frequency based on all tested 75 markers specific to chromosome 2D was estimated as 83% and suggested that for any given marker there would be 17% deletion events in a panel of 100 lines. Figure 3 shows the distribution of club-RH lines for chromosomal breaks specific to chromosome 2D (Fig. 3). About 63% of the lines did not show any chromosomal breaks even after applying 75 chromosome-wide markers from chromosome 2D. Only 37% of lines show breaks for chromosome 2D (Fig. 3). The results indicated that this club-wheat RH panel might contain more than 50 overlapping deletion events. These overlapping deletions on chromosome 2D could allow the precise mapping of the targeted C-locus. Supplementary Fig. 1 shows the distribution of club and non-club phenotypes in the RH panel (SF. 1). Visual phenotyping is quite easy in assessing the deletions in the targeted C-locus region. If there was a deletion of the chromosome fragment containing the C-locus, a clear laxed spike phenotype was observed, but if the locus was retained, the phenotype would show compact spikes (SF.1). The phenotypic data were used in the format of dominant markers (present or absent), and RH mapping software Carthagene was used to map the C-locus using our integrated phenotyping and genotyping datasets.
Grouping of the 75 markers yielded two distinct RH groups (2D-RH-group-1 and 2D-RH-group-2) and 22 singletons. Each RH group was then individually used to generate RH maps using a threshold distance of 0.3 and LOD of 10 and default commands ).
The C-locus was associated with 2D-RH-group-1 with 39 markers and presented by 35 unique loci. The RH group containing C-locus spanned wheat deletion bins 2DS-5 to some part of C-2DL-1 with a total map distance of 360.29cR 1500 (Fig. 4c). The RH map of the C-locus spanned a total physical distance of 339.9 Mb starting from anchored marker BARC168 (44.0 Mb) to AT34976_1 (383.9 Mb) and showed excellent collinearity with the reference map of chromosome 2D with tested regions (Fig. 4a-e). Our RH mapping approach precisely mapped the C-locus on the short arm of the chromosome 2D between gene markers CS_2D_231 and CS_2D_242 spanning an RH distance of 1.6 cR 1500. The flanking markers on the C-locus spanned a physical distance of 11.0 Mb on the wheat reference genome chromosome 2D (Fig. 4c-d).
The mapping resolution of an RH panel can be calculated as a ratio between the total true physical distance covered by the RH map and the total map length of the targeted region. Since the RH map of the C-locus spanned a physical distance of 339.9 Mb and RH map length of 360.29cR 1500, the expected map resolution of the C-locus region can be calculated to be 339.9 Mb/360.29cR 1500 , which can be presented as 1.06 Mb/cR 1500 . The second RH group (2D-RH-group-2) contained 18 markers and 17 unique loci with an RH distance of 184.5 cR 1500. The anchored markers confirmed that the RH-group (2D-RH-group-2) is represented by the long arm region of chromosome 2D. The RH map of the long arm spanned between the 2D deletion bins 2DL-3 to 2DL-9. The map of the RH-group covers a total physical distance of 241.2 Mb and an RH mapping distance of 184.5 cR 1500 . The map resolution of the 2D-RH-group-2 is 1.3 Mb/ cR 1500, which is very similar to the 2D-RH-group-1.
We then merged the 2D-RH-group-1 and 2D-RH-group-2 using groupmerge command and generated a combined map by relaxing the grouping parameters. Overall, by combining 2D-RH-group-RH-group-1 and 2D-RH-group-2, this RH panel assayed a total of 601.0 Mb physical region as per wheat reference genome (IWGSC 2018) using wellspaced 57 mapped molecular markers (SF. 2.). Both the RH groups together occupied a combined RH map length of 594.6 cR 1500 (SF.2.). The average map resolution of the entire RH panel (by combining 2D-RH-group-1 and 2D-RHgroup-2) was calculated as 0.99 Mb/cR 1500 . The total number of obligate breaks identified in 2D-RH-group-1 and 2D-RH-group-2 maps was calculated to be 385 and 215, respectively. A total of 600 obligate breaks (similar to recombination breaks) suggested a similar mapping resolution of 1.01 Mb per breakpoint (601.0/600 obligate breaks).
We plotted the marker retention pattern across chromosome 2D (generated by combining 2D-RH-group-1 and 2D-RH-group-2), and it was found to be uniform with slight exception toward the centromeric region (SF. 3.). It suggested that although the number of chromosomal breaks was fewer close to the centromeric region there were enough breaks to perform high-resolution mapping of this region.

High-resolution genetic mapping of C-locus
RH mapping allowed the precise localization of the C-locus on the chromosome 2D short arm. In the next step, we used the molecular markers from the flanking C-locus region to cross-validate the mapping results using the genetic mapping approach. Since previous studies have indicated lowrecombination events surrounding C-locus, we used an F 2 genetic mapping population (club-wheat x synthetic wheat) of 1404 gametes (702 F 2 lines) to address this. We recorded a high rate of marker polymorphism between club-wheat and synthetic hexaploid wheat.  In general, this particular synthetic hexaploid wheat (TA8051) was highly polymorphic (for D-genome chromosomes) relative to modern cultivars because of its reconstituted AABBDD genome that includes a very distant accession of ancient wheat progenitor (Ae. tauschii, 2n = 2x = 14). High polymorphism between TA8051 and club wheat facilitated the development and use of SSR, single nucleotide polymorphisms (SNPs), and gene-based markers for mapping C-locus. Phenotypes were used as a stand-in marker to estimate the position of C-locus using a genetic mapping population. Progeny testing of selected genotypes was performed to confirm the mapping interval.
Ten markers between 180 and 311.0 Mb (from deletion bins C-2DS to C-2DL) of chromosome 2D from the flanking C-locus region were mapped on the F 2 population (Fig. 5). The mapping of this 131 Mb region using F 2 population generated a genetic map of 18.40 cM with 10 distinct unique loci and the genetic map was highly collinear with RH map, Ae. tauschii genetic map and reference chromosome 2D (Luo et al. 2017;IWGSC 2018).
Similar to our RH mapping results, the genetic mapping study placed the C-locus on the short arm of the chromosome 2D between flanking markers club_2D_229 and club_2D_242 and these flanking markers span a genetic distance of 2.2 cM. The C-locus spanned a physical region of 229.0 Mb -242.0 Mb as per reference genome chromosome 2D (IWGSC 2018) with a total physical distance of 13.0 Mb.
Our genetic mapping results further validated and confirmed the localization of C-locus on chromosome 2D short arm. Since two independent approaches placed C-locus on chromosome arm 2DS and a shorter distance was provided with RH mapping results, we used our consensus results and concluded that C-locus is flanked by the gene markers CS_2D_231 and CS_2D_242 and physically located between 231.0 and 241.0 Mb interval as per wheat reference genome chromosome 2D (Fig. 5).

The physical C-locus region and putative candidate genes
Our two independent approaches (Figs. 4 and 5) placed the C-locus at a physical interval of 11.0 Mb (IWGSC 2018). This consensus physical region is large, but both the mapping panels (RH and genetic populations) may be used to reduce this interval by adding more markers and more individuals from unutilized portions of the mapping populations.
Although the flanking 11.0 Mb region on a wheat chromosome is expected to harbor several candidate genes, since this locus is mapped in the low-recombination and low-gene density region of wheat chromosome (IWGSC 2018) we expect a limited number of genes in this region. We extracted all the high-confidence genes present in the flanking 11.0 Mb of the C-locus region from wheat reference genome 2D (IWGSC 2018; Fig. 6). Our analysis of Chinese Spring reference genome from the flanking 11.0 Mb region identified only 11 high-confidence candidate genes (Fig. 6). Since wheat pangenomes are also available, we extracted all candidate genes from the homoeologous 11.0 Mb region. This analysis yielded a range of 11-26 genes in the Deletion bin map of chromosome 2D a.
Genetic map of the C-locus b.
Anchored markers on reference chromosome 2D c.

Fig. 5
Relationship of wheat deletion bin map a with the genetic map of C-locus b and its respective alignment with reference genome chromosome 2D of wheat (c) Fig. 6 Graphical presentation of the 11.0 Mb long C-locus physical region delineated by RH mapping panel. a Partial RH map of the C-locus physical region; b gene id, gene family name, and high-confidence genes present in the 11.0 Mb region of the physical map targeted 11.0 Mb region (ST-2). Of these candidate genes, two specific genes TraesCS2D02G236100 and TraesC-S2D02G237200 (IWGSC 2018) have been described to affect the spike development. However, more work is needed to narrow down the causal gene associated with C-locus. Further narrowing down this region using available RH panels and genetic mapping lines will allow us to identify the underlying gene associated with the C-locus.

Analysis of comparative map resolution and mapping potential of the RH approach
We estimated resolution using markers mapped in the 2D physical region with significant marker coverage. One cR 1500 is defined as the distance at which there is a 1% probability of a break between two markers in a given RH panel that was developed using 15 Gy irradiation dosage (Tiwari et al. 2012b We also compared the rate of recombination and chromosomal breakage using wheat deletion bins and common markers anchored on both maps (Luo et al. 2009;Kumar et al. 2012;Consortium (IWGSC) et al. 2018). The rate of recombination defines the mapping resolutions in a given genetic map, whereas, in the RH map, it depends upon the rate of chromosomal breaks in a given physical interval. To compare the mapping potential of the RH panel and genetic maps using another approach, we used the RH map of the C-locus generated in this study using 279 lines and a genetic map of the chromosome 2D of the Ae. tauschii that was generated using 1102 AL8/78 × AS75 F 2 lines (Luo et al. 2009). Common anchored gene markers with the knowledge of the physical location on the reference genome chromosome 2D of Chinese Spring wheat were used for this particular evaluation (Consortium (IWGSC) et al. 2018).
The physical region between the markers AT2D1056 and BE490204 was compared using a total of 20 anchored gene markers (Fig. 4). On the Ae. tauschii genetic map, the interval between markers AT2D1056 and BE490763 spanned a genetic map length of 92.78 cM (extracted from Luo et al. 2009) and covered a physical length of 296.0 Mb. The overall map resolution of this region using the genetic map was observed as 0.25 cM/Mb.
The genetic map generated in this study (Corrigin × SHW TA8051 F2 population; 702 lines) to map C-locus (Fig. 5) also showed a genetic distance of 18.4 cM and first and last anchored markers on this map spanned a total of 131 Mb on chromosome 2D. Using a similar calculation, the overall map resolution of this genetic mapping population was estimated as 0.14 cM/Mb.
For the same physical region of 296.0 Mb, the RH map (based on 279 lines) spanned a total map length of 294.51cR 1500 . The map resolution of the RH panel was estimated as 0.99 cR 1500 /Mb, which was 4-7 times higher than both the genetic maps used in the comparative analysis (Fig. 4).
Although we did not have good coverage of markers in representative wheat deletion bins, we tried to measure the map resolutions in 4 different wheat deletion bins using these anchored markers (Table 2.). We found that the map resolution of the genetic map decreased continuously from the terminal bin to the centromeric bins from 1.44 to 0.03 cM/Mb (Table 2). In contrast, the resolution of the RH map in the same tested intervals showed comparable resolution in terminal deletion bin 2DS-5 (1.59cR 1500 / Mb) and was highest in the pericentromeric deletion bin 2DS1 and then decreased in the centromeric deletion bins 0.63cR 1500 /Mb, but still was 21 times higher than the genetic map (Table 2). Overall, these results indicate that RH maps can provide a much better option to map genes in low-recombination or highly suppressed recombination regions on wheat chromosomes.

Discussion
Wheat has a completely sequenced genome and very often its telomeric regions are referred to as distal parts while the centromeric and pericentromeric regions are referred to as the proximal parts of the genome Walkowiak et al. 2020). The proximal part of the wheat chromosomes shows highly reduced recombination events and much lower gene density with more than 30% of wheat genes residing in the proximal or pericentromeric regions (Saintenac et al. 2009;Tiwari et al. 2016;IWGSC 2018). It is well documented that precise mapping of wheat genes in low-recombination region (2/3 of a given chromosome) of wheat chromosomes poses a big challenge (Johnson et al. 2008;Saintenac et al. 2009;Luo et al. 2009;Tiwari et al. 2016). RH mapping is a recombination-independent approach that provides an alternative to genetic mapping approaches (Tiwari et al. 2012bBassi et al. 2013;Mahlandt et al. 2021). Since RH mapping uses randomly induced chromosome breaks for developing a marker scaffold, it is independent of recombination-based events and provides more or less uniform chromosomal breaks or deletions across the length of a given chromosome (Saintenac et al. 2009;Riera-Lizarazu et al. 2010;Tiwari et al. 2016). In an earlier study, Mahlandt et al (2021) reported the application of RH mapping in the high-resolution mapping of a dominant gene controlling grain number per spikelet in wheat (Mahlandt et al. 2021). This study observed that this deletion-based method is suited to map dominant traits in the telomeric region (terminal region of chromosome 2DL) of the wheat chromosomes (Mahlandt et al. 2021).
It is important to note that the entire RH mapping was done on RH 1 plants (similar to F 1 plants in normal crosses) and it suggests that RH mapping offers a much quicker approach to mapping wheat genes than genetic mapping that requires F 2 or further advanced RIL populations. For a given dominant gene, RH mapping can be concluded in a 6-7-month timeline confirming that RH mapping provides a quick and high-resolution mapping approach. Our club-RH mapping panel was generated in 8-9 weeks. Another 8-9 weeks is required to phenotype the panel for the targeted traits. Due to its shorter generation time, and requirement of a smaller number of independent crossing events, pollen irradiation panels are preferred for gene and genome mapping studies in large and complex genome crops such as wheat.
Similar to other reported studies, we found that 15 Gy is the optimal dosage (Tiwari et al. 2012bMahlandt et al. 2021) for developing a club-RH panel. The marker retention frequency of the club-RH panel was closer to 85 for all the d-genome chromosomes, but was a bit lower for chromosome 2D (83%), which may be because the addition of more markers reveals more deletions. With about 17% of the chromosomal deletion (83% marker retention), we estimate that the club-RH panel will generate about 50 overlapping deletions for a targeted physical region (Fig. 3). Tiwari et al. (2012a, b) reported that the percentages of interstitial deletion are higher than the terminal deletions in a given RH panel, especially toward the proximal part of the chromosome. It suggests that due to higher percentages of interstitial deletions and since these deletions are expected to be overlapping, they provide an excellent mapping resource for a dominant gene (Bassi et al. 2013;Mahlandt et al. 2021).
In the previous attempt to map the C-locus, Johnson et al. 2008 reported that due to highly suppressed recombination rates on chromosome 2D, the flanking markers from two deletion bins across the centromere spanned more than 80 Mb, but were tightly linked on the genetic map (Johnson et al. 2008). To date, the putative location of the C-locus was identified as C-2DS or C-2DL (Johnson et al. 2008;Yu et al. 2022;Wen et al. 2022). Some recent studies have also used different approaches to localize the C-locus without any consensus on a single arm of the chromosome 2D. Since the source of the club-phenotype was similar to the reports by Johnson et al. 2008 and our study, we used mapping information presented here (Johnson et al. 2008) as the base for our analysis (Fig. 4).
The distribution of marker retention frequencies of the entire club-RH panel across chromosome 2D indicated that the chromosomal breaks are distributed across the length of the chromosome (SF. 3.). As described in many published studies, we also observed that marker retention frequency was evenly distributed across the chromosome with a slight increase toward the centromeric region of chromosome 2D (Tiwari et al. 2012aBassi et al. 2013;Mahlandt et al. 2021). Combined with the higher marker retention, deletion bin and the genomic sequence data, our results correctly indicated that centromeric repeats might be located at 260-280 Mb region on the chromosome 2D of the reference genomes of CS (Consortium (IWGSC) et al. 2018) and Aegilops tauschii (Luo et al. 2017).
In this study, to map the C-locus we primarily focused on designing markers from these two bins (C-2DS and C-2DL). To compare the mapping potential of the RH map for a different part of the chromosome, we included representative markers from each of the wheat deletion bins of chromosome 2D.
Since all the mutations or deletion events in pollen irradiation take place in gametic cells, it eliminates the possibility of a chimeric plant, and the event of the entire deletion will be hemizygous, it provides an excellent approach to map a dominant gene that has a clear phenotype in RH 1 plants Mahlandt et al. 2021).
Using our high-resolution deletion-based RH mapping approach, we precisely localized the C-locus to the centromeric bin C-2DS-1 on the short arm of the wheat chromosome 2D, and by adding some more markers with the C-locus region, we can still reduce the physical interval for the C-locus. Using just 279 lines and a quick generation time, our results strongly support the benefits of RH mapping in localizing dominant genes in low-recombination regions in wheat.
To rule out any RH mapping-based artifact or background deletions, we developed a high-resolution genetic mapping population of 702 F2 lines generated by crossing club-wheat and synthetic wheat TA8051. Our genetic mapping results validated the physical interval of the C-locus and confirmed that random deletion and translocation events do not affect our ability to map the dominant trait in the RH mapping panel (Fig. 6). Several other studies have already shown the potential of RH mapping in trait mapping and our results also strongly support it (Bassi et al. 2013;Mahlandt et al. 2021).
Although both the studies (RH and genetic mapping) allowed us to localize C-locus, we observed significant differences in the map resolution of both mapping panels. RH map and genetic maps both were quite comparable in the terminal region of the chromosomes, but as expected map resolution of the genetic map was reduced drastically in the proximal part of the wheat chromosomes, especially in deletion bins in pericentromeric and centromeric regions (Saintenac et al. 2009).
Several studies have indicated that the mapping potential of RH mapping and genetic mapping was equivalent at the terminal region of the wheat chromosomes (Kumar et al. 2012;Tiwari et al. 2012bTiwari et al. , 2016Bassi et al. 2013;Mahlandt et al. 2021). It is due to higher recombination events at the tip of the wheat chromosomes and a lower number of interstitial deletion events toward the end of the wheat chromosomes (Tiwari et al. 2012b, a;Mahlandt et al. 2021). However, the mapping potential of the RH panel is very high in the low-recombination-prone proximal part (pericentromeric regions) of the wheat chromosomes. It may be due to an almost two-threefold higher number of small to large interstitial deletions in the proximal part of the wheat chromosomes than in the terminal parts (Tiwari et al. 2012b). Tiwari et al. (2012a, b) observed that most of the deletions in the pollen irradiation panel were smaller (< 20 Mb), and the frequency of interstitial deletions was higher in the proximal part of the wheat chromosomes than in the terminal parts (Tiwari et al. 2012b).
Overall, the map resolution of the RH panel was found to be ~ 4 times higher across the length of chromosome 2D. However, in pericentromeric (2DS1-0.33-0.47) and centromeric bins (C-2DS1-0.33 to C-2DL3-0.49), the RH map (based on 279 lines) showed 10.4-and 21-time higher map resolution (Table. 2) than the Ae. tauschii genetic map. Very similar results were observed when we compared the genetic map and RH map of the C-locus (Figs. 4 and Fig. 5). The improvement of the mapping resolution of RH on 2DS may be due to the accumulation of repeat sequences on the low-recombination regions, which appeared to be highly sensitive to breakage fusion by the radiation (Bassi et al. 2013).
The RH map of the C-locus region showed excellent collinearity with reference wheat and Ae. tauschii genomes and provide confidence and validation for the mapping approach. Using two independent approaches, we located the C-locus, which opens up opportunities for the molecular understanding of this important trait. The wheat spike is the most important part of the wheat that directly influences grain yield. The identification and molecular characterization of genes and regulatory factors controlling wheat spike architecture will allow us to engineer high-yielding wheat cultivars. Our ability to precisely map the locus was greatly facilitated by the availability of the wheat reference genome as well as the Ae. tauschii physical map (Luo et al. 2017;Appels et al. 2018) for marker development and their comparative analysis. The use of a very distant and diverse D-genome in the synthetic wheat derived from Ae. tauschii in developing the mapping population further aided by the rapid development of polymorphic markers.
This study used two independent approaches to precisely map C-locus on the short arm of chromosome 2D. Although the flanking markers spanned an 11.0 Mb region, they contain only 11-26 genes high-confidence annotated genes (Fig. 6) based on the wheat reference genome as well as wheat pangenomes (Consortium (IWGSC) et al. 2018;Walkowiak et al. 2020). Both the mapping panels have not yet been fully utilized, and by performing high-resolution mapping of recombinants and RH lines, we will be able to reduce the physical region as well as will reduce the number of candidate genes for validation of the candidate genes.
Precise mapping of the C-locus opens up the opportunity for significant progress toward understanding this important gene at the molecular level and demonstrates an exciting route to target dominant genes in low-recombination regions of wheat using RH mapping. It can be achieved by selecting using right set of germplasm as the female parent (such aswheat deletion stocks, ditelosomic lines and nullisomic lines) and using irradiated pollen samples of male parent with major QTL or gene located in the low-recombination regions on any wheat chromosome (Bassi et al. 2013;Mahlandt et al. 2021).