New Male Specific Markers for Hop and Application in Breeding Program

Male specific DNA sequences were selected from a Diversity Arrays Technology (DArT) mapping study to evaluate their suitability for determination of the sex phenotype among young seedlings in a hop (Humulus lupulus L.) breeding program. Ten male specific DArT markers showed complete linkage with male sex phenotype in three crossing families. Following optimization, four were successfully converted into PCR markers and a multiplex PCR approach for their use was developed. Among 197 plants (97 from the world collection; 100 from three segregating families), 94–100% positive correlation with sex phenotypic data was achieved for the single PCR amplification, whereas the multiplex approach showed 100% correlation. To develop a fast and low-cost method, crude sample multiplex PCR was evaluated in 253 progenies from 14 segregating populations without losing accuracy. The study describes, for the first time, the routine application of molecular markers linked to male sex in an intensive Slovenian hop breeding program. The methods described could be employed for screening of sex at the seedling stage in other hop programs worldwide, thereby saving resources for desirable female plants.

plants at least at one locus on a pair of chromosomes 10 . Renner and Ricklef 11 also reported that dioecy is more common in wind-pollinated species in comparison to animal pollinated plants.
Molecular markers are a useful tool for determination of sex in dioecious plants before the phenotypic differences become detectable as the plants enter their reproductive phase. Several reports on identification of sex-specific markers in other plants have been published, such as in rattan Calamus guruba Buch.-Ham 12 , Momordica dioica 13 , Coccinia grandis 14 , Carica papaya 15 , and Cannabis sativa 16 .
Hop (H. lupulus var. lupulus) is normally diploid (2n = 2x = 20) with nine autosomal bivalents and two heteromorphic sex chromosomes. Female plants have a pair of X chromosomes, while male plants have an XY pair with the Y chromosome being smaller 8,[17][18][19] . In all other species of the genus Humulus, such as H. lupulus var. cordifolius, H. japonicus, H. yunnanensis and H. acetosa, the sex is expressed based on a multi-sex chromosome system [20][21][22] . Expression of sex phenotype in hop is also influenced by the X: autosomes ratio 8 , indicating that the genes controlling pollen development are in the sex determining region and structural genes located on autosomes 17,21 . In hop, monoecious plants occasionally occur, especially among progeny of specific crosses. Škof et al. 23 revealed that a high percentage of monoecious hop plants were triploids. Experimental work of this study focusses on diploid germplasm.
Since the hop karyotype contains sex chromosome Y, which is transmitted from generation to generation only by the male line, it was first assumed that majority of polymorphisms on Y chromosome should be male specific. Polley et al. 6 , on the basis of one cross, isolated a male molecular marker linked to sequences on the Y chromosome. Čerenak and Javornik 24 tested the published marker in the Slovenian breeding program, but the marker did not multiply in diploid and tetraploid forms of the Japanese male hop No3-38, and therefore, there was an assumption that the marker developed was specific to European hops. Patzak et al. 5 further tested its suitability in four families from the Czech breeding program and they also observed that the marker was not absolutely linked to the male phenotype. Seefelder et al. 25 were the first to construct a genetic linkage map of the male hop, and they found 24 molecular markers which were sex specific. Danilova and Karlov 26 were successful in the development of male specific molecular markers with the inter-simple sequence repeat polymerase chain reaction (ISSR-PCR) method on hop plants of Russian and European origin. Čerenak et al. 27 published a genetic linkage map where simple-sequence repeat (SSR) marker HlAGA7 mapped to the male locus. Jakše et al. 28 further described the HlAGA7 marker which appeared to distinguish male, female and monoecious plants in two Slovenian populations. In the study of McAdam et al. 29 the marker HlAGA7 showed complete linkage to the male sex phenotype in a New Zealand mapping population. Even though the marker proved to be perfectly associated with the male sex, the use of this particular marker was limited due to the technical difficulties of using SSR with multiple alleles. Buck et al. 30 developed four sex linked markers; three sequenced characterised amplified region (SCAR) markers, and one high-resolution melting (HRM) curve analysis marker the latter of which has been successfully applied within the New Zealand hop breeding programme 31 . Recently, Hill et al. 22 identified a pseudoautosomal region (PAR), and male-specific regions of the Y-chromosome, along with genes located in these parts of the sex chromosomes.
In order to develop an efficient method to determine the sex phenotype among young seedlings in a hop breeding program, we developed a multiplex PCR approach amplifying four male specific markers and a chloroplast specific DNA fragment. Male specific markers were idenitified from a DArT (Diversity Arrays Technology 32 genetic mapping study 33 . Selected DArT markers and the developed multiplex PCR approach were validated across a broad spectrum of hop genetic resources, in progeny of specific crosses, and for routine application in the Slovene breeding program. Furthermore, we evaluated these markers on crude sap samples to avoid the tedious step of DNA isolation, with promising results.

Materials and Methods
Phenotypic determination of sex in hop plants. Sex  DArT male linked markers. In the framework of the hop DArT consortium mapping project, all markers were sequenced by Sanger technology. The genotyping data for the Slovenian population were searched for DArT markers showing complete linkage with male sex phenotype, while being absent in female plants (i.e., no recombination between the DArT marker and trait of interest). The resulting set of markers was compared against the phenotypic data from the other two mapping populations, to infer male plants and to identify any additional male sex-linked markers.
Identified male linked sequences were edited and assembled in CodonCode Aligner (ver. 7.1.2) and submitted to GenBank (MG744425-MG744432). A hop chloroplast specific sequence named contig18 (GenBank MG744433) was obtained from our recent hop transcriptome project 36 .
Primer construction. Primer pairs for use in a single PCR were constructed using PRIMER3 37 web version using the default program parameters. Primers for the multiplex approach were designed using the MPprimer tool 38 specially developed to account for multiplex conditions, using default program options. Developed primers are presented in Tables 1 and 2. Single PCR marker reaction. Plant material. 203 different genotypes (including cultivars, wild female and male plants, and monoecious plants; Supplementary Information 1) and 100 breeding lines (Table 3) with known sex phenotypes from three different families were used in PCR amplification. Only two primers, hPb-CONT and hPb-719005 (Table 1) were further used for amplification across three crossing families (Table 3). Amplified PCR products were separated on 2% agarose gel and visualized by ethidium bromide staining.  (Table 3) were used for comparison of the results obtained from single and multiplex PCR reactions.

DArT marker
Primer sequence for (5′-3′) Primer sequence rev (5′-3′)   www.nature.com/scientificreports www.nature.com/scientificreports/ Multiplex PCR amplification. Single primer-pair PCRs were carried out initially, for each of the five multiplex primer pairs listed in Table 2. After initial confirmation by single pair amplification, the multiplex amplification was optimized by varying the primer concentration. Optimized PCR conditions in 15 μl reactions were as follows: 40 ng DNA, 1x QIAGEN Multiplex PCR Master Mix, 1x Q-Solution and primers at following concentrations: 0.2 μM for primers hPb-CONT, hPb-365890 and hPb-719005, 0.4 μM for primer hPb-718821 and 0.04 μM for primer contig18. Different primer concentrations are crucial to achieving a comparable rate of amplification of five different fragments. For example, contig18 which is of chloroplast origin requires much lower primer concentration, due to higher number of copies present in nucleic acid extract compared to the nuclear DNA. Amplification was carried out using the following thermal cycling touchdown protocol: 95 °C for 15 min, followed by 8 cycles of 30 sec at 94 °C, 90 sec at 65 °C (decreased 1 °C each subsequent cycle) and 90 sec at 72 °C. The amplification continued for 27 cycles at 94 °C for 30 sec, 57 °C for 90 sec and 72 °C for 90 sec. The reactions were completed by incubation at 72 °C for 10 min. PCR products were separated on 2% agarose gel and visualized by ethidium bromide staining.
Crude sample multiplex PCR amplification. Plant material. As a first optimisation step, 10 samples of female varieties and 10 male plants were used for amplification of crude sample PCR. Afterwards, 253 hop plants at the seedling stage from Slovenian hop breeding program representing 14 crossing families were analysed by using crude sample multiplex PCR amplification (Table 4).
Crude Sample multiplex PCR amplification. Crude sample multiplex PCR amplification was developed using Kapa 3 G Plant PCR kit (Kapa Biosystems) utilizing fast extraction of crude DNA extract. Leaf disc circles (1 cm diameter) were excised by a puncture tool and immersed in 200 μl of extraction buffer (0.5 M Tris-HCl, 1 mM EDTA (pH = 8.0), 2% β-mercaptoethanol) with two steel beads (5 mm). Tissue was homogenized in TissueLyser (Qiagen) with 10 rotations per second for 30 sec. Samples were heated at 95 °C for 5 min, cooled (-20 °C) for 10 min and centrifuged at 12,000 g for 10 min. Supernatant of crude extract was diluted in ratio 1:9 in sterile dH 2 O. Two μl of diluted crude DNA extract was used in 10 μl PCR solution containing 1x PCR buffer, 1.25 mM MgCl 2 , Enhancer (diluted 1:50), 0.2 U of KAPA3G Plant DNA polymerase and primers in same concentration as determined previously. The PCR amplification profiles and agarose gel electrophoresis analysis were the same as described for multiplex reactions (see Multiplex PCR amplification).

Results
DArT male linked markers discovery. In the Slovenian mapping family ('Hallertauer Magnum' × 'SBL2/1') represented by 92 plants, 9 were phenotypically male. Based on this observation, 10 DArT markers (hPb-361327, hPb-363461, hPb-365890, hPb-366371, hPb-715987, hPb-716314, hPb-718821, hPb-718886, hPb-719005, and hPb-716926) were discovered that were present in the male parent, absent in the female parent and present in all male siblings. Based on this information these markers were searched in the New Zealand ('Nugget' × 'SBL3/3') and USA ('Perle' × 'USDA19058M') mapping populations for being present in male parent and in male siblings. Three markers (hPb-365890, hPb-716314 and hPb-719005) were confirmed to be present in all three families, 5 were common between Slovenian and New Zealand's families (hPb-718886, hPb-718821, hPb-715987, hPb-363461, and hPb-361327), while one marker was unique to each of the Slovenian family (hPb-716926) and USA family (hPb-366371). Therefore, there were a total of 10 DArT markers specific to the male phenotype among those three families (Table 5).   Initial single PCR screening of male linked markers. Single PCR primers were developed (Table 1) and PCR conditions optimized on a set of four female and four male hops comprising cultivars 'Magnum' , 'Perle' , 'Comet' , and 'Fuggle' and male breeding lines 2/1, 3/3, 19058, and 29-70-54 for four DArT markers, including the contiguous sequence hPb-CONT (representing hPb-361327, hPb-363461, hPb-718886), hPb-365890, hPb-718821 and hPb-719005. These four male-specific markers were further screened in 122 female genotypes (117 cultivars and 5 wild hops), 44 male genotypes and 37 monoecious genotypes (20 predominantly male phenotype -Mf; 16 predominantly female phenotype -Fm and 1 plant in which neither male nor female flowers clearly predominate -FM - Supplementary Information 1). In summary, no male specific marker was successfully amplified in female genotypes, while the success of identifying males varied. For example, markers hPb-CONT and hPb-365890 each failed to amplify in two male genotypes (hPb-CONT: No3-38 and 284/113; hPb-365890: 19058 and 120/13). The marker hPb-718821 was not amplified in one male genotype (19058), and hPb-719005 did not amplify in four males (No3-38, 85/169, 19058 and 120/13). Interestingly, male specific markers were amplified in Mf hermaphroditic plants while not in Fm plants. For two DArT marker sequences (hPb-366371 and hPb-716926) amplification could not be achieved without varying the PCR conditions. Further analysis was performed on 100 breeding lines from 3 crossing families (Table 3) by using primers hPb-719005 and hPb-CONT (Fig. 1). By using primers hPb-719005 and hPb-CONT the correct sex determination was achieved between 94% and 100% of cases, depending upon the marker used and the family, showing the importance of using all four male specific markers. Therefore, all 5 markers are required to maximise the likelihood that the assay is effective throughout the biogeographic range of hop. This is important as the large majority of hop in commercial production and breeding programs around the world is some form of multi-generation hybrid between European and North American germplasm.

Development of multiplex PCR approach.
Since neither hPb-CONT nor hPb-719005 were able to detect 100% known males independently, we aimed to develop a multiplex PCR approach where all four markers could be amplified simultaneously. In multiplex primer development the fifth sequence of hop chloroplast origin  Table 5. Presence of male linked DArT markers in three different mapping families 29 . * DArT markers with almost identical sequences (sequence comparison in Supplementary Information 2). www.nature.com/scientificreports www.nature.com/scientificreports/ (contig18) was included to confirm amplification of isolated DNA and to exclude the risk of false negatives being identified as females. The developed multiplex primer set is presented in Table 2.
The verification of multiplex PCR amplification was initially checked on 97 genotypes (73 female and 24 male plants) representatives of globally significant hop accessions (Supplementary Information 1) by multiplying all five primers (hPb-CONT, hPb-365890, hPb-719005, hPb-718821, contig18). The markers successfully predicted sex phenotype in 97/97 cases (Fig. 3). In one male plant, a wild male genotype (63012 - Fig. 3, lane 20) of North American biogeographic origin, the primer hPb-718821 failed to amplify the fragment, even when the reaction was repeated. Nevertheless, the sex phenotype of the plant was correctly distinguished based on amplification of the other three male specific markers.
The results obtained from single PCR amplification and multiplex PCR amplification were compared. By using hPb-CONT and hPb-719005 primers, the results of sex determination in three crossing families coincided with phenotypic determination from 94% to 96% and 94% to 100%, respectively. Comparatively, when multiplex of five primers for PCR amplification was used in same progeny, 100% success was achieved. Furthermore, correct results were obtained by using multiplex PCR in the analyses of 96 world hop accessions ( Supplementary  Information 1).
Direct PCR reaction using crude nucleic acid extract. With the aim of accelerating sex determination in the hop breeding process, the possibility of using crude sample multiplex PCR amplifications was investigated on an initial 20 samples. In this initial test, the multiplexed markers correctly diagnosed the sex phenotype of each sample (Fig. 4).
Furthermore, in 14 families (a total of 253 genotypes - Table 4) the sex phenotype was diagnosed by crude sample multiplex PCR at seedling stage. After molecular analysis the plants were separately planted according to their marker determined sex genotype, in field trials within the context of the Slovenian hop breeding program. In the subsequent 1-2 years, phenotypic observations were collected and the comparison with molecular data was obtained. As can be seen from Table 4, the results from crude multiplex PCR amplification coincided perfectly with the in-field determination of plant sex phenotype.

Discussion
In dioecious species, where plants of one sex phenotype are preferred for commercial use, breeding or cultivation, various DNA fingerprinting methods have been employed for sex determination, as described in species such as Cannabis sativa 39 , Asparagus officinalis 40,41 , Actinidia chinensis 42 , and Ficus fulva 43 . In some species such as Garcinia gummi-gutta 44 , Simmondsia chinensis 45 , and Calamus guruba 11 , the time between germination and  www.nature.com/scientificreports www.nature.com/scientificreports/ the onset of reproductive maturity may be several years, imposing a strong limitation on the use of phenotypic determination of sex 9 . In hop, male plants cannot be differentiated phenotypically from female plants before flowering (sexual maturity), early sex determination via molecular markers is appealing. Beatson et al. 31 reported the economic benefits of using a HRM sex marker for elimination of male seedlings at the nursery stage in several New Zealand triploid breeding populations. The sex determination system is similar in Humulus, Cannabis and Rumex species, where the XX/XY system exists and the ratio between X chromosomes and autosomes effects sex expression 8,19 . In the family Cannabaceae, the three species -Cannabis sativa, Humulus lupulus and Humulus japonicus -are dioecious with heterogametic sex chromosomes. In H. lupulus, Ono 46 described six different systems, where 1 to 3 pairs of sex chromosomes with different sizes of Y chromosome exist. Divashuk et al. 47 reported a cytogenetic marker for the identification of sex chromosomes in H. lupulus. They revealed pseudo autosomal regions on the long arms of the X and Y chromosomes. Furthermore, Hill et al. 22 identified a set of loci that are sex-linked and probably located in the pseudo autosomal region. Their study identified a 1.3 Mb section of DNA that appears unique to male hop genotypes. They proposed that the identified region could be utilised for the development of molecular markers for diagnosis of sex phenotype.
All reported sex determining molecular markers in hop, have been linked to the maleness. Polley et al. 6 were the first to publish a sex-specific DNA sequence in hop, developed from random amplified polymorphic DNA (RAPD) molecular markers by using bulk segregant analysis (BSA), which is predominantly present on the Y chromosome and hybridized only weakly to female DNA. Upon testing 5,23 , this first sex-linked marker did not appear to coincide completely with the phenotypic assessments. Jakše et al. 28 reported an SSR marker tightly linked to the male sex in hop, which appeared to show complete linkage to the male character but technical difficulties associated with SSR genotyping meant that this marker was not applied in marker assisted selection (MAS).
Hop breeding programs are long lasting, with a development timeline of 10 to 15 years from crossing to registration of a new cultivar. To improve the efficiency of traditional selection procedures, several different molecular markers have been used to analyse genetic distances among hop breeding genotypes 33,35,[48][49][50] , and to detect QTLs (quantitative trait loci) [26][27][28]51,52 . Nevertheless, implementation of developed techniques in marker-assisted selection has remained unpublished up to this point. In order to avoid the delayed identification of male plants within the experimental systems of hop genetic improvement, we developed and applied selected DArT molecular markers to determine sex at an early hop seedling stage.
The present study was based on the research of Howard et al. 33 , where 730 polymorphic markers from 92 hop accessions were discovered using diversity arrays technology (DArT), which were further used in linkage studies 29 . There have been few studies describing the transferability of DArT markers identified through research, into breeding programs. In most cases markers were used in variability studies and genome mapping in different species, such as olive (Olea europaea L.) 53,54 , sugar beet (Beta vulgaris L.) 55 , apple (Malus domestica Borkh.) 56 , eucalypt (Eucaplyptus spp.) 57,58 , hexaploid wheat (Triticum aestivum L.) 59,60 , perennial ryegrass (Lolium perenne L.) 61 , and tomato (Solanum lycopersicum L.) 62 .
By studying sex linked marker performance in 197 plants (97 from the global hop genotype collection and 100 from crossing families), sex determined by multiplex PCR amplification appears to completely coincide with in-field phenotypic assessment. Furthermore, by using a crude sample multiplex PCR technique, the determination of sex in hop seedlings was made more efficient, without losing accuracy. The present research verified both single, and multiplex PCR sex-linked markers in 203 genotypes from world collection and three experimental families and developed and demonstrated the efficacy of a system of rapid, crude sample extract multiplex PCR sex-linked markers among progeny of 14 crosses.
To conclude, selected markers combined into a crude sample multiplex PCR assay, were tested in a broad spectrum of hop genetic resources, in progeny of crosses, and for routine application around 4.000 seedlings were annualy tested in the period of 2015-2018 in an active hop breeding program. It is important to note that taking into consideration all technical hours previously occupied in preventing male plants from flowering near commercial fields and the fact that accurate determination of sex at the seedling stage reduces the trial area required for screening by about one third, the financial cost of MAS has been recovered. Further, the laboratory analysis can be performed during winter, using leaf samples collected the previous growing season, spreading research activity away from seasonal labour peaks associated with hop production. The methods described above appear to Full-length gel is presented in Supplementary Information 3.