POLYMORPHISMS IN SSR-LOCI ASSOCIATED WITH E GENES IN SOYBEAN MUTANT LINES OFFER PERSPECTIVE FOR BREEDING

Flowering time is critical for successful seed production of plants, and with time to maturity it determines geographic adaptation, seed quality and yield. In contrast to most cereals, soybean (Glycine max (L.) Merr.) cultivars are confi ned to comparatively narrow ranges of geographical latitudes [1]. For each degree of changes in latitude (corresponding to 100 – 150 km), it is better to develop a new (better adapted to environmental conditions) cultivar [2]. ISSN: 2312-3370, Agricultural Science and Practice, 2019, Vol. 6, No. 3


INTRODUCTION
Flowering time is critical for successful seed production of plants, and with time to maturity it determines geographic adaptation, seed quality and yield. In con-trast to most cereals, soybean (Glycine max (L.) Merr.) cultivars are confi ned to comparatively narrow ranges of geographical latitudes [1]. For each degree of changes in latitude (corresponding to 100 -150 km), it is better to develop a new (better adapted to environmental conditions) cultivar [2]. Across the world, cultivars of soybean can grow in a wide range of latitudes from 55° N to 35° S [3,4]. Northern than 55 ° N latitude most genotypes do not mature [4]. In Ukraine breeders identify the area most favorable for growing this crop, the so-called «Soybean's Belt», which is located on irrigated soils and is between 46° N and 51° N. According to [4] in 5 regions of the Forest-Steppe Zone: Vinnitsa, Kyiv, Poltava, Cherkasy and Khmelnitsky and Steppe-Kropyvnytsky, more than 60 % of all soybean in Ukraine is grown. For all regions of Ukrainian «Soybean's Belt» extra-early and early maturity groups of soybeans allow to expand the area of this crop considerably, to get dry grain without drying, to use soybean as intermediate and repeated crops are especially important [5].
The development of early-ripeness (short season) soybeans for different regions of Ukraine requires effective use of early maturity genes. Detection of alleles of E genes that are involved in the control of plant response to photoperiod and determination of the days to fl owering (DFT) and days to maturation (DTM) with microsatellite markers could help in the evaluation of adaptive capacity of soybean cultivars under different growth conditions [2].
Soybean production in Ukraine increases every year since its wide introduction in 2007, and now Ukraine is one of the leaders in the production of soybean in Europe [6]. There are few investigations of molecular genetic polymorphism in the genome and genes that affect time to fl owering (involved in mechanism of photoperiodic sensitivity of plants) with molecular markers for Ukrainian soybean cultivars [7][8][9][10][11]. But optimization of soybean breeding and development of new cultivars with good adaptability to the conditions of Ukraine are important.
Breeders are interested in developing new material with effective alleles of E genes (early maturity genes) for Ukraine [4,5,12]. By using experimental mutagenesis, it is possible to get a high level of genetic variability. But it is impossible to predict to which changes chemical mutagenesis will lead to. The mutagenic factors in soybean selection are most often used for production of new forms that differ from the original cultivars according to individual characteristics: seed coloration, plant height, seed size, leaf shape, duration of the vegetative period, content and quality of protein and seed oil, resistance to pathogens and increasing productivity elements, especially: main stem nodes, pods per plant, seeds per plant, weight of thousand seeds [13,14]. The main aim of modern soybean breeding is improving the productivity, technological qualities of the seeds, increasing resistance to biotic and abiotic factors, yield, fertility, optimizing the growing season [15].
The purpose of our work was to analyze genetic diversity in mutant lines of soybean by using the microsatellite (MS) markers Satt100, Satt229, Satt319, Satt354, Satt365, Sat_038 linked with genes that determine sensitivity of soybean to photoperiod and time to maturation. These markers are recommended for E1, E3, E4 and E7-genes detection by Molnar et al. [16]. The markers Satt100, Satt319, we used as recommended by Rosenzweig et al. [17].

MATERIALS AND METHODS
A set of new soybean lines created with the help of chemical mutagenesis from cultivars Femida, Oksana, Podils'ka 416, Zolotysta, that belong to different groups of maturation [18][19][20], was used as a material for investigation.
The mutagens were applied on seeds in aqueous solution at concentrations of 0.05 g/l, 0.5 g/l, 5 g/l, and 10 g/l. Seeds were exposed to the mutagens for 2, 4, 8, 16 hours [21]. After mutagenic treatment lines were grown in an experimental fi eld during 5-7 agronomical seasons in the Institute of Feeds and Agriculture of Podillia of NAAS (IFAP; Vinnitsa, Ukraine). Then, mutant lines perspective for breeding with internodes less than 5 cm were selected in order to reduce the length of internodes, to decrease the vegetative mass of plants and to increase the generative mass of seeds and beans (Dr. S.V. Ivanyuk, personal communication (IFAP)).
DNA was isolated from soybean seeds of 10 mutant lines and parental cultivars using the DNA-Neo-Prep100 kit (Neogen Laboratory, Kiev, Ukraine). Five randomly chosen seeds from each of the parents and mutant lines were used for DNA isolation. PCR with primers specifi c for the microsatellite loci Satt100, Satt229, Satt319, Satt354, Satt365, Sat_038 was performed according to the method of Monlar et al. [16]. For fragment analysis and detection of the alleles of MS-loci the ABI PRISM® Genetic Analyzer  [22].
Agronomical traits, such as days to fl owering (DTF), days to maturation (DTM), length of the vegetative period (LV -days), duration of the period shoots-fl owering (S-F -days) and yield (t/ha) were investigated in three year trials (2016-2018) under fi eld conditions of IFAP, 49 o 13`N (Vinnitsa, Ukraine). Parental cultivars, mutant lines, and the controls, cultivars Vilana and Maple Arrow were sown all at the same day.
Parental cultivars and control isolines belong to different maturity groups: (00 − early maturity (91-110 days)) − Harosoy OT 89-5, Ros', Zolotysta; (0 − middle maturity (111-130 days)) − Maple Arrow, Cormoran AC, Podilska 416; (I − middle late maturity (131-150 days)) − Vilana [18,23]. According to the Catalog of varieties of forage and fi eld crops of IFAP, cultivar Femida, which has LV − 116-124 days, belongs to the middle maturity group I and cultivar Oksana, which LV is 125-132 days, belongs to the middle-late maturity group II [24]. So, cultivars in the same group of maturity could have different alleles of E-genes. Classifi cation of maturity groups (MGs) for soybean cultivars have been developed in the 1940s, it was revised and improved from these times and now includes 13 MGs (000, 00, 0, I -X), according to Liu et al. [25]. In different countries scientists have adopted this system to local conditions, for example, in Japan local soybeans differentiate in 8 MGs (0-VII), in India local cultivars are mainly in V-VIII MGs, in Italy soybean cultivars are from 0 to II MGs, in France -from 000 to II MGs. At the same time the system of MGs is a major approach in characterizing ecological properties and possible growing areas of cultivars and lines [25], there is a mind that the difference in maturity date between two adjacent groups is approximately 10 to 15 days in adapted area [25]. We used the one-way analysis of variance (ANOVA) and the Statistica 10 software for analysis of the results of the fi eld experiments. The signifi cance of interline variations for each trait was determined using Fisher's exact test (F-test) and LSD of the corresponding level of signifi cance for the investigated factors.

RESULTS AND DISCUSSION
In total 25 alleles of 6 MS-loci for the investigated parental, mutant and control genotypes were detected ( Table 1). The number of alleles per locus ranged from 3 to 6 with an average of 4.2. In the genotypes of 10 mutant lines, which were selected as promising for breeding, we have revealed new alleles of microsatellite loci that were not present in parental cultivars (bold in the Table 1). The changes of allele size in mutant lines were detected in 52 % of the cases of comparisons with the alleles of the same loci in the parental cultivars.
These changes can be explained by the action of mutagenic substances applied at different concentrations on soybean by inducing instability and mutations in «hot spots» of the genome, among which microsatellite loci can act. We supposed that mutagenic treatment, however, could have infl uence on generative apparatus of soybean [26], that leads to open fl owering and possible pollination by pollen of other cultivars and soybean lines grown adjacent to the experimental fi eld. In general, soybean is self-pollinating plant, but crosspollina-tion may still occur at a low percentage (average 1.8 %) [27] under natural conditions, it is still not enough investigations to say whether the percentage of crosspollination can increase after mutagenic treatment, but we can assume that it can be possible. Whatever the cause of the observed variation is, breeders found and selected lines, which varied at MS-loci, linked with genes that control photoperiod sensitivity.
Five alleles were detected at the Satt100 locus in the mutant lines and parental cultivars and an additional allele for cultivar Ros' that was used as control. In comparison with the control varieties we identifi ed dominant and recessive alleles of E7 gene according to alleles of Satt100 locus. Three alleles were revealed for the Satt319 locus. According to our results cultivar Oksana, mutant lines Oksana M12 and Femida M29, which have a 167 bp amplifi cation fragment at locus Satt100 and a 175 bp amplifi cation fragment at locus Satt319, have the dominant allele of E7 gene. Genotypes Vilana and Harosoy OT 89-5 were used as the control, they have the same alleles (167 bp at Satt100 and 175 bp at Satt319 loci). We should mention that according to Molnar et al. [16], alleles 167 bp at Satt100 and 175 bp at Satt319 are associated with the dominant allele of E7 gene.
In our investigation an allele of 175 bp at Satt319 locus was detected in Femida, Femida M32, Podils'ka 416, Podilska 416 M40, but there was no confi rmation of the presence of the dominant E7 al- lele, when we analyzed the Satt100 locus for these lines.
The presence of E3 dominant allele was established for Podils'ka 416, because for this cultivar we identifi ed an amplifi cation fragment of 215 bp at locus Satt229, the same as we showed to be present in the control cultivar Maple Arrow -a carrier of the dominant E3 gene. We did not detect any dominant alleles for E2 among investigated parental and mutant genotypes. The control allele for E2 − fragment of 243 bp at locus Sat_038 was found only in cultivar Ros', which is the carrier of E2.
One of the purposes of soybean breeding in Ukraine is creation of ultra-early cultivars with high yield potential in order to originate a cost-effective cultivar, which produce high harvest as early as possible and permit, to prepare a fi eld for sowing winter crops. For the creation of such genotypes, according to Dr. S.V. Ivanyuk, it is necessary to combine the recessive alleles e1, e2, e3, e4, e5, e7 in one genotype of new cultivar by crossing different donor sources of soybean. Golovenko et al. [28] stated on the basis of their research with the 7 alleles, that a recessive genotype for all these genes contributed to higher soybean yields in Belarus. Miladinovich et al. [29] have shown that the combina- Table 3. Data of dispersion analysis of trait variations Notes: Rd -Remaining dispersion; * Signifi cant at p = 0.05, ** p = 0.01 and *** p = 0.001. df -degree of freedom.

MS-loci/Trait
Variation source, MS Notes: *-for DTF days from the beginning of June, ** -for DTM days from the beginning of August. tion of alleles e1-as/e2/E3/E4 was the most common for high yielding soybean genotypes in Novi Sad. They stated that this specifi c allele combination possibly is the optimal one for the climatic zone of Central-Eastern Europe.
In our experiment E1 dominant allele has been detected in Zolotysta M 16, because this line has a 270 bp allele at locus Satt365, the same allele was present in the control cultivars Cormoran AC and Ros' in which dominant E1 is present.
Signifi cant differences between investigated lines were detected in three years fi eld trials for traits DTM (Fig. 1, a) and LV (Table 2).
There were no signifi cant differences in the date of fl owering, period shoots − fl owering and yield ( Table  2, Fig. 1, b) between all investigated genotypes.  (Table 3).
On the other hand, alleles of MS-locus Satt100 affected all traits except DTF; alleles of Satt319 and Satt354 MS-loci affected DTM and LV; and alleles of Sat_038 MS-locus affected DTF and S-F. Lines with alleles 167 bp at Satt100 and 175 bp at Satt319 loci (that marks E7) were shown to have a longer vegetation period and later maturity than other ones. There were no signifi cant differences in DTF for these lines.
According to Molnar et al. [16] MS-marker Satt354 can be used to detect alleles of E4; the dominant allele of E4 is marked by a 251 bp amplifi cation fragment. According to personal communication of Dr. E.A. Aksyonova (Institute of Genetics and Cytology, Minsk) the amplifi cation fragment for Vilana has smaller molecular size, than that of the control line L64-4830 with the dominant allele. In our investigation all amplifi cation fragments sizes were smaller than those of cultivar Vilana. Thus, we can conclude that lines with dominant allele E4 were not present in our experiment. The mutant lines with allele 232 bp at Satt354 locus reached maturity later than lines with other alleles at this locus ( Fig. 2). In literature, there is, however, no information that a 232 bp fragment marks the dominant allele of E4.
The lines with allele 247 bp at the Sat_038 locus fl owered earlier than the lines with a 245 bp allele in our fi eld experiment.
Genotypes with the same alleles for each of the mentioned E-genes were grouped together. Dominant E1 allele, which is present in Zolotysta M16, could be involved in earlier fl owering (about 10 days), than the other lines and cultivars demonstrated and the S-F period was 10 days shorter for this line in comparison with all investigated material. But of course, these shorter periods could also be caused/infl uenced additionally by interaction of the genotype by E-genes with environmental conditions and genetic background of cultivars.
We did not observe signifi cant E3 effects on the investigated agronomical traits.
Alleles of the E7-gene signifi cantly infl uenced on the DTF, DTM, LV and S-F. Genotypes with the dominant allele were characterized by a longer DTF, DTM and LV. Genotypes with E7 and e7 alleles had not signifi cant differences in the duration of S-F period, but for the lines with an allele 175 bp at locus Satt319 the characteristics S-F was shorter for 9 or 6 days for E7 dominant and e7 recessive alleles respectively. Furthermore, for lines with e7 and an allele of 175 bp at locus Satt319 there were no differences in the traits for LV and DTM and for soybean genotypes with the E7 allele the maturity was later and LV was longer for 10 days.
Both groups with E1 and E3 dominant alleles contained only one genotype each. According to our results of MS analysis we predicted E1 to be present in Zolotysta M16 and E3 in Podils'ka 416. Рresumably, lines created on the basis of one cultivar had smaller differences than those created on the basis of another cultivar. Perhaps in this case, it will be possible to determine the infl uence of the genotype on the DTF, DTM, LV, S-F and yield, which is also of importance.
Inside a group created from one ancestral genotype, for example, the group created on the base of Podils'ka 416 (included Podils'ka 416, Podils'ka 416 M33, Podils'ka 416 M38, Podils'ka 416 M40) we have not detected signifi cant effects of the E3 on the investigated traits. The same situation was with E1 in the group produced on the base of cultivar Zolotysta. Inside the group of genotypes created on the base of cultivar Oksana dominant E7 allele signifi cantly decreased yield and did not signifi cantly affect other traits. Also we observed difference in yield between both genotypes with E7 Oksana and Oksana M12.
It should be noted that in our three year fi eld experiment cultivar Zolotysta, which belongs to maturity group 00, was earlier among the genotypes tested but its mutant lines have a tendency to be later. Oksana was the latest genotype, which we included to maturity group I. This cultivar has mutant derivates with a shorter LV on 18 days. For cultivar Zolotysta the length of the vegetative period was shorter for 28 days (almost 1 month) as compared to that of Oksana.

CONCLUSIONS
Our results show that the mutagens D-6, DMSSO-11, DMSSO-12, DMSNPIR-11, DUDMS12, D12DMC-11B induced changes in soybean genome. By using these mutagens, it is possible to effectively increase genetic diversity in loci associated with the genes that determined photoperiod sensitivity of soybean. The obtained mutant lines could be involved in the breeding of soybean cultivars with different levels of photoperiodic sensitivity, terms of maturity, length of vegetation period and therefore adaptation ability. For example, mutant lines created on the basis of variety Oksana have signifi cantly shorter LV (P = 0.05 and P = 0.01) for about 15-18 days than Oksana because of the earlier maturity. Line Femida M29 reached maturity significantly later (P = 0.05) then parental cultivar Femida.
The three year observations of the 10 mutant lines, parental cultivars and controls obtained under fi eld conditions of the Vinnitsa region of Ukraine permit to make the following conclusion: the duration of the shootsfl owering (S-F) period for soybean lines with E7 and e7 alleles differed not signifi cantly, but the lines with an allele of 175 bp at locus Satt319 the S-F period was 6-9 days shorter. Lines with e7 and an allele of 175 bp at locus Satt319, however, did not show differences in LV and DTM. For soybean genotypes with the E7 allele the DTF was longer for 3-9 days and LV for 10-11 days. We can conclude, that the changes in DTF and LV, which we have detected for mutant lines in comparison with parental cultivars, are on the measure of variation in MG groups and between the MGs, when we compared with the studies Liu et al. [25], but if the IFAP's soybean mutant lines is grown in other ecological condition the differences in DTM can became signifi cant. In general, mutant lines that have been developed in the IFAP are interesting not only for breeding process, but also for investigation of molecular mechanism of changes in the soybean genome that permit to create lines with different times of maturity and LV.