Inheritance of Upper Node Tillering in Rice
Tsuneo Kato
Faculty of Biology-Oriented Science and Technology, Kindai University (930 Nishimitani, Kinokawa, Wakayama, 649-6493 Japan)
Summary: In rice (Oryza sativa L.), tiller buds on the uppermost and/or a few successive phytomers in a stem usually show dormancy, whereas those on phytomers on basal parts develop tillers and contribute to the determination of panicle number, and fi nally yield. However, the tiller buds on upper phytomers can eventually develop in response to genetic and non-genetic factors and generate tillers from upper node (upper node tillers, UNTs). In a population of recombinant inbred lines (RILs) derived from a cross between wild-type parents, Nakateshinsenbon and Milyang 23, the occurrence of upper node tillering (UNTing) was apparently segregated among the RILs at a ratio of 3 wild-type: 1 UNTing type, suggesting some genetic control for UNTing. The present study was conducted to identify the mode of inheritance of UNTing using this RIL population and several segregating populations after crossing of the RILs and the parents. The results suggested strongly that UNTing expression was controlled by two independent loci, UNT1 and UNT2, and also moderately infl uenced by non-genetic factors, resulting in inconsistency of expression among the RILs between growing years. UNTing should be expressed when the two recessive alleles, unt1 and unt2, are present simultaneously at their two respective loci. Further studies are needed to defi ne these duplicate recessive alleles for UNTing, their genome location and physiological function, toward understanding the mechanism of tiller development.
Keywords: duplicate action of two recessive alleles, phytomer, rice, tillering, upper node
Introduction
In rice (Oryza sativa L.) and many other cereals, a stem or shoot consists of a number of phytomers, which are successively piled up under a shoot apical meristem. A phytomer in rice consists of a leaf blade, a leaf sheath, a culm, lateral root primordia and a tiller bud. Ordinary tiller buds, which are located at the base of every phytomer, are activated, with outgrowth and result in tillers, only at the non-elongating phytomers located at the base of the stem. On the other hand, in the uppermost and/or a few successive phytomers from the top, the tiller buds show dormancy and usually do not generate any tillers.
In some cases, however, the tiller buds of phytomers on upper positions in a stem develop and generate tillers. In this study, these tillers are designated as upper node tillers (UNTs) and the occurrence of these tillers as upper node tillering (UNTing) . This UNTing is not a rare phenomenon in rice and has been reported in several rice studies: in ratoons after early cutting of hills (Gotoh and Hoshikawa 1988), in several very early heading cultivars ( Matsuba 2003), in some direct-seeded well-drained paddy fields ( Okabe et al. 2005), particularly in the low establishment of hills in these fi elds (Nagoshi et al. 2010). Kato
(2014) showed that this UNTing could be induced nearly every time in a stem when the panicle on the stem was deleted immediately after heading. In addition, a mutant line of rice showing UNTing was isolated after a chemical mutagen treatment (Hobo et al. 2011).
The underlying mechanisms for UNTing, however, are not fully understood. The panicles on these UNTs emerged much later than ordinary panicles and mostly showed sterility and no additional or significant contribution to grain production. However, understanding of the UNTing mechanism, namely to clarify why the dormant upper node tiller buds can be activated, could lead to the understanding in ordinary cases of why these upper node tiller buds are suppressed, why only the tiller buds at non-elongating basal nodes are activated, and how to regulate the panicle numbers in a rice plant. This could be a signifi cant viewpoint of studies on UNTing.
In a rice population of recombinant inbred lines (RILs) derived from Nakateshinsenbon/Milyang 23, UNTing was clearly segregated among the RILs (Kato 2014). The present study fi rst examined the degree of inheritance of UNTing in this segregating RIL population over several generations in different growing years, and then analyzed the mode of inheritance for UNTing using several other segregating populations after crossing lines and/or cultivars showing different degrees of UNTing expression. The objective of this study was to evaluate
Acccepted: September 27, 2018
Corresponding author: Tsuneo Kato ([email protected])
the mode of inheritance of UNTing. The locus/loci for the expression of UNTing, if identifi ed, could contribute to the full understanding of the physiological mechanism for this character.
Materials and Methods
A total of 97 RILs derived from a cross of rice (Oryza sativa L.) cultivars, Nakateshinsenbon (japonica-type)/Milyang 23 (indica-type), were used, because the apparent segregation for UNTing has been observed among the RILs. To examine the consistency of UNTing expression in the RILs between different growing years, this population were cultivated from F12 to F16 generations (from 2013 to 2017), together with their parents. In every growing year, they were sown in nursery boxes fi lled with fertilized soil in early May, and grown in a glasshouse during about one month. In early June, they were transplanted into a paddy field of the Faculty of Biology-Oriented Science and Technology, Kindai University, Kinokawa, Wakayama, Japan. Each RIL and parent consisted of six plants. Inter-plant and inter-row spacing were 15 cm and 30 cm, respectively. Fertilizer was applied at the time of transplanting at a rate of N:P2O5:K2O = 6:6:6 g m-2. Ordinary cultivation practices for irrigation and pest
management were done after transplanting.
In early September (about 30 days after heading for most of the RILs) in every growing years, the expression of UNTing was judged in each RIL with the following criteria: an UNTing RIL was defi ned as a RIL in which more than half of plants within the RIL had UNTs emerging from lamina joints of the fi rst and/ or second leaves from the top. Wild-type (WT, non-UNTing) RIL was defi ned as a RIL showing no UNTs in any stems within the RIL. The parents were also examined for their UNTing using the same criteria. As a control character for UNTing, spreading stub, which also segregated among the same RIL population, was examined in 2015, 2016 and 2017. For this trait, the degree of spreading stub for each RIL was classified as erect (non-spreading) type similar to Nakateshinsenbon or spreading-stub type similar to Milyang 23. No apparent segregation for this trait was observed within the RIL.
After the observation for UNTing over several growing years, two RILs (#209 and #222) and one RIL (#217) expressing consistently UNTing and WT, respectively, were selected as the parents of the following crossing experiment, together with the parents of the RILs. In 2014, eight crossings were carried out: #209 (UNTing)/#217 (WT), Nakateshinsenbon (WT)/#209, # 2 0 9 / M i l y a n g 2 3 ( W T ) , # 2 2 2 ( U N T i n g ) / # 2 1 7 , Nakateshinsenbon/#222, #222/Milyang 23, #222/#209, and Nakateshinsenbon/Milyang 23. F1s in 2015 and F2 populations in 2016 from the crossings, which consisted of six and about 150 plants, respectively, were grown in the same paddy field, together with the parents using the same cultivation conditions,
as described above. For the four cross combinations of Nakateshinsenbon/#209, #209/Milyang 23, #222/#209 and Nakateshinsenbon/Milayng 23, F2:3 lines randomly selected from the respective F2 populations were also cultivated using the same condition as above for pedigree tests. In early September in each growing year, every plant was judged individually for the expression of UNTing, in which an UNTing plant was defi ned as a plant with emerging UNTs from the lamina joints both of the fi rst and second leaves from the top.
To evaluate preliminary the effects of UNTing on other agronomic traits, particularly on the number of panicles per plant, the RIL population grown in 2017 were measured for their number of panicles per plant excluding UNTs, the longest culm length in a plant and length of the panicle on the longest culm, for four plants at the central part of the individual RILs. Two RILs were omitted because they consisted of less than four plants. In every trait, an analysis of variance in one-way classifi cation with four replications (four plants) were conducted. Sum of squares for RILs (df=94) (this was highly signifi cant in every trait) was divided into that of comparison between WT-RIL group and UNTing-WT-RIL group (df=1) and residual sum of squares after subtracting that of the comparison (df=93) (within-group sum of squares), and the mean of squares for the comparison was divided by that of within-group, and that of the within-group by that of residual (df=285), to obtain F-values.
Results and Discussion
Table 1 shows the differences in the number of panicles per plant, culm length and panicle length between the means of WT-RIL group and UNTing-WT-RIL group grown in 2017. In every trait, means of both groups were not signifi cantly different, while the within-group variations were highly significant. Although the effects of UNTing on other traits should be examined using near-isogenic lines or other appropriate materials, the present preliminary result indicated that UNTing could not affect strongly the expression of other traits.
To estimate the degree of inheritance of the UNTing in rice, the consistency of UNTing expression in the same RILs between two growing years was tested statistically using a two-by-two contingency table. This contingency table compared the numbers of RILs with and without UNTing expression between two different growing years. If there was no tendency of inheritance for UNTing (no consistency between years), we can detect the fitness between observed values and expected values from the assumption of independence. If, on the other hand, UNTing was completely inherited, the RILs should show an identical expression of UNTing in both years. Therefore, a χ2 test for
independence with one degree of freedom was conducted based on the contingency table. Table 2 shows the results of χ2 values
and their probabilities in every combination over fi ve growing years. This table also shows consistency indices (CIs), which are the ratio of the number of RILs showing the same expression for UNTing in both years against the total number of RILs. We can, thus, evaluate a kind of heritability of UNTing from this index. In the case of complete inheritance, the CI value should be unity, whereas in the case of no-inheritance, the CI value (the minimum value) should be p2 + (1-p)2, where p is an average frequency of
individuals with the target phenotype randomly generated by non-genetic factors.
Table 2 shows that χ2 values for the contingency tables were
all signifi cant at the 0.01 or much lower probability level, and the CIs ranged from 0.722 to 0.784. This result strongly indicated that the expression of UNTing was inherited and dependent on the RILs, although non-genetic factors also affected the result. To compare the degree of the inheritance for UNTing with another trait, the same procedure was applied to spreading stub in the same RIL population. In comparison with UNTing, Table 3 shows that higher χ2 values and lower levels of
probability were obtained in all combinations of growing years for this trait. The CIs ranged from 0.804 to 0.866. Therefore, it was suggested that the expression of UNTing was more sensitive to non-genetic factors than that of spreading stub, which is regulated by a dominant allele on a single locus, TAC1 (Yu et al. 2007) in Milyang 23.
In every growing year, the ratio of WT RILs to UNTing RILs fitted significantly to a 3:1 ratio at the 0.01 probability level, with non-signifi cant heterogeneity among years (Table 4). Since there are generally rare lines involving heterozygotes in a RIL population, the 3:1 ratio is apparently expected from a genetic model that two independent loci are involved in the regulation of UNTing, with some genetic interaction between these two loci. For the spreading-stub phenotype in the same RIL population as a control, the ratio of erect-type RILs to spreading-stub RILs were signifi cantly fi tted to a 1:1 ratio, which is expected from the segregation in the single locus, TAC1 (data not shown). For
,WHP 3DQLFOHVSODQW &XOPOHQJWK FP 3DQLFOHOHQJWK FP 0HDQRI:75,/JURXS 0HDQRI817LQJ5,/JURXS FYDOXHIRU:7YV817LQJ FYDOXHIRUZLWKLQJURXS <HDU <HDU
Table 1. Diff erences in the number of panicles per plant, culm length and panicle length between the means of WT-RIL group and UNTing-RIL group, and an analysis of variance for between-group and within-group variation
Table 2. Consistency of the expression of upper node tillering in rice recombinant inbred lines from Nakateshinsenbon/Milyang 23 in every combination across diff erent growing years
Table 3. Consistency of the expression of spreading stub in rice recombinant inbred lines from Nakateshinsenbon/Milyang 23 in every combination across growing years, as a control for upper node tillering
Values in ( ) indicate the probability for F-values.
1) This was calculated from dividing mean of squares for the comparison between wild-type-RIL group and UNTing-RIL
group by that for within-group variation.
2) This was calculated from dividing mean of squares for within-group variation by residual mean of squares.
Upper, middle and lower lines in each combination of growing years
indicate χ2 value of independence in the contingency table, its P value
and consistency index (CI), respectively. See text for the CI.
Upper, middle and lower lines in each combination of growing years have the same meaning as in Table 2.
the expected minimum CIs for UNTing and spreading stub, thus, these can be estimated as 0.625 and 0.500, respectively
The results of F1 plants and F2 populations derived from eight cross combinations among WTs and UNTing types (Table 5) clearly indicated that two independent loci were involved in UNTing expression, as expected from the segregation in RILs. Moreover, it was clearly demonstrated that UNTing was expressed only when two recessive alleles were present simultaneously at the respective loci. These two loci were tentatively designated as UNT1 and UNT2, and the respective recessive alleles with duplicate action as unt1 and unt2. This is typically appeared from the results of the pedigree of Nakateshinsenbon/Milyang 23, i.e., the F1 plants were WT, and the F2 showed segregation of WT:UNTing signifi cantly fi tted to a ratio of 15:1 (Table 5). Therefore, the genotype of UNTing RILs, #209 and #222, should be both unt1unt1unt2unt2, and the wild-type parents, Nakateshinsenbon and Milyang 23, were Unt1Unt1unt2unt2 and unt1unt1Unt2Unt2 or vice versa. In addition, the facts that all F2 segregations of WT:UNTing significantly fitted to a ratio of 3:1, when the parents were crossed with UNTing RILs (Table 5), clearly consistent with the model of duplicate recessive alleles for this trait. The genotype of the WT RIL, #217, was expected as Unt1Unt1unt2unt2 or unt1unt1Unt2Unt2, because of the WT F1 phenotypes and F2 segregation ratios of 3:1, when this WT RIL was crossed with UNTing RILs (Table 5). The genotype of the #217 will be fi nally determined from the results of the crosses with the two parents.
The results of F2:3 lines strongly supported the above mode of inheritance for UNTing: The segregating ratio of 7:8:1 and the ratio of 1:2:1 for WT fi xed line:segregating line:UNTing fi xed line, were obtained from WT/WT (Unt1Unt1unt2unt2/ unt1unt1Unt2Unt2) and WT/UNTing (Unt1Unt1unt2unt2/ unt1unt1unt2unt2 or unt1unt1Unt2Unt2/unt1unt1unt2unt2) cross combinations, respectively (Table 6). In addition, no segregation of WT F2:3 line was also confi rmed the genotypes of the parental RILs in UNTing/UNTing (Table 6). The ratio of 7:8:1 from the
cross between original parents was expected from that the genotypes and their frequencies in F2 parents should be Unt1Unt1Unt2Unt2 (1/16), Unt1Unt1unt2unt2 (1/16), unt1unt1Unt2Unt2 (1/16), Unt1Unt1Unt2unt2 (2/16) and U n t 1 u n t 1 U n t 2 U n t 2 ( 2 / 1 6 ) f o r W T f i x e d F2 : 3 l i n e s , Unt1unt1unt2unt2 (2/16), unt1unt1Unt2unt2 (2/16) and Unt1unt1Unt2unt2 (4/16) for segregating lines, and finally unt1unt1unt2unt2 (1/16) for UNTing fixed lines. On the other hand, the phenotypes of individual F2 plants were not completely consistent with the segregating patterns found in their F2:3 progenies, e.g., UNTing fi xed F2:3 lines sometimes derived from WT F2 plants. This might be due to environmental fl uctuation of UNTing expression, particularly in F2 individuals, as observed with the lower consistency between growing years (Table 2).
The physiological mechanism underlying the expression of UNTing in rice is not yet understood. Kato (2014) conducted decapitation experiments in rice where the panicles were cut at fi ve days after heading. These treated stems inevitably generated UNTs not only in UNTing RILs but also in WT RILs or cultivars. However, the contents of non-structural carbohydrates (NSC) in upper part of a stem did not signifi cantly increase after panicle decapitation in any RILs. Therefore, the expression of UNTing after panicle decapitation could not result from the increase in NSC (Kato unpublished data). In addition, the NSC content in a stem was not associated with the degree of UNTing in non-treated materials (Kato 2014).
Another possible factor may be phytohormones. Cytokinins are generally required when lateral buds in a plant shoot, like tiller buds in rice, develop and elongate ( Kurokawa et al. 2007). On the other hand, auxin distributed from the apical part of a shoot clearly inhibits the growth of lateral buds, particularly those in the upper part of the shoot ( Nordstrom et al. 2004). This is well known as a cause of apical dominance (Taiz and Zeiger 1998) or primigenic dominance (Bangerth 1989), and probably underlies the suppression of UNTing in ordinary WT rice. Recently, another phytohormone, strigolactone, has also been
1XPEHURI5,/V <HDU :7 817LQJ Ȥ P +HWHURJHQHLW\
Table 4. Segregation for upper node tillering (UNTing) in the recombinant inbred lines (RILs) of rice from Nakateshinsenbon/ Milyang 23 over fi ve growing years and their heterogeneity
Total number of RILs were all 97. WT indicates non-UNTing (wild-type).
shown to be involved in the suppression of lateral buds ( Umehara et al. 2000, Gomez-Roldan et al. 2008, Brewer et al. 2009, Ferguson and Beveridge 2009, Dun et al. 2012). The present results strongly indicated that two recessive alleles, unt1 and unt2, at independent two loci, UNT1 and UNT2, could cause the expression of UNTing in a duplicate recessive genotype, although the degree of inheritance of this phenotype was not high. These two loci have not yet determined in detail in term of their genomic locations or nucleotide and amino acid sequences, etc. However, it is plausible that these two loci regulate the metabolism of a phytohormone, auxin, strigolactone, cytokinins, and others. In rice, several loci that affected the development and growth of tiller buds contributed to the biosynthetic pathways of phytohormones, for example, D10 ( Arite et al. 2007), DWARF27 ( Lin et al. 2009) and OsTB1 (Takeda et al. 2003, Minakuchi et al. 2010). However, all of these previous studies addressed the growth of ordinary tiller buds at the basal part of a stem, not UNTing. Further studies will be needed to clarify the understanding of UNTing in rice, both from genetic and physiological approaches. This could also lead to the understanding of ordinary tillering mechanisms in rice.
Acknowledgements
I wish to express sincere thanks to Dr. A. Horibata, the Faculty of Biology-Oriented Science and Technology, Kindai University to perform this study. This study was supported by a Grant from the Faculty of Biology-Oriented Science and Technology, Kindai University, 13-IV-1.
References
Arite, T., H. Iwata, K. Ohshima, M. Maekawa, M. Nakajima, M. Kojima, H. Sakakibara and J. Kyozuka (2007) DWARF10, an RMS1/MAX4/DAD1 ortholog, controls lateral but outgrowth in rice. The Plant J. 51: 1019-1029.
Bangerth, F. (1989) Dominance among fruits/sinks and the search for a correlative signal. Physiol. Plant. 76: 608-614. Brewer, P. B., E. A. Dun, B. J. Ferguson, C. Rameau and C. A.
Beveridge (2009) Strigolactone acts downstream of auxin to regulate bud outgrowth in pea and Arabidopsis. Plant Physiol. 150: 482-493.
Dun, E. A., A. de Saint Germain, C. Rameau and C. A. Beveridge (2012) Antagonistic action of strigolactone and
1XPEHURI)SODQWV &URVVFRPELQDWLRQ ) :7 817LQJ 7RWDO Ȥ P 817LQJ:7 :7 1. :7817LQJ :7 0/ 817LQJ:7 :7 817LQJ:7 :7 1. :7817LQJ :7 0/ 817LQJ:7 :7 817LQJ817LQJ 817LQJ 1.0/ :7:7 :7 1XPEHURI) OLQHV
&URVVFRPELQDWLRQ :7IL[HG 6HJUHJDWHG 817LQJIL[HG 7RWDO Ȥ P
1.0/ :7:7 1. :7817LQJ 0/ 817LQJ:7 817LQJ817LQJ
Table 5. Expression and segregation for upper node tillering (UNTing) in F1 plants and F2 populations from several cross
combinations among wild-type (WT) and UNTing parents
Table 6. Segregation for upper node tillering (UNTing) in the F2:3 lines from several cross combinations, as detailed in Table 5
#209, #217 and #222 were RILs from Nakateshinsenbon (NK)/Milyang 23 (ML).
cytokinin in bud outgrowth control. Plant Physiol. 158: 487-498.
Ferguson, B. I. and C. A. Beveridge (2009) Roles for auxin, cytokinin, and strigolactone in regulating shoot branching. Plant Physiol. 149: 1929-1944.
Gomez-Roldan, V., S. Fermas, P. B. Brewer, V. Puech-Pages, E. A. Dun, J. P. Pillot, F. Letisse, R. Matusova, S. Danoun, J. C. Portais, H. Bouwmeester, G. Becard, C. A. Beveridge, C. Rameau and S. Rochange (2008) Strigolactone inhibition of shoot branching. Nature 455: 189-194.
Lin, H., R. Wang, Q. Qian, M. Yan, X. Meng, Z. Fu, C. Yan, B. Jiang, Z. Su, J. Li and Y. Wang (2009) DWARF27, an iron-containing protein required for the biosynthesis of strigolactones, regulates rice tiller bud outgrowth. Plant Cell 21: 1512-1525.
Gotoh, Y. and K. Hoshikawa (1988) Studies on the regrowth of rice shoots II. On characteristics of the new effective tillers appeared after cuttings. Jpn. J. Crop Sci. 57: 59-64 (in Japanese).
Hobo, T., Y. Nagato and H. Kitano (2011) Phenotypic characterization of a rice mutant NM3-631 expressing hyper-tillering in the late vegetative and reproductive stages. Breed. Res. 13 (Suppl. 2): 227 (in Japanese).
Kato, T. (2014) Occurrence and inheritance of upper nodal tillering in rice recombinant inbred lines. Jpn. J. Crop Sci. 83 (Suppl. 1): 154-155 (in Japanese).
Kurokawa, T., N. Ueda, M. Maekawa, K. Kobayashi, M. Kojima, Y. Nagato, H. Sakakibara and J. Kyozuka (2007) Direct control of shoot meristem activity by a cytokinin-activating enzyme. Nature 445: 652-655.
Matsuba, K. (2003) Tillering system of extremely early rice varieties - especially development of upper nodal tillers. Jpn.
J. Crop Sci. 72: 62-67 (in Japanese).
Minakuchi, K., H. Kameoka, N. Yasuno, M. Umehara, L. Luo, K. Kobayashi, A. Hanada, K. Ueno, T. Asami, S. Yamaguchi and J. Kyozuka (2010) FINE CULM1 (FC1) works downstream of strigolactones to inhibit the outgrowth of axillary buds in rice. Plant Cell Physiol. 51: 1127-1135.
Nagoshi, T., R. Uchida, F. Tamai, S. Hirano, T. Hirose, Y. Motoda and M. Fukuyama (2010) Productivity of upper nodal tillers of rice plants compared to those of mother stems under low establishment density in direct sowing. Jpn. J. Crop Sci. 79: 424-430 (in Japanese).
Nordstrom, A., P. Tarkowski, D. Tarkowska, R. Norbaek, C. Astot, K. Dolezal and G. Sandberg (2004) Auxin regulation of cytokinin biosynthesis in Arabidopsis thaliana: A factor of potential importance for auxin-cytokinin-regulated development. Proc. Natl. Acad. Sci. USA 101: 8039-8044. Okabe, M., F. Tamai, Y. Motoda, T. Nagoshi and G. Takeda
(2005) Effect of tillering and seeding methods on seedling emergence, growth and yield of rice direct-seeded in well-drained paddy field. Jpn. J. Crop Sci. 74: 125-133 (in Japanese).
Taiz, L. and E. Zeiger (1998) Plant Physiology 2nd ed. , Sinauer
Associations Inc. Publishers, Sunderland, Massachusetts, USA.
Takeda, T., Y. Suwa, M. Suzuki, H. Kitano, M. Ueguchi-Tanaka, M. Ashikari, M. Matsuoka and C. Ueguchi (2003) The OsTAB1 gene negatively regulates lateral branching in rice. The Plant J. 33: 513-520.
Umehara, M., A. Hanada, H. Magome, N. Takeda-Kamiya and S. Yamaguchi (2000 ) Contribution of strigolactones to the inhibition of tiller bud outgrowth under phosphate defi ciency in rice. Development 127: 4971-4980.
