Acccepted:May 22, 2017
Corresponding author: Tsuneo Kato ([email protected])
Inheritance of a Novel Panicle Type, ‘Short Tip Panicle’, found in a Rice Cultivar Gimbozu
Tsuneo Kato
Faculty of Biology-Oriented Science and Technology, Kindai University （930 Nishimitani, Kinokawa, Wakayama 649-6493, Japan）
Summary: A novel panicle type short tip panicle (STP) in rice was characterized by its short primary branches and fewer spikelets on primary branches only at the top of the panicle, compared with on other branches. The objective of this study was to examine the mode of inheritance of STP. To examine STP quantitatively, the mean spikelet number per primary branch on the top two primary branches (STP value) was measured as an index of the degree of STP. To search for variation in STP values among rice cultivars, 41 cultivars were examined. Only the japonica-type cultivar Gimbozu showed values consistent with STP, indicating that this cultivar was unique
in its STP expression. F1 plants, F2 populations and F2:3 lines were generated from crosses in a half-diallel mode
of Gimbozu (STP)/Chugoku 117 (wild type), Gimbozu/RY50 (wild type) and Chugoku 117/RY50. The results strongly suggested that the STP phenotype was regulated at a dominant allele, Stp, at STP locus, and the expression of Stp was inhibited by an incomplete dominant allele, I-Stp, at an independent I-STP locus. Therefore, Gimbozu, Chugoku 117 and RY50 have genotypes of StpStp i-Stpi-Stp, stpstp I-StpI-Stp and stpstp i-Stpi-Stp, respectively. This STP phenotype could contribute to understand genetic regulation of panicle architecture in rice.
Key words: panicle architecture, rice, inhibitor, segregation analysis, primary branch
Introduction
The japonica-type rice (Oryza sativa L.) cultivar, Gimbozu, was developed through pure-line selection by Mr. I. Ishiguro from an indigenous Japanese cultivar, Aikoku, in early the 1900 s in Toyama Prefecture, Japan (https://www.jataff.jp/senjin4/8. html). Gimbozu has been used as a parent of many cross-breeding programs in Japan, resulting in major Japanese cultivars, e.g., Norin 8, Norin 22, Koshihikari and Sasanishiki, in the progenies. From the viewpoint of molecular biology, Gimbozu contains a non-autonomous, Type-II transposon mPing with several hundred copies in the rice genome, as well as several copies of autonomous Ping and Pong, which are the sources of transposase for mPing (Nakazaki et al. 2003, Jiang et al. 2003, Kikuchi et al. 2003). These transposons are still active not only in rice genome but also in Arabidopsis (Yang et al. 2007), and they are available as useful tools for transposon tagging, which can contribute to studies of functional genomics of rice.
Gimbozu shows mid or late maturing, medium plant mass, resistant against rice blast, and it has been a leading cultivar in the 1930 s. In addition, Gimbozu, at least in the material examined in this study, also showed a peculiar panicle morphology: its primary branches at the top and the second top
of a panicle have fewer spikelets and were shorter than the other primary branches in the same panicle (Fig. 1). I termed this characteristic Short Tip Panicle (STP) . STP may be controlled by genetic factors, because many of the progenies derived from Gimbozu, such as near-isogenic lines for grain length under the background of Gimbozu (Kato 2010), also show STP, whereas the expression of STP might be influenced by non-genetic factors.
Fig. 1. Tip of a rice panicle of (A) Gimbozu (STP), (B) Chugoku 117 (wild-type) and (C) RY50 (wild-type). Arrows indicate the top two primary branches measured for their spikelet numbers.
The present study tried to clarify the mode of inheritance of STP using segregating populations derived from a half-diallel crossing among of three genotypes involving Gimbozu. Although STP itself could not contribute directly to higher yielding ability in rice, due to the decreased spikelet number, this characteristic could be important in understand rice panicle morphogenesis, particularly in the formation of primary
branches. Hence, this information might indirectly contribute to higher yielding ability through the development of better panicle type.
Materials and Methods
To deal with the STP phenotype quantitatively, an index of the degree of STP was measured in this study as the average number of spikelets on primary branches per primary branch located at the top and the second top of a panicle (STP value). A total of 41 cultivars involving Gimbozu, which are conserved in the Plant Breeding Laboratory, Faculty of Biology-Oriented Science and Technology, Kindai University (Table 1), were sown in nursery boxes in early May, 2014, grown in a greenhouse, and transplanted in the experimental paddy field of the Faculty of Biology-Oriented Science and Technology in mid June. Each cultivar was planted as a single plant per hill at a density of 15 cm inter-hill and 30 cm inter-row. Fertilizers were applied only as basal dressing at a rate of 6:6:6 g m-2. Usual cultivation
procedures, e.g., pest and irrigation control, were conducted. Panicles on the tallest culm of fi ve plants for every cultivar were measured to determine their STP values after harvesting.
Table 1. Rice cultivars and their STP values examined in the present study.
&XOWLYDU 673YDOXH &XOWLYDU 673YDOXH &XOWLYDU 673YDOXH *LPER]X .DUDVXPRFKL -DNRX +DEDWDNL 'DZ'DP 1DQMLQJNRKWR% $]XFHQD 6DEDUPDWL $LNRNX +DPDVDUL $NHQRKRVKL 0RXUHWVX ,5 +LURVKLPD +LHUL $NLVKLQRPRFKL 0DQJRNXPRFKL ,5 .\XVKX8QLY %LQLFRO 1DNDWHVKLQVHQERQ 6XUMDPNKL 0XVNDNXGDQWL 0RPLURPDQ %HLMLQJNRKWR %DVPDWL 1DQMLQJ 7DQFKRPRFKL .RVKLKLNDUL ,5 +DWWDQQLVKLNL )XVD\RVKL 6HVLD 7DNDQDUL 0L\DNDRUL 0LO\DQJ 0DQJRNXXUXFKL 6DVDQLVKLNL 2KLWDNRKWR 2WVXQH *XL&KDR .DEDVKLNR %*
Three cross combinations in a half-diallel mode were made in 2011: Gimbozu (STP)/Chugoku 117 [non-STP wild type (WT)], Gimbozu/RY50 (WT), and Chugoku 117/RY50. Of these, Chugoku 117 was developed at the Western Agricultural Research Center, NARO, as a high yield, strong lodging resistant cultivar, and selected as a parent for the feed-rice cultivar Leaf Star. RY50 is a derived line from Gimbozu, and one of the lines presumably experienced mPing transposition. Chugoku 117 and RY50 have gh2 alleles at OsCAD2 locus (Zhang et al. 2006), and express gold hull and internode phenotype, whereas Gimbozu does not. Originally, these three crosses were part of an allelism test for gh2 alleles between Chugoku 117 and RY50.
an F2 population of 100 plants per cross combination, and F3
lines (F2:3) of 12 plants derived from F2 plants were grown in
2012, 2013, and 2014, respectively. The growing conditions were the same as for the 41 cultivar evaluation in 2014 described above. After maturation, the STP value was measured for the panicle on the tallest culm of every plant. For Chugoku 117/
RY50, ten F3:4 lines, which were derived from fi ve F3 lines, with
the fi ve top STP mean values by single seed decent and fi ve lines showing the lowest STP mean values, were also grown in 2015 to evaluate the selection response for STP value in this cross
combination. These F3:4 lines consisted of 12 plants.
Results and Discussion
Figure 2 shows the frequency distribution of STP values among 41 rice cultivars examined. Only one cultivar, Gimbozu, exhibited a distinctly lower mean STP value of 3.1, with a standard deviation of 0.42, whereas other cultivars showed a normal distribution with a mean of 5.6 and standard deviation of 0.51. The difference between Gimbozu and Habataki which showed the second smallest value (mean 4.7) among the cultivars (Table 1), was significant (t=5.84, df=8, P=0.0004). Therefore, Gimbozu was a unique cultivar in terms of STP among the cultivars examined.
Fig. 2. Frequency distribution of STP value among 41 rice cultivars.
Figure 3 shows the frequency distributions of STP values of
the three segregating populations, involving F1, F2 and F2:3
generations from the same cross combination, together with their
parents. F2 populations from Gimbozu/Chugoku 117 and
Gimbozu/RY50 showed at least two peaks in their frequency
distributions. In the F2 populations from Gimbozu/Chugoku 117,
the group with the lower STP values was less than the other
group with higher STP values. On the contrary in the F2
population from Gimbozu/RY50, the lower-STP group was more
STP phenotype. In the F2 population from Gimbozu/Chugoku
117, one more additional peak was observed within the
higher-STP group. The mean higher-STP value of F1 plants from Gimbozu/
Chugoku 117 was intermediate between the parents, whereas that from Gimbozu/RY50 was similar to that of Gimbozu.
Fig. 3. Frequency distribution of STP value in F2 plants derived from crosses of rice cultivars/line between Gimbozu and Chugoku 117, Gimbozu and RY50 and Chugoku 117 and RY50. STP values of F1 plants and parents are shown as arrows (horizontal bars indicate S.D.). The data of F2:3 lines are also overlaid on the respective F2 distributions.
These results of F1 plants and F2 populations from the three
cross combinations strongly suggested that two loci were responsible to the expression of STP: a complete dominant allele (hereafter designated as Stp) at one locus (STP) essentially caused the STP phenotype, and an incomplete dominant allele (I-Stp) at the other independent locus (I-STP) inhibited the
expression of Stp, because the STP values of F1 plants of
Gimbozu/Chugoku 117 and Gimbozu/RY50 were intermediate between those of the parents and the same as Gimbozu, respectively. According to this two-locus model with one inhibitor, the genotypes of Gimbozu, Chugoku 117 and RY50 were postulated as StpStp i-Stpi-Stp, stpstp I-StpI-Stp and stpstp i-Stpi-Stp, respectively. Table 2 shows the expected genotypes
for STP phenotype in the F1, F2 and also F2:3 generations, and
their expected segregation ratios, derived from these three parents.
In this table, the phenotypes of F2 plants were classifi ed as WT,
WT , and STP depending on the expected genotypes. In addition,
F2:3 lines were classifi ed as four types depending on their expected
segregating patterns: fixed as WT (WT type), segregating (STP was more than WT) (SegI type), segregating (STP was less than WT) (SegII type), and fi xed as STP (STP type).
Table 2. Expected genotypes and phenotypes for STP in F1 plants, F2 populations, and F2:3 lines from (A) Gimbozu/Chugoku 117, (B) Gimbozu/RY50, and (C) Chugoku 117/RY50, according to the two-locus model involving one inhibitor for this trait.
$ *LPER]X StpStp i-Stpi-Stp &KXJRNX stpstp I-StpI-Stp )SODQW )SRSXODWLRQ ) OLQH
*HQRW\SH 3KHQRW\SH *HQRW\SH 3KHQRW\SH *HQRW\SH 3KHQRW\SH
Stpstp I-Stpi-Stp :7¶ StpStp I-StpI-Stp :7 >Stp I-Stp@ :7IL[HG StpStp I-Stpi-Stp :7¶ >Stp I-Stp@ >Stp i-Stp@ 6HJ,, StpStp i-Stpi-Stp 673 >Stp i-Stp@ 673IL[HG Stpstp I-StpI-Stp :7 >Stp I-Stp@ >stp I-Stp@ :7IL[HG Stpstp I-Stpi-Stp :7¶ >Stp I-Stp@ >Stp i-Stp@ >stp I-Stp@ >stp i-Stp@ 6HJ,, Stpstp i-Stpi-Stp 673 >Stp i-Stp@ >stp i-Stp@ 6HJ, stpstp I-StpI-Stp :7 >stp I-Stp@ :7IL[HG stpstp I-Stpi-Stp :7 >stp I-Stp@ >stp i-Stp@ :7IL[HG stpstp i-Stpi-Stp :7 >stp i-Stp@ :7IL[HG % *LPER]X StpStp i-Stpi-Stp 5< stpstp i-Stpi-Stp
)SODQW )SRSXODWLRQ ) OLQH
*HQRW\SH 3KHQRW\SH *HQRW\SH 3KHQRW\SH *HQRW\SH 3KHQRW\SH
Stpstp i-Stpi-Stp 673 StpStp i-Stpi-Stp 673 >Stp i-Stp@ 673IL[HG Stpstp i-Stpi-Stp 673 >Stp i-Stp@ >stp i-Stp@ 6HJ, stpstp i-Stpi-Stp :7 >Vtp i-Stp@ :7IL[HG & &KXJRNX stpstp I-StpI-Stp 5< stpstp i-Stpi-Stp
)SODQW )SRSXODWLRQ ) OLQH
*HQRW\SH 3KHQRW\SH *HQRW\SH 3KHQRW\SH *HQRW\SH 3KHQRW\SH
stpstp I-Stpi-Stp :7 stpstp I-StpI-Stp :7 >Vtp I-Stp@ :7IL[HG stpstp I-Stpi-Stp :7 >stp I-Stp@ >stp i-Stp@ :7IL[HG stpstp i-Stpi-Stp :7 >Vtp i-Stp@ :7IL[HG :7:7¶DQG673LQ)DQG)LQGLFDWHSKHQRW\SHVRIZLOGW\SHZLWKKLJKHU673YDOXHVZLOGW\SHZLWKORZHU673YDOXHVDQG673 UHVSHFWLYHO\ LQGLFDWHVWKHVHJUHJDWLRQUDWLR:7IL[HG6HJ,6HJ,,DQG673IL[HGLQ) LQGLFDWHOLQHVIL[HGDV:7VHJUHJDWLQJ 673 ZDVPRUHWKDQ:7 VHJUHJDWLQJ 673ZDVOHVVWKDQ:7 DQGIL[HGDV673UHVSHFWLYHO\>@LQ) LQGLFDWHJHQRW\SHFRUUHVSRQGLQJWKH UHVSHFWLYHSKHQRW\SH HJ>Stp I-Stp@LQFOXGHVJHQRW\SHVRIStpStp I-StpI-Stp Stpstp I-StpI-Stp StpStp I-Stpi-Stp DQG Stpstp I-Stpi-Stp
In the F2 population from Gimbozu/Chugoku 117, the
segregation ratio of phenotypes, i.e., STP:WT :WT (if the thresholds between STP and WT , and WT and WT were set at STP values of 3.0 and 5.0, respectively) should be 3:6:7 (Table
3), and the χ2 value for goodness of fi t was 2.45 (P=0.294) (Table
3). In the F2 population from Gimbozu/RY50, the segregation
ratio of SPT:WT :WT (if the threshold between STP and others
was set at 4.0) should be 3:0:1, and the χ2 value was 1.33
(P=0.248) (Table 3). In the F2:3 lines, the expected segregation
ratios of STP:SegI:SegII:WT in Gimbozu/Chugoku 117 and Gimbozu/RY50 were 1:2:6:7 and 1:2:0:1, respectively (Table 4).
The χ2 values for Gimbozu/Chugoku 117 and Gimbozu/RY50
were 3.97 (P=0.265) and 2.88 (P=0.237), respectively (Table 4).
In Chugoku 117/RY50, all F2:3 lines were fi xed as WT phenotype,
as expected from Table 2.
Table 3. Segregation of STP phenotypes in F2 populations from Gimbozu/Chugoku 117, Gimbozu/RY50, and Chugoku 117/ RY50, and their goodness of fi t test.
&URVV FRPELQDWLRQ
3KHQRW\SHRI)SODQWV 7RWDO Ȥ
GI P 673 :7¶ :7 *LPER]X 2EVHUYHG &KXJRNX ([SHFWHG *LPER]X 2EVHUYHG 5< ([SHFWHG &KXJRNX 2EVHUYHG 5< ([SHFWHG **LPER]X &&KXJRNX 55< 7KHWKUHVKROG673YDOXHVRI:7DQG:7¶DQG:7¶DQG673ZHUHVHWDWDQG UHVSHFWLYHO\
Table 4. Segregation of STP phenotypes in F2:3 lines from Gimbozu/Chugoku 117, Gimbozu/RY50 and Chugoku 117/ RY50, and their goodness of fi t test.
&URVV FRPELQDWLRQ 3KHQRW\SHZLWKLQDQ) OLQH 7RWDO Ȥ GI P 673 6HJ, 6HJ,, :7 *LPER]X 2EVHUYHG &KXJRNX ([SHFWHG *LPER]X 2EVHUYHG 5< ([SHFWHG &KXJRNX 2EVHUYHG 5< ([SHFWHG **LPER]X &&KXJRNX 55<
:7 IL[HG DV :7 QRQ673 6HJ, VHJUHJDWLQJ 673 ZDV PRUH WKDQ :7 6HJ,, VHJUHJDWLQJ 673ZDVOHVVWKDQ:7 673IL[HGDV673
To examine whether the segregation for the STP phenotype in
the F2 population from Chugoku 117/RY50 was not a genetic
nature, a parent-offspring correlation (between F2 plants values
and F2:3 lines means) was calculated as r=0.304 (P<0.01).
Although the difference between the means of F2:3 lines of the
top fi ve and bottom fi ve was signifi cant, the difference between
F3:4 lines was not (Table 5), indicating there was no genetic
nature for the variation in STP values in the derived population from Chugoku 117/RY50. In addition, a realized heritability for
the difference between those in F3:4) value was very low
(hR2=0.092). The significant parent-offspring correlation
coeffi cient was probably due to the large population size (df=98).
Table 5. Selection response experiment for STP values in F3 and F3:4 lines from Chugoku 117/ RY50.
Line Mean STP value t/F1) P Top five lines Bottom five lines
F3 5.67 4.41 19.183 <0.0001 F3:4 4.91 4.87 0.094 - 1)t and F were for the tests of F3 and F3:4 lines, respectively. F was calculated as the ratio of mean of squares of the comparison between the top five group and the bottom five group to that of within-group variation.
The STP phenotype was a unique character found in only one rice cultivar, Gimbozu, at least in the present search. Several reports have been published for genetic regulation of the numbers of primary branches and of secondary branches in a panicle (Kato 2004, Terao et al. 2010). However, STP expression is restricted only at the top of a rice panicle. On the other hand, in common wheat (Triticum aestivum L.), a similar spike characteristic could be cited as squareheaded spike , in which spikelets around the top-most rachis are arranged more compactly than those at other positions along the rachis. This character, which is found ubiquitously in many species and genus of wheat, is regulated by the Q locus at 5A chromosome. The Q allele at this locus plays an important role in domestication process of wheat, because of its pleiotropic effects on glume shape, tenacity, rachis fragility, etc. (Simons et al. 2006, Pereg et al. 2011). Simons et al. (2006) demonstrated that the Q allele has a similarity to members of the AP2 transcription factor. It could be inferred that the STP locus in rice also has an important role on agronomic performance, if it shows a similar function to the Q locus in wheat.
The results of F1 plants, F2 populations and F2:3 lines from the
present three cross combinations strongly suggested that the two-locus model involving one inhibitor was adequate to explain the expression of the STP phenotype. These two loci, however, have not yet been elucidated to determine their positions in rice genome or their gene structures. In the present cross combination of Gimbozu/RY50, alleles only at STP (Stp and stp) should be segregated. Because RY50 is a derived line from Gimbozu, which could experience the transposition of mPing, the STP locus, at least, might be searched for by the procedure of mPing transposon tagging (Yasuda et al. 2013, Horibata and Kato 2015). Further experiments will be needed for a full understand of the regulation of STP.
イネ品種銀坊主が示す新規穂型「先端短縮穂」の遺伝様式
加藤恒雄
近畿大学生物理工学部（〒 649-6493 和歌山県紀の川市西三谷 930）
要旨：イネにおいては，一部の品種で，穂の先端部分でのみ他の部分と比べて 1 次枝梗が短くかつ 1 次枝梗上穎花数が少ないこ とで特徴づけられる穂型が見られる．これを，新規の穂型，「先端短縮穂」（short tip panicle，STP）と呼び，本研究では，この STP の遺伝様式を検討した．STP の程度を定量化するため，穂先端 2 本の 1 次枝梗当たり平均 1 次枝梗上穎花数を STP 値として 測定した．まず，イネ 41 品種間での STP 値に関する変異を調査した．その結果，日本型品種銀坊主のみが特異的に STP を発現 することが判った．次に，銀坊主（STP）/ 中国 117 号（野生型），銀坊主 /RY50（野生型）および中国 117 号 /RY50 の 3 交雑組
み合わせに由来する F1個体，F2集団および F2:3系統を用いて STP 値の分離を検討した．その結果，STP 表現型は STP 座上の優
性アレル Stp によって制御されるが，Stp の発現は STP 座とは独立の I-STP 座上の不完全優性アレル I-Stp によって抑制されるこ とが強く示唆された．したがって，銀坊主，中国 117 号および RY50 の遺伝子型は，それぞれ StpStp I-StpI-Stp，stpstp I-StpI-Stp および stpstp i-stpi-stp であると推定された．この STP は，イネにおける穂形成に関する遺伝的制御の理解に寄与できると考えら れる． キーワード：1 次枝梗，イネ，草型，分離分析，抑制遺伝子 作物研究 62 号（2017） 連絡責任者：加藤恒雄（[email protected]）
Acknowledgements
The author sincerely thank to Dr. A. Horibata, T. Hanaoka and S. Takeda, the Faculty of Biology-Oriented Science and Technology, Kindai University, for their technical assistance.
References
Horibata, A. and T. Kato (2015) Distorted genetic segregation of the transposon mPing at the long arm of chromosome 12 in rice. Breed. Sci. 65: 340-344.
Jiang, N., Z. Bao, X. Zhang, H. Hirochika, S.R. Eddy, S.R. McCouch and S.R. Wessler (2003) An active DNA transposon family in rice. Nature 421: 163-167.
Kato, T. (2004) Quantitative trait loci controlling the number of spikelets and component traits in rice: Their main effects and interaction with years. Breed. Sci. 54: 125-132.
Kato, T. (2010) Main effects of individual QTLs and interactions between two QTLs controlling grain length of rice under near-isogenic lines. J. Crop Res. 55: 33-38.
Kikuchi, K., K. Terauchi, M. Wada and H.Y. Hirano (2003) The plant MITE mPing is mobilized in anther culture. Nature 421: 167-170.
Nakazaki, T., Y. Okumoto, A. Horibata, S. Yamahira, M. Teraishi, H. Nishida, H. Inoue and T. Tanisaka (2003) Mobilization of a transposon in the rice genome. Nature 421:
170-172.
Pereg, Z., T. Fahima, A.B. Korol, S. Abbo and Y. Saranga (2011) Genetic analysis of wheat domestication and evolution under domestication. J. Exp. Bot. 62: 5051-5061.
Simons, K.J., J.P. Fellers, H.N. Trik, Z. Zhang, Y.-S. Tai, B.S. Gill and J.D. Faris (2006) Molecular characterization of the major wheat domestication gene Q. Genetics 172: 547-555. Terao, T., K. Nagata, K. Morino and T. Hirose (2010) A gene
controlling the number of primary rachis branches also controls the vascular bundle formation and hence is responsible to increase the harvest index and grain yield in rice. Theor. Appl. Genet. 120: 875-893.
Yang, G., F. Zhang, C.N. Hancock and S.R. Wessler (2007) Transposition of the rice miniature inverted repeat transposable element mPing in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 104:10962-10967.
Yasuda, K., M. Ito, T. Sugita, T. Tsukiyama, H. Saito, K. Naito, M. Teraishi, T. Tanisaka and Y. Okumoto (2013) Utilization of transposable element mPing as a novel genetic tool for modification of the stress response in rice. Mol. Breed. 32: 505-516.
Zhang, K., Q. Qian, Z. Huang, Y. Wang, M. Li, L. Hong, D. Zeng, M. Gu, C. Chu and Z. Cheng (2006) GOLD HULL AND INTERNODE2 encodes a primarily multifunctional cinnamyl-alcohol dehydrogenase in rice. Plant Physiol. 140:972-983.
