- Email: [email protected]
Quantitative trait loci analysis of lateral shoot growth in tomato
Scientia Horticulturae 192 (2015) 117–124
Contents lists available at ScienceDirect
Scientia Horticulturae journal homepage: www.elsevier.com/locate/scihorti
Quantitative trait loci analysis of lateral shoot growth in tomato O New Lee a,∗ , Yusuke Uchida a , Keisuke Nemoto a , Yoko Mine b , Nobuo Sugiyama b a b
The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, 113-8657 Tokyo, Japan Tokyo University of Agriculture, Funako, Atsugi, 243-0034 Kanagawa, Japan
a r t i c l e
i n f o
Article history: Received 4 March 2015 Received in revised form 22 May 2015 Accepted 23 May 2015 Keywords: Tomato Lateral shoot development Branching Backcross inbred lines QTL analysis Epistasis
a b s t r a c t Branching is an important agronomic trait to determine plant architecture and fruit yield. In tomato, the regulatory mechanisms of shoot branching patterns have been studied mainly in mutants defective in meristem identity. Little is known about quantitative trait loci (QTLs) and their genetic interaction (epistasis) in lateral shoot development. Here, the genetic control of lateral bud outgrowth initiation and subsequent lateral bud development into branches were investigated. We used a BC1 F7 population developed from Solanum lycopersicum (SL) and its close wild relative Solanum pimpinellifolium (SP) to measure the number of leaves on primary shoot (LN), the proportion of nodes with lateral buds (LS0) and lateral branches longer than 5 cm (LS5) and 10 cm (LS10) on main axis, the proportion of nodes with secondary lateral branches longer than 1 cm (LSLS1) on primary lateral branches, and the lengths of the longest primary and secondary lateral branches (LLSL1 and LLSL2, respectively). Composite interval mapping detected 17 additive QTLs; those for LS5, LS10, and LLSL1 clustered near C2 At5g49480 or LEATPACb on chromosome 1 and LEOH361 on chromosome 4. One epistatic (QTL × QTL) interaction was identified for LS10 QTLs on chromosomes 1 and 12, where recombinant-type alleles increased lateral bud development into primary lateral branches. We analyzed further the QTLs clustered on chromosome 1 using two BC3 F4 populations containing introgression region from the SP, which revealed the co-location of QTLs for lateral shoot development and days to flowering time. These results suggest that lateral shoot development is regulated by some additive and epistatic QTLs, some of which possibly exhibit a pleiotropic effect on flowering time. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The patterns of axillary/lateral bud formation and their development determine plant architecture (Schmitz and Theres, 1999). Lateral buds have two developmental modes, monopodial and sympodial (Pnueli et al., 1998). The model plant Arabidopsis has monopodial shoot architecture, and its shoot development can be separated into three phases, vegetative, inflorescence, and flowering (Schmitz and Theres, 1999). The vegetative shoot consists of a short internode and rosette leaves. The inflorescence shoot is extended by internode expansion after the transition from the vegetative to reproductive stage. At the same time, the axillary buds become visible. Later, these axillary buds develop into lateral inflorescence shoots. The flowering shoots consist of an intermediate-length internode and solitary flowers with or without subtending bracts (Pnueli et al., 1998). By contrast, tomato has sympodial shoot architecture, and its development can be separated into two phases, vegetative and reproductive, which alternate
∗ Corresponding author. Tel.: +81 3 5841 5046; fax: +81 3 5841 1131. E-mail address: [email protected] (O. New Lee). http://dx.doi.org/10.1016/j.scienta.2015.05.026 0304-4238/© 2015 Elsevier B.V. All rights reserved.
regularly (Pnueli et al., 1998). The vegetative phase is terminated by the development of a cymose inflorescence after the production of 8–12 leaves. When reproductive growth is initiated by the outgrowth of the inflorescence, vegetative growth restarts with the development of lateral buds just below the inflorescence. This generates three more leaves until terminating with the next inflorescence initiation. In tomato, several mutants defective in meristem identity has been used to study the genetic control of shoot development and branching habit. The bushy (bu) mutant stimulates branching, resulting in a characteristic bushy appearance (Young and MacArthur, 1947). Campbell and Nonnecke (1974) reported another unusual branching mutant, lateral promoter (Lp), in which buds in the axils of the cotyledons appeared early, causing a bushier appearance. The self-pruning (sp) gene in tomato replaces flowers with leaves in the inflorescence and suppresses the transition of the vegetative shoot apex to a reproductive shoot (Pnueli et al., 1998). The lateral suppressor (ls) mutant does not form most of the axillary meristem (Schumacher et al., 1999) and is related to two proteins involved in negative regulation of GA signal transduction (GAI, RGA). In the blind (bl) mutant, the sympodial meristem is absent, and the number of flowers per inflorescence is reduced
118
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
commercial cultivar Solanum lycopersicum (M570018, normal branched tomato cultivar) and its close wild relative Solanum pimpinellifolium (PI124039, highly branched wild tomato) and a backcross of the F1 to the ‘M570018’ cultivar. The resultant BC1 F1 population was advanced using the single-seed descent method to obtain the BC1 F6 population, which was used for genotyping. The subsequent BC1 F7 generation was used for phenotypic evaluation. To confirm the effects of detected QTLs in chromosome 1 on flowering time and branch development, two BC3 F1 lines containing a 45 cM heterozygous fragment on chromosome 1 were selected and advanced to BC3 F4 populations by selfing. In addition to a fragment on chromosome 1, these two lines contained 13 and 16.2 cM heterozygous fragments near C2 At3g55120 and SSR115 to C2 At3g55120 markers on chromosome 5. A total of 100 plants from two BC3 F4 populations (1762050-7-16 and 1762050-7-34) were used for the QTL analysis. Fig. 1. Illustration of inflorescence and lateral branch development in tomato. To evaluate the lateral shoot development in a BC1 F7 and two BC3 F4 populations, the number of leaves on primary shoot (LN), the number of leaves on the main axis (TLN) was measured. The proportion of buds to LN was calculated as an indicator of the proportion of lateral buds released from apical dominance (LS0). The proportions of the number of these primary lateral branches to TLN were designated LS5 and LS10, respectively. The proportion of the number of secondary lateral branches longer than 1 cm to the leaf number of the longest primary branches (LSLS1), the length of the longest primary lateral branches (LLSL1) and the longest secondary lateral branches (LLSL2) were measured. The total number of inflorescences (IN) and the average number of flowers (AFN) on the longest primary lateral branch were measured 62 days after sowing.
(Schmitz et al., 2002). The jointless (j) mutant suppresses sympodial meristem identity, which reverts the inflorescence meristem to vegetative growth after forming 1–2 flowers (Mao et al., 2000; Szymkowiak and Irishi, 2006). Although genes regulating branching habit and shoot development in these mutants have been identified, lateral shoot development is also controlled by the cumulative effects of quantitative trait loci (QTLs), like most important agronomic traits such as yield, plant height, and flowering time. QTL analysis using phenotypic values and genetic marker information could enable us to determine the number of QTLs associated with lateral shoot development, their locations and epistatic effects between QTLs. In cereal crops, many QTLs that affect branching have been identified. Teosinte branched (tb1) is a major QTL that controls the reduction in axillary branching from teosinte to maize (Doebley et al., 1995). In rice, the identification of genes that underlie QTLs affecting branching type has been advanced using rice genome information (Doust, 2007). Peng et al. (2014) also reported that QTLs for primary panicle branch number was regulated by a gene for plant height and tiller number. In spray cut chrysanthemum, 16 additive QTLs for branching traits were identified (Peng et al., 2015). By contrast, QTLs controlling lateral shoot development has been little studied in tomato although the number of lateral shoots affects vegetative and reproductive traits, such as leaf mass, canopy structure, the number of flowers and yield (Navarrete and Jeannequin, 2000). In the present study, therefore, we tried to identify QTLs related to the initiation of lateral bud outgrowth and subsequent lateral bud development in tomato using 111 BC1 F7 lines and 100 BC3 F4 lines (Fig. 1). Our results are useful for understanding the genetic basis of lateral shoot development and branching type in tomato. 2. Materials and methods 2.1. Mapping population One hundred and eleven BC1 F7 lines (hereafter referred to as BILs) were derived from an initial cross between the
2.2. Phenotyping For phenotypic evaluation, the 111 BC1 F7 lines, along with their parents, were grown in a greenhouse with natural daylight in The University of Tokyo, Japan. Seven seeds from each of the BILs were sown into 200 mL plastic pots (7 cm diameter) containing mixed commercial growth medium (Engeibaido, Kureha, Tokyo, Japan; Soilmix, Sakata Seed Co., Yokohama, Japan) on 15 July 2010. On 12 August (four weeks later), five out of seven seedlings were selected and transplanted into 1 L plastic pots (14.5 cm diameter). The pots were arranged in two double rows with 130 cm spacing between the centers of the rows. The plants were spaced 30 cm apart within each row and 45 cm apart between rows. For QTL analysis using BC3 F4 families, two seeds from 50 lines, each of the two BC3 F4 families were sown into 500 mL plastic pots (9 cm diameter), with five pots per line, on 7 October 2010, and thinned to one plant per pot after emergence. On 29 October, the plants were transplanted into 1 L plastic pots (14.5 cm diameter). The plants were spaced 27 cm apart within rows and 45 cm apart between rows. The plants were watered daily throughout the experiment. A randomized complete block design was used in both experiments. No lateral branches were pruned in either experiments. To evaluate the number of axillary buds (lateral buds) in the 111 BC1 F7 lines, the number of leaves before the inflorescence (LN), i.e., the number of leaves on the primary shoot was measured 46 days after sowing the seeds, at which time the macroscopic appearance of the first inflorescence had occurred. At the same time, the number of visible lateral buds was counted, and the proportion of buds to LN was calculated as an indicator of the proportion of lateral buds released from apical dominance (LS0). The numbers of primary lateral branches longer than 5 cm and 10 cm and the number of leaves on main axis (monopodial and sympodial shoots), TLN, were measured 62 days after sowing. The proportions of the number of these primary lateral branches to TLN were designated LS5 and LS10, respectively. The proportion of the number of secondary lateral branches longer than 1 cm to the leaf number of the longest primary branches (LSLS1), the length of the longest primary lateral branches (LLSL1), and the longest secondary lateral branches (LLSL2) were measured 64 days after sowing. The total number of inflorescences (IN) and the average number of flowers (AFN) on the longest primary lateral branch were measured 62 days after sowing. In QTL analysis using the two BC3 F4 families, the traits for which QTLs were detected in the experiment using BC1 F7 BILs were measured; LS5 and LS10 were measured 72 days after sowing, while LSLS1, LSLS2, LLSL1, and LLSL2 were measured 74 days after sowing. The number of leaves (LN) before inflorescence was measured 46 days after sowing, and days to flowering (DTF) were measured as well. However, LS0 was not measured in this experiment because
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
119
Fig. 2. Frequency distribution for each trait in the BC1 F7 population. Means for the parental and BC1 F7 lines are shown by arrowheads and an arrow, respectively. () Solanum lycopersicum cv. ‘M570018’ (n = 5); () S. pimpinellifolium (PI124039) (n = 5); (↓) BC1 F7 (n = 111). No data for the parental lines SL and SP for AFN.
lateral buds were visible in most axils of the primary shoot even in SL plants. 2.3. Genotyping One hundred and seven polymorphic markers (44 SSR, 23COSII, 37CAPS, and three SNPs) were used to construct a linkage map. Marker analysis was carried out as described by Cagas et al. (2008) using DNA from the 111 BC1 F6 plants and parental lines. In brief, DNA was extracted from 0.1 g of bulked leaf material of 7 individuals in each BIL using the Nucleon PhytoPure Plant DNA extraction kit (Amersham Biosciences, Buckinghamshire, UK), according to the manufacturer’s protocol. PCR mixtures consisted of 200 ng of template DNA, 2 M forward and reverse primers, 0.2 mM dNTP mixture, 1.8 mM MgCl2 , 1 L 10× NH4 buffer and 0.25 units of Biotaq DNA polymerase (Bioline, London, UK) in 10 L of total reaction solution. PCR conditions were one cycle at 94 ◦ C for 5 min, followed by 35 cycles at 94 ◦ C for 30 s, 48–55 ◦ C for 45 s and 72 ◦ C for 45 s, and one cycle at 72 ◦ C for final extension for 5 min. For COSII and CAPS markers, the amplified products were digested using restriction enzymes at 37 ◦ C for 12 h. PCR products were separated by polyacrylamide gel electrophoresis for 3 h at 120 V. Gels were visualized under UV light after staining with SYBR Green I (Lonza, Rockland, ME, USA). 2.4. QTL analysis The MAPMAKER/EXP v. 3.0b software was used for linkage analysis in the ‘RI self’ mode (Lander et al., 1987). The Kosambi function was used to convert the recombination frequencies (r) to map distances in centiMorgans (Kosambi, 1944). QTL detection was performed with composite interval mapping (CIM) using Windows QTL Cartographer ver. 2.0 (Wang et al., 2003). The experimentwise LOD threshold level of significance was evaluated by
performing 1000 permutations of each character (P < 0.05) (Churchill and Doerge, 1994). Epistatic (QTL × QTL) effects of the QTLs were tested using the QTLNetwork 2.0 software, which utilizes a mixed model-based composite interval mapping approach (Yang et al., 2008). The QTLs detected were designated by the trait name followed by the chromosome number. If more than one QTL from the same trait was detected on a chromosome, serial letters were added after the chromosome number. 2.5. Statistical analysis In both experiments, the Kolmogorov–Smirnov test was conducted to compare the observed cumulative frequency distributions of each trait using Microsoft Excel. In the experiment using BILs, single-factor ANOVA and Pearson correlation analysis were performed using the SPSS for Windows v. 11.01 software package. In the experiment using two BC3 F4 families, associations between genotype class at the closest flanking marker and phenotypic values were tested with the non-parametric Mann–Whitney U test. 3. Results 3.1. Trait variation and correlations among traits The Kolmogorov–Smirnov test indicated that none of the traits followed a normal distribution (Fig. 2). The LLSL2 trait distribution was left-skewed because the length of the secondary lateral shoots for most lines was shorter than 5 cm. For all of the traits except AFN, the average values were greater in S. pimpinellifolium (SP) than in S. lycopersicum (SL). On the other hand, the average values of the BILs lay in between SP and SL and the differences were significant among them, except for LLSL2 (Table 1; Fig. 2); the difference in LLSL2 was not significant between SL and the BILs. Statistical analysis of AFN could not be carried out because of missing data for the
120
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
Table 1 Averages (±SD) and statistical analysis of traits related with lateral shoot development and floral development in Solanum lycopersicum (SL), S. pimpinellifolium (SP), and 111 BC1 F7 lines (BILs). Comparisonsb Traitsa
SL
SP
RIL
SL:SP
SL:RILs
SP:RILs
LN LS0 LS1 LS5 LS10 LSLS1 LLSL1 LLSL2 IN AFNc
6.8 ± 0.89 0.62 ± 0.17 0.14 ± 0.14 0.03 ± 0.04 0.02 ± 0.03 0.03 ± 0.09 6.39 ± 7.83 0.09 ± 0.15 1.82 ± 0.40 –
11.75 ± 0.77 0.93 ± 0.07 0.77 ± 0.09 0.58 ± 0.08 0.50 ± 0.07 0.55 ± 0.08 69.47 ± 8.57 33.21 ± 11.45 5.8 ± 0.77 –
8.46 ± 0.65 0.76 ± 0.15 0.39 ± 0.18 0.15 ± 0.065 0.10 ± 0.05 0.19 ± 0.12 29.43 ± 14.40 3.73 ± 4.50 2.93 ± 0.66 4.96 ± 0.81
** ** ** ** ** ** ** ** ** –
** ** ** ** ** ** * NS * –
** ** ** ** ** ** ** ** ** –
a LN, leaf number on the primary shoot; LS0, proportion of nodes of lateral buds released from paradormancy; LS5 and LS10, ratios of the number of primary lateral branches longer than 5 cm and 10 cm to LN; LSLS1, ratio of secondary lateral branches longer than 1 cm to the number of nodes at the longest primary lateral branches; LLSL1 and LLSL2, lengths of the longest primary and secondary lateral branches (cm); IN, total number of inflorescences; AFN, average number of flowers on the longest lateral branches. b NS = not significant at P < 0.05; * and ** = significant at P < 0.05 and 0.01, respectively. c No statistical analysis was carried out for the trait because of a lack of data for the parental lines SL and SP.
Trait
LN
LS0
0.176
LS5
0.074
LS10
0.117
LS0
LS5
LS10
LSLS1
LLSL1
LLSL2
IN
parent plants. All of the traits except LN and LS0 were positively and significantly correlated with each other (Fig. 3). LN was only correlated with IN, and LS0 correlated with LSLS1.
presence of the SP allele stimulated elongation of the secondary lateral branches. For IN, only one additive QTL was detected in the vicinity of LEATPACb on chromosome 1, explaining 20% of the phenotypic variation (Table 2). For AFN, three additive QTLs were detected on chromosomes 2, 5, and 10, and together explained 37% of the phenotypic variation in this trait (9%, 19%, and 9%, respectively). The presence of SP alleles in the detected QTLs for IN and AFN increased the total number of inflorescences and the average number of flowers per inflorescence. One epistatic (QTL × QTL) interaction was identified for LS10 on chromosomes 1 and 12 (Table 3). The epistasis between ls101 and ls10-12 explained 12% of the phenotypic variation and the recombinant-type of the parental alleles (Qi Qi qj qj or qi qi Qj Qj ) stimulated elongation of primary lateral branches.
3.2. QTLs detected in 111 BC1 F7 BILs
3.3. QTLs detected in two BC3 F4 families
Seventeen QTLs with additive effects and one epistatic QTL pair were detected for the eight traits related with lateral shoot development (Fig. 4; Table 2). Among 17 QTLs with additive effects, seven were co-located between LEATPACb and C2 At5g49480 on chromosome 1 and three were near the marker LEOH361 on chromosome 4, while no co-location was found for the other seven additive QTLs mapped to chromosomes 1, 2, 5, 6, and 10. Only one epistatic interaction pair was detected, for LS10 on chromosomes 1 and 12. For LN, three additive QTLs were detected on chromosomes 1, 6, and 10, explaining 22%, 9%, and 7% of the phenotypic variation, respectively (Table 2). The presence of SP alleles for ln-1 on chromosome 1 increased LN by 0.54, while the presence of SP alleles on chromosome 6 and 10 decreased LN by 0.37 and 0.33, respectively. The ls5-1, ls10-1, and llsl2-1 QTLs were co-located near the marker C2 At5g49480, while the lsls1-1 and llsl1-1 QTLs were colocated near the marker LEATPACb on chromosome 1. These QTLs, which formed a cluster between LEATPACb and C2 At5g49480 on chromosome 1, explained 13%, 16%, 19%, 31%, and 13% each of the phenotypic variation, respectively (Fig. 4; Table 2). The ls5-4, ls104, and llsl1-4 QTLs were co-located in the vicinity of LEOH361 on chromosome 4, explaining 10%, 25%, and 19% of the phenotypic variation, respectively. The presence of SP alleles in the detected QTLs on chromosomes 1 and 4 increased the development of lateral shoots in terms of LS5, LS10, LSLS1, LLSL1, and LLSL2. Another QTL, llsl2-1, detected in the vicinity of SSR478 on chromosome 1 explained 41% of the phenotypic variation of the trait, and the
The two BC3 F4 families (1762050-7-16 and 1762050-7-34) were grown and evaluated for flowering time and lateral shoot development. The two BC3 F4 families, chromosome segment substitution lines (CSSLs) containing two introgression regions from the SP allele (45 cM in chromosome 1 and about 15 cM in chromosome 5). No QTLs related flowering time and branch development were detected on chromosome 5, except for AFN QTL. At the closest flanking marker on chromosome 1, the Mann–Whitney test revealed that DTF was significantly lower in plants homozygous for SP alleles than in plants homozygous for SL alleles, while other traits related to lateral shoot development and LN were significantly higher in plants homozygous for SP alleles (Table 4). All of the traits were positively and significantly correlated with each other (data not shown). A single QTL was detected for all of the traits except LS10 (Table 5). These QTLs formed two clusters; QTLs for LN, LLSL1 and LLSL2 were co-located in the vicinity of C2 At5g49480 while QTLs for LS5 and LSLS1 were co-located in the vicinity of SSR105 (Table 5; Fig. 5a). By contrast, dtf1 was located 4–5 cm upstream of llsl1 and llsl2. Furthermore, LOD score profile of DTF QTL was different from those of QTLs for LS5 and LSLS1, and LOD score profile of QTLs for LLSL1 and LLSL2 was intermediate (Fig. 5b). However, the one-LOD support confidence intervals of these QTLs overlapped. The SL allele for dtf-1 increased DTF, while the SP alleles for ln-1, ls5-1, lsls-1, llsl1-1 and llsl2-1 increased LN, LS5, LSLS1, LLSL1 and LLSL2, respectively (Table 5). dtf-1, ls5-1, and lsls1-1 with high LOD scores (4.62 to 8.08) explained phenotypic variations ranging from 21 to 39%, while ln-1, llsl1, llsl2 with low LOD scores (2.19–3.64)
0.053 0.120
0.860**
LSLS1 0.151
0.198* 0.320** 0.292**
LLSL1 0.065
0.099
0.365** 0.438** 0.571**
LLSL2 0.179
0.004
0.220*
0.314** 0.473** 0.688**
IN
0.257** 0.140
0.393** 0.362** 0.336** 0.344** 0.308**
AFN
0.133
0.310** 0.281** 0.237** 0.239*
0.025
0.252** 0.396**
Fig. 3. Correlation between the traits in the BC1 F7 population. * and ** indicate significance at P < 0.05 and P < 0.01, respectively.
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
121
Table 2 Additive QTLs related with lateral shoot development and floral development. Traitsa
Chromosome
Molecular marker
QTL positionb
LOD score
Additive effectc
R2
LN
1 6 10 1 4 1 4 1 4 1 1 1 2 1 2 5 10
C2 At5g49480 SSR578 SSR248 C2 At5g49480 LEOH361 C2 At5g49480 LEOH361 LEATPAC LEOH361 SSR478 C2 At5g49480 LEATPAC SSR32 LEATPAC TG337 C2 At3g55120 C2 At5g06430
41.49 23.11 37.96 47.55 6.26 46.55 10.26 43.66 11.56 10.01 52.02 38.55 59.25 41.04 68.83 32.15 0.01
8.31 3.04 2.65 4.03 3.32 5.90 6.55 4.17 3.79 4.21 5.05 6.87 4.43 5.43 3.02 5.05 3.08
−0.54 0.37 0.33 −0.028 −0.026 −0.027 −0.033 −65.25 −79.58 −53.77 −25.3 −0.10 −0.06 −0.36 −0.32 −0.52 −0.33
0.22 0.09 0.07 0.13 0.10 0.16 0.25 0.13 0.19 0.41 0.19 0.31 0.13 0.20 0.09 0.19 0.09
LS5 LS10 LLSL1 LLSL2 LSLS1 IN AFN
a
See the caption of Fig. 1 for trait abbreviations. Position on chromosome in cM. c Positive values for additive effects indicate that S. lycopersicum alleles increased the phenotypic value more than S. pimpinellifolium alleles, whereas negative values indicate that S. pimpinellifolium alleles increased the phenotypic value. b
Table 3 Epistatic QTLs related with lateral shoot development. Trait
QTLi
Chra
Interval (SIb )
Position
QTLj
Chra
Interval
(SIb )
LS10
ls101
1
SSR117 −SSR65 (99.9−115.2)
105.9
ls1012
12
SSR345−C2 At1g48300 (35.8−57.9)
Position
AAc
H2 AA
d
44.3
−0.0239 0.1201
P-valuee 0.000001
a
Chromosome number. Support interval. AA, a positive value indicates that the presence of parental type alleles (Qi Qi Qj Qj or qi qi qj qj ) in the epistatic loci increased the phenotypic value more than the recombinant type (Qi Qi qj qj or qi qi Qj Qj ).Likewise, a negative value indicates that the recombinant type alleles increased the phenotypic value more than the parental type did. d H2 , the heritability of the additive effect. e P-value, the probability of the estimated or predicted QTL effect. b c
Table 4 Averages (±SD) and statistical analysis of traits related with lateral shoot development and days to flowering (DTF) in BC3 F4 lines. Phenotypic values were compared between plants homozygous for S. lycopersicum (SL) alleles, homozygous for S. pimpinellifolium (SP) alleles and heterozygous for S. lycopersicum and S. pimpinellifolium (SLSP) alleles in the BC3 F4 families, respectively. Comparisonsb
Genotype class at the closest flanking markers Traitsa
SLSL
SPSP
SLSP
SL:SP
DTF LN LS5 LS10 LSLS1 LLSL1 LLSL2
57.6 ± 2.47 5.64 ± 0.48 0.11 ± 0.07 0.52 ± 0.03 0.72 ± 0.14 18.6 ± 16.2 1.66 ± 2.48
54.1 ± 1.70 6.44 ± 1.70 0.20 ± 0.08 0.06 ± 0.02 0.21 ± 0.10 45.0 ± 23.2 14.24 ± 22.87
57.6 ± 2.47 6.20 ± 0.63 0.14 ± 0.07 0.06 ± 0.01 0.18 ± 0.10 45.5 ± 31.5 4.50 ± 6.74
* * ** ** ** ** **
a
See the caption of Fig. 1 for trait abbreviations. Significance of marker association was determined for genotypic classes using the Mann–Whitney test between Solanum lycopersicum (SL) and S. pimpinellifolium (SP). NS = not-significant at P < 0.05; * and ** = significant at P < 0.05 and 0.01, respectively. b
Table 5 Additive QTLs related with lateral shoot development in the experiment on BC3 F4 lines. Traitsa
Chromosome
Molecular marker
QTL positionb
LOD score
Additive effectc
R2
DTF LN LS5 LSLS1 LLSL1 LLSL2
1 1 1 1 1 1
C2 At5g49480 SSR105 SSR105 SSR105 C2 At5g49480 C2 At5g49480
12.01 17.2 19.2 21.2 16.01 17.01
8.08 2.19 4.62 7.18 3.64 2.78
1.98 −0.26 −0.05 −0.10 −13.97 −6.07
0.39 0.09 0.21 0.29 0.17 0.13
a
See the caption of Fig. 1 for trait abbreviations. Position on chromosome in cM Positive values for additive effects indicate that S. lycopersicum alleles increased the phenotypic value more than S. pimpinellifolium alleles, whereas negative values indicate that S. pimpinellifolium alleles increased the phenotypic value. b c
122
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
Fig. 4. QTL map for traits related with lateral shoot development, based on a BC1 F7 population derived from a cross between Solanum lycopersicum (SL) and its close wild relative Solanum pimpinellifolium (SP). Names of the markers and QTLs are shown on the right and left of chromosomes, respectively. Arrowheads indicate the F-statistic peak of each QTL. Black and white symbols indicate that the QTL exhibited additive effects and epistatic effects, respectively. The associated bars denote the support confidence interval of each QTL.
explained smaller phenotypic variations ranging from 9 to 17%. 4. Discussion Tomato plants exhibit a sympodial growth pattern; when the shoot apical meristem of the primary shoot is terminated by the initiation of the first inflorescence, the lateral bud just below the first inflorescence develops into a sympodial shoot (Vegetti and Pilatti, 1998). However, some lateral buds formed at middle and lower nodes can also develop into sympodial shoots, showing a large genetic variation in the number of lateral branches. Indeterminate tomatoes are usually trained to maintain a single stem; lateral branches just below inflorescences are retained but other lateral branches are removed to enable easy handling and harvesting and improve air flow through the canopy (Navarrete and Jeannequin, 2000). Therefore, cultivars that produce a small number of lateral branches have attracted the interest of tomato breeders because manual pruning is labor intensive (Reinhardt
and Kuhlemeier, 2002). On the other hand, determinate tomato cultivars do not require pruning (Kemble and Gardner, 1992) or need only minimal pruning (Olson et al., 2006) because of their compact plant type. However, a large number of lateral branches are a prerequisite for high yield, and cultivars with the dominant lateral promoter gene (Lp) have been used for breeding high yielding determinate tomatoes (Ito et al., 1990). Therefore, the genetic basis for the variation of lateral shoot development needs to be studied in determinate as well as indeterminate tomatoes. Many QTL analyses in tomato have focused on flowering time and fruit related traits, e.g., fruit weight and soluble solids (Weller et al., 1988; De Vicente and Tanksley, 1993; Grandillo and Tanksley, 1996; Jimenez-Gomez et al., 2007; Cagas et al., 2008; Sumugat et al., 2010; Barrios-Masias and Jackson, 2014). However, there have been few reports on QTLs related to lateral shoot development in tomatoes. This study showed that additive QTLs related to lateral bud outgrowth and subsequent development into branches were mainly clustered on two chromosomes, chromosomes 1 and
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
123
Fig. 5. (A) Genetic location of QTLs for DTF, LN, LS5, LSLS1, LSLS1, and LLSL2 in the BC3 F4 populations (B) LOD score profiles for QTLs within the introgression region from the SP allele on chromosome 1.
4 (Fig. 4). On chromosome 1, QTLs for seven traits (ln, ls5, ls10, lsls1, llsl1, llsl2, in) were detected concurrently between the markers LEATPACb and C2 A5g49480, explaining 13–41% of the phenotypic variation. In this region, several QTLs related to flowering time and root growth have been identified in previous studies using BILs derived from the same parental accessions (Cagas et al., 2008; Sumugat et al., 2010, 2011). Sumugat et al. (2010) assumed that a single QTL acts as a ‘master gene’ to control multiple developmental processes, e.g., DTF, days to macroscopic flower bud appearance and number of leaves initiated, through pleiotropic effects. The co-localization of LS5, LSLS1, LLSL1, and LLSL2 QTLs suggests that the developmental processes of primary and secondary lateral shoots are regulated by the same ‘master gene’ that Sumugat et al. (2010) suggested for several flowering time traits. However, we could not clarify whether LS5, LSLS1, LLSL1, and LLSL2 QTLs were regulated by a pleiotropic QTL or closely linked QTLs in the experiment using the BC1 F7 population. Thus, QTL analysis for these traits as well as DTF was performed using two BC3 F4 families in this study (Table 5; Fig. 5). The results showed that QTLs related with lateral shoot development were within the cluster region near C2 At5g49480 and SSR105, but it appears that this cluster can be divided into two sub-clusters and single QTL; llsl11 and llsl2-1 in the first sub-cluster, and ls5-1 and lsls1-1 in the second. On the other hand, DTF QTL was identified at a slightly different position of chromosome 1. Sub-clustering of QTLs related with lateral shoot development and separate detection of DTF QTL support the idea that the initiation of lateral bud outgrowth, extension of lateral shoots and flowering time are controlled by closely linked QTLs. However, the possibility that QTLs for LS5 and LSLS1 is located at the same position as QTLs for DTF, LLSL1 and LLSL2 cannot be ruled out because their support confidence intervals overlapped. To confirm these QTL positions more precisely, it is necessary to increase the number of polymorphic markers and narrow the gap or to develop single-chromosome segment substitution lines. In connection with the existence of a pleiotropic QTL in this region, it is noteworthy that Cagas et al. (2008) reported that a QTL for the number of lateral buds longer that 0.5 cm was co-located with the DTF QTL in the vicinity of C2 At5g49480 and that this QTL had a negative additive effect similar to LS5 and LSLS1.
Werner et al. (2003) reported that cytokinins have important, but opposite, functions in the regulation of the activities of lateral shoot meristems and root meristems; overexpression of the cytokinin oxidase (CKX1) gene retarded leaf formation but increased root elongation in Arabidopsis. Sumugat et al. (2011) raised 110 BC1 F6 lines derived from the same parents as this study using 128-cell trays and evaluated the root ball weight of seedlings collected 19 days after sowing. They identified a QTL for root ball development between SSR105 and SSR134 on chromosome 1, in a different position to the QTLs for lateral shoot development (LEATPACb, C2 A5g49480 to SSR 105) identified in this study. Therefore, it is not clear whether lateral shoot development is negatively correlated with root growth. Paran et al. (1997) reported that the QTL for leaf length at node number 3 was co-located with the QTL for the number of side branches longer than 2 cm on chromosome 3, using an inbred population from a cross between S. lycopersicum and Solanum cheesmaniae. However, we did not detect any QTLs related to lateral shoot development on chromosome 3 in this study. On the other hand, we detected one independent QTL cluster related to lateral shoot development on chromosome 4, in which additive QTLs for three traits (LS5, LS10, and LLSL1) were co-located near the marker LEOH 361 (Table 2; Fig. 4). The presence of an SP allele in this region could enhance the initiation of lateral branch outgrowth (LS5; LS10) and the elongation of primary lateral branches (LLSL1). In this study, experiment using a BC1 F7 population was carried out in summer and experiment using two BC3 F4 populations was carried out in autumn. As a result, QTLs for LN, LS5, LSLS1, LLSL1, and LLSL2 on chromosome 1 were detected across seasons, indicating these QTLs were stable QTL. However, LS10 on chromosome 1 was only detected in experiment using a BC1 F7 population. This suggested that some QTLs have QTL × environment interaction effects. Thus, it seems necessary to carry out QTL analysis under different seasons for better understanding the genetic control of lateral shoot development. In this study, only one epistatic interaction was detected; one QTL pair on chromosomes 10 and 12 for LS10 (Table 3; Fig. 4). The epistatic interaction of the ls10-10 and ls10-12 QTLs accounted for 12% of the phenotypic variation and the recombinant-type parental alleles increased LS10 by 0.024. Two additive QTLs for LS10 detected
124
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
on chromosome 1 and 4 accounted for 16% and 25%, respectively, and their additive values were around −0.03. These results suggest that any attempt to use these QTLs in breeding programs needs to take into account such epistatic effects but also additive effects. 5. Conclusion In conclusion, this is the first study on QTLs that control the vigorous development of lateral branches in tomatoes. We suggest that shoot lateral branching is predominantly controlled by closely linked QTLs near the C2 At5g49480 marker on chromosome 1 and those near LEOH361 on chromosome 4, along with one pair of epistatic QTLs. Furthermore, our results suggest that a small target segment derived from S. pimpinellifolium could shorten DTF and stimulate lateral branching because the QTLs were co-located near the C2 At5g49480 marker on chromosome 1. References Barrios-Masias, F.H., Jackson, L.E., 2014. California processing tomatoes: morphological, physiological and phonological traits associated with crop improvement during the last 80 years. Eur. J. Agron. 53, 45–55. Cagas, C.C., Lee, O., Nemoto, K., Sugiyma, N., 2008. Quantitative trait loci controlling flowering time and related traits in a Solanum lycopersicum × S. pimpinellifolium cross. Sci. Hortic. 116, 144–151. Campbell, C.G., Nonnecke, I.L., 1974. Inheritance of and enhanced branching character in the tomato (Lycopersicon esculentum Mill.). J. Am. Soc. Hortic. Sci. 99, 358–360. Churchill, G.A., Doerge, R.W., 1994. Empirical threshold values for quantitative trait mapping. Genetics 138, 963–971. De Vicente, M.C., Tanksley, S.D., 1993. QTL analysis of transgressive segregation in an interspecific tomato cross. Genetics 134, 585–596. Doebley, J., Stec, A., Gustu, C., 1995. Teosinte branched 1 and the origin of maize: evidence for epistasis and the evolution of dominance. Genetics 141, 333–346. Doust, A.N., 2007. Grass architecture: genetic and environmental control of branching. Curr. Opin. Plant Biol. 10, 21–25. Grandillo, S., Tanksley, S.D., 1996. QTL analysis of horticultural traits differentiating cultivated tomato from the closely related species Lycopersicon pimpinellifolium. Theor. Appl. Genet. 92, 935–951. Ito, K., Kamimura, S., Mochizuki, T., Ishiuchi, D., Kanno, A., 1990. New line tomato for processing, ‘Morioka No. 22’. Bull. Natl. Res. Inst. Veg. Ornament. Plants Tea (C) 1, 11–20, Japanese with English Summary. Jimenez-Gomez, J.M., Alonso-Blanco, C., Borja, A., Anastasio, G., Angosto, T., Lozano, R., Martinez-Zapater, J.M., 2007. Quantitative genetic analysis of flowering time in tomato. Genome 50, 303–315. Kemble, J.M., Gardner, R.G., 1992. Inheritance of shortened fruit maturation in the cherry tomato Cornell 871213-1 and its relation to fruit size and other component of earliness. J. Am. Soc. Hortic. Sci. 117, 464–650. Kosambi, D.D., 1944. The estimation of map distances from recombination values. Ann. Eugen. 12, 172–175. Lander, E.S., Green, P., Abrahamson, J., Barlow, A., Daly, M.J., Lincoln, S.E., Newburg, L., 1987. MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1, 174–181.
Mao, L., Begum, D., Chuang, H.W., Budiman, M.A., Szymkowiak, E.J., Irish, E.E., Wing, R.A., 2000. JOINTLESS is a MADS-box gene controlling tomato flower abscission zone development. Nature 406, 910–913. Navarrete, M., Jeannequin, B., 2000. Effect of frequency of axillary bud pruning on vegetative growth and fruit yield in greenhouse tomato crops. Sci. Hortic 86, 197–210. Olson, S.M., Stall, W.M., Momol, M.T., Webb, S.E., Taylor, T.G., Smith, S.A., Simonne, E.H., McAvoy, E., 2006. In: Olson, S.M., Simonne, E.H. (Eds.). Vegetable production handbook for Florida, 2006–2007: Tomato production in Florida. University of Florida, Institute of Food and Agricultural Science and Vance Publishing, Lincolnshire. pp. 407–426 Cited by Santos, B.M (2008). Early pruning effects on ‘Florida-47 and ‘Sungard’ tomato HortTech. 18, 2006, 467–470. Paran, I., Goldman, I., Zamir, D., 1997. QTL analysis of morphological traits in a tomato recombinant inbred line population. Genome 40, 242–248. Peng, Y., Gao, Z., Zhang, B., Liu, C., Xu, J., Ruan, B., Hu, J., Dong, G., Guo, L., Linag, G., Qian, Q., 2014. Fine mapping and candidate gene analysis of a major QTL for panicle structure in rice. Plant Cell Rep. 33, 1843–1850. Peng, H., Zhang, F., Jiang, J., Chen, S., Fang, W., Guan, Z., Chen, F., 2015. Identification of quantitative trait loci for branching traits of spray cut chrysanthemum. Euphytica 202, 385–392. Pnueli, L., Carmel-Goren, L., Hareven, D., Gutfinger, T., Alvarez, J., Ganal, M., Zamir, D., Lifschitz, E., 1998. The SELF-PRUNING gene of tomato regulates vegetative to reproductive switching of sympodial meristems and is the ortholog of CEN and TFL1. Development 125, 1979–1989. Reinhardt, D., Kuhlemeier, D., 2002. Plant architecture. EMBO Rep. 3, 846–851. Schmitz, G., Theres, K., 1999. Genetic control of branching in Arabidopsis and tomato. Curr. Opin. Plant Biol. 2, 51–55. Schmitz, G., Tillman, E., Carriero, F., Fiore, C., Cellini, F., Theres, K., 2002. The tomato Blind gene encodes a MYB transcription factor that controls the formation of lateral meristems. Proc. Natl. Acad. Sci. U. S. A. 99, 1064–1069. Schumacher, K., Schmitt, T., Rossberg, M., Schmitz, G., There, K., 1999. The lateral suppressor (Ls) gene of tomato encodes a new member of the VHIID protein family. Proc. Natl. Acad. Sci. U. S. A. 96, 290–295. Sumugat, M.R., Lee, O., Nemoto, K., Sugiyama, N., 2010. Quantitative trait loci analysis of flowering-time-related traits in tomato. Sci. Hortic. 123, 343–349. Sumugat, M.R., Lee, O., Mine, Y., Sugiyama, N., 2011. Quantitative trait loci analysis of transplanting time and other root-growth-related traits in tomato. Sci. Hortic. 129, 622–628. Szymkowiak, E.J., Irishi, E.E., 2006. JOINTLESS suppresses sympodial identity in infloresecence meristems of tomato. Planta 223, 646–658. Vegetti, A.C., Pilatti, R.A., 1998. Structural patterns of tomato plants grown in glasshouse. Genet. Resour. Crop Evol. 45, 87–93. Wang, S., Basten, C.J., Zeng, Z.B., 2003. Windows QTL CARTOGRAPHER 2.0. Department of Statistics, North Carolina State University, Raleigh, NC http:// statgen.ncsu.edu/qtlcart/WQTLCart.htm Weller, J.I., Soller, M., Brody, T., 1988. Linkage analysis of quantitative traits in an interspecific cross of tomato (Lycopersicon esculentum × Lycopersicon pipinelifolium) by means of genetic markers. Genetics 118, 329–339. Werner, T., Motyka, V., Laucou, V., Smets, R., Van Onclelen, H., Schmülling, T., 2003. Cytokinin-deficient transgenic Arabidopsis plants show multiple developmental alterations indicating opposite functions of cytokinins in regulation of shoot and root meristem activity. Plant Cell 15, 2532–2550. Yang, J., Hu, C.C., Hu, H., Yu, R.D., Xia, Z., Ye, X.Z., Zhu, J., 2008. QTLNetwork: mapping and visualizing genetic architecture of complex traits in experimental populations. Bioinformatics 24, 721–723 http://ibi.zju.edu.cn/ software/qtlnetwork/ Young, P.A., MacArthur, J.W., 1947. Horticultural Characters of Tomatoes. Texas Agricultural Experiment Station Bulletin, pp. 698.
Contents lists available at ScienceDirect
Scientia Horticulturae journal homepage: www.elsevier.com/locate/scihorti
Quantitative trait loci analysis of lateral shoot growth in tomato O New Lee a,∗ , Yusuke Uchida a , Keisuke Nemoto a , Yoko Mine b , Nobuo Sugiyama b a b
The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, 113-8657 Tokyo, Japan Tokyo University of Agriculture, Funako, Atsugi, 243-0034 Kanagawa, Japan
a r t i c l e
i n f o
Article history: Received 4 March 2015 Received in revised form 22 May 2015 Accepted 23 May 2015 Keywords: Tomato Lateral shoot development Branching Backcross inbred lines QTL analysis Epistasis
a b s t r a c t Branching is an important agronomic trait to determine plant architecture and fruit yield. In tomato, the regulatory mechanisms of shoot branching patterns have been studied mainly in mutants defective in meristem identity. Little is known about quantitative trait loci (QTLs) and their genetic interaction (epistasis) in lateral shoot development. Here, the genetic control of lateral bud outgrowth initiation and subsequent lateral bud development into branches were investigated. We used a BC1 F7 population developed from Solanum lycopersicum (SL) and its close wild relative Solanum pimpinellifolium (SP) to measure the number of leaves on primary shoot (LN), the proportion of nodes with lateral buds (LS0) and lateral branches longer than 5 cm (LS5) and 10 cm (LS10) on main axis, the proportion of nodes with secondary lateral branches longer than 1 cm (LSLS1) on primary lateral branches, and the lengths of the longest primary and secondary lateral branches (LLSL1 and LLSL2, respectively). Composite interval mapping detected 17 additive QTLs; those for LS5, LS10, and LLSL1 clustered near C2 At5g49480 or LEATPACb on chromosome 1 and LEOH361 on chromosome 4. One epistatic (QTL × QTL) interaction was identified for LS10 QTLs on chromosomes 1 and 12, where recombinant-type alleles increased lateral bud development into primary lateral branches. We analyzed further the QTLs clustered on chromosome 1 using two BC3 F4 populations containing introgression region from the SP, which revealed the co-location of QTLs for lateral shoot development and days to flowering time. These results suggest that lateral shoot development is regulated by some additive and epistatic QTLs, some of which possibly exhibit a pleiotropic effect on flowering time. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The patterns of axillary/lateral bud formation and their development determine plant architecture (Schmitz and Theres, 1999). Lateral buds have two developmental modes, monopodial and sympodial (Pnueli et al., 1998). The model plant Arabidopsis has monopodial shoot architecture, and its shoot development can be separated into three phases, vegetative, inflorescence, and flowering (Schmitz and Theres, 1999). The vegetative shoot consists of a short internode and rosette leaves. The inflorescence shoot is extended by internode expansion after the transition from the vegetative to reproductive stage. At the same time, the axillary buds become visible. Later, these axillary buds develop into lateral inflorescence shoots. The flowering shoots consist of an intermediate-length internode and solitary flowers with or without subtending bracts (Pnueli et al., 1998). By contrast, tomato has sympodial shoot architecture, and its development can be separated into two phases, vegetative and reproductive, which alternate
∗ Corresponding author. Tel.: +81 3 5841 5046; fax: +81 3 5841 1131. E-mail address: [email protected] (O. New Lee). http://dx.doi.org/10.1016/j.scienta.2015.05.026 0304-4238/© 2015 Elsevier B.V. All rights reserved.
regularly (Pnueli et al., 1998). The vegetative phase is terminated by the development of a cymose inflorescence after the production of 8–12 leaves. When reproductive growth is initiated by the outgrowth of the inflorescence, vegetative growth restarts with the development of lateral buds just below the inflorescence. This generates three more leaves until terminating with the next inflorescence initiation. In tomato, several mutants defective in meristem identity has been used to study the genetic control of shoot development and branching habit. The bushy (bu) mutant stimulates branching, resulting in a characteristic bushy appearance (Young and MacArthur, 1947). Campbell and Nonnecke (1974) reported another unusual branching mutant, lateral promoter (Lp), in which buds in the axils of the cotyledons appeared early, causing a bushier appearance. The self-pruning (sp) gene in tomato replaces flowers with leaves in the inflorescence and suppresses the transition of the vegetative shoot apex to a reproductive shoot (Pnueli et al., 1998). The lateral suppressor (ls) mutant does not form most of the axillary meristem (Schumacher et al., 1999) and is related to two proteins involved in negative regulation of GA signal transduction (GAI, RGA). In the blind (bl) mutant, the sympodial meristem is absent, and the number of flowers per inflorescence is reduced
118
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
commercial cultivar Solanum lycopersicum (M570018, normal branched tomato cultivar) and its close wild relative Solanum pimpinellifolium (PI124039, highly branched wild tomato) and a backcross of the F1 to the ‘M570018’ cultivar. The resultant BC1 F1 population was advanced using the single-seed descent method to obtain the BC1 F6 population, which was used for genotyping. The subsequent BC1 F7 generation was used for phenotypic evaluation. To confirm the effects of detected QTLs in chromosome 1 on flowering time and branch development, two BC3 F1 lines containing a 45 cM heterozygous fragment on chromosome 1 were selected and advanced to BC3 F4 populations by selfing. In addition to a fragment on chromosome 1, these two lines contained 13 and 16.2 cM heterozygous fragments near C2 At3g55120 and SSR115 to C2 At3g55120 markers on chromosome 5. A total of 100 plants from two BC3 F4 populations (1762050-7-16 and 1762050-7-34) were used for the QTL analysis. Fig. 1. Illustration of inflorescence and lateral branch development in tomato. To evaluate the lateral shoot development in a BC1 F7 and two BC3 F4 populations, the number of leaves on primary shoot (LN), the number of leaves on the main axis (TLN) was measured. The proportion of buds to LN was calculated as an indicator of the proportion of lateral buds released from apical dominance (LS0). The proportions of the number of these primary lateral branches to TLN were designated LS5 and LS10, respectively. The proportion of the number of secondary lateral branches longer than 1 cm to the leaf number of the longest primary branches (LSLS1), the length of the longest primary lateral branches (LLSL1) and the longest secondary lateral branches (LLSL2) were measured. The total number of inflorescences (IN) and the average number of flowers (AFN) on the longest primary lateral branch were measured 62 days after sowing.
(Schmitz et al., 2002). The jointless (j) mutant suppresses sympodial meristem identity, which reverts the inflorescence meristem to vegetative growth after forming 1–2 flowers (Mao et al., 2000; Szymkowiak and Irishi, 2006). Although genes regulating branching habit and shoot development in these mutants have been identified, lateral shoot development is also controlled by the cumulative effects of quantitative trait loci (QTLs), like most important agronomic traits such as yield, plant height, and flowering time. QTL analysis using phenotypic values and genetic marker information could enable us to determine the number of QTLs associated with lateral shoot development, their locations and epistatic effects between QTLs. In cereal crops, many QTLs that affect branching have been identified. Teosinte branched (tb1) is a major QTL that controls the reduction in axillary branching from teosinte to maize (Doebley et al., 1995). In rice, the identification of genes that underlie QTLs affecting branching type has been advanced using rice genome information (Doust, 2007). Peng et al. (2014) also reported that QTLs for primary panicle branch number was regulated by a gene for plant height and tiller number. In spray cut chrysanthemum, 16 additive QTLs for branching traits were identified (Peng et al., 2015). By contrast, QTLs controlling lateral shoot development has been little studied in tomato although the number of lateral shoots affects vegetative and reproductive traits, such as leaf mass, canopy structure, the number of flowers and yield (Navarrete and Jeannequin, 2000). In the present study, therefore, we tried to identify QTLs related to the initiation of lateral bud outgrowth and subsequent lateral bud development in tomato using 111 BC1 F7 lines and 100 BC3 F4 lines (Fig. 1). Our results are useful for understanding the genetic basis of lateral shoot development and branching type in tomato. 2. Materials and methods 2.1. Mapping population One hundred and eleven BC1 F7 lines (hereafter referred to as BILs) were derived from an initial cross between the
2.2. Phenotyping For phenotypic evaluation, the 111 BC1 F7 lines, along with their parents, were grown in a greenhouse with natural daylight in The University of Tokyo, Japan. Seven seeds from each of the BILs were sown into 200 mL plastic pots (7 cm diameter) containing mixed commercial growth medium (Engeibaido, Kureha, Tokyo, Japan; Soilmix, Sakata Seed Co., Yokohama, Japan) on 15 July 2010. On 12 August (four weeks later), five out of seven seedlings were selected and transplanted into 1 L plastic pots (14.5 cm diameter). The pots were arranged in two double rows with 130 cm spacing between the centers of the rows. The plants were spaced 30 cm apart within each row and 45 cm apart between rows. For QTL analysis using BC3 F4 families, two seeds from 50 lines, each of the two BC3 F4 families were sown into 500 mL plastic pots (9 cm diameter), with five pots per line, on 7 October 2010, and thinned to one plant per pot after emergence. On 29 October, the plants were transplanted into 1 L plastic pots (14.5 cm diameter). The plants were spaced 27 cm apart within rows and 45 cm apart between rows. The plants were watered daily throughout the experiment. A randomized complete block design was used in both experiments. No lateral branches were pruned in either experiments. To evaluate the number of axillary buds (lateral buds) in the 111 BC1 F7 lines, the number of leaves before the inflorescence (LN), i.e., the number of leaves on the primary shoot was measured 46 days after sowing the seeds, at which time the macroscopic appearance of the first inflorescence had occurred. At the same time, the number of visible lateral buds was counted, and the proportion of buds to LN was calculated as an indicator of the proportion of lateral buds released from apical dominance (LS0). The numbers of primary lateral branches longer than 5 cm and 10 cm and the number of leaves on main axis (monopodial and sympodial shoots), TLN, were measured 62 days after sowing. The proportions of the number of these primary lateral branches to TLN were designated LS5 and LS10, respectively. The proportion of the number of secondary lateral branches longer than 1 cm to the leaf number of the longest primary branches (LSLS1), the length of the longest primary lateral branches (LLSL1), and the longest secondary lateral branches (LLSL2) were measured 64 days after sowing. The total number of inflorescences (IN) and the average number of flowers (AFN) on the longest primary lateral branch were measured 62 days after sowing. In QTL analysis using the two BC3 F4 families, the traits for which QTLs were detected in the experiment using BC1 F7 BILs were measured; LS5 and LS10 were measured 72 days after sowing, while LSLS1, LSLS2, LLSL1, and LLSL2 were measured 74 days after sowing. The number of leaves (LN) before inflorescence was measured 46 days after sowing, and days to flowering (DTF) were measured as well. However, LS0 was not measured in this experiment because
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
119
Fig. 2. Frequency distribution for each trait in the BC1 F7 population. Means for the parental and BC1 F7 lines are shown by arrowheads and an arrow, respectively. () Solanum lycopersicum cv. ‘M570018’ (n = 5); () S. pimpinellifolium (PI124039) (n = 5); (↓) BC1 F7 (n = 111). No data for the parental lines SL and SP for AFN.
lateral buds were visible in most axils of the primary shoot even in SL plants. 2.3. Genotyping One hundred and seven polymorphic markers (44 SSR, 23COSII, 37CAPS, and three SNPs) were used to construct a linkage map. Marker analysis was carried out as described by Cagas et al. (2008) using DNA from the 111 BC1 F6 plants and parental lines. In brief, DNA was extracted from 0.1 g of bulked leaf material of 7 individuals in each BIL using the Nucleon PhytoPure Plant DNA extraction kit (Amersham Biosciences, Buckinghamshire, UK), according to the manufacturer’s protocol. PCR mixtures consisted of 200 ng of template DNA, 2 M forward and reverse primers, 0.2 mM dNTP mixture, 1.8 mM MgCl2 , 1 L 10× NH4 buffer and 0.25 units of Biotaq DNA polymerase (Bioline, London, UK) in 10 L of total reaction solution. PCR conditions were one cycle at 94 ◦ C for 5 min, followed by 35 cycles at 94 ◦ C for 30 s, 48–55 ◦ C for 45 s and 72 ◦ C for 45 s, and one cycle at 72 ◦ C for final extension for 5 min. For COSII and CAPS markers, the amplified products were digested using restriction enzymes at 37 ◦ C for 12 h. PCR products were separated by polyacrylamide gel electrophoresis for 3 h at 120 V. Gels were visualized under UV light after staining with SYBR Green I (Lonza, Rockland, ME, USA). 2.4. QTL analysis The MAPMAKER/EXP v. 3.0b software was used for linkage analysis in the ‘RI self’ mode (Lander et al., 1987). The Kosambi function was used to convert the recombination frequencies (r) to map distances in centiMorgans (Kosambi, 1944). QTL detection was performed with composite interval mapping (CIM) using Windows QTL Cartographer ver. 2.0 (Wang et al., 2003). The experimentwise LOD threshold level of significance was evaluated by
performing 1000 permutations of each character (P < 0.05) (Churchill and Doerge, 1994). Epistatic (QTL × QTL) effects of the QTLs were tested using the QTLNetwork 2.0 software, which utilizes a mixed model-based composite interval mapping approach (Yang et al., 2008). The QTLs detected were designated by the trait name followed by the chromosome number. If more than one QTL from the same trait was detected on a chromosome, serial letters were added after the chromosome number. 2.5. Statistical analysis In both experiments, the Kolmogorov–Smirnov test was conducted to compare the observed cumulative frequency distributions of each trait using Microsoft Excel. In the experiment using BILs, single-factor ANOVA and Pearson correlation analysis were performed using the SPSS for Windows v. 11.01 software package. In the experiment using two BC3 F4 families, associations between genotype class at the closest flanking marker and phenotypic values were tested with the non-parametric Mann–Whitney U test. 3. Results 3.1. Trait variation and correlations among traits The Kolmogorov–Smirnov test indicated that none of the traits followed a normal distribution (Fig. 2). The LLSL2 trait distribution was left-skewed because the length of the secondary lateral shoots for most lines was shorter than 5 cm. For all of the traits except AFN, the average values were greater in S. pimpinellifolium (SP) than in S. lycopersicum (SL). On the other hand, the average values of the BILs lay in between SP and SL and the differences were significant among them, except for LLSL2 (Table 1; Fig. 2); the difference in LLSL2 was not significant between SL and the BILs. Statistical analysis of AFN could not be carried out because of missing data for the
120
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
Table 1 Averages (±SD) and statistical analysis of traits related with lateral shoot development and floral development in Solanum lycopersicum (SL), S. pimpinellifolium (SP), and 111 BC1 F7 lines (BILs). Comparisonsb Traitsa
SL
SP
RIL
SL:SP
SL:RILs
SP:RILs
LN LS0 LS1 LS5 LS10 LSLS1 LLSL1 LLSL2 IN AFNc
6.8 ± 0.89 0.62 ± 0.17 0.14 ± 0.14 0.03 ± 0.04 0.02 ± 0.03 0.03 ± 0.09 6.39 ± 7.83 0.09 ± 0.15 1.82 ± 0.40 –
11.75 ± 0.77 0.93 ± 0.07 0.77 ± 0.09 0.58 ± 0.08 0.50 ± 0.07 0.55 ± 0.08 69.47 ± 8.57 33.21 ± 11.45 5.8 ± 0.77 –
8.46 ± 0.65 0.76 ± 0.15 0.39 ± 0.18 0.15 ± 0.065 0.10 ± 0.05 0.19 ± 0.12 29.43 ± 14.40 3.73 ± 4.50 2.93 ± 0.66 4.96 ± 0.81
** ** ** ** ** ** ** ** ** –
** ** ** ** ** ** * NS * –
** ** ** ** ** ** ** ** ** –
a LN, leaf number on the primary shoot; LS0, proportion of nodes of lateral buds released from paradormancy; LS5 and LS10, ratios of the number of primary lateral branches longer than 5 cm and 10 cm to LN; LSLS1, ratio of secondary lateral branches longer than 1 cm to the number of nodes at the longest primary lateral branches; LLSL1 and LLSL2, lengths of the longest primary and secondary lateral branches (cm); IN, total number of inflorescences; AFN, average number of flowers on the longest lateral branches. b NS = not significant at P < 0.05; * and ** = significant at P < 0.05 and 0.01, respectively. c No statistical analysis was carried out for the trait because of a lack of data for the parental lines SL and SP.
Trait
LN
LS0
0.176
LS5
0.074
LS10
0.117
LS0
LS5
LS10
LSLS1
LLSL1
LLSL2
IN
parent plants. All of the traits except LN and LS0 were positively and significantly correlated with each other (Fig. 3). LN was only correlated with IN, and LS0 correlated with LSLS1.
presence of the SP allele stimulated elongation of the secondary lateral branches. For IN, only one additive QTL was detected in the vicinity of LEATPACb on chromosome 1, explaining 20% of the phenotypic variation (Table 2). For AFN, three additive QTLs were detected on chromosomes 2, 5, and 10, and together explained 37% of the phenotypic variation in this trait (9%, 19%, and 9%, respectively). The presence of SP alleles in the detected QTLs for IN and AFN increased the total number of inflorescences and the average number of flowers per inflorescence. One epistatic (QTL × QTL) interaction was identified for LS10 on chromosomes 1 and 12 (Table 3). The epistasis between ls101 and ls10-12 explained 12% of the phenotypic variation and the recombinant-type of the parental alleles (Qi Qi qj qj or qi qi Qj Qj ) stimulated elongation of primary lateral branches.
3.2. QTLs detected in 111 BC1 F7 BILs
3.3. QTLs detected in two BC3 F4 families
Seventeen QTLs with additive effects and one epistatic QTL pair were detected for the eight traits related with lateral shoot development (Fig. 4; Table 2). Among 17 QTLs with additive effects, seven were co-located between LEATPACb and C2 At5g49480 on chromosome 1 and three were near the marker LEOH361 on chromosome 4, while no co-location was found for the other seven additive QTLs mapped to chromosomes 1, 2, 5, 6, and 10. Only one epistatic interaction pair was detected, for LS10 on chromosomes 1 and 12. For LN, three additive QTLs were detected on chromosomes 1, 6, and 10, explaining 22%, 9%, and 7% of the phenotypic variation, respectively (Table 2). The presence of SP alleles for ln-1 on chromosome 1 increased LN by 0.54, while the presence of SP alleles on chromosome 6 and 10 decreased LN by 0.37 and 0.33, respectively. The ls5-1, ls10-1, and llsl2-1 QTLs were co-located near the marker C2 At5g49480, while the lsls1-1 and llsl1-1 QTLs were colocated near the marker LEATPACb on chromosome 1. These QTLs, which formed a cluster between LEATPACb and C2 At5g49480 on chromosome 1, explained 13%, 16%, 19%, 31%, and 13% each of the phenotypic variation, respectively (Fig. 4; Table 2). The ls5-4, ls104, and llsl1-4 QTLs were co-located in the vicinity of LEOH361 on chromosome 4, explaining 10%, 25%, and 19% of the phenotypic variation, respectively. The presence of SP alleles in the detected QTLs on chromosomes 1 and 4 increased the development of lateral shoots in terms of LS5, LS10, LSLS1, LLSL1, and LLSL2. Another QTL, llsl2-1, detected in the vicinity of SSR478 on chromosome 1 explained 41% of the phenotypic variation of the trait, and the
The two BC3 F4 families (1762050-7-16 and 1762050-7-34) were grown and evaluated for flowering time and lateral shoot development. The two BC3 F4 families, chromosome segment substitution lines (CSSLs) containing two introgression regions from the SP allele (45 cM in chromosome 1 and about 15 cM in chromosome 5). No QTLs related flowering time and branch development were detected on chromosome 5, except for AFN QTL. At the closest flanking marker on chromosome 1, the Mann–Whitney test revealed that DTF was significantly lower in plants homozygous for SP alleles than in plants homozygous for SL alleles, while other traits related to lateral shoot development and LN were significantly higher in plants homozygous for SP alleles (Table 4). All of the traits were positively and significantly correlated with each other (data not shown). A single QTL was detected for all of the traits except LS10 (Table 5). These QTLs formed two clusters; QTLs for LN, LLSL1 and LLSL2 were co-located in the vicinity of C2 At5g49480 while QTLs for LS5 and LSLS1 were co-located in the vicinity of SSR105 (Table 5; Fig. 5a). By contrast, dtf1 was located 4–5 cm upstream of llsl1 and llsl2. Furthermore, LOD score profile of DTF QTL was different from those of QTLs for LS5 and LSLS1, and LOD score profile of QTLs for LLSL1 and LLSL2 was intermediate (Fig. 5b). However, the one-LOD support confidence intervals of these QTLs overlapped. The SL allele for dtf-1 increased DTF, while the SP alleles for ln-1, ls5-1, lsls-1, llsl1-1 and llsl2-1 increased LN, LS5, LSLS1, LLSL1 and LLSL2, respectively (Table 5). dtf-1, ls5-1, and lsls1-1 with high LOD scores (4.62 to 8.08) explained phenotypic variations ranging from 21 to 39%, while ln-1, llsl1, llsl2 with low LOD scores (2.19–3.64)
0.053 0.120
0.860**
LSLS1 0.151
0.198* 0.320** 0.292**
LLSL1 0.065
0.099
0.365** 0.438** 0.571**
LLSL2 0.179
0.004
0.220*
0.314** 0.473** 0.688**
IN
0.257** 0.140
0.393** 0.362** 0.336** 0.344** 0.308**
AFN
0.133
0.310** 0.281** 0.237** 0.239*
0.025
0.252** 0.396**
Fig. 3. Correlation between the traits in the BC1 F7 population. * and ** indicate significance at P < 0.05 and P < 0.01, respectively.
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
121
Table 2 Additive QTLs related with lateral shoot development and floral development. Traitsa
Chromosome
Molecular marker
QTL positionb
LOD score
Additive effectc
R2
LN
1 6 10 1 4 1 4 1 4 1 1 1 2 1 2 5 10
C2 At5g49480 SSR578 SSR248 C2 At5g49480 LEOH361 C2 At5g49480 LEOH361 LEATPAC LEOH361 SSR478 C2 At5g49480 LEATPAC SSR32 LEATPAC TG337 C2 At3g55120 C2 At5g06430
41.49 23.11 37.96 47.55 6.26 46.55 10.26 43.66 11.56 10.01 52.02 38.55 59.25 41.04 68.83 32.15 0.01
8.31 3.04 2.65 4.03 3.32 5.90 6.55 4.17 3.79 4.21 5.05 6.87 4.43 5.43 3.02 5.05 3.08
−0.54 0.37 0.33 −0.028 −0.026 −0.027 −0.033 −65.25 −79.58 −53.77 −25.3 −0.10 −0.06 −0.36 −0.32 −0.52 −0.33
0.22 0.09 0.07 0.13 0.10 0.16 0.25 0.13 0.19 0.41 0.19 0.31 0.13 0.20 0.09 0.19 0.09
LS5 LS10 LLSL1 LLSL2 LSLS1 IN AFN
a
See the caption of Fig. 1 for trait abbreviations. Position on chromosome in cM. c Positive values for additive effects indicate that S. lycopersicum alleles increased the phenotypic value more than S. pimpinellifolium alleles, whereas negative values indicate that S. pimpinellifolium alleles increased the phenotypic value. b
Table 3 Epistatic QTLs related with lateral shoot development. Trait
QTLi
Chra
Interval (SIb )
Position
QTLj
Chra
Interval
(SIb )
LS10
ls101
1
SSR117 −SSR65 (99.9−115.2)
105.9
ls1012
12
SSR345−C2 At1g48300 (35.8−57.9)
Position
AAc
H2 AA
d
44.3
−0.0239 0.1201
P-valuee 0.000001
a
Chromosome number. Support interval. AA, a positive value indicates that the presence of parental type alleles (Qi Qi Qj Qj or qi qi qj qj ) in the epistatic loci increased the phenotypic value more than the recombinant type (Qi Qi qj qj or qi qi Qj Qj ).Likewise, a negative value indicates that the recombinant type alleles increased the phenotypic value more than the parental type did. d H2 , the heritability of the additive effect. e P-value, the probability of the estimated or predicted QTL effect. b c
Table 4 Averages (±SD) and statistical analysis of traits related with lateral shoot development and days to flowering (DTF) in BC3 F4 lines. Phenotypic values were compared between plants homozygous for S. lycopersicum (SL) alleles, homozygous for S. pimpinellifolium (SP) alleles and heterozygous for S. lycopersicum and S. pimpinellifolium (SLSP) alleles in the BC3 F4 families, respectively. Comparisonsb
Genotype class at the closest flanking markers Traitsa
SLSL
SPSP
SLSP
SL:SP
DTF LN LS5 LS10 LSLS1 LLSL1 LLSL2
57.6 ± 2.47 5.64 ± 0.48 0.11 ± 0.07 0.52 ± 0.03 0.72 ± 0.14 18.6 ± 16.2 1.66 ± 2.48
54.1 ± 1.70 6.44 ± 1.70 0.20 ± 0.08 0.06 ± 0.02 0.21 ± 0.10 45.0 ± 23.2 14.24 ± 22.87
57.6 ± 2.47 6.20 ± 0.63 0.14 ± 0.07 0.06 ± 0.01 0.18 ± 0.10 45.5 ± 31.5 4.50 ± 6.74
* * ** ** ** ** **
a
See the caption of Fig. 1 for trait abbreviations. Significance of marker association was determined for genotypic classes using the Mann–Whitney test between Solanum lycopersicum (SL) and S. pimpinellifolium (SP). NS = not-significant at P < 0.05; * and ** = significant at P < 0.05 and 0.01, respectively. b
Table 5 Additive QTLs related with lateral shoot development in the experiment on BC3 F4 lines. Traitsa
Chromosome
Molecular marker
QTL positionb
LOD score
Additive effectc
R2
DTF LN LS5 LSLS1 LLSL1 LLSL2
1 1 1 1 1 1
C2 At5g49480 SSR105 SSR105 SSR105 C2 At5g49480 C2 At5g49480
12.01 17.2 19.2 21.2 16.01 17.01
8.08 2.19 4.62 7.18 3.64 2.78
1.98 −0.26 −0.05 −0.10 −13.97 −6.07
0.39 0.09 0.21 0.29 0.17 0.13
a
See the caption of Fig. 1 for trait abbreviations. Position on chromosome in cM Positive values for additive effects indicate that S. lycopersicum alleles increased the phenotypic value more than S. pimpinellifolium alleles, whereas negative values indicate that S. pimpinellifolium alleles increased the phenotypic value. b c
122
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
Fig. 4. QTL map for traits related with lateral shoot development, based on a BC1 F7 population derived from a cross between Solanum lycopersicum (SL) and its close wild relative Solanum pimpinellifolium (SP). Names of the markers and QTLs are shown on the right and left of chromosomes, respectively. Arrowheads indicate the F-statistic peak of each QTL. Black and white symbols indicate that the QTL exhibited additive effects and epistatic effects, respectively. The associated bars denote the support confidence interval of each QTL.
explained smaller phenotypic variations ranging from 9 to 17%. 4. Discussion Tomato plants exhibit a sympodial growth pattern; when the shoot apical meristem of the primary shoot is terminated by the initiation of the first inflorescence, the lateral bud just below the first inflorescence develops into a sympodial shoot (Vegetti and Pilatti, 1998). However, some lateral buds formed at middle and lower nodes can also develop into sympodial shoots, showing a large genetic variation in the number of lateral branches. Indeterminate tomatoes are usually trained to maintain a single stem; lateral branches just below inflorescences are retained but other lateral branches are removed to enable easy handling and harvesting and improve air flow through the canopy (Navarrete and Jeannequin, 2000). Therefore, cultivars that produce a small number of lateral branches have attracted the interest of tomato breeders because manual pruning is labor intensive (Reinhardt
and Kuhlemeier, 2002). On the other hand, determinate tomato cultivars do not require pruning (Kemble and Gardner, 1992) or need only minimal pruning (Olson et al., 2006) because of their compact plant type. However, a large number of lateral branches are a prerequisite for high yield, and cultivars with the dominant lateral promoter gene (Lp) have been used for breeding high yielding determinate tomatoes (Ito et al., 1990). Therefore, the genetic basis for the variation of lateral shoot development needs to be studied in determinate as well as indeterminate tomatoes. Many QTL analyses in tomato have focused on flowering time and fruit related traits, e.g., fruit weight and soluble solids (Weller et al., 1988; De Vicente and Tanksley, 1993; Grandillo and Tanksley, 1996; Jimenez-Gomez et al., 2007; Cagas et al., 2008; Sumugat et al., 2010; Barrios-Masias and Jackson, 2014). However, there have been few reports on QTLs related to lateral shoot development in tomatoes. This study showed that additive QTLs related to lateral bud outgrowth and subsequent development into branches were mainly clustered on two chromosomes, chromosomes 1 and
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
123
Fig. 5. (A) Genetic location of QTLs for DTF, LN, LS5, LSLS1, LSLS1, and LLSL2 in the BC3 F4 populations (B) LOD score profiles for QTLs within the introgression region from the SP allele on chromosome 1.
4 (Fig. 4). On chromosome 1, QTLs for seven traits (ln, ls5, ls10, lsls1, llsl1, llsl2, in) were detected concurrently between the markers LEATPACb and C2 A5g49480, explaining 13–41% of the phenotypic variation. In this region, several QTLs related to flowering time and root growth have been identified in previous studies using BILs derived from the same parental accessions (Cagas et al., 2008; Sumugat et al., 2010, 2011). Sumugat et al. (2010) assumed that a single QTL acts as a ‘master gene’ to control multiple developmental processes, e.g., DTF, days to macroscopic flower bud appearance and number of leaves initiated, through pleiotropic effects. The co-localization of LS5, LSLS1, LLSL1, and LLSL2 QTLs suggests that the developmental processes of primary and secondary lateral shoots are regulated by the same ‘master gene’ that Sumugat et al. (2010) suggested for several flowering time traits. However, we could not clarify whether LS5, LSLS1, LLSL1, and LLSL2 QTLs were regulated by a pleiotropic QTL or closely linked QTLs in the experiment using the BC1 F7 population. Thus, QTL analysis for these traits as well as DTF was performed using two BC3 F4 families in this study (Table 5; Fig. 5). The results showed that QTLs related with lateral shoot development were within the cluster region near C2 At5g49480 and SSR105, but it appears that this cluster can be divided into two sub-clusters and single QTL; llsl11 and llsl2-1 in the first sub-cluster, and ls5-1 and lsls1-1 in the second. On the other hand, DTF QTL was identified at a slightly different position of chromosome 1. Sub-clustering of QTLs related with lateral shoot development and separate detection of DTF QTL support the idea that the initiation of lateral bud outgrowth, extension of lateral shoots and flowering time are controlled by closely linked QTLs. However, the possibility that QTLs for LS5 and LSLS1 is located at the same position as QTLs for DTF, LLSL1 and LLSL2 cannot be ruled out because their support confidence intervals overlapped. To confirm these QTL positions more precisely, it is necessary to increase the number of polymorphic markers and narrow the gap or to develop single-chromosome segment substitution lines. In connection with the existence of a pleiotropic QTL in this region, it is noteworthy that Cagas et al. (2008) reported that a QTL for the number of lateral buds longer that 0.5 cm was co-located with the DTF QTL in the vicinity of C2 At5g49480 and that this QTL had a negative additive effect similar to LS5 and LSLS1.
Werner et al. (2003) reported that cytokinins have important, but opposite, functions in the regulation of the activities of lateral shoot meristems and root meristems; overexpression of the cytokinin oxidase (CKX1) gene retarded leaf formation but increased root elongation in Arabidopsis. Sumugat et al. (2011) raised 110 BC1 F6 lines derived from the same parents as this study using 128-cell trays and evaluated the root ball weight of seedlings collected 19 days after sowing. They identified a QTL for root ball development between SSR105 and SSR134 on chromosome 1, in a different position to the QTLs for lateral shoot development (LEATPACb, C2 A5g49480 to SSR 105) identified in this study. Therefore, it is not clear whether lateral shoot development is negatively correlated with root growth. Paran et al. (1997) reported that the QTL for leaf length at node number 3 was co-located with the QTL for the number of side branches longer than 2 cm on chromosome 3, using an inbred population from a cross between S. lycopersicum and Solanum cheesmaniae. However, we did not detect any QTLs related to lateral shoot development on chromosome 3 in this study. On the other hand, we detected one independent QTL cluster related to lateral shoot development on chromosome 4, in which additive QTLs for three traits (LS5, LS10, and LLSL1) were co-located near the marker LEOH 361 (Table 2; Fig. 4). The presence of an SP allele in this region could enhance the initiation of lateral branch outgrowth (LS5; LS10) and the elongation of primary lateral branches (LLSL1). In this study, experiment using a BC1 F7 population was carried out in summer and experiment using two BC3 F4 populations was carried out in autumn. As a result, QTLs for LN, LS5, LSLS1, LLSL1, and LLSL2 on chromosome 1 were detected across seasons, indicating these QTLs were stable QTL. However, LS10 on chromosome 1 was only detected in experiment using a BC1 F7 population. This suggested that some QTLs have QTL × environment interaction effects. Thus, it seems necessary to carry out QTL analysis under different seasons for better understanding the genetic control of lateral shoot development. In this study, only one epistatic interaction was detected; one QTL pair on chromosomes 10 and 12 for LS10 (Table 3; Fig. 4). The epistatic interaction of the ls10-10 and ls10-12 QTLs accounted for 12% of the phenotypic variation and the recombinant-type parental alleles increased LS10 by 0.024. Two additive QTLs for LS10 detected
124
O. New Lee et al. / Scientia Horticulturae 192 (2015) 117–124
on chromosome 1 and 4 accounted for 16% and 25%, respectively, and their additive values were around −0.03. These results suggest that any attempt to use these QTLs in breeding programs needs to take into account such epistatic effects but also additive effects. 5. Conclusion In conclusion, this is the first study on QTLs that control the vigorous development of lateral branches in tomatoes. We suggest that shoot lateral branching is predominantly controlled by closely linked QTLs near the C2 At5g49480 marker on chromosome 1 and those near LEOH361 on chromosome 4, along with one pair of epistatic QTLs. Furthermore, our results suggest that a small target segment derived from S. pimpinellifolium could shorten DTF and stimulate lateral branching because the QTLs were co-located near the C2 At5g49480 marker on chromosome 1. References Barrios-Masias, F.H., Jackson, L.E., 2014. California processing tomatoes: morphological, physiological and phonological traits associated with crop improvement during the last 80 years. Eur. J. Agron. 53, 45–55. Cagas, C.C., Lee, O., Nemoto, K., Sugiyma, N., 2008. Quantitative trait loci controlling flowering time and related traits in a Solanum lycopersicum × S. pimpinellifolium cross. Sci. Hortic. 116, 144–151. Campbell, C.G., Nonnecke, I.L., 1974. Inheritance of and enhanced branching character in the tomato (Lycopersicon esculentum Mill.). J. Am. Soc. Hortic. Sci. 99, 358–360. Churchill, G.A., Doerge, R.W., 1994. Empirical threshold values for quantitative trait mapping. Genetics 138, 963–971. De Vicente, M.C., Tanksley, S.D., 1993. QTL analysis of transgressive segregation in an interspecific tomato cross. Genetics 134, 585–596. Doebley, J., Stec, A., Gustu, C., 1995. Teosinte branched 1 and the origin of maize: evidence for epistasis and the evolution of dominance. Genetics 141, 333–346. Doust, A.N., 2007. Grass architecture: genetic and environmental control of branching. Curr. Opin. Plant Biol. 10, 21–25. Grandillo, S., Tanksley, S.D., 1996. QTL analysis of horticultural traits differentiating cultivated tomato from the closely related species Lycopersicon pimpinellifolium. Theor. Appl. Genet. 92, 935–951. Ito, K., Kamimura, S., Mochizuki, T., Ishiuchi, D., Kanno, A., 1990. New line tomato for processing, ‘Morioka No. 22’. Bull. Natl. Res. Inst. Veg. Ornament. Plants Tea (C) 1, 11–20, Japanese with English Summary. Jimenez-Gomez, J.M., Alonso-Blanco, C., Borja, A., Anastasio, G., Angosto, T., Lozano, R., Martinez-Zapater, J.M., 2007. Quantitative genetic analysis of flowering time in tomato. Genome 50, 303–315. Kemble, J.M., Gardner, R.G., 1992. Inheritance of shortened fruit maturation in the cherry tomato Cornell 871213-1 and its relation to fruit size and other component of earliness. J. Am. Soc. Hortic. Sci. 117, 464–650. Kosambi, D.D., 1944. The estimation of map distances from recombination values. Ann. Eugen. 12, 172–175. Lander, E.S., Green, P., Abrahamson, J., Barlow, A., Daly, M.J., Lincoln, S.E., Newburg, L., 1987. MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1, 174–181.
Mao, L., Begum, D., Chuang, H.W., Budiman, M.A., Szymkowiak, E.J., Irish, E.E., Wing, R.A., 2000. JOINTLESS is a MADS-box gene controlling tomato flower abscission zone development. Nature 406, 910–913. Navarrete, M., Jeannequin, B., 2000. Effect of frequency of axillary bud pruning on vegetative growth and fruit yield in greenhouse tomato crops. Sci. Hortic 86, 197–210. Olson, S.M., Stall, W.M., Momol, M.T., Webb, S.E., Taylor, T.G., Smith, S.A., Simonne, E.H., McAvoy, E., 2006. In: Olson, S.M., Simonne, E.H. (Eds.). Vegetable production handbook for Florida, 2006–2007: Tomato production in Florida. University of Florida, Institute of Food and Agricultural Science and Vance Publishing, Lincolnshire. pp. 407–426 Cited by Santos, B.M (2008). Early pruning effects on ‘Florida-47 and ‘Sungard’ tomato HortTech. 18, 2006, 467–470. Paran, I., Goldman, I., Zamir, D., 1997. QTL analysis of morphological traits in a tomato recombinant inbred line population. Genome 40, 242–248. Peng, Y., Gao, Z., Zhang, B., Liu, C., Xu, J., Ruan, B., Hu, J., Dong, G., Guo, L., Linag, G., Qian, Q., 2014. Fine mapping and candidate gene analysis of a major QTL for panicle structure in rice. Plant Cell Rep. 33, 1843–1850. Peng, H., Zhang, F., Jiang, J., Chen, S., Fang, W., Guan, Z., Chen, F., 2015. Identification of quantitative trait loci for branching traits of spray cut chrysanthemum. Euphytica 202, 385–392. Pnueli, L., Carmel-Goren, L., Hareven, D., Gutfinger, T., Alvarez, J., Ganal, M., Zamir, D., Lifschitz, E., 1998. The SELF-PRUNING gene of tomato regulates vegetative to reproductive switching of sympodial meristems and is the ortholog of CEN and TFL1. Development 125, 1979–1989. Reinhardt, D., Kuhlemeier, D., 2002. Plant architecture. EMBO Rep. 3, 846–851. Schmitz, G., Theres, K., 1999. Genetic control of branching in Arabidopsis and tomato. Curr. Opin. Plant Biol. 2, 51–55. Schmitz, G., Tillman, E., Carriero, F., Fiore, C., Cellini, F., Theres, K., 2002. The tomato Blind gene encodes a MYB transcription factor that controls the formation of lateral meristems. Proc. Natl. Acad. Sci. U. S. A. 99, 1064–1069. Schumacher, K., Schmitt, T., Rossberg, M., Schmitz, G., There, K., 1999. The lateral suppressor (Ls) gene of tomato encodes a new member of the VHIID protein family. Proc. Natl. Acad. Sci. U. S. A. 96, 290–295. Sumugat, M.R., Lee, O., Nemoto, K., Sugiyama, N., 2010. Quantitative trait loci analysis of flowering-time-related traits in tomato. Sci. Hortic. 123, 343–349. Sumugat, M.R., Lee, O., Mine, Y., Sugiyama, N., 2011. Quantitative trait loci analysis of transplanting time and other root-growth-related traits in tomato. Sci. Hortic. 129, 622–628. Szymkowiak, E.J., Irishi, E.E., 2006. JOINTLESS suppresses sympodial identity in infloresecence meristems of tomato. Planta 223, 646–658. Vegetti, A.C., Pilatti, R.A., 1998. Structural patterns of tomato plants grown in glasshouse. Genet. Resour. Crop Evol. 45, 87–93. Wang, S., Basten, C.J., Zeng, Z.B., 2003. Windows QTL CARTOGRAPHER 2.0. Department of Statistics, North Carolina State University, Raleigh, NC http:// statgen.ncsu.edu/qtlcart/WQTLCart.htm Weller, J.I., Soller, M., Brody, T., 1988. Linkage analysis of quantitative traits in an interspecific cross of tomato (Lycopersicon esculentum × Lycopersicon pipinelifolium) by means of genetic markers. Genetics 118, 329–339. Werner, T., Motyka, V., Laucou, V., Smets, R., Van Onclelen, H., Schmülling, T., 2003. Cytokinin-deficient transgenic Arabidopsis plants show multiple developmental alterations indicating opposite functions of cytokinins in regulation of shoot and root meristem activity. Plant Cell 15, 2532–2550. Yang, J., Hu, C.C., Hu, H., Yu, R.D., Xia, Z., Ye, X.Z., Zhu, J., 2008. QTLNetwork: mapping and visualizing genetic architecture of complex traits in experimental populations. Bioinformatics 24, 721–723 http://ibi.zju.edu.cn/ software/qtlnetwork/ Young, P.A., MacArthur, J.W., 1947. Horticultural Characters of Tomatoes. Texas Agricultural Experiment Station Bulletin, pp. 698.