sented a combination of endogenous GA plus some. GA20 that accompanied the applied [3H]GA2a or metabolite from this exogenously applied GA20.
Plant Growth Regulation @ 1995 Kluwer Academic
271
16: 21 l-278,1995. Publishers,
Printed
in the Netherlands.
Heterosis and the metabolism of gibberellin A20 in sorghum Stewart B. Rood Department of Biological Sciences, University of Lethbridge, Alberta, Tl K 3M4, Canada Received 7 July 1994; accepted 29 September 1994
Key words: gibberellins, growth, heterosis, hybrid vigor, metabolism, Sorghum bicolor
Abstract The correlation between gibberellin (GA) metabolism and growth rate was investigated using two Sorghum bicolor inbred lines, Hegari and ATx 623, and their heterotic Fi hybrid. Previous studies have demonstrated that this hybrid is taller and has substantially greater shoot dry weights and leaf areas than either parental inbred. [3H]GA2c was applied to the leaf whorl of seedlings and after 24 hours, plants were harvested and separated into roots, shoot cylinders containing the apical meristems, and leaf blades. Chromatographic analyses of metabolites indicated the conversions of [3H]GAza to [3H]GAi sand29.The conversion of [*H]GAza to [*HIGAl was demonstrated by gas chromatography-selected ion monitoring (GC-SIM). Putative glucosyl conjugates of all of the [3H]GAs were also produced and GAs was identified by GC-SIM following enzymic cleavage of the putative [3H]GAs glucosyl conjugate fraction. Comparing the genotypes, [3H]GAza metabolism was more rapid in the shoot cylinders of the hybrid than in the shoot cylinders from inbreds. In the hybrid samples, there was a three-fold increase in the putative conjugate(s) of [3H]GAi which was the principal metabolite, and increased production of [3H]GAs and the putative conjugates of [3H]GAzs and [3H]GAs. Conversely, levels of the remaining precursor, [3H]GAza, and its putative conjugate(s) were reduced in the hybrid. The rate of GA20 metabolism was thus positively correlated with growth rate across these sorghum genotypes. This correlation supports a promotive role of GA in the regulation of shoot growth and in the expression of heterosis (hybrid vigor) in sorghum. 1. Introduction Despite its agricultural, horticultural and silvicultural importance, the physiological basis for heterosis, the rapid growth of some hybrids, remains unclear. An understanding of its physiological basis would assist with the systematic utilization of heterosis in crop breeding and would also contribute to the understanding of the physiological regulation of crop growth rate. Gibberellins (GAS) are probably involved in the regulationofheterosisofmaize[16,17,18],poplar[l], black spruce [26], and possibly, wheat [7]. The broader involvement of GAS in the regulation of geneticallydetermined growth rate is also supported by studies with Plantago major [3, 41 and various other plants. The observed slow growth of GA-deficient dwarf mutants of maize, rice, pea, Brassica and numerous other plants [ 141 indicates that GAS are involved in the regulation of growth rate in a broad range of plants.
Further, the application of plant growth retardants that block GA biosynthesis generally retard shoot growth in many crop plants [13]. In sorghum, the exogenous application of GA3 can promote shoot growth [9, 10, 271. The majR (maturity 3) gene of sorghum influences GA content and produces correlative changes in growth rate [2]. GA content is elevated in some sorghum hybrids, and slower-growing inbreds that contain lower levels of endogenous GAS may be particularly promoted by exogenous GA3 [21]. This suggests that the reduced GA content is a limitation to shoot growth rate in some sorghum genotypes including some inbreds. These previous studies with sorghum and with other plants collectively indicate that GAS are involved in the regulation of shoot growth rate and particularly, the control of height growth. It is consequently of interest to investigate the mechanism(s) by which genetic differences regulate the changes in GA content that
272 underlie genotypic differences in the shoot growth rate of crop plants. Thus, the present study was conducted to investigate aspects of GA metabolism that are correlated with growth rate across sorghum genotypes. The study involved a pair of slow-growing sorghum inbreds and their fast-growing F1 hybrid, an experimental system that utilizes heterosis as the genetic basis for major differences in shoot growth rate [ 121. Growth characteristics of the three genotypes have previously been described; the hybrid plants are consistently taller and have greater shoot dry weights and leaf areas than plants of either parental inbred [21].
2. Materials and methods 2.1 Plant materials Seeds of two parental lines of Sorghum bicolor, Hegari (PI 34911, forage type, male) and ATx 623 (grain type, female) and their FI hybrid, Hegari x AT x 623 (a commercially important forage type) were obtained from Professor F.R. Miller, Texas A & M Univ., Texas, USA [21]. Six seeds of a genotype were planted in each 14 x 13 cm pot filled with Metro-mix 200 (W.R. Grace & Co., Ajax, Ontario, Canada), a peat and vermiculite medium. Pots were watered to saturation daily, thinned to three uniform seedlings per pot following emergence, fertilized weekly with 0.25 g 28-l 4-14 with added micronutrients (Plant Prod 28-l 4-14, Plant Products Co. Ltd., Bramalea, Ontario) and positioned in a greenhouse at the University of Lethbridge (latitude 49.68” N) at about 22 OC (night) and 25 “C (day). Supplemental lighting for 14 h daily was presented with cool-white fluorescent tubes (140 pmol s- ’ me2 photosynthetic photon flux density, determined with a Li-Cor quantum sensor (Li-Cor Inc., Lincoln, Nebraska, USA). Plants of all three genotypes were grown, commencing Sept. 22, 1987 and again Jan. 3, 1989. Hybrid seed only was planted October 15,199l to investigate the occurrence of specific endogenous GAS. 2.2 Application of 3H and 2H-gibberellins [1,2,3-3H]GA2c (62 TBq mol-1) was obtained from Professor R.P. Pharis (Univ. of Calgary, Canada). [17,17-2Hz]GAzc (greater than 99% enrichment) was obtained from Professor L.N. Mander (Australian Nat’l. Univ., Canberra, Australia). GAS were dissolved in 95% ethanol @OH) and the solution was
then diluted to 10% EtOH. 0.1 ml of 10% aqueous EtOH with 3.3 kBq [3H]GA20 was applied by pipette into the leaf whorl of each of fifteen 13 day-old (days after seeding) plants at about 10 a.m., 4 h into the 14 h photoperiod. For the [2H]GA2a application, 3.3 kBq [3H]GA2e plus 140 ng [2H]GA2c were applied to each of five plants. After 24 h, plants were removed from the growing medium and cross-section cuts were made at the root crown and two cm above the root crown. No solution remained in the leaf whorls at harvest and samples were not rinsed prior to freezing. The sectioning produced three plant samples: (i) roots, (ii) shoot cylinders with the apical meristems, and (iii) leaf blades. Three replicates with five plants each were harvested, samples were frozen in liquid N2, lyophilized, and stored at -20 OC prior to analysis. 2.3 Analyses of GA metabolites Lyophilized samples were ground with a mortar and pestle and extracted with -20 ‘C 80% aqueous methanol (MeOH). Following vacuum filtration, extracts were evaporated in vacua to dryness with MeOH being added to assist in the removal of H20. Extract residues were dissolved in 50:50 MeOH:ethyl acetate (EtOAc) and aliquots were loaded onto glass fiber filter discs prior to step-elution silicic acid (SiO2) partition chromatography [22], eluted with a sequence of hexane, EtOAc and MeOH as previously described [20]. Regions of radioactivity from SiO2 columns were further resolved on gradient-elution reversed-phase Cis HPLC [S]. One-quarter of the MeOH eluate was incubated with cellulase (EC 3.2.1.4 from Aspergillus niger, Sigma) in 0. I MpH 4.2 sodium phosphate buffer for 24 h at 35 ‘C and rechromatographed on SiO2 and then HPLC in an attempt to cleave and consequently indirectly detect GA glucosyl conjugates [20,24]. Following selected analyses, HPLC peaks were dried, converted to the methyl ester, trimethylsilyl ether (MeTMSi) derivative [23] and the analyzed by GC-mass spectrometry with selected ion monitoring (GC-SIM) using a Hewlett-Packard 5890 GC coupled to a Hewlett-Packard 5970B mass selective detector. Chromatographic and mass spectrometer conditions were as previously described [5, 11,231. For quantitative comparisons of 3H-metabolites following the [3H]GA2a feed to the three genotypes, substances were assigned to metabolite groups based on sequential SiO2 and HPLC analyses. The proportions of various metabolites were determined for each
273
E
3o
SiO 2
Region
SiO,
II
Fraction
substances from a step-wise Fig. I. Elution of 3H-labelled SiO2 partition column following the application of [3H]GA2e leaf whorl of sorghum hybrid Hegari x ATx623.
20 eluted to the 10
of the three replicate samples (each sample containing material from five plants) and data were subsequently analyzed by one-factor analyses of variance (ANOVA), with 2 degrees of freedom between groups and 6 degrees of freedom within groups. Subsequent pair-wise comparisons were based on the Scheffe test.
3. Results and discussion
0 0
10
20 HPLC
30
40
Fraction
Fig. 2. Elution of 3H-labelled substances from gradient eluted reversed phase Cts HPLC columns loaded with SiO2 regions I (top) or II (bottom) from the SiO2 partition chromatography shown in Fig. 1. HFLC peak designations are included and fractions from which endogenous GAS were detected are indicated.
3.1 Gibberellin A20 metabolism Following the applications of [3H]GAaa to sorghum shoots, the distribution of metabolites was very similar to that previously observed following the application of [3H]GA2~ to maize 116, 191. SiOa chromatography separated three principal radioactive peaks and the first two, which were eluted with EtOAc and hexane, were grouped together to produce SiO2 Region I (Fig. 1). Since these substances were soluble in EtOAc, they would probably be principally non-conjugated (free) [3H]GAs. SiOa Region II was eluted with MeOH and would contain the more polar [3H]GA glucosyl conjugates (Fig. 1). Subsequent resolution of SiO;! Region I by HPLC revealed principal peaks eluting at the retention times (Rt) of [3H]GA2a and GAS that are metabolites of GA20 in maize and other plants [6, 19, 251 (Fig. 2). The largest [3H]GA metabolite peak (peak Ic) eluted at the Rt of GA1 while smaller peaks eluted at the earlier Rts of GA29 (Ib) and GAS (Ia) (Fig. 2). Two unknown metabolites, peaks Id and Ie, were also frequently observed (Fig. 1B). Corresponding HPLC Peaks from four extracts were bulked and analyzed by GC-SIM.
HPLC peaks Ic and If were found to contain GA1 and GA20, respectively (data not presented), GAS that have been repeatedly detected in the shoots of sorghum [2, 231. The (protio) GAS would probably have represented a combination of endogenous GA plus some GA20 that accompanied the applied [3H]GA2a or metabolite from this exogenously applied GA20. feed, GA1 was again Following the [2H2/3H]GA20 detected (SiO2 fraction 14-18; HPLC fraction 20-22; GC Rt 15.48 min (authentic GA,-MeTMSi, Rt 15.47 mm), coincidental peaks for ions 5 10, 508, 506 (ions from the molecular ion clusters), 491,448,450 a.m.u.). Abundances of the 508, 510 and 450 ions indicated enrichment of the GA1 with [2H2]GAi. In particular, the ion 508 abundance was 80.5% of that of the 506 ion, substantially greater than the 13 to 16% for the 508 ion from analyses of endogenous or standard GA1 . The GC-SIM analysis of HPLC peak Ib that eluted at the Rt of GA29 revealed a prominent 506 a.m.u. ion peak at the gas chromatography Rt of MeTMSi-GA29 but contaminants obscured fragment ion peaks, preventing confirmation of the occurrence of GA29. GA29 was not
274 identified in previous analyses of sorghum shoots [2, 231 suggesting that this GA occurs in low abundance. GC-SIM analysis of HPLC peak Ia, the putative GAs peak from SiOz Region I, failed to detect GAs. However, following the enzymic cleavage of an aliquot of SiOz Region II, and HPLC peak eluted at the same Rt as Ia and GC-SIM analysis revealed the occurrence of MeTMSi-GA8 (authentic GAs: HPLC fraction 911; GC Rt 17.33 min.; ions (m/z (% abundance)): 594 (lOO), 579 (7), 535 (12), 448 (60); sorghum GA& HPLC fraction 9-11, GC Rt 17.34 min.; ions: 594 (loo), 579 (7), 535 (15), 448 (72)). This indicates the occurrence of GAs, a GA that had been proposed to occur in sorghum based on the presence of other endogenous 13-hydroxylated GAS and comparative GAS of other plant species [2,6,23]. As was the case with the GAi and GAzo, the GAs detected may have represented endogenous GAs and/or metabolite from the GA20 that accompanied the applied [3H]GA2~. The failure to previously detect GAs from sorghum shoots and the failure to detect free GAs in the present study suggests that this GA may occur predominantly in the conjugated form, presumably as GAs-glucoside [24]. While peak Ha, the apparent conjugate of r3H]GAs, was a minor metabolite, most of the radioactivity from SiO2 Region II eluted coincidental with or slightly prior to r3H]GAi (Fig. 2). This radioactive peak would be likely to contain glucosyl conjugate(s) of [3H]GAi since glucosyl conjugates elute from Cis HPLC coincidental with or slightly prior to corresponding free GAS [8]. The putative [3H]GAi conjugate(s) peak, IIc, was the largest metabolite peak occurring in extracts from the sorghum hybrid (Fig. 2). The apparent sequence of hydroxylations and conversions into polar, putative glucosyl conjugates is similar to the metabolic fate of GA20 in other plants and particularly in maize (Fig. 3) [6, 16, 19, 24, 251. The similarity with maize is expected since both are tropical, C4, monocotyledonous plants and are physiologically similar. 3.2 Quantitative comparison of [3H]GA2~ metabolism across sorghum genotypes Following the addition of r3H]GA2c to the leaf whorl of the three sorghum genotypes, most of the radioactivity was recovered in the filtered extracts. In Hegari, ATx623 and the hybrid, 86, 86 and 79%, of the total radioactivity was recovered, respectively, in the sum of the root, shoot and leaf samples. Since the 3H was largely located at the C-2 and C-3 positions, hydrox-
ylations at those positions would results in the loss of 3H. The possibly reduced recovery of 3H in the hybrid samples provided initial evidence that the hybrid metabolized the [3H]GA2~ more quickly than the parental inbreds. The loss of the 3H due to 2- or 3hydroxylation will result in an underestimation of the rates of the metabolic conversions. The highest proportion of the recovered radioactivity was most commonly associated with the shoot cylinder samples (mean 47.7%) and these samples provided the focus of subsequent analyses. In all shoot cylinder samples, two groupings eluted from the SiO2 columns and those were subsequently resolved by HPLC and generally yielded six principal radioactive peaks (Fig. 2). The qualitative pattern of elution of the HPLC peaks was generally similar from SiO;?regions I (free GAS) and II (putative conjugates), an expected pattern since glucosyl conjugates elute from gradient eluted Cis HPLC coincidental with or slightly earlier than the corresponding free GAS [8, 221. The qualitative profiles of metabolites were generally similar across genotypes although ATx623 did not produce detectable amounts of [3H]GAs or its putative conjugate in the experiment (Fig. 4). Quantitative differences across the genotypes were consistent across replicates resulting in significant (p c 0.05) differences in the proportions of certain 3Hlabelled substances in the shoot cylinders (Fig. 4). Firstly, the proportion of non-metabolized [3H]GAzo was consistently higher in the parental inbreds than in the hybrid (ANOVA, F = 25.54, p = 0.001). Shoot cylinders from the hybrid contained less than one-half of the proportion of r3H]GA2a that remained in the shoot cylinders from the inbred lines. Similar to this pattern for free [3H]GAzc, the proportion of putative conjugate [3H]GA2~ also tended (ANOVA, F=3.75, p = 0.088) to be lower in shoot cylinders of the hybrid (Figs. 2 and 4). While the putative [3H]GA2~ conjugate(s) was the principal metabolite in the inbreds, the most abundant metabolite in the hybrid was the putative conjugate(s) of [3H]GAi (Figs. 2 and 4). The proportion of this metabolite was three- to five-fold higher in the hybrid samples than in the inbred tissues (ANOVA, F=5.541, p=O.O43). However, the proportion of the free [3H]GAi was relatively constant across the three genotypes (Fig. 4). It is likely that the putative conjugate(s) of [3H]GAi originated from [3H]GAi and consequently, although [3H]GAi was equivalent at the time of harvest, previous metabolism through [3H]GAi would have occurred
275
GA19
-w-t
@$$
-
Ho@
G&o
GA1
;I
;I
-&
$YgK
GAS
GA29 in sorghum shoots based on results Fig. 3. Probable pathway of GA metabolism from sorghum shoots except GAzg. Dashed lines indicated probable interconversions solid line indicates the confirmed conversion based on a [*H]GA2o feed.
to a greater extent in the hybrid than in the parental inbreds. The hybrid shoot cylinders contained higher proportions of putative conjugate(s) of [3H]GA2a than shoot cylinders of either parent (Fig. 4) and higher proportions of [3H]GA2s than the shorter parental inbred, ATx623. Free [3H]GAs was detected only in the shoot cylinders from the hybrid, providing a substantial difference across the genotypes (Fig. 4, ANOVA, F = 28.84, p = 0.0008). There was no evidence for the occurrence of the putative conjugate(s) of [3H]GAs in inbred ATx623 and its occurrence in inbred Hegari tended to be lower than in the hybrid (Fig. 4). The patterns across the genotypes were consistent for most of the metabolites, with increased proportion of hydroxylated metabolites in the hybrid, intermediate levels in Hegari and lower levels in ATx 623. Thus, metabolism of [3H]GA~a, and particularly 2- and 3-hydroxylation occurred most rapidly in the hybrid. Very little radioactivity was detected in the root samples, with 2.7, 5.6 and 2.1% of the total recovered 3H in the roots of Hegari, ATx623 and the hybrid, respectively. Consequently, only a single root sample replicate was analyzed for each genotype and the distributions of metabolites in these root samples were relatively similar to those in the shoot cylinders.
of the present study. The GAS shown have been identified based on native occurrence and [3H]GA studies and the
For example, roots from ATx623 contained: 58.5% [3H]GA2a, 7.5% [3H]GAr, 27.2% putative [3H]GA~a conjugate and 6.3% putative [3H]GAr conjugate. As was observed for the ATx623 shoot cylinders, no [3H]GAs or putative 13H]GAs conjugate was detected in ATx623 roots. Due to the minimal levels of radioactive substances, more detailed comparisons of [3H]GA~~ metabolism in the roots of the three genotypes could not be made. The lack of radioactivity in the roots indicates that GAS are not transported basipetally (at least beyond the root crown) and correspondingly suggests that sorghum roots are not sinks for GAS, at least for applied GA20 and its metabolites. Variable proportions of recovered radioactivity were detected in the leaf samples with consistently higher proportions associated with the larger leaves that are typical of the hybrid (Hegari x ATx 623, mean = 86%, Hegari, 34%; ATx623, 27%) [12, 211. This increased level of radioactivity in the leaves of the hybrid occurred in spite of the larger diameter of the leaf whorl that would have increased the proportion of [3H]GA~a directly applied to the apical meristem and decreased the amount applied to the leaf surfaces. Thus, the export of the applied GA, whether passively with the transpirational stream or actively through other transport routes, was more rapid in the
276 60
25
50
20
40
15
30
10
20 10
5
0 12
0 50 40
9
30 6
20
z
3
10
-$ ii-
0 5
0 10
‘
