Logo of plntphysLink to Publisher's site
Plant Physiol. 1998 Mar; 116(3): 1003–1011.

Photoperiod Control of Gibberellin Levels and Flowering in Sorghum1


Regulation of rhythmic peaks in levels of endogenous gibberellins (GAs) by photoperiod was studied in the short-day monocot sorghum (Sorghum bicolor [L.] Moench). Comparisons were made between three maturity (Ma) genotypes: 58M (Ma1Ma1, Ma2Ma2, phyB-1phyB-1, and Ma4Ma4 [a phytochrome B null mutant]); 90M (Ma1Ma1, Ma2Ma2, phyB-2phyB-2, and Ma4Ma4); and 100M (Ma1Ma1, Ma2Ma2, PHYBPHYB, and Ma4Ma4). Plants were grown for 14 d under 10-, 14-, 16-, 18-, and 20-h photoperiods, and GA levels were assayed by gas chromatography-mass spectrometry every 3 h for 24 h. Under inductive 10-h photoperiods, the peak of GA20 and GA1 levels in 90M and 100M was shifted from midday, observed earlier with 12-h photoperiods, to an early morning peak, and flowering was hastened. In addition, the early morning peaks in levels of GA20 and GA1 in 58M under conditions allowing early flowering (10-, 12-, and 14-h photoperiods) were shifted to midday by noninductive (18- and 20-h) photoperiods, and flowering was delayed. These results are consistent with the possibility that the diurnal rhythm of GA levels plays a role in floral initiation and may be one way by which the absence of phytochrome B causes early flowering in 58M under most photoperiods.

The maturity gene Ma3 contributes to the timing of floral initiation in sorghum (Sorghum bicolor [L.] Moench). This gene has three known alleles, Ma3, ma3, and ma3R (Quinby, 1973). Plants containing ma3R differ from those homozygous for the other alleles in that they flower early, are relatively insensitive to a wide range of photoperiodic conditions, and elongate more rapidly and tiller less than non-ma3R plants. In contrast, genotypes containing Ma3 or ma3 are very sensitive to photoperiod, and flowering is delayed by long (12-h) days and hastened by short (10-h) days (Pao and Morgan, 1986). Other than relatively small differences in flowering dates, Ma3 and ma3 are phenotypically indistinguishable.

Recently, Ma3 was shown to code for phytochrome B and ma3R was shown to contain a deletion mutation that results in a transcription stop codon upstream of the proposed second dimerization site (Childs et al., 1997). Therefore, ma3R is an apparent null mutant and the three maturity gene alleles were renamed: Ma3 to PHYB, ma3R to phyB-1 (Childs et al., 1997), and ma3 to phyB-2 (present paper). Abbreviations for phytochrome genes are according to the proposal of Quail et al. (1994).

Increasing the photoperiod increases GA levels in many long-day plants, presumably as the result of daylength perception by phytochrome (Metzger and Zeevaart, 1980a, 1980b; Gianfagna et al., 1983; Davies et al., 1986; Rood et al., 1986). Photoregulation of GA metabolism may be the result of the presence or absence of light, the transition between light and dark, the duration or alteration of photoperiod length, the quality of the light (particularly red and far-red light), and the quantity of light.

Several of the steps in the early C-13β-hydroxylation pathway have been reported to be regulated by the light environment, including those for steps upstream of GA12 (Zeevaart et al., 1993), GA12→GA53 (Davies et al., 1986), GA53→GA44 (Graebe, 1987), GA19→GA20 (Metzger and Zeevaart, 1980b; Gianfagna et al., 1983), and GA20→GA1 (Campell and Bonner, 1986; Gilmour et al., 1986). Assays of GA1 at the end of the 8-h light period and at the end of the 16-h dark period showed that its level is diurnally regulated in long-day spinach (Talon et al., 1991). The same pattern for GA1 was observed when 8-h high-light days were extended with 16 h of dim far-red light, but GA20, the precursor to GA1, continued to increase during the extension, and the highest levels occurred at the end of the daily far-red light period.

GA biosynthesis in sorghum is diurnally regulated (Foster and Morgan, 1995). Under 12-h photoperiods, rhythmic patterns of GA12 and GA53 levels in phyB-1 and non-phyB-1 genotypes were similar, peaking at midday and having their minima at night. However, GA20 and GA1 exhibited a different pattern between phyB-1 and non-phyB-1 genotypes; levels peaked near dawn in phyB-1 and at midday in the non-phyB-1 genotypes (Foster and Morgan, 1995). This reveals that relative GA1 level differences between phyB-1 and non-phyB-1 genotypes depend on the time of day; the maximum difference occurs around dawn. The differences in GAs between genotypes are not in the absolute level, but in a shift in the timing of rhythmic peaks or pulses of GA20 and GA1. The timing of the rhythmic pulse of GA1 is altered or possibly uncoupled in 58M, and this change is correlated with the absence of a functional phytochrome B, altered photoperiodic sensitivity, early flowering, inhibition of tillering, and promotion of shoot growth (Pao and Morgan, 1986; Childs et al., 1991, 1992, 1995, 1997).

Production of a plant hormone in rhythmic diurnal pulses is not a unique discovery. Ethylene has been shown to be produced in rhythmic, often diurnal pulses in a number of studies (Lipe and Morgan, 1973; Rikin et al., 1984; Morgan et al., 1990; Ievinsh and Kreicbergs, 1992; Machácková et al., 1997). However, the observations on GA20 and GA1 are the first, to our knowledge, in which the rhythmic levels of hormones have been observed in a context that suggests a possible linkage with both a physiological process (promotion of floral initiation) and a regulatory mechanism (phytochrome perception of daylength) (Talon et al., 1991; Foster and Morgan, 1995). The existence of diurnal rhythms of GA levels in sorghum, and the change in the rhythm of the levels of GA20 and GA1 in plants deficient in phytochrome B, offer opportunities to further our understanding of photoperiodic control of both GA metabolism and flowering.

If the distinctive pattern of GA20 and GA1 levels seen under 12-h photoperiods is linked to early floral initiation, that pattern should be altered under long-day conditions, which delay floral initiation in the phyB-1 genotype. It should be of interest to determine whether the pattern of GA concentration rhythms associated with floral induction and noninduction are the same for phyB-1 and non-phyB-1 genotypes. We report here the characterization of the photoperiodic regulation of GA content in genotypes containing phyB-1 (58M) and those without phyB-1 (phyB-2 [90M] and PHYB [100M]).

In our studies with sorghum (for review, see Morgan, 1994), photoperiods have been presented as high-intensity light periods of varying length because photoperiodicity of sorghum had already been established when the work began. In addition, brief daylength extensions with far-red light had been shown to hasten flowering (Lane, 1963; Williams and Morgan, 1977), making the use of dim far-red light to extend a basic photoperiod questionable for this species. Sorghum also appears to be strongly thermoperiodic, because synchrony of photoperiods with thermoperiods influences flowering date (Morgan et al., 1987). Thus, to avoid confusion in interpreting data and to provide a strong, unambiguous cue to the plant, the experiments reported here involved variations in the length of the light period, which conformed to the timing of a 30°C thermoperiod. Unless specified otherwise, we use the term photoperiod to refer to treatments of varying duration of a high-intensity light period with a matching high-temperature period (30°C).


Plant Materials and Growth Conditions

Three of the maturity genotypes of sorghum (Sorghum bicolor [L.] Moench) used in this study are nearly isogenic except for the following differences at the third maturity-gene locus as recently redesignated: 100M (Ma1Ma1, Ma2Ma2, PHYBPHYB, and Ma4Ma4); 90M (Ma1Ma1, Ma2Ma2, phyB-2phyB-2, and Ma4Ma4); and 58M (Ma1Ma1, Ma2Ma2, phyB-1phyB-1, and Ma4Ma4) (Quinby, 1973; Childs et al., 1997). Seeds supplied by Dr. Fred Miller and Dr. Bill Rooney, Department of Soil and Crop Sciences, Texas A&M University, were germinated and grown in a fertilized peat moss-perlite mixture (Beall et al., 1991) in growth chambers (EGC, Chagrin Falls, OH) equipped with a mixture of cool-white fluorescent and incandescent lights, yielding a light intensity of 250 to 300 μmol m−2 s−1 (300–800 nm) measured at the pot surface with a portable spectroradiometer (model LI-1800, Li-Cor, Lincoln, NE). Temperatures were maintained at 30°C day/20°C night, with 10-, 14-, 16-, 18-, and 20-h photoperiods. Thermoperiod consistently corresponded to the day/night transition. To minimize the environmental variation, all three genotypes were grown at the same time in the same growth chamber in each photoperiod duration (except 20 h, which is described below).

Harvests started at the beginning of the light period 14 d after seeding at 3-h intervals for 24 h. When sampling times decreased at the transition between lights-on and lights-off, samples were taken immediately after lights-on at the beginning of the photoperiod and immediately before lights-off at the end of the photoperiod. For harvests during the dark period, plants were removed from the growth chamber in complete darkness and cut under a dim-green safelight. Plants were cut at the root-shoot junction and at the top of the tallest leaf collar. Samples containing the part of the culm between the root crown and the tallest leaf collar were obtained from each genotype. The basal three leaf blades and sheaths were removed (two at harvest and the third after freeze-drying). Because samples of 58M grown under 18- and 20-h photoperiods were tall and had more expanded leaves, a fourth leaf was removed. Two replicate samples were obtained at each harvest time. Samples were frozen in liquid N2 immediately after harvesting. After lyophilization, samples were stored at −20°C until extracted for GAs. To minimize the variation due to differential moisture absorption by tissue during storage, every sample was relyophilized just before weighing and extraction.

Gas Analysis

After methanolic extraction, GAs were purified using a combination of preparatory column chromatography, solvent partitioning, and reverse-phase HPLC (Foster et al., 1994; Foster and Morgan, 1995). Deuterated (25 ng each of [17,17-2H2]GA1, -GA8, -GA19, -GA20, -GA44, and -GA53, and 20 ng of [17,17-2H2]GA12) internal standards were added. Tritiated (1500 Bq each of [1,2-3H2]GA1 and [1,2-3H2]GA4) standards were also added to the combined extract to monitor recovery through the purification procedure. GAs were quantified using GC-MS selected ion monitoring by calculating the area ratio of endogenous GA to the deuterated standard GA that had been added during the extraction step, and the contribution from the deuterated standard to the nondeuterated GA was corrected (Beall et al., 1991). Data are given on a content or level basis. Rhythmic patterns of GA levels, expressed as nanograms per plant, were similar to those expressed as nanograms per gram dry weight; data on a per-plant basis are not given.

As discussed by Beall et al. (1991) and Foster et al. (1994), sorghum contains the early C-13β-hydroxylation pathway GAs demonstrated to occur in maize (Phinney, 1984), which presumably follow the biosynthetic pathway established in maize: precursor → GA12 → GA53 → GA44 → GA19 → GA20 → GA1 → GA8. Analysis focused on these compounds, but because the amounts of GA44 detected were very small, these data were not reported.

Experimental Design and Replication

All of the plants for the individual experiments (differing in photoperiod duration) reported in Figures Figures114 were grown in the same growth chamber at the same time, and multiple plants were harvested every every33 h. Two subsamples of separate plants were extracted and analyzed for each sample time for each genotype for each experiment. Means of the two assays are plotted in the figures, with bars showing the range. In addition to the replicate samples, there are several internal indications of validity of the data: (a) close agreement of data from 90M and 100M, which both exhibit a wild-type phenotype when grown in growth chambers (Pao and Morgan, 1986); (b) the general stepwise progress of levels with sampling times during the day; (c) the general agreement of patterns of GA levels and absolute levels from experiment to experiment and to previous experiments (Foster and Morgan, 1995); (d) the general difference between data from 58M (which exhibits a unique phenotype) and 90M and 100M; (e) the very small range of variability in levels of GAs in 58M, which has a lower chlorophyll and anthocyanin content than 90M and 100M (Childs et al., 1991), to be separated from GAs; and (f) the close agreement (small range) of GA1 levels in 58M grown in two different growth chambers at different times (Fig. (Fig.5,5, 19 d after seeding).

Figure 1
Diurnal regulation of GA levels in shoots of 14-d-old seedlings of 58M (•), 90M (▪), and 100M (▴) grown under 10-h photoperiods. The dark period is indicated by the solid bars at the top and bottom of the figure. GA levels were ...
Figure 3
Diurnal regulation of GA levels in shoots of 14-d-old seedlings of 58M (•), 90M (▪), and 100M (▴) grown under 16-h photoperiods. The dark period is indicated by the solid bars at the top and bottom of the figure. GA levels were ...
Figure 5
Diurnal regulation of GA levels in shoots of 14- (•) and 19-d-old seedlings (▪) of 58M, a phytochrome B mutant, grown under 20-h photoperiods. The dark period is indicated by the solid bars at the top and bottom of the figure. GA levels ...
Figure 4
Diurnal regulation of GA levels in shoots of 14-d-old seedlings of 58M (•), 90M (▪), and 100M (▴) grown under 18-h photoperiods. The dark period is indicated by the solid bars at the top and bottom of the figure. GA levels were ...

The experiment reported in Figure Figure55 involved only 58M to determine whether the patterns of GA1 levels exhibited by 58M under 18-h photoperiods persisted with a longer photoperiod (20 h) and with older plants (19 d old). For this 20-h-photoperiod experiment, we grew 58M in two different growth chambers to verify that microenvironmental differences in different growth chambers did not have a major effect on the rhythms of GA levels. Some of the samples from one set of 14-d-old plants (58M, 2:00 am, Fig. Fig.1)1) were lost after harvest. The data shown in Figure Figure66 are averages of all values in Figures Figures114, and sd values are plotted (see legend of Fig. Fig.6). 6).

Figure 6
Levels of endogenous GAs in shoots of 14-d-old seedlings of 58M (•), 90M (▪), and 100M (▴) under four different photoperiodic conditions. GA levels were measured by GC-MS selected ion monitoring using deuterated internal standards. ...


The results of earlier experiments suggest that GA12 levels are regulated by photoperiod in 58M (phyB-1phyB-1), 90M (phyB-2phyB-2), and 100M (PHYBPHYB) (Foster and Morgan, 1995). In all three genotypes, GA12 levels were high during the light periods and low during the dark periods. The absence of phytochrome B (Ma3 gene product) in 58M (Childs et al., 1992, 1997) had no effect on the rhythmic fluctuation of GA12 levels, since the timing of the maximum and minimum levels of GA12 were similar between phyB-1 and non-phyB-1 genotypes. In the present study the pattern of pulses of GA12 levels was not changed to a major degree by different photoperiod and thermoperiod durations ranging from 10 to 18 h (Figs. 1–4). The pattern was expressed most strongly in a 10-h photoperiod/thermoperiod (Fig. (Fig.1). 1).

With 14- and 16-h photoperiods, GA12 levels tended not to vary much during the day, especially in the non-phyB-1 genotypes under 14 h, and if levels did vary discernibly, peaks occurred later in the photoperiod. The daily peaks in GA12 were a little more strongly expressed in an 18-h photoperiod, although still peaking late in the day. Genotypes 58M, 90M, and 100M all possess a working circadian clock, as demonstrated by rhythmic pulses in chlorophyll a/b binding protein mRNA that continue in constant light (Childs et al., 1995). Zeevaart and Gage (1993) have demonstrated that ent-kaurene synthesis is regulated by photoperiod, and this may be a result of phytochrome action. ent-Kaurene levels were not measured in this study, and the participation of phytochrome in the control of GA-related metabolism upstream of GA12 or ent-kaurene was not evaluated.

However, if a phytochrome is responsible for diurnal regulation of GA12 level or biosynthetic steps upstream of GA12, it is unlikely to be phytochrome B, which is missing in 58M. The reason for this conclusion is that the pattern of GA12 levels is similar in phyB-1 and non-phyB-1 genotypes across photoperiods ranging from 10 to 18 h, except for the differences previously noted at 14 h (Figs. (Figs.114).

Levels of GA53 also showed a daytime peak in all three genotypes and all photoperiods tested (Figs. (Figs.114). This pattern is similar to that of GA12 and suggests that GA53 levels are determined by the levels of its precursor. However, the changes in levels of GAs observed in this study were not demonstrated to be caused by synthesis and could also result from breakdown or conjugation reactions. Levels of GA53 increased gradually after lights-on, peaked later in the light period than GA12 levels, and decreased to minimum levels at the end of the dark period. These diurnal patterns were lost as the presumed pathway of metabolism proceeded through GA44 to GA19. Neither genotype exhibited a distinctive pattern of GA19 levels. It should be noted that pool sizes of GA19 are several times higher than those of GA12 and GA53; therefore, daily pulses in the metabolism of GA53 and GA19 might have less impact on the total GA19 pool size. These patterns of GA12 and GA53 and the absence of a daily peak in GA19 levels are consistent with the observed patterns of these GAs under a 12-h photoperiod (Foster and Morgan, 1995).

In plants that are wild type (Quinby, 1973; Pao and Morgan, 1986) for all aspects of phenotype except flower initiation dates in some environments (90M and 100M), GA20 levels followed a discernible diurnal pattern for all photoperiods tested (Figs. (Figs.114). The peaks in GA20 levels, although small in magnitude, occurred 6 to 9 h after lights-on. Note, however, that a peak in levels of GA20 did not occur in 90M under 10-h photoperiods, making its GA20 pool behave more like 58M at the 2:00 and 5:00 pm harvests. The uniformity of the data for these non-phyB-1 genotypes indicates that the increase in GA20 is probably controlled by photoperiod. Although the rhythmic pattern of GA20 levels was not altered by variation of the photoperiod (10–18 h) in 90M and 100M, it was altered by photoperiod in the phytochrome B null mutant 58M.

In photoperiods of 16 h or less (Figs. (Figs.113), GA20 levels in 58M decreased gradually during the day, but in photoperiods of 18 h or more (Figs. (Figs.44 and and5),5), GA20 levels increased gradually to a peak and then declined. These results were reinforced by the rhythmic pattern of GA20 levels (increasing during the day) of 19-d-old 58M under a 20-h photoperiod (Fig. (Fig.5).5). Another way of viewing the pattern of GA20 levels is that under photoperiods of 16 h or less, the highest level during the light period occurred at the time of lights-on in 58M and 6 to 9 h later in 90M and 100M (Figs. (Figs.113).

Under the 18-h photoperiod, the pattern of GA20 levels was similar in all three genotypes (Fig. (Fig.4).4). The phase shift in the regulation of GA20 levels in 58M compared with 90M and 100M strongly suggests that the missing phytochrome B is required for proper wild-type regulation of the metabolism controlling the GA20 pool size (Figs. (Figs.113). This reaction in the mutant 58M was affected by photoperiod duration; a noninductive, long photoperiod (18 and 20 h) altered this pattern to one similar to that of wild type (100M) (Figs. (Figs.44 and and55).

Floral initiation of 58M was delayed by very long days (Childs et al., 1995). Under 18-h photoperiods, 58M initiated a floral meristem at about d 50 (Foster et al., 1997), whereas under 12-h photoperiods, it initiated at d 20 (Pao and Morgan, 1986). In this study under 18-h photoperiods, 58M did not initiate florally until d 60, and four to five internodes elongated before floral initiation. The same response was noted by Childs et al. (1995). It is interesting that in 58M under the 18-h photoperiod (Fig. (Fig.4),4), in which floral initiation was delayed, the pattern of GA20 levels was changed to one similar to that of 90M and 100M (wild type), which normally exhibit delayed floral initiation with a photoperiod only 12 h long. However, the correlation with flowering appears not to be absolute because in some experiments floral initiation is delayed by 16-h photoperiods (Childs et al., 1995), and that condition did not shift the GA20 content pattern here (Fig. (Fig.33).

In this study GA20 levels were regulated in a different manner compared with the levels of GA53 and GA19. Assuming that biosynthesis is a major factor in the fluctuation in GA20 levels, the results here may imply that different enzymes catalyzed the steps before and after GA19.

There is considerable evidence indicating that the steps catalyzed by GA 20-oxidase are important regulatory steps in GA biosynthesis. In maize seedlings bioactive GAs were shown to regulate their own biosynthesis through feedback control of GA 20-oxidase activity (Hedden and Croker, 1992). In addition, expression of GA 20-oxidase genes in Arabidopsis thaliana was remarkably reduced after the application of GA3 (Phillips et al., 1995). Higher levels of accumulation of GA20 precursors (GA19 and GA53) in non-phyB-1 genotypes (Figs. (Figs.114) may be the result of the low activity of GA 20-oxidase. Furthermore, it may be speculated that the apparent higher enzyme activity of this enzyme in the mutant phyB-1 may partially account for the rapid growth of this genotype. Correlation between an increase in GA20 and a decrease in its precursor, GA19, was fairly well kept in 58M (Figs. (Figs.114). However, there was no clear relationship between these two GAs in 90M and 100M.

The level of GA1 also appears to be controlled by photoperiod. In the non-phyB-1 genotypes (100M and 90M), rhythmic production of GA1 occurs in plants grown under 10-h photoperiods (Fig. (Fig.1),1), which is different from that of plants grown under longer photoperiods (Figs. (Figs.224; see Foster and Morgan [1995] for data on 12-h photoperiods). In photoperiods of 12 h or longer, GA1 levels consistently increased during the early portion of the light period, with the peak concentration occurring 6 h or occasionally 9 h after lights-on. Under a 10-h photoperiod, GA1 levels were highest at lights-on and decreased gradually during the day. In 90M this downward trend started after 3 h rather than at lights-on (Fig. (Fig.1).1). This decreasing pattern is similar to that observed in 58M under 12-h photoperiods, when it flowers early (Foster and Morgan, 1995).

Figure 2
Diurnal regulation of GA levels in shoots of 14-d-old seedlings of 58M (•), 90M (▪), and 100M (▴) grown under 14-h photoperiods. The dark period is indicated by the solid bars at the top and bottom of the figure. GA levels were ...

Under 10-h photoperiods, there are no differences in floral initiation between these three genotypes (Pao and Morgan, 1986). In the present study, 58M, 90M, and 100M initiated floral primordia at d 20, 23, and 22, respectively, under a 10-h photoperiod. In contrast to the behavior of the non-phyB-1 genotypes under 10-h photoperiods, the patterns of GA1 levels in 58M grown under photoperiods of 18 h or more (Figs. (Figs.44 and and5)5) were different from those of plants grown under photoperiods of 16 h or less (Figs. (Figs.113; Foster and Morgan, 1995), with GA1 levels starting low and increasing during the day in the former cases of 18-h or longer photoperiods.

Under photoperiods of 16 h or less, GA1 levels in 58M decreased during the day. In addition, this decreasing pattern of GA1 during the first 3 to 6 h of the light period was sharper in the 10-h photoperiod (Fig. (Fig.1)1) than in the other photoperiods (Figs. (Figs.224). Again, these patterns can also be viewed as a peak level during the light period, at either lights-on or 6 to 9 h later. This decreasing pattern (morning peak) of GA1 in 58M under a 10-h photoperiod eventually changed to an increasing pattern (midday peak) observed under an 18-h photoperiod. In 58M, GA1 levels in all photoperiods followed the pattern of GA20 levels. However, GA1 levels in 100M and 90M grown under a 10-h photoperiod did not follow the GA20 pattern. Absolute levels of GA1 in 90M and 100M were slightly lower than those observed for GA20. The magnitude of the differences in levels of GA1 between peaks and valleys is often small, and the pattern is less distinct at mid-range photoperiods (14 and 16 h) than at those that distinguish differences in genotypes (10 and 18 h).

In some cases, the GA levels exhibited in the first sample (8:00 am, 14 d after seeding) were not reproduced in the sample taken 24 h later (8:00 am, 15 d after seeding). This was most often expressed in low levels of GA20 and GA1 in 58M at the end of the dark period, as shown in Figures Figures115. This was also noted earlier when the harvest period was 36 h, ending at the end of a light period rather than a dark period (i.e. after two daily patterns were exhibited) (Foster and Morgan, 1995). Possible explanations for this inconsistency have been discussed previously (Foster and Morgan, 1995). There are no data to explain this occasional difference; however, the decreased levels may be related to the frequent opening of the growth chamber during sampling, which might briefly change the temperature, or to a change in the light environment as plant density changed because of thinning of the population during successive harvests. In fact, at the first 8:00 am sampling the growth chamber had not been opened for several hours, but at the last sampling it had been opened every 3 h for 24 h.

The patterns of GA levels found in this study, (a) being high at lights-on and declining or low at lights-off, and (b) increasing to a peak during the day, occurred in multiple experiments in harvests at 14 and 19 d in this study (Fig. (Fig.5)5) and at 14 and 25 d in a previous study (Foster and Morgan, 1995), which indicates that the results are not unique to a certain plant age. Thus, the occasional failure of GA levels to track back to a starting level from 24 h previously is concluded to be of less significance than the more dependable occurrence of pulses or peaks in levels of GA12, GA53, GA20, and GA1, which are altered by phyB-1 and by the duration of photoperiod (Figs. (Figs.114; Foster and Morgan, 1995).

The phyB-1 mutant 58M was initially classified as a GA overproducer based on the GA content of sorghum samples collected during the first few hours after lights-on (Beall et al., 1991; Childs, 1993; Foster et al., 1994). The discovery of rhythmic peaks in GA concentrations at different times of the day indicated that this overproducer classification was an oversimplification of the effect of the absence of phytochrome B (Foster and Morgan, 1995). This conclusion is strongly supported by the present study when the data from nine sampling points (two replications) for 1 d were averaged for each photoperiod (Fig. (Fig.6).6). Some of the trends do not represent unquestionable differences, as indicated by the overlap of the sd values. However, if the data for GA12 are set aside, the overall pattern is for the non-phyB-1 genotypes to contain higher levels of early-pathway GAs (GA53 and GA19), whereas the average levels of GA20 are clearly higher in the phyB-1 genotype (58M). Differences in levels of GA1 between 58M and the non-phyB-1 genotypes tend to follow those of GA20, except at 10 h, when average levels in all three genotypes are about the same. These results suggest that the conversion of GA19 to GA20 is a rate-limiting step in sorghum and may be a major control point of phytochrome B.

Both shoot length and dry weight are known to be higher in 58M than in 90M and 100M when grown under 12-h photoperiods (Pao and Morgan, 1986; Beall et al., 1991). However, under long photoperiods (23 or 24 h) shoot elongation is inhibited (compared with 12-h photoperiods) in 90M and 100M but not in 58M (Childs et al., 1995; K.L. Childs and P.W. Morgan, unpublished data). In the present study increasing daylength promoted shoot dry weight until 18 h, when the dry weight of 90M and 100M but not 58M was reduced (Fig. (Fig.7). 7).

Figure 7
Effect of photoperiod on shoot dry weight (DW) of 14-d-old seedlings of 58M, 90M, and 100M. Dry weight represents the shoot between the root crown and tallest leaf collar without the basal three leaves. Error bars show sd.

These inhibitory effects of very long photoperiods on shoot growth of wild-type plants is not explained by average GA1 levels, which do not differ at 18 h from those at 14 or 16 h (Fig. (Fig.6).6). Under shade, shoot growth is favored at the expense of root growth (Smith, 1992), but 58M plants (and other phytochrome B mutants; Childs et al., 1997) grow as if they were continuously in shade. The phyB-1-containing 58M exhibits a higher shoot-to-root ratio of dry weight than the non-phyB-1 genotypes (Lee, 1996; Morgan et al., 1996). The data shown in Figure Figure77 suggest the possibility that extremely long photoperiods may have the opposite effect of shade on partitioning of photosynthate in the wild type.

It has been clearly demonstrated in a number of rosette long-day plants that photoperiodic control of stem elongation is mediated by GAs (Jones and Zeevaart, 1980; Metzger and Zeevaart, 1980b; Gianfagna et al., 1983; Talon et al., 1990, 1991; Zeevaart et al., 1993). The promotive effect of long days on stem growth and floral initiation appears to be the result of enhanced levels of endogenous GAs, especially the GAs of the late steps of the early C-13β-hydroxylation pathway (GA20 and GA1). However, in short-day sorghum the anticipated increase in height with increasing photoperiod length (Childs et al., 1995) did not parallel the average GA1 level. This suggests that part of the height advantage that 58M has over 90M and 100M is the result of increased sensitivity to GAs (Weller et al., 1994; Reed et al., 1996) or to a non-GA-mediated mechanism.

As in 58M, short days shifted the peak GA1 concentration in 100M from midday to near dawn (Fig. (Fig.1).1). As in 90M and 100M, long days shifted GA20 and GA1 peaks in 58M from early morning to midday (Fig. (Fig.4).4). Thus, GA metabolic steps leading to active GAs appear to be controlled differently in the phytochrome B mutant 58M. These results are consistent with the hypothesis that the rhythm of bioactive GA production may play a role in flowering. The pulses of GAs (especially GA1) may have different effects on floral initiation according to the time of day that they occur. Plant hormones, including GAs, are well known to regulate gene expression (Fox and Jacobs, 1987; Jacobsen and Gubler, 1992). Changing the timing of a rhythmic peak in the level of a hormone has the potential to alter its relationship to transcription or translation regulators rhythmically produced by circadian oscillator genes.

The actual differences in levels of GAs in this study at different times of day and night are small. This suggests that the pulses or peaks are not physiologically important. On the other hand, it is possible that the mass of tissue harvested and extracted masks larger differences in GA levels in the apical meristems and young leaves. If large differences in GA pulse levels and timing do occur, then the relations noted in this paper between photoperiod and behavior of GA levels in the early-flowering 58M and the late-flowering 90M and 100M may be physiologically important.


We thank Drs. Fred Miller and Bill Rooney for supplying the sorghum seed needed for these experiments.


1This work was supported in part by U.S. Department of Agriculture competitive grant no. 91-37304-6582 (to P.W.M.), a Korean Government Overseas Scholarship (to I.-J.L.), a postdoctoral fellowship from the Natural Sciences and Engineering Research Council of Canada (to K.R.F.), and the Texas Agricultural Experiment Station.


  • Beall FD, Morgan PW, Mander LN, Miller FR, Babb KH. Genetic regulation of development in Sorghum bicolor. V. The ma3R allele results in gibberellin enrichment. Plant Physiol. 1991;95:116–125. [PMC free article] [PubMed]
  • Campell BR, Bonner BA. Evidence for phytochrome regulation of gibberellin A20 3β-hydroxylation in shoots of dwarf (lele) Pisum sativum L. Plant Physiol. 1986;82:909–915. [PMC free article] [PubMed]
  • Childs KL (1993) The effect of the ma3R mutation on phytochrome, gibberellin metabolism and photoperiodism in Sorghum bicolor (L.) Moench. PhD thesis, Texas A&M University, College Station
  • Childs KL, Cordonnier-Pratt MM, Pratt LH, Morgan PW. Genetic regulation of development in Sorghum bicolor. VII. ma3R flowering mutant lacks a phytochrome that predominates in green tissue. Plant Physiol. 1992;99:765–770. [PMC free article] [PubMed]
  • Childs KL, Lu J-L, Mullet JE, Morgan PW. Genetic regulation of development in Sorghum bicolor. X. Greatly attenuated photoperiod sensitivity in a phytochrome-deficient sorghum possessing a biological clock but lacking a R-HIR. Plant Physiol. 1995;108:345–351. [PMC free article] [PubMed]
  • Childs KL, Miller FR, Cordonnier-Pratt MM, Pratt LH, Morgan PW, Mullet JE. The Sorghum bicolor photoperiod sensitive gene, Ma3, encodes a phytochrome B. Plant Physiol. 1997;113:611–619. [PMC free article] [PubMed]
  • Childs KL, Pratt LH, Morgan PW. Genetic regulation of development in Sorghum bicolor. VI. The ma3R allele results in abnormal phytochrome physiology. Plant Physiol. 1991;97:714–719. [PMC free article] [PubMed]
  • Davies PJ, Birnberg PR, Maki SL, Brenner ML. Photoperiod modification of [14C]gibberellin A12 aldehyde metabolism in shoots of pea line G2. Plant Physiol. 1986;81:991–996. [PMC free article] [PubMed]
  • Foster KR, Lee IJ, Pharis RP, Morgan PW. Effect of ring D-modified gibberellin on gibberellin levels and development in selected Sorghum bicolor maturity genotypes. Plant Growth Regul. 1997;16:79–87.
  • Foster KR, Miller FR, Childs KL, Morgan PW. Genetic regulation of development in Sorghum bicolor. VIII. Shoot growth, tillering, flowering, gibberellin biosynthesis, and phytochrome levels are differentially affected by dosage of the ma3R allele. Plant Physiol. 1994;105:941–948. [PMC free article] [PubMed]
  • Foster KR, Morgan PW. Genetic regulation of development in Sorghum bicolor. IX. The ma3R allele disrupts diurnal control of gibberellin biosynthesis. Plant Physiol. 1995;108:337–343. [PMC free article] [PubMed]
  • Fox JE, Jacobs M (1987) Molecular Biology of Plant Growth Control. Alan R Liss, New York
  • Gianfagna T, Zeevaart JAD, Lusk WL. Effect of photoperiod on the metabolism of deuterium labeled gibberellin A53 in spinach. Plant Physiol. 1983;72:86–89. [PMC free article] [PubMed]
  • Gilmour SJ, Zeevaart JAD, Schwenen L, Graebe JE. Gibberellin metabolism in cell-free extracts from spinach leaves in relation to photoperiod. Plant Physiol. 1986;82:190–195. [PMC free article] [PubMed]
  • Graebe JE. Gibberellin biosynthesis and control. Annu Rev Plant Physiol. 1987;38:419–465.
  • Hedden P, Croker SJ. Regulation of gibberellin biosynthesis. In: Karssen CM, Van Loon LC, Vreugdenhil D, editors. Progress in Plant Growth Regulation. Dordrecht, The Netherlands: Kluwer Academic Publishers; 1992. pp. 534–544.
  • Ievinsh G, Kreicbergs O. Endogenous rhythmicity of ethylene production in growing intact cereal seedling. Plant Physiol. 1992;100:1389–1391. [PMC free article] [PubMed]
  • Jacobsen JV, Gubler F. Hormonal control of gene expression. In: Karssen CM, Van Loon LC, Vreugdenhil D, editors. Progress in Plant Growth Regulation. Dordrecht, The Netherlands: Kluwer Academic Publishers; 1992. pp. 116–127.
  • Jones MG, Zeevaart JAD. Gibberellins and photoperiodic control of stem elongation in the long-day plant Agrostemma githago L. Planta. 1980;149:269–273. [PubMed]
  • Lane HC. Effect of light quality on maturity in the milo group of sorghum. Crop Sci. 1963;3:496–499.
  • Lee IJ (1996) Photoperiodic control of gibberellin metabolism, flowering and circadian ethylene production in a phytochrome B mutant of Sorghum bicolor. PhD thesis, Texas A&M University, College Station
  • Lipe JA, Morgan PW. Ethylene, a regulator of young fruit abscission. Plant Physiol. 1973;51:949–953. [PMC free article] [PubMed]
  • Machácková I, Chauvaux N, Dewitte W, Van Onckelen H. Dirunal fluctuations in ethylene formation in Chenopodium rubrum. Plant Physiol. 1997;113:981–985. [PMC free article] [PubMed]
  • Metzger JD, Zeevaart JAD. Identification of six endogenous gibberellins in spinach shoots. Plant Physiol. 1980a;65:623–626. [PMC free article] [PubMed]
  • Metzger JD, Zeevaart JAD. Effect of photoperiod on the levels of endogenous gibberellins in spinach as measured by combined gas chromatography-selected ion current monitoring. Plant Physiol. 1980b;66:844–846. [PMC free article] [PubMed]
  • Morgan PW. Regulation of flowering in sorghum. Flowering Newsletter. 1994;17:3–11.
  • Morgan PW, Finlayson SA, Lee IJ, Childs KL, He CJ, Creelman RA, Drew MC, Mullet JE. Regulation of circadianly rhythmic ethylene production by phytochrome B in sorghum. In: Kanellis AK, editor. Biology and Biotechnology of the Plant Hormone Ethylene. Dordrecht, The Netherlands: Kluwer Academic Publishers; 1996. pp. 105–111.
  • Morgan PW, Guy LW, Pao CI. Genetic regulation of development in Sorghum bicolor. III. Asynchrony of thermoperiods with photoperiods promotes floral initiation. Plant Physiol. 1987;83:448–450. [PMC free article] [PubMed]
  • Morgan PW, He CJ, De Greef JA, De Proft MP. Does water deficit stress promote ethylene synthesis by intact plants? Plant Physiol. 1990;94:1616–1624. [PMC free article] [PubMed]
  • Pao CI, Morgan PW. Genetic regulation of development in Sorghum bicolor. I. Role of the maturity genes. Plant Physiol. 1986;82:575–580. [PMC free article] [PubMed]
  • Phillips AL, Ward DA, Uknes S, Appleford NEJ, Lange T, Huttly AK, Gaskin P, Graebe JE, Hedden P. Isolation and expression of three gibberellin 20-oxidase cDNA clones from Arabidopsis. Plant Physiol. 1995;108:1049–1057. [PMC free article] [PubMed]
  • Phinney BO. Gibberellin A1, dwarfism and the control of shoot elongation in higher plants. In: Crozier A, Hillman JR, editors. The Biosynthesis and Metabolism of Plant Hormones. Society Experimental Biology Seminar Series, Vol 23. Oxford, UK: Oxford University Press; 1984. pp. 17–41.
  • Quail PH, Briggs WR, Chory J, Hangarter RP, Harberd NP, Kendrick RE, Koornneef M, Parks B, Sharrock RA, Schäfer E. and others. Spotlight on phytochrome nomenclature. Plant Cell. 1994;6:468–469. [PMC free article] [PubMed]
  • Quinby JR (1973) The genetic control of flowering and growth in sorghum. In NC Brady, ed, Advances in Agronomy, Vol 25. Academic Press, New York, pp 125–162
  • Reed JW, Foster KR, Morgan PW, Chory J. Phytochrome B affects responsiveness to gibberellins in Arabidopsis. Plant Physiol. 1996;112:337–342. [PMC free article] [PubMed]
  • Rikin A, Chalutz E, Anderson JD. Rhythmicity in ethylene production in cotton seedlings. Plant Physiol. 1984;75:493–495. [PMC free article] [PubMed]
  • Rood SB, Larsen KM, Mander LN, Abe H, Pharis RP. Identification of endogenous gibberellins from sorghum. Plant Physiol. 1986;82:330–332. [PMC free article] [PubMed]
  • Smith H. The ecological functions of the phytochrome family: clues to a transgenic programme of crop improvement. Photochem Photobiol. 1992;56:816–822.
  • Talon M, Koornneef M, Zeevaart JAD. Endogenous gibberellins in Arabidopsis thaliana and possible steps blocked in the biosynthetic pathways of the semidwarf ga4 and ga5 mutants. Proc Natl Acad Sci USA. 1990;87:7983–7987. [PMC free article] [PubMed]
  • Talon M, Zeevaart JAD, Gage DA. Identification of gibberellins in spinach and effects of light and darkness on their levels. Plant Physiol. 1991;97:1521–1526. [PMC free article] [PubMed]
  • Weller JL, Ross JJ, Reid JB. Gibberellins and phytochrome regulation of stem elongation in pea. Planta. 1994;192:489–496.
  • Williams EA, Morgan PW. Floral initiation in sorghum hastened by gibberellic acid and far-red light. Planta. 1977;145:269–272. [PubMed]
  • Zeevaart JAD, Gage DA. ent-Kaurene biosynthesis is enhanced by long photoperiod in the long-day plant Spinacia oleracea L. and Agrostemma githago L. Plant Physiol. 1993;101:25–29. [PMC free article] [PubMed]
  • Zeevaart JAD, Gage DA, Talon M. Gibberellin A1 is required for stem elongation in spinach. Proc Natl Acad Sci USA. 1993;90:7401–7405. [PMC free article] [PubMed]

Articles from Plant Physiology are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...