Genetic studies, particularly with Pisum, in combination with grafting experiments have identified genes which appear to affect sensitivity to daylength through the synthesis of inhibitors (Murfet, 1985) and possibly also promoters (Murfet, 1987). If the major growth substances were involved in the balance of promoters and inhibitors controlling flowering, it might be expected that mutations affecting production or sensitivity to them would affect daylength-dependent flowering. Once again, Arabidopsis is the photoperiodic species in which a range of relevant (i.e. growth substance) mutations has been identified. The fact that the major growth substances are involved in multiple aspects of development makes it difficult to separate out any phenotype which shows modified growth from specific alterations in the daylength response system. Gibberellins have a significant influence on flowering in Arabidopsis. Exogenous GA partially overcomes the delaying effect of SD in WT (Langridge, 1957). A series of gibberellin biosynthesis mutants, such as gal, and sensitivity mutants, such as gai (Koornneef et al., 1985), both result in a similar, dwarfed phenotype in which flowering is slightly delayed in LD but considerably delayed or prevented in SD (Wilson et al., 1992; Coupland et al., 1993). The gal-3 allele exhibits normal flowering in continuous light but does not flower in SD unless it is treated with exogenous gibberellin. This suggests the operation of two independent pathways to flowering in Arabidopsis. A gibberellin-dependent pathway which operates in SD, i.e. the autonomous flowering pathway, and a LD-dependent pathway which does not depend on gibberellins (Wilson et al., 1992). Double mutants of co and gai flower much later than either parental line and in some cases not at all. This suggests that CO and GAI can partially compensate for each other, perhaps because they operate in the two pathways leading to flowering; CO in the LD-dependent pathway and GAI in the gibberellin-dependent autonomous pathway. This is not certain, however, because the double mutant also showed a more extreme phenotype than gai with respect to dwarfing and aberrant flower development, which could suggest that they operate within a common pathway.
The amp-1 (altered meristem programme) mutant is pleiotrophic, showing a phenotype altered in three different aspects of plant development; spatial pattern of development, constitutive photomorphogenetic development in darkness and an altered flowering phase. The amp-1 mutants have about six times as much cytokinin as WT plants. Chaudhury et al. (1993) describe the mutant as being precocious for flowering, consistent with cytokinin being a floral stimulus in Arabidopsis. This is based on the observation that it took about 16 d before flower buds were visible in WT plants in LD, but they were visible after about 12 d in the mutant. On the other hand, the number of rosette leaves in the amp-1 mutant at flowering was about 20 as compared with about 7 for the WT plants. Based on the criterion of leaf number, amp-1 is actually a late-flowering rather than an early-flowering mutant. The effect of cytokinin therefore appears to be largely on leaf production and it is difficult to reconcile this with a specific role for cytokinin in floral promotion in Arabidopsis. Mutants affecting the ethylene system in Arabidopsis have been reported in which flowering time in LD is delayed (Bleeker et al., 1988; Guzman and Ecker, 1990) but there is no evidence that photoperiodic responses as such are altered.
Was this article helpful?