Within the Arabidopsis inflorescence, two distinct developmental phase
s exist. The early inflorescence phase is characterized by nodes beari
ng coflorescences and leaves, and the late inflorescence phase by node
s bearing flowers. Four genes, TERMINAL FLOWER 1, LEAFY, APETALA1 and
APETALA2 are necessary to initiate the switch from formation of early
to formation of late inflorescence nodes at the appropriate time. We h
ave investigated the relative roles of these genes in development by i
solating and characterizing new alleles of TERMINAL FLOWER 1, LEAFY an
d APETALA1, and by constructing double mutants to test gene interactio
ns. We suggest that the TERMINAL FLOWER 1 gene product is part of a me
chanism that controls the timing of phase-switching in Arabidopsis. We
propose that this mechanism involves factor(s) whose activity changes
in response to shoot development and environmental variation. TERMINA
L FLOWER 1 influences phase transitions in Arabidopsis, and appears to
regulate the timing of expression of LEAFY, APETALA1 and APETALA2. LE
AFY, APETALA1 and APETALA2 have partially redundant functions in initi
ating the floral program. In the absence of any one of the three genes
, there is a gradual transition from coflorescence to flower-like late
ral shoots. This suggests that (1) LEAFY, APETALA1 and APETALA2 are re
quired in combination to ensure that the floral program is initiated r
apidly and completely and (2) in the absence of one of the three genes
, the others are activated slowly in response to the mechanism control
ling timing of phase switching. Besides their role in establishing the
floral program; phenotypes of flower-like lateral shoots in mutant in
florescences suggest that ail three, LEAFY, APETALA1 and APETALA2, inf
luence expression of whorl identity genes. Loss of LEAFY results in de
creased Class B gene expression, as well as altered expression pattern
s of Class A and Class C genes. In the absence of either APETALA2 or A
PETALA1, reproductive organs develop in the perianth whorls, suggestin
g that both genes should be considered Class A organ identity genes, r
estricting Class C gene expression to inner whorls.