Cell, Vol. 93, 593–603, May 15, 1998, Copyright ©1998 by Cell Press
The indeterminate Gene Encodes a Zinc Finger
Protein and Regulates a Leaf-Generated Signal
Required for the Transition to Flowering in Maize
signal the time of transition (Lang, 1965; Bernier, 1988;
McDaniel et al., 1996). Most plants integrate both environmental
and developmental signals to elicit flowering.
In Arabidopsis, at least 12 late-flowering mutants have
been described, showing that many genes function to
Joseph Colasanti,*†§ Zhuang Yuan,†
and Venkatesan Sundaresan†‡
*University of California, Berkeley
Plant Gene Expression Center
Albany, California 94710
†Cold Spring Harbor Laboratory augment the transition to reproductive growth in this
species (Koornneef et al., 1991; Martinez-Zapater et al.,Cold Spring Harbor, New York 11724
‡Institute of Molecular Agrobiology 1994), and three of these genes have been cloned and
characterized (Lee et al., 1994; Putterill et al., 1995;National University of Singapore
Singapore 117604 Macknight et al., 1997). Genetic analysis has supported
a model for two pathways of floral induction in Arabidopsis;
one pathway is constitutive, dependent on developmental
signals, and the other pathway relies onSummary
environmental signals to induce flowering (MartinezZapater
et al., 1994; Coupland, 1995). A combination ofFlowering in plants is a consequence of the transition
grafting and genetic experiments with pea have identi-of the shoot apex from vegetative to reproductive
fied genes that act at the shoot apex to make it compe-growth in response to environmental and internal sigtent
to receive the floral stimulus and other genes thatnals. The indeterminate1 gene (id1) controls the transiact
outside the apex to produce graft-transmissible sig-tion to flowering in maize. We show by cloning the id1
nals that either induce or inhibit the transition to flow-gene that it encodes a protein with zinc finger motifs,
ering (reviewed in Weller et al., 1997). In contrast, muchsuggesting that the id1 gene product functions as a
less is known about genes controlling this process intranscriptional regulator of the floral transition. id1
monocotyledonous plants, including the agronomicallymRNA expression studies and analyses of transpoimportant
cereals such as maize.son-induced chimeric plants indicate that id1 acts
Maize grown in temperate climates is a vegetativelynon-cell-autonomously to regulate the production of
determinate plant that makes the transition to floweringa transmissible signal in the leaf that elicits the transafter
initiating a particular number of leaves. indetermi-formation of the shoot apex to reproductive developnate
(id1) is the only mutation known to specifically andment. These results provide molecular and genetic
severely alter the ability of maize to undergo the transi-data consistent with the florigen hypothesis derived
from classical plant physiology studies. tion to reproductive growth (Singleton, 1946). Homozygous
id1 mutants develop normally at firstbut eventually
Introduction produce many more leaves than their wild-type siblings
(Figure 1A). As with late flowering mutants from other
The shoot apical meristem of higher plants embodies a species, id1 mutants are unable to undergo a normal
population of undifferentiated cells that gives rise to both transition to flower development and remain in a provegetative
and reproductive structures. Leaves and axil- longed state of vegetative growth (Galinat and Naylor,
lary meristems emerge from the flanks of the shoot apex 1951). But in contrast to the well-characterized lateduring
vegetative growth, and inflorescences and flow- flowering mutants in Arabidopsis and pea, severe id1
ers are formed during reproductive growth. The point mutants produce aberrant inflorescences with vegetaof
transition, when the shoot apex begins to produce tive characteristics (Figures 1B and 1C). Determinate
reproductive structures instead of vegetative structures, flowers that form on the flanks of the id1 inflorescence
is a critical developmental process in flowering plants. revert to vegetative growth and whole plantlets emerge
In some plants such as maize, the shoot apical meristem from each spikelet (Figures 1C and 1D). Axillary meriundergoes
a transformation during the transition marked stems of id1 mutants either fail to form a female infloresby
a rapid increase in cell division and restructuring of cence or they are converted into branch-like vegetative
the apex. The end results are the formation of the male structures (not shown). These mutant phenotypes deminflorescence,
or tassel, from which determinate flowers onstrate that the id1 gene has an important role in conemerge,
and the cessation of vegetative growth. After trolling the transition to reproductive development as
the transition, one or two axillary meristems develop well as in maintaining the florally determined state.
into female inflorescences, or ears, which also bear de- In this study we describe the molecular characterizaterminate
flowers. tion of the id1 gene and analyze its expression pattern.
Plants have evolved intricate schemes to coordinate The deduced ID1 protein has two zinc finger motifs that
the transition to flowering with optimal environmental implicate id1 as a regulator of genes that mediate the
conditions and developmental states. Some plants are transition to flowering. Analysis of plants with transpocompletely
dependent upon environmental signals to son excisions that generate somatic chimeras provides
evoke flowering, whereas others rely on internal devel- evidence that id1 acts non-cell-autonomously; i.e., id1
opmental cues perhaps correlated with plant size to can act outside the shoot apical meristem to mediate
the transition. Further, expression studies show that id1
is expressed in immature leaves and is not detectable§To whom correspondence should be addressed.
Cell
594
Figure 1. Comparison of Wild-Type Maize and an id1-m1/id1-m1
Mutant
(A) Plants were grown at the same time in the greenhouse. The wildtype
plant on the left (Id1/Id1) has 13 leaves, a normal tassel, and
Figure 2. Isolation of an id1 Genomic Fragment by Transposon
two ears; its id1-m1/id1-m1 mutant sibling (right) has 20 visible
Tagging
leaves and shows no signs of flowering.
(A) Illustration of tagging scheme using a Ds2 transposon inserted(B) Inflorescence of a severe id1-m1 mutant that produced 32 leaves
within the nearby bz2-m allele (indicated by spotted kernel). Selec-and a “tassel” with plantlets emerging from every spikelet.
tion for germinal excisions (solid kernel) that restored Bz2 function(C) Comparison of tassel branches from normal and id1-m1 mutant
identified an F2 family that segregated the id1-m1 allele caused byplants. The normal branch (left) shows anthers emerging from some
Ds2 insertion.flowers (florets). The mutant tassel branch (right) has produced flo(B)
Southern blot of SacI-digested DNA from a family of plantsrets showing initial stages of proliferation of plantlets from within
segregating the newly identified id1 mutation (id1 ϭ id1-m1). Theeach spikelet.
upper blot was hybridized with a Ds2-specific probe; the lower panel(D) Spikelet from an id1-m1 tassel showing an emerging plantlet.
shows the same blot hybridized with a genomic probe derived fromArrowheads indicate each glume of the spikelet. Scale bar ϭ 1 cm.
Ds2-flanking sequence. The 4.2 kb and 2.9 kb bands are indicated.
(C) Schematic showing relative positions of the probes within the
4.2 kb SacI fragment. Triangle indicates the Ds2 transposon.
at the shoot apex. Thus, id1 appears to mediate the
transition to reproductive development by regulating the
locus containing the wild-type Id1 allele (Figure 2A). Outsynthesis or transmission of a long-distance signal from
of 600 independent transposition events, one family seg-the leaves to the shoot apex.
regated plants with characteristics of indeterminate
plants; i.e., they continued to make leaves long after
normal siblings had flowered and died. The new mutantResults
was designated id1-m1 after it was found to be allelic to
the first id1 mutation to be described, id1-R (Singleton,Identification of a Transposon Insertion
into the id1 Gene 1946), and to another allele, id1-Compeigne.
Southern analysis of several hundred plants revealedThe id1 gene is located on chromosome 1 near the Bz2
gene that conditions kernel coloration; wild-type Bz2 ker- a 4.2 kb SacI Ds2-hybridizing band that consistently
cosegregated with the id1-m1 allele and was absent innels are purple, whereas mutant bz2 kernels are bronze
colored. Transpositions of a Ds2 transposon from the plants that did not contain this allele (Figure 2B). The
cloned 4.2 kb fragment was found to have a 1.3 kbbz2-m allele were used for localized mutagenesis of the
Cloning and Analysis of the Maize id1 Gene
595
Ds2 element inserted 165 bp from one of the SacI sites
(Figure 2C). Reprobing the blots with a genomic fragment
flanking the insertion showed the expected pattern
for Ds2 insertion; i.e., a 4.2 kb band in id1-m1 homozygotes,
a 2.9 kb band in plants homozygous for the wildtype
allele, and both bands in heterozygotes (Figure
2B, lower panel). Southern analysis of other id1 alleles
revealedthat id1-R/id1-Rplants had no hybridizingband
and that the id1-Compeigne allele was associated with
an insertion of approximately 3 kb in this region (J.C. and
V. S., unpublished data). Therefore, the id1-R mutation is
caused by a deletion of all or part of the id1 gene, and
the id1-Compeigne allele is likely the result of an undetermined
transposon insertion.
Determination of the id1 Coding Region
and Gene Structure
Sequence analysis indicated an open reading frame
(ORF) near the Ds2 insertion site with similarities to
Kru¨ppel-like zinc finger proteins from several species
(not shown). Three different cDNA libraries made from
3-leaf seedlings, vegetative apices, and immature ears,
respectively, were screened with an ORF probe. However,
while cDNAs that encoded zinc finger proteins
closely related to id1 could be isolated, none of them
corresponded to the id1 gene itself (not shown). This
suggests that id1 mRNA may be present only at very
low levels or that it is not represented in the libraries
thatwere screened.SubsequentNorthern blotting showed
Figure 3. Characterization of the id1 Gene Structure, Deduced Proid1
expression only in immature leaf tissue (see below). tein Sequence, and Comparison to id-like Genes
The complete id1 cDNA sequence was derived by RT–
(A) Intron/exon structure of the id1 gene. Open boxes represent
PCR reconstruction of the id1 transcribed region using exons and thin lines show intron position with size indicated below.
information based on the id1 genomic sequence and on Filled double circles show positions of the two zinc finger motifs
within the second and third exons. Position of the Ds2 element nearsequences of cDNA obtained from immature leaves (see
the third exon/intron junction is indicated.Experimental Procedures). The intron/exon structure
(B) Deduced amino acid sequence of the ID1 protein. The two zinc
and deduced amino acid sequence of the id1 gene are
finger sequences are underlined; conserved cysteine (C) and histishown
in Figures 3A and 3B, respectively. dine (H) residues are in bold and conserved hydrophobic residues
are double underlined. Position of a potential SV40-like nuclear localization
signal is boxed (Raikhel, 1992). A filled triangle shows the
The id1 Gene Is a Member of a Zinc
relative position of the Ds2 element within the id1 coding sequence.
Finger Gene Family (C) Comparison of the id1 zinc fingers and intervening region with
The sequence of the id1 transcript reveals a second that of an id-like gene from maize, “p1”, and to PCP1 from potato
zinc finger motif upstream of the first zinc finger motif (Kuhn and Frommer, 1995).
identified near the Ds2 insertion site at the 3Ј exon/intron
junction. Comparison of id1 to the id-like genes isolated
from screening cDNA libraries showed a region of exten- similar level of homology to the zinc finger motifs above
(data not shown). The similarity between the productssive similarity that included both zinc finger motifs and
the intervening sequence (Figure 3C). Southern analysis of id1, the id-like genes from maize, and the genes from
potato and Arabidopsis extends beyond the zinc fingerwith a conserved region probe indicates that there may
be as many as ten id-like genes in the maize genome region by approximately 30 amino acids on either side
of the sequence shown in Figure 3C. The overall identity(J. C. and V. S., unpublished data).
More recent database searches picked up another of these genes in this 160 amino acid region is approximately
70% (data not shown). No sequence could begene from potato encoding a protein with striking similarity
in the zinc finger region to the id1 and id-like gene classed as being more similar to any other in the conserved
region, although the id1 gene is unique in thatproducts (Figure 3C). This protein, PCP1, is a putative
RNA/DNA-binding protein that was isolated by its abil- the region between the zinc fingers has an additional
24 amino acids (Figure 3C).ity to complement a sucrose transport deficiency in
Saccharomyces cerevisiae (Kuhn and Frommer, 1995).
Searchesof Arabidopsis expressedsequence tags (EST) Effect of Ac on the id1-m1 Phenotype
The Ds2 transposon in theid1 gene isexpected to exciseand genomic databases revealed an EST (Genbank accession
number T04539) and a region of a BAC clone in the presence of an Ac element, giving rise to revertant
tissues in the plant as well as derivative alleles that arefrom chromosome IV (R. Martienssen and R. McCombie,
personal communication) that encode proteins with a transmitted through the germ line. Differences in the
Cell
596
Figure 4. Effects of Ac on Flowering of id1m1
Mutants and Generation of Derivative Al-
leles
(A) Flowering time of wild-type (left) and id1m1/id1-m1
mutant plants (right) with Ac
(hatched boxes) and with no Ac (open boxes).
Each box indicates the range of flowering
time denoted as time of initial pollen shedding.
The number of plants in each group is
shown in parentheses. Thedashed line designates
the end of the growing season. Plants
that did not flower during the growing season
are shown above the dashed line.
(B) Effect of Ac dosage on flowering time of
id1-m1 mutants. All plants were grown in a
greenhouse fromkernels generated bycrosses
between id1-m1 homozygous mutants with
one Ac element. Segregation of Ac results in
kernels with three distinct spotting patterns
that are diagrammed at the bottom of the
graph; the number of Ac elements in the corresponding
plant grown from each type of
kernel is shown in parentheses. Plants that
made normal flowers are shown by open circles,
and plants that made proliferous tassels
(with plantlets emerging from spikelets) are
shown by filled circles. Shaded circles represent
three plants in which Ds2 had excised
and left 3 bp footprints (see below). The revertant
allele Id1-Rev is indicated by the circle
with a dot.
(C) Footprint sequence for eight types of germinal
derivative alleles generated by excision
of Ds2 from the id1-m1 locus. The 8 bp target
site for Ds2 insertion into the wild-type
gene is underlined. The inverted “V” indicates
the exon/intron splice junction. Nucleotides
changed by Ds2 excision are shown in bold.
The phenotype of each mutant is shown on
the right. Normal plants (double plus) made
12–14 leaves and wild-type inflorescences;
severe id1 mutants (double minus) made
more than 25 leaves and often produced proliferous
tassels. Moderate id1 mutant phenotypes
(minus) were observed with the derivative
alleles id1-X3 and id1-X31, which made
19 or 20 leaves and normal tassels.
id1-m1 phenotype related to the absence or presence tics. When the season ended 25 weeks after planting,
id1-m1(ϪAc) plants had produced over 25 leaves andof Ac could give clues about how the id1 gene functions
to mediate the transition to reproductive growth. For showed no signs of flowering. At this time, all but three
of the id1-m1(ϩAc) plants had made tassels that shedexample, it may be possible to determine whether or
not id1 acts cell-autonomously from analysis of sectored pollen (Figure 4A). Some of these plants produced ears
that could be crossed with pollen from other id1-m1plants generated by Ds2 excisions. In order to monitor
Ac activity in sectored plants, we constructed a recom- (ϩAc) mutants. Most of the ears, however, exhibited extensive
vegetative characteristics that precluded crossing.binant in which the id1-m1 allele was linked to the bz2-m
marker allele. In id1-m1 bz2-m recombinants, Ac activity We conclude that id1-m1 plants (with three exceptions,
see below) had a significantly attenuated pheno-is indicated by the mutability of bz2-m as visible sectors
of purple pigmentation in the kernel. In early stages of type in response to the presence of Ac. This partial
restoration of wild-type characteristics is consistentgrowth, mutant plants with Ac, i.e., id1-m1(ϩAc), were
indistinguishable from id1 mutants with no Ac. Both with excisions of Ds2 in response to Ac, resulting in
reversion of id1-m1 to a functional Id1 gene. However,classes of mutant plants continued to produce leaves
11 weeks after planting, whereas normal siblings, re- we never observed plants that displayed identifiable
revertant sectors from id1 mutant phenotype to wildgardless of the presence of Ac, made between 11 and
13 visible leaves before producing a tassel inflorescence type. Rather, attenuation of the id1 mutation by Ac resulted
inan overall reduction inthe severity of the mutantat this stage (Figure 4A). At 15 weeks after planting,
some of the id1-m1(ϩAc) plants produced normal tas- phenotype. No significant differences were observed
between the expression of the mutant phenotype in thesels, and a few initiated ears withvegetative characteris-
Cloning and Analysis of the Maize id1 Gene
597
tassel and in the ear from any one id1-m1(ϩAc) plant, identical footprints (id1-X3) and the other had a different
sequence (id1-X31), but the overall result was the addi-even though the two inflorescences arise from different
populations of cells in the shoot meristem (McDaniel tion of a single serine residue to the putative coding
sequence. All three of these plants made fewer leavesand Poethig, 1988).
As described above, several id1-m1(ϩAc) mutants than the most severe id1-m1 mutants, and they did not
show reversion to vegetative growth in the tasselsflowered early enough and made flowers that were sufficiently
normal, so a few mutant siblings could be (shaded circles in Figure 4B). Loss of function due to
an extra serine residue 12 amino acids downstream ofcrossed to each other. These plants produced ears with
viable kernels that segregated spotted and bronze ker- the second zinc finger motif suggests that even minor
changes in this region of the ID1 protein have substantialnels 3:1, indicating that each parent had a single Ac
element. The Ac-Ds transposon family exhibits a nega- effects on gene function.
Finally, while all the derivatives exhibit mutant pheno-tive dosage phenomenon that is unique to maize; i.e.,
transposons excise laterin development with increasing types, one derivative allele was completely normal; i.e.,
it made 12 leaves and produced fully normal flowers.copies of Ac, resulting in smaller somatic sectors (McClintock,
1949). The spotting pattern on bz2-m kernels was Southern blots showed that this plant was heterozygous
for the id1-m1 locus. Sequence analysis revealed thatused as an indicator of the number of Ac elements present
in the progeny of each kernel. Kernels with large Ds2 had excised precisely, completely restoring the
functional Id1 gene structure (Figure 4C). This plant ispurple sectors have a single Ac element, whereas kernels
with very few, tiny spots have two or three Ac a true germinal revertant and not a contaminant, because
its parents had been crossed after all normalelements present in the endosperm and generally two
elements in the embryo. The flowering times of mutant plants had completely shed their pollen. The possibility
that the revertant phenotype is due to large somaticplants with different doses of Ac are compared in the
experiment shown in Figure 4B. Mutants with a single revertant sectors can be eliminated because an Ac element
was not present. Although excision of Ds from aAc produced significantly fewer leaves and flowered
earlier than mutants with no Ac. More importantly, id1- locus more often causesa change in the DNA sequence,
precise excisions of Ds elements from various genesm1 mutants with two Ac elements made almost as many
leaves as plants with no Ac and many exhibited rever- have been reported previously (Scott et al., 1996; Goodrich
et al., 1997).sion to vegetative growth. Plants with no Ac had the
most severe mutant phenotype, producing more leaves
than siblings with one or two Ac elements and also id1 Is Expressed Early in Development in Immature
Leaf Tissueexhibiting floral reversion in nearly all tassels (Figure
4B). This suggests that Ac-mediated attenuation of the The expression pattern of id1 in different parts of the
plant and at different times of development was exam-id1-m1 mutant phenotype exhibits a negative dosage
effect with respect to Ac copy number. ined by Northern hybridization with an id1-specific
probe. PolyAϩ
RNA was prepared from plants early in
development (3 days postgermination) to late in devel-Derivative Alleles of id1-m1 Have Altered
Flowering Times opment (after flower formation) and from various parts
of the plant. Temporally, id1 mRNA was detected inIn the previous experiment, three plants did not show
attenuation of the id1-m1 phenotype by Ac (Figure 4A). vegetative shoots as early as 3 days after germination
(Figure 5A, longer exposure), and its expression in-These three plants were siblings that came from spotted
kernels, indicating the presence of an Ac element, but creased as plants approached the floral transition stage.
Expression of id1 was also detected later in develop-they resembled plants with no Ac since they did not
flower. Sequencing revealed that these plants were new ment in plants that had undergone the floral transition
as well as in plants that had formed flowers (Figurederivative alleles caused by Ds2 excision from the id1m1
locus, which left behind a 5 bp duplication resulting 5C, lanes I and LL). Therefore, id1 mRNA is present
throughout the postembryonic life of the sporophyticin a potential frame shift in the third exon of id1 (Figure
4C). Such a Ds2 footprint is expected to generate a plant.
Although id1 expression is not specific to a particularstable null allele, which would explain the nonresponsiveness
to the effects of Ac. time of development, we found that id1 mRNA is present
only in particular structures (specifically, immature leavesOther alleles derived by Ds2 excision from id1-m1
were found by examining progeny of id1-m1 plants in [Figures 5A and 5B]). Analysis of different parts of 3-leaf
and 5-leaf seedlings (i.e., before the floral transition)which the Ac had been segregated away. Absence of
Ac was important in this experiment to ensure that the showed that id1 is not expressed in leaf blades but is
expressed within shoots that contain a whorl of imma-observed phenotypes were due to germinal excision
events and not large somatic revertant sectors. The ture leaves and the shoot apical meristem (Figure 5A).
We determined by dissection of standard B73 inbredflowering time for some of these plants is shown in
Figure 4B. Several derivatives had either 5 bp or 7 bp seedlings that the shoot apical meristem of plants with
eight visible leaves is at, or near, a stage of transitioninsertions that would result in a frameshift in the Id1
ORF (Figure 4C). The derivative id1-X10 had a 10 bp to flowering. The stems of 8-leaf plants were dissected
into five equal 2 cm pieces from which mRNA was ex-deletion that removed the splice donor site, which probably
causes a disruption of the normal transcript and tracted and expression of id1 analyzed (Figure 5B).
Northern analysis showed that id1 is expressed at rela-results in a severe phenotype. Three independent excision
alleles with 3 bp insertions were isolated; two had tively high levels in the whorl of unexpanded leaves
Cell
598
above the shoot apex, but it is not detectable in the part
of the shoot that contained the apex nor is it expressed
in the internode stem below the apex. Expression of id1
mRNA therefore is confined to immature leaves and not
the shoot apex.
Northern analysis with posttransition plants (ten or
more visible leaves) reveals that id1 continues to be
expressed in immature leaves and remains detectable,
although at lower levels, in the last immature leaf that
surrounds the tassel before it emerges from the whorl
of leaves (Figure 5C, lane LL). No id1 mRNA was detected
in mature leaves or in roots or floral tissues at
any stage of plant development. Conversely, the id-like
gene “p1” is expressed in both mature and immature
leaves as well as in the apical region, but it is absent in
root and floral tissues (Figures 5B and 5C).
In situ hybridization experiments provided further evidence
that id1 is expressed in immature leaf tissue and
not at the apex. Transverse sections through the stems
of B73 inbred seedlings with eight visible leaves showed
id1-specific mRNA in the inner, immature leaf of the
whorl and not in the outer leaf (Figures 6A and 6B).
Expression was not detectable at the shoot apical meristem
and the bases of surrounding leaves (Figure 6C),
although expression of the Knotted (Kn) gene was confined
to the shoot apical meristem (Figure 6D) as reported
previously (Jackson et al., 1994). Longitudinal
sections through the shoot apical meristem similarly
showed no id1 expression (unpublished data). As expected,
id1 mRNA was not detected by in situ hybridization
in any tissue sections of id1-R mutant plants (data
not shown). Although id1 expression appears to be
greater in the outer epidermal layers of the inner leaves,
there is also detectable staining in the leaf mesophyll
cells, which might beless intense because of large vacuoles
(Figure 6B).
In contrast, expression analysis of the Arabidopsis LD
gene by Northern hybridization (Lee et al., 1994) and CO
by RT–PCR analysis (Putterill et al., 1995) showed that
these genes are expressed in all aerial parts of the plant,
and in situ hybridization experiments showed that CO
expression is localized to the vegetative shoot apical
Figure 5. Northern Analysis of id1 Gene Expression meristem and several leaf primordia (Simon et al., 1996).
(A) Expression of id1 in shoots of 3- and 7-day-old plants and in FCA transcripts were detected in roots as well as shoots
entire 3-leaf seedling shoots (all) or from leaves 1 and 2 only (1 ϩ (Macknight et al., 1997). Therefore, the expression pat-
2). Plants with five visible leaves were dissected into leaves 1 ϩ 2, tern of id1 differs from previously characterized flowleaf
3, leaf 4, andleaf 5 ϩ shoot (including apex). Blotswere reprobed
ering time genes in that id1 expression is restricted tosequentially with the maize cdc2ZmA gene and then with hcf106,
immature leaves and is absent from the shoot apex.a gene required for chloroplast biogenesis (Martienssen et al., 1989),
and a cDNA specific for the maize id-like gene “p1”.
(B) Expression of id1 in different portions of a B73 inbred seedling Discussion
at the transition to flowering stage (eight visible leaves). The first
four leaves, includingsheath, were removed andthe remaining shoot The deduced amino acid sequence of the ID1 protein
was partitioned into five equal 2 cm portions as illustrated (left).
shows two zinc finger motifs with the pattern HH-CC
Segment A contains the shoot apical meristem (asterisk); sba refers
and HH-HC, respectively. Since zinc finger proteins areto the stem below the apex, which contains no leaf tissue.
associated with DNA/RNA binding activity and the par-(C) Expression of id1 in a plant that has initiated all of its leaves and
ticular type of zinc finger in ID1 has some similarity tohas begun floral development (ten visible leaves) is shown in the
left four lanes. The two right lanes are mRNA samples from a mature a class of animal transcription factors, we suggest that
plant with a fully formed tassel that has not yet emerged. M, mature ID1 could act as a regulatory protein that controls the
leaf derived from fully expanded leaves; I, immature leaf; A, apex
transcription of genes required for the transition to flowtransformed
into tassel primordium; R, root; IT, immature tassel; LL,
ering. The existence of related genes in maize with ex-the last leaf enclosing the immature tassel. The id1-specific probe
tensive similarities in the zinc finger and surroundinghybridized to a 1.6 kb band, the approximate size predicted from
regions suggests the possibility of some functional re-the cDNA. A minor band of 1.9 kb that consistently hybridizes with
the id1 probe may be an unspliced precursor RNA. dundancy with id1. However, the one id-like transcript
Cloning and Analysis of the Maize id1 Gene
599
regulatory function, the id1 gene product has no significant
similarity to these three Arabidopsis proteins. id1
is the only flowering transition gene isolated from a
monocot, and there is presently no evidence for or
against the existence of orthologs of this gene in dicots.
The id1 mutation causes other defects in reproductive
development besides delayed flowering, i.e., floral aberrations
caused by reversion to vegetative growth that
have not been reported for flowering time mutants described
in dicots. We note that a floral reversion phenotype
is associated with mutation of floral meristem identity
genes such as floricaula of Antirrhinum and LEAFY
of Arabidopsis (Coen et al., 1990; Weigel et al., 1992).
In these cases, loss of gene function causes a transformation
of flowers into inflorescence shoots. In addition,
under conditions of noninductive short-day photoperiods,
plants heterozygous for a leafy null mutation show
a high rate of floral reversion, with shoots emerging from
within flowers (Okamuro et al., 1996). In some species
such as Arabidopsis, meristem identity genes such as
LEAFY may play an important role in establishing as
well as maintaining floral meristem identity (Bla´ zquez et
al., 1997). Therefore, there likelyare differences between
monocots and dicots in the genetic regulatory networks
for the induction of flowering and for the maintenance
of the florally differentiated state.
Evidence for Non-Cell-Autonomous Action
of the id1 Gene
The overall attenuation of the id1-m1 phenotype by Ac
together with the absence of visible phenotypically revertant
sectors is characteristic of non-cell-autonomous
gene action. An alternative explanation would be that
the presence of Ac partiallyrestores Id1 function in everyFigure 6. In Situ Hybridization in Seedlings with Eight Visible Leaves
cell carrying the id1-m1 mutation. The phenotypes of(A) Transverse section through the stem 5 cm above the apex with
mutations generated by transposon insertions can de-id1 anti-sense probe. Two concentric leaves are shown wrapped
within the whorl; id1 staining is evident in the inner leaf whorl. Scale pend upon the presence of regulatory transposons
bar ϭ 300 ␮m. (McClintock, 1955). In the Ac-Ds system, such an effect
(B) Same section as (A). Scale bar ϭ 100 ␮m.
was shown to occur with the Kn gene of maize (Hake
(C) Transverse section through the shoot apical meristem hybridized
et al., 1989). However, we believe that this explanationwith id1 anti-sense probe. Scale bar ϭ 300 ␮m.
is not applicable to the effects of Ac on id1-m1 for the(D) Transverse section through the apex of same plant as (C) hybridfollowing
two reasons. First, attenuation of the id1-m1ized with a Kn-specific anti-sense probe showing staining in the
shoot meristem (Jackson et al., 1994). Scale bar ϭ 300 ␮m. mutation exhibited a negative dosage effect with the Ac
copy number. Plants with one Ac had a less severe id1
phenotype than plants with two Ac elements. Such a
negative dosage effect is typical of excision events inthat we characterized in detail (“p1”) had a markedly
different expression pattern from that of id1 (Figure 5), the Ac-Ds system; i.e., Ds2 transposes from the mutant
locus later with more Ac elements, resulting in smallerso it is likely that some of the id-like genes of maize
regulate developmental processes other than flowering. revertant sectors. By contrast, the Ds2-induced gainof-function
Kn mutation exhibits a positive dosage effectArabidopsis is the only other plant species in which
genes that specifically regulate the transition to flow- with Ac (Hake et al., 1989). Second, for all suppressible
transposon-induced mutations that have been molecu-ering have been isolated. The LUMINIDEPENDENS (LD)
gene encodes a possible homeodomain protein (Lee larly characterized to date, the transposon is inserted
in a noncoding region of the gene, so that transcriptset al., 1994; Aukerman and Amasino, 1996), and the
CONSTANS (CO) gene product contains GATA factor– encoding a functional protein can be generated without
transposon excision. In id1-m1, the Ds2 element is in-like zincfingermotifs (Putterillet al., 1995). More recently
the FCA gene product was reported to have similarities serted into a coding exon of the id1 gene near the zinc
finger domain at a location where even minor alterationsto RNA-binding proteins, suggesting a potential role in
controlling gene expression at the level of RNA pro- of the amino acid sequence severely affect function.
Therefore, it is likely that attenuation of the id1-m1 phe-cessing (Macknight et al., 1997). All three of these genes
have a postulated role in controlling the transition to notype is due to excisions of Ds2 rather than suppression
by Ac.flowering. Although id1 is similar in that it likely has a
Cell
600
Another relevant observation is that even though the analysis of theid1 gene providesevidence that is consisflowers
of many Ac-containing id1-m1 plants appeared tent with the florigen model. We find that id1 is exnormal,
presumably due to reversion by Ds2 excision, pressed in specific vegetative structures such as immathe
gametophytes within these flowers always carried ture leaves, even though its effect is manifested at the
the mutant allele. Generation of a revertant sector that shoot apex. In addition, genetic evidence based on
completely restores normal tassel formation would re- transposon-mediated effects on the mutant phenotype
quire an early excision event that covers the whole shoot suggest that the action of id1 is non-cell-autonomous.
apical meristem including both epidermal (L1) and sub- We propose that revertant somatic sectors in immature
epidermal (L2) layers of the developing inflorescence leaves act as sites of ID1 production. Functional ID1
(Dawe and Freeling, 1990). Therefore, pollen derived protein in the leaf then mediates the synthesis of a transfrom
a revertant tassel would necessarily be revertant missible substance(s) that migrates to the apex to signal
as well, resulting inwild-type progeny inthe next genera- the transition to flowering. It is also possible that the
tion. Such was not the case; with one exception, the id1 gene product itself migrates from itssite of synthesis
progeny of id1-m1(ϩAc) mutants were all mutants. The in the leaves to the apex. Intercellular migration was
exception was a germinal revertant that resulted from reported for the KNOTTED homeobox protein (Lucas et
a precise excision of the Ds2 element. Since the siblings al., 1995). However, we consider this possibility to be
of this revertant were still mutant, we assume that rever- unlikely, since KNOTTED migration is limited to a few
sion occurred late in gametophyte development and cell layers and cannot account for the type of longthat
it did not affect tassel or ear initiation. These results distance signaling proposed for id1.
are most simply explained by proposing that the mode Our results suggest a model for the regulation of flowof
action of theid1 geneproduct is non-cell-autonomous ering time by id1. Maize plants generally flower after
and that excision events outside the apex can restore making a fixed number of leaves, indicating that flownormal
tassel formation. ering is initiated by an endogenous signal that is depenThe
earlier flowering times observed with one Ac ele- dent upon leaf number. Detection of id1 mRNA in 3-dayment
as opposed to two Ac elements can now be inter- old seedlings suggests that ID1 protein is expressed in
preted as follows. We anticipate that more functional ID1 leaf tissue throughout shoot development. In addition,
would be generated by the frequent excisions expected studies with id1-m1 plants with different Ac doses indiwith
one Ac element than by the infrequent excisions cate that flowering of id1-m1 plants is accelerated by
with two Ac elements. If a critical level of ID1 is required increasing levels of functional id1 gene product. We
for signaling the transition of the apex, this level will suggest that the amount of ID1 in vegetative tissues
be reached earlier in id1-m1 plants that have frequent must increase to a critical level to signal the transition
excisions of the Ds2 element. Similarly, the CO gene and that this critical level is reached after a particular
of Arabidopsis displays a dose-dependent induction of number of leaves are generated. If id1 regulates the
flowering, since extra copies of CO or ectopic expres- synthesis of a diffusible factor, as we propose, then the
sion promotes earlier flowering (Putterill et al., 1995; level of ID1 in leaves would reflect the concentration
Simon et al., 1996). It should be noted that whereas CO of diffusible factor that is received by the apex, and a
is semidominant, id1 is completely recessive, and there threshold level for id1 product may reflect the requireappears
to be no difference in flowering time between ment for a critical concentration of this factor for the
plants with one or two functional id1 genes (Singleton,
initiation of reproductive growth. The idea of a critical
1946), perhaps due to dosage compensation.
or threshold level for id1 is consistent with observations
Since plants with no Ac do ultimately flower, although
from studies with excised maize shoot apices showing
imperfectly, the id1 gene may be partially redundant
that the meristem relies on signals from other parts of
with a second gene or floral-induction pathway. It is
the plant to determine the extent of vegetative growth
possible that one of the id-like genes is partially able to
(Irish and Nelson, 1988, 1991; Irish and Jegla, 1997).
substitute for id1 function. We note that mutations other
Similar experiments intobacco demonstrated that a critthan
id1 that severely affect the flowering transition in
ical leaf number is required for shoot apices to become
maize have not been isolated, suggesting that redundetermined
for reproductive development (Singer and
dant genes may be detectable only in a loss-of-function
McDaniel, 1986).id1 mutant background.
The observation that id1 continues to be expressed
late in development after flowers are formed suggests
an additional role in maintaining the florally determinedModel of id1 Gene Action: Support
state. Consistent with this observation, even severe id1for the Florigen Model
mutants eventually undergo a transition in which theNumerous studies of the last century have provided
shoot apex is converted to an inflorescence-like struc-insights into the physiology of the floral transition. One
ture, but the floral meristems that form are capable ofof the most significant conclusions from these studies
further vegetative growth as evidenced by the emer-is that the flowering transition is triggered by a signal
gence of plantlets containing shoots and roots fromthat originates inthe leaves(Bernier, 1988; O’Neill, 1992).
within the spikelets. In some plants such as ImpatiensThis led to the proposal that a diffusible substance,
and wheat, an inflorescence or flower will revert to vege-sometimes referred to as “florigen,” is made in leaves in
tative development when the floral stimulus is removedresponse to environmental and/or developmental cues
(Fisher, 1972; Battey and Lyndon, 1990; Pouteau et al.,and is diffused or transported to the apex where it triggers
the transition to reproductive development. Our 1997).
Cloning and Analysis of the Maize id1 Gene
601
maize cDNA libraries. Approximately 2 ϫ 106
plaqueswere screenedConclusions
from a 3-leaf seedling library (made by Alice Barkan, University ofWe have reported the isolation and characterization of
Oregon, Eugene, OR), and 1 ϫ 106
plaques each were screened
the id1 gene from maize. id1 is the first gene isolated
from an immature ear library and a library made with mRNA from
from a monocot that has an important role in signaling 3- to 4-week-old vegetative apices (from Bruce Veit, Plant Gene
the transition from vegetative to reproductive growth, Expression Center, Albany, CA). Phage clones that hybridized to
the probe were isolated and plasmids with inserts were excised byand it has some distinctive features as compared to
an in vivo protocol recommended by the manufacturer (Stratagene).genes isolated from dicots that regulate the same process.
The action of id1 is non-cell-autonomous, and a
Isolation of a Genomic Clone and Determination
critical level of id1 may be required for the floral transiof
the id1 Transcribed Region
tion. id1 could regulate the floral transition either by A genomic clone containing the entireid1 coding region was isolated
acting at the shoot apex to make it competent to receive from a B73 DNA library (from David Jackson, Plant Gene Expression
the floral stimulus or outside the apex to regulate the Center, Albany, CA) by screening with a 1.0 kb probe derived from
the 1.4 kb intron of id1 (see Figure 3). One recombinant phage, ␭1a,production or transmission of a floral signal that is sent
was isolated from approximately 1 ϫ 106
plaques screened. Twoto the apex. Since id1 is expressed in immature leaves
overlapping fragments from the genomic insert of ␭1a were subbut
not at the shoot apex, we suggest that id1 acts
cloned into pLITMUS29. Plasmid pZY12 contained a 4.0 kb BamHI
through the second mechanism. Together with genetic fragment that included most of the id1 coding region, and pZY15
evidence that suggests id1 acts in a non-cell-autono- contained a 1.3 kb SacI fragment that overlapped by 110 bp with the
mous manner, we propose that id1 directly or indirectly 5Ј end of the BamHI fragment. RT–PCR identified the id1 transcribed
region. Total RNA was isolated from immature leaves of 3-week-regulates the synthesis of a florigenic substance or subold
plants by grinding in liquid nitrogen and extraction with Trizolstances that is transmitted to the apex, consistent with
reagent (Bethesda Research Laboratories). PolyAϩ
RNA was isophysiological
studies of the last century.
lated by Oligotex spin elution (Qiagen). Reverse transcriptase (SuperscriptII)
was used to synthesizecDNA from 1 ␮g of purified mRNA
Experimental Procedures
using the manufacturer’s protocol (Bethesda Research Laboratories).
Oligonucleotides spanning the genomic region were used
Maize Stocks and Genetic Analysis
as primers in PCR reactions with cDNA as template. Amplification
Maize bz2-m stocks were a gift from Kelly Dawe (University of Georproducts
that consistently reamplified with nested primers and that
gia). Other maize stocks were obtained from the Maize Genetics
were smaller than products amplified from genomic DNA templates
Stock Center (Urbana, IL) and Virginia Walbot (Stanford University).
were cloned into pLITMUS29 and sequenced. Comparison of ampliid1-R
and id1-Compeigne alleles were obtained from Benjamin Burr
fied sequences to genomic sequences revealed the intron positions.
(Brookhaven National Laboratory, Upton, NY). Inbred B73 seed was
a gift of Zuo-Yu Zhao (Pioneer Hi-Bred International). The strategy
Northern Blot Analysis
used to obtain a transposon-tagged allele of id1 is reported in detail
PolyAϩ
RNA was isolated from tissues using the method described
elsewhere (Colasanti and Sundaresan, 1992). The effect of Ac dosabove,
and 1 ␮g of each sample was electrophoresed on 1.1%
age on the id1-m1 phenotype was observed by recording the flowformaldehyde
agarose gels and transferred to nylon membranes
ering time of id1 mutants with zero, one, or two copies of Ac based
(Genescreen). A 650 bp fragment of genomic DNA specific to the id1
on the kernel spotting pattern. Kernels derived from crosses of
transcribed region was used to probe Northern blots using standard
two id1-m1 homozygous plants were classified as having one Ac
conditions (Sambrook et al., 1989). A 900 bp DNA fragment unique
element (large spots), two Ac elements (tiny spots), or no Ac eleto
the id-likegene “p1” was amplifiedfrom a cDNAclone andused as
ments (no spots). These were grown in greenhouses so that a comprobe.
The cdc2ZmA probe was prepared as described previously
plete life cycle could be observed. The number of leaves made by
(Colasanti et al., 1991), and the hcf106 probe was made from a 1.1
wild-type and id1 mutant plants was determined by counting from
kb fragment of cDNA clone (provided by Mark Settles, Cold Spring
the first node visible above the ground. Plants were grown outdoors
Harbor Laboratory).
and in greenhouses at Uplands Farm Agricultural Field Station, Cold
Spring Harbor, New York.
In Situ Hybridization
Plant tissues were prepared from maize B73 or id1 mutant seedlingsSouthern Blot Analysis and Isolation of Genomic Subclones
with eight visible leaves and fixed and hybridized by the method ofDNA was extracted from maize leaves as described by Chen and
Jackson (1991) except that FAA (3.5% formaldehyde, 10% aceticDellaporta (1994), digestedwith restriction enzymes, and transferred
acid, 50% ethanol) was used as fixative. Fixed tissue was dehy-to nitrocellulose membranes according to standard protocols (Samdrated,
embedded in Paraplast, sectioned at 10 ␮m thickness, andbrook et al., 1989). Probes were 32
P-labeled using a random priming
attached to ProbeOnPlus slides (Fisher Biotech). Two overlappingDNA labeling kit (Boehringer-Mannheim), and hybridization was perregions
of id1-specific sequence were cloned into transcription vec-formed using standard 50% formamide hybridization buffer with
tor pSPT18 (Boehringer-Mannheim) to create plasmids p390 and10% dextran sulfate. A 109 bp Ds2-specific probe (from Sarah Hake
p400 (with 390 bp and 400 bp inserts, respectively). RiboprobesPlant Gene Expression Center, Albany, CA) was used to identify
were prepared following directions provided by the manufacturercosegregating Ds2-hybridizing bands. A 4.2 kb SacI fragment con(Boehringer-Mannheim).
Anti-sense id1 riboprobe was preparedtaining a portion of id1 gene was isolated from a subgenomic library
from p390 by linearizing with HindIII and in vitro transcription withconstructed from 100␮g of SacI-digested genomic DNAfrom an id1T7
RNA polymerase in the presence of 11-digoxigenin-dUTP; anti-m1 mutant.Purified DNA was ligated with SacI-digested pLITMUS29
sense riboprobe from p400 was prepared by linearizing with BamHI(New England Biolabs) and transformed into DH10B cells by electroand
transcription with SP6 RNA polymerase. Riboprobes wereporation. Approximately 60,000 transformants were plated on nitromixed
and used at a total concentration of 2 ng/␮l/kb for each slide.cellulose membranes and screened with a labeled Ds2 probe. One
A Kn anti-sense riboprobe was provided by the Hake Laboratory.recombinant plasmid containing the 4.2 kb SacI fragment was isoImmunological
digoxigen–nucleic acid detection followed recom-lated (Figure 2C). The id1 genomic region was sequenced by primer
mended protocols (Boehringer Mannheim) except that antibodywalking; i.e., oligonucleotides complementary to every 200–300 bp
conjugate was diluted 1:1250.of genomic DNA were synthesized and used in sequencing reac-
tions.
Acknowledgments
Screening cDNA Libraries for id1 and id-like Sequences
We thank Rob Martienssen for suggestions and advice concerningA 165 bp Ds2-flanking genomic DNA fragment with a putative open
reading frame (ORF) was labeled and used to screen three B73 the mutagenesis experiments, Kelly Dawe and Virginia Walbot for
Cell
602
providing seeds, and Janet Ross for help with the initial planting. predicts patterns of morphogenesis in the vegetative shoot. Development
120, 405–413.Cathy Andorfer, ScottColasanti, Jill Nemachek, Ed Chan, andLouise
McNitt contributed excellent technical assistance, and Tim Mulligan Koornneef, M., Hanhart, C.J., and van den Veen, J.H. (1991). A geprovided
expert plant care. We are grateful to Lauren Hubbard, David netic and physiological analysis of late-flowering mutants in ArabiJackson,
Debbie Laudencia-Chingcuanco, Paula McSteen, and dopsis thaliana. Mol. Gen. Genet. 229, 57–66.
Leonore Reiser for help with in situ hybridization experiments and
Kuhn, C., and Frommer, W. (1995). A novel zinc finger protein enSarah
Hake and Leonore Reiser for comments on the manuscript.
coded by a couch potato homologue from Solanum tuberosum
Thanks to Kristen Briggs, Benjamin Burr, Mary Byrne, Paul Chomet,
enables a sucrose transport-deficient yeast strain to grow on suErin
Irish, Yukiko Mizukami, and Michael Muszynski for helpful discrose.
Mol. Gen. Genet. 247, 759–763.
cussions. This research was supported by grants from the National
Lang, A. (1965). Physiology of flower initiation, differentiation and
Science Foundation (MCB-9630622) to J. C. and from the United
development. In Encyclopedia of Plant Physiology, W. Ruhland, ed.
States Department of Agriculture (94–37304–1004) and the National
(Berlin: Springer), pp. 1380–1536.
Science Foundation (MCB-9208646) to V. S.
Lee, I., Aukerman, M.J., Gore, S.L., Lohman, K.N., Michaels, S.D.,
Weaver, L.M., John, M.C., Feldman, K.A., and Amasino, R.M. (1994).
Received February 9, 1998; revised March 31, 1998. Isolation of LUMINIDEPENDENS: a gene involved in the control of
flowering time in Arabidopsis. Plant Cell 6, 75–83.
Lucas, W.J., Bouche-Pillon, S., Jackson, D.P., Nguyen, L., Baker,
References
L., Ding, B., and Hake, S. (1995). Selective trafficking of KNOTTED1
homeodomain protein and its mRNA through plasmodesmata. SciAukerman,
M.J., and Amasino, R.M. (1996). Molecular genetic analyence
270, 1980–1983.
sis of flowering time in Arabidopsis. Semin. Cell Dev. Biol. 7,
Macknight, R., Bancroft, I., Page, T., Lister, C., Schmidt, R., Love,427–433.
K., Westphal, L., Murphy, G., Sherson, S., Cobbett, C., and Dean,
Battey, N.H., and Lyndon, R.F. (1990). Reversion of flowering. Bot.
C. (1997). FCA, a gene controlling flowering time in Arabidopsis,
Rev. 56, 162–189.
encodes a protein containing RNA-binding domains. Cell 89,
Bernier, G. (1988). The control of floral evocation and morphogene- 737–745.
sis. Ann. Rev. Plant Physiol. Plant Mol. Biol. 39, 175–219.
Martienssen, R.A., Barkan, A., Freeling, M., and Taylor, W. (1989).
Bla´ zquez, M.A., Soowal, L.N., Lee, I., and Weigel, D. (1997). LEAFY Molecular cloning of a maize gene involved in photosynthetic memexpression
and flower initiation in Arabidopsis. Development 124, brane organization that is regulated by Robertson’s Mutator. EMBO
3835–3844. J. 8, 1633–1639.
Chen, J., and Dellaporta, S. (1994). Urea-based plant DNA miniprep. Martinez-Zapater, J.M., Coupland, G., Dean, C., and Koornneef, M.
In The Maize Handbook, M. Freeling and V. Walbot, eds. (New York: (1994). The transition to flowering in Arabidopsis. In Arabidopsis,
Springer-Verlag), pp. 526–538. E.M. Meyerowitz and C.R. Somerville, eds. (Cold Spring Harbor, NY:
Cold Spring Harbor Laboratory Press), pp. 403–434.Coen, E.S., Romero, J.M., Doyle, S., Elliot, R., Murphy, G., and Carpenter,
R. (1990). FLORICAULA: a homeotic gene required for flower McClintock, B. (1949). Mutable loci in maize. Carnegie Inst. Wash.
Year Book 48, 142–154.development in Antirrhinum majus. Cell 63, 1311–1322.
McClintock, B. (1955). Controlled mutation in maize. Carnegie Inst.Colasanti, J., and Sundaresan, V. (1992). Isolation of new alleles of
Wash. Year Book 54, 242–255.anther ear and indeterminate. Maize Genet. Coop. Newslett. 66,
30–31. McDaniel, C.N., and Poethig, R.S. (1988). Cell-lineage patterns in
the shoot apical meristem of the germinating maize embryo. PlantaColasanti, J., Tyers, M., and Sundaresan, V. (1991). Isolation and
175, 13–22.characterization of cDNA clones encoding a functional p34cdc2
homologue from Zea mays. Proc. Natl. Acad. Sci. USA 88, 3377– McDaniel, C.N., Hartnett, L.K., and Sangrey, K.A. (1996). Regulation
3381. of node number in day-neutral Nicotiana tabacum: a factor in plant
size. Plant J. 9, 55–61.Coupland, G. (1995). Genetic and environmental control of flowering
time in Arabidopsis. Trends Genet. 11, 393–397. O’Neill, S.D. (1992). The photoperiodiccontrol of flowering: progress
toward understanding the mechanism of induction. Photochem.Dawe, R.K., and Freeling, M. (1990). Clonal analysis of the cell linPhotobiol.
56, 789–801.eage in the male flower of maize. Dev. Biol. 142, 233–245.
Okamuro, J.K., den Boer, B.G.W., Lotys-Prass, C., Szeto, W., andFisher, J.E. (1972). The transformation of stamens to ovaries and of
Jofuku, K.D. (1996). Flowers into shoots:photo andhormonal controlovaries to inflorescences in Triticum aestivum L. under short-day
of a meristem identity switch in Arabidopsis. Proc. Natl. Acad. Sci.treatment. Bot. Gaz. 133, 78–85.
USA 93, 13831–13836.
Galinat, W.C., and Naylor, A.W. (1951). Relation of photoperiod to
Pouteau, S., Nicholls, D., Tooke, F., Coen, E., and Battey, N. (1997).inflorescence proliferation in Zea mays L. Am. J. Bot. 38, 38–47.
The induction and maintenance of flowering in Impatiens. DevelopGoodrich,
J., Puangsomlee, P., Martin, M., Long, D., Meyerowitz,
ment 124, 3343–3351.
E.M., and Coupland, G. (1997). A Polycomb-group gene regulates
Putterill, J., Robson, F., Lee, K., Simon, R., and Coupland, G. (1995).
homeotic gene expression in Arabidopsis. Nature 386, 44–51.
The CONSTANS gene of Arabidopsis promotes flowering and enHake,
S., Vollbrecht, E., and Freeling, M. (1989). Cloning Knotted, codes a protein showing similarities to zinc finger transcription facthe
dominant morphological mutant inmaize using Ds2 asa transpo- tors. Cell 80, 847–857.
son tag. EMBO J. 8, 15–22.
Raikhel, N. (1992). Nuclear targeting in plants. Plant Physiol. 100,
Irish, E.E., and Nelson, T.M. (1988). Development of maize plants 1627–1632.
from cultured shoot apices. Planta 175, 9–12.
Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular ClonIrish,
E.E., and Nelson, T.M. (1991). Identification of multiple stages ing: A Laboratory Manual (Cold Spring Harbor, New York: Cold
in the conversion of maize meristems from vegetative to floral devel- Spring Harbor Press).
opment. Development 112, 891–898. Scott, L., LaFoe, D., and Weil, C.F. (1996). Adjacent sequences influIrish,
E., and Jegla, D. (1997). Regulation of extent of vegetative ence DNA repair accompanying transposon excision in maize. Gedevelopment
of the maize shoot meristem. Plant J. 11, 63–71. netics 142, 237–246.
Jackson, D. (1991). In situ hybridization in plants. In Molecular Plant Simon, R., Igeno, M.I., and Coupland, G. (1996). Activation of floral
Pathology: A Practical Approach, D.J. Bowles, S.J. Gurr, and M. meristem identity genes in Arabidopsis. Nature 384, 59–62.
McPherson, eds. (Oxford: Oxford Univ. Press), pp. 163–174. Singer, S.R., and McDaniel, C.N. (1986). Floral determination in the
Jackson, D., Veit, B., and Hake, S. (1994). Expression of maize terminal and axillary buds of Nicotiana tabacum L. Devel. Biol. 118,
587–592.KNOTTED1 related homeobox genes in the shoot apical meristem
Cloning and Analysis of the Maize id1 Gene
603
Singleton, W.R. (1946). Inheritance of indeterminate growth in maize.
J. Hered. 37, 61–64.
Weigel, D., Alvarez,J., Smyth, D.R., Yanofsky, M.F., and Meyerowitz,
E.M. (1992). LEAFY controls floral meristem identity in Arabidopsis.
Cell 69, 843–859.
Weller, J.L., Reid, J.B., Taylor, S.A., and Murfet, I.C. (1997). The
genetic control of flowering in pea. Trends Plant Sci. 2, 412–418.
GenBank Accession Number
The accession number for the id1 gene reported in this paper is
AF058757.