Development of the Inflorescences

Erik Vollbrecht and Robert J. Schmidt

Abstract The maize tassel and ear are complex structures that arise from a shared
program of development involving a number of meristem identities and fates, yet
emerge as distinct inflorescences that bear exclusively male and female flowers,
respectively. Careful phenotypic and genetic studies of mutants that perturb meristem
initiation, size, determinacy and identity or pathways of organogenesis such as sex
determination, are elaborating the intricacies of these developmental programs by
providing important insights into the underlying genes and gene interactions. Our
understanding at the molecular level includes information from expression patterns
of cloned genes coupled with knowledge of the gene products they encode and
interact with, coalescing into molecular bases for the mechanisms underlying the
formation of these remarkable structures. These studies show that some mechanisms
are at least partially conserved with those elucidated in other model systems such
as Arabidopsis, but many others are unique to the grasses. Permeating advances in
our understanding of the maize inflorescence is a rich collection of mutants, some
identified at the turn of the last century and new ones recorded in databases of
productive EMS- and transposon-based screens, constituting a key genetic resource
that will provide fertile ground for maize geneticists and developmental biologists
for years to come.

1 Introduction

The transition from vegetative growth to flowering sets in motion one of the most
dramatic developmental processes to occur during the life of a plant. This process
in maize is additionally dramatic in that it produces two morphologically distinct
inflorescences that bear separate male and female flowers. Many other grasses
produce perfect flowers in a single inflorescence, yet the maize inflorescence
remains an important and excellent model for the grasses, including other grain
crops. The purpose of this review is to highlight some of the features that have
made the study of maize inflorescence development productive. These attributes
include meristem fates and identities that are unique to grass species, advancing
molecular tools and a wealth of informative mutants that affect those meristem

J.L. Bennetzen and S.C. Hake (eds.), Handbook of Maize: Its Biology, 13
DOI: 10.1007/978-0-387-79418-1_2, © Springer Science + Business Media, LLC 2009
14 E. Vollbrecht and R.J. Schmidt

activities. Many of the underlying genes are now cloned, providing an emerging
molecular framework for the developmental pathways regulating inflorescence
architecture. In addition, maize manifests remarkable natural diversity in the tassel
and ear among different inbreds. These features, and of course, the economic
importance of inflorescence structures and the intense artificial selection that has
influenced their morphology, all make the study of inflorescence development in
maize an extremely exciting pursuit.

2 Features of the Mature Inflorescence

Modern maize produces two distinct mature inflorescences commonly referred to

as the tassel and ear (Fig. 1a, b). The tassel bears staminate flowers and is borne
at the apex of the mature plant, whereas the ear bears pistillate flowers. Ears are
produced at the apex of a compressed branch originating in the axil of one or more
leaves (Fig. 1b) and terminate all lateral branches, with the exception of branches
in the tassel and a few vegetative basal branches called tillers (Kiesselbach, 1949).
One obvious difference in morphology between the two inflorescences is the pres-
ence of a variable number of long branches originating at the base of the typical
tassel. Ears lack these long branches. To produce flowers, or analogous floral
shoots called florets in grasses, both inflorescences first make a pair of short
branches called spikelets. The spikelet is the basic unit of grass inflorescence
architecture and is characterized by an outer and inner glume (sterile bracts or
modified leaves) that together enclose a variable number of florets. In maize, there
are two florets (Fig. 1a). Each floret includes a floral shoot that originates from the
axil of an additional bract called a lemma. The floret consists of a single lemma
and palea, two lodicules, three stamens and a central pistil consisting of three
fused carpels, surrounding a single ovule. During floral development, subsequent
to initiating all floral organ types, the developing pistil aborts in tassel florets, and
the developing stamens undergo a similar fate in ear florets. Additionally, the
entire lower floret of each ear spikelet aborts, again soon after initiating floral
organ primordia (reviewed in Dellaporta and Calderon-Urrea, 1994; Irish, 1996).
The end result is a monoecious plant with distinct male and female inflorescences
(Fig. 1a, b).
One of the intriguing aspects of maize inflorescence development is the wide
range of natural variation that exists among different inbred lines, no doubt
reflecting the great depth of allelic diversity that has been documented among these
inbreds (Liu et al., 2003). For example, tassel branch number varies from an
average of 3 to nearly 20, and the length and angle of tassel branches varies
considerably ([Link] Similarly, the size and number of ears per
plant varies considerably. This natural variation is being exploited to uncover the
quantitative trait loci (QTL) underlying a variety of phenotypes (Upadyayula et al.,
2006; Zhao et al., 2006) and through association mapping to identify statistical
Development of the Inflorescences 15

Fig. 1 General morphology of maize inflorescences. (a) a typical mature male inflorescence, or
tassel, with several long branches at the base of the central spike. Long branches and the central
spike produce rows of short branches or paired spikelets (lower inset); the pedicellate spikelet is
atop a pedicel that is considerably longer than that which distinguishes the sessile spikelet. Each
tassel spikelet produces two glumes, an outer and inner glume, that enclose two florets, an upper
and lower floret (upper inset), each consisting of one lemma, one palea, two lodicules and three
stamens. (b) a typical mature female inflorescence, or ear, with husk leaves removed to illustrate
the ear’s position atop several compressed internodes on an axillary branch. The ear is formed at
the tip of the branch and produces rows (black rectangle) of paired spikelets; the distinction
between sessile and pedicellate spikelet is subtle. Each ear spikelet consists of an upper and lower
floret enclosed by a pair of glumes (inset). The lower floret and the stamens in the upper floret,
degenerate. Only the upper floret matures to produce carpels, two of which fuse to form the long
silk that emerges from within the surrounding glumes (inset). Thus each spikelet appears in (b) as
glumes surrounding one pistillate floret. The lemma, palea and lodicules are present but obscured
by the surrounding glumes.

associations between traits and genetic markers. Some 5,000 recombinant inbred
lines have been generated between a core set of 25 diverse inbred lines and the
common elite inbred B73. This resource is generating a wealth of genetic informa-
tion that will be utilized to understand the genetic control underlying inflorescence
diversity and many other traits. Identifying the genes controlling natural variation
is of interest to developmental and evolutionary biologists alike. The importance of
many of these inflorescence traits to seed yield underlies the intense interest by
maize breeders as well.
16 E. Vollbrecht and R.J. Schmidt

3 Features of the Developing Inflorescence

The morphology of a mature structure in an organism is the end product of an array

of contributing developmental processes. In plants, many of those core processes
reflect the activity of meristems, small groups of undifferentiated, self-regenerating
cells (Steeves and Sussex, 1989; Weigel and Jurgens, 2002). Two main meristems,
the shoot apical meristem and the root apical meristem, establish the main axes of
plant growth. Subsequently secondary meristems, formed in the axils of leaf pri-
mordia during both vegetative and reproductive development, may be either quiescent
or active. When active, they are directly responsible for the formation of secondary
axes of growth. In maize, these include lateral branches called tillers, and several
specialized axes producing the tassel and ear. The ear and tassel have distinct mor-
phologies at maturity, yet for much of the early stages of their development they
appear remarkably similar. This similarity is due to sharing many common devel-
opmental processes, manifest as a shared set and arrangement of meristem types.
Normal maize inflorescence development has been described from a variety of
perspectives (Collins, 1919; Bonnett, 1940, 1948; Mangelsdorf, 1945; Cutler and
Cutler, 1948; Kiesselbach, 1949) including electron microscopy (Cheng et al.,
1983). When the plant transitions from vegetative to reproductive development the
terminal SAM becomes an inflorescence meristem (IM), committed to the forma-
tion of the tassel. The inflorescence meristem starts to produce lateral meristems
called branch meristems (BMs), which develop into major branches at the base of
the mature tassel (Fig. 2a, b). Coincident with their proximity to the vegetative
shoot, BMs may be produced in the two-rowed phyllotaxy of the vegetative shoot
(Giulini et al., 2004), or in the multiple rows that characterize the rest of the inflo-
rescence. Regardless, only these first few lateral meristems are indeterminate and
committed to form branches; the subsequent ones are initiated in multiple rows and
acquire a different identity, that of the spikelet-pair meristem (SPM). Tassel
branches form spikelet-pair meristems as well, but in a distichous pattern (Fig. 2b).
Branch meristems are said to have less determinacy due to their capacity for con-
tinued growth of the axis, while spikelet-pair meristems have more determinacy and
thus remain short (McSteen et al., 2000; Vollbrecht et al., 2005). Each spikelet-pair

Fig. 2 (continued) then converts to another SM. Each SM uniquely initiates two glume primordia
(gp, panel c) followed by a lemma primordium in whose axil the lower floral meristem (Lower
FM) forms (e). The upper SM then produces a second lemma and converts into a second FM
(Upper FM, panel e) and the upper floret matures in advance of the lower floret. Each FM initiates
one palea, two lodicule (lod), and three stamen primordia and then differentiates into the ovule
and surrounding carpel wall or gynoecial ridge (f), which becomes the pistil (h). The schematics
of the mature florets of the tassel (g) and ear (h) are oriented approximately in parallel to the
microscope images in the panels above. Floral organs that degenerate as part of the sex determina-
tion program are shown in grey. The upper floret in each spikelet is depicted slightly larger than
the lower floret. Lo, lodicules; st, stamens; ps, pistil. Black dots represent relative position of stem
(rachilla). White scale bars are 200 µM.
Development of the Inflorescences 17

Fig. 2 Meristem identities during inflorescence development. (a) Schematic of the developmental
progression of meristem identities during inflorescence development. (b–f) Scanning electron
micrographs of developing tassel and ear through floral organ initiation. (b) developing tassel
primordium, (c) close up of developing tassel primordium spike, (d) developing ear primordium,
(e) floral meristems forming in spikelets, (f) initiation of floral organ primordia, (g) schematic of
a mature staminate (tassel) spikelet, (h) schematic of a mature pistillate (ear) spikelet. The vegeta-
tive shoot meristem transitions to an inflorescence meristem (IM) that produces on its flanks the
spikelet pair meristems (SPM) (b–d). In the tassel, the IM first produces a number of branch
meristems (BM) that like the remainder of the tassel central spike, then recapitulate the develop-
mental sequence indicated below. Each SPM produces a spikelet meristem (SM) on its flank and
18 E. Vollbrecht and R.J. Schmidt

meristem gives rise to a short branch that bears two spikelet meristems (SMs). The
SM produces a glume primordium, a distinguishing feature of this particular meristem
(Fig. 2c), and then forms two floral meristems (Fig. 2d, e), the upper and lower
floral meristems (UFM and LFM). Each floral meristem subsequently forms the
floral organs: a palea, two lodicules, three stamens and one pistil, all subtended by
the lemma (Fig. 2e, f). Gynoecial development arrests and soon the gynoecium
degenerates, resulting in an imperfect, staminate flower (Figs. 1a and 2g).
Each ear also originates from a meristem at the tip of a shoot, in this case a
lateral shoot in the axil of a leaf (Fig. 1b). The lateral SAM becomes an ear IM
within a few weeks after the floral transition at the tassel. One notable difference
between the ear and tassel is that the ear’s IM, while producing multiple rows of
spikelet-pair meristems, does not form any basal branches. Spikelet initiation and
development follows the same steps as in the tassel until an obvious divergence
during floral development. As discussed earlier, in the ear the lower floret aborts as
do the stamens of the upper floret, leaving one pistillate flower per spikelet with
floral organs that appear less conspicuous then their tassel counterparts (Figs. 1b
and 2h). In summary, four types of axillary meristems with different identity and
fate, branch meristems (BMs), spikelet-pair meristems (SPMs), spikelet meristems
(SMs) and floral meristems (FMs), give rise to structures in the maize inflorescence
(Cheng et al., 1983; Irish, 1997; McSteen et al., 2000).

4 Significance as a Developmental System

Many classical mutants were described in the first half of the last century that spe-
cifically perturb aspects of maize inflorescence development (see Coe et al., 1988).
Some of these altered the program of sex determination, whereas others affected
various aspects of inflorescence morphology or some combination of features.
The dramatic effect that many of these mutations have on inflorescence development
(see examples, Figs. 3 and 4) is a consequence of changes in the activity of the various
meristems that function during inflorescence development, in the differentiation of
organs produced by meristems, or both. Broadly speaking then, these changes may
affect basic properties of meristem initiation, size and/or maintenance, meristem
identity or determinacy, or aspects of sex determination and floral organ specification
(Table 1).
A central quest among maize developmental biologists is to understand mecha-
nisms controlling these processes. How are the sites for new meristem initiation
determined? Through what mechanisms are the identities of particular meristems
established? What are the genes and gene interactions that dictate the fate of each
meristem type? Answering those questions is key to understanding how the archi-
tecture of the maize inflorescence is orchestrated, and likely contains clues regarding
those genes that were selected during grass evolution and the subsequent domes-
tication of cereal crops. There are some 10,000 grass species, a few of which have
been domesticated into our modern cereals. These domestications encompass a
Development of the Inflorescences 19

Fig. 3 Examples of mutant tassel phenotypes. (a) branched silkless, bd1, (b) ramosa1, ra1, (c)
unbranched, ub1, (d) a tasselseed, (e) a tasselless, (f) barren stalk1, ba1

wide variety of inflorescence architectures, many of which resemble some of the

abnormal phenotypes that accompany maize inflorescence mutants. One can
speculate that subtle changes in the timing, levels or location of expression of key
regulatory genes could underlie the architectural differences observed in the inflo-
rescences of different grass species (Kellogg, 2000b). Understanding the genetic and
molecular control of these developmental mechanisms in maize may therefore
20 E. Vollbrecht and R.J. Schmidt

Fig. 4 Examples of mutant ear phenotypes. (a) normal ear (left) and branched silkless1, bd1, (b)
ramosa2, ra2, (c) a fasciated ear, (d) ramosa1, ra1, (e) a barren inflorescence, (f) a silkless ear

provide valuable insights into the evolution of morphological variation. Tantalizing

evidence to this effect has come from the molecular characterization of the ram-
osa1 gene of maize and the examination of its expression in other grass species
(Vollbrecht et al., 2005) where its timing of expression in different species is
consistent with the variation in inflorescence branching among those species and
the deduced role of this gene from genetic and molecular studies of its action in
maize (see below).
Genetic and molecular analyses of maize flower development have already
proven fertile ground for examining conservation in the developmental programs
for angiosperm floral organ specification. Comparative studies of floral organ
Development of the Inflorescences 21

Table 1 Selected mutants in maize that affect the inflorescence

Mutant Meristem Meristem Organ Sex
symbol Map location function identity dev determ Gene product
an1 1.08 Ö Ö ent-kaurene synthase
bde 5.06 Ö Ö Ö
Bif1 8.02 Ö
bif2 1.05 Ö S-T kinase
baf1 9.01 Ö
ba1 3.06 Ö bHLH (TF)
ba2 2.04 Ö
ba3 Unplaced Ö
bd1 7.04 Ö AP2-domain (TF)
Clt1 8.05 Ö
ct2 1.02 Ö Ö
Cg1 3.02 Ö Ö Micro RNA
dlf1 7.06 Ö bZIP (TF)
Dvd1 5.03 Ö
d1 3.02 Ö Ö
d2 9.03 Ö Ö ent-kaurenoic acid
hydroxylase
d5 2.02 Ö Ö
D8 1.10 Ö Ö DELLA (TF)
D9 5.02 Ö Ö DELLA (TF)
eg1 5.07 Ö
fea2 4.05 Ö Ö LRR receptor-like
kinase
fea3 3.04 Ö Ö
Fas1 9.05 Ö Ö
Fbr1 Unplaced Ö
Gn1 2.10 Ö KNOX (TF)
ifa1 1.02 Ö Ö
ids 1.11 Ö AP2-domain (TF)
id1 1.08 Ö Ö Zinc-finger (TF)
kn1 1.10 Ö Ö KNOX (TF)
Lfy1 3.09 Ö
lg1 2.02 Ö Ö SBP-domain (TF)
lg2 3.06 Ö Ö bZIP (TF)
ns1; ns2 2.05; 4.08 Ö WOX (TF)
Pn1 7.05 Ö
Pt1 6.05 Ö Ö
rel1 1.07 Ö
rel2 10.03 Ö
ra1 7.02 Ö C2H2 Zinc-finger
(TF)
ra2 3.04 Ö LOB domain (TF)
ra3 7.04 Ö Ö Trehalose phos-
phatase
rgo1 9.03 Ö
Sdw1 8.04 Ö
sk1 2.03 Ö Ö
si1 6.01 Ö MADS box (TF)
spi1 3.08 Ö YUCCA-like
(continued)
22 E. Vollbrecht and R.J. Schmidt

Table 1 (continued)
Mutant Meristem Meristem Organ Sex
symbol Map location function identity dev determ Gene product
Sos1 4.02 Ö
tls1 1.11 Ö
tru1 3.06 Ö Ö
ts1 2.03 Ö
ts2 1.03 Ö Alcohol dehydrogenase
Ts3 1.09 Ö
ts4 3.05 Ö Ö Micro RNA
Ts5 4.03 Ö
Ts6 1.11 Ö Ö AP2-domain (TF)
tsh1 6.07 Ö Ö
tsh2 Unplaced Ö Ö
tsh3 6L Ö Ö
tsh4 7.03
Tp1 7.03 Ö Ö
Tp2 10.04 Ö Ö
tb1 1.09 Ö Ö TCP (TF)
tga1 4.05 Ö SBP-domain (TF)
te1 3.05 Ö Ö Ö RNA binding
td1 5.03 Ö Ö LRR R-like kinase
Tu1 4.07 Ö
ub1 Unplaced Ö
Vg1 1.07 Ö
wcr 2L Ö Ö
zag1 6.05 Ö Ö MADS box (TF)
zfl1; zfl2 10.06; 2.02 Ö Ö Ö LEAFY class (TF)

Mutant Mutant
symbol Gene name symbol Gene name
an1 anther ear1 rel1 ramosa enhancer locus1
Bif1 barren inflorescence1 rel2 ramosa enhancer locus2
bif2 barren inflorescence2 ra1 ramosa1
baf1 barren stalk ra2 ramosa2
fastigiate1
ba1 barren stalk1 ra3 ramosa3
ba2 barren stalk2 rgo1 reversed germ orientation1
ba3 barren stalk3 Sdw1 semi-dwarf plant1
bde bearded ear sk1 silkless ears1
bd1 branched silkless1 si1 silky1
Clt1 clumped tassel1 spi1 sparse inflorescence1
ct2 compact plant2 Sos1 supressor of sessile spikelets1
Cg1 corngrass1 tsh1 tassel sheath1
dlf1 delayed flowering1 tsh2 tassel sheath2
Dvd1 developmental disaster1 tsh3 tassel sheath3
d1 dwarf1 tsh4 tassel sheath4
d2 dwarf2 tls1 tasselless1
d5 dwarf5 tru1 tassels replace upper-ears1
D8 dwarf8 ts1 tasselseed1
(continued)
Development of the Inflorescences 23

Table 1 (continued)
Mutant Mutant
symbol Gene name symbol Gene name
D9 dwarf9 ts2 tasselseed2
eg1 expanded glumes1 Ts3 tasselseed3
fea2 fasciated ear2 ts4 tasselseed4
fea3 fasciated ear3 Ts5 tasselseed5
Fas1 fascicled ear1 Ts6 tasselseed6 (ids1)
Fbr1 few-branched1 Tp1 teopod1
Gn1 gnarley1 Tp2 teopod2
ifa1 indeterminate floral apex1 tb1 teosinte branched1
ids indeterminate spikelet1 (ts6) tga1 teosinte glume architecture1
id1 indeterminate1 te1 terminal ear1
kn1 knotted1 td1 thick tassel dwarf1
Lfy1 leafy1 Tu1 Tunicate1
lg1 liguleless1 ub1 unbranched1
lg2 liguleless2 Vg1 vestigial glume1
ns1; ns2 narrow sheath1 and 2 wcr wandering carpel
Pn1 papyrescent glumes1 zag1 zea AGAMOUS homolog1
Pt1 polytypic ear1 zfl1; zfl2 zea floricaula/leafy1 and 2
Key to columns: Mutant symbol or Mut symbol, the abbreviated symbol for the most com-
mon mutant alleles (dominant alleles begin uppercase and recessive alleles begin lowercase);
Gene name, the full name of the gene; Map location, chromosome and bin in which the gene
has been mapped (for cloned genes, the bin is assigned by placement on the October 2007
sequence assembly of the maize genome). Each gene is classified according to our interpre-
tation of its primary developmental implication(s): Meristem function, affects basic meris-
tem properties such as initiation, maintenance and/or size; Meristem identity, affects
processes of meristem determinacy and/or identity; Organ dev, affects organ development
and/or identity; Sex determ, affects sex determination either specifically at the organ level
(e.g., tasselseed genes) or generally by altering the sex of whole or parts of the
inflorescence(s); Gene product, the functional protein or RNA encoded by the gene (TF puta-
tive Transcription Factor)

specification between grass species like maize and rice with those of core eudicots
(Arabidopsis and Antirrhinum) have established conserved mechanisms of MADS-
box gene expression and genetic control of floral organ specification (reviewed in
Whipple and Schmidt, 2006). Finally, an understanding of the genetic and molecu-
lar mechanisms underlying maize inflorescence development will likely have
important implications to the ongoing efforts to improve maize as one of the
world’s most important cereal crops. The links between inflorescence morphology
and grain yield are readily apparent. Although rapid progress has been made dur-
ing the past 5 years in cloning and characterizing the genes underlying a number
of very informative inflorescence mutants, as is apparent from the discussion
below and the list of mutants in Table 1, this is an area of study wide open with
opportunities for research at the intersection of development, evolution, and
applied agriculture.
24 E. Vollbrecht and R.J. Schmidt

5 Insights from Analyses and Gene Cloning of Mutants

Our intent here is to introduce mechanisms of inflorescence development by

describing what we feel are a few of the more informative mutants, and by providing
an update on the rapidly expanding pool of recently cloned genes and what is
known regarding how their gene products function. For economy we will concen-
trate on work in maize, noting that recent, excellent review articles compare and
contrast our current understanding of the genetic and molecular control of inflores-
cence development in maize to that in other grass species and the model Arabidopsis
(Bommert et al., 2005a; Malcomber et al., 2006; Bortiri and Hake, 2007; Kellogg,
2007). Considering the striking similarities between many phases of tassel and ear
development, it is perhaps not surprising that the majority of mutations affecting
structural aspects of the inflorescence affect tassels and ears similarly. However,
this should not be assumed to be the case for every mutant listed in Table 1.

5.1 Mutants Affecting the Transition to Flowering

In order to produce inflorescences a plant must undergo the transition to flowering,

a process that integrates a combination of internal and external signals at the shoot
apical meristem. This integration occurs after the SAM produces leaf primordia and
stem tissue for a period of about 4–5 weeks post germination. At this point in time
the SAM elongates and transitions into an inflorescence meristem. The IM still
generates leaf primordia but in normal tassels the leaves remain greatly reduced, as
miniscule bracts that are barely visible only near the inflorescence apex (Fig. 2c).
In their axils are new meristems that quickly overtop the bracts and form the lateral
structures that ultimately comprise the tassel and ear. In mutants like barren stalk1,
where those axillary meristems are absent (Fig. 3f), or like tassel sheath or the
dominant Few-branched1 where bracts are derepressed, the bracts are clearly evident.
From studies in Arabidopsis, the transition to flowering involves a complex network
of interacting genes (Boss et al., 2004; Bernier and Perilleux, 2005; Kobayashi and
Weigel, 2007), and a similar complexity can be expected in maize (see Colasanti
and Muszynski). Indeed, more than 80 flowering time QTLs have been identified
in maize (Chardon et al., 2004).
Among the several mutants that affect the transition to flowering, indeterminate1
(id1), and zea floricula/leafy1 (zfl1) and its duplicate zfl2, are cloned genes that in
addition to showing a delay in the time to flowering show dramatic effects on the
inflorescence (Colasanti et al., 1998; Bomblies et al., 2003). The id1 gene encodes
a Zn-finger putative transcription factor that is produced in young leaf primordia and
is hypothesized to regulate production of a compound that acts elsewhere, at the
meristem (Colasanti et al., 1998; Kozaki et al., 2004). id1 mutants produce aberrant
tassels with increased branch number accompanied by unusual reversion of tassel
branch meristems to a vegetative state, including shoots and roots. id1 mutants rarely
Development of the Inflorescences 25

produce an ear at all. The zfl1 and zfl2 genes have a shared role in regulating the
transition to flowering and aspects of inflorescence architecture, as well as in speci-
fying floral meristem identity and organ development (Bomblies et al., 2003). Their
floral patterning activities suggests that gene function is at least partially conserved
with that of the eudicot LEAFY homologs, but the inflorescence role may be exclu-
sive to maize or the grasses. On the other hand, id1 appears to be a gene unique to
the grasses with no apparent ortholog identified in eudicots (Colasanti et al., 2006).
Delayed flowering (dlf1), like the other mutants affecting this transition, shows
a delay in the time to flowering and typically produces more leaves than in wild
type. Tassel and ear development are also affected with extra branching and some
feminization appearing in the tassel. The dlf1 gene encodes a member of the basic
leucine zipper family of DNA-binding proteins and appears to be a homolog of
FLOWERING LOCUS D (FD) of Arabidopsis (Muszynski et al., 2006). dlf1 is
expressed in the shoot apex before, during and after the transition, with a peak of
expression near the time of transition, consistent with the idea that a critical level
of dlf1 expression is required for the floral transition. Genetic analysis of id1 dlf1
double mutants places dlf1 downstream of id1, and suggests that dlf1 may mediate
the floral inductive signal transmitted between leaves and the shoot apex.

5.2 Mutants Affecting Meristem Size

Considerable progress has been made over the past decade in elucidating a number
of the key genes regulating meristem size and maintenance. Prominent among these
are members of the CLAVATA pathway (see Clark, 2001), originally described in
Arabidopsis where it comprises three genes, CLV1–3. These work in concert in a
signaling pathway in the central zone of the shoot apical meristem to restrict stem
cell proliferation programmed through the transcription factor, WUSCHEL
(reviewed in Williams and Fletcher, 2005). Because the BMs and SPMs are pro-
duced from the flanks of the inflorescence meristem, it is not surprising that more
than a few mutants have been described in maize that affect inflorescence development
by impacting the size or maintenance of the IM. Several of these encode homologs
of the genes that regulate meristem size in Arabidopsis, suggesting conservation of
this pathway across the angiosperms.
A CLAVATA1-like gene product is encoded by the maize thick tassel dwarf1
(td1) gene, whose mutant phenotype results in a plant with shorter stature, a tassel
with a distinctly thickened central spike, and variably fasciated ears. The sequence
and developmental role of the TD1 protein appears similar to that of CLAVATA1
(Bommert et al., 2005b). Another mutant in maize called fasciated ear 2 (fea2)
encodes a CLAVATA2 like gene product (Taguchi-Shiobara et al., 2001), a mem-
brane associated leucine rich repeat (LRR) receptor-like protein. Mutations in fea2
result in enlarged tassel and ear IMs and FMs and typically lead to ears with a
markedly fasciated phenotype (Fig. 4c). The vegetative shoot meristem appears
normal in fea2 mutants, distinct from the effects observed in clavata2 mutants in
26 E. Vollbrecht and R.J. Schmidt

Arabidopsis where all shoot meristems are affected. Thus, fea2 appears to have
evolved an inflorescence-specific role in regulating meristem size. Several other
mutants in maize such as fea3, compact plant2 and Fascicled ear1 also affect
aspects of meristem size (Table 1).

5.3 Mutants Affecting Meristem Initiation and Maintenance

Mutants that have an effect on a fundamental element of meristem function, may

lead to a failure to initiate and/or maintain meristems. In a branch system like the
inflorescence that involves multiple meristems, such failure may have various
implications. The homeodomain transcription factor knotted1 (kn1) plays a role in
maintaining all SAMs, perhaps by positively regulating meristem size (Kerstetter et al.,
1997; Vollbrecht et al., 2000). Thus, in recessive kn1 mutants any of the different
inflorescence meristems may fail, leading to inflorescences that are smaller, sparser
due to missing structures or, in the case of the ear, sometimes absent entirely when
the IM itself fails at an early developmental stage. Other mutants affect the initiation
and/or maintenance of axillary meristems without significantly affecting the SAM
or IM. These include barren stalk1 (ba1), barren inflorescence2 (bif2), unbranched1
(ub1) and the dominant mutant Suppressor of sessile spikelets1 (Sos1) among
others (see Table 1).
Mutations in ba1 prevent the formation of all axillary meristems, resulting in
mutant plants lacking vegetative tillers, ears, tassel branches and spikelets (Fig. 3f),
but otherwise have normal vegetative growth (Ritter et al., 2002). bif2 mutants have
a similar phenotype but typically produce rudimentary ears and tassels that occa-
sionally produce spikelets (McSteen and Hake, 2001). Both genes have been cloned.
ba1 encodes a putative transcription factor belonging to the bHLH class of DNA
binding proteins (Gallavotti et al., 2004) whereas bif2 specifies a serine/threonine
protein kinase (McSteen et al., 2007). Recent evidence indicates that application of
inhibitors of polar auxin transport like NPA at or near the time of the floral transition
phenocopies the bif2 and ba1 inflorescences phenotypes, suggesting that they act in
a similar, auxin-related pathway. In addition, ba1 expression is dependent on polar
auxin transport whereas bif2 transcript was still present in NPA-treated plants (Wu
and McSteen, 2007). This suggests a model in which bif2 is upstream of polar auxin
transport and polar auxin transport is required for ba1 expression. The BA1 protein
may be a target of BIF1 activity (A. Skirpan and P. McSteen, personal communica-
tion), thus providing a link between their similar phenotypes. Consistent with this
prediction, their patterns of expression overlap. ba1 is expressed in a narrow arc of
cells adaxial to the sites of new axillary meristem initiation (Gallavotti et al., 2004),
while bif2 is more globally expressed in all axillary meristems (McSteen et al.,
2007). The restricted pattern of ba1 expression adjacent to sites of new meristem
initiation (see Fig. 5), and the failure in ba1 mutants to form any axillary meristems,
suggests an essential and general role for ba1 in the formation of all axillary meristems,
be they vegetative or inflorescence.
Development of the Inflorescences 27

Fig. 5 An illustration of the expression patterns for several cloned genes affecting inflorescence
meristem initiation and/or identity. The upper portion of a developing inflorescence is repre-
sented with the approximate expression patterns of ba1, ra2, ra3, and bd1 portrayed in different
colors, beginning with the onset of each gene’s expression and continuing through formation of
SMs. For clarity, expression is only depicted in some primordia, although the patterns are the
same in every row. The overlay illustrates how the discrete patterns of each gene overlap in a
pattern that surrounds the base of the SPM and/or SM.

ub1, lg2, and sos1 are examples of genes that affect initiation of specific meristems.
ub1 mutants fail to initiate branch meristems, resulting in tassels that lack or produce
only vestigial long branches but retain normal spikelet development along the central
spike (Fig. 3c). lg2 mutants also have altered BM function. They either lack long tas-
sel branches altogether or initiate only one or two relatively normal ones, without the
vestigial structures seen in mature ub1 mutants. The tassel sheath mutants are similar,
except that bract derepression accompanies the failed BMs. lg2 encodes a bZIP tran-
scription factor (Walsh et al., 1998), while ub1 has not been cloned. A dominant,
mutant allele of Sos1, another classic mutation awaiting molecular characterization,
affects the formation of the sessile spikelet only, resulting in inflorescences with rows
of single, unpaired and only pedicellate spikelets. Unpaired spikelets appeared in
NPA-treated plants, again suggesting links between polar auxin transport and early
events in inflorescence development (Wu and McSteen, 2007).

5.4 Mutants Affecting Meristem Identity and Determinacy

Apart from general properties such as meristem initiation and maintenance that are
essential to all functioning meristems, the way a meristem behaves in a certain
context is controlled by distinct properties commonly described as meristem identity.
28 E. Vollbrecht and R.J. Schmidt

Reflecting its array of meristem types, the maize inflorescence employs multiple
genetic mechanisms to regulate meristem identity. Moreover, meristem identity,
which classically defines qualities such as the number and types of primordia that
a meristem will produce, is closely related in the inflorescence to the property of
determinacy, which relates to the capacity for continued meristem activity. Branch
meristems and the main IM possess indeterminacy as reflected in their continued
cycles of initiation and generation of a stem axis, while each of the other meristem
types in the inflorescence (SPM, SM and FM) show relative determinacy in that
they produce only short branches with a small, defined number of organs. Genetic
analysis suggests that indeterminacy is something of a ground state, because several
mutants are known that clearly increase meristem indeterminacy but there are arguably
few, if any, known mutants with increased determinacy. Put another way, the deter-
minate meristem identities seem to be sequential and imposed by a number of genes
(Irish, 1997) on a default, indeterminate fate (Vollbrecht et al., 2005).
For example, a normal function of the three ramosa genes (ra1, ra2 and ra3) is
to specify the determinate identity of the SPM, and thereby limit branch outgrowth.
In each of the ramosa mutants SPMs have increased indeterminacy and assume fates
more like a BM, leading to branched ears and to tassels with increased degrees of
branching (Figs. 3b and 4b, d). Moreover, the mutants make novel branches whose
length and identity is transitional between BM and SPM (Vollbrecht et al., 2005;
Satoh-Nagasawa et al., 2006). This blurring of the normally sharp distinction
between long branches and spikelet pairs demonstrates the interplay between iden-
tity and determinacy, where evolution has led to fates such as BM and SPM in maize,
but to different permutations of determinacy, meristem order and other basic proper-
ties to produce related but distinct branching architectures across the grasses
(Kellogg, 2000a; Vollbrecht et al., 2005; Malcomber et al., 2006). From work on the
ramosa and other genes in maize, we now have clues about the underlying molecular
mechanisms. ra2 encodes a LOB domain transcription factor (Bortiri et al., 2006b)
and is expressed in the incipient primordia of all axillary inflorescence meristems
except the FMs, well before the developing meristems protrude as bumps on the
lateral surface of the inflorescence (Fig. 5). This suggests that ra2 has a fundamental
and general role in axillary meristem specification, perhaps distinct from its effects
on meristem determinacy. Consistent with this, ra2 mutants have elongated branch
and spikelet internodes, a feature not exhibited by other determinacy mutants. ra1
also encodes a plant-specific transcription factor, one belonging to the C2H2 zinc
finger protein class (Vollbrecht et al., 2005). Its expression initiates after but overlaps
that of ra2, in a discrete domain associated with the base of initiating SPMs (Fig. 5).
Consistent with the gene’s initial expression, the ra1 mutant phenotype is the most
restricted among the ramosa mutants, being confined to effects on SPM determi-
nacy. ra3 encodes a trehalose 6-phosphate phosphatase whose expression also initi-
ates in a basal domain concomitant with the emergence of SPMs, but like ra2, ra3
is also expressed in later stages, and ra3 affects determinacy of both SPMs and SMs
(Satoh-Nagasawa et al., 2006). Each of the pairwise double mutants shows a syner-
gistic effect on SPM determinacy and ra1 expression is reduced in ra2 single
mutants and in ra3 single mutants, but ra2 and ra3 expression is unaltered in the
Development of the Inflorescences 29

other single mutants. Taken together, these data imply one genetic pathway in which
ra1 is downstream of ra2 and ra3 to impose determinacy on the SPM, while in paral-
lel pathways, ra2 and ra3 regulate additional functions including determinacy of
higher order meristem types. Whether or not ra1 has other roles outside of ra2 and
ra3, awaits careful analysis of the triple mutant.
Identity and determinacy of one of the higher order meristems, the SM, is regu-
lated in part by the branched silkless1 (bd1) and indeterminate spikelet1 (ids1)
genes, which encode distinct transcription factors of the AP2-domain class (Chuck
et al., 1998, 2002). In bd1 mutants the SM is indeterminate in both the tassel and the
ear, but with slightly different consequences. In the mutant tassel, the converted SM,
now indeterminate, bears numerous spikelets as in the ramosa mutants, and eventu-
ally produces fertile flowers. The extra spikelets result in a mutant tassel (Fig. 3a)
whose branches and central spike appear thicker in comparison to a normal tassel
(Fig. 1a). But in the ear, the indeterminate SMs do not produce florets. Instead, the
mutant SMs near the base of the ear often appear to take on an identity like that of
the SPM, initiating new meristems in a distichous fashion, some of which in turn
appear to again adopt a SPM fate, resulting in complex, compound branching
(Fig. 4a). Interpretation of the phenotype is further complicated by the occurrence of
additional indeterminate, SM-like meristems arising in a novel location, the axil of
the normally sterile glume that subtends the SM on the main axis (Colombo et al.,
1998; Chuck et al., 2002). On one hand, this mutant phenotype suggests a regression
of SMs to BM identity, and therefore that bd1 connotes SM identity (Chuck et al.,
2002). However, it also suggests a more general role for bd1 in SM determinacy and
suppression of meristem outgrowth in the axil of the glume. This complex mutant
ear phenotype further blurs the distinction between SM determinacy and identity.
Intriguingly, like the ramosa genes, bd1 is expressed in a specific domain, in an arc
at the base of the meristem it regulates, the SM (Fig. 5).
ids1 is expressed more broadly in the SPM, SM and some floral primordia,
perhaps reflecting its distinct role as a target of regulation by microRNAs (see
below). But in ids1 mutants only the spikelet axis shows increased indeterminacy,
by producing multiple florets per spikelet rather than the requisite two (Chuck et al.,
1998). A strikingly similar defect is shown by reversed germ orientation1 (rgo1)
mutants. Intriguingly, recessive alleles of rgo1 and ids1 exhibit nonallelic noncomple-
mentation and double mutants show a synergistic, indeterminate phenotype in both SM
and SPM, demonstrating a function for ids1 in other meristems and fueling the
hypothesis that meristem identity and determinacy may be regulated by factors that
function in a dosage-sensitive manner (Kaplinsky and Freeling, 2003).
A novel mechanistic theme emerges from recent analyses of determinacy and
other inflorescence gene expression patterns. For several genes, the domain of RNA
expression is not within the meristem per se that the gene regulates, but in a boundary
domain that surrounds the base and/or adaxial side of the meristem, as to visually
separate the meristem from the main inflorescence axis. Such an expression pattern
is observed for genes including ba1, a meristem initiation/maintenance gene, and the
determinacy genes ra1, ra2, ra3 and bd1 (Fig. 5). In the cases of ra2 and ba1, the
genes are expressed even before the corresponding meristems are initiated. Further
30 E. Vollbrecht and R.J. Schmidt

characterization of the functions and locations of the encoded proteins as well as the
transcriptional targets and corresponding gene networks, should help elucidate this
mechanism. Nevertheless, the expression patterns suggest that in the inflorescence,
disparate properties such as meristem initiation and determinacy are controlled by a
mobile signal(s) whose activity may be gated or regulated spatially at the junction
with the main inflorescence axis (Gallavotti et al., 2004; Vollbrecht et al., 2005).
With the exception of ra3, all of these boundary domain genes encode putative
transcription factors. It is tempting to speculate about the nature of the mobile signal.
For example, the putative transcription factors are all small proteins that may themselves
be mobile. Alternatively, sugar or hormone signaling or signal transduction may be
involved, upstream and/or downstream of the transcription factors.

5.5 Mutants Affecting Sex Determination

The process by which strictly staminate (male) florets are produced in the tassel and
pistillate (female) florets in the ear is termed sex determination, and a number of
mutants affect this pathway (Table 1). The genes underlying the many andromonoe-
cious dwarfs (an1, d1, d2, d5, D8, and D9) play a role in promoting stamen abortion
in ear florets. Mutants typically produce anthers in the ear florets although these do
not usually shed pollen. In addition, the mutants are generally characterized as
dwarfs due to shortened internodes that produce broad and shortened leaves. The
tassel branch number can also be reduced. The recessive andromonoecious dwarfs
fail to synthesize normal levels of gibberellic acid (GA) and normal stature and
floret development can be rescued by GA application (Phinney, 1956; Phinney and
West, 1960). The specific enzymatic activity encoded by many of the dwarf genes
has been predicted by careful feeding experiments (Spray et al., 1984); indeed, the
cloning of an1 (Bensen et al., 1995) confirmed its role in the biosynthetic pathway.
The dominant D8 and D9 mutants, however, have defects in a DELLA repressor
protein that interacts with the GA receptor and are hence unresponsive to GA appli-
cations (Harberd and Freeling, 1989; Peng et al., 1999). Clearly activity of the plant
hormone, GA, is important to the formation of pistillate flowers in maize ears, but
the mechanism remains unresolved. Insights into the connection between GA and
sex determination should be facilitated by the molecular characterization of those
genes with more specific effects on the sex determination pathway.
The group of tasselseed mutants (Table 1) shares in common the partial to nearly
complete conversion of the normally male tassel spikelets into ones that bear female
florets (Nickerson and Dale, 1955). This often results in seeds being formed on the tassel,
sometimes as profusely as on the ear. Two groups of tasselseeds were established based
on genetic analyses and morphological considerations (Irish et al., 1994), with ts1, ts2,
Ts3, and Ts5 falling in the group where sex reversal was not accompanied by extra
branching, and ts4 and Ts6 affecting sex determination and causing proliferative inflo-
rescence branching. ts2 was the first tasselseed gene to be cloned (DeLong et al.,
1993). It encodes a short chain alcohol dehydrogenase that is expressed in both ear and
Development of the Inflorescences 31

tassel spikelets. This leads to an apoptotic-based degeneration of the gynoecium in the

tassel florets as well as in the lower floret of the ear (Calderon-Urrea and Dellaporta,
1999). The ts1 gene product is required for ts2 mRNA accumulation placing ts1
upstream of ts2. Interestingly, the silkless ears1 (sk1) gene appears to be a key player
in this pathway. sk1 mutants fail to make a gynoecium and so produce ear florets and
mature ears that lack silks (Fig. 4f). Surprisingly, the double mutant between sk1 and
ts2 results in a silk-containing, essentially ts2 phenotype. Thus ts2 is epistatic to sk1,
suggesting a model (Calderon-Urrea and Dellaporta, 1999) of sex determination in
which TS1 promotes expression of ts2 mRNA within cells of the developing pistils.
The TS2 dehydrogenase then modifies a yet unidentified steroid that promotes pistil
cell death. sk1 protects the pistil in the upper ear floret by either interfering with TS2
activity directly, or by blocking a downstream step in the cell death response. The vali-
dation of this model will have to wait the cloning of sk1, but it predicts that the SK1
product will be specifically expressed in the upper floret only. Genetic analyses have
shown ts2 to be additive with d1 although, as already indicated, epistatic to sk1. Thus,
the formation of unisexual florets in maize involves an interplay of at least two appar-
ently independent pathways, one involving GA and one involving TS1, TS2, and SK1
in an apoptotic cell death program.
The complexity of the sex determination process is highlighted by the recent
cloning of two of the genes responsible for the proliferative tasselseed mutants. The
recessive ts4 and the dominant Ts6 alleles affect tassel and ear development simi-
larly (Irish et al., 1994; Irish, 1997), causing proliferative branching due to acquired
indeterminacy of the SPM and SM, and the production of tassel florets bearing pis-
tils. Molecular cloning of these genes (Chuck et al., 2007b) reveals ts4 to encode a
microRNA of the miR172 family. One of the targets of this microRNA is the AP2
domain-encoding gene, ids1, discussed above with respect to SM determinacy.
Remarkably, the dominant Ts6 allele contains a single nucleotide change in the ids1
gene, in the binding site for the ts4-encoded microRNA. Hence, ids1 and Ts6 are
allelic. The nucleotide change renders the dominant Ts6 allele free of ts4-mediated
regulation and therefore leads to ectopic accumulation of ids1 RNA and IDS1 protein.
Double mutants of ts4 with a recessive allele of ids1 have restored, near normal tas-
sel sex determination, but the SPMs and SMs still show some indeterminacy. This
double mutant phenotype indicates that most of the ts4 mutant phenotype is due to
misregulation of ids1. However, the residual SPM and SM indeterminacy, indicates
that there is at least one other target of ts4. The nature of the link between the
branching pathway and sex determination is still unclear, but may be clarified by the
cloning of the remaining tasselseed mutants and identification of downstream targets
of ids1. However, it seems that additional links between inflorescence development
and epigenetic regulation are likely. For example, mutants of the required to maintain
repression6 (rmr6) gene, initially identified on the basis of rmr6’s role in paramutation
of anthocyanin regulatory genes, show a tasselseed phenotype. Genetic analysis
showed that rmr6 maintains maize’s monoecious pattern of sex determination by
restricting the activity of sk1 (Parkinson et al., 2007). Although not demonstrated as
for ts4, rmr6 may also implement epigenetic repression by a mechanism that involves
noncoding small RNA molecules.
32 E. Vollbrecht and R.J. Schmidt

5.6 Mutants Affecting Organ Specification

and Floral Meristem Identity

There are several mutants of maize that act very late in the program of inflorescence
development, affecting specifically the determinacy of the floral meristem or the
specification of floral organs. One mutant with effects confined to a specific spikelet
organ is Vestigial glume1 (Vg1). In these mutants the tassel and ear appear normal
with the exception that the glumes are dramatically reduced in size, leaving the
stamens exposed on developing tassel florets, resulting in reduced fertility due to
desiccation and premature sun exposure.
Other mutants are specific to development of floral organs. The sk1 mutant,
discussed above for its role in sex determination, has ears without silks, but other-
wise produces a morphologically normal ear and tassel. The opposite phenotype is
conditioned by the silky1 (si1) mutant which produces numerous silks emerging
from both tassel and ear spikelets. A mutant tassel producing silks is reminiscent of
the tasselseed class of mutants. However, in the case of si1 mutants the silks arise
in place of stamens due to homeotic floral organ conversions. si1 encodes a MADS-
box gene related to the B-class floral homeotic MADS-box genes APETALA3 of
Arabidopsis and DEFICIENS of Antirrhinum (Ambrose et al., 2000). In the si1
mutant tassel spikelets, the central gynoecium aborts as in wild type, but sterile
carpels replace the stamens, and lemma/palea-like structures replace the lodicules.
Unlike in the tasselseeds, the silks that form in place of stamens are sterile, and so
never set seed. Similar organ conversions occur in si1 ears such that the mutant ear
spikelets each contain three additional silks in place of the stamens, and a func-
tional primary silk.
zea agamous 1 (zag1) is another MADS-box gene that has been functionally
characterized in maize. The zag1-mum1 mutant was obtained through a reverse
genetics approach that targeted one of the first MADS-box genes identified in
maize, specifically a candidate ortholog of the Arabidopsis C-class floral homeotic
gene, AGAMOUS (Schmidt et al., 1993). Mutants produce ears with greatly reduced
fertility, resulting from production of sterile silks emerging from the indeterminate
floral meristem of ear spikelets (Mena et al., 1996), indicating a role for zag1 in
floral meristem determinacy, a conserved function of C-class genes. No effects
were observed on stamen identity, as might have been predicted for a loss of func-
tion mutation in an AGAMOUS homolog. This surprising result was explained
based on the partitioning of the predicted C-function activities between zag1 and its
duplicate gene zmm2. The expression patterns of these two genes within developing
stamens and carpels are distinct (Mena et al., 1996), suggesting that this floral
organ identity function has been subfunctionalized during maize evolution. The
double mutant between zag1 and si1 produced striking mutant tassels and ears with
spikelets consisting of proliferating bracts having lemma and palea character
(Ambrose et al., 2000), a phenotype reminiscent of the B- and C-function double
mutants in Arabidopsis where a proliferation of sepals replaces normal floral organ
development (Bowman et al., 1991). These and other data (reviewed in Whipple
Development of the Inflorescences 33

and Schmidt, 2006) provide support for the conservation of the ABC program of
flower development within maize.
Finally, a number of mutants discussed earlier (bif2, kn1, zfl1, and zfl2) and
others like indeterminate floral apex1 (Laudencia-Chingcuanco and Hake, 2002)
have effects on floral organ development, but unlike the mutants discussed above,
their effects are not restricted to this late stage of inflorescence development, but
instead more globally impact all the meristems of the inflorescence. Worth noting
here is the dramatic phenotype of double mutants between ifa1 and zag1 in which
floral meristems within the developing ear revert to a BM identity producing ear
spikelets from which a discrete inflorescence branch emerges (Laudencia-
Chingcuanco and Hake, 2002). Since ifa1 alone has only modest affects on floral
organ development and floral meristem determinacy, the double mutant pheno-
type suggests a redundant role of both ifa1 and zag1 in promoting floral meristem
identity, independent of their separate roles in floral meristem determinacy.

5.7 Mutants Lacking an Inflorescence

Lastly, there exists in maize several curious mutants where one or the other inflores-
cence fails to form altogether. tasselless1 (tl1) mutants lack a tassel at maturity, but can
produce ears and may otherwise appear normal (Fig. 3e). Anecdotal evidence suggests
that this is a phototropic response that can be ameliorated by growth under short days,
but detailed studies on this mutant are lacking. Similarly, the mutants barren stalk2 and
3 (Pan and Peterson, 1992; Neuffer et al., 1997) lack ears and tillers but produce an
otherwise normal plant including a normal tassel. In the case of ba2, at least, ears
shoots are initiated but then arrest their development at such an early stage that the
shoot never emerges from the axil of the leaf. Unlike ba1, which affects initiation of
all axillary meristems, ba2 and ba3 are specific to the axillary shoot meristem, but
affect the continued growth of the shoot rather than its initiation or formation.

6 Relationship of the Inflorescence to the Whole Plant

Thus far we have discussed the inflorescences more or less separately from the rest
of the plant, but they develop in the context of and in response to signals generated
by the whole plant. At one extreme are heterochronic mutants with pervasive
effects on vegetative shoot architecture that have correlative alterations in inflores-
cences (e.g. Teopod1 and 2, Corngrass1). The recent cloning of the dominant
Corngrass1 (Cg1) mutation provides some insight into the basis of such a global
effect on development. Cg1 mutants (Fig. 6a) are a consequence of the ectopic
expression of a noncoding regulatory RNA locus, miR156, whose apparent misex-
pression throughout development perturbs the transcript levels of at least 13 puta-
tive target genes belonging to the SPL family of transcription factors, with several
34 E. Vollbrecht and R.J. Schmidt

Fig. 6 Examples of mutants that affect whole plant and inflorescence architecture. (a) a normal
plant (left) and a sibling Corngrass1 (Cg1) mutant. (b) a tassels replace upper-ears1 (tru1) mutant
and (c) a schematic showing the fate of axillary meristems in the mutant shown in (b). In Cg1
mutants axillary meristems are derepressed and proliferate throughout the plant, accompanied by
heterochronic changes in shoot and organ identity. tru1 mutants produce a relatively normal main
shoot (represented in green in panel c) but axillaries are differentially derepressed. The normally
pistillate ear shoot and a few axillaries below it develop instead as long, staminate branches (sb,
blue axes) tipped by tassel-like inflorescences. An ear shoot (red axis) forms at a lower position,
while the lowest axillary buds (yellow circles) remain repressed

of these normally expressed during both vegetative and inflorescence development

(Chuck et al., 2007a). Considering the similar molecular bases of Ts6 (discussed
above) and Cg1, it seems likely that other dominant mutants of maize (see Chapter
by Neuffer et al) with global effects on development may turn out to be alteration
in pathways involving regulatory RNAs. As mentioned in Sect. 5.1, one integrative
signal in plant ontogeny cues the transition to flowering and some mutants that
affect this transition in maize, also show dramatic effects on inflorescence develop-
ment (e.g., id1). When these mutants with pleiotropic phenotypes are considered as
a group, it is evident that particular suites of inflorescence-related characteristics
are affected, with different subsets changing in different mutants. Rather than list
mutants with all of their phenotypes, in this section we will consider them as a
group and discuss some of the themes that emerge from a synthetic perspective.
One unifying phenotype among mutants that demonstrate integration of whole
plant and inflorescence development, is a derepression of axillary meristems. Just
as inflorescence architecture arises from patterns of meristem determinacy, overall
plant architecture reflects substantially the activity of vegetative axillary meristems
(McSteen and Leyser, 2005). In maize, several axillary meristems are present on the
main shoot at flowering but typically only the uppermost few grow out, to form the
ear shoots (see Fig. 1b). Traversing down the plant (basipetally) from the leaf
axil(s) that contain ears, sequential axillary meristems are repressed and remain
quiescent, except that the lowermost may be released early in ontogeny to form
tillers that recapitulate the main plant axis. The TCP transcription factor gene teosinte
Development of the Inflorescences 35

branched1 (tb1) regulates the determinacy of these axillary meristems on the main
shoot. In plants containing tb1 alleles with reduced or no function, all axillary mer-
istems up to and including those at the ear node are derepressed and grow out to
produce long axes, each one much like a tiller (Hubbard et al., 2002). Thus a lateral
branch that would normally be compressed and tipped by a pistillate ear, in tb1
mutants is elongated with a staminate, tassel-like structure at its terminus. The tassels
replace upper-ears1 (tru1) mutant (Sheridan, 1988) similarly transforms the erstwhile
ear branch to a long branch tipped by a staminate inflorescence (Fig. 6b, c), but
does not strongly affect overall branching.
Indeed, the tru1 mutation derepresses only a few axillary meristems, at and below
the normal ear position. Below the uppermost elongated, tassel-tipped branch, the
branches are progressively shorter and the lower axillaries do not grow out; tru1
mutants do not show a marked tillering phenotype. The similarity of transformed
lateral branches of tru1 mutants to the staminate ear tips frequently seen in many
normal maize lines, and often attributed to genotype by environment interaction (Coe
et al., 1988) suggests an involvement of phytohormones in this phenotype. Flowering
time mutants such as those corresponding to the transcription factor genes dlf1 and
zfl1 and 2, show a more subtle degree of derepression, of axillary meristems on the
ear shoot itself. These mutants typically produce at the normal position, an shoot that
is ear-tipped but whose derepressed axillaries lead to multiple, closely spaced ears
attached to the same shank (Bomblies et al., 2003; Muszynski et al., 2006).
These and other mutants also suggest a general correlation between axis length and
inflorescence sex, in which inflorescences that tip long axes tend to be staminate, while
those on compressed axes tend to be pistillate, perhaps correlating with a physiology
inherent to short branches (Iltis, 1983; Irish, 1996). tb1 mutants provide the long
branch example, typically with staminate inflorescences on all lateral branch termini.
A series of transitional inflorescence forms may be evident in tru1 mutants, where the
uppermost and longest lateral branch shows the strongest and most consistent stami-
nate transformation, sequentially lower branches are progressively shorter and more
feminized and a normal ear forms at a somewhat lower-than-normal node (Fig. 6b, c).
The same trend manifests in terminal ear1 and zfl mutants but with respect to the
length of the main shoot itself; both mutants have shortened vegetative internodes
immediately preceding the tassel, so that the tassel is borne on a relatively shortened
main axis, and both form strikingly similar feminized branches at the tip of this shortened
axis, in the positions normally occupied by lower tassel branches (Veit et al., 1998;
Bomblies et al., 2003). In analogous fashion, the long, staminate-tipped branches of
tb1 mutants bear small pistillate inflorescences laterally, especially in axillary positions
of the shorter, uppermost branches off the main axis. Even ts1 and ts2 main stems,
which are tipped by a pistillate inflorescence, tend to be shortened. Interestingly, the
GA-related dwarf mutants provide a conspicuous counter-example; axis length is
reduced in mutants, but inflorescences are always staminate or andromonoecious.
Finally, one group of mutants may be seen as altering the spacing or developmental
coordination between the tassel and ear. This group includes mutants with intact axillary
meristems but that otherwise do not elaborate a mature ear, discussed in Sect. 5.7. A
similar ear phenotype occurs in some flowering time mutants. For example, dominant
36 E. Vollbrecht and R.J. Schmidt

Leafy1 mutants modestly increase flowering time and increase, by a few, the number
of nodes below the ear and between the ear and the tassel, while strong mutants of the
zinc-finger protein gene id1 rarely form an ear at all, even when they do eventually
produce tassels. These phenotypes demonstrate that coordinating whole plant archi-
tecture and placement of the ear relative to the tassel is regulated in part by an integrative
component of the floral transition mechanism. Considered in this context, mutants like
ba2 and ba3 that flower and make axillary meristems normally but do not elaborate
an ear, may have ear development blocked in a related way.

7 Concluding Remarks

Considerable progress has been made in the last 5 years in the cloning and characteri-
zation of genes underlying many of the inflorescence mutants. This has provided
valuable information on patterns of gene expression and the nature of the gene products,
and enabled researchers to begin assembling testable models of gene networks.
Interestingly, an overwhelming majority of these cloned genes encode transcriptional
regulators, or regulatory molecules like small RNAs that have transcription factors as
their targets. This lends support to the idea that subtle changes in the timing or pattern
of expression of these genes may well underlie morphological variation in inflores-
cence structure that has evolved within the grasses (Doebley and Lukens, 1998;
Vollbrecht et al., 2005). Establishing those associations will require the development
of genomic resources in other grass species. The advent of positional cloning in maize
(Bortiri et al., 2006a), made possible by the emerging genomic resources in maize and
rice, has revolutionized the pace of cloning genes identified by mutant phenotypes.
An additional challenge now will be to keep pace with the genetic analyses, so impor-
tant to understanding the hierarchical relationships between genes. Linkages between
plant hormones and inflorescence gene activities are becoming ever more apparent.
Databases of cataloged mutants are providing new alleles of existing mutants and
uncovering new genes for further study and integration into existing genetic frame-
works. An expanding platform of reverse genetic resources (see Volume 2) is facilitating
the identification of mutations in candidate genes that are identified through studies
in other organisms and/or through expression and other analyses. Thus, we can antici-
pate that continued mutant analysis and genomics and other molecular approaches
will flesh out the complex interplay of genes, ultimately leading to the identification
of downstream targets and an understanding of relevant gene and protein interactions.
These efforts will lead us closer to an understanding of the mechanisms underlying
patterns of meristem activities that ultimately lead to the production of the tassel
and ear.

Acknowledgments The authors wish to acknowledge contributions by colleagues and lab mem-
bers (past and present) in the form of stimulating discussions, and toward images and figures
appearing in this chapter. Special thanks to G. Chuck for Fig. 6a, D. Hall for work on Fig. 2g, h.
We also acknowledge the generous support of the National Science Foundation and the U.S.
Department of Agriculture CSREES.