Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. 2005 Jun 28; 102(26): 9412–9417.
Published online 2005 Jun 15. doi:  10.1073/pnas.0503927102
PMCID: PMC1166634
Plant Biology

microRNA172 down-regulates glossy15 to promote vegetative phase change in maize


Shoot development in many higher plant species is characterized by phase change, where meristems and organs transition from one set of identities to another. The transition from a juvenile to adult leaf identity in maize is regulated by the APETALA2-like gene glossy15 (gl15). We demonstrate here that increasing gl15 activity in transgenic maize not only increases the number of leaves expressing juvenile traits, but also delays the onset of reproductive development, indicating that gl15 plays a primary role in the maintenance of the juvenile phase. We also show that the accumulation of a maize microRNA homologous to miR172 increases during shoot development and mediates gl15 mRNA degradation. These data indicate that vegetative phase change in maize is regulated by the opposing actions of gl15 and miR172, with gl15 maintaining the juvenile phase and miR172 promoting the transition to the adult phase by down-regulation of gl15. Our results also suggest that the balance of activities between APETALA2-like genes and miR172 may be a general mechanism for regulating vegetative phase change in higher plants.

Keywords: juvenile-to-adult transition, flowering time

Vegetative phase change has been characterized in many plant species (1, 2). The most obvious features that distinguish the basal juvenile and upper adult phases are the capacity for reproductive development (an adult trait) and heteroblasty, where leaves with distinct physiological and morphological identities are produced during shoot development (3, 4). Studies of vegetative phase change suggest that this process is regulated independently from the transition to reproductive development, although there may be overlap and coordination among vegetative and reproductive programs (5).

In maize (Zea mays L.), the basal juvenile leaves (the first five or six in most genotypes) differ from upper adult leaves in their expression of a suite of epidermal cell characteristics, including epicuticular waxes, cell wall features, and the presence of specialized cell types such as leaf hairs. Mutations in at least seven maize genes and the gibberellic acid class of plant growth regulators have been shown to alter the relative expression of juvenile and adult leaf identity (2). Among these genes is glossy15 (gl15), an APETALA2-like gene that is expressed in juvenile leaves and plays a central role in regulating the epidermal cell traits that define leaf identity (68).

The progressive nature of vegetative phase change in maize and other plant species suggests that this process is regulated by the relative levels of both juvenile and adult regulatory factors (6, 7). Other than gl15 being required to promote juvenile leaf identity in maize, little is known about the factors controlling vegetative phase change or their regulation. Recently, two studies in Arabidopsis have demonstrated that the microRNA miR172 antagonizes the activity of AP2 and a specific subset of AP2-like genes to regulate floral organ identity and flowering time (9, 10). The target sequence for miR172 is also present in the gl15 transcript, suggesting that a homolog of miR172 may down-regulate gl15 to promote adult vegetative development in maize.

To test the above hypothesis, we first generated transgenic maize lines with increased gl15 gene dosage and activity. These lines produced a greater number of juvenile leaves and delayed flowering, indicating gl15 functions to maintain both leaf identity and shoot meristem aspects of the juvenile phase. We also measured gl15 mRNA expression and miR172 accumulation during shoot development and showed that as in Arabidopsis, maize miR172 is not expressed during early shoot development. Maize miR172 is first detected during the transition to the adult phase, concurrent with a sharp decline in gl15 mRNA expression. Recovery of gl15 transcripts that were cleaved within the putative miR172 target site suggests that miR172 mediates gl15 transcript degradation, which differs from the apparent translational repression of AP2-like genes by miR172 in Arabidopsis (9, 10). The results presented here indicate that vegetative phase change in maize is regulated by miR172-mediated down-regulation of the juvenility gene gl15. Our findings also suggest that reductions in AP2-like gene activities by miR172 may be a general mechanism for regulating phase change in higher plants.

Materials and Methods

Maize Transformation. A 5,967-bp EcoRI genomic fragment (Fig. 1), containing the gl15 gene, was isolated from the maize inbred line W64A (8), released from its plasmid vector by EcoRI digestion, and gel purified. A DNA fragment from the pCAMBIA2200 plasmid (11) containing the cauliflower mosaic virus 35S promoter driving the expression of the npt II selectable marker gene was similarly prepared. The DNA fragments were coprecipitated on 0.6-μm gold particles at a concentration of 15 ng/mg of particles and introduced into immature embryos from the maize inbred line H99 through particle bombardment (12). Transgenic calli were selected on media containing 300 μg/ml paromomycin (13) and regenerated as described in ref. 12. Transgenic R0 lines were propagated by crossing to both the H99 inbred and a genetic stock containing the null gl15-Hayes (gl15-H) mutation and the genetically linked (10 centimorgans) endosperm marker waxy1 (wx1) in a W64A inbred background.

Fig. 1.
gl15 transgenic lines. (A) Map of the gl15 genomic DNA fragment used in transformation experiments. Exons are thick black arrows or vertical lines; the promoter region is a white arrow. Positions of miR172 target site, AP2 DNA-binding domain, and primers ...

Analysis of Gl15 Transgenic Lines. Plants that carried gl15 transgenes (Gl15-TG) were identified among R1 progeny from crosses between individual Gl15-TG events and either H99 or the wx1,gl15-H stock based on their visual phenotype (additional juvenile leaves) and amplification of the cointegrated npt II transgene by PCR with the following primers: 5′-ATCTCCTGTCATCTCACCTTGC and 5′-CAAGCTCTTCAGCAATATCACG. Hemizygous Gl15-TG/0 progeny from the cross to wx1,gl15-H were backcrossed again to the wx1,gl15-H stock, which generated ears segregating for Gl15-TG, gl15-H, and the wx1 endosperm marker. At least 50 wx1 kernels from each of these ears were planted in the greenhouse or summer nursery, and plants were scored for leaf identity traits (visible juvenile waxes and macrohairs).

Lines homozygous for Gl15-TG7 in either the H99 inbred or the wx1,gl15-H backgrounds were obtained by selfing R1 plants hemizygous for Gl15-TG7. Homozygosity at Gl15-TG7 was confirmed by outcrossing putative homozygotes to either H99 or the wx1,gl15-H stock and observing the Gl15-TG7 phenotype (2–3 additional juvenile leaves) in all outcross progeny (at least 30 plants scored). Confirmed H99: Gl15-TG7 homozygotes were then crossed as males onto B73 inbred ears to produce transgenic hybrids hemizygous for Gl15-TG7.

A minimum of 50 seeds from different Gl15-TG genotypes as well as the H99 inbred, B73 inbred, wx1,gl15-H stock, and B73 × H99 hybrid were planted and evaluated in the 2003 summer nursery at the Crop Sciences Research and Education Center in Urbana, IL. Bluish-gray juvenile waxes were used as a visual marker for juvenile leaf identity, whereas macrohairs served as a marker for adult leaf identity. To simultaneously monitor both juvenile and adult leaf epidermal traits at the cellular level, leaf epidermal peels were prepared and stained with toluidine blue O as described in ref. 8.

DNA gel blots were prepared with ≈10 μg of HindIII-digested genomic DNA isolated from leaf tissue of Gl15-TG and H99 lines. The blots were hybridized by using a gl15-specific probe as described in ref. 8.

Tissue Sampling and RNA Isolations. Shoot tissues were harvested for RNA isolation at various developmental stages according to the number of days after sowing (DAS). All samples were taken from pools of 10 to 30 plants for a given stage, genotype, and tissue type. Each sample shown in Fig. 2 included the shoot apices and the basal 2 cm of any leaves that were <4 cm in total length. For the genotypes examined, DAS staging corresponded closely to leaf number staging, such that all of 12 DAS samples excluded tissue from leaves 1–3 but not from leaf 4. “Apex” samples shown in Fig. 3 included some leaf tissue but not from any leaves that were >2 mm in total length. “Leaf” samples shown in Fig. 3 were taken from leaves that were ≈3 cm in length. RNA was isolated by using TRI REAGENT (Molecular Research Center, Cincinnati) according to manufacturer's protocols. Quantitation and quality checks of total RNA were performed by A260/A280 spectrophotometry and electrophoresis in denaturing 1.5% agarose gels.

Fig. 2.
gl15 mRNA and miR172 expression during shoot development. (a) Real-time PCR was used on cDNA to assay gl15 mRNA expression in shoot tissue at various DAS. Relative fold expression (2-(ΔΔcT)) in tissue from all three genotypes is reported ...
Fig. 3.
gl15 mRNA and miR172 expression during shoot and leaf development. (a) Northern analysis of miR172 accumulation in H99 shoot apex only (A) or leaf only (L) tissues harvested between 7 and 14 DAS. The blot (Right) shows 3-cm leaves further divided into ...

Profiling gl15 mRNA Expression. Reverse transcription reactions were performed on 8 μg of total RNA at 52°C in 50 μl of volume by using oligo(dT) primer and SuperScriptIII (Invitrogen). Relative gl15 transcript abundance in the RNA samples was estimated by performing real-time PCR on the cDNA samples by using maize polyubiquitin1 (ubi1) as a control gene. gl15 primers (5′-TTCGTCAGGGGCAGCTCCAGGT, 5′-TCCGGCAGACCGAGCGTCA, spans junction of exons 8 and 9) that failed to amplify a PCR product from H99 and W64A genomic DNA were used to amplify a 371-bp product from H99 and H99:Gl15-TG7 cDNA samples. The 209-bp fragment from the 3′UTR of ubi1 was amplified by using primers described in ref. 14.

Real-time PCR reactions were performed by using Dynamo SYBR-green reagents (Finnzymes, Helsinki) in the Opticon2 DNA Engine (MJ Research, Cambridge, MA) and used the following cycling profile: 94°C for 5 min, 40 cycles of (94°C for 10 sec, 61°C for 20 sec, 72°C for 15 sec with fluorescence reads at 72, 80, and 83°C), final extension at 72°C for 4 min, and melt curve analysis at 0.2°C increments from 65°C to 95°C. Multiple fluorescence reads were necessary because the melting temperatures of the ubi1 and gl15 amplicons were 81°C and 87°C, respectively, and a gl15 primer dimer that formed in the absence of gl15 cDNA template melted at 82°C. The melt curve analysis allowed us to assess the uniformity of the PCR products amplified in each well; we saw no evidence of contamination or mispriming.

In each 96-well PCR run, 14 samples plus a no template control and a no reverse transcriptase control were analyzed in triplicate by using gl15 and ubi1 primers. For each sample, a ΔcT value (the difference in number of PCR cycles required to cross the threshold into the linear amplification phase in the gl15 versus ubi1 reactions) was calculated by averaging the three technical replicates and then subtracting the ubi1 cT from the gl15 cT. ΔΔcT values were obtained by subtracting the ΔcT of a given sample from the ΔcT of the reference sample and were only calculated for samples analyzed in the same run. Fold relative mRNA expression differences between the reference sample and the other samples are calculated as --(ΔΔcT) for each sample, reflecting the theoretical doubling of amplicons from one PCR cycle to the next.

Three pooled plant RNA samples for each stage and genotype were analyzed in triplicate to generate the data presented in Fig. 2a. Propagation of among-biological-replicate standard errors for the transformation of ΔcT values to ΔΔ-fold expression differences was performed as described in ref. 15. The data presented in Fig. 3a were generated from triplicate PCRs of single pooled plant RNA samples for each time point and tissue type of H99, so no error bars are shown. The results presented are consistent with observations from similar experiments performed with two other genotypes (data not shown).

Profiling miR172 Accumulation. Forty-five micrograms of total RNA per sample were electrophoresed in denaturing 15% polyacrylamide gels and electroblotted to a charged nylon membrane. An antisense miR172 DNA oligonucleotide (5′-AGAATCTTGATGATGCTGCAT-3′) was end-labeled as described in ref. 9 and hybridized to RNA gel blots overnight at 37.5°C. Blots were washed twice for 15 min in 3 × SSC (1× SSC = 0.15 M sodium chloride/0.015 M sodium citrate, pH 7), 0.5% SDS at 37.5°C before being exposed to a phosphor screen, and scanned on a Typhoon phosphorimager (Molecular Dynamics). Blots were stripped of miR172 probe by incubation in 3 × SSC, 0.5% SDS at 42°C, and reprobed with an end-labeled antisense miR166 oligonucleotide (5′-GGGGAATGAAGCCTGGTCCGA-3′) or a PCR-amplified and radiolabeled 225-bp fragment from maize ubi1 (14).

Analysis of Truncated gl15 Transcripts. RNA ligase mediated 5′ RACE was performed by using the GeneRacer Kit (Invitrogen) on RNA samples subjected to two rounds of poly(A) purification. RNAs were isolated from 30 shoot apices of H99 plants at 6 and 10 DAS and from H99:Gl15-TG7/0 plants at 14 DAS. We omitted the dephosphorylation and decapping steps, such that only 5′ ends of truncated transcripts could be ligated to the GeneRacer RNA oligo. Forward and nested forward GeneRacer primers complementary to the GeneRacer RNA oligo were used with either a reverse gene specific primer (gl15 gsp1, 5′-CCAACAGCGCTAGAGCAGCATCA) or the nested reverse gene specific primer (gl15 gsp2, 5′-CTGTCGCTAGCTCGCCCCTCTGCCACTGCTTT). Both of these gl15 gsp primers anneal 3′ to the target site for miR172. The cDNA populations were independently sampled 12 times for an initial round of PCR, followed by nested PCR. The nested products were pooled and cloned without size selection before sequencing 48 clones from each of the two experiments that produced RACE products.


Increasing gl15 Activity Delays Vegetative and Reproductive Phase Change. The phenotype of gl15 mutants demonstrates that gl15 is required for expression of juvenile traits in the leaf epidermis. To learn more about how gl15 operates to regulate leaf epidermal cell identity in the context of vegetative phase change, we generated transgenic maize lines that attempted to increase gl15 activity in its appropriate developmental context. Because we had no prior knowledge of whether gl15 regulatory sequences were restricted to sequences upstream of the transcribed region or might also reside within introns or the 3′ untranslated region, we introduced an ≈6-kbp genomic fragment that spanned the entire gl15 transcribed region as well as ≈1.5-kbp of 5′ and 1.1-kbp of 3′ flanking sequences (Fig. 1; ref. 8). Sequencing this fragment suggested that it is likely to contain the entire functional gl15 gene, because both the 5′ and 3′ termini were highly repetitive sequences, with the 3′ end sharing significant similarity to the pol gene within the Opie class of retrotransposons.

Two separate transformation experiments produced seven transgenic lines that contained from 3 to >10 Gl15 transgene (Gl15-TG) copies when analyzed by DNA gel blots (Fig. 1). Among these R0 lines, four (1, 2, 5, and 7) exhibited a phenotype suggesting increased GL15 activity, where additional leaves expressed juvenile leaf epidermal cell identity (e.g., leaf waxes and purple staining with toluidine blue) compared with the parental inbred line H99 (Fig. 1). Each R0 transgenic plant was outcrossed to H99, which maintained the transgenes in a common inbred genetic background but avoided potential deleterious phenotypic effects due to the fixation of potential recessive mutations caused by the transgene insertion or somaclonal variation.

Observation of epidermal leaf identity traits in the transgenic R1 progeny for each of the four lines showed that the number of leaves with juvenile waxes increased and the onset of macrohairs was delayed compared with H99 (Table 1). Although the four events differed in the extension of juvenile leaf identity (with Gl15-TG5 being the strongest), they each had a greater effect on juvenile compared with adult traits. None of the transgenic lines affected the expression of other traits associated with vegetative phase change, such as leaf shape or shoot axillary meristem identity. Thus, Gl15-TG functioned similarly to the endogenous gl15 gene, with specific effects on leaf epidermal cell differentiation that were opposite in polarity to those observed in loss-of-function gl15 mutants.

Table 1.
Quantitation of phase change phenotypes conditioned by Gl15-TG

We confirmed that Gl15-TG loci could complement gl15 gene function by crossing the Gl15-TG lines twice to a null gl15-H mutant stock that also carried the linked waxy1 (wx1) endosperm mutation. Homozygous wx1 kernels were selected, 90% of which were also expected to be homozygous for the gl15-H mutation. For each of the Gl15-TG1, Gl15-TG2, Gl15-TG5, and Gl15-TG7 events, approximately half of the progeny showed a gl15 mutant phenotype (three or fewer juvenile leaves) and the other half exhibited the Gl15-TG phenotype (five or more juvenile leaves). DNA gel blot analyses demonstrated that the increased number of juvenile leaves cosegregated with Gl15 transgenes (data not shown), indicating that the Gl15 transgenes were inherited as single dominant loci that showed no linkage with the wx1-gl15 chromosomal interval. Gl15-TG2 appeared to show only partial complementation of the gl15 mutant phenotype, suggesting it may restore GL15 activity to less than wild-type levels. Conversely, the greater number of juvenile leaves in Gl15-TG1, Gl15-TG5, and Gl15-TG7 compared with H99 indicates that GL15 activity from these transgene loci is more than that present in normal H99 plants.

Gl15-TG7 was selected for further genetic analysis because of its strong stable phenotype. Self-pollination of hemizygous Gl15-TG7/0 plants produced progeny homozygous for Gl15-TG7 in both the H99 and gl15-H backgrounds. Relative to hemizygous plants, homozygous Gl15-TG7 plants prolonged the expression of juvenile and delayed the expression of adult leaf identity. To investigate whether Gl15-TG7 could alter leaf identity in a hybrid with relevance to commercial maize production, homozygous Gl15-TG7 plants in the H99 inbred background were crossed to the B73 inbred. This hybrid produced two additional leaves with juvenile traits and one fewer leaf with adult traits relative to its nontransgenic genotype.

Each of the Gl15-TG lines also delayed the onset of reproductive development (Table 1), as measured by days from sowing to anthesis (50% of plants shedding pollen) and the total number of vegetative nodes. This phenotype was initially unexpected, because gl15 mutants have not previously been reported to affect reproductive maturity (6, 7). However, we believe the delayed flowering phenotype indeed reflects increased gl15 activity because it is observed in three independent Gl15-TG lines, the severity in delayed flowering is correlated with the effects of each Gl15-TG locus on prolonging juvenile leaf identity, and flowering time is further delayed in homozygous compared with hemizygous Gl15-TG7 plants.

To confirm gl15 overexpression in the Gl15-TG7 line, gl15 mRNA levels were measured by real-time PCR of reverse transcribed RNA from plants hemizygous or homozygous for Gl15-TG7 maintained in the H99 inbred background, as well as from H99 itself. At 12 days after sowing, which coincides with the initiation of the first adult leaves in the H99 genotype, gl15 mRNA levels were higher in the Gl15-TG7 shoot apices compared with H99 (Fig. 2a). Plants hemizygous for Gl15-TG7 exhibited an ≈3-fold increase and homozygous Gl15-TG7 plants a 6-fold increase in gl15 mRNA relative to H99, suggesting Gl15-TG7 conditions a dosage-dependent increase in gl15 mRNA expression. By 15 DAS and thereafter, gl15 mRNA expression in H99 plants decreased dramatically to near or below the limit of detection by this assay. gl15 mRNA continued to be expressed in Gl15-TG7 plants at 30 DAS, which coincides with the initiation of the male inflorescence (tassel) and the end of vegetative development. However, the apparent dosage effect of Gl15-TG7 on gl15 mRNA expression observed at 12 DAS was not evident later in shoot development, raising the possibility gl15 is subject to posttranscriptional regulation.

Maize miR172 Mediates gl15 mRNA Degradation. Two studies (9, 10) recently showed that the microRNA miR172 antagonizes the activity of AP2 and a small subfamily of AP2-like genes to regulate flowering time and floral organ identity in Arabidopsis. Aukerman and Sakai (9) suggested that the progressive accumulation of miR172 during shoot development eventually reduces AP2-like gene activity below a critical threshold needed to repress flowering. The target sequence for miR172 is also present in the gl15 transcript, suggesting that a maize homolog of miR172 may similarly down-regulate gl15 to promote adult vegetative development.

We examined this possibility by profiling the accumulation of small RNAs homologous to miR172 in maize shoots during vegetative development. RNA gel blot analysis by using a probe complementary to Arabidopsis miR172a identified a 21- to 25-nucleotide single-stranded RNA in maize shoots (Fig. 2b). Just as in Arabidopsis, maize miR172 is not detected early in shoot development but is present at 12 DAS and later stages. Control hybridizations with a maize miR166 probe (16) demonstrated that microRNAs were indeed present in the samples where miR172 was not detected (Fig. 4, which is published as supporting information on the PNAS web site) and that miR166 is expressed at similar levels in both juvenile and adult shoot apices. The onset of miR172 expression between 6 and 12 DAS coincides with the period when the identity of the first leaves expressing adult traits is specified (17), indicating that increased miR172 expression is associated with the transition to the adult vegetative phase. Importantly, miR172 accumulation is not affected by either increased (Gl15-TG7) or decreased (gl15-H mutant) gl15 activity (Fig. 2b), which is consistent with gl15 being a downstream molecular target of miR172.

The potential interaction between miR172 and gl15 mRNA was further investigated by assaying their relative expression between 7 and 14 DAS, the developmental window when newly initiated leaves transition from a juvenile to adult identity. Because leaf identity in maize is determined when leaves are between 0.3 and 3 mm in length (17), shoots were divided into apex (including leaf primordia <2 mm) and basal leaf samples. miR172 could be detected in the shoot apex, but not leaves, at 11 DAS (Fig. 3a). miR172 expression increased between 11 and 13 DAS and, by 14 DAS, could be detected in the basal 1 cm, but not the medial 1 cm, of a 3-cm leaf. We again performed control hybridizations with miR166, which showed miR166 is detected in all samples (Fig. 4). Conversely, gl15 mRNA expression could be detected at seven DAS in both leaf and apex samples but declined steadily thereafter (Fig. 3b). gl15 mRNA expression was higher in the shoot apex compared with leaf samples, consistent with the expectation that gl15 acts very early in leaf development to program juvenile identity.

A direct role for miR172 in reducing gl15 mRNA levels is demonstrated by the identification of gl15 cDNA fragments whose termini are consistent with products expected from miR172-directed mRNA cleavage (Fig. 3c). A majority (51 of 96) of gl15 transcript cleavage sites occurred at precisely the same nucleotide within the miR172 target site as previously observed for Arabidopsis AP2 and AP2-like transcripts (9). Forty-two of the gl15 transcript fragments terminated at positions 3′ to the putative miR172 cleavage site, likely representing rapid and progressive transcript degradation that has been observed for other mRNAs degraded by microRNAs (18). The remaining three RACE products terminated in close proximity but 5′ to the putative miR172 cleavage site, and could represent random breakage or aberrant directed cleavage of gl15 mRNA. Importantly, miR172-directed cleavage products were detected in both 10 DAS H99 and 14 DAS H99: Gl15-TG7/0 apex RNA samples (Fig. 3c) but not in RNA samples from six DAS H99 apices (Fig. 5, which is published as supporting information on the PNAS web site). Because gl15 transcript are not cleaved at 6 DAS and their abundance declines between 7 and 9 DAS (Fig. 3b), we suggest that miR172 is initially expressed at ≈7 DAS but remains below the limit of detection by our gel blot analysis until 11 DAS.


The results presented here indicate that the balance of activities between miR172 and the AP2-like gene gl15 is a major mechanism regulating vegetative phase change in maize. High GL15 activity during early shoot development maintains the juvenile phase, whereas the accumulation of miR172 between 10 and 12 DAS down-regulates GL15 to promote the transition to the adult phase. Such a mechanism is analogous to that proposed for the regulation of flowering time and floral organ identity in Arabidopsis (9, 10), where miR172 promotes the transition to reproductive development by reducing the activity of AP2-like genes below a critical threshold needed to repress flowering. In contrast to the regulation of reproductive phase change in Arabidopsis, where it is likely that the cumulative expression of multiple AP2-like genes contribute to repress flowering, only the gl15 gene appears to be necessary to maintain the juvenile vegetative phase in maize.

Analysis of both decreased and increased GL15 activity relative to wild-type indicates that gl15 functions as a primary juvenility gene in maize. Loss-of-function gl15 mutations demonstrate that gl15 is required for juvenile leaf identity. Conversely, increasing gl15 activity in the Gl15-TG lines (Fig. 2) showed that gl15 is not only sufficient to maintain the expression of juvenile leaf identity but can also delay reproductive maturation (Table 1), which is the other major defining feature of the juvenile phase. Previous characterizations of gl15 mutations did not identify any affects on flowering time (6, 7), but these prior studies only examined gl15 mutants in relatively early flowering backgrounds with a short juvenile phase (e.g., the inbred line W23). When the null gl15-m1:dSpm allele is introgressed into early flowering lines with a prolonged juvenile phase (e.g., the inbred line A188), flowering time is accelerated by two to four days in gl15 mutants compared with the parental genotype (S.P.M., unpublished data). We suggest that in genetic backgrounds such as A188 and the Gl15-TG lines, gl15 activity overlaps sufficiently with that of other AP2-like genes to contribute to the repression of flowering, whereas in lines like W23, there is little overlap between the activities of gl15 and other AP2-like genes. Candidates for these other AP2-like genes would be maize orthologs of the Arabidopsis TOE genes (9), or possibly indeterminate spikelet1 (ids1; ref. 17), which also contains a miR172 binding site.

Different expression patterns for functionally redundant AP2-like genes could also explain why only gl15 appears to regulate leaf identity in maize. However, given that the closely related ids gene is expressed in juvenile leaves but ids mutations have no effect on leaf identity (19), we favor the idea that individual AP2-like genes regulate different target genes, with gl15 specifically controlling those that determine phase-specific leaf epidermal cell differentiation. Even within the leaf epidermis, it appears that juvenile traits are more sensitive to gl15 activity compared with adult traits (6, 7), which is supported by the greater effect of each Gl15-TG line on the expression of juvenile waxes than adult macrohairs (Table 1).

The degradation of gl15 mRNA by miR172 (Fig. 3c) is likely to be an important downstream molecular effector of vegetative phase change in maize. The onset of miR172 accumulation (Figs. 2b and Fig. 3a) coincides with decreases in gl15 mRNA expression (Fig. 2a and Fig. 3b), and the developmental window where the identity of transition leaves is determined. Although not quantitative as performed, our analysis of miR172 expression in maize suggests that its expression may be more abundant during early leaf development compared with expanded leaves (Figs. (Figs.3a3a and 4), which is consistent with the finding that maize leaf identity is determined soon after leaf initiation (17). Further evidence that microRNAs may mediate vegetative phase change in maize is derived from the phenotypic and likely molecular similarities between the maize early phase change1 (epc1) and Arabidopsis HASTY (HST) genes. Both epc1 and hst mutations accelerate vegetative phase change (20, 21). HST encodes an exportin-5 protein that may mediate the transport of microRNA precursors (22) and epc1 probably encodes the maize ortholog of HAST Y (www.maizegdb.org/cgi-bin/displaylocusrecord.cgi?id=486968).

MicroRNAs have been shown to down-regulate their molecular targets through two apparently distinct mechanisms, translational repression or directed mRNA cleavage (23). In Arabidopsis, miR172 acts primarily as a translational repressor of AP2, although target mRNA cleavage products are also detected (9, 10). We show here that unlike the AP2 and TOE genes in Arabidopsis, gl15 transcripts rapidly decrease after the onset of miR172 accumulation, suggesting that miR172 down-regulates gl15 primarily through RNA cleavage. Schwab et al. (24) have recently shown that the observed reduction in AP2 protein but not mRNA levels in Arabidopsis could be due to AP2 negatively regulating its own expression, leading to increased AP2 transcription when miR172 is acting to reduce AP2 protein. These authors present a unified view of miR172 action in Arabidopsis, whereby the amount of miR172 relative to its target transcripts, the degree of complementarity between the microRNA and its target(s), and feedback regulatory mechanisms each contribute to whether transcript degradation or translational repression appears to predominate. We have informatically identified five putative maize miR172 genes, as well as at least seven AP2-like genes with highly conserved miR172 target sites in their 3′ untranslated regions (N.L. and S.P.M., unpublished results). Therefore, we anticipate that a similar level of complexity in miR172 action may operate in maize. For gl15, however, it appears that transcript degradation by miR172 is the favored mechanism.

The Gl15-TG lines indicate that gl15 can repress both vegetative and reproductive phase change in a dosage-dependent manner, but the precise mechanism(s) remain unclear. Gl15 does not directly antagonize miR172 accumulation, because miR172 expression was unaffected by either increases or decreases in gl15 (Fig. 2b). An increased abundance of GL15 protein in the Gl15-TG lines may extend the developmental period where overall AP2-like gene activity remains above the threshold required to repress phase change. Alternatively, the increased abundance of gl15 RNA in the Gl15-TG lines may competitively limit miR172 binding to other AP2-like genes that repress flowering (10), effectively increasing the activities of those genes.

Aukerman and Sakai (9) propose a threshold model where the opposing activities of AP2-like and miR172 regulate flowering in Arabidopsis. We have shown that a similar mechanism operates to control vegetative phase change in maize, where gl15 expression during early shoot development maintains the juvenile phase, and is down-regulated by miR172 to promote the adult phase. An attractive feature of this model is that it can account for the progressive nature of heteroblasty seen in many plant species and the intermediate phenotypes commonly observed in their transition leaves (2, 5). Consistent with this view, the discovery that miR172 inhibits AP2-like genes controlling organ identity and reproductive competence in both Arabidopsis (a eudicot) and maize (a monocot) suggests that this antagonism may be a general mechanism regulating phase change in plants.

The finding that the balance of activities between gl15 and miR172 regulates vegetative phase change also places other maize factors that alter this process within a regulatory context. The Corngrass1 (Cg1), Teopod1 (Tp1), and Teopod2 (Tp2) mutations prolong the juvenile phase (25, 26) and may do so by increasing the activity of AP2-like genes. Cg1 is already known to increase gl15 mRNA levels (8). The Cg1, Tp1, and Tp2 loci could either directly enhance the transcription of AP2-like genes, decrease miR172 accumulation, or reduce the sensitivity of AP2-like gene transcripts to microRNA-mediated repression, as observed for the maize Rolled1-O mutation (16). Conversely, the maize epc1 and viviparous8 genes, as well as gibberellic acids, promote the transition to adult development (21, 27, 28). Each of these factors act upstream of gl15 and could affect vegetative phase change by increasing miR172 expression. Gibberellic acids have already been shown to regulate miR159 abundance and flowering time in Arabidopsis (29) and, thus, could also conceivably affect miR172 levels.

Finally, we wish to point out here that reintroduction of genomic clones into a homozygous genotype can generate a dosage series of overexpression phenotypes in their appropriate developmental contexts, circumventing potential problems that may occur when driving transgenes from strong constitutive promoters. The Gl15-TG lines generated in this study essentially harbor duplications of the gl15 gene and, therefore, mimic one of the major mechanisms of evolutionary change in higher plants (30). Furthermore, transformation of an inbred line aids in the observation and interpretation of subtle phenotypic effects, such as those observed for the Gl15-TG lines, and permits the direct testing of transgene effects on hybrid performance (Table 1). These features illustrate the advantages of transforming inbred lines with genomic clones for the functional analysis of transgenes in maize.

Supplementary Material

Supporting Figures:


We thank a group of undergraduate students who helped with this project: Stephanie Halbig, Lisa Haney, Josh Inman, Nadia Shakoor, and Magen Starr, and Kris Lambert for sharing equipment and assistance with photomicroscopy. This research was supported by funding from the Illinois-Missouri Biotechnology Alliance. N.L. was supported by U.S. Department of Agriculture National Research Initiative Competitive Grants Program Postdoctoral Grant 2003-35304-13239.


Author contributions: N.L. and S.P.M. designed research; N.L., A.K., M.R.G., and S.P.M. performed research; S.R.C. contributed new reagents/analytic tools; N.L. and S.P.M. analyzed data; and N.L. and S.P.M. wrote the paper.

Abbreviation: DAS, days after sowing.

Data deposition: The gl15 genomic sequence reported in this paper has been deposited in the GenBank database (accession no. AY714877).


1. Lawson, E. J. R. & Poethig, R. S. (1995) Trends Genet. 11, 263-268. [PubMed]
2. Kerstetter, R. A. & Poethig, R. S. (1998) Annu. Rev. Cell Dev. Biol. 14, 373-398. [PubMed]
3. Goebel, K. (1900) Organography in Plants. Part I. General Organography (trans. Balfour, I. B.) (Clarendon, Oxford) (English).
4. Allsopp, A. (1967) Adv. Morphog. 6, 127-171. [PubMed]
5. Poethig, R. S. (2003) Science 301, 334-336. [PubMed]
6. Moose, S. P & Sisco, P. H. (1994) Plant Cell 6, 1343-1355. [PMC free article] [PubMed]
7. Evans, M. M., Passas, H. J. & Poethig, R. S. (1994) Development (Cambridge, U.K.) 120, 1971-1981. [PubMed]
8. Moose, S. P. & Sisco, P. H. (1996) Genes Dev. 10, 3018-3027. [PubMed]
9. Aukerman, M. & Sakai, H. (2003) Plant Cell 15, 2730-2741. [PMC free article] [PubMed]
10. Chen, X. (2004) Science 303, 2022-2025. [PubMed]
11. Hajdukiewicz, P., Svab, Z. & Maliga, P. (1994) Plant Mol. Biol. 25, 989-994. [PubMed]
12. Wan, Y., Widholm, J. M & Lemaux, P. G. (1995) Planta 196, 7-14.
13. Armstrong, C. L., Parker, G. B., Pershing, J. C., Brown, S. M., Sanders, P. R., Duncan, D., Stone, T., Dean, D. A., DeBoer, D. L., Hart, J., et al. (1995) Crop Sci. 35, 550-557.
14. Schneeberger, R., Tsiantis, M., Freeling, M. & Langdale, J. A. (1998) Development (Cambridge, U.K.) 125, 2857-2865. [PubMed]
15. Grozinger, C. M., Sharabash, N., Whitfield, C. & Robinson, G. (2003) Proc. Natl. Acad. Sci. USA 100, Suppl. 2, 14519-14525. [PMC free article] [PubMed]
16. Juarez, M. T., Kui, J. S., Thomas, J., Heller, B. A. & Timmermans, M. C. (2004) Nature 428, 84-88. [PubMed]
17. Orkwiszewski, J. A. J. & Poethig, R. S. (2000) Proc. Natl. Acad. Sci. USA 97, 10631-10636. [PMC free article] [PubMed]
18. Shen, B. & Goodman, H. M. (2004) Science 306, 997. [PubMed]
19. Chuck, G., Meeley, R. B. & Hake, S. (1998) Genes Dev. 12, 1145-1154. [PMC free article] [PubMed]
20. Telfer, A. & Poethig, R. S. (1998) Development (Cambridge, U.K.) 125, 1889-1898. [PubMed]
21. Vega, S. H., Sauer, M., Orkwiszewski, J. A. J. & Poethig, R. S. (2002) Plant Cell 14, 133-147. [PMC free article] [PubMed]
22. Bollman, K. M., Aukerman, M. J., Park, M. Y., Hunter, C., Berardini, T. Z. & Poethig, R. S. (2003) Development (Cambridge, U.K.) 130, 1493-1504. [PubMed]
23. Rhoades, M. W., Reinhart, B. J., Lim, L. P., Burge, C. B., Bartel, B. & Bartel, D. P. (2002) Cell 110, 513-520. [PubMed]
24. Schwab, R., Palatnik, J. F., Riester, M., Schommer, C., Schmid, M., Weigel, D. (2005) Dev. Cell 8, 1-11.
25. Galinat, W. C. (1966) Maize Genetics Cooperation News Letter 40, 102.
26. Poethig, R. S. (1988) Genetics 119, 959-973. [PMC free article] [PubMed]
27. Evans, M. M. S. & Poethig, R. S. (1995) Plant Physiol. 108, 475-487. [PMC free article] [PubMed]
28. Evans, M. M. S. & Poethig, R. S. (1997) Plant J. 12, 769-779.
29. Achard, P., Herr, A., Baulcombe, D. C. & Haberd, N. (2004) Development (Cambridge, U.K.) 131, 3357-3365. [PubMed]
30. Moore, R. C. & Purugganan, M. D. (2003) Proc. Natl. Acad. Sci. USA 100, 15682-15687. [PMC free article] [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Compound
    PubChem chemical compound records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records. Multiple substance records may contribute to the PubChem compound record.
  • Gene
    Gene records that cite the current articles. Citations in Gene are added manually by NCBI or imported from outside public resources.
  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence and PMC links.
  • MedGen
    Related information in MedGen
  • Nucleotide
    Primary database (GenBank) nucleotide records reported in the current articles as well as Reference Sequences (RefSeqs) that include the articles as references.
  • Protein
    Protein translation features of primary database (GenBank) nucleotide records reported in the current articles as well as Reference Sequences (RefSeqs) that include the articles as references.
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...