Identification of the Teopod1, Teopod2, and Early Phase Change genes in maize

Abstract Teopod1 (Tp1), Teopod2 (Tp2), and Early phase change (Epc) have profound effects on the timing of vegetative phase change in maize. Gain-of-function mutations in Tp1 and Tp2 delay all known phase-specific vegetative traits, whereas loss-of-function mutations in Epc accelerate vegetative phase change and cause shoot abortion in some genetic backgrounds. Here, we show that Tp1 and Tp2 likely represent cis-acting mutations that cause the overexpression of Zma-miR156j and Zma-miR156h, respectively. Epc is the maize ortholog of HASTY, an Arabidopsis gene that stabilizes miRNAs and promotes their intercellular movement. Consistent with its pleiotropic phenotype and epistatic interaction with Tp1 and Tp2, epc reduces the levels of miR156 and several other miRNAs.


Introduction
Shoot development in plants occurs in discrete phases.Upon germination, the shoot enters a juvenile vegetative phase.In most, but not all, plants, the shoot then progresses to an adult vegetative phase before entering a reproductive phase in which it produces structures involved in sexual reproduction (Poethig 2013).This process is regulated by 2 families of miRNAs, miR156/157 and miR172, which target, respectively, squamosa promoter binding/ SBP-like (SBP/SPL) transcription factors and Apetala2-like (AP2-like) transcription factors (Wu and Poethig 2006;Chuck et al. 2007;Wu et al. 2009).miR156 regulates most phase-specific vegetative traits, whereas miR172 regulates phase-specific epidermal traits and flowering time (Lauter et al. 2005;Jung et al. 2007;Wu et al. 2009;Lian et al. 2021;O'Maoileidigh et al. 2021;Zhao et al. 2023).
Genetic analysis of vegetative phase change began with studies of 3 phenotypically similar dominant gain-of-function mutations in maize-Teopod1 (Tp1), Teopod2 (Tp2), and Corngrass/Teopod3 (Cg/ Tp3).These spontaneous mutations initially attracted interest because their phenotype strongly resembles teosinte, the ancestor of maize (Lindstrom 1925;Weatherwax 1929;Galinat 1954).This phenotype was later attributed to the prolonged expression of juvenile vegetative traits (Poethig 1988a), and subsequent analyses of Tp1, Tp2, and Cg/Tp3 provided key insights into the nature and regulation of vegetative phase change (Poethig 1988b;Dudley andPoethig 1991, 1993;Bassiri et al. 1992;Evans and Poethig 1995;Bongard-Pierce et al. 1996).The phenotype of Cg is attributable to an insertion of a transposable element that causes overexpression of a transcript encoding 2 copies of miR156, Zma-miR15b/c (Chuck et al. 2007); however, the basis for the phenotype of Tp1 and Tp2 is still unknown.Recessive mutations that accelerate the appearance of adult traits in maize have also been identified.glossy15 (gl15) causes the epidermis of juvenile leaves to resemble an adult epidermis (Evans et al. 1994;Moose and Sisco 1994) and is the result of a loss-of-function mutation in a AP2-like transcription factor regulated by miR172 (Moose and Sisco 1996).Consistent with this result, overexpression of miR172-which targets Gl15-produces an early phase change phenotype (Lauter et al. 2005).early phase change (epc) mutations have a pleiotropic phenotype that-depending on genetic background-ranges from seed lethality to the precocious expression of a wide range of adult vegetative traits (Vega et al. 2002).An analysis of natural variation in vegetative phase change in maize has also uncovered several QTLs that affect this transition (Foerster et al. 2015).
Here we show that Tp1 and Tp2 map close to and cause the overexpression of, respectively, Zma-miR156j and Zma-miR156h, strongly suggesting that they are mutations in these genes.We also demonstrate that Epc is the maize ortholog of the Arabidopsis gene, HASTY (HST).

Plant material
Stocks of Tp1 and Tp2 were originally obtained from the Maize Genetics Cooperation and propagated by crossing plants heterozygous for these mutations to various inbred lines.The origin and phenotypes of epc-W23, epc-1S2P, epc-nl4, epc-Mo, and epc-Mo are described in Vega et al. (2002).epc-3 and epc-4 were identified in a noncomplementation screen for EMS-induced alleles.For this purpose, pollen of B73 was treated with EMS (Neuffer and Coe 1978) and crossed onto epc-W23/B73 plants.Two plants with a precocious "glossy" phenotype were identified in the ∼5,000 progeny of this cross, and subsequent sequencing of the ZmHST gene in these plants revealed the presence of point mutations in this

Nucleic acid analysis
Genomic DNA was isolated from leaf tissue using the CTAB protocol (Murray and Thompson 1980).Twelve to 15 µg of genomic DNA was separated by electrophoresis in 0.8% agarose and then transferred to Hybond N+ membranes.Hybridizations were performed in Church and Gilbert (7% SDS) buffer (Church and Gilbert 1984) with 32-P labeled, random-primed probes.Filters were visualized using a phosphoimager (Molecular Dynamics).RNA was isolated using TRIzol (GibcoBRL), and poly(A) RNA was isolated using PolyATract (Promega).Five micrograms of poly(A)-enriched RNA was separated on 1.2% agarose gels with 3% formaldehyde and transferred to Hybond N+ membranes.cDNA was made from 50 µg of total RNA using the SuperScript System for cDNA Synthesis (Life Technologies).The genomic sequence of mutant alleles of ZmHST alleles was determined by sequencing PCR products generated with the primers listed in Supplementary Table 1.
Northern analysis of miRNAs was conducted using 30 µg of total RNA from leaf 7 run on a 15% denaturing polyacrylamide gel containing 8 M urea.RNA was transferred to Hybond N + membranes (Amersham) electrophoretically and hybridized with [γ− 32 P]-labeled probes.Oligonucleotide probes were labeled with T4 polynucleotide kinase (New England Biolabs).Hybridization was performed at 40 °C using ULTRAhyb-oligo hybridization buffer (Ambion) using probes complementary to mature miRNAs (https://www.mirbase.org).A probe to U6 snRNA (AGG GGC CAT GCT AAT CTT CTC) was used as a loading control.
The abundance of the primary transcripts of genes encoding miR156 was measured by qRT-PCR in mutant and wild-type progeny from the cross of Tp/W22×W22, using the primers listed in Supplementary Table 1.Total RNA was isolated from the middle of leaf 7 with TRIzol (Invitrogen), treated with Turbo DNA-Free Kit (Invitrogen) and reverse transcribed using the SuperScript III First-Strand Synthesis System (Invitrogen) with the Oligo(dT) 21 primer.qRT-PCR was performed using the primers listed in Supplementary Table 1 in a C1000 Touch Thermo Cycler (Bio-Rad).ACTIN was used as endogenous control.Each data point represents 3 biological replicates, each with 3 technical replicates.

Phylogenetic analysis
Protein sequences of exportin-like genes were collected from Arabidopsis thaliana genome and Zea mays B73 genome.The sequences were aligned with the MAFFT on XSEDE tool, and phylogeny was further determined using the RAxML-HPC on XSEDE tool on the CIPRES Gate Way (Miller et al. 2010).

The identity of Tp1 and Tp2
Based on the phenotypic and genetic similarity between Tp1, Tp2, and Cg, we predicted that Tp1 and Tp2 were likely mutations that caused the overexpression of miR156.Northern analysis of mutant plants and their wild-type siblings confirmed this prediction (Fig. 1a).To determine the basis for the effect of these mutations on miR156 expression, we examined mapping data we had generated for Tp1 and Tp2 (Bongard-Pierce et al. 1993).Three of the markers we used have since been sequenced, making it possible to compare the position of these sites to genes encoding miR156.Zma-miR156j is 666 kb distal to php20569(bhlh86), which is consistent with the location of Tp1.UMC1507 showed no recombination with Tp2 and is located only 170 kb distal to Zma-miR156h (Fig. 1b).To determine if Tp1 and Tp2 affect the expression of these miR156 genes, we compared the expression of the primary transcripts of Zma-miR156j and Zma-miR156h in the 7th leaf of mutant and wild-type siblings in families segregating Tp1 and Tp2 (Fig. 1c).As a control, we measured the abundance of the primary transcript of Zma-miR156g, which is located ∼15 Mb proximal to Zma-miR156j on chromosome 7. Zma-miR156j was approximately 3-fold more abundant in Tp1/+ plants than in their wild-type siblings, whereas Zma-miR156h was ∼2.75-fold more abundant in Tp2/+ than in wild-type plants.Tp1 had no effect on the expression of Zma-miR156h, and Tp2 had no effect on the expression of Zma-miR156j, and neither mutation affected the expression of Zma-miR156g.To determine if the elevated level of these transcripts is functionally significant, we measured the abundance of the mRNA of the SPL gene, Zm0001d49824.1,which has a target site for miR156.This gene was repressed by about 50% in Tp1/+ and by more than 80% in Tp2/+.These results are consistent with a recent study of Tp2 (Li et al. 2023) and strongly suggest that Tp1 and Tp2 are cis-acting mutations that increase the expression of, respectively, Zma-miR156j and Zma-miR156h.

Identifying ZmHST
In the W23 background, epc-w23 nearly completely eliminates the juvenile phase (Vega et al. 2002) (Fig. 2).This phenotype is similar to that of hst mutants in Arabidopsis (Bollman et al. 2003), which suggested that Epc might be the ortholog of HST.Phylogenetic analysis of exportin proteins in Arabidopsis (Merkle 2011) and maize revealed that HST is most closely related to the maize gene Zm00001d009270 (Fig. 3), hereafter referred to as ZmHST.ZmHST contains 22 exons and spans 57 kb.Its relatively large size is attributable to 4 large introns: intron 2, which is 20 kb in size, and introns 15, 18, and 19, which are all around 8 kb (Fig. 4).The majority of intron 2 consists of 3 retrotransposons inserted sequentially within one another.Introns 15, 18, and 19 also consist primarily of sequences related to various types of transposable elements.
To determine if Epc corresponds ZmHST, we genotyped the F2 progeny derived from the self-pollination of an Oh43/epc-W23 plant using a probe that recognizes an EcoRV polymorphism near the 5′ end of ZmHST.We observed no recombination between Fig. 3. Phylogenetic tree of exportin proteins in maize (Zm) and Arabidopsis (At).This tree was generated using the maximum likelihood method.Bootstrap values from 1,000 replications are shown at the branch points.
Identification of the Teopod1, Teopod2, and early phase change genes in maize | 3 this polymorphism and epc-W23 in 32 mutant plants, making ZmHST an excellent candidate for Epc.We then sequenced transcripts of 4 previously described mutant alleles of Epc (epc-Mu, epc-Mo, epc-W23, and epc-1s2p) (Vega et al. 2002) (Fig. 5a).Two splice forms of Epc were identified in epc-Mu.Both contained a premature stop codon: 1 lacked exon 2, while the second was derived from a cryptic splice site in intron 1 and contained a portion of intron 1 and exon 2. Sequencing of genomic DNA from the 5′end of epc-Mu revealed a 16-bp deletion at the border of intron 1 and exon 2, explaining the splicing defects produced by this mutation.This deletion was absent in 15 wild-type progeny from the family in which epc-Mu arose, indicating that it is responsible for the mutant phenotype of epc-Mu.The epc-Mo cDNA lacked exon 20, and sequence analysis of the corresponding genomic region showed that this exon is deleted in epc-Mo and replaced with a non-LTR transposon.The cDNA sequences of epc-1s2p and epc-W23 were identical to that of B73.However, Southern blots of these alleles revealed the presence of a ca.1.5-kb insertion at the 5′ end of these alleles that is also present in epc-nl4 and in the W23 inbred line in which epc-W23 was identified (Fig. 5b).Sequencing of the corresponding genomic region from W23, epc-W23, epc-1s2p, and nl4 demonstrated that these mutations contain an identical insertion in intron 1 consisting of a Loner retrotransposon and a duplication of a portion of exon 1 (Fig. 5a).This polymorphism is fixed in the wildtype stock of W23 that we obtained from Ed Coe, Jr, suggesting that epc-1s2p and nl4 reflect the use of this W23 stock (or derivative lines) in the research programs of the investigators who identified these mutations.Two additional alleles of Epc-epc-4 and epc-5were identified in a noncomplementation screen using epc-W23/ B72 as a female parent (see Materials and Methods) (Fig. 5a).The 2 alleles identified in this screen have a phenotype that is similar to that the previously described epc alleles (Supplementary Fig. 1).
We used Northern analysis to determine if ZmHST expression is altered in the mutants described above.mRNA from 3-leaf seedlings from both W23 and mutant alleles introgressed into W23 was hybridized with the ZmHST cDNA (Fig. 5c).All 4 of the mutants we examined had reduced levels of ZmHST mRNA.The observation that the ZmHST transcript is less abundant in epc-W23 and epc-1s2p than in W23-despite the fact that W23 has the same intronic insertion as these mutants-could either mean that this Loner retrotransposon causes sporadic silencing of this locus (producing the epc-W23 and epc-1s2p alleles) or that the causative polymorphism in these alleles lies in a regulatory sequence outside of the genomic region examined in our experiments.
In Arabidopsis, hst mutations decrease the abundance of many different miRNAs (Park et al. 2005).To determine if this is also true for epc, we examined the effect of epc-W23 on the abundance of several miRNAs by Northern analysis (Fig. 6).Four of the 6 miRNAs we examined were present at reduced levels in epc-W23 compared to W23.This result is consistent with the morphological similarity between epc and hst and supports the conclusion that these genes have similar functions.

Discussion
Genetic analysis of the mechanism of vegetative phase change began with the characterization of 3 dominant gain-of-function mutations-Tp1 and Tp2 and Cg/Tp3-that prolong the expression of juvenile vegetative traits (Poethig 1988a).We subsequently showed that the phenotype of Tp1 and Tp2 is partially suppressed by epc-W23 (Vega et al. 2002), suggesting that these 3 genes are functionally related and that Epc acts either in parallel or downstream of Tp1 and Tp2.Here we show that Tp1 and Tp2 cause the overexpression of, respectively, Zma-miR156j and Zma-miR156h and that Epc is the maize ortholog of HST, an Arabidopsis gene required for the stability and/or processing of miRNAs and their intercellular movement (Park et al. 2005;Zhu et al. 2019;Brioudes et al. 2021;Cambiagno et al. 2021).These results provide a reasonable explanation for the phenotypes and the genetic interaction between these genes.
In our original analysis of Tp1 and Tp2, we used dosage analysis to determine if these dominant mutations were haplo-insufficient or gain-of-function (Poethig 1988a).For this purpose, we created plants of the genotypes, Tp/−, Tp/+, and Tp/+/+, using B-jA translocations.This experiment revealed that both mutations were gain-of-function but produced different predictions for the specific nature of the gain-of-function allele in each case.We and others (Li et al. 2023) have now shown that both mutations cause overexpression of genes encoding miR156, which raises the question of why this experiment produced different results for Tp1 and Tp2.One possible explanation is that the A segments on the TB-7L  and TB-10L translocations contain SPL/SBP genes targeted by miR156 as well as the wild-type alleles of Tp1(Zma-miR156j) and Tp2(Zma-miR156h).Specifically, tsh4 (Zm00001eb316740) and ZmSBP29 (Zm00001eb322280) are distal to Tp1 on chromosome 7, and ZmSBP21 (Zm00001eb429730) is distal to Tp2 on chromosome 10 ( Mao et al. 2016).All 3 of these SPL/SBP genes are in the same family as AtSPL9 and AtSPL15, which have major but slightly different effects on vegetative phase change in Arabidopsis (Xu et al. 2016).As plants with duplications or deficiencies of 7L or 10L have altered doses of both miR156 and genes repressed by miR156, it is not surprising that the phenotypes of the Tp/−, Tp/ +, and Tp/+/+ dosage series were not entirely consistent with the predicted effect of these genotypes on miR156 expression.
The mutant phenotype of epc (Vega et al. 2002) is similar to that of hst in Arabidopsis (Telfer and Poethig 1998) and mutations in the rice ortholog of HST, CRD1 (Zhu et al. 2019).Like epc, hst mutations reduce the number of juvenile leaves without dramatically affecting the number of adult leaves, cause leaf curling, and reduce the growth of adventitious roots.Although the effect of crd1 on shoot development has not been described, this mutation is similar to epc and hst in reducing adventitious (crown) root development (Zhu et al. 2019).In this regard, it is interesting that ZmHST/EPC maps within 1 cM of the maize QTLs for adventitious (brace) root development (Hostetler et al. 2021), which suggests that variation in the expression of EPC may contribute to natural variation in this trait.
One of the interesting features of epc is its variable expressivity.This is particularly evident in the case of epc-W23.In the W23 background in which it arose, the phenotype of plants homozygous for epc-W23 can range from nearly wild type to a very strong early vegetative phase change phenotype (Vega et al. 2002).However, in an Oh43 or A632 background, epc-W23 seeds fail to germinate or display shoot abortion immediately following germination.The variable phenotype of epc-W23 in different genetic backgrounds could reflect natural variation in the expression level of the miRNAs regulated by Epc, differences in the expression of genes regulated by these miRNAs, or the presence of modifiers of these genes.The variable expressivity of epc-W23 in a W23 genetic background is more difficult to explain.Northern analysis indicates that epc-W23 is not completely null, and it may be that the processes for which it is required are hypersensitive to small variation in its activity.In Arabidopsis, small changes in the level of miR156 can have dramatic effects on the expression of its targets when miR156 is present at low levels (He et al. 2018).epc-W23 reduces the abundance of miR156, and stochastic or environmentally induced variation in the remaining miR156 transcripts could explain the variable expressivity of epc-W23 in a W23 background.A more interesting possibility is that the Loner retrotransposon in epc-W23 causes stochastic, but reversible, silencing of this gene.
The evidence that mutations in Epc reduce the abundance of many miRNAs likely explains the pleiotropic phenotype of these mutations but also complicates efforts to determine the genes most directly responsible for the different components of this phenotype.The effect of Epc on vegetative phase change is likely attributable to its effect on miR156 expression, as epc corrects the delayed phase change phenotype of Tp1 and Tp2 (Vega et al. 2002), which cause the overexpression of miR156.The basis for the effect of epc on other aspects of maize development is a subject for future research.

Fig. 1 .
Fig. 1.Tp1 and Tp2 are miR156 genes.a) Northern blot of miR156 levels in Tp1/+ and Tp2/+ and their wild-type siblings.b) Recombination (left line) and physical (right line) maps of the regions on chromosome 7 and chromosome 10 containing Tp1 and Tp2.Three of the markers used for recombination mapping are located on the physical map and reveal that Tp1 and Tp2 are close to genes encoding miR156.c) qRT-PCR analysis of the abundance of the precursors of Zma-miR156j, Zma-miR156h, and Zma-miR156g and an SPL gene with a miR156 binding site (Zm00001d049824.1) in Tp1/+, Tp2/+, and their wild-type siblings.

Fig. 2 .
Fig. 2. The phenotype of epc-W23.a) The shoot and b) first leaf of wild-type W23 and epc-w23.

Fig. 4 .
Fig. 4. Genomic structure of ZmHST.Intron-exon organization of ZmHST and the identity of the transposable elements in introns.

Fig. 5 .
Fig. 5. Epc is ZmHST.a) Genomic structure of epc alleles.b) Southern blot of HinDIII digested genomic DNA from B73, W23, and several epc alleles, probed with the ZmHST cDNA.c) Northern blot of mRNA from W23 and epc mutant seedlings probed with the ZmHST cDNA.Ubiquitin was used as a loading control.