Stem cells are crucial in morphogenesis in plants and animals. Much is known about the mechanisms that maintain stem cell fates or trigger their terminal differentiation. However, little is known about how developmental time impacts stem cell fates. Using Arabidopsis floral stem cells as a model, we show that stem cells can undergo precise temporal regulation governed by mechanisms that are distinct from, but integrated with, those that specify cell fates. We show that two microRNAs, miR172 and miR165/166, through targeting APETALA2 and type III homeodomain-leucine zipper (HD-Zip) genes, respectively, regulate the temporal program of floral stem cells. In particular, we reveal a role of the type III HD-Zip genes, previously known to specify lateral organ polarity, in stem cell termination. Both reduction in HD-Zip expression by over-expression of miR165/166 and mis-expression of HD-Zip genes by rendering them resistant to miR165/166 lead to prolonged floral stem cell activity, indicating that the expression of HD-Zip genes needs to be precisely controlled to achieve floral stem cell termination. We also show that both the ubiquitously expressed ARGONAUTE1 (AGO1) gene and its homolog AGO10, which exhibits highly restricted spatial expression patterns, are required to maintain the correct temporal program of floral stem cells. We provide evidence that AGO10, like AGO1, associates with miR172 and miR165/166 in vivo and exhibits "slicer" activity in vitro. Despite the common biological functions and similar biochemical activities, AGO1 and AGO10 exert different effects on miR165/166 in vivo. This work establishes a network of microRNAs and transcription factors governing the temporal program of floral stem cells and sheds light on the relationships among different AGO genes, which tend to exist in gene families in multicellular organisms.

Document Type


Publication Date


Notes/Citation Information

Published in PLOS Genetics, v. 7, issue. 3, e1001358.

© 2011 Ji et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Digital Object Identifier (DOI)


Figure_S1.tif (4035 kB)
Diagrams of AG, AGO10, and PHV genes and multiple sequence alignments of Arabidopsis AGO proteins. (A) Diagrams of AG, AGO10, and PHV. In ag-10, the replacement of the guanine at nucleotide 3618 (position 1 being the “A” in the ACG start codon) by adenine causes an E-to-K substitution. ago10-12 has a C-to-T transition at nucleotide 2991 (position 1 being the “A” in the ATG start codon), which results in an L-to-F substitution at amino acid 674 in the protein. ago10-13 has a G-to-A transition at nucleotide 833, which introduces a premature stop codon in the second exon. ago10-14 is a G-to-A mutation at nucleotide 3323 causing a D-to-N substitution. phv-5d contains a G-to-A transition at nucleotide 1410 (position 1 being the “A” in the ATG start codon), causing an G-to-D substitution. (B) Multiple sequence alignment of all Arabidopsis AGO proteins indicates that 674 L is conserved. (C) A diagram showing that the phv-5d mutation disrupts the binding site for miR165/166.

Figure_S2.tif (4801 kB)
The accumulation of miRNAs, ta-siRNAs and their target mRNAs in ago10 mutants. (A) Northern blotting to detect six miRNAs and three ta-siRNAs. The ago10-13 mutation resulted in slightly elevated levels of miR165/166 and slightly reduced levels of miR164. The other examined miRNAs were not obviously affected. The levels of siR255 (from TAS1), siR1511 (from TAS2), and 5D8(+) (from TAS3), were not affected by ago10 mutations. The U6 blots served as loading controls for the overlying small RNA blots. The numbers below the small RNA blots indicate the relative abundance of the small RNAs. (B) Realtime RT-PCR to determine the levels of miRNA- and siRNA-targeted mRNAs. Most of the examined mRNAs were not significantly different between ago10-13 and Ler inflorescences. A small elevation was observed for CUC1 mRNA targeted by miR164. Bars represent standard deviation of three technical replicates. Three biological replicates yielded similar results.

Figure_S3.tif (15073 kB)
Floral phenotypes of various genotypes. (A) An ago10-13 flower; ago10-13 flowers resembled wild-type flowers except for the narrower petals. (B) Siliques from ago10-13 plants. ago10-13 siliques were shorter and wider than those of wild-type plants. (C) An SEM image of an ago1-11 flower. (D) An ago1-11 ago10-13/+ inflorescence with obviously enlarged floral meristems (arrow). (E) A wus-1 flower. (F, G) hua1 hua2 ago10-12 and clv3-1 exhibited synergistic effects in terms of floral determinacy. (F) A hua1 hua2 ago10-12 clv3-1 flower with an enlarged gynoecium. (G) A hua1 hua2 ago10-12 clv3-1 silique with a massive amount of internal stigma tissue bursting out of the primary gynoecium. (H) Third whorl organs in hua1 hua2 ago10-12 flowers. Some had petaloid features (the two on the left) while others resembled stamens. (I-L) SEM images of anthers and anther epidermal cells in hua1 hua2 ago10-12 (I and J) and hua1 hua2 ago10-12 ap2-2 (K and L). (M) Siliques of ag-10 ap2-2 plants. (N) Siliques of ag-10 ago10-13 plants. Note the elongated gynophore (arrow). (O) Siliques of ag-10 ago10-13 ap2-2 plants. (P) Quantification of gynophore length in the two genotypes. (Q) Siliques of phb-1d/+ plants. (R) ag-10 ago10-13/+ phb-1d/+ siliques showing gynoecia enlargement and the presence of ectopic organs (indicated by the arrows). (S) Bulged siliques and ectopic floral organs, indicated by the arrows, were found in ag-10 phv-5d/+ plants. (T) An ag-10 amiR165/166 plant showing SAM defects reminiscent of ago10 mutants. Scale bar, 500 µm in (C), 300 µm in (I), (K), 50 µm in (J), (L), and 1 mm in the rest of the panels.

Figure_S4.tif (573 kB)
AGO10 does not bind miR390. Northern blots for miR166 and miR390 were performed for total RNAs from wild type (Ler; the three lanes on the right) and immunoprecipitated samples (the first two lanes from the left) using anti-GFP antibodies from Ler (a negative control) and the YFP-ZLL transgenic line. YFP-ZLL was associated with miR166 but not miR390 in vivo.

Figure_S5.tif (856 kB)
The expression of ZPR3 in wild type (Ler) and ago10-13 as determined by realtime RT-PCR.

Figure_S6.jpeg (116 kB)
A diagram of amiR165/166. A modified, partial pri-miR168 is highlighted in black. The amiR165 and amiR166 sequences (in green) were introduced into the pri-miR168 backbone through two rounds of PCR cloning using the SwaI and PmeI sites, respectively. The whole structure was inserted between a 2X35S promoter and a 35S terminator through the restriction sites HindIII and EcoRI.

Table_S1.pdf (57 kB)
Floral organ counts in various genotypes.

Table_S2.doc (105 kB)
Sequences of oligonucleotides used in this study.