The tropical monocot Canna indica (Cannaceae, Zingiberales) has an androecium in which most stamen-derived organs develop as sterile petaloid staminodes and a labellum, while a single fertile stamen is reduced to one theca bearing a petaloid appendage. Fu et al. (2014) isolated the B-class GLOBOSA homolog CiGLO and the C-class AGAMOUS homolog CiAG and found both transcribed in the petaloid staminodes, the labellum, and the fertile stamen, so organ identity gene expression alone cannot explain why these organs diverge in morphology and fertility. A decade of follow-up work — temporal-spatial transcriptomes and in situ hybridization (localized detection of transcripts in tissue sections) in Canna, adaxial-abaxial polarity and CYC/TB1 gene analyses, comparative B-, C-, E-class, and YABBY gene-family studies across the Zingiberales, dosage evidence from petaloid-stamen mutants in other families, and a chromosome-scale C. indica genome assembly — has sharpened the competing hypotheses (expression timing, dosage, protein-protein interactions, post-transcriptional regulation, downstream polarity programs) without resolving them, and no functional perturbation of any B- or C-class gene has been performed in Canna itself. The open problem is to determine the molecular-genetic mechanism that converts broadly similar B+C identity-gene presence into sterile petaloid versus fertile stamen fates: a complete Canna GLO/AG paralogue (duplicate gene copy) inventory and its sequence evolution, stage- and tissue-resolved expression during early organogenesis, the regulatory network (E-class/SEP and other interacting regulators) specifying the distinct androecial identities, and functional discrimination among the candidate mechanisms.

Background

Floral organ identity is classically described by the ABC model (Coen and Meyerowitz 1991): in the model eudicots Arabidopsis and Antirrhinum, A-class function alone specifies sepals in the outermost whorl, A plus B specifies petals, B plus C specifies stamens, and C alone specifies carpels. The B and C functions are encoded by MADS-box genes — transcription factors named for their conserved DNA-binding domain — with B-class activity contributed by one APETALA3/DEFICIENS-lineage and one PISTILLATA/GLOBOSA-lineage protein (GLO is the Antirrhinum GLOBOSA homolog of Arabidopsis PISTILLATA), and C-class activity by AGAMOUS (AG); these act together with E-class SEPALLATA proteins as combinatorial transcription-factor complexes that activate organ-specific developmental programs.

Canna indica belongs to the Zingiberales, a tropical monocot order that also includes bananas (Musaceae), gingers (Zingiberaceae), and spiral gingers (Costaceae). Zingiberales flowers show a graded transformation of the androecium (the male floral whorl): basal families retain five or six fertile stamens, whereas in the derived ginger clade most or all androecial members become petaloid — flat, showy, petal-like structures called staminodes (sterile, stamen-derived organs). In Canna the androecium consists of three to four petaloid staminodes, a labellum (a lip-shaped structure interpreted as a staminode of the inner androecial whorl), and a single fertile stamen reduced to one theca (one half of a typical two-theca anther) bearing a prominent petaloid appendage; the style is also flattened and petaloid.

Two studies framed the question addressed here. Almeida, Brown and Specht (2013) showed by developmental series and RT-PCR (reverse-transcription PCR, which measures transcript abundance) that the petaloid appendage of the Canna fertile stamen arises from the filament/connective side of a half-stamen primordium, that B-class genes are expressed in all floral organs, and that two C-class AGAMOUS copies show expression expanded beyond the canonical pattern, concluding that the canonical ABC model is insufficient to explain androecial petaloidy. Fu et al. (2014) then isolated full-length CiGLO and CiAG and quantified their expression in dissected floral organs: both genes are transcribed in the petaloid staminodes, the labellum, the fertile stamen, and the carpels. By the ABC model, simultaneous B and C expression should specify stamen identity, so the petaloid staminodes and the fertile stamen carry indistinguishable identity-gene transcripts — yet the former are sterile and petaloid while the latter produces pollen in a single theca. Earlier, Bartlett and Specht (2010) had documented Zingiberales-specific GLOBOSA duplications (ZinGLO1-4) with divergent expression patterns, proposing that labellum identity reflects the combined activity of several GLO homologs. The unresolved question is thus not whether B- and C-class genes are present in the petaloid organs — they are — but what distinguishes the developmental fates of androecial organs that all express them.

Problem Statement

Determine the molecular-genetic mechanism — or combination of mechanisms — by which the Canna indica androecium produces sterile petaloid organs (petaloid staminodes and the labellum) alongside a fertile stamen, despite co-expression of the B-class GLOBOSA homolog CiGLO and the C-class AGAMOUS homolog CiAG in all of these organs (Fu et al. 2014). This single objective comprises jointly supporting parts: (1) the full inventory of GLO and AG family paralogues in Canna indica and their sequence evolution, since only single or partial copies had been isolated when the question was posed; (2) stage-specific and spatially resolved expression of these genes and their regulators across early floral primordia and organogenesis, because organ-level RT-PCR cannot resolve temporal dynamics or sub-organ localization; (3) the identity and roles of the additional regulators — E-class/SEP genes, other B-class paralogues, CYC/TB1-like symmetry genes (TCP-lineage transcription factors, named after the Antirrhinum CYCLOIDEA and maize TEOSINTE BRANCHED1 genes, that regulate floral symmetry and organ fate), polarity genes, and downstream effector genes — and the structure of the regulatory network specifying a sterile petaloid staminode versus a fertile stamen; and (4) which candidate mechanism — differences in expression timing or dosage, altered protein-protein interactions among the MADS-box proteins, post-transcriptional regulation, or downstream morphogenesis programs such as adaxial-abaxial polarity (the developmental contrast between the axis-facing, adaxial, and outward-facing, abaxial, sides of a lateral organ) — actually causes the divergent morphology. The question is answerable from the existing published evidence without new functional perturbation of Canna B- or C-class genes.

References

  1. Fu Q, Liu H, Almeida AMR, Kuang Y, Zou P, Liao J (2014). Molecular basis of floral petaloidy: insights from androecia of Canna indica. AoB PLANTS 6: plu015. doi:10.1093/aobpla/plu015, https://doi.org/10.1093/aobpla/plu015
  2. Coen ES, Meyerowitz EM (1991). The war of the whorls: genetic interactions controlling flower development. Nature 353: 31-37. doi:10.1038/353031a0, https://doi.org/10.1038/353031a0
  3. Almeida AM, Brown A, Specht CD (2013). Tracking the development of the petaloid fertile stamen in Canna indica: insights into the origin of androecial petaloidy in the Zingiberales. AoB PLANTS 5: plt009. doi:10.1093/aobpla/plt009, https://doi.org/10.1093/aobpla/plt009
  4. Bartlett ME, Specht CD (2010). Evidence for the involvement of GLOBOSA-like gene duplications and expression divergence in the evolution of floral morphology in the Zingiberales. New Phytologist 187: 521-541. doi:10.1111/j.1469-8137.2010.03279.x, https://doi.org/10.1111/j.1469-8137.2010.03279.x
  5. Specht CD, Yockteng R, Almeida AM, Kirchoff BK, Kress WJ (2012). Homoplasy, Pollination, and Emerging Complexity During the Evolution of Floral Development in the Tropical Gingers (Zingiberales). The Botanical Review 78: 440-462. doi:10.1007/s12229-012-9111-6, https://doi.org/10.1007/s12229-012-9111-6
  6. Yockteng R, Almeida AMR, Morioka K, Alvarez-Buylla ER, Specht CD (2013). Molecular Evolution and Patterns of Duplication in the SEP/AGL6-Like Lineage of the Zingiberales: A Proposed Mechanism for Floral Diversification. Molecular Biology and Evolution 30: 2401-2422. doi:10.1093/molbev/mst137, https://doi.org/10.1093/molbev/mst137
  7. Almeida AMR, Yockteng R, Otoni WC, Specht CD (2015). Positive selection on the K domain of the AGAMOUS protein in the Zingiberales suggests a mechanism for the evolution of androecial morphology. EvoDevo 6: 7. doi:10.1186/s13227-015-0002-x, https://doi.org/10.1186/s13227-015-0002-x
  8. Rocha de Almeida AM, Yockteng R, Specht CD (2015). Evolution of petaloidy in the Zingiberales: An assessment of the relationship between ultrastructure and gene expression patterns. Developmental Dynamics 244: 1121-1132. doi:10.1002/dvdy.24280, https://doi.org/10.1002/dvdy.24280
  9. Morioka K, Yockteng R, Almeida AM, Specht CD (2015). Loss of YABBY2-Like Gene Expression May Underlie the Evolution of the Laminar Style in Canna and Contribute to Floral Morphological Diversity in the Zingiberales. Frontiers in Plant Science 6: 1106. doi:10.3389/fpls.2015.01106, https://doi.org/10.3389/fpls.2015.01106
  10. Piñeyro-Nelson A, Rocha De Almeida AM, Sass C, Iles WJD, Specht CD (2017). Change of Fate and Staminodial Laminarity as Potential Agents of Floral Diversification in the Zingiberales. Journal of Experimental Zoology Part B: Molecular and Developmental Evolution 328: 41-54. doi:10.1002/jez.b.22724, https://doi.org/10.1002/jez.b.22724
  11. Tian X, Yu Q, Liu H, Liao J (2016). Temporal-Spatial Transcriptome Analyses Provide Insights into the Development of Petaloid Androecium in Canna indica. Frontiers in Plant Science 7: 1194. doi:10.3389/fpls.2016.01194, https://doi.org/10.3389/fpls.2016.01194
  12. Tian X, Zou P, Miao M, Ning Z, Liao J (2018). RNA-Seq analysis reveals the distinctive adaxial-abaxial polarity in the asymmetric one-theca stamen of Canna indica. Molecular Genetics and Genomics 293: 391-400. doi:10.1007/s00438-017-1392-3, https://doi.org/10.1007/s00438-017-1392-3
  13. Tian X, Li X, Yu Q, Zhao H, Song J, Liao J (2020). Irregular adaxial-abaxial polarity rearrangement contributes to the monosymmetric-to-asymmetric transformation of Canna indica stamen. AoB PLANTS 12(5): plaa051. doi:10.1093/aobpla/plaa051, https://doi.org/10.1093/aobpla/plaa051
  14. Yu Q, Tian X, Lin C, Specht CD, Liao J (2020). Expression and Function Studies of CYC/TB1-Like Genes in the Asymmetric Flower Canna (Cannaceae, Zingiberales). Frontiers in Plant Science 11: 580576. doi:10.3389/fpls.2020.580576, https://doi.org/10.3389/fpls.2020.580576
  15. Tian X, Li X, Yu Q, Zhao H, Liao J (2021). Asymmetric expression patterns of B- and C-class MADS-box genes correspond to the asymmetrically specified androecial identities of Canna indica. Plant Biology 23(4): 540-545. doi:10.1111/plb.13231, https://doi.org/10.1111/plb.13231
  16. Liu J, Li CQ, Dong Y, Yang X, Wang YZ (2018). Dosage imbalance of B- and C-class genes causes petaloid-stamen relating to F1 hybrid variation. BMC Plant Biology 18: 341. doi:10.1186/s12870-018-1562-4, https://doi.org/10.1186/s12870-018-1562-4
  17. Yu Q, Zhao T, Zhao H, Specht CD, Tian X, Liao J (2022). Correlation between Inflorescence Architecture and Floral Asymmetry—Evidence from Aberrant Flowers in Canna L. (Cannaceae). Plants 11(19): 2512. doi:10.3390/plants11192512, https://doi.org/10.3390/plants11192512
  18. Li X, Kuang Y, Ye Y, Chen Z, Zhang M (2022). Diverse function of the PISTILLATA, APETALA 3, and AGAMOUS-Like MADS-box genes involved in the floral development in Alpinia hainanensis (Zingiberaceae). Gene 839: 146732. doi:10.1016/j.gene.2022.146732, https://doi.org/10.1016/j.gene.2022.146732
  19. Liu M, Li C, Jiang T, Wang R, Wang Y, Zhang W, Pan X (2023). Chromosome-scale genome assembly provides insights into flower coloration mechanisms of Canna indica. International Journal of Biological Macromolecules 251: 126148. doi:10.1016/j.ijbiomac.2023.126148, https://doi.org/10.1016/j.ijbiomac.2023.126148
  20. Appleton AD, Kramer EM (2024). Diversifying floral organ identity. Current Opinion in Plant Biology 80: 102550. doi:10.1016/j.pbi.2024.102550, https://doi.org/10.1016/j.pbi.2024.102550
  21. Appleton AD, Kramer EM (2024). Genetic Architecture of Novel Floral Organs. International Journal of Plant Sciences 185(3): 211-217. doi:10.1086/729359, https://doi.org/10.1086/729359
  22. Chen Q, Zhou Y, Long L, Zhang L, Liao H (2025). Comparative analyses of morphology and temporal floral organ transcriptome provide insights into the development of staminodes in Globba racemosa (Zingiberaceae). Biochemical and Biophysical Research Communications 760: 151690. doi:10.1016/j.bbrc.2025.151690, https://doi.org/10.1016/j.bbrc.2025.151690
  23. Gao W, Zheng W, Bai J, Zhang W, Zhang H, Zhang J, Wu Z (2022). Transcriptome analysis in Alcea rosea L. and identification of critical genes involved in stamen petaloid. Scientia Horticulturae 293: 110732. doi:10.1016/j.scienta.2021.110732, https://doi.org/10.1016/j.scienta.2021.110732
  24. Luo Y, Li Y, Yin X, Deng W, Liao J, Pan Y, Jiang B, Yang H, Ding K, Jia Y (2024). Transcriptomics analyses reveal the key genes involved in stamen petaloid formation in Alcea rosea L. BMC Plant Biology 24(1): 551. doi:10.1186/s12870-024-05263-6, https://doi.org/10.1186/s12870-024-05263-6