<scp><i>STERILE APETALA</i></scp> regulates nectar volume and nectary size in <i>Mimulus verbenaceus</i>

Nectaries are specialized plant glands that secrete nectar, a mixture of water, sugars, amino acids, and various other compounds. This secretion provides selective advantages by attracting pollinators and deterring herbivorous insects (Nicolson et al., 2007; Brandenburg et al., 2009; Liao et al., 2021). Variation in nectary-related floral traits may drive pollinator-dependent reproductive isolation (Bradshaw et al., 1995; Katzer et al., 2019; Vandelook et al., 2019). Beyond facilitating plant–animal interactions and speciation, nectar contributes vital ecosystem services, sustaining insect biodiversity crucial for modern agricultural productivity (Roy et al., 2017; Slavković et al., 2021). Despite the ecological, evolutionary, and agricultural importance of nectar, its molecular regulation remains poorly understood (Brandenburg et al., 2009; Roy et al., 2017; Liao et al., 2021; Slavković et al., 2021). Nectary-associated traits are inherently complex and polygenic (Bradshaw et al., 1998; Galliot et al., 2006). Investigating the genetic basis of these traits has been challenging due to limited mapping resolution (Barstow et al., 2022), a lack of functional genomic tools (Lambert et al., 2022), and significant gene–environment interactions (Mitchell, 2004; Leiss & Klinkhamer, 2005; Kaczorowski et al., 2008). Furthermore, phylogenetic analyses indicate that nectaries are evolutionarily labile, having independently evolved multiple times across different plant lineages (Bernardello, 2007). This suggests that divergent genetic mechanisms may be at play in different lineages, further complicating the study of nectary-associated traits. Despite these challenges, genes involved in nectary development, such as CRABS CLAW (CRC), which patterns both ovary and nectary development in Arabidopsis thaliana, appear to play a conserved role, particularly within core eudicots (Bowman & Smyth, 1999; Lee et al., 2005a,b; Morel et al., 2018; Sinjushin, 2022). However, this conservation does not necessarily extend to basal angiosperms (Min et al., 2019). Several other genes have been shown to regulate nectary development, such as the C2H2 zinc-finger (Cys2-His2) transcription factor POPOVICH in Aquilegia coerulea (Ballerini et al., 2020), NEC1 in Petunia hybrida (Ge et al., 2000), MYB305 in ornamental tobacco (Nicotiana langsdorffii × N. sanderae) (Liu et al., 2009), and GoNe (nectary development gene in Gossypium) in Gossypium (Pei et al., 2021). By contrast, genes including CELL WALL INVERTASE (Ruhlmann et al., 2010; Minami et al., 2021), GIBBERELLIN 2-OXIDASE6 (Wiesen et al., 2016), and PIN-FORMED6 from cruciferous plants (Bender et al., 2013) have been implicated in nectar production. However, how these genes interact within a shared regulatory network, or whether they act in parallel or convergent pathways to control nectary development and nectar secretion, remains largely unknown. To identify the molecular networks underlying nectary development, we employed Mimulus verbenaceus (Erythranthe verbenacea) (Phrymaceae), a hummingbird-pollinated species as our model system. Through ethyl methanesulfonate (EMS) mutagenesis, we identified three allelic mutants exhibiting reduced floral organ size, decreased nectar production, and smaller nectaries. Sequencing analysis revealed that these mutants carry mutations in the ortholog of A. thaliana STERILE APETALA (SAP). Interestingly, the SAP gene appears to be associated with the evolutionary transition from hummingbird-pollinated Mimulus cardinalis to self-pollinating Mimulus parishii. Furthermore, these mutations also led to the downregulation of MvCRC1 and MvCRC2, which are critical for nectary development. Our study establishes a genetic framework for nectary development in Mimulus and provides the foundation for understanding the genetic basis of pollination syndrome transitions. Wild-type (WT) M. verbenaceus (MvBL) flowers produce abundant nectar and are primarily pollinated by hummingbirds, although they retain the ability to self-pollinate when pollinators are scarce (Vickery, 2008). To investigate the genetic basis underlying nectar production, we conducted a forward genetics screen on EMS-mutagenized MvBL plants exhibiting reduced nectar volume. We identified three morphologically similar mutants: MV00081, MV00017, and NJ000356-1, each exhibiting reduced floral organ size (Fig. 1a). More importantly, nectar volumes in these mutants were significantly lower than MvBL, with MvBL flowers producing 53.45 ± 7.59 μl (mean ± SD) of nectar, while the mutants produced 14.55 ± 2.50 μl (MV00081), 13.50 ± 1.69 μl (MV00017), and 25.50 ± 4.04 μl (NJ00356-1) (Fig. 1b). Nectar volume is often inversely proportional to its concentration, with exceptions reported in hummingbird-pollinated species, and its nutritional value is primarily derived from three simple sugars: sucrose, glucose, and fructose (Nicolson et al., 2007; Nicolson, 2022). To assess whether mutation impacts nectar sugar composition, we used high-performance liquid chromatography to compare the concentrations of sucrose, glucose, and fructose between the MvBL and the MV00081 mutant, which was selected for analysis due to its pronounced floral defects and markedly reduced nectar volume. Intriguingly, sucrose levels were comparable. However, glucose and fructose showed significant differences (Table 1), suggesting mutation differentially influences the production of these sugars. To determine whether the three morphologically similar mutants resulted from mutations in the same gene, we conducted a complementation cross among all three mutants. The F1 progeny failed to rescue the mutant floral phenotypes, indicating allelism (Supporting Information Fig. S1). We then focused on the more severe MV00081 mutant. Segregation ratio analysis of the F2 progeny resulting from a cross between the MvBL and the MV00081 mutant conformed to the expected 3 : 1 ratio (324 : 106; chi-square = 0.028; P = 0.87), suggesting that the mutation in MV00081 is governed by a single recessive locus. We further characterized its leaf and floral morphology. This mutant exhibited significantly reduced leaf length and width, as well as delayed flowering time (Fig. S2). Additionally, petal and corolla widths were markedly diminished (Fig. S3a,b), and the lengths of the calyx, corolla, stamen, pistil, and ovary were reduced (Fig. S3c). To determine whether the reduction in floral organ size was due to changes in cell division or cell expansion, we measured cell number and width in the ventral petals of both MvBL and MV00081 mutant plants. The MvBL ventral petals contained 292 ± 10 cells, while the MV00081 mutant had 231 ± 6 cells (Fig. S3d–h), indicating that the MV00081 mutation primarily affects cell division. The function of nectaries is evolutionarily conserved, conferring selective advantages to plants, although their position and morphology vary widely among plant species (Pacini et al., 2003; Nicolson et al., 2007). For instance, in A. thaliana, the nectary develops at the base of the stamens (Baum et al., 2001), while in A. coerulea, it originates from petal spurs (Min et al., 2019; Zhang et al., 2020), and in Mimulus, it initiates from the base of the pistil (Hazle & Canne-Hilliker, 2005). To investigate how the MV00081 mutation affects nectary development, we performed scanning electron microscopy analysis on both MvBL and mutant nectaries. In MvBL, the nectary is located below the ovary. As the style initiates, a bulge forms at the base of the abaxial side of the pistil, with a few epidermal cells differentiating into a spherical protrusion (Stage 1, 2 in Fig. 1c, Fig. S5a,b). This protrusion then separates from the ovary as the region between the ovary and gland begins to indent (Stage 3 in Fig. 1c, Fig. S5c). Modified stomata develop on the surface of the nectary glands, and their number increases. Eventually, a triangular protrusion forms at the front of the glands, completing nectary morphogenesis (Stage 4, 5 in Fig. S5d–h). Although the nectary developmental trajectory in the MV00081 mutant resembles that of MvBL, the mutant nectaries are significantly smaller (Fig. S4a–c) and exhibit no other obvious morphological defects (Figs 1d, S4a). Notably, the nectary-to-ovary area ratio remains unchanged between the mutant and WT (Fig. S4d), suggesting that the smaller nectary size in all three allelic mutants results from a reduced growth rate. These observations imply that MV00081 disrupts nectary gland development via a cellular mechanism analogous to its role in floral petals – likely by suppressing cell division (Fig. S4b–d). Cyclin D1 and cyclin D3 are well-established markers of cell division in plants (Gaudin et al., 2000). Analysis of cell division-related gene expression in RNA samples extracted from pistils dissected from 2-mm flower buds revealed significant downregulation in the mutant compared with MvBL, aligning with our hypothesis (Fig. S6). To identify the gene responsible for the MV00081 mutant phenotypes, we carried out a bulk segregant analysis, which revealed 12 single nucleotide polymorphisms in protein-coding regions (Table S1). Among these, we identified a single point mutation in the ortholog of A. thaliana STERILE APETALA (SAP) (Byzova et al., 1999), which introduces a premature stop codon and is likely to disrupt the function of the SAP protein (Figs 1e, S7a, S8). SAP was initially identified in A. thaliana (Byzova et al., 1999), in which its loss-of-function results in severe male and female sterility. It is an F-box protein functioning within a SKP1/Cullin/F-box E3 ubiquitin ligase complex and regulates organ size by mediating the proteasome-dependent degradation of substrate proteins, thus promoting the proliferation of meristemoid cells (Sicard et al., 2016; Wang et al., 2016; Yordanov et al., 2017; Li et al., 2018; Yang et al., 2018; Liu et al., 2019; Huang et al., 2020; Yin et al., 2020; Azeez & Busov, 2021; Zhou et al., 2021). Further Sanger sequencing of the SAP coding sequences from the MV00017 and NJ00356-1 mutants revealed independent mutations. The MV00017 mutant carries two nonsynonymous substitutions: alanine-to-valine (A213V) and proline-to-leucine (P289L) in the WD40 domain, respectively (Figs 1e, S7b, S8). The NJ00356-1 mutant contains a glycine-to-aspartate (G239D) substitution in the same domain, underscoring its functional significance (Figs 1e, S7c, S8). Collectively, these findings support SAP as the gene responsible for the observed mutant phenotypes. Consequently, we renamed the mutants MV00081, MV00017, and NJ00356-1 to sap-1, sap-2, and sap-3, respectively. Floral traits are integrated to optimize pollination, with distantly related plant species exhibiting similar floral phenotypes, known as pollination syndromes. Transitions between pollination syndromes are frequent and rapid, involving the correlated evolution of multiple traits to maintain phenotypic integration. Strong selection for specific trait combinations promotes linkage disequilibrium among loci – either unlinked or tightly linked – and involves pleiotropy, enabling coordinated trait changes over short evolutionary timescales (Wessinger & Hileman, 2016). Loss of function in SAP, which simultaneously reduces multiple floral traits, prompted us to investigate its role in transitions between pollination syndromes. A recent study identified four quantitative trait loci (QTLs) associated with nectar volume and floral dimensional traits between the self-pollinating M. parishii and the hummingbird-pollinated M. cardinalis (Chen et al., 2024), a close relative of M. verbenaceus, which is also hummingbird-pollinated and displays similar floral morphology. Remarkably, SAP was localized within one of these major QTLs, suggesting its potential involvement in the shift between pollination syndromes (Fig. S9). In A. thaliana, CRC is essential for initiating nectary development (Bowman & Smyth, 1999). CRC activation involves both general floral homeotic genes and specific transcription factors (Lee et al., 2005a; Reeves et al., 2012; Wuest et al., 2012). Lee et al. (2005b) demonstrated that CRC-mediated nectary development is highly conserved across core eudicot species. The M. verbenaceus genome contains two CRC homologs: MvCRC1 and MvCRC2 (Fig. S10). Therefore, we investigated whether reduced nectar volume in sap mutants correlates with altered expression of MvCRC1 and MvCRC2. We first compared the two genes' expression levels in pistils dissected from 2-mm floral buds between the MvBL and sap-1 mutant using quantitative reverse transcription polymerase chain reaction (RT-qPCR). Their expression was significantly downregulated in the sap-1 mutant compared with that in MvBL (Fig. S11). Furthermore, RT-qPCR analysis performed on S2-stage nectaries microdissected from MvBL and sap-1 mutants revealed significant downregulation of both genes (Figs 2a, S12). To further elucidate the regulatory relationship between MvSAP and MvCRC1 and MvCRC2, we examined the spatiotemporal expression patterns of these genes in MvBL floral buds at various developmental stages using in situ hybridization with the S phase-associated gene histone H4 (MvH4) serving as a positive control. MvSAP is ubiquitously expressed in all floral organs during early floral development. As the nectary differentiates from the base of the pistil, MvSAP exhibits a diffuse expression pattern within the developing nectary. Due to the low endogenous expression of MvSAP in the sap mutant, we were unable to reliably resolve its spatiotemporal expression pattern using conventional in situ hybridization. By contrast, MvCRC1 and MvCRC2 are expressed early and initially localize to the pistil walls; however, their expression becomes increasingly prominent in the differentiating nectary as development progresses (Fig. 2b). This spatiotemporal pattern mirrors observations in P. hybrida (Lee et al., 2005b), in which CRC orthologs also accumulate in maturing nectaries. Notably, MvCRC1 expression is higher than that of MvCRC2, suggesting a more significant role for MvCRC1 in nectary development in Mimulus. Interestingly, the spatial expression patterns of MvSAP, MvCRC1, and MvCRC2 show only partial overlap within the nectary, implying that MvSAP may function in regulating other developmental processes beyond those shared with MvCRC1/2. Given that MvSAP lacks a DNA-binding domain, future studies should focus on identifying its interacting partner to elucidate the direct regulatory mechanisms governing MvCRC1 and MvCRC2 in Mimulus. Transgenic studies will be essential to further validate the function of MvSAP. It remains to be determined whether the SAP-CRC regulatory program is involved in nectary development across eudicots and whether it is similarly utilized in monocots or other noncore eudicot lineages. In conclusion, we identified the SAP gene as a key regulator of nectary development in M. verbenaceus, demonstrating that mutations in SAP lead to reduced nectary size and nectar production likely through the downregulation of CRC genes. Our findings highlight the importance of SAP and CRC in nectary formation and have significant implications for understanding plant–pollinator interactions and the evolution and development of floral traits. We apologize to colleagues whose work could not be fully cited owing to space constraints. We thank Dr. Yao-Wu Yuan (University of Connecticut) for providing the MV00081 and MV00017 mutant seeds. This work was financially supported by grants from the National Natural Science Foundation of China to B.D. (32122078) and the Fundamental Research Funds for the Central Universities (YDZX2025035, QTPY2025005). This project was supported by the Bioinformatics Center of Nanjing Agricultural University. We are grateful to Dr. Yuehua Ma (Central Laboratory of the College of Horticulture, Nanjing Agricultural University) for assistance in using the Scanning Electron Microscope FlexSEM 1000II (Hitachi). None declared. LZ, YH and BD designed the research. LZ, YH, YG, PZ, YS, DZ, JH, LY, YK, RW, JW and WW performed experiments and data analysis. FC and SC provided insightful advice, discussions and support throughout. LZ and BD wrote the manuscript with input from all authors. LZ and YH contributed equally to this work. The short read data are deposited in NCBI SRA (accession no.: PRJNA1130631). Phenotypic measurement data are available at Dryad (DOI: 10.5061/dryad.3ffbg79vf). The data that support the findings of this study are available in the article and in the Supporting Information. Fig. S1 Complementation crosses indicate MV00017 and NJ00356-1 are allelic to MV00081. Fig. S2 Vegetative and flowering phenotypes of the MvBL and the MV00081 mutant. Fig. S3 Measurements of the floral organs at the tissue and cellular levels. Fig. S4 Quantitative comparison of nectary and ovary sizes between the MvBL and the MV00081 mutant. Fig. S5 Scanning electron micrographs of nectary development in MvBL. Fig. S6 Relative transcript level of Cyclin in MvBL and MV00081. Fig. S7 Sanger sequencing chromatograms of the MvBL and sap mutants. Fig. S8 Alignment of the STERILE APETALA amino acid sequences. Fig. S9 Collinearity analysis of the chromosome segment underlying the nectar volume quantitative trait loci across four Mimulus species and quantification of nectar volume associated with selected genetic markers near SAP. Fig. S10 Phylogenetic analysis of STERILE APETALA and CRABS CLAW. Fig. S11 Quantitative measurement of MvCRC1 and MvCRC2 expression levels in pistils of MvBL and sap-1 at the 2-mm stage. Fig. S12 Nectary microdissected at the S2 stage in MvBL and sap-1 mutants. Methods S1 Materials and Methods. Table S1 Candidate single nucleotide polymorphisms (SNPs) from the mutant genome comparisons. The SNP highlighted in bold is the causal mutation. Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.