Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Mar 2, 2004; 101(9): 3281–3285.
Published online Feb 18, 2004. doi:  10.1073/pnas.0306778101
PMCID: PMC365781
Plant Biology

FRIGIDA-related genes are required for the winter-annual habit in Arabidopsis


In temperate climates, the prolonged cold temperature of winter serves as a seasonal landmark for winter-annual and biennial plants. In these plants, flowering is blocked before winter. In Arabidopsis thaliana, natural variation in the FRIGIDA (FRI) gene is a major determinate of the rapid-cycling vs. winter-annual flowering habits. In winter-annual accessions of Arabidopsis, FRI activity blocks flowering through the up-regulation of the floral inhibitor FLOWERING LOCUS C (FLC). Most rapid-flowering accessions, in contrast, contain null alleles of FRI. By performing a mutant screen in a winter-annual strain, we have identified a locus, FRIGIDA LIKE 1 (FRL1), that is specifically required for the up-regulation of FLC by FRI. Cloning of FRL1 revealed a gene with a predicted protein sequence that is 23% identical to FRI. Despite sequence similarity, FRI and FRL1 do not have redundant functions. FRI and FRL1 belong to a seven-member gene family in Arabidopsis, and FRI, FRL1, and at least one additional family member, FRIGIDA LIKE 2 (FRL2), are in a clade of this family that is required for the winter-annual habit in Arabidopsis.

Keywords: FLOWERING LOCUS C, vernalization, natural variation

To coordinate reproductive development with seasonal change, many plant species found in temperate climates have evolved a biennial or winter-annual growth habit. The distinguishing feature of this growth habit is an obligate or facultative requirement for vernalization before flowering can occur. Thus flowering in biennials and winter annuals is blocked before winter, and exposure to winter permits flowering the next spring. In Arabidopsis thaliana, there exist both winter-annual and rapid-flowering types (1-4), which flower rapidly without vernalization. The vernalization requirement of naturally occurring winter-annual types is due to the dominant alleles of two genes, FRIGIDA (FRI) and FLOWERING LOCUS C (FLC) (5-7). FLC encodes a MADS-domain-containing transcription factor that acts to delay flowering in a dosage-dependent manner, and FRI is required for high levels of FLC expression (8, 9).

Rapid-flowering types contain either null alleles of FRI (1-3) or alleles of FLC that are not up-regulated by FRI (1, 10). However, rapid-flowering types can acquire a winter-annual habit by loss-of-function mutations in any of six genes [LUMINIDEPENDENS (LD), FCA, FLOWERING LOCUS D (FLD), FPA, FY, and FVE] that define the autonomous floral-promotion pathway (11). Like FRI, the autonomous-pathway genes also affect flowering time through the regulation of FLC (8, 9, 12). In rapid-flowering accessions, FLC expression is repressed by the autonomous pathway; consequently, autonomous-pathway mutants have high levels of FLC and are late-flowering. In winter-annual accessions, FRI acts epistatically to the autonomous pathway to up-regulate FLC levels and block flowering. Regardless of whether increased FLC expression is due to FRI or autonomous-pathway mutations, vernalization promotes flowering by causing the epigenetic repression of FLC expression (8, 9).

Screens for flowering-time mutants in rapid-flowering backgrounds have been effective in genetically defining several pathways that regulate flowering time. Because rapid-flowering backgrounds lack FRI activity, however, genes that specifically interact with FRI to confer the winter-annual habit have not been identified. Loss-of-function mutations in the recently identified genes, VERNALIZATION INDEPENDENCE 3 (VIP3) (13), VERNALIZATION INDEPENDENCE 4 (VIP4) (14), and PHOTOPERIOD INDEPENDENT EARLY FLOWERING 1 (PIE1) (15), suppress the effects of FRI on FLC expression and flowering time. However, these mutations also suppress the up-regulation of FLC expression by autonomous-pathway mutants and have additional pleiotropic phenotypes. Thus these genes appear to affect gene expression more generally and do not act specifically in conjunction with FRI. In this paper, we describe the identification and characterization of FRIGIDA LIKE 1 (FRL1), a gene that is specifically required for the FRI-mediated block to flowering, but not for the late-flowering phenotype of autonomous-pathway-mutants. FRL1 is related to FRI, and FRI, FRL1, and another related gene, FRIGIDA LIKE 2 (FRL2), represent a clade of the family of FRI-related genes that is required for the winter-annual habit.


Plant Material and Mutagenesis. FRI-SF2 in the Columbia (Col) background (16), ld-1 (17), fpa-7, flc-3, gi-13 (12) have been described previously. frl2-1 was obtained from the SALK collection, (SALK 098628) (18). T-DNA and fast-neutron mutagenized populations have been described (8).

Growth Conditions. Plants were grown under long days (16 h light/8 h dark) or short days (8 h light/16 h dark) at 22°C under cool-white fluorescent lights. For experiments involving vernalization, seeds were plated on agar-solidified medium containing 0.65 g/liter Peters Excel 15-5-15 fertilizer (Grace Sierra, Milpitas, CA) and were kept at room temperature overnight to allow seeds to become metabolically active before being transferred to 2°C on for 40 days. During cold treatment, samples were kept under short-day conditions (8 h light/16 h dark).

Gene Expression Analysis. RNA blot analysis of FLC was performed as described (10). For RT-PCR analysis, first-strand cDNA synthesis was performed on 2 μg of RNA by using the SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen Life Technologies) by using a primer containing the M13 primer sequence with an oligo dT extension (5′-GTAAAACGACGGCCAGTCCCTTTTTTTTTTTTTTTTTT-3′).

PCR amplification was performed with Platinum TaqDNA Polymerase (Invitrogen Life Technologies) according to the manufacturer's recommendations. FLC (5′-TTCTCCAAACGTCGCAACGGTCTC-3′ and 5′-GATTTGTCCAGCAGGTGACATCTC-3′), SOC1 (5′-TGAGGCATACTAAGGATCGAGTCAG-3′ and 5′-GCGTCTCTACTTCAGAACTTGGGC-3′), FRI (5′-GCAAAACGGAAAGCCCAGTC-3′ and 5′-CGATGAGGAAAAGATGTTGACGG-3′), and UBIQUITIN (5′-GATCTTTGCCGGAAAACAATTGGAGGATGGT-3′ and 5′-CGACTTGTCATTAGAAAGAAAGAGATAACAGG-3′) were amplified by using the indicated primers. Because FRL1 is intronless, PCR was performed with a gene-specific primer 5′-AATCCTCCCAAGCAAGAGCCACAG-3′ and the M13 Forward primer (5′-GTAAAACGACGGCCAGT-3′), whose sequence was incorporated into the template during first-strand synthesis. Cycling was performed as follows: 95°C for 4 min followed by 20 (for UBQ) or 24 cycles (for FRL1, FRI, FLC, and SOC1) of 95°C for 30 sec, 65°C for 30 sec, and 72°C for 30 sec. Amplified fragments were separated on a 1.2% agarose gel, blotted onto a Biotrace HP nylon membrane (Gelman, Ann Arbor, MI), hybridized with radiolabeled probes, and visualized on a Molecular Dynamics PhosphorImager.

Results and Discussion

Identification of a Suppressor of FRI-Mediated Late Flowering. To identify components of the FRI-mediated vernalization requirement of winter-annual accessions, T-DNA and fast-neutron mutagenesis were carried out in a winter-annual line containing the dominant San Feliu-2 (SF2) allele of FRI in a Col background (16). In this background, which is late flowering due to high levels of FLC, loss of a positive regulator of FLC would be predicted to cause reduced levels of FLC and an early-flowering phenotype. Three alleles of a locus subsequently named FRL1 (see below) were identified in these screens. frl1 mutants behave recessively and flower much earlier than the FRI-containing parent under both long and short days similar to the Col rapid-flowering line, which is a fri null (Fig. 1A). Also, similar to the situation in Col, frl1 mutants have reduced levels of FLC expression (Fig. 1B). Thus FRL1 is required for the winter-annual habit conferred by FRI.

Fig. 1.
Effect of FRL1 on flowering time and FLC expression. (A) Total leaves formed by the primary meristem of plants grown under long (black bars) or short days (gray bars). The shaded and open portions of the bars indicate the number of rosette leaves and ...

frl1 Mutations Do Not Affect Flowering in Autonomous-Pathway-Mutant Backgrounds or in the Absence of FRI. Because the winter-annual habit can be conferred by either dominant alleles of FRI or by autonomous-pathway mutations, we evaluated whether the frl1 mutation would also suppress the late flowering of the autonomous-pathway mutants ld and fpa. The presence of the frl1 mutation does not affect the late-flowering phenotype of either autonomous-pathway mutant (nor does it affect the phenotype of the photoperiod-pathway mutant gigantea) (Fig. 2A). Consistent with this FRI specificity, the frl1-1 mutation does not affect flowering behavior in the rapid-flowering Col background, which lacks FRI activity (Fig. 2A), and frl1 mutants also lack any of the pleiotropic phenotypes associated with vip3, vip4, and pie1 mutations (13-15).

Fig. 2.
Phenotype of FRL1 in various late-flowering backgrounds. (A) Flowering time of frl1-1 in fri, autonomous-pathway (ld and fpa), and photoperiod-pathway (gi) mutant backgrounds. Flowering time is expressed as total leaves formed by the primary meristem. ...

The FRI specificity of FRL1 is also apparent at the RNA level (Fig. 2B). In the absence of FRI, FLC is expressed at low levels and SUPPRESSOR OF OVEREXPRESSION OF CONSTANS1 (SOC1), a promoter of flowering that is negatively regulated by FLC (19-21), is highly expressed. In the presence of FRI, FLC expression is increased and leads to the suppression of SOC1 (12, 20). The frl1 mutation eliminates the effect of FRI on FLC and SOC1 expression. Similar to the effect of FRI, autonomous-pathway mutations, such as ld, also lead to an up-regulation of FLC RNA expression and a corresponding decrease in SOC1 (12, 20, 21). In contrast to FRI, however, frl1 mutations have no effect on FLC and SOC1 regulation in an ld mutant background (Fig. 2B). Thus the effects of frl1 mutations on flowering time and FLC and SOC1 expression appear to be specific to FRI-containing backgrounds.

FRL1 Encodes a Protein with Homology to FRI. The frl1-1 allele was created by T-DNA mutagenesis, and DNA flanking the site of T-DNA insertion was isolated by thermal asymmetric interlaced PCR (22). The sequence of the flanking DNA revealed that the T-DNA was inserted 14 bp upstream of the predicted translational start site of At5g16320 (Fig. 3A). To determine whether At5g16320 was FRL1, frl1-1 was transformed with a genomic clone containing At5g16320. Late flowering was restored in the T1 plants, demonstrating that At5g16320 was indeed FRL1 (data not shown). Furthermore, sequencing of the fast-neutron alleles frl1-2 and frl1-3 also revealed lesions in At5g16320 (Fig. 3A).

Fig. 3.
Sequence analysis of FRL1 and related proteins. (A) Schematic representation of the FRL1 locus with the sites of lesions indicated. frl1-2 contains a 40-nt deletion (CTGTTGATACGAGAAATAGAGCTAAGAAACTGGCTTACCA) and frl1-3 contains a 4-nt deletion (GTGTTTCTGCAT ...

FRL1 is predicted to encode a plant-specific 470-aa protein. A blast search of the FRL1 predicted amino acid sequence to the Arabidopsis proteome showed that FRL1 belongs to a seven-member gene family located on chromosomes I, II, IV, and V (Fig. 3 B and C). The Arabidopsis FRL1 gene family can be grouped into two clades (Fig. 3C). One contains At5g48385, At1g14900, and At2g22440. Microsynteny exists in the regions surrounding At1g14900 and At2g22440; together with the high degree of amino acid sequence identity (81%), indicating that At1g14900 and At2g22440 were derived from a relatively recent duplication event. The other clade contains FRL1, At1g31814 (54% identical to FRL1 and designated FRL2, see below), At5g27230 (25% identical to FRL1), and interestingly, FRI (23% identical). Although FRI and FRL1 are related, FRI contains additional N- and C-terminal domains, and there are no large blocks of primary sequence identity that might suggest a functional domain shared between FRI and FRL1 (Fig. 3D).

FRL1 and FRI Have Distinct Roles in the Regulation of Flowering. The observation that loss-of-function mutations in either frl1 or fri strongly suppress the late-flowering phenotype of a winter-annual line and that their phenotypes are not additive in double mutants (Fig. 2A) indicates that FRL1 and FRI are likely to have nonredundant roles. To further investigate whether FRI and FRL1 might have unique functions, the effects of increased expression of each gene were evaluated (Fig. 4A). When placed under the control of the strong constitutive 35S Cauliflower Mosaic virus promoter (23), both 35S::FRI and 35S::FRL1 are able to restore late flowering in their respective mutant backgrounds, demonstrating that the constructs are functional. However, overexpression of FRI cannot substitute for a loss of FRL1 activity and vice versa; i.e., 35S::FRL1 does not delay flowering in a fri null mutant and conversely 35S::FRI causes only a slight delay in flowering in a frl1 null mutant. Thus despite sequence similarity, FRL1, and FRI have nonredundant roles in the regulation of flowering time. It is also interesting to note that overexpression of 35S::FRI and 35S::FRL1 in their respective mutant backgrounds did not give rise to plants that were significantly later than wild-type plants containing FRI in the T1 generation. Along with the fact that fri (16) and frl1 mutations are fully recessive, this indicates that the levels of FRI and FRL1 are not limiting for the inhibition of flowering.

Fig. 4.
Effects of overexpression of FRL1 and FRI. (A) Flowering time of transgenic lines containing 35S::FRL1 or 35S::FRI in frl1 and fri-mutant backgrounds and after vernalization. All lines are in the Col genetic background. Data are presented for two independent ...

The frl1 mutants, however, do not suppress the late-flowering phenotype of winter-annual accessions as strongly as fri or flc null mutants (i.e., frl1 mutants flower later that fri or flc loss-of-function mutants, Fig. 1 A). Thus FRI is able to delay flowering independently of FRL1 activity. This may indicate that FRI acts to inhibit flowering through FRL1-dependent and -independent pathways or, alternatively, that frl1 mutants may not represent a total loss in FRL1 activity due to functional redundancy with other related genes (see below).

Although FRL1 and FRI appear to have unique activities, it is important to note that frl1 mutants flower slightly later than fri or flc loss-of-function mutants (Fig. 1 A; frl1 mutants produce approximately five more leaves than a fri null). Some degree of functional redundancy between FRL1 and a related gene family member could explain the incomplete suppression of the late-flowering phenotype of FRI by frl1 mutants. As mentioned above, At1g31814 is the gene most closely related to FRL1 and is thus a candidate for a functionally redundant gene. Although At1g31814 single mutants did not have a significant effect on flowering time in a FRI-containing background (data not shown), the frl1 At1g31814 double mutant flowered earlier than the frl1 single (Fig. 1 A), suggesting that At1g31814 may play a role similar to that of FRL1. We have therefore named At1g31814 FRIGIDA LIKE 2 (FRL2). The frl1 frl2 double mutant, however, remains slightly later flowering (approximately two leaves) than Col, a fri null, raising the possibility that higher-order mutants among family members may be required to completely suppress FRI. Alternatively, FRI may have a slight effect on flowering that is independent of FRL1 and FRL2 activity; in this case, fri mutants will always flower earlier than a frl1 frl2 double mutant.

Vernalization Acts Downstream of FRL1 and FRI. In winter-annual accessions of Arabidopsis, vernalization results in an epigenetic repression of FLC that is necessary for early flowering. For example, lines constitutively expressing FLC remain late flowering after vernalization (8, 9). Thus it is possible that constitutive expression of FRI or FRL1, which act as positive regulators of FLC, would also result in a vernalization-insensitive phenotype. Plants containing constitutively expressed FRI or FRL1, however, showed a wild-type vernalization response (Fig. 4A). This indicates that vernalization must act downstream of FRI and FRL1 transcription or through a separate pathway (Fig. 4B). Consistent with this model, FRI and FRL1 mRNA levels are not affected by vernalization (Fig. 2B).


These results demonstrate that FRL1 is required for the naturally occurring winter-annual habit of Arabidopsis but not for the winter-annual phenotype of autonomous-pathway mutants. Thus FRL1 is specifically required for the FRI-mediated up-regulation of FLC. The FRI/FRL1/FRL2 clade may represent a group of genes that have evolved to confer a winter-annual habit to Arabidopsis. Whether related genes play a similar role in other vernalization-requiring species remains to be determined. It is interesting to note that, in The Institute for Genomic Research's rice genome sequence, several FRI/FRL1-related genes are present (Fig. 2C), including one gene that is more closely related to FRI than to other Arabidopsis genes. Although rice is a summer-annual species, which does not have a vernalization requirement, these FRI-related genes may play a role in the flowering time in other monocots, such as biennial varieties of wheat, barley, and rye. It should be noted, however, that although FRI-like genes are present in other species, no clear homologs of FLC have been identified in cereals (24). Thus it is unclear whether the regulation of flowering time by FLC- and FRI-like genes is widespread.

It is interesting to note that, although loss-of-function mutations in FRI or FRL1 result in similar early-flowering phenotypes, only natural variation at the FRI locus has been reported as a major determinate of flowering habit in Arabidopsis. The majority of rapid-flowering accessions examined contain nonfunctional alleles of FRI and, because several different lesions have been identified among rapid-flowering accessions (1-3), it appears that FRI activity was lost multiple times over the course of evolution. It is unclear why extensive natural variation has occurred in FRI but not in FRL1. Perhaps loss of FRL1 activity confers more of a selective disadvantage than loss of FRI activity. Alternatively, accessions that are early flowering due to natural variation in FRL1 may exist but may have yet to be identified. Previous work examining DNA polymorphism in FRI identified several accessions that are early flowering despite containing potentially functional FRI genes (i.e., these FRI alleles do not contain insertions, deletions, or premature stop codons). Thus these rapid-flowering accessions are candidates for lines that have become early flowering due to loss of FRL1 function.


We are grateful to Shinhan Shiu for assistance in phylogenetic analysis. We thank the Salk Institute Genome Analysis Laboratory and the Arabidopsis Biological Resource Center at Ohio State University (Columbus) for providing T-DNA-insertion lines. This work was supported by the College of Agricultural and Life Sciences and the Graduate School of the University of Wisconsin, and by grants from the U.S. Department of Agriculture National Research Initiative Competitive Grants Program and the National Science Foundation (0133663 to R.M.A.).


This paper was submitted directly (Track II) to the PNAS office.

Abbreviation: Col, Columbia.

Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. BK004884 and BK004885).


1. Gazzani, S., Gendall, A. R., Lister, C. & Dean, C. (2003) Plant Physiol. 132, 1107-1114. [PMC free article] [PubMed]
2. Corre, V. L., Roux, F. & Reboud, X. (2002) Mol. Biol. Evol. 19, 1261-1271. [PubMed]
3. Johanson, U., West, J., Lister, C., Michaels, S., Amasino, R. & Dean, C. (2000) Science 290, 344-347. [PubMed]
4. Napp-Zinn, K. (1979) in La Physiologie de la Floraison, eds. Champagnat, P. & Jaques, R. (Collogues Internationaux Centre National de la Recherche Scientifique, Paris), pp. 217-220.
5. Lee, I., Bleecker, A. & Amasino, R. (1993) Mol. Gen. Genet. 237, 171-176. [PubMed]
6. Burn, J. E., Smyth, D. R., Peacock, W. J. & Dennis, E. S. (1993) Genetica 90, 147-155.
7. Clarke, J. H. & Dean, C. (1994) Mol. Gen. Genet. 242, 81-89. [PubMed]
8. Michaels, S. & Amasino, R. (1999) Plant Cell 11, 949-956. [PMC free article] [PubMed]
9. Sheldon, C. C., Burn, J. E., Perez, P. P., Metzger, J., Edwards, J. A., Peacock, W. J. & Dennis, E. S. (1999) Plant Cell 11, 445-458. [PMC free article] [PubMed]
10. Michaels, S. D., He, Y., Scortecci, K. C. & Amasino, R. M. (2003) Proc. Natl. Acad. Sci. USA 100, 10102-10107. [PMC free article] [PubMed]
11. Simpson, G. G. & Dean, C. (2002) Science 296, 285-289. [PubMed]
12. Michaels, S. D. & Amasino, R. M. (2001) Plant Cell 13, 935-942. [PMC free article] [PubMed]
13. Zhang, H., Ransom, C., Ludwig, P. & Van Nocker, S. (2003) Genetics 164, 347-358. [PMC free article] [PubMed]
14. Zhang, H. & van Nocker, S. (2002) Plant J. 31, 663-673. [PubMed]
15. Noh, Y. S. & Amasino, R. M. (2003) Plant Cell 15, 1671-1682. [PMC free article] [PubMed]
16. Lee, I., Michaels, S. D., Masshardt, A. S. & Amasino, R. M. (1994) Plant J. 6, 903-909.
17. Lee, I., Aukerman, M. J., Gore, S. L., Lohman, K. N., Michaels, S. D., Weaver, L. M., John, M. C., Feldmann, K. A. & Amasino, R. M. (1994) Plant Cell 6, 75-83. [PMC free article] [PubMed]
18. Alonso, J. M., Stepanova, A. N., Leisse, T. J., Kim, C. J., Chen, H., Shinn, P., Stevenson, D. K., Zimmerman, J., Barajas, P., Cheuk, R., et al. (2003) Science 301, 653-657. [PubMed]
19. Borner, R., Kampmann, G., Chandler, J., Gleissner, R., Wisman, E., Apel, K. & Melzer, S. (2000) Plant J. 24, 591-599. [PubMed]
20. Lee, H., Suh, S. S., Park, E., Cho, E., Ahn, J. H., Kim, S. G., Lee, J. S., Kwon, Y. M. & Lee, I. (2000) Genes Dev. 14, 2366-2376. [PMC free article] [PubMed]
21. Samach, A., Onouchi, H., Gold, S. E., Ditta, G. S., Schwarz-Sommer, Z., Yanofsky, M. F. & Coupland, G. (2000) Science 288, 1613-1616. [PubMed]
22. Liu, Y. G., Mitsukawa, N., Oosumi, T. & Whittier, R. F. (1995) Plant J. 8, 457-463. [PubMed]
23. Odell, J. T., Nagy, F. & Chua, N. H. (1985) Nature 313, 810-812. [PubMed]
24. Becker, A. & Theissen, G. (2003) Mol. Phylogenet. Evol. 29, 464-489. [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


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene
    Gene links
  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene links
  • MedGen
    Related information in MedGen
  • Nucleotide
    Published Nucleotide sequences
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • Protein
    Published protein sequences
  • PubMed
    PubMed citations for these articles
  • Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...