- Top of page
- Materials and Methods
- Supporting Information
The cold shock domain (CSD) is a highly conserved nucleic acid-binding domain found in bacteria, animals and higher plants (Yamanaka et al., 1998; Karlson & Imai, 2003; Mihailovich et al., 2010). The CSD structure contains a five-stranded β-barrel sheet with two consensus RNA binding motifs (RNP-1 and RNP-2) that are essential for single-stranded DNA/RNA binding (Schroder et al., 1995; Hillier et al., 1998; Wang et al., 2000). Bacterial cold shock proteins (CSPs) are composed of a single CSD and act as RNA chaperones that destabilize secondary structures in RNAs (Jiang et al., 1997; Bae et al., 2000; Phadtare et al., 2002). In Escherichia coli, nine members of the CSP gene family (cspA to cspI) have been identified and four of them (cspA, cspB, cspG and cspI) are highly induced after cold shock (Yamanaka et al., 1998; Wang et al., 1999). A quadruple deletion mutant that lacks cspA, cspB, cspG and cspE exhibits growth defects at low temperature (Xia et al., 2001). At low temperatures, RNA molecules form secondary structures that may inhibit RNA functions including translation and transcription (Rajkowitsch et al., 2007). Therefore, RNA chaperone activity is likely to be critical for RNA function at low temperatures. In addition, CspA, CspC and CspE also function as transcription antiterminators and regulate the expression of a set of cold-responsive genes (Bae et al., 2000).
Vertebrate Y-box proteins are the most extensively studied eukaryotic CSD proteins and Y-box CSDs show about 40% amino acid sequence identity to bacterial CSDs (Wistow, 1990; Wolffe et al., 1992). All vertebrate Y-box proteins contain a single N-terminal CSD and C-terminal auxiliary domains (Kohno et al., 2003). The vertebrate Y-box binding protein-1 (YB-1) acts in transcriptional and translational regulation, and plays critical roles in many biological processes such as stress responses, cell proliferation and DNA repair (Kohno et al., 2003). YB-1-deficient mice die during late embryonic development, indicating its essential function in animals (Lu et al., 2005). In addition, YB-1-depleted chicken cells cease cell division after a temperature downshift (Matsumoto et al., 2005). YB-1 binds to DNA and RNA, and shows concentration-dependent melting and annealing activities (Matsumoto & Wolffe, 1998; Skabkin et al., 2001). These activities are probably necessary for the pleiotropic functions of YB-1. A Xenopus Y-box binding protein, FRGY2, is involved in translational repression of maternal mRNAs that accumulate in oocytes, preventing premature mRNA translation (Matsumoto et al., 1996). In Caenorhabditis elegans, another CSD protein, LIN-28, was first characterized as a critical regulator of larval development (Moss et al., 1997). LIN-28 is an evolutionarily conserved RNA-binding protein containing an N-terminal CSD and retroviral-like CCHC zinc fingers at the C-terminus. Mammalian LIN-28, together with other factors, can reprogram human somatic cells into pluripotent stem cells (Yu et al., 2007; Liao et al., 2008). LIN-28 post-transcriptionally inhibits the biogenesis of let-7 miRNA, which is a regulator of cell growth and differentiation in embryonic cells (Roush & Slack, 2008; Viswanathan et al., 2008). Taken together, these studies suggest that animal CSD proteins are essential regulators of a variety of biological processes.
Plant CSD proteins typically contain an N-terminal CSD and a glycine-rich region interspersed with various numbers of retroviral-like CCHC zinc fingers at the C-terminus (Sasaki & Imai, 2012). The first studied plant CSD protein was wheat cold shock domain protein 1 (WCSP1; Karlson et al., 2002), which accumulates in crown tissue during prolonged cold acclimation. WCSP1 mRNA is not modulated by other environmental stresses such as salt, drought and heat, or treatment with ABA (Karlson et al., 2002), which suggests that the function of WCSP1 is specific to cold adaptation. WCSP1 has nucleic acid-binding activity (Karlson et al., 2002; Nakaminami et al., 2005) and unwinds nucleic acid duplexes in vitro and in vivo (Nakaminami et al., 2006). These studies indicated that, similar to bacterial CSPs, WCSP1 functions as an RNA chaperone to destabilize RNA secondary structures during cold acclimation.
In Arabidopsis, four CSD protein genes (AtCSP1–AtCSP4) have been identified in the genome sequence (Karlson & Imai, 2003; Sasaki & Imai, 2012) and AtCSP3 (At2 g17870) is essential for the acquisition of freezing tolerance in Arabidopsis (Kim et al., 2009). C-repeat binding factors (CBFs) regulate freezing tolerance through activation of cold-regulated (COR) gene expression (Jaglo-Ottosen et al., 1998; Liu et al., 1998; Thomashow, 1998). However, AtCSP3 does not affect expression of CBFs or COR genes (Kim et al., 2009). AtCSP3 regulates expression of stress-related genes but their roles in freezing tolerance are unknown (Kim et al., 2009). Transgenic plants overexpressing AtCSP1 (CSDP1, At4 g36020) show delayed seed germination under dehydration and salt stress conditions, and plants overexpressing AtCSP2 (AtGRP2/CSDP2, At4 g38680) exhibit accelerated seed germination under salt stress conditions (Park et al., 2009). Overexpression of AtCSP1 or AtCSP2 did not enhance freezing tolerance of wild-type plants but did complement the freezing sensitive phenotype of grp7, a mutant of glycine-rich RNA binding protein 7 (Park et al., 2009). Expression of bacterial CSPs in plants confers stress tolerance against drought and cold (Castiglioni et al., 2008). These studies support the hypothesis that plant and bacterial CSD proteins have conserved biochemical functions and are involved in stress adaptation.
Plant CSD proteins also regulate developmental processes similar to animal CSD proteins. AtCSP2 is highly expressed in meristematic tissues and ovules (Fusaro et al., 2007; Sasaki et al., 2007; Nakaminami et al., 2009). AtCSP2 knockdown plants showed early flowering, reduced number of stamens and high rates of abnormal development of seeds/embryos (Fusaro et al., 2007). Recently, Yang & Karlson (2011) demonstrated that overexpression of AtCSP4 (AtGRP2b, At2 g21060) reduces silique length and induces embryo lethality. Overexpression of AtCSP4 affects the expression of several MADS box and endosperm development genes during floral and silique development (Yang & Karlson, 2011).
In this paper, we describe the functional characterization of AtCSP2 in planta. Genetic studies with overexpressor and mutant lines indicate that AtCSP2 is a negative regulator of freezing tolerance. In addition, AtCSP2 is functionally redundant with AtCSP4 in freezing tolerance and flowering time.
- Top of page
- Materials and Methods
- Supporting Information
Differences in freezing tolerance between the atcsp2-3 atcsp4-1 mutant and wild-type were observed after cold acclimation but not in nonacclimated conditions (Fig. 2b). In addition, 35S:AtCSP2 plants showed reduced freezing tolerance only when cold treatment was applied (Fig. 4b). These data clearly demonstrate that AtCSP2 negatively regulates cold acclimation in Arabidopsis. AtCSP2 may directly regulate the expression of CBFs or control transcription or translation processes upstream of CBFs. It has been demonstrated that several proteins negatively regulate expression of CBFs (Chinnusamy et al., 2007). HOS1, an ubiquitin E3 ligase, represses expression of CBFs and confers increased sensitivity to freezing under cold-acclimated conditions (Dong et al., 2006). MYB15 (an R2R3 type MYB family protein) is another negative regulator of the CBF pathway and suppresses expression of CBFs (Agarwal et al., 2006). Interestingly, AtCSP2 and MYB15 are both induced during cold acclimation, even though they negatively regulate the CBF pathway. In our study, transcripts of CBFs were transiently up-regulated after 6 h cold treatment and the transcript abundances decreased thereafter (Fig. 3), while expression of AtCSP2 increased after 6 h cold treatment and the increased expression level was maintained for up to 48 h of cold treatment (Fusaro et al., 2007). This may suggest that AtCSP2 is involved in the attenuation of the CBF pathway to fine-tune the levels of freezing tolerance at the later stage of cold acclimation.
Although AtCSP2 and AtCSP3 are both induced by cold treatment and have an RNA chaperone activity in common, their physiological functions clearly differ. AtCSP3 positively regulates freezing tolerance through a CBF-independent pathway under both non-acclimated and cold-acclimated conditions (Kim et al., 2009). 35S:AtCSP3 plants show no morphological or developmental phenotypes. By contrast, AtCSP2 negatively regulates freezing tolerance through a CBF-dependent pathway (Figs 3, 5) and overexpression of AtCSP2 affects plant development (Figs 6, S3). These data suggest that there is no functional interaction between AtCSP2 and AtCSP3 in regulating freezing tolerance and plant development.
The atcsp2 and atcsp4 single mutants did not show differences from the wild-type in freezing tolerance and plant development (Figs 2a, S2). However, the atcsp2 atcsp4 double mutants showed increased freezing tolerance after cold acclimation and early flowering (Figs 2b, S2b). AtCSP2 shows the highest sequence similarity with AtCSP4 and the two genes also show very similar tissue expression patterns and protein subcellular localization (Sasaki et al., 2007; Nakaminami et al., 2009; Yang & Karlson, 2011). These data suggest that AtCSP2 and AtCSP4 function redundantly in freezing tolerance and flowering time. By contrast, neither atcsp2 nor atcsp4 single mutants, nor the double mutant, showed an altered silique phenotype. Therefore, their functional relationship in silique development remains unclear.
Fusaro et al. (2007) demonstrated that AtCSP2 RNAi plants in the C24 ecotype showed early flowering, reduced number of stamens and high rates of abnormal seeds/embryos. The atcsp2-3 atcsp4-1 mutants (Col-0 background) also showed an early flowering phenotype but no significant alterations in the number of stamens and seed development (Fig. S4). Since the atcsp2-3 atcsp4-1 mutant has some remaining AtCSP2 expression (Fig. 1d), it may be that this low level of AtCSP2 expression is sufficient to complete stamen and embryo development in Col-0.
In addition to plant growth and flowering time, AtCSP2 negatively regulates silique length without affecting the number of seeds (Fig. 6). In the wild-type, there is a positive correlation between seed number and silique length (Cox & Swain, 2006). In addition, it was reported that increased seed size in Arabidopsis lba1/atupf1-1 mutant coincides with a decrease in seed numbers within a unit length of silique (Yoine et al., 2006). Therefore, it is possible to hypothesize that there is a mechanism that maintains proper packing of seeds within a silique. This mechanism is probably impaired in 35S:AtCSP2 plants. Since AtCSP2 expression is detected in ovules and transmitting tissue (Fusaro et al., 2007; Sasaki et al., 2007), AtCSP2 may play a role in regulating the distance between funiculi. It is interesting to note that the siliques of ga20ox2 mutants, which have reduced GA4 levels, are shorter than that of wild-type (Rieu et al., 2008). However, the number of seeds per silique is not reduced in the ga20ox2 mutant (Rieu et al., 2008), which indicates that the seeds are more closely packed in the ga20ox2 siliques. Since AtCSP2 negatively regulates GA biosynthesis genes in seed germination (K. Sasaki et al., unpublished), the short silique phenotype of 35S:AtCSP2 plants may possibly involve GA signaling.
The mechanism by which AtCSP2 regulates RNA functions is currently unknown. AtCSP2 may enhance translation of mRNAs that encode negative regulators of freezing tolerance and plant development. YB-1, a well studied CSD protein in animals, is a major component of cytoplasmic messenger ribonucleoprotein particles (mRNPs) and binds to mRNA to inhibit the access of cap-binding factor, resulting in translational arrest, which is called RNA masking (Kohno et al., 2003; Mihailovich et al., 2010). YB-1 stimulates translation when the YB-1/mRNA ratio is low, but inhibits translation when the ratio increases (Pisarev et al., 2002; Nekrasov et al., 2003). RNA masking activity was also described for another CSD protein, FRGY2, in Xenopus oocytes (Matsumoto et al., 1996). In Chlamydomonas, a CSD protein NAB1 from Chlamydomonas reinhardtii stabilizes mRNAs of LHCBM (major light-harvesting complex of photosynthesis II) genes and represses translation at the pre-initiation stage (Mussgnug et al., 2005). Based on these studies, we can propose a putative function of AtCSP2 in translational repression of specific mRNAs during cold acclimation and various developmental stages. In addition to a role in translational regulation, YB-1 plays a role in pre-mRNA splicing (Chansky et al., 2001; Stickeler et al., 2001). The Arabidopsis U11/U12-31K protein harboring one CCHC-type zinc finger motif has RNA chaperone activity and is involved in proper U12 intron splicing and plant development (Kim et al., 2010). Since AtCSP2 localizes to the nucleolus and nucleoplasm in addition to cytoplasm (Sasaki et al., 2007), it is also possible to speculate that AtCSP2 plays a role in pre-rRNA processing and pre-mRNA splicing.
Our data revealed that AtCSP2 has regulatory functions in several physiological processes in addition to cold acclimation. Regulatory mechanisms of these processes can be different. Our study with AtCSP3 revealed that AtCSP3 interacts with a variety of RNA-processing proteins within the nucleus and cytoplasm (Kim et al., 2013). Considering the structural conservation, we speculate that AtCSP2 also functions as a versatile modulator of gene expression within a variety of RNA processing bodies. Therefore, it will be important to determine coregulators and downstream factors of AtCSP2 for each phenotype in the future study.