Cold-shock response is elicited by the transfer of exponentially growing cells from their optimum temperature to a significantly lower growth temperature and is characterized by the induction of several cold-shock proteins. As cold shock leads to several cellular processes being modified, slowed, or inhibited, these proteins, which presumably possess a variety of different activities prove to be critical for survival and continued growth of the cell at low temperature. NusA (Friedman et al. 1984), IF2 (Gualerzi & Pon 1990), RbfA (Dammel & Noller 1995), CspA (Goldstein et al. 1990) and its homologues such as CspB (Lee et al. 1994), CspG (Nakashima et al. 1996), CspI (Wang et al. 1999), CsdA (Toone et al. 1991; Turner et al. 2007) and PNPase (Reiner 1969a,b; Donovan & Kushner 1986) are among the various major cold-shock proteins produced in Escherichia coli after transfer of the cells from 37 to 15 °C (Jones et al. 1987).
Cold-shock response is well studied from various bacteria (Graumann & Marahiel 1999; Lopez et al. 2001; Balhesteros et al. 2010) to higher organisms (Bradley et al. 2007). One of the main consequences of cold shock is stabilization of the secondary structures in nucleic acids (Rajkowitsch et al. 2007; Phadtare & Severinov 2010; Phadtare 2011) leading to hindrance of (i) transcription and translation and (ii) RNA degradation. The role of CspA and its homologues as RNA chaperones that act as transcription antiterminators and essentiality of this activity for the cold acclimation of cells has been well studied and the Csp-responsive, promoter-proximal sequences that can block the transcript elongation have been identified in several target genes of Csp homologues (Jiang et al. 1997; Bae et al. 2000; Phadtare et al. 2002, 2006; Phadtare & Inouye 2004). However, cold-shock proteins such as RNA helicase, CsdA and 3′-5′ processing exoribonucleases such as PNPase and RNase R are presumably involved in facilitating the RNA metabolism at low temperature. In the present study, we will focus on these three proteins in the context of the cellular cold-shock response.
CsdA is a highly conserved, DEAD-box RNA helicase (Linder et al. 1989) and is essential only at low temperature (Jones et al. 1996; Charollais et al. 2004). It has been suggested to be involved in the biogenesis of the small ribosomal subunits (Toone et al. 1991; Moll et al. 2002) and the 50S ribosomal subunits (Charollais et al. 2004), promotion of translation initiation of structured mRNAs (Lu et al. 1999), low-temperature riboregulation of RpoS mRNA (Resch et al. 2010) and stabilization and degradation of mRNAs (Tamura et al. 2012; Khemici et al. 2004; Prud'homme-Genereux et al. 2004; Yamanaka & Inouye 2001). Its role in mRNA decay and ribosome biogenesis has been studied in detail. Interestingly, the unwinding (helicase) activity of CsdA may be important for both of these functions. It was suggested that CsdA may help 50S assembly by modulating RNA or RNP (ribonucleoprotein) structures, and its unwinding activity may be required to facilitate structural transitions within the RNA and may also allow proper binding of ribosomal protein(s) (Iost & Dreyfus 2006). It may alternatively prevent and/or resolve misfolding, which may provide assistance to rRNA to reach its active conformation. Our studies showed that CsdA-mediated mRNA decay may be critical during cold shock, and the helicase activity of CsdA is crucial for promoting degradation of mRNAs stabilized at low temperature (Awano et al. 2007). We showed that a target mRNA was significantly stabilized in the csdA null mutant cells at 15 °C, and this effect was counteracted by over-expression of wild-type CsdA protein but not by a helicase-deficient mutant of CsdA. Furthermore, our in vivo genetic screening of an E. coli csdA null mutant strain and results from other research groups showed that RhlE, another DEAD-box RNA helicase, can complement the cold sensitivity of the csdA null mutant strain (Iost & Dreyfus 2006; Awano et al. 2007; Jain 2008). We also observed that two cold-shock proteins, CspA and RNase R can also substitute for CsdA at low temperature, albeit to a somewhat weaker degree. It is also interesting to note that combined absence of CsdA and RNase R results in increased sensitivity of the cells to moderate temperature downshifts.
RNase R, PNPase and RNase II are the three major 3′-to-5′ processing exoribonucleases in E. coli. These enzymes are primarily involved in RNA metabolism in E. coli. PNPase (Jones et al. 1987; Zangrossi et al. 2000) and RNase R (Cairrao et al. 2003) are induced by cold shock and are suggested to be the universal degraders of structured RNA in vivo (Li et al. 2002; Cheng & Deutscher 2003). Escherichia coli PNPase (polynucleotide phosphorylase) is essential for growth at low temperatures (Luttinger et al. 1996; Piazza et al. 1996). However, its exact role in this essential function is not fully elucidated. In addition to promoting the processive degradation of RNA, PNPase is also responsible for residual RNA tailing observed in E. coli mutants devoid of the main polyadenylating enzyme PAP I (Mohanty & Kushner 2000a,b, 2003).
We have made several interesting observations during further studies on these proteins, (i) of these three exoribonucleases, only RNase R can substitute for CsdA at low temperature (Awano et al. 2007), (ii) upon in vivo domain analysis of RNase R, we showed that it also has helicase activity, and this activity is essential to substitute for the cold-shock function of CsdA. RNase R possesses ribonuclease and helicase activities, which are distinct from each other as the mutant RNase R proteins lacking the ribonuclease activity retained their ability to substitute for CsdA. It was observed that the CSD2 domain of RNase R is not essential for its ribonuclease activity; however, it is important for its helicase activity (Awano et al. 2010), (iii) ribonuclease activity of PNPase is critical at low temperature (Awano et al. 2008); however, in spite of its ability to act as the universal degrader of the structured RNAs in vivo (Li et al. 2002; Cheng & Deutscher 2003), it cannot substitute for CsdA at low temperature (Awano et al. 2007) and (iv) RNase II and RNase R belong to the RNR family and exhibit approximately 60% similarity in their secondary structures, but RNase II can substitute for PNPase and not CsdA, whereas RNase R can substitute for CsdA, but not PNPase at low temperature (Awano et al. 2007, 2008).
These observations suggest that (i) CsdA and RNase R share common target mRNAs, whose degradation they facilitate by virtue of their unwinding activity. Note that the target mRNA, which was significantly stabilized in the csdA null mutant cells at 15 °C was destabilized by over-expression of CsdA as well as RNase R (Awano et al. 2007). (ii) Some of these targets, which are apparently important for the cold acclimation of cells, cannot be degraded by PNPase leading to the lack of suppression of the cold sensitivity of the csdA null mutant strain by the PNPase over-expression. (iii) There exists a distinct set of target mRNAs degraded by PNPase, which may be relevant to the cold-shock response of the cells. However, these mRNAs are not degraded by RNase R which would explain the lack of suppression of the cold sensitivity of the Δpnp strain by RNase R. Thus, it can be hypothesized that these proteins have different substrate specificities. Previously, it was reported that although CsdA does not have strong specificity for RNA substrates in vitro, it may have substrate preferences in vivo (Iost & Dreyfus 2006). The helicase activity of RNase R is independent of its ribonuclease activity; thus, the RNA substrates for these activities may be different or it may share a common subset of mRNAs, which it can both unwind and degrade. Although RNase R may have some preference for RNA substrates, it can degrade (Awano et al. 2008); no data are available at present as to the preferred targets for its unwinding activity. A comparative DNA microarray analysis of the targets of these three proteins at 15 °C should provide clues to elucidate how RNA metabolism is mediated at low temperature. Thus, in the present study, DNA microarray analysis of the effect of deletion and over-expression of CsdA, RNase R and PNPase at 15 °C as compared to the wild-type cells was carried out.
To get a comprehensive picture of the changes occurring at low temperature, these data were also compared with our published DNA microarray data (Phadtare & Inouye 2004) profiling the genes which are affected by the temperature downshift in the wild-type cells as well as cold-sensitive cells bearing deletions of the four csp genes.