To describe the Gram-positive RNA degradosome better, understanding the well-studied E. coli RNA degradosome first would be helpful. The E. coli RNA degradosome complex includes the endoribonuclease RNase E, the exoribonuclease polynucleotide phosphorylase (PNPase), the DEAD-box RNA helicase RhlB, and the glycolytic enzyme enolase (Table 1). The degradosome is the major player involved in degrading most RNA transcripts and maintains the typical half-lives of RNA transcripts (Carpousis, 2007). The endoribonuclease RNase E is responsible for the decay of bulk mRNA. Mutations in RNase E influence the half-lives of ~60% of E. coli mRNA (Bernstein et al., 2004; Stead et al., 2011). Also, it acts as a scaffold, recruiting the other proteins and organizing the structure of the RNA degradosome (Górna et al., 2012).

RNase E has two major domains: N-terminal ribonuclease domain (NTD) and C-terminal scaffolding domain (CTD; Górna et al., 2012). The N-terminal half of RNase E has endoribonuclease activity, which is the major activity initiating the decay of bulk mRNA in the RNA degradosome and essential for the growth of E. coli. The C-terminal half of RNase E is intrinsically unstructured except six microdomains mostly involved in interactions with other components in the degradosome. The secondary structure of this half is not much predicted by in silico analysis, indicating that this part is unstructured and does not form a stable globular structure (Gunasekaran et al., 2003). Even though most sequences of the C-terminal half are not much conserved among RNase E orthologs, some parts are more conserved than others. These conserved microdomains range from 20 to 70 residues in size and are binding sites for other protein components in the RNA degradosome, RNA substrates, and the cytoplasmic membrane (Vanzo et al., 1998; Callaghan et al., 2004; Chandran et al., 2007; Khemici et al., 2008). The first microdomain in the C-terminal half, the membrane targeting sequence (MTS), forms an amphipathic helix that interacts with the membrane.

B. subtilis is a spore-forming soil bacterium whose RNA degradosome was discovered first among Gram-positive bacteria (Commichau et al., 2009). B. subtilis, a facultative anaerobe, has been studied intensively, especially in the area of chromosome replication and sporulation, and considered as a Gram-positive model organism. The B. subtilis RNA degradosome contains similar or the same components as the E. coli RNA degradosome such as a DEAD-box helicase, the glycolytic enzyme enolase, and the 3′–5′ exoribonuclease PNPase. Like in E. coli, these components are also found in the cytoplasm (Cascante-Estepa et al., 2016), suggesting that these proteins fulfill independent functions in the cytoplasm and only a fraction of them interact with the degradosome. However, the B. subtilis degradosome does not have a homolog of RNase E, the scaffold protein in E. coli. Instead, the degradosome contains new endoribonucleases that are not found in the E. coli degradosome. These novel enzymes are RNase Y, RNase J1, and RNase J2. Among these three endoribonucleases, RNase Y and RNase J1 are involved in the degradation of bulk mRNA (Even et al., 2005; Shahbabian et al., 2009; Durand et al., 2012).

When RNase Y is depleted, the average half-life of bulk mRNA increases more than 2-fold (Shahbabian et al., 2009). Also, RNase Y influences the steady-state level of about 25% of transcripts (Lehnik-Habrink et al., 2011a; Durand et al., 2012), indicating that RNase Y plays a crucial role in global mRNA turnover. Like RNase E, RNase Y is an endoribonuclease with a preference for 5′ monophosphorylated ends (Shahbabian et al., 2009). In addition, it has 5′ end independent endoribonuclease activity against certain transcripts (Yao et al., 2011). RNase Y also has the ability to interact with all other enzymes in the degradosome except RNase J2 (Commichau et al., 2009). At first, RNase Y was considered essential for growth. However, a recent study shows that it is not essential but the null mutant pays a very high fitness cost for growth (Figaro et al., 2013).

B. subtilis RNase Y has four major domains: N-terminal transmembrane domain, a coiled-coil domain, an RNA-binding KH domain, and a catalytic HD domain, in that order (Figure 1). The interaction with the membrane occurs through the N-terminal transmembrane domain and is important for the function of the RNA degradosome (Figure 2; Lehnik-Habrink et al., 2011b). The following domain is a coiled-coil domain, which seems to be intrinsically disordered in solution and important for self-oligomerization. Similar to the C-terminal half of RNase E, the disordered region might play an important role in protein-protein interactions. Structured microdomains in disordered regions can be relatively free from the structural and thermodynamic constraints of protein folding (Lehnik-Habrink et al., 2011b). The disordered region in RNase Y has not yet been studied in detail. The KH domain in RNase Y is an RNA binding domain that contains ~70 amino acids. KH domains can bind single stranded RNA or DNA and are involved in different biological processes such as transcriptional controls, translational controls, or splicing (Valverde et al., 2008). KH domains are classified into two types: Type I KH domains are found in multiple copies in eukaryotic proteins and Type II domains are found in generally a single copy in prokaryotic proteins. Interestingly, the KH domain in RNase Y is Type I with a single copy, so the KH domain in RNase Y is uncommon in bacteria. KH domains usually have broad substrate specificity. The typical binding surface of KH domains can accommodate only four consecutive RNA bases. Since the depletion of RNase Y increases the half-life of bulk mRNA significantly, RNase Y appears to have broad substrate specificity (Shahbabian et al., 2009).

In addition, RNase Y is often used for specific cellular events in B. subtilis. RNase Y cleaves between the cggR and gapA transcripts in the gapA operon that encodes six genes in the glycolytic pathway (cggR, gapA, pgk, tpiA, pgm, and eno), and this cleavage separates the expression level between the regulator CggR and the glycolysis structural genes by conferring differential stability to the transcripts (Commichau et al., 2009). Also, it initiates the degradation of all 11 SAM-dependent riboswitches only when the transcripts are prematurely terminated by binding to S-adenosylmethionine (SAM; Shahbabian et al., 2009).

The other ribonucleases found only in the B. subtilis degradosome, but not in the E. coli one, are RNase J1 and J2 (Figure 2). When RNase J1 was discovered, it was thought to be the functional homolog of RNase E because RNase J1 is a new class enzyme that can initiate RNA degradation. RNase J1 possesses two enzymatic activities: 5′–3′ exoribonuclease activity and 5′ end independent endoribonuclease activity (Even et al., 2005). Recent biochemical and structural studies show that the major enzymatic activity of RNase J1 is the 5′–3′ exoribonuclease activity (Condon, 2010; Newman et al., 2011). It has robust 5′–3′ exoribonuclease activity in vitro and the exoribonuclease activity is required to process 16S rRNA from the 5′ end in vivo (Mathy et al., 2007). However, the endoribonuclease activity was detected only in vitro with high protein concentration, so the endoribonuclease activity of RNaseJ1 is highly controversial (Condon, 2010). RNase J1 and J2 were shown to affect gene transcription on a global scale but not as much as RNase E (Mäder et al., 2008). Recently, RNase Y has been proposed to be the functional homolog of RNase E because of (i) its localization on the membrane, (ii) its interaction with other protein members in the RNA degradosome, (iii) its endoribonuclease activity with the same substrate preference as RNase E that prefers 5′ monophosphorylated RNA, and (iv) its significant influence on average half-life of bulk mRNA.

Before RNase J1 and J2 were discovered, it had long been believed that, unlike eukaryotes, bacteria did not possess enzymes with 5′–3′ exoribonuclease activity. For example, E. coli has three 3′–5′ exoribonucleases: PNPase, RNase R, and RNase II, but no 5′–3′ exoribonucleases. However, the discovery of RNase J1 changed this notion because many Gram-positive bacteria have RNase J1 orthologs. The substrates for the RNase J1 exoribonuclease activity are 5′ non- or monophosphorylated RNA (Mathy et al., 2007). Therefore, the substrates can be generated by the action of endoribonucleases that produce monophosphorylated 5′ end or the pyrophosphohydrolase RppH that removes diphosphate from 5′ triphosphate of nascent transcripts (Richards et al., 2011).

In B. subtilis, both RNase J1 and RNase J2 are not essential for growth but RNase J1 null mutants display more severe growth defect than RNase J2 null mutants (Figaro et al., 2013). The exoribonuclease activity of RNase J1 is 25–100 fold higher than that of RNase J2. (Mathy et al., 2010). RNase J1 and J2 form a heterotetramer (Mathy et al., 2010), and RNase J2 attaches to the degradosome only through the interaction with RNase J1 (Commichau et al., 2009). Many Gram-positive bacteria have both RNase J1 and J2, implying that these two proteins have their own unique roles in vivo, even though some of their roles could be overlapped.

PNPase, which is a 3′–5′ phosphorolytic exoribonuclease, is also a member of the RNA degradosome. B. subtilis has four 3′–5′ exoribonucleases: PNPase, RNase R, RNase PH, and YhaM. Among these, PNPase is the major exoribonuclease in the RNA degradosome. Unlike the other 3′–5′ exoribonucleases that use a hydrolytic mechanism, PNPase uses a phosphorolytic mechanism to degrade RNA, so PNPase activates an inorganic phosphate to attack the 3′ phosphodiester bond in RNA and produces a single diphosphate nucleoside as the result. PNPase is not essential for growth but appears to be crucial for cellular and metabolic activities. A pnpA mutant is cold sensitive, forms filaments, and loses competence (Luttinger et al., 1996; Wang and Bechhofer, 1996). When endoribonucleases cleave internal regions of transcripts, PNPase binds the endonucleolytically generated 3′ sites and degrades the transcripts (Figure 2; Deikus and Bechhofer, 2007). PNPase is processive, so it continues to cleave along the same RNA substrate. However, since the phosphorolytic mechanism generates little free energy change in RNA degradation, PNPase alone often cannot degrade structured RNA with extensive stem and loop structures. Hence, RNA helicases are required for the complete digestion of structured RNA (Coburn et al., 1999; Khemici and Carpousis, 2004).

Another component of the RNA degradosome is a DEAD-box RNA helicase. DEAD-box RNA helicases, which have D-E-A-D amino acid residues within RecA-like domains, act locally to unwind only a few base pairs of duplex RNA at the expense of ATP hydrolysis. Four ATP-dependent DEAD box helicases exist in B. subtilis. Among these, CshA can interact with RNase Y, PNPase, enolase, and phosphofructokinase (PFK). These interactions suggest that CshA is a member of the degradosome and the functional equivalent of RhlB that is the major DEAD-box helicase found in the E. coli degradosome (Lehnik-Habrink et al., 2010).

Among the major components of the RNA degradosome, only glycolytic enzymes are not directly involved in the degradation of RNA. In the E. coli degradosome, enolase is the only glycolytic enzyme in the RNA degradosome. However, two glycolytic enzymes, enolase, and PFK, are found in the B. subtilis degradosome. PFK is involved in one of the initial steps of glycolysis, which converts fructose-6-phosphate to fructose 1,6–bisphosphate. This step is a key regulatory step operated by the cell's energetic needs because it is activated by AMP and inhibited by ATP. Enolase converts 2-phospho-D-glycerate (2PG) to phosphoenolpyruvate (PEP), which not only produces ATPs but also transfers the energy-bearing phosphate to the sugar transport system in bacteria. The exact function of the glycolytic enzymes in the RNA degradosome is not yet known.

Recently, the intracellular localization of all proteins that are thought to be a part of the RNA degradosome has been studied (Cascante-Estepa et al., 2016). It was found that except RNase Y, which was localized to the membrane, all of other proteins were localized in the cytoplasm. This might indicate that the interaction among the RNA degradosome components is transient.

S. aureus is a Gram-positive coccal bacterium and a member of Firmicutes. Even though it is a part of human normal microflora found frequently in the nostril, respiratory tract, and on the skin, it can cause diverse human diseases with immense health care concern. Major infections caused by S. aureus range from skin infections such as abscesses and impetigo to life-threatening diseases such as pneumonia, endocarditis, toxic shock syndrome, and osteomyelitis. It is also a major nosocomial infection agent, and ~500,000 patients in hospitals contract this organism each year in the United States.

The interactions among the RNA degradosome components in S. aureus were studied through the bacterial two-hybrid analysis that is performed in E. coli (Roux et al., 2011). The S. aureus degradosome complex contains the same components as those of B. subtilis, such as RNase Y, RNase J1 and J2, polynucleotide phosphorylase (PNPase), enolase, PFK, and a DEAD-box RNA helicase. In addition, the RNase RnpA appears to be a member of the RNA degradosome complex in both S. aureus and B. subtilis (Roux et al., 2011). The interactions between these components showed conservation between the two organisms with some differences. In the S. aureus RNA degradosome, strong interactions appear between RNase Y and enolase, PFK and CshA, CshA and RnpA, CshA and Enolase, and RNase J1 and RNase J2. Thus, the DEAD-box RNA helicase CshA appears to play a crucial role in the S. aureus degradosome formation.

Khemici et al. examined the cleavage sequence by S. aureus RNase Y using the novel method EMOTE (Exact Mapping Of Transcriptome Ends; Khemici et al., 2015). The EMOTE assay is a mapping technique using high throughput sequencing to identify 5′ monophosphorylated end created by the action of endoribonucleases. The first step of this assay is a ligation of a synthetic RNA oligo to the 5′-ends of total RNA. Then, cDNA is synthesized using the sequence of the RNA synthetic oligo and sequenced through a high-throughput sequencing. This assay revealed that the preferred cleavage site by RNase Y is the 3′ side of a guanosine (58% of total cleaved sites) in adenosine and uridine-rich regions (Khemici et al., 2015). Both requirements of G and the A/U-rich region are not so strict, thus sites immediately next to other nucleotides can be cleaved (A, 30%; U, 10%; and C, 2%). Since the genome of S. aureus is AT rich (A+T = 67%), there can be many AU rich regions in transcripts. These results imply that other factors such as RNA structures or RNA-binding proteins might influence RNase Y to recognize its substrates.

S. aureus RNase Y is not essential for growth and the mutation of its gene rny stabilizes only some mRNAs (Marincola et al., 2012). Also, the growth of both the RNase Y deletion strain and the strain with the mutation in the active site in the HD-domain are similar to that of the wild-type strain (Khemici et al., 2015). These results suggest that the role of RNase Y in S. aureus might be not for the degradation of bulk RNA. The location of RNase Y appears important for its role because the strain with the transmembrane domain-deleted RNase Y grows markedly slower, implying that the degradosome formation in the cytosol is deleterious for the growth of S. aureus (Khemici et al., 2015).

In S. aureus, both RNase J1 and J2 are not essential for viability but deletion of one or both genes confers high fitness cost for growth, especially under the condition of nutritional or temperature stress (Linder et al., 2014). Interestingly, an active site mutation in RNase J1 displays the same growth defect as an RNase J1 deletion mutation, but an active site mutation in RNase J2 exhibits no growth defect. Based on this result, RNase J2 has been proposed to be a stabilizer or modulator of RNase J1 (Linder et al., 2014).

Another Gram-positive pathogen whose RNA degradosome components have been studied is S. pyogenes. S. pyogenes is a human pathogen that causes diverse diseases from superficial infections such as pharyngitis and impetigo to life-threatening systemic and invasive diseases such as rheumatic heart disease, toxic shock syndrome and necrotizing fasciitis (Cunningham, 2000). Almost 20% of the population carries this pathogen in their throat asymptomatically.

Before the existence of RNase Y was known, other RNase components in the degradosome such as PNPase, RNase J1, and RNase J2 were studied to examine the stability of virulence gene transcripts in S. pyogenes (Barnett et al., 2007; Bugrysheva and Scott, 2010). S. pyogenes produces many virulence factors and the regulation of these virulence factors are strictly controlled according to cellular or environmental signals. One of the signals that determine virulence gene expression in vitro is the growth phase. When the stability of transcripts of a M3 serotype strain, MGAS315, was examined, most virulence gene transcripts were rapidly degraded at the stationary phase. However, the stability of some transcripts of virulence genes such as sagA, sda, and arcT are greatly enhanced at the stationary phase (Barnett et al., 2007). These genes encode the precursor of the hemolysin streptolysin S, a DNase or Streptodornase, and a peptidase, respectively. In a PnpA null mutant, the transcripts of sagA, sda, and arcT are more stabilized at the late exponential phase. The transcripts of sagA, sda, and arcT show biphasic decay: slow decay at the initial period, then rapid decay after that. The pnpA mutation abolished the second phase of rapid decay, indicating that PNPase is responsible for the second decay at the late exponential phase.

In B. subtilis, RNase J1 and RNase J2 are paralogous and form a heterotetramer in vivo (Mathy et al., 2010). In S. pyogenes, both RNase J1 and RNase J2 are essential for growth, suggesting that, at least partially, their roles are not redundant and their substrates could be different in vivo (Bugrysheva and Scott, 2010). When RNase J1 and J2 are minimally expressed to support the growth of S. pyogenes, the half-lives of transcripts showing rapid decay at both exponential and stationary phase growth increase. However, in the case of the transcripts showing biphasic decay at the exponential phase and increased stability at the stationary phase, the period of the first phase of slow decay increases, but the second phase of rapid decay still maintains, unlike the pnpA mutation. This result indicates that PNPase and RNase J1/J2 work cooperatively and the action of RNase J1/J2 precedes that of PNPase in RNA decay.

The average half-life of S. pyogenes mRNAs is 1.26 min, so the turnover rate is very high even for a bacterium (Chen et al., 2013). In an RNase Y null mutant, mRNA stability increases two-fold and this indicates that RNase Y plays a major role in the global decay of mRNA in S. pyogenes (Chen et al., 2013). S. pyogenes RNase Y is also involved in the processing of certain mRNA such as the transcripts of speB, a cysteine protease secreted by S. pyogenes, and its positive regulator RopB. (Chen et al., 2012, 2013).

The structure of S. pyogenes degradosome and interactions among components have not yet been studied, except for the interaction between RNase Y and enolase (Kang et al., 2010).

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.