If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Conditional proteolytic degradation is an irreversible and highly regulated process that fulfills crucial regulatory functions in all organisms. As proteolytic targets tend to be critical metabolic or regulatory proteins, substrates are targeted for degradation only under appropriate conditions through the recognition of an amino acid sequence referred to as a “degron”. DEAD-box RNA helicases mediate all aspects of RNA metabolism, contributing to cellular fitness. However, the mechanism by which abiotic-stress modulation of protein stability regulates bacterial helicase abundance has not been extensively characterized. Here, we provide in vivo evidence that proteolytic degradation of the cyanobacterial DEAD-box RNA helicase CrhR is conditional, being initiated by a temperature upshift from 20 to 30 °C in the model cyanobacterium, Synechocystis sp. PCC 6803. We show degradation requires a unique, highly conserved, inherently bipartite degron located in the C-terminal extension found only in CrhR-related RNA helicases in the phylum Cyanobacteria. However, although necessary, the degron is not sufficient for proteolysis, as disruption of RNA helicase activity and/or translation inhibits degradation. These results suggest a positive feedback mechanism involving a role for CrhR in expression of a crucial factor required for degradation. Furthermore, AlphaFold structural prediction indicated the C-terminal extension is a homodimerization domain with homology to other bacterial RNA helicases, and mass photometry data confirmed that CrhR exists as a dimer in solution at 22 °C. These structural data suggest a model wherein the CrhR degron is occluded at the dimerization interface but could be exposed if dimerization was disrupted by nonpermissive conditions.
Although energetically costly and irreversible, proteolytic degradation offers a rapid mechanism for adjusting the proteome under specific conditions when the continued presence of select proteins would be deleterious (
). For free living bacteria, AAA+ protease activity serves a critical role in cellular adaptation in response to the numerous and rapidly fluctuating conditions experienced in their natural environments. Accordingly, bacterial proteases contribute to the regulation of several crucial pathways including the cell cycle, induction of virulence, circadian clock synchronization, and biofilm formation (
). Recognition of specific protein targets in conditional proteolytic pathways is often achieved through amino acid motifs termed “degrons” that direct proteins to one or more bacterial AAA+ proteases under nonpermissive conditions (
). Although proteomic and genetic approaches have expanded the catalog of bacterial proteins regulated by proteolytic degradation, the amino acid composition which constitute protein-specific degrons have rarely been experimentally verified. Detailed analysis of bacterial degrons has been hampered by their extended nature and general lack of conservation across bacteria, either in sequence or functionality (
). Degron sequences have been observed to vary considerably in the pathways targeted and also their length, composition, and location within a protein, indicating that our understanding of how protein degradation is regulated in bacteria is not complete. Examples illustrating bacterial degron diversity include an N-terminal 21 amino acid sequence containing a core region of hydrophobic residues surrounded by polar residues in the oxidative stress response regulator SoxS and, in contrast, a single histidine at the C terminus of the SOS response protein SulA required for targeting Lon-mediated turnover in Escherichia coli (
Sequence requirements for Lon-dependent degradation of the Escherichia coli transcription activator SoxS: Identification of the SoxS residues critical to proteolysis and specific inhibition of in vitro degradation by a peptide comprised of the N-terminal 21 amino acid residues.
Cyanobacteria are Gram-negative bacteria that constitute the only bacterial phylum that perform oxygenic photosynthesis, thereby representing a valuable platform for autotrophic production of bioproducts (
), only three examples of sequence-specific degron-mediated proteolytic degradation have been reported. UmuD possesses a dual N-terminal degron consisting of an extended 19 amino acid sequence that targets proteolytic degradation by both N-degron and ClpX/P-associated pathways in Synechocystis sp. PCC 6803 (hereafter Synechocystis) (
). It can be deduced that degron-mediated proteolytic degradation likely plays a significantly expanded role in cyanobacteria, since they encode multiple protease duplications with nonoverlapping and essential physiological functions (
). However, these characteristics of the cyanobacterial proteolytic complement make it difficult to identify both degron sequences and the interacting proteases. Thus, identification of degron-regulation of posttranslational gene expression deserves further attention in cyanobacteria.
Previously, we reported that enhanced protein stability contributed significantly to the constantly elevated abundance of the cyanobacterial RNA helicase redox (CrhR), the only DEAD (Asp-Glu-Ala-Asp)-box RNA helicase encoded in Synechocystis, observed at low temperature (
). In response to the permissive abiotic stress, cold shock at 20 °C, crhR mRNA rapidly and transiently accumulates resulting in a 15-fold enhancement in CrhR protein abundance. Furthermore, CrhR continues to remain elevated at low temperature, despite the absence of crhR transcript, resulting from a significant increase in CrhR protein half-life. Repression of the low temperature induction was initiated by a temperature upshift to 30 °C resulting in conditional destabilization of CrhR and rapid degradation, leading to the basal levels observed at higher, nonpermissive temperatures (
). This study also showed that a truncated, biochemically inactive mutant of CrhR, CrhRTR, was resistant to degradation during temperature upshift, suggesting an autoregulatory mechanism. While the degradation mechanism was not investigated further, the data suggested that either a targeting sequence necessary for proteolytic degradation had been deleted in CrhRTR and/or that CrhR RNA helicase activity was a requirement to activate the proteolysis machinery.
Physiologically, although initial transcript and proteomic analysis of WT, crhRTR, and ΔcrhR strains of Synechocystis have implicated CrhR in gene expression vital for supporting phototrophic metabolism, particularly at low temperatures, it remains unclear how dynamic regulation of CrhR at the posttranslational level is achieved or ultimately why it is necessary. The biochemical, genetic, and evolutionary principles governing control of CrhR abundance by the cell are therefore of prime interest in contextualizing RNA helicases within stress biology programming (
Here, we utilized an in vivo strategy to systematically identify a degron within the unique C-terminal extension (CTE) of CrhR that is essential for temperature upshift-mediated proteolytic degradation. Additionally, bioinformatics analysis indicated evolutionary conservation of the identified degron sequence in RNA helicases that are only encoded in the phylum Cyanobacteria. Proteolytic regulation of such “CrhR-type” RNA helicases may share a common mechanism and/or biological role, as they exhibited evolutionarily conserved temperature regulated induction and degradation kinetics in diverse cyanobacteria. Interestingly, although the CrhR degron sequence is necessary, it is not sufficient for proteolytic degradation as helicase function and de novo protein synthesis are also required. The results imply protease degradation of CrhR is autoregulatory through a positive feedback mechanism whereby CrhR enables translation of an unknown degradation factor(s). Induction and repression of proteolysis is also exquisitely sensitive to the magnitude of temperature shift, resulting in fine-tuning of CrhR abundance. Finally, mass photometry indicated CrhR exists as a dimer at 22 °C. AlphaFold 3D structural modeling predicted that the CTE is responsible for this dimerization behavior resulting in the degron being buried inside the interface, likely impeding recognition and degradation. Identification of this cyanobacterial-specific degron may provide a biotechnological tool to fine-tune posttranslational gene expression across multiple cyanobacterial genera.
The CTE of CrhR-type RNA helicases contains an evolutionarily conserved temperature-responsive degron in cyanobacteria
It was of interest to determine if CrhR-related proteins were temperature regulated in other cyanobacteria. Alignment of the CrhR amino acid sequence downstream of the conserved HRIGR box indicated the presence of an ∼50 amino acid sequence spanning amino acids K386–W435 that was highly conserved in 22 orthologous helicases (Fig. 1A). The strength of this cross-species conservation is illustrated by the sequence logo generated from this region using Multiple Em for Motif Elicitation (MEME) software (Fig. 1B), produced using the cyanobacterial species listed in Table S1. Similar cold-induced accumulation combined with temperature upshift repression of CrhR-related RNA helicases was demonstrated in four representative cyanobacterial species (Fig. 1C). Note that the temperature shift response observed for CrhR in WT cells shown in Figure 1CSynechocystis exhibits a linear pattern of degradation, as observed previously (
). The kinetics of temperature downshift induced changes in CrhR-related protein abundance are presented in Figure 1D. Importantly, CrhR-like proteins in the four cyanobacterial species tested responded in a similar manner to temperature shift, increasing at 20 °C or 10 °C and decreasing at 30 °C. While the timelines and induction temperatures differed slightly between the four strains, the kinetics of induction and repression were consistent. The results suggest that an evolutionarily conserved, temperature responsive degradation motif is located in the CTE of CrhR-like cyanobacterial RNA helicases.
CrhR degradation depends on residues within the CTE
DEAD-box RNA helicases share a number of conserved motifs, including the characteristic DEAD motif, which comprise the highly conserved RNA helicase core (
). To explore the possibility that a degron sequence within the CTE or NTE of CrhR is responsible for temperature upshift-mediated degradation, a series of deletions were generated in the crhR gene (Fig. 2B). Since deletion-induced alteration of protein folding could contribute to mutant polypeptide degradation or aggregation, we produced a diverse range of five N- or C-deletion constructions to examine. The most extensive C-terminal deletion, CrhRΔR228-Q492, corresponds to the CrhRTR mutation, which produces a truncated polypeptide that can be detected readily in purified membrane and soluble Synechocystis cell lysis fractions (
), implying that even gross removal of the CTE does not produce instability or insoluble precipitates. Examination of temperature upshift induced degradation of the protein products expressed from the five constructs in WT Synechocystis indicated that an amino acid sequence between E393 and K449 was required for proteolytic degradation of the mutant protein (Fig. 2, C–F). This domain is located within the highly conserved C-terminal region identified in Figure 1, A and B.
Insightfully, the presence of mutant, nondegradable versions of CrhR did not affect degradation of WT CrhR nor did WT CrhR rescue degradation of mutated CrhR (Fig. 2, C–F). Again, a similar pattern of WT, genome-encoded CrhR degradation as shown in Figure 1CSynechocystis was observed, irrespective of the fate of the mutant form of CrhR present in each strain (Fig. 2, C–F).
Overall, the results shown in Figure 2, C–F and Fig. S1 provide crucial insights since they indicated that WT, biochemically active CrhR is required to activate degradation, but only of polypeptides that contain the degron. In addition, they demonstrate that plasmid-encoded CrhR polypeptides containing the degron are actively degraded, indicating analysis of the fate of CrhR polypeptides expressed from plasmids provide valid observations.
To better facilitate comparisons across strains or growth conditions, the linear nature of the decay curve for genome-encoded CrhR from Synechocystis (Fig. 1D) prompted us to assign linear degradation rates to all WT and mutant CrhR polypeptides. This analysis was performed by fitting linear regression models to abundance curves for either chromosomal expressed CrhR (Fig. S1A) or plasmid-encoded CrhR C-terminal deletions (Fig. S1B) in WT cells expressing functional CrhR. Linear regression models encompassing each time point from the entire 30 °C time courses during which degradation was occurring were found to closely correlate with abundance in all circumstances (Fig. S1A). Slopes of these regression lines, thus representing independent measurements of WT or mutant CrhR change in relative abundance over the entire time course, then enabled simultaneous visualization of pairwise comparisons in subsequent western analysis.
Examination of these degradation rate results indicated that the rate of change of chromosomal, WT CrhR does not vary appreciably in the presence of the CrhR-deletion constructs (Fig. 2G WT CrhR and Fig. S1A). In contrast, while degradation of the CrhRΔK449-Q492 mutant matched WT CrhR, abundance changes of the CrhRΔE393-Q492, CrhRΔA347-Q492, and the CrhRΔR228-Q492 deletion polypeptides varied significantly from the WT rate, indicating the degron motif had been deleted in these constructs (Fig. 2G Deletion CrhR constructs and Fig. S1B). Overall, the data indicated that the CrhR degron motif is located between E393 and K449 (Figs. 2, C–G and S1).
Delineation of a minimal core degron
A series of short, internal deletions were constructed within the 50 amino acid conserved region within the C-terminal motif to further delineate the location of the degron. Deletions in this region included 26 amino acids from the N terminus of this domain, CrhRΔK386-S411, 24 amino acids from the C-terminal portion, CrhR ΔD412-W435, and the entire domain, CrhRΔK386-W435. As a control, temperature-dependent CrhR accumulation in WT cells revealed the expected low temperature induction and subsequent temperature upshift repression (Fig. 3A), as observed in Figure 1CSynechocystis. Expression of the deletion constructs in WT cells indicated that while CrhR accumulation was relatively normal from the CrhRΔK386-S411 construct (Fig. 3B), temperature regulation of the CrhR ΔD412-W435 (Fig. 3C) and CrhRΔK386-W435 (Fig. 3D) constructs was significantly diminished.
Quantification of CrhR abundance changes in response to temperature upshift is provided in Figure 3E. As observed in Figures 1 and 2, C–F, the rate of change of genomic, WT CrhR did not vary appreciably in the presence of the fine CrhR-deletion constructs (Fig. 3E WT CrhR). In contrast, while degradation of the CrhRΔK386-S411 polypeptide did not differ from the plasmid-encoded CrhR, abundance change trends of the CrhRΔD412-W435 and CrhRΔK386-W435 deletion polypeptides were abnormal, and degradation was significantly impaired (Fig. 3E mutant CrhR). These results suggested unequal contributions of the two segments of the 50 amino acid motif to CrhR regulation and that the CrhRΔD412-W435 motif contains a sequence that, when absent, results in muted polypeptide turnover at 30 °C. Interestingly, the CrhR deletion mutants resistant to proteolytic degradation, CrhR ΔD412-W435 and CrhRΔK386-W435, unexpectedly limited cold-induced induction of genomic-encoded CrhR at 20 °C but had no effect on temperature upshift degradation (Fig. 3, B–D).
While deletion mutant suppression of WT CrhR accumulation was observed at all time points, we selected the point of maximum induction, 3 h at 20 °C, to quantitate the relative accumulation of plasmid-encoded mutant CrhR polypeptides to the abundance of the genomic-encoded WT polypeptide (Fig. 3E Ratio plasmid:genomic CrhR). As the functionality of the degradation domain decreased from CrhRΔK386-S411 to CrhRΔD412-W435 and CrhRΔK386-W435, the ratio of mutant to WT CrhR increased significantly (Fig. 3E Ratio plasmid:genomic CrhR). Insightfully, this observation indicated that chromosomal CrhR abundance is proportionately reduced in response to the presence of increasing nondegradable CrhR. These results are the inverse of the stability seen for the respective mutations and suggested that the deletion polypeptides negatively feedback on WT, genome-encoded CrhR expression through an unknown mechanism. Overall, the results indicate that the primary degron motif was located between D412 and W435.
These intriguing observations prompted further examination of temperature regulation of the internal deletion constructs in a complete crhR deletion mutant, ΔcrhR, to determine if proteolytic profiles differed in the absence of WT, genome-encoded RNA helicase activity. Similar to the pattern observed in Figure 1CSynechocystis, normal temperature-shift induction and repression were observed when the sole source of WT CrhR was expressed from a plasmid (Fig. 4A). In comparison to degradation observed in Figure 1CSynechocystis and Figure 2, C–F, for WT, genome-encoded CrhR, while essentially identical temperature regulation was observed for the CrhRΔK386-S411 mutant (Fig. 4B), both temperature downshift and upshift induction and repression of the system were absent in the CrhRΔD412-W435 (Fig. 4C) and CrhRΔK386-W435 (Fig. 4D) strains. Given the high degree of conservation and location within the essential C-terminal portion of the CrhR motif, a further deletion encompassing the region between Q417 and Q425 (CrhRΔQ417-Q425) was generated in an attempt to define a minimal degron. The importance of the Q417-Q425 region for the functional degron was emphasized by the partial loss of temperature regulation (Fig. 4E). The NTE was not significantly associated with degradation since absence of the NTE2–26 had no effect on induction or repression in ΔcrhR cells (Fig. 4F). Quantification indicated that while the CrhRΔK386-S411 and NTE2–26 polypeptides degraded at a rate similar to WT CrhR, the rate of CrhRΔQ417-Q425 degradation was intermediate and CrhRΔD412-W435 and CrhRΔK386-W435 degraded at a significantly reduced rate (Fig. 4G).
Taken together, the data indicated the presence of an extended degron within the CrhR CTE, CrhRΔD412-W435, with residues spanning Q417 and Q425 specifically performing a major role in the degradation mechanism.
RNA helicase activity is required for degradation
We then asked if CrhR RNA helicase activity was required to initiate a robust proteolytic response. The location of three point mutations, K57A (Walker ATPase box A), E156Q (DEAD box II), or R335A (HRIGR box VI), known to inhibit RNA helicase biochemical activity (
) are depicted in Figures 2, A and B and 5A. Addition of an N-terminal 6X-His tag was utilized to differentiate the biochemical mutants from full-length CrhR in WT cells. Fusion of an N-terminal 6X-His tag is known to have no discernable effect on CrhR RNA helicase activity (
). None of the biochemical mutants affected temperature induction or repression of either mutant or WT CrhR expression in WT cells expressing native, functionally active CrhR (Fig. 5B). In sharp contrast, all three biochemical mutants abolished both induction and repression in ΔcrhR cells (Fig. 5C). These results suggest an autoregulatory mechanism whereby RNA helicase activity is a co-requisite for degradation in addition to the presence of the degron. Separate quantification of the abundance of WT, genome-encoded CrhR and the plasmid-expressed 6X-His-CrhR point mutants is shown in Figure 5D. As shown in analysis presented above, WT CrhR abundance was again not altered in the presence of the CrhR point mutant polypeptides (Fig. 5D WT CrhR). Similarly, abundance of the point mutants decreased in the presence of WT CrhR although at a slower rate (Fig. 5D MT CrhR). In comparison, temperature shift induction and repression of the three-point mutants was abolished in the absence of WT CrhR (Fig. 5D ΔcrhR cells MT CrhR).
CrhR repression can be activated by removal of CrhR inducing abiotic stresses
Previous evidence from our lab indicated that CrhR induction occurs in response to a variety of abiotic stresses that regulate induction via a common effect on the redox potential of the electron transport chain, independent of temperature shift (
). As a result, it was of interest to determine if temperature-independent degradation of CrhR, similar to that observed during temperature upshift, occurred in response to removal of these abiotic stresses. CrhR abundance was assessed following removal of NaCl and sorbitol stress at 30 °C, in the absence of low temperature stress (Fig. 6).
A similar rate of degradation as that observed in control WT cells expressing genome encoded in response to a 20 to 30 °C upshift (Fig. 6A) was observed upon removal of both NaCl and sorbitol stress in the absence of temperature stress (Fig. 6B NaCl and Fig. 6C sorbitol), as quantified in Figure 6D Quantification. Overall, the results agree with and extend previous work suggesting that the sensor for CrhR turnover can respond to a range of abiotic stress conditions in addition to temperature fluctuation.
CrhR proteolysis is adaptive and relies on sustained translation elongation
We have previously shown that antibiotics blocking either translation initiation or elongation have divergent effects on the in vivo degradation of CrhR (
). Here, it was explored whether translational interference could influence proteolytic degradation once cold shocked cells had already initiated degradation in response to an upshift to 30 °C (Fig. 7). First, in the absence of antibiotics, the expected pattern of rapid temperature-downshift induction and temperature-upshift induced degradation was observed (Fig. 7A). Based on the decrease in CrhR abundance after 30 min observed in Figure 7A, translational inhibitors were added at this time point for all subsequent treatments. Looking initially at the temperature upshift effects under these conditions, it was observed that further degradation of CrhR could be effectively blocked by chloramphenicol but not by kanamycin treatment (compare Fig. 7, A–C). Therefore, continued translation elongation but not initiation was essential for maintaining CrhR degradation. Furthermore, addition of the transcriptional inhibitor rifampicin failed to prevent CrhR turnover, signifying de novo transcription is not required to maintain proteolysis but was required for subsequent temperature downshift induction (Fig. 7D). These interpretations were confirmed by quantification of the relative rate of degradation that indicated a similar rate to WT CrhR in cells treated with kanamycin and rifampicin, but a rate that was significantly reduced in chloramphenicol-treated cells (Fig. 7E).
These cultures were also used to extend this analysis where the interplay between CrhR degradation and induction was examined for translational inhibitor treated and untreated cells in response to subsequent temperature downshift to 20 °C. Interestingly, proteolysis was rapidly reversed in response to temperature downshift in control cells in the absence of inhibitors with normal CrhR induction recovering within 15 min after temperature downshift to 20 °C (Fig. 7A). Kanamycin permitted a minimal level of low temperature induction, suggesting protein synthesis was incompletely blocked at the inhibitor concentration used (Fig. 7B), indicating that translation was not completely inactivated. In contrast, chloramphenicol and rifampicin abolished low temperature induction (Fig. 7, C and D). The rate of decline in rifampicin matches the degradation rate observed in WT cells at 30 °C (Fig. 7E). This illustrates that de novo transcription and proteolysis act in concert to produce growth condition appropriate levels of CrhR.
These observations crucially indicated that proteolytic degradation was not an all or none process and is susceptible to rapid inactivation at the permissive temperature, 20 °C. Overall, CrhR degradation is a dynamic process contingent on continued transcription and protein synthesis coordinated by a signal propagated by incubation at the higher, nonpermissive temperature.
The rate of CrhR degradation is temperature dependent
Further insights into the temperature dependence of CrhR proteolysis were obtained by determining the rate of degradation in response to increasing magnitude of temperature upshift as shown in Figure 8. Although CrhR performs a vital physiological role in the cell at 20 °C, CrhR also functions at 30 °C as evidenced by the growth defect in ΔcrhR cells observed at 30 °C (
). The kinetics of CrhR turnover have only been exhaustively characterized for the 20 to 30 °C transition. Here, aliquots of a 10 °C cold shocked culture were subjected to a time course analysis after temperature upshift directly to either 20 °C, 30 °C, or 40 °C to evaluate degradation dynamics (Fig. 8A). A shift from 10 to 20 °C did not elicit a significant proteolytic response (compare Fig. 8A 10 and 20 °C). Temperature upshift to 30 °C or 40 °C were significant enough, however, to elicit degradation of CrhR, suggesting that a threshold temperature exists between 20 °C and 30 °C that must be exceeded to elicit CrhR destabilization. Above this threshold, cells are sensitive to the degree of upshift, permitting establishment of a higher basal level of CrhR at 30 °C versus 40 °C (Fig. 8A). How this basal level is reached could be a product of either the amount of time proteolysis is active or differential magnitude of the degradation activity. Quantification of CrhR abundance for cells shifted from 10 to 20 °C reveals a modest rate of decline, while a 10 to 30 °C shift significantly enhances degradation (Fig. 8B). This rate was further stimulated in response to a 10 to 40 °C shift, resulting in the enhanced basal level observed at 30 °C versus 40 °C (Fig. 8B). Thus, the mechanism responsible for CrhR degradation must convey a detailed input regarding the magnitude of temperature upshift into an appropriate level of proteolytic activity to achieve the required CrhR abundance at each temperature.
AlphaFold predicts the CrhR CTE is a dimerization domain
To better understand the CTE of CrhR, the structure of residues 375 to 427 was predicted by the protein folding algorithm AlphaFold2 using the ColabFold implementation (
). AlphaFold returned a dimeric model that consisted of two bundles of three helices each, that fold against each other (Fig. 9A). Side and top depictions of the model structure are shown in Figure 9, A and B with the interacting helices H1–H3 labeled from the N to C terminus.
AlphaFold predicts this model with high confidence by both predicted local-distance difference test (Fig. S2A) and the predicted alignment error (Fig. S2B). Indicating high local confidence around each amino acid and position/orientation confidence between the monomers, respectively (
). The model was also compared to known structures of DEAD-box helicase dimerization domains from cold-shock DEAD-box protein A (CsdA), cold shock helicase A, and heat resistant RNA-dependent ATPase (Fig. 9, C and D) (
). And, in spite of low sequence identity (Fig. S3), the model aligns well with an average RMSD of 3.8 Å (Fig. S2). To verify the prediction that CrhR is a homodimer, the oligomeric state in solution was determined by mass photometry. Mass photometry estimates molecular weights of proteins by detecting light scattering as single molecules interact with a glass coverslip. The intensity of the light scattering is compared to known molecular weight standards to obtain the estimates. Over time, a population is generated, and gaussian curves are fit to the population to identify individual species (
). The major species observed accounted for 93% of the protein in solution and had a MW of 107 kDa ± 15 kDa (Fig. 9E). This is within ∼9% error of the deduced dimer molecular weight (111.88 kDa) indicating that CrhR exists predominantly as a homodimer in solution.
Finer AlphaFold modeling of residues 215 to 445, corresponding to the second RecA domain (RecA2) and the entire CTE, was performed to predict the level of occlusion of the C-terminal degron sequence D412-W435. The degron is shown in red and is more solvent exposed than the subsequence Q417-Q425, shown in orange (Fig. 9F). However, despite extending past the end of H3, residues S431-W435 are predicted to pack against the RecA2 domain with an average predicted local-distance difference test score of 83.6 (Fig. S2, C and D). This tight packing would pose a steric hindrance for recognition of the degron.
Taken together, the AlphaFold predicted structure suggests that the CrhR C-terminal degron forms a three-helix bundle that homodimerizes to form a structure similar to the dimerization domains observed in the bacterial DEAD-box RNA helicases CsdA, cold shock helicase A, and Hera.
Expression of CrhR, the sole DEAD-box RNA helicase encoded in the Synechocystis genome, is induced by a variety of abiotic stresses and repressed by proteolysis, a conditional mechanism induced by removal of the stress (
). Proteolytic degradation represents a powerful yet flexible method for shaping the proteome in all cells, not just through routine protein turnover but also by way of targeted degradation influencing crucial cellular pathways. Through the remodeling of RNA secondary structure, RNA helicases act as important regulators of all aspects of RNA metabolism, impacting processes ranging from transcription to translation and degradation (
). Here we show in vivo that conditional degradation of CrhR requires an inherent 24 amino acid sequence, CrhRD412-W435, located in a conserved 50 amino acid CrhR-specific motif, CrhRK386-W435 within the CTE. AlphaFold structural analysis predicted that access and thus activity of this degron motif would be restricted by occlusion in a dimeric species. In addition, CrhR was required for the translation of an unknown degradation factor indicating that activity of the degradation machinery was autoregulatory. Thus, from an evolutionary perspective, the CrhR CTE is unique, encompassing the identified degron within a highly conserved 50 amino acid domain that defines a separate clade of DEAD-box RNA helicases found only in cyanobacteria (
Using a phylogenetically guided deletion approach, we show that the CrhR degron is composed of an inherent, bipartite structure whose regions do not contribute equally to degradation. CrhRK386-W411 appears to function in a supportive role, possibly as a regulatory or structural element while CrhRD412-W435 contains the degron sequence. The presence of both domains was required for WT levels of degradation.
The dual CrhR degron therefore appears to function in a manner similar to the DnaA (
) degrons, composed of functionally independent regions. Additional structure-function analysis, including fusion of the CrhR degron regions identified here to reporter genes, will be essential to further dissect the amino acid sequence participation in the proteolytic mechanism.
CrhR is a homodimer
AlphaFold in silico analysis predicted and mass photometry in vitro at 22 °C confirmed that CrhR is a homodimer. AlphaFold structural modeling also revealed significant structural similarities between the CrhR CTE motif, CrhRK386-W435, and the CsdA dimerization domain (
). In CsdA, dimer stability is temperature dependent owing to differential activity of the CsdA dimerization domain, with stable associations observed at 24 °C that discourage monomer switching while incubation at 37 °C resulted in rapid monomer switching. The observed structural conservation between CsdA and CrhR is expected to be indicative of the functionality of the CrhR CTE motif as well and thereby suggests a regulatory mechanism by which the CrhR degron could be exposed and thus active under nonpermissive conditions. The dimer species was stable over all conditions assayed, but we predict that temperature upshift or salt concentration could result in protein unfolding, revealing the degron sequence to initiate degradation. Dimerization-mediated autoregulation of CrhR stability would constitute an additional level of control, allowing rapid fine-tuning of CrhR abundance and thus activity, enhancing cyanobacterial fitness in response to changing environmental conditions. However, these are predictions, and further study is required to elucidate the dynamics of CrhR structure under these conditions.
Conditional degradation of CrhR involves an autoregulatory mechanism
CrhR autoregulation of its own expression has previously been reported at a number of levels (
). Here, we extend these observations by showing that autoregulation performs crucial roles in the degradation mechanism. Physiologically, alteration of CrhR degradation by both antibiotic inhibition and biochemical inactivation of helicase activity supports a scenario that implicates CrhR autoregulated translation of a factor required for degradation.
Biochemically, translation autoregulation may involve CrhR-mediated changes in gene expression initiated by RNA secondary structure rearrangements catalyzed by dsRNA unwinding, ssRNA annealing, and strand exchange (
). Related, it is also insightful that plasmid-based enhanced CrhR abundance proportionately decreased WT CrhR expression from the genome. Although the mechanism is unknown, it appears to involve an autoregulatory, negative feedback system and suggests that the cell can actively monitor levels of CrhR. This unexpected outcome further emphasizes that CrhR abundance is tightly regulated within an appropriate window as dictated by environmental conditions. These predictions were substantiated by the observation that CrhR interacts with the crhR transcript in pull-down experiments (
). A pathway that resembles the proposed thermal-induced structural regulation of CrhR is observed for the DNA binding protein RovA that acts as a dimeric thermosensor to modulate virulence in Yersinia pestis (
). RovA functions as a transcription factor between 20 °C and 25 °C, enacting virulence programming necessary for host invasion and infection. A temperature increase to 37 °C, indicative of a mammalian host, causes a conformational shift that impairs RovA DNA binding and promotes proteolysis by Lon/Clp proteases. Dimerization of RovA monomers is achieved through extensive interactions between α helices, similar to the CsdA dimerization domain. While a number of thermosensing proteins have been described, the mechanisms generally do not involve degradation but are reversible and most frequently observed in pathogenic bacteria (
Although their applicability to Synechocystis remains to be fully elucidated, insights from other model systems reveal that DEAD-box helicases are dynamic regulators that perform fundamental roles in assembly and/or activity of duplex RNA and/or RNP complexes thereby influencing all stages in the lifecycle of an RNA (
). In examining the physiological importance of CrhR, omics analysis has indicated that the absence of functional CrhR RNA helicase activity disrupts the expression of genes associated with photosynthetic and translation competency, particularly during cold stress, leading to a severe cold stress phenotype (
). Such a role explains the necessity for CrhR induction, but not necessarily degradation. Clearly however, our data reveal that Synechocystis is highly sensitive to CrhR abundance, and modulation of genomic CrhR expression can occur in response to stable, persistent forms of CrhR.
Functionally, recent reports indicating that deletion of crhR decreased clpP and clpR transcript accumulation at 30 °C (
) potentially implicate these subunits as components of the proteolytic machinery. The known localization of CrhR and Clp/FtsH proteases on the thylakoid membrane could then catalyze degradation at the site of CrhR activity (
From an evolutionary perspective, the CrhR degron and the associated degradation machinery define a unique regulatory mechanism specific to cyanobacteria. Characterization of this inherent degron, controlling the autoregulatory proteolytic degradation of CrhR-like RNA helicases, identifies a novel mechanism of conditional proteolysis in bacteria. In addition, RNA helicase activity has not previously been implicated in proteolytic pathways, nor has degron-mediated degradation been identified as a regulator of DEAD-box RNA helicase abundance in other systems. In conjunction, and from an evolutionary perspective, low-resolution small angle X-ray scattering data have recently identified that the eukaryotic RNA helicase DDX21 contains an AAXL motif and displays dimerization in solution (
). Thus, similar domains exist in other Kingdoms and the potential for thermal regulation of oligomer state and polypeptide stability in these RNA helicases remain to be elucidated.
Since CrhR turnover exhibits considerable flexibility in both the environmental inducing signal and magnitude of the response, the CrhR degron represents a promising biotechnological module for fine-tuning heterologous protein stability in cyanobacteria containing CrhR-related proteins. Although our data connect temperature shift-induced changes in CrhR helicase abundance to the sequence and structure of the CTE, future work will be required to further clarify how the CrhR degron is connected to the physiological function of CrhR type helicases and why the intracellular concentration of these proteins is so tightly regulated.
Bacterial strains and culturing
Cyanobacterial and E. coli strains used in this study are listed in Table 1. The freshwater photoautotrophic cyanobacterial strains used in this study, Synechocystis sp. strain PCC 6803, Nostoc sp. strain PCC 7120, and Leptolyngbya sp. 696 were maintained on BG-11, while the marine Synechococcus sp. strain PCC 7002 was maintained on A+ medium supplemented with Vitamin B12 (
). Cells were grown in liquid BG-11 or A+ media at 30 °C with continuous shaking (150 rpm) coupled with bubbling with humidified air at an illumination of 50 μmol photons m−2 s−1. Transformed Synechocystis cultures were supplemented with either gentamycin (10 μg/ml) or with a combination of gentamycin and kanamycin (50 μg/ml) for ΔcrhR strains. Plasmids were propagated in either DH5α or XL-1 Blue strains of E. coli using gentamycin (10 μg/ml) supplemented Luria-Bertani broth.
Table 1Oligonucleotide, plasmid, and bacterial strain information
). To remove the constitutive nirA promoter, pMON 36456 was digested with NdeI and StuI, overhangs filled in with Klenow DNA pol and self-ligated to create ΔPpMON. A region containing the entire crhR ORF, including the native 5′ and 3′ UTRs, was cloned into ΔPpMON from cs0096-9 using EcoRI and XbaI to create ΔPpMON-CrhR (
). The 316 base pair operon promoter from upstream of slr0082 was amplified using dcsr30f and dcsr30r (Table 1), digested with NotI and EcoRI and ligated into ΔPpMON-CrhR using the same restriction sites to create plasmid B1.
The B1 plasmid was used as a template to generate the N- and C-terminal crhR deletions by inverse PCR. For the N-terminal deletion, the initiator Met was left intact and T2-G26 were deleted, leaving a phenylalanine as the second amino acid that did not destabilize the mutant protein. For inactivation of RNA helicase core motifs, pMON-His-CrhR was used as a template in QuikChange PCR to introduce point mutations (
). Plasmids were first transformed into E. coli DH5α carrying the helper plasmid pRL623 prior to transfer to Synechocystis using E. coli DH5α cells containing the conjugative plasmid RP4. The transformation mixture was spread on BG-11 agar plates supplemented with 5% LB and incubated under low light for 2 to 3 days. Bacteria were washed off these plates with BG-11 and spread on selective BG-11 plates supplemented with increasing gentamycin concentration (2, 5, and 10 μg/ml).
Sequences of 22 cyanobacterial DEAD-box RNA helicases with homology to CrhR (Table S1) were identified by a BLASTp search of the NCBI nonredundant protein sequences database, restricted to cyanobacteria (taxid:1117), with Synechocystis CrhR (BAA10556.1) as the query sequence. These sequences were aligned in MEGA 7 (version 7.0.21) (
To examine the degradation characteristics of CrhR mutants within Synechocystis, WT or ΔcrhR transformants were grown at 30 °C to mid-log phase (A750 ∼ 0.4), transferred to 20 °C for 3 h to induce maximal CrhR accumulation and then returned to 30 °C to initiate CrhR degradation. For analysis of CrhR ortholog stability, Synechocystis, Nostoc, Lyngbya, and Synechococcus were grown to mid-log phase at 30 °C and transferred either to 20 °C for 3 h (Synechocystis, Synechococcus) or 10 °C for 24 h (Nostoc, Lyngbya) to induce helicase protein accumulation. Cultures were returned to 30 °C and sampled at the indicated times. Cells were harvested by centrifugation at the growth temperature, and cell pellets stored at −80 °C prior to Western blot analysis.
To assess the effect of a temperature gradient on CrhR degradation kinetics, a WT Synechocystis culture grown at 30 °C to mid-log phase was placed at 10 °C for 24 h. The culture was aliquoted, and the flasks transferred immediately to either 20 °C, 30 °C, or 40 °C for 6 h. Cells were collected by centrifugation at the indicated times at the stated growth temperature, and cell pellets were stored at −80 °C.
Influence of additional stresses on CrhR repression
Aliquots of a mid-log phase, WT Synechocystis culture grown at 30 °C were subjected to salt (600 mM) or sorbitol (600 mM) (
) stress in the absence of temperature stress at 30 °C for 3 h. A single aliquot was transferred to 20 °C to act as a control. Cells were collected on Durapore 0.22 μM GV membrane filters by vacuum using an EMD Millipore filter apparatus connected to a Trivac D4A pump. Cells were extensively washed with and suspended in fresh BG-11 media and incubation continued at 30 °C for 3 h to observe CrhR degradation in the absence of stress.
Effect of translational and transcriptional inhibitors on CrhR degradation
Synechocystis WT cells were grown to mid-log phase at 30 °C and transferred to 20 °C for 3 h to induce maximal CrhR protein accumulation. Aliquots were transferred to 30 °C for 30 min, at which time chloramphenicol (250 μg/ml), kanamycin (200 μg/ml), rifampicin (400 μg/ml) or nothing (control) was added. Cultures were subsequently incubated for an additional hour at 30 °C prior to transfer to 20 °C for 1 h. Aliquots for protein isolation were harvested at the indicated times before and after either cold induction or antibiotic treatment, and the cell pellets were stored at −80 °C.
Protein extraction and western detection
Soluble protein extraction and western analysis were performed as described previously (
). After Bradford standardization, aliquots of the clarified soluble fraction corresponding to the masses indicated were separated by SDS–10% PAGE and immunoblot detection of CrhR performed using anti-CrhR antisera (1:5000) and secondary antibody (anti-rabbit IgG at 1:20,000 dilution; Sigma-Aldrich) using enhanced chemiluminescence (ECL) (
). When indicated, Rps1 was used as a protein loading control. Representative data are shown from a minimum of three biological replicates.
Imaging of CrhR protein levels was conducted using either X-ray film (Fujifilm) or a ChemiDoc MP Imaging system (Bio-Rad). Quantification was performed using Image Studio Lite software (LI-COR), with Coomassie stained gels to correct for protein loading. A ratio representing the relative intensity of CrhR in each lane was obtained by dividing the corrected signal in that lane by the signal in the lane corresponding to the maximum induction, typically 3 h at 20 °C. Maximum induction required 24 h at 10 °C in the case of Leptolyngbya and Nostoc.
Statistical assessment of degradation
GraphPad Prism, version 9.1.1, for Windows (GraphPad Software) was used to visualize and then statistically analyze CrhR abundance data. Relative degradation rates were derived from averaging linear regression model slopes produced for all biological replicates from the corrected abundance values corresponding to the downregulation portion of each time course observed at 30 °C and using the level observed at 20 °C for 3 h or 24 h at 10 °C, as the maximum induction (T = 0 set to 1.0). Mean degradation rates were assessed for difference from the control condition using one-way ANOVA followed by a Dunnett’s multiple comparisons test.
). The sequence of either residues 375 to 427 or 215 to 445 was input along with the input parameters listed in Table S2 into the AlphaFold2_MMseqs2 Google Colab notebook. Multiple sequence alignments were found using the MMseqs2 algorithm and were input along with the protein sequence to generate five models. The models were ranked by pTM score, and the highest scoring model was used for analysis.
Pymol (Version 2.4.0, Schrödinger, LLC) was used to compare known RNA helicase structures (PDB codes: 5IVL, 3EAS, 5GI4) with the AlphaFold models and for creating images. Confidence metrics were analyzed using python3 scripts and plotted with Gnuplot (Version 5.4, http://www.gnuplot.info/).
Model 375 to 427: This region was selected based on initial homology models to encompass the rigid core of the CTE. Colabfold, version 1.0, was used.
Model 215 to 445: This region was selected to encompass the environment around the CTE, through the relationships between the dimerization domain, the RecA2 domain, and the degron sequence. Colabfold, version 1.2, was used.
Purification of full-length CrhR
The crhR ORF cloned as a 6XHis tag in pRSET-A was overexpressed in BL21 DE3 cells using ampicilin (100 μg/ml) and 1 mM IPTG and grown for 18 h at 18 °C (
). The 2 l cell pellet was suspended in 100 ml of lysis buffer (50 mM Borate pH 9, 550 mM KCl, 1 mM EDTA, 1 mM DTT, cOmplete Protease Inhibitor Cocktail [Sigma-Aldrich Canada]) and disrupted by sonication and clarified by centrifugation. PEI (final concentration 0.5%) was added to the cleared lysate followed by centrifugation to remove DNA. The supernatant was subjected to a 60% ammonium sulphate precipitation followed by centrifugation. The pellets were suspended in a buffer consisting of 5 ml Hic A (50 mM Tris pH 8, 1 mM EDTA, 1 mM DTT) and 5 ml Hic B (50 mM Tris pH 8, 1.5 M ammonium sulphate, 1 mM EDTA, 1 mM DTT) and diluted to 135 ml with equal parts Hic A and Hic B. The protein was applied to a 25 ml butyl-sepharose reverse phase column (Amersham Biosciences, Inc). The column was washed with 20 ml of Hic B, and the protein eluted with a 100 ml linear gradient of Hic B to Hic A. CrhR containing fractions were pooled and diluted to 135 ml with a 2:1 ratio of Cat A buffer (50 mM Hepes pH 7.3, 1 mM EDTA, 1 mM DTT) and Cat B buffer (50 mM Hepes pH 7.3, 750 mM KCl, 1 mM EDTA, 1 mM DTT). Protein was loaded onto a 25 ml SP-Sepharose fast flow cation exchange column (Amersham Biosciences, Inc). The column was washed with 20 ml of Cat A, and the protein eluted with a 100 ml linear gradient from Cat A to Cat B. CrhR containing fractions were pooled and loaded onto a Superdex 200 16/60 size exclusion column (Amersham Biosciences, Inc) where the protein was eluted and stored in 10 mM Hepes pH 7.3, 500 mM KCl, 1 mM DTT.
Commercial mass photometer (Refeyn Ltd) was used to determine the oligomeric state of purified 6xHis CrhR (
). Protein (5.5 μM) in SEC buffer (10 mM Hepes pH 7.3, 500 mM KCl, 1 mM DTT) was kept on ice, then diluted 90 times in the PBS buffer at 22 °C just before the experiment, and 15 μl of the mixture was placed on the well created by clean gaskets on top of clean microscope coverslip. Six thousand frames of data containing the contrast from different oligomeric state of the protein binding at the liquid glass interface were recorded by Refeyn AcquireMP and analyzed by DiscoverMP softwares. Mass of different oligomeric states of protein was determined from the calibration curve obtained from the known masses of BSA and Apoferritin. Data were plotted using a 5.2 kDa bin size in gnuplot. Gaussian curves were used to fit the data.
All data described are included within the article or supporting information. All raw data from figures have been included as a single pdf file for reference.
The authors declare that they have no conflicts of interest with the contents of this article.
We wish to thank Reem Skeik and Dr Patrick Linder for advice during the initial CrhR dimerization experiments, Dr Jeff Elhai and Dr Donald Bryant for supplying E. coli and cyanobacterial strains, Dr Michael T. Woodside for use of equipment and Dr Paul Baumann for supplying the E. coli S1 antisera.
B. T. W., C. R. A. M., M. J. N. G., R. P. F., and G. W. O. conceptualization; B. T. W., C. R. A. M., D. S. W., S. S. P., and M. J. N. G. investigation; B. T. W., D. S. W., C. R. A. M., S. S. P., M. J. N. G., R. P. F., and G. W. O. formal analysis; B. T. W., C. R. A. M., M. J. N. G., R. P. F., and G. W. O. writing-original draft; B. T. W., D. S. W., C. R. A. M., S. S. P., M. J. N. G., R. P. F., and G. W. O. writing-review and editing.
Funding and additional information
This work was supported by the Natural Sciences and Engineering Research Council of Canada grant RGPIN-2016-05448 to G. W. O., 1705792 to R. P. F. and 2016–05163 to M. J. N. G.
Sequence requirements for Lon-dependent degradation of the Escherichia coli transcription activator SoxS: Identification of the SoxS residues critical to proteolysis and specific inhibition of in vitro degradation by a peptide comprised of the N-terminal 21 amino acid residues.