Studies on Escherichia coliHflKC suggest the presence of an unidentified λ factor that influences the lysis-lysogeny switch
- 3.1k Downloads
The lysis-lysogeny decision in the temperate coliphage λ is influenced by a number of phage proteins (CII and CIII) as well as host factors, viz. Escherichia coli HflB, HflKC and HflD. Prominent among these are the transcription factor CII and HflB, an ATP-dependent protease that degrades CII. Stabilization of CII promotes lysogeny, while its destabilization induces the lytic mode of development. All other factors that influence the lytic/lysogenic decision are known to act by their effects on the stability of CII. Deletion of hflKC has no effect on the stability of CII. However, when λ infects ΔhflKC cells, turbid plaques are produced, indicating stabilization of CII under these conditions.
We find that CII is stabilized in ΔhflKC cells even without infection by λ, if CIII is present. Nevertheless, we also obtained turbid plaques when a ΔhflKC host was infected by a cIII-defective phage (λcIII 67 ). This observation raises a fundamental question: does lysogeny necessarily correlate with the stabilization of CII? Our experiments indicate that CII is indeed stabilized under these conditions, implying that stabilization of CII is possible in ΔhflKC cells even in the absence of CIII, leading to lysogeny.
We propose that a yet unidentified CII-stabilizing factor in λ may influence the lysis-lysogeny decision in ΔhflKC cells.
KeywordsMG1655 Cell Bacterial Lawn Strain AK990 Clear Plaque Luria Agar
After it infects host E. coli cells, bacteriophage λ follows either of two fates, lytic or lysogenic. How the virus decides which pathway to follow after infection depends upon a complex genetic circuit. An increase in the number of infecting phages converts the decision making process from a deterministic to a stochastic one, with the cell fate depending on the number of phages deciding in favour of lysogeny [1, 2]. There are phage coded proteins and transcription factors [3, 4, 5] dedicated for this decision making process, but host factors are also involved [6, 7, 8, 9]. Mutations in the cI, cII and cIII genes of λ  enhances the lytic frequency (leading to clear plaque formation, hence the names) and therefore the products of these genes were thought to be responsible for the establishment of lysogeny. CII, the key tetrameric transcription factor for lysogenic establishment, is a very unstable protein [7, 11, 12] and its presence in sufficient amounts is crucial for the lysogenic choice [13, 14, 15]. Other factors such as λCIII and the host hfl proteins that influence the lysis-lysogeny switching affect the stability of CII in one way or the other. λCIII promotes lysogeny by acting as a general inhibitor of E. coli HflB that degrades CII .
Mutations in the host hfl loci cause an infecting λ particle to follow the lysogenic mode. These genes therefore encode factors that somehow destabilize CII. Primarily from mutational studies, two such loci, hflA and hflB, were initially identified. The product of the latter gene, HflB, is an ATP-dependent metalloprotease known as a 'quality control' protease that removes misfolded proteins produced due to rapid translation during good nutrient conditions [17, 18]. CII is also a substrate of HflB  and thus acts as a sensor for cellular nutrient conditions of the host. Rapid degradation of CII in cells growing in rich media thus favors the lytic development [13, 14]. The hflA locus consists of the genes hflX, hflK and hflC that are under the control of the same promoter [19, 20, 21, 22]. Of these, hflX has been demonstrated to have no role in lambda lysogeny . The products of the other two, HflK and HflC, are tightly associated with each other and copurify as the 'HflKC' complex, which was earlier thought to be a protease . Subsequently, HflKC was found only to act as a 'modulator' of HflB by forming a complex with the latter [25, 26, 27]. The only other known E. coli factor in this process, HflD , has been shown to inhibit CII-mediated activation of transcription by impairing the DNA-binding ability of CII .
HflKC antagonizes the action of HflB towards the membrane associated substrates of the latter [18, 25]. The behavior of HflKC with respect to the cytosolic substrates of HflB (such as λCII), however, remains unclear. Likewise, the role of HflKC in the lysis-lysogeny decision of λ is not well understood. Though an 'hfl' protein, mutations in whose gene(s) causes an increase in the lysogenic frequency of λ , the deletion of these genes has little effect on the in vivo stability of exogenous CII . CII expressed from a plasmid is found to be stabilized in an hflKC-deleted cell, only if the host is simultaneously infected with a lambda phage . On the other hand, E. coli cells overexpressing HflKC exhibit an enhanced frequency of lysogenization . These results lead to a paradox: lysogeny increases both upon deletion or overexpression of HflKC. Therefore, the role of HflKC in the λ lysis-lysogeny switch merits further investigations.
Plasmids, bacterial strains and phages
Plasmid pQKC was constructed by PCR cloning of the hflK and hflC ORFs (not fused, because the genomic region between these two contains the stop codon for hflK and the RBS for hflC) between the BamHI and SalI sites of pQE30 (purchased from Qiagen, contains the phage T5 promoter under the control of a Lac operator). Construction of pKP219 (which contains the cII gene under the lac promoter LacP and a P15A replication origin) has been described earlier . Plasmid pC2C3 (containing the cII and cIII genes) was constructed in three steps. First, the NdeI-BamHI fragment of pAB905 containing the cIII gene  was cloned into pKP07  and was named pLaCIII (containing the cIII gene under LacP). Then the BglII-XhoI fragment of pLaCIII (i.e. the cIII gene along with the LacP) was cloned into the compatible BamHI-XhoI sites of pKP106 (which already contained the cII gene under LacP) . The resulting plasmid was named pLaC2C3. In the final step the BamHI-BglII fragment of pLaC2C3 (containing both cII and cIII under individual LacP promoters) was cloned into the linearized arm of pK109 (having a P15A origin of replication)  at the BglII site.
For wild type E. coli, the strain MG1655 (F- λ- ilvG rfb-50 rph-1) was used. The strain AK990  (ΔhflKC:: Kan) served as cells with mutant hflKC.
The phage strain λcIII67 was used as the CIII-defective phage. In this strain, a G to T mutation in the 23rd nucleotide of the cIII ORF leads to an alternative structure of the cIII mRNA that is incapable of translation . This is one of the most effective cIII mutants  and has been used as cIII- by many workers.
Purification of proteins
For the purification of the HflKC complex, XL1Blue cells carrying pQKC was used and 100 μg/ml of ampicillin was used for selection. 7.5 ml of the overnight saturated culture was inoculated into 750 ml of fresh M9 medium with the appropriate antibiotic and allowed to grow on a 37°C shaker incubator till the culture O.D. (at 600 nm) was 0.4-0.5. The culture was then cooled to 18°C and induced by 500 μM IPTG, followed by further growth at 18°C with constant shaking (at 100 rpm) for 20 hours. After induction, bacterial cells were recovered by centrifugation at 3000 g for 10 minutes in Sorvall RC5C, using an SA600 rotor, at 4°C. The medium was decanted out and the pellet was washed with 0.9% NaCl and dissolved in 20 ml of lysis buffer (20 mM TRIS-HCl, pH 8.0, 100 mM KCl, 10% glycerol, 5 mM imidazole, 0.5% NP40, bacterial protease inhibitor cocktail (MBI Fermentas) and 200 μg/ml lysozyme). Cells were then lysed by sonication with 5 pulses (at a pulse rate of 10 mV/30 seconds), followed by centrifugation at 26000 g for 30 minutes at 4°C. The supernatant was collected into a fresh tube and loaded on to a Ni+2-NTA column, pre-equilibrated with the lysis buffer. After loading, the column was washed with wash buffer (20 mM TRIS-HCl, pH 8.0, 600 mM KCl, 10% glycerol, 15 mM imidazole). Proteins were eluted from the column using the elution buffer (20 mM TRIS-HCl, pH 8.0, 100 mM KCl, 10% glycerol, 0.1% NP40, 300 mM imidazole). Imidazole was removed by dialysis in 20 mM TRIS-HCl, pH 8.0, 100 mM KCl, 10% glycerol, 0.1% NP40).
In vitroproteolysis of CII
HflB mediated proteolysis of CII was carried out in buffer P (50 mM Tris-acetate, 100 mM NaCl, 5 mM MgCl2, 25 μM Zn-acetate, 1.4 mM β-ME; pH 7.2). ATP was added to a concentration of 5 mM in all the reaction mixtures. 8 μM of CII was taken with 1 μM of purified GST-HflB in a 30 μl reaction mix. The reactions were incubated at 37°C for the specified time intervals followed by the addition of SDS-PAGE loading buffer and heating in a boiling water bath for 5 minutes. The samples were analyzed on a 15% SDS-PAGE. The effect of HflKC on the proteolysis of CII was observed by the addition of His-HflKC (up to 2 μM) to GST-HflB prior to the addition of CII. The band corresponding to CII was quantitated by volume analysis (software used: Versadoc (Bio rad) Quantity-1) and used as the amount of CII remaining (expressed as the percentage of the amount of CII at zero time) after the specified time.
In vivoproteolysis of CII
In vivo proteolysis of CII was carried out in E. coli MG1655 cells (having wild type HflB) transformed with pKP219 or pC2C3, both of which contained cII under Lac promoter. In addition, pC2C3 contained cIII under a second Lac promoter. Cells carrying pKP219 or pC2C3 were inoculated in 10 ml of LB medium supplemented with 50 μg/ml kanamycin. Expression of CII was induced by 1 mM IPTG after the O.D. of the culture (at 600 nm) had reached 0.6. The culture was further grown at 37°C for another 30 minutes, followed by the addition of 10 μg/ml spectinomycin to arrest further protein synthesis. Samples were taken out at regular intervals after spectinomycin addition, and immediately centrifuged to pellet the cells. 30 μl of sterile water and 8 μl of SDS gel loading dye were added to each sample, followed by immediate boiling and loading onto a 15% SDS-PAGE. The gel was transferred to a PVDF membrane (Pierce Biotech) and was blotted with anti-CII antibody. Each CII band was quantitated by volumetric analysis as described above.
The effect of overexpression of hflKC was observed by transformation of MG1655 cells by plasmid pQKC (plus pKP219 or pC2C3). The transformed cells were grown in the presence of both kanamycin and ampicillin. Promoters in both the plasmids are inducible with IPTG. The effect of deletion of hflKC was observed by transformation of AK9990 cells by pKP219 or pC2C3.
For measurement of the stability of CII under conditions of infection by λcIII67, MG1655 or AK990 cells carrying pKP219 were grown in Luria broth supplemented with 0.4% maltose and were infected with the phage (at an MOI of 10 to ensure that all the cells were infected), 20 minutes after the addition of IPTG. Spectinomycin was added after another 25 minutes to ensure the entry of phage DNA and the expression of phage factors. Samples were then taken out at regular intervals and analyzed as described above.
Assay of plaque morphology
The plaque morphology of λcIII67 was assayed in E. coli MG1655 (wild type), in MG1655 cells carrying pQKC, and in strain AK990 (ΔhflKC::Kan). Cells were grown up to an O.D. (at 600 nm) of 0.6 in Luria broth supplemented with 0.4% maltose, and were induced with 500 μM IPTG. A bacterial lawn was made by pouring 5 ml of soft top agar (0.5% Luria agar supplemented with 0.4% maltose) mixed with 300 μl of these cells onto a 2% Luria agar plate. Another 100 μl of the above liquid culture was infected with λcIII67 at an M.O.I. of 0.1. It was further incubated at 32°C for 10 minutes to allow adsorption of the phage. Appropriate dilutions were then plated onto the prepared bacterial lawn and the plates were incubated overnight at 32°C. The turbidity of plaques formed in AK990 cells or in cells overexpressing HflKC were compared with the clear plaques formed in wild type cells upon infection by λcIII67.
Results and Discussion
Role of HflKC on the proteolysis of CII in vivo
These results pose an intriguing question. Why does the deletion of an inhibitor of CII proteolysis promote lysogeny? One can think of the following possibilities: (i) A proper assembly of HflB that is necessary for its activity against cytosolic substrates, may require HflKC; or (ii) In the absence of HflKC, HflB is guided towards its membrane-associated substrates , indirectly stabilizing the cytosolic substrate CII. However, from in vivo proteolysis experiments we found that in AK990 cells (ΔhflKC), exogenous CII was not stabilized (Figure 1), confirming that HflB was active against CII even in the absence of hflKC. This result rules out both the possibilities mentioned above. It may be noted that similar results were also obtained by Kihara et al . Therefore, an increase in lambda lysogeny upon overexpression of host HflKC  is not at all surprising, since HflKC inhibits the proteolysis of CII.
Effect of increasing concentrations of HflKC on the proteolysis of CII in vitro
Role of HflKC on in vivoproteolysis of CII: the effect of CIII
Deletion of hflKC genes lead to an increase in the lysogenic frequency of λ. Nevertheless, the stability of cloned CII remained unaffected in ΔhflKC cells. An interesting phenomenon, however, was observed in ΔhflKC cells that were infected by λ. CII expressed from a plasmid was stabilized in these cells . Thus it appears that some additional factors, supplied by the infecting phage, caused a stabilization of CII in the absence of HflKC. The only known phage factor that favors lysogeny by inhibiting the proteolysis of CII by HflB, is CIII [29, 36]. We therefore tested the possible involvement of CIII as the λ factor responsible for the above result, viz. stabilization of CII in λ-infected ΔhflKC cells.
CIII is a general inhibitor of CII proteolysis [29, 36, 37]. It is therefore expected that between a wild type strain alone and one that carries CIII, CII would exhibit a greater stability in the latter. A comparison of figures 1 and 3 (open circles) shows that this is indeed the case. Nonetheless, a greater stability of CII in ΔhflKC cells compared to wild type (both carrying the CIII-expressing plasmid) is surprising, since the absence of hflKC does not affect the stability of CII. CIII is itself a substrate of HflB . If HflKC facilitated the proteolysis of CIII, the above effect could be explained by the preferential stabilization of CIII in ΔhflKC cells. However, there was no difference in the in vitro proteolysis of CIII by HflB in the presence or absence of purified HflKC (data not shown). Therefore the role of CIII in this paradoxical effect is indirect.
Are there additional λ factors that influence the lysis-lysogeny decision?
Plaque morphology upon infection with λcIII67
Genotype of host E. colicell
Wild Type + pQKC
The unknown factor CSF could have been a non-protein factor (i.e, DNA) and lambda DNA would have been a good candidate for the same, since CII may be stabilized by binding to its cognate promoter. However, in our in vivo experiments, the plasmid pKP219 (used for the expression of exogenous CII) contained the promoter sequence PE, ruling out such a possibility.
Stabilization of CII in cells overexpressing hflKC is not surprising since HflKC is an inhibitor of CII-proteolysis. It is worthwhile to note that the effect of HflKC deletion is epistatic over the effect of cIII deletion, since even the absence of CIII cannot produce clear plaques in a ΔhflKC host. It is possible that CIII (and the hypothesized CIII-like factor CSF) works better in the absence of HflKC (Figure 5B). Therefore CII is better stabilized under these conditions and produces turbid plaques in ΔhflKC cells. cI, cII and cIII were first described as phage mutations which led to clear plaques in a wild type host. On the other hand, λ gives very turbid plaques in a ΔhflKC host. Our study thereby raises the possibility of finding novel phage mutations that would give clear plaques in an hflKC-deleted host.
1. E. coli HflKC inhibits the proteolysis of λCII by HflB and hence the overexpression of the former results in an increase in the lysogenic frequency.
2. In the absence of HflKC, λCII is stabilized upon infection by cIII-defective λ, suggesting the presence of a yet unidentified phage factor CSF (CII-stabilizing factor).
We thank S. Adhya (NIH, Bethesda) for λcIII 67 and for his comments on the manuscript, K. Ito and Y. Akiyama (Kyoto University) for E. coli AK990 strain, S. K. Dasgupta (Bose Institute) for the plasmid pSD5b and K. Shearwin (University of Adelaide) for anti-CII antibody. This work was funded by Institutional Project 5 (Microbial Genomics) of Bose Institute. KB was supported by CSIR, India (F. No. 9/15 (302)/2004-EMR-I).
- 8.Herman C, Ogura T, Tomoyasu T, Hiraga S, Akiyama Y, Ito K, Thomas R, D'Ari R, Bouloc P: Cell growth and lambda phage development controlled by the same essential Escherichia coli gene, ftsH/hflB. Proc Natl Acad Sci USA. 1993, 90 (22): 10861-10865. 10.1073/pnas.90.22.10861.PubMedPubMedCentralCrossRefGoogle Scholar
- 20.Noble JA, Innis MA, Koonin EV, Rudd KE, Banuett F, Herskowitz I: The Escherichia coli hflA locus encodes a putative GTP-binding protein and two membrane proteins, one of which contains a protease-like domain. Proc Natl Acad Sci USA. 1993, 90 (22): 10866-10870. 10.1073/pnas.90.22.10866.PubMedPubMedCentralCrossRefGoogle Scholar
- 21.Tsui HC, Feng G, Winkler ME: Transcription of the mutL repair, miaA tRNA modification, hfq pleiotropic regulator, and hflA region protease genes of Escherichia coli K-12 from clustered Esigma32-specific promoters during heat shock. J Bacteriol. 1996, 178 (19): 5719-5731.PubMedPubMedCentralGoogle Scholar
- 27.Kihara A, Akiyama Y, Ito K: Different pathways for protein degradation by the FtsH/HflKC membrane-embedded protease complex: an implication from the interference by a mutant form of a new substrate protein, YccA. J Mol Biol. 1998, 279 (1): 175-188. 10.1006/jmbi.1998.1781.PubMedCrossRefGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.