Phage Therapy Application to Counteract Pseudomonas aeruginosa Infection in Cystic Fibrosis Zebrafish Embryos.

Antimicrobial resistance, a major consequence of diagnostic uncertainty and antimicrobial overprescription, is an increasingly recognized cause of severe infections, complications, and mortality worldwide with a huge impact on our society and on the health system. In particular, patients with compromised immune systems or pre-existing and chronic pathologies, such as cystic fibrosis (CF), are subjected to frequent antibiotic treatments to control the infections with the appearance and diffusion of multidrug resistant isolates. Therefore, there is an urgent need to address alternative therapies to counteract bacterial infections. Use of bacteriophages, the natural enemies of bacteria, can be a possible solution. The protocol detailed in this work describes the application of phage therapy against Pseudomonas aeruginosa infection in CF zebrafish embryos. Zebrafish embryos were infected with P. aeruginosa to demonstrate that phage therapy is effective against P. aeruginosa infections as it reduces lethality, bacterial burden and pro-inflammatory immune response in CF embryos.

Phage therapy against Pseudomonas aeruginosa infections in a cystic fibrosis zebrafish model.

Cystic fibrosis (CF) is a hereditary disease due to mutations in the CFTR gene and causes mortality in humans mainly due to respiratory infections caused by Pseudomonas aeruginosa. In a previous work we used phage therapy, which is a treatment with a mix of phages, to actively counteract acute P. aeruginosa infections in mice and Galleria mellonella larvae. In this work we apply phage therapy to the treatment of P. aeruginosa PAO1 infections in a CF zebrafish model. The structure of the CFTR channel is evolutionary conserved between fish and mammals and cftr-loss-of-function zebrafish embryos show a phenotype that recapitulates the human disease, in particular with destruction of the pancreas. We show that phage therapy is able to decrease lethality, bacterial burden, and the pro-inflammatory response caused by PAO1 infection. In addition, phage administration relieves the constitutive inflammatory state of CF embryos. To our knowledge, this is the first time that phage therapy is used to cure P. aeruginosa infections in a CF animal model. We also find that the curative effect against PAO1 infections is improved by combining phages and antibiotic treatments, opening a useful therapeutic approach that could reduce antibiotic doses and time of administration.

Two Faces of CwlM, an Essential PknB Substrate, in Mycobacterium tuberculosis.

Tuberculosis claims >1 million lives annually, and its causative agent Mycobacterium tuberculosis is a highly successful pathogen. Protein kinase B (PknB) is reported to be critical for mycobacterial growth. Here, we demonstrate that PknB-depleted M. tuberculosis can replicate normally and can synthesize peptidoglycan in an osmoprotective medium. Comparative phosphoproteomics of PknB-producing and PknB-depleted mycobacteria identify CwlM, an essential regulator of peptidoglycan synthesis, as a major PknB substrate. Our complementation studies of a cwlM mutant of M. tuberculosis support CwlM phosphorylation as a likely molecular basis for PknB being essential for mycobacterial growth. We demonstrate that growing mycobacteria produce two forms of CwlM: a non-phosphorylated membrane-associated form and a PknB-phosphorylated cytoplasmic form. Furthermore, we show that the partner proteins for the phosphorylated and non-phosphorylated forms of CwlM are FhaA, a fork head-associated domain protein, and MurJ, a proposed lipid II flippase, respectively. From our results, we propose a model in which CwlM potentially regulates both the biosynthesis of peptidoglycan precursors and their transport across the cytoplasmic membrane.

Design of a Broad-Range Bacteriophage Cocktail That Reduces Pseudomonas aeruginosa Biofilms and Treats Acute Infections in Two Animal Models.

The alarming diffusion of multidrug-resistant (MDR) bacterial strains requires investigations on nonantibiotic therapies. Among such therapies, the use of bacteriophages (phages) as antimicrobial agents, namely, phage therapy, is a promising treatment strategy supported by the findings of recent successful compassionate treatments in Europe and the United States. In this work, we combined host range and genomic information to design a 6-phage cocktail killing several clinical strains of Pseudomonas aeruginosa, including those collected from Italian cystic fibrosis (CF) patients, and analyzed the cocktail performance. We demonstrated that the cocktail composed of four novel phages (PYO2, DEV, E215 and E217) and two previously characterized phages (PAK_P1 and PAK_P4) was able to lyse P. aeruginosa both in planktonic liquid cultures and in biofilms. In addition, we showed that the phage cocktail could cure acute respiratory infection in mice and treat bacteremia in wax moth (Galleria mellonella) larvae. Furthermore, administration of the cocktail to larvae prior to bacterial infection provided prophylaxis. In this regard, the efficiency of the phage cocktail was found to be unaffected by the MDR or mucoid phenotype of the pseudomonal strain. The cocktail was found to be superior to the individual phages in destroying biofilms and providing a faster treatment in mice. We also found the Galleria larva model to be cost-effective for testing the susceptibility of clinical strains to phages, suggesting that it could be implemented in the frame of developing personalized phage therapies.

The external PASTA domain of the essential serine/threonine protein kinase PknB regulates mycobacterial growth.

PknB is an essential serine/threonine protein kinase required for mycobacterial cell division and cell-wall biosynthesis. Here we demonstrate that overexpression of the external PknB_PASTA domain in mycobacteria results in delayed regrowth, accumulation of elongated bacteria and increased sensitivity to ß-lactam antibiotics. These changes are accompanied by altered production of certain enzymes involved in cell-wall biosynthesis as revealed by proteomics studies. The growth inhibition caused by overexpression of the PknB_PASTA domain is completely abolished by enhanced concentration of magnesium ions, but not muropeptides. Finally, we show that the addition of recombinant PASTA domain could prevent regrowth of Mycobacterium tuberculosis, and therefore offers an alternative opportunity to control replication of this pathogen. These results suggest that the PknB_PASTA domain is involved in regulation of peptidoglycan biosynthesis and maintenance of cell-wall architecture.

Mycobacterium tuberculosis RNA polymerase-binding protein A (RbpA) and its interactions with sigma factors.

RNA polymerase-binding protein A (RbpA), encoded by Rv2050, is specific to the actinomycetes, where it is highly conserved. In the pathogen Mycobacterium tuberculosis, RbpA is essential for growth and survival. RbpA binds to the ß subunit of the RNA polymerase where it activates transcription by unknown mechanisms, and it may also influence the response of M. tuberculosis to the current frontline anti-tuberculosis drug rifampicin. Here we report the solution structure of RbpA and identify the principle sigma factor σ(A) and the stress-induced σ(B) as interaction partners. The protein has a central ordered domain with a conserved hydrophobic surface that may be a potential protein interaction site. The N and C termini are highly dynamic and are involved in the interaction with the sigma factors. RbpA forms a tight complex with the N-terminal domain of σ(B) via its N- and C-terminal regions. The interaction with sigma factors may explain how RbpA stabilizes sigma subunit binding to the core RNA polymerase and thereby promotes initiation complex formation. RbpA could therefore influence the competition between principal and alternative sigma factors and hence the transcription profile of the cell.

WhiB5, a transcriptional regulator that contributes to Mycobacterium tuberculosis virulence and reactivation.

The proteins belonging to the WhiB superfamily are small global transcriptional regulators typical of actinomycetes. In this paper, we characterize the role of WhiB5, a Mycobacterium tuberculosis protein belonging to this superfamily. A null mutant was constructed in M. tuberculosis H37Rv and was shown to be attenuated during both progressive and chronic mouse infections. Mice infected with the mutant had smaller bacillary burdens in the lungs but a larger inflammatory response, suggesting a role of WhiB5 in immunomodulation. Most interestingly, the whiB5 mutant was not able to resume growth after reactivation from chronic infection, suggesting that WhiB5 controls the expression of genes involved in this process. The mutant was also more sensitive than the wild-type parental strain to S-nitrosoglutathione (GSNO) and was less metabolically active following prolonged starvation, underscoring the importance of GSNO and starvation in development and maintenance of chronic infection. DNA microarray analysis identified 58 genes whose expression is influenced by WhiB5, including sigM, encoding an alternative sigma factor, and genes encoding the constituents of two type VII secretion systems, namely, ESX-2 and ESX-4.

Genome-wide discovery of small RNAs in Mycobacterium tuberculosis.

Only few small RNAs (sRNAs) have been characterized in Mycobacterium tuberculosis and their role in regulatory networks is still poorly understood. Here we report a genome-wide characterization of sRNAs in M. tuberculosis integrating experimental and computational analyses. Global RNA-seq analysis of exponentially growing cultures of M. tuberculosis H37Rv had previously identified 1373 sRNA species. In the present report we show that 258 (19%) of these were also identified by microarray expression. This set included 22 intergenic sRNAs, 84 sRNAs mapping within 5'/3' UTRs, and 152 antisense sRNAs. Analysis of promoter and terminator consensus sequences identified sigma A promoter consensus sequences for 121 sRNAs (47%), terminator consensus motifs for 22 sRNAs (8.5%), and both motifs for 35 sRNAs (14%). Additionally, 20/23 candidates were visualized by Northern blot analysis and 5' end mapping by primer extension confirmed the RNA-seq data. We also used a computational approach utilizing functional enrichment to identify the pathways targeted by sRNA regulation. We found that antisense sRNAs preferentially regulated transcription of membrane-bound proteins. Genes putatively regulated by novel cis-encoded sRNAs were enriched for two-component systems and for functional pathways involved in hydrogen transport on the membrane.

Mycobacterium smegmatis RNase J is a 5'-3' exo-/endoribonuclease and both RNase J and RNase E are involved in ribosomal RNA maturation.

The presence of very different sets of enzymes, and in particular the presence of RNase E and RNase J, has been used to explain significant differences in RNA metabolism between the two model organisms Escherichia coli and Bacillus subtilis. However, these studies might have somewhat polarized our view of RNA metabolism. Here, we identified a RNase J in Mycobacterium smegmatis that has both 5'-3' exo- and endonucleolytic activity. This enzyme coexists with RNase E in this organism, a configuration that enabled us to study how these two key nucleases collaborate. We demonstrate that RNase E is responsible for the processing of the furA-katG transcript in M. smegmatis and that both RNase E and RNase J are involved in the 5' end processing of all ribosomal RNAs. In contrast to B. subtilis, the activity of RNase J, although required in vivo for 23S rRNA maturation, is not essential in M. smegmatis. We show that the pathways for ribosomal RNA maturation in M. smegmatis are quite different from those observed in E. coli and in B. subtilis. Studying organisms containing different combinations of key ribonucleases can thus significantly broaden our view of the possible strategies that exist to direct RNA metabolism.

Isolation of conditional expression mutants in Mycobacterium tuberculosis by transposon mutagenesis.

In Mycobacterium tuberculosis identification of essential genes has been hampered by the scarcity of suitable genetic tools for genome wide screenings. We constructed two Himar1 transposon derivatives in which the Streptomyces pristinamycin I-inducible ptr promoter was inserted at one transposon end in outward orientation. These transposons, Tn-pip/pptr (which harbours the promoter and its repressor pip gene) and Tn-pptr (which depends on a host expressing the pip gene), were inserted in the thermosensitive mycobacteriophage phAE87. After transduction into M. tuberculosis H37Rv, hygromycin resistant clones were selected in the presence of pristinamycin, screened for inducer dependent growth, and the transposon insertion point mapped by sequencing. Out of 3530 Hyg(R) mutants tested, we obtained 14 (0.4%) single insertion conditional mutants. In three (leuA, mazE6, rne) pptr was located upstream of genes whose function had been assessed by experimental evidence, whereas in seven the transposon targeted genes (ftsK, glf, infB, metC, pyrD, secY, and tuf) whose function had been assigned by similarity with homologous genes and four ORFs of unknown function (Rv0883c, Rv1478, Rv2050 and Rv2204c). These results validate our mutagenesis system and provide previously unavailable conditional expression mutants in genes of known, putative and unknown functions for genetic and physiological studies.

Pristinamycin-inducible gene regulation in mycobacteria.

In this work the Pip-inducible system, already used in eukaryotes, was tested in mycobacteria. This system is based on the Streptomyces coelicolor Pip repressor, the Streptomyces pristinaespiralis ptr promoter and the inducer pristinamycin I. By cloning in an integrative plasmid the ptr promoter upstream of the lacZ reporter gene and the pip gene under the control of a constitutive mycobacterial promoter, we demonstrated that the ptr promoter activity increased up to 50-fold in Mycobacterium smegmatis and up to 400-fold in Mycobacterium tuberculosis, in dependence on pristinamycin I concentration, and that the promoter was fully repressed in the absence of the inducer. Three mycobacterial genes were cloned under pptr-Pip control, both in sense and antisense direction; both proteins and antisense RNAs could be over-expressed, the antisenses causing a partial reduction of the amount of the targeted proteins. This system was used to obtain two M. tuberculosis conditional mutants in the fadD32 and pknB genes: the mutant strains grew only in the presence of the inducer pristinamycin I. Thus it showed to be an effective inducible system in mycobacteria.

The katG mRNA of Mycobacterium tuberculosis and Mycobacterium smegmatis is processed at its 5' end and is stabilized by both a polypurine sequence and translation initiation.

BACKGROUND: In Mycobacterium tuberculosis and in Mycobacterium smegmatis the furA-katG loci, encoding the FurA regulatory protein and the KatG catalase-peroxidase, are highly conserved. In M. tuberculosis furA-katG constitute a single operon, whereas in M. smegmatis a single mRNA covering both genes could not be found. In both species, specific 5' ends have been identified: the first one, located upstream of the furA gene, corresponds to transcription initiation from the furA promoter; the second one is the katG mRNA 5' end, located in the terminal part of furA. RESULTS: In this work we demonstrate by in vitro transcription and by RNA polymerase Chromatin immunoprecipitation that no promoter is present in the M. smegmatis region covering the latter 5' end, suggesting that it is produced by specific processing of longer transcripts. Several DNA fragments of M. tuberculosis and M. smegmatis were inserted in a plasmid between the sigA promoter and the lacZ reporter gene, and expression of the reporter gene was measured. A polypurine sequence, located four bp upstream of the katG translation start codon, increased beta-galactosidase activity and stabilized the lacZ transcript. Mutagenesis of this sequence led to destabilization of the mRNA. Analysis of constructs, in which the polypurine sequence of M. smegmatis was followed by an increasing number of katG codons, demonstrated that mRNA stability requires translation of at least 20 amino acids. In order to define the requirements for the 5' processing of the katG transcript, we created several mutations in this region and analyzed the 5' ends of the transcripts: the distance from the polypurine sequence does not seem to influence the processing, neither the sequence around the cutting point. Only mutations which create a double stranded region around the processing site prevented RNA processing. CONCLUSION: This is the first reported case in mycobacteria, in which both a polypurine sequence and translation initiation are shown to contribute to mRNA stability. The furA-katG mRNA is transcribed from the furA promoter and immediately processed; this processing is prevented by a double stranded RNA at the cutting site, suggesting that the endoribonuclease responsible for the cleavage cuts single stranded RNA.

Bacteriophage P4 sut1: a mutation suppressing transcription termination.

In the Escherichia coli satellite phage P4, transcription starting from PLE is prevalently controlled via premature termination at several termination sites. We identified a spontaneous mutation, P4 sut1 (suppression of termination), in the natural stop codon of P4 orf151 that, by elongating translation, suppresses transcription termination at the downstream t151 site. Both the translational and the transcriptional profile of P4 sut1 differed from those of P4 wild-type. First of all, P4 sut1 did not express Orf151, but a higher molecular mass protein, compatible with the 303 codon open reading frame generated by the fusion of orf151, cnr and the intervening 138 nt. Moreover, after infection of E. coli, the mutant expressed a very low amount of the 1.3 and 1.7 kb transcripts originating at PLE and PLL promoters, respectively, and terminating at the intracistronic t151 site, whereas correspondingly higher amounts of the 4.1 and 4.5 kb RNAs arising from the same promoters and covering the entire operon were detected. Thus the sut1 mutation converts a natural stop codon into a sense codon, suppresses a natural intracistronic termination site and leads to overexpression of the downstream cnr and alpha genes. This correlates with the inability of P4 sut1 to propagate in the plasmid state. By cloning different P4 DNA fragments, we mapped the t151 transcription termination site within the 7633-7361 region between orf151 and gene cnr. A potential stem-loop structure, resembling the structure of a Rho-independent termination site, was predicted by mfold sequence analysis at 7414-7385.

DNA replication in phage P4: characterization of replicon II.

The genetic element P4 propagates in its host Escherichia coli both as a satellite phage and as a plasmid. Two partially overlapping replicons coexist, namely replicon I and replicon II. The former is composed of two sites, ori1 and crr, and depends on P4 alpha gene product for replication. The P4 alpha protein has primase and helicase activities, and binds specifically to both ori1 and crr. Replicon II is composed of two sites, ori2 and crr, and its replication also depends on P4 alpha primase and helicase activities. In replicon II, the alpha protein binds only crr. Here we show that for replicon II the relative orientation of ori2 and crr is essential for replication to occur. Furthermore we delimit ori2 to a 22 bp region (6234-6255), internal to the alpha gene, sufficient for replicon II replication. We mutagenized this region and identified two mutants, which carry one and two base substitutions, respectively, that prevent replicon II replication. In electrophoretic mobility shift experiments of ori2, ori1, and crr DNA fragments with E. coli extracts, ori2 was not shifted, whereas both ori1 and crr were specifically bound, suggesting that other host protein(s), beside P4 alpha, are able to bind to these cis essential regions. Apparently, no binding to ori2 could be identified, thus suggesting that neither alpha nor other bacterial proteins specifically bind to this region.

Bacteriophage P4 Vis protein is needed for prophage excision.

Upon infection of its host Escherichia coli, satellite bacteriophage P4 can integrate its genome into the bacterial chromosome by Int-mediated site-specific recombination between the attP and the attB sites. The opposite event, excision, may either occur spontaneously or be induced by a superinfecting P2 helper phage. In this work, we demonstrate that the product of the P4 vis gene, a regulator of the P4 late promoters P(LL) and P(sid), is needed for prophage excision. This conclusion is supported by the following evidence: (i) P4 mutants carrying either a frameshift mutation or a deletion of the vis gene were unable to excise both spontaneously or upon P2 phage superinfection; (ii) expression of the Vis protein from a plasmid induced P4 prophage excision; (iii) excision depended on a functional integrase (Int) protein, thus suggesting that Vis is involved in the formation of the excision complex, rather than in the excision recombination event per se; (iv) Vis protein bound P4 DNA in the attP region at two distinct boxes (Box I and Box II), located between the int gene and the attP core region, and caused bending of the bound DNA. Furthermore, we mapped by primer extension the 5' end of the int transcript and found that ectopic expression of Vis reduced its signal intensity, suggesting that Vis is also involved in negative regulation of the int promoter.

Characterization of Escherichia coli uridine phosphorylase by single-site mutagenesis.

The Escherichia coli udp gene encodes uridine phosphorylase (UP), which catalyzes the reversible phosphorolysis of uridine to uracil and ribose-1-phosphate. The X-ray structure of E. coli UP resolved by two different groups produced conflicting results. In order to cast some light on the E. coli UP catalytic site, we mutagenized several residues in UP and measured by RP-HPLC the phosphorolytic activity of the mutant UP proteins in vitro. Mutations Thr94Ala, Phe162Ala, and Tyr195Gly caused a drastic decrease in UP activity. These three residues were suggested to be involved in the nucleoside binding site. However, surprisingly, Tyr195Ala caused a relative increase in enzymatic activity. Both Met197Ala and Met197Ser conserved low activity, suggesting a minor role for this residue in the UP active site. Glu196Ala completely lost UP activity, whereas the more conservative Glu196Asp mutation was still partially active, confirming the importance of maintaining the correct charge in the surroundings of this position. Glu198 was mutated to either Gly, Asp and Gln. All three substitutions caused complete loss of enzymatic activity suggesting an important role of Glu198 both in ribose binding and in interaction with phosphate ions. Arg30Ala and Arg91Ala eliminated UP activity, whereas Arg30Lys and Arg91Lys presented a very low activity, confirming that these residues might interact with and stabilize the phosphate ions. Ile69Ala did not decrease UP activity, whereas His8Ala lowered the activity to about 20%. Both amino acids were suggested to take part in subunit interactions. Our results confirm the structural similarity between E. coli UP and E. coli purine nucleoside phosphorylase (PNP).

A mutation in polynucleotide phosphorylase from Escherichia coli impairing RNA binding and degradosome stability.

Polynucleotide phosphorylase (PNPase), a 3' to 5' exonuclease encoded by pnp, plays a key role in Escherichia coli RNA decay. The enzyme, made of three identical 711 amino acid subunits, may also be assembled in the RNA degradosome, a heteromultimeric complex involved in RNA degradation. PNPase autogenously regulates its expression by promoting the decay of pnp mRNA, supposedly by binding at the 5'-untranslated leader region of an RNase III-processed form of this transcript. The KH and S1 RNA-binding domains at the C-terminus of the protein (amino acids 552-711) are thought to be involved in pnp mRNA recognition. Here we show that a G454D substitution in E.coli PNPase impairs autogenous regulation whereas it does not affect the catalytic activities of the enzyme. Although the mutation maps outside of the KH and S1 RNA-binding domains, analysis of the mutant protein revealed a defective RNA binding, thus suggesting that other determinants may be involved in PNPase-RNA interactions. The mutation also caused a looser association with the degradosome and an abnormal electrophoretic mobility in native gels. The latter feature suggests an altered structural conformation of PNPase, which may account for the properties of the mutant protein.

Mycobacterium tuberculosis FurA autoregulates its own expression.

The furA-katG region of Mycobacterium tuberculosis, encoding a Fur-like protein and the catalase-peroxidase, is highly conserved among mycobacteria. Both genes are induced upon oxidative stress. In this work we analyzed the M. tuberculosis furA promoter region. DNA fragments were cloned upstream of the luciferase reporter gene, and promoter activity in Mycobacterium smegmatis was measured in both the presence and absence of oxidative stress. The shortest fragment containing an inducible promoter extends 45 bp upstream of furA. In this region, -35 and -10 promoter consensus sequences can be identified, as well as a 23-bp AT-rich sequence that is conserved in the nonpathogenic but closely related M. smegmatis. M. tuberculosis FurA was purified and found to bind upstream of furA by gel shift analysis. A ca. 30-bp DNA sequence, centered on the AT-rich region, was essential for FurA binding and protected by FurA in footprinting analysis. Peroxide treatment of FurA abolished DNA binding. Three different AT-rich sequences mutagenized by site-directed mutagenesis were constructed. In each mutant, both M. tuberculosis FurA binding in vitro and pfurA regulation upon oxidative-stress in M. smegmatis were abolished. Thus, pfurA is an oxidative stress-responsive promoter controlled by the FurA protein.

RNase E and polyadenyl polymerase I are involved in maturation of CI RNA, the P4 phage immunity factor.

Bacteriophage P4 immunity is controlled by a small stable RNA (CI RNA) that derives from the processing of primary transcripts. In previous works, we observed that the endonuclease RNase P is required for the maturation of CI RNA 5'-end; moreover, we found that polynucleotide phosphorylase (PNPase), a 3' to 5' RNA-degrading enzyme, is required for efficient 5'-end processing of CI RNA, suggesting that 3'-end degradation of the primary transcript might be involved in the production of proper RNase P substrates. Here, we demonstrate that another Escherichia coli nuclease, RNase E, would appear to be involved in this process. We found that transcripts of the P4 immunity region are modified by the post-transcriptional addition of short poly(A) tails and heteropolymeric tails with prevalence of A residues. Most oligoadenylated transcripts encompass the whole cI locus and are thus compatible as intermediates in the CI RNA maturation pathway. On the contrary, in a polynucleotide phosphorylase (PNPase)-defective host, adenylation occurred most frequently within cI, implying that such transcripts are targeted for degradation. We did not find polyadenylation in a pcnB mutant, suggesting that the pcnB-encoded polyadenyl polymerase I (PAP I) is the only enzyme responsible for modification of P4 immunity transcripts. Maturation of CI RNA 5'-end in such a mutant was impaired, further supporting the idea that processing of the 3'-end of primary transcripts is an important step for efficient maturation of CI RNA by RNase P.

Characterization of the small antisense CI RNA that regulates bacteriophage P4 immunity.

In the immune state bacteriophage P4 prevents expression of the replication functions by premature termination of transcription. A small RNA, the CI RNA, is the trans acting factor that regulates P4 immunity, by pairing to complementary target sequences and causing premature transcription termination. The CI RNA is matured by RNAse P and PNPase from the leader region of the same operon it regulates. In this work we better characterize this molecule. CI RNA copy number was determined to be around 500 molecules per lysogenic cell. By S(1) mapping we defined the 3'-end at 8423(+/-1); thus CI RNA is 79(+/-1) nt long. The minimum region for correct processing requires two bases upstream of the CI RNA 5'-end and the CCA sequence at the 3'-end. Computer analysis by FOLD RNA of CI RNA sequence predicts a cloverleaf-like structure formed by a double-stranded stalk, a minor and a major stem loop, and a single-stranded bulge. We analysed several cI mutations, which fall either in the single or double-stranded CI RNA regions. Base substitutions in the main loop and in the single-stranded bulge apparently did not change CI RNA structure, but affected its activity by altering the complementarity with the target sequences, whereas a mutation in the secondary stem had a disruptive effect on CI RNA secondary structure. The effects of this latter mutation were suppressed by a base substitution that restored the complementarity with the corresponding base in the stem. Base substitutions in the main stem caused only local alterations in the secondary structure of CI. However, when the substitutions concerned either G8501 or its complementary base at the bottom of the stem, CI RNA was not correctly processed.