Microorganisms usually coexist with a wide variety of polymicrobial communities, not only on abiotic surfaces in their natural habitat, but also on mucosal sites in the human body. These polymicrobial populations from resident microbiota and/or invading microbes can be regarded as important determinants of the human health and physiology, whose imbalances may lead to the pathological states of the human bodies. In recent years, increased attention has been paid to the complex interactions that occur in the polymicrobial populations, especially during bacterial infections. One of the best studied polymicrobial infections caused by complex communities of bacterial pathogens is the respiratory infections in the patients with cystic fibrosis (CF), where four major bacterial species (i.e., Pseudomonas aeruginosa, Staphylococcus aureus, Haemophilus influenzae, and Burkholderia cepacia complex) have been focused on [1, 2], with the two most prevalent species being P. aeruginosa and S. aureus [3]. They are highly notorious ESKAPE pathogens and the well-studied model bacteria in various microbiological aspects [4]. P. aeruginosa is a Gram-negative bacterium that is commonly found in soil and water as well as in plants, animals, and humans. P. aeruginosa has become an emerging opportunistic pathogen with multiple antibiotic resistance and tolerance in the clinics. S. aureus is a Gram-positive bacterium that is identified from warm-blooded animals, being a common cause of food poisoning and skin infections such as abscesses. Methicillin-resistant S. aureus (MRSA) is a worldwide concern in clinical medicine.

It is recently known that both P. aeruginosa and S. aureus affect each other in mixed cultures in vitro [5, 6]. P. aeruginosa produces 2-heptyl-4-hydroxyquinoline-N-oxide (HQNO) that kills S. aureus cells most likely to survive physiological conditions depleted in irons [7, 8]. HQNO-mediated S. aureus-killing is attributed to the apparent disturbance of electron transport in S. aureus. Given that functional electron transport chains are required not only to support the rapid growth of S. aureus, but also to render S. aureus cells susceptible to the bactericidal antibiotics [9], HQNO could select S. aureus small colony variants with altered respiratory activity in the presence of antibiotics during the interaction with P. aeruginosa [10]. They suggest the complex interactions between these two bacterial species during human infections and antibiotic treatment.

In this study, we first observed that the P. aeruginosa quorum-sensing (QS) mutant (lasRmvfR) devoid of HQNO production still displayed some residual killing activity against S. aureus mutants with altered respiratory activity. The QS mutant suffers from earlier autolysis than the wild type, which was suppressed by the presence of S. aureus cells as well as their culture supernatants. We have also uncovered the genetic link between the S. aureus-killing activity of P. aeruginosa and the autolysis suppression by S. aureus, based on the identification of the genes from both bacterial species, which might contribute to iron-dependent metabolism in P. aeruginosa and exoprotein secretion in S. aureus.

Bacterial Strains and Culture Conditions

The bacterial strains and plasmids used in this study are described in Table 1. Luria-Bertani (LB) (1% tryptone, 0.5% yeast extract and 1% NaCl) broth, LB broth supplemented with 50 mM KNO₃ (LBN), Tryptic soy broth (Difco, USA), 2% Bacto-agar (Difco) LB plates, and cetrimide agar (CA) (Difco) plates were used. Overnight-grown cultures were used as inoculum (1% sub-culture) into fresh broth and grown at 37°C in a shaking incubator until the logarithmic growth phase (i.e., OD₆₀₀ of 1.0), and then the cell cultures were used for the experiments described herein.

Table 1 . Bacterial strains and plasmids used in this study..

Strain or plasmid	Relevant characteristics or purposea	Reference or source
Pseudomonas aeruginosa
PA14	Wild type laboratory strain; RifR	Laboratory collection
PA14 lasR	PA14 with in-frame deletion of lasR; RifR	[18]
PA14 mvfR	PA14 with in-frame deletion of mvfR; RifR	This study
PA14 pqsA	PA14 with in-frame deletion of pqsA; RifR	[37]
PA14 lasRpqsA	PA14 with in-frame deletion of lasRpqsA; RifR	This study
PA14 lasRmvfR	PA14 with in-frame deletion of lasRmvfR; RifR	This study
PA14 lasRmvfRcysB	PA14 with in-frame deletion of lasRmvfRcysB; RifR	This study
PA14 lasRmvfRcysG	PA14 with in-frame deletion of lasRmvfRcysG; RifR	This study
PA14 lasRmvfRtruB	PA14 with in-frame deletion of lasRmvfRtruB; RifR	This study
Staphylococcus aureus
Newman	Wild type laboratory strain (MSSA); methicillin-sensitive	Laboratory collection
SA3	Wild type laboratory strain (MRSA); McR	Laboratory collection
m5	Respiratory mutant of SA3; McR	[17]
m5 saeS	m5 with Tn917 insertion in saeS; McR EmR	This study
m5 comEC	m5 with Tn917 insertion in comEC; McR EmR	This study
Escherichia coli
DH5α	Multipurpose cloning	Laboratory collection
S17-1(λpir)	Conjugal transfer of mobilizable plasmids	Laboratory collection
S17-1(λpir)(pBTK30)	S17-1(λpir) harboring pBTK30 for mariner transposon insertion; GmR CbR	Laboratory collection
Plasmids
pEX18T	Allelic exchange by homologous recombination	Laboratory collection
pTV1	Insertional mutagenesis by the transposon Tn917; CmR EmR	[35]
pEX18T-cysB	pEX18T with in-frame deletion in the cysB gene; CbR	This study
pEX18T-cysG	pEX18T with in-frame deletion in the cysG gene; CbR	This study
pEX18T-truB	pEX18T with in-frame deletion in the truB gene; CbR	This study

^aRif^R, rifampicin-resistant; Mc^R, methicillin-resistant; Gm^R, gentamicin-resistant; Cb^R, carbenicillin- and ampicillin-resistant; Cm^R, chloramphenicol-resistant; Em^R, erythromycin-resistant.

Construction of Deletion Mutants

All the deletion constructs were created using pEX18T as described elsewhere [16]. Oligonucleotide primers were designed using the PA14 genome sequence. SOEing (splicing by overlap extension) PCR was conducted by using four oligonucleotide primers for in-frame deletions as listed in Table 2. The resulting constructs were introduced into the wild type PA14 or the relevant mutants like the QS (lasRmvfR) mutant and the double-crossover recombinants were obtained by sucrose selection from the cointegrates, all of which were verified by PCR at each stage.

Table 2 . Primers and probes used in this study..

Primer or probe	Relevant characteristics or purpose	Oligonucleotide sequence (5'–3')a
mvfR-OF	SOEing PCR for mvfR in-frame deletion	GAATTCACGAGCAATATGA
mvfR-IR	SOEing PCR for mvfR in-frame deletion	CGCGCAGGCGCTGGGCGATGACCTGGAGGAA
mvfR-IF	SOEing PCR for mvfR in-frame deletion	CCAGGTCATCGCCCAGCGCCTGCGCGAACTGG
mvfR-OR	SOEing PCR for mvfR in-frame deletion	CTGCAGCATGGCAAGAGC
cysB-OF	SOEing PCR for cysB in-frame deletion	GAATTCGCAGGCTGGATGGTC
cysB-IR	SOEing PCR for cysB in-frame deletion	GGAGGACGAACTGGGCGGCTTCATGTGCGACT
cysB-IF	SOEing PCR for cysB in-frame deletion	AGTCGCACATGAAGCCGCCCAGTTCGTCCTCC
cysB-OR	SOEing PCR for cysB in-frame deletion	GGATCCTCGCCGGCAGCCATA
cysG-OF	SOEing PCR for cysG in-frame deletion	GGTACCCAGCCAGGACAAGTAC
cysG-IR	SOEing PCR for cysG in-frame deletion	GCTCAGCCACCAGTTGCGCGCCGGCGTCGGCC
cysG-IF	SOEing PCR for cysG in-frame deletion	GGCCGACGCCGGCGCGCAACTGGTGGCTGAGC
cysG-OR	SOEing PCR for cysG in-frame deletion	GGATCCTGCGGCGCATCGAAGAC
truB-OF	SOEing PCR for truB in-frame deletion	GGATCCTGTTGATGTTGGCGG
truB-IR	SOEing PCR for truB in-frame deletion	GAGCGTGGCCCAGGTGTGATGCCGGAAGACAG
truB-IF	SOEing PCR for truB in-frame deletion	CTGTCTTCCGGCATCACACCTGGGCCACGCTC
truB-OR	SOEing PCR for truB in-frame deletion	AAGCTTACAGCCGTACCCAGC
PA-ArbM1	Arbitrary PCR for transposon insertion site mapping	CTTACCAGGCCACGCGTCGACTAGTACNNNNNNNNNNGATAT
PA-ArbM2	Arbitrary PCR for transposon insertion site mapping	CTTACCAGGCCACGCGTCGACTAGTAC
Arb1-BTK	Arbitrary PCR for transposon insertion site mapping	CACCGCTGCGTTCGGTCAAG
Arb2-BTK	Arbitrary PCR for transposon insertion site mapping	CGAACCGAACAGGCCTTATGTTCAATTC
Seq-BTK	Sequencing of arbitrary PCR amplicons	GGATGAAGTGGTTCGCATCCTC
SA-ArbA1	Arbitrary PCR for transposon insertion site mapping	GGCCACGCGTCGACTAGTCANNNNNNNNGATCA
SA-ArbA2	Arbitrary PCR for transposon insertion site mapping	GGCCACGCGTCGACTAGTCA
Arb1-Tn917	Arbitrary PCR for transposon insertion site mapping	CACCTGCAATAACCGTTACCTG
Arb2-Tn917	Arbitrary PCR for transposon insertion site mapping	TCACAATAGAGAGATGTCACCG
Seq-Tn917	Sequencing of arbitrary PCR amplicons	CCAATCACTCTCGGACAATAC

^aUnderlining denotes the engineered restriction enzyme sites.

Autolysis Assay

P. aeruginosa autolysis was examined in 24-h or 48-h LBN cultures. Briefly, freshly grown (OD₆₀₀ of 1.0) cells (~10⁶ CFU) of P. aeruginosa or S. aureus were inoculated into the 48-well plate wells containing 400 μl LBN broth. The plates are incubated on a rotatory shaker at 37°C for either 24 or 48 h. Autolysis is monitored by visual inspection of aggregated cell debris, which was verified by Live/Dead-Baclight staining (Invitrogen, USA) [33].

S. aureus Killing Assay

S. aureus killing was assessed either by plate killing or by growth competition in 16-h liquid culture. Plate killing assay was previously described [18]. Briefly, LBN plates were overlaid with 0.7% top agar containing 100 μl of S. aureus cultures that had been grown to OD₆₀₀ of 1.0 and then dried for 1 h under sterile air blowing. P. aeruginosa bacterial suspensions (3 μl) containing 10⁶ CFU of early stationary growth phase (OD₆₀₀ of 3.0) were spotted onto the S. aureus lawns. Plates were incubated at 37°C for 16 h. The killing activity is scored as the visible halo around the cell spots.

Growth competition was monitored by separate viable counts of each strain after 16-h coculture of P. aeruginosa and S. aureus. Freshly grown (OD₆₀₀ of 1.0) cells (10⁶ CFU) of P. aeruginosa and S. aureus were inoculated into the culture tubes containing 3 ml LBN broth. After 16-h incubation, culture suspensions were 10-fold serially diluted, and the diluted samples (3 μl) were spotted onto the LB agar plates containing either 5% NaCl (for S. aureus selection) and 50 μg/ml rifampicin (for P. aeruginosa selection).

Transposon Experiments

Two plasmids (pBTK30 with Himar1 for P. aeruginosa and pTV1 with Tn917 for S. aureus) were used for transposon mutagenesis [34, 35]. pBTK30 was introduced into the P. aeruginosa QS mutant by conjugation for 5 h at 37°C. CA plates containing gentamicin (50 μg/ml) were used for Himar1 transposant selection. A total of 1,734 transposon insertion clones were screened for the mutants devoid of the residual S. aureus-killing activity of the QS mutant. Three mutants were chosen out of the 67 primary candidates. pTV1 was introduced into S. aureus m5 by transformation. Temperature induction and selection was performed by growing the cells overnight in LB broth supplemented with 10 μg/ml erythromycin at 42°C. A total of 1,765 chloramphenicol-sensitive and erythromycin-resistant Tn917 transposant clones were screened for the mutants incapable of the autolysis suppression, resulting in 2 transposon clones as the final candidates. The transposon insertion sites were determined by arbitrary PCR followed by sequencing using the appropriate primers listed in Table 2.

Protein Experiments

Exoprotein profiles were analyzed by SDS-PAGE. S. aureus cells were grown, and the culture supernatants were precipitated by 10% (v/v) trichloroacetic acid and separated on a 12% (vol/vol) polyacrylamide gel at 100 V for 110 min. The gels were stained with Coomassie Brilliant blue R 250 for 30 min as described elsewhere [36]. Ammonium sulfate (AS) precipitation was used for partial fractionation of the exoproteins, by altering the AS concentrations. The culture supernatant from 500 ml of the S. aureus m5 culture was subjected to filtration (0.22-μm membrane filter) and the filtrate was mixed with cOmplete^TM ULTRA Tablets EASYpack (Roche, Switzerland) and then kept at 4°C for 16 h. The sample was transferred to a beaker and AS powder was gradually added while agitating the sample to reach a final concentration of 50%. The sample was subjected to centrifugation at 12 K for 30 min to separate the pellet and the supernatant. The pellet was dissolved in PBS buffer (2.7 mM KCl, 137 mM NaCl, 10 mM Na₂HPO₄, and 2 mM KH₂PO₄, pH 7.0) as the 50% AS sample. The supernatant was further subjected to the AS addition to reach 65% and the 65% AS sample was obtained after dissolution of the 65% AS pellet. Likewise, the 80% and the 90% AS samples were obtained.

P. aeruginosa Autolysis Is Suppressed by S. aureus

Autolysis phenotype of P. aeruginosa has long been known, which are associated with the overproduction of the PQS-related signaling molecules [11]. This phenotype is frequently observed in the clinical isolates from the CF patients, which is attributed to the mutations in the lasR gene encoding the master QS regulator in P. aeruginosa. Fig. 1 represents the involvement of the QS regulators, LasR and MvfR (PqsR), in regulation of the biosynthetic pathway for the PQS-related signaling molecules: the pqsABCD and pqsE genes are positively regulated by MvfR, whereas the pqsH gene is under the LasR control. Autolysis is often observed in the prolonged incubation of the lasR mutant of the laboratory strains [12], which undergoes accumulation of the PQS precursor, 2-heptyl-4-hydroxyquinoline (HHQ) and/or HQNO [13]. It is known that the wild-type P. aeruginosa laboratory strains (PA14 and PAO1) exhibits autolysis after 48-h growth in the planktonic cultures, which is triggered by HQNO-mediated self-poisoning of the electron transport chains [14]. HQNO also poisons S. aureus at the level that could not poison P. aeruginosa. It can select for the small colony variants (SCVs) of S. aureus in condition that the growth of both P. aeruginosa and S. aureus could be affected by antibiotics in the mixed culture [10, 15].

Figure 1. Schematic overview of the PQS biosynthesis. The biosynthetic enzymes for the PQS-related signaling molecules are encoded by the pqs genes as indicated, and their expression is positively regulated by the QS-regulators LasR and MvfR as highlighted in red and blue, respectively. Abbreviations for the molecules: AA, anthranilic acid; 2-ABA, 2'-aminobenzoylacetic acid; 2-HABA, 2'-hydroxylaminobenzoylacetatic acid; HHQ, 4-hydroxy-2-heptylquinoline; HQNO, 2-heptyl-4-hydroxyquinoline- N-oxide; PQS, Pseudomonas quinolone signal.

To better understand the autolysis phenotype of P. aeruginosa regarding QS-dependent and/or conditional HQNO poisoning of both species, we used the QS mutants of P. aeruginosa PA14 during the coculture with S. aureus strains under aerobic nitrate-respiration condition, which might promote alternative respiration mode [16]. Fig. 2 shows that the wild type (WT) underwent visible autolysis with pigment overproduction by 48-h incubation, but not by 24-h incubation, whereas the QS mutants such as lasR, mvfR, and lasRmvfR displayed autolysis phenotypes by 24-h incubation. This result suggests that the P. aeruginosa QS involving LasR and MvfR is required to delay the autolysis, although they are required to generate the known auto-poisoning molecule, HQNO. This observation of the time-dependent relationship between autolysis and HQNO would be attributed to the differential susceptibility of the QS mutants to HQNO and/or to the other unknown mechanisms by which the earlier autolysis should occur in the QS mutants.

Figure 2. Autolysis suppression of the QS mutants. Autolysis of P. aeruginosa PA14 (WT) and its QS mutants (lasR, mvfR, and lasRmvfR) was monitored from the 48-well LBN cultures in the presence or absence (-) of S. aureus strains (Newman, SA3, and m5), which were grown at 37°C for either 24 h or 48 h. Cell debris by autolysis is indicated by arrowheads.

Michelsen et al. [15] reported the commensal-like interaction between P. aeruginosa and S. aureus, in which S. aureus was not killed by a certain P. aeruginosa isolate and its autolysis was suppressed by S. aureus. This study prompted us to evaluate the QS mutants for the S. aureus-mediated autolysis suppression, in that the HQNO-directed S. aureus-killing activity should be clearly reduced in the QS mutants. The WT S. aureus, Newman, an MRSA strain, SA3, and its respiratory mutant (m5) were used for the coculture with P. aeruginosa. The growth of m5 is comparable to that of SA3, and the whole genome sequencing revealed that m5 contains the ubiE mutation for the ubiquinone metabolism and two other mutations (atl_2 and lytN_1) [17]. We revealed that all the S. aureus strains were able to suppress the autolysis of the QS mutants and the 48-h culture autolysis of the WT. This and the fact that the WT P. aeruginosa can kill S. aureus, not in such commensal-like interactions, led us to hypothesize that the QS mutants could still kill S. aureus, which might be associated with the autolysis suppression of S. aureus.

P. aeruginosa Autolysis Suppression Is associated with the Residual S. aureus-Killing Activity

P. aeruginosa mvfR and pqsA mutants are impaired in killing Gram-positive bacteria [18]. It is noted that some residual killing activity happened to be observed on the respiratory mutant, m5 (Fig. 3). The killing activity was simply assessed by spotting P. aeruginosa cells on the lawns of Gram-positive bacteria. m5 is one of the five mutants that are resistant to naphthoquinone-generated reactive oxygen species (ROS) and have a ubiE mutation in common [17]. Those five mutants showed reduced respiratory activity and subsequently reduced ROS generation [17], where ROS would be the key to HQNO-poisoning of S. aureus. Fig. 3A shows that the S. aureus-killing activity was observed by the WT P. aeruginosa and, to the lesser extent, by the lasR mutant. However, the mutations in mvfR and pqsA completely abolished the visible killing activity on SA3. This is consistent with the previous observation that HQNO is crucial to the S. aureus-killing activity [10]. However, we highlighted some residual killing activity on the m5 mutant for the killing assay (Fig. 3B). The killing by the WT as well as that by the lasR mutant were also enhanced on m5, suggesting HQNO is the major, but not the sole killing activity against S. aureus. The residual killing activity was scarcely detected previously, but evident when P. aeruginosa interacts with the m5 mutant of S. aureus. The residual killing activity must have been unseen in the S. aureus with normal respiration. Although it needs to be further verified that the altered respiration as observed in m5 could occur during the natural interaction between P. aeruginosa and S. aureus, we hypothesized that the P. aeruginosa autolysis suppression by S. aureus might be associated with the residual S. aureus-killing activity by P. aeruginosa.

Figure 3. S. aureus killing of the QS mutants. S. aureus killing of P. aeruginosa PA14 (WT) and its mutants (pqsA, lasRpqsA, and lasRmvfR) was monitored on the cell lawns of S. aureus SA3 (A) and m5 (B). The residual killing activity is indicated by arrowheads.

P. aeruginosa Autolysis Suppression and S. aureus-Killing Activity Are Genetically Linked

To validate the association between the autolysis suppression and the residual S. aureus-killing activity, we attempted to isolate the mutants from both P. aeruginosa and S. aureus, which are defective in the residual S. aureus-killing activity and the autolysis suppression, respectively. Random transposon mutagenesis as described in Materials and Methods enabled us to identify three mutants of P. aeruginosa lasRmvfR (cysB, cysG, and truB) and two mutants of S. aureus m5 (saeS and comEC) (Fig. 4). It is important at this stage that we just wanted to gain an insight into the relationship between the autolysis suppression and the residual S. aureus-killing activity. Therefore, we did not delve into characterization of the individual genes, but just wanted to validate the genetic link therebetween.

Figure 4. S. aureus killing of the isolated mutants. (A) Positions of P. aeruginosa PA14 (WT) and its mutant (lasR, lasRmvfR, lasRmvfRcysB, lasRmvfRcysG, and lasRmvfRtruB) cells spots in B that had been applied to S. aureus killing plate assay as in Fig. 3. (B) S. aureus killing of P. aeruginosa cells as in A was monitored on the cell lawns of S. aureus m5 and its mutants (saeS and comEC). The disappearance of the residual killing activity of lasRmvfR is indicated by dotted arrowheads.

The involvement of the truB gene that encodes a subunit of pseudouridine synthase in tRNA modification [19] might be surprising. It should be noted, however, that the truA gene encoding another subunit of pseudouridine synthase is required for the optimal expression of type III secretory genes presumably by affecting the tRNA-mediated translation efficiency [20], suggesting that some TruB-dependent tRNA functions might be required for the optimal expression of the genes involved in the residual killing activity in P. aeruginosa.

The cysB gene encodes the master regulator (CysB) in sulfur uptake and cysteine biosynthesis [21], whereas the cysG encodes a methyltransferase for siroheme biosynthesis [22], despite its sequence and functional similarity to methyltransferases, NirE and CobA, in heme d1 and cobalamin synthesis, respectively [23]. Although we have not performed complementation experiments and do not fully understand whether these genes are indeed involved in the residual S. aureus-killing activity, the involvement of CysG was noteworthy in that it is associated with nitrate respiration and sulfur metabolism, in that siroheme is the prosthetic group for both nitrite reductase and sulfite reductase [24]. The impact of siroheme in addition to the previous finding that P. aeruginosa uses S. aureus as the iron source upon S. aureus killing [5] suggest the importance of iron-dependent metabolism of P. aeruginosa in the residual S. aureus-killing activity.

Identification of the S. aureus m5 mutants (saeS and comEC) also revealed the importance of virulence exoproteins (for saeS) and membrane functions (for comEC) in support for the S. aureus-killing. SaeS is a histidine kinase of the two-component regulatory system (SaeRS) that is required for expression and secretion of various extracellular virulence factors [25], whereas comEC encodes a large integral membrane protein forming a large channel for passage of DNA and/or peptides for competence [26]. Although we did not experimentally verify either whether these are indeed required for the residual S. aureus-killing activity, it is evident that the isolated mutant clones were devoid of the residual S. aureus-killing activity (Fig. 4).

To verify the S. aureus killing quantitatively, we designed the mixed culture of these two species in liquid broth. After 16-h culture, the remaining bacteria were enumerated by separate viable counts on the selective media as described in Materials and Methods. Fig. 5A shows the viable counts of S. aureus and P. aeruginosa after the coculture: in all cultures, the growth of P. aeruginosa was not affected at all by the presence of S. aureus. As expected, however, the killing activity against m5 was highest in the WT and no killing activity was observed in the lasRmvfRcysG mutant. In contrast, only the residual killing activity of the lasRmvfR mutant disappeared in either P. aeruginosa mutation (cysG) or S. aureus mutations (saeS and comEC), suggesting that both CysG and SaeS (or ComEC) are simultaneously required for the m5-killing activity of the lasRmvfR mutant.

Figure 5. S. aureus-killing and autolysis suppression of the isolated mutants. (A) S. aureus killing of P. aeruginosa PA14 (WT) and its mutants (lasR, lasRmvfR, and lasRmvfRcysG) cells was monitored after growth competition between one of them and S. aureus (m5, saeS, and comEC) in 24-h liquid culture. Culture suspensions were diluted and spotted on LB plates amended with either 5% NaCl (to select S. aureus) or 50 μg/ml rifampicin (to select P. aeruginosa). The numbers indicate the dilution folds of the culture suspension. (B) Autolysis of P. aeruginosa PA14 (WT) and its mutants (lasRmvfR, lasRmvfRcysB, lasRmvfRcysG, and lasRmvfRtruB) was monitored from the 48-well LBN cultures in the presence or absence (-) of S. aureus strains (SA3, m5, saeS and comEC), which were grown at 37°C for 24 h.

To verify the genetic link between the autolysis suppression and the residual S. aureus-killing activity, these mutants were tested for their ability to suppress the autolysis (Fig. 5B). It is noted that saeS and comEC bacteria could not suppress the autolysis of the m5-killing QS mutant and that the m5 bacteria could not suppress the autolysis of the unkilling QS mutants (especially with cysG). These results substantiate the genetic link between the autolysis suppression and the residual S. aureus-killing activity.

P. aeruginosa Autolysis Is Suppressed by S. aureus Exoproteins or Iron Treatment

Based on the genetic link between autolysis suppression and the residual S. aureus-killing activity featuring the identified genes especially the P. aeruginosa cysG and the S. aureus saeS, we postulated that the iron metabolism of P. aeruginosa and the secreted exoproteins of S. aureus could be involved in those phenotypes. It is evident that the extracellular protein profiles differed between the m5 and the mutant bacteria (Fig. 6A), in that some bands in the m5 sample were missing in those of the mutants, suggesting that some extracellular proteins could suppress the autolysis of the QS mutant and the subsequent S. aureus killing. To confirm this, we obtained the culture supernatant from m5 and prepared its ammonium sulfate (AS) fractions, which were tested for their ability to suppress the autolysis. As shown in Fig. 6B, the most prominent autolysis suppression was observed in the 80% AS fraction, suggesting that the suppressing activity was enriched in this fraction. The red color of the fraction led us to hypothesize that some iron-containing protein(s) could be the key to the autolysis suppression. Under consideration of the aforementioned mutant screens from P. aeruginosa and S. aureus in addition to the assumption that the exoprotein(s) would be important in the autolysis suppression, we examined if the treatment with only iron could suppress the autolysis phenotype (Fig. 6C). As a result, both iron (II) and iron (III) could suppress the autolysis of the QS mutants, whereas copper (II), calcium, and magnesium could not. It should be noted that copper (I) could partially suppress the autolysis of the mvfR and the lasRmvfR mutants, which could be further verified in comparison with other monovalent cations.

Figure 6. Autolysis suppression of the extracellular proteins and metals. (A) Profiles of extracellular proteins from S. aureus m5 and its mutants (saeS and comEC) were analyzed by 12% SDS-PAGE. The size markers (M) are included with the molecular weight (kDa) of representative bands. (B and C) Autolysis of P. aeruginosa PA14 was monitored from the 24-h 48- well LBN cultures in the presence of either ammonium sulfate ((NH₄)₂SO₄) precipitate fractions (B) at the indicated concentrations (50%, 65%, 80%, and 95%) or metal ions (C).

Iron is an essential element for growth and survival of most microorganisms. It is involved in many cellular processes as a cofactor tightly coordinated by hemes or amino acid residues of iron-containing proteins. Mashburn et al. [7] suggested that P. aeruginosa can utilize the iron-containing proteins of S. aureus as an iron source, which could be released from the lysed S. aureus cells. However, the S. aureus-killing activity could vary in the contexts of the interactions between these two species as well as their interactions with the human host in vivo. The involvement of the secreted exoproteins of S. aureus in the residual self-killing activity through iron-dependent metabolisms in P. aeruginosa needs to be further elucidated by characterizing the chemical identity of both the residual S. aureus-killing substance(s) of P. aeruginosa and the secreted exoprotein(s) of S. aureus that is enriched in the 80% AS fraction.

Polymicrobial infections can have profound effects on the course, severity, and treatment of microbial infections [27]. In many cases, different microorganisms within a polymicrobial community can lead to facilitated host colonization, enhanced pathogenic potential, and differential immune response [28, 29]. One of the well-studied examples is the interaction between P. aeruginosa and S. aureus that are highly prevalent in the CF lung and chronic wound infections [3, 30]. In the present study, we demonstrated the apparent association between the residual (not the major) S. aureus-killing activity of P. aeruginosa and the P. aeruginosa autolysis-suppression by S. aureus. We first elucidated the residual S. aureus-killing activity of the P. aeruginosa QS mutant (lasRmvfR) by exploiting the respiratory mutant of S. aureus, which had been resistant to chemical-generated ROS under aerobic conditions [17].

The connection between electron transport chains and ROS susceptibility is understandable, given that ROS can be generated during the respiration processes on molecular oxygen. It is noted that HQNO, whose production and secretion are controlled by the P. aeruginosa QS system, is implicated in both the S. aureus-killing and the P. aeruginosa autolysis. The finding of the residual S. aureus-killing activity in the absence of HQNO enabled us to highlight the importance of iron metabolism during the interaction between P. aeruginosa and S. aureus, which require the iron-related (i.e., siroheme) metabolism of P. aeruginosa and presumably the iron-containing exoprotein(s) of S. aureus. It is well known that P. aeruginosa can use S. aureus as an iron source [7]. The recent identification of pyochelin biotransformation by a secreted enzyme of S. aureus [31] also substantiates the importance of iron-dependent metabolism in the P. aeruginosa-S. aureus interaction. It is still likely that the details of iron availability will vary depending on the structure of the polymicrobial community of P. aeruginosa and S. aureus, which are more complicated by generation of the SCVs of both species [32]. Nevertheless, this study suggests that the interaction between P. aeruginosa and S. aureus might be governed by sophisticated mechanisms that necessitate the P. aeruginosa QS circuitry and the polymicrobial metabolism involving the extracellular iron resources during the coexistence with S. aureus in human airways.

This work was supported by the National Research Foundation of Korea (NRF) Grant (NRF-2022R1A2C3003943).

Y.-H.C. conceived and designed the research. S.-Y.C. and I.-Y.C. designed and performed the experiments, and collected and analyzed the experimental data. S.-Y.C., H.-W.B. and Y.-H.C. wrote the manuscript. All authors reviewed the manuscript.

The authors have no financial conflicts of interest to declare.