Solonamides, a Group of Cyclodepsipeptides, Influence Motility in the Native Producer Photobacterium galatheae S2753

ABSTRACT

The marine bacterium Photobacterium galatheae S2753 produces a group of cyclodepsipeptides, called solonamides, which impede the virulence but not the survival of Staphylococcus aureus. In addition to their invaluable antivirulence activity, little is known about the biosynthesis and physiological function of solonamides in the native producer. This study generated a solonamide-deficient (Δsol) mutant by in-frame deletion of the sol gene, thereby identifying the core gene for solonamide biosynthesis. By annotation from antiSMASH, the biosynthetic pathway of solonamides in S2753 was also proposed. Mass spectrometry analysis of cell extracts found that deficiency of solonamide production influenced the production of a group of unknown compounds but otherwise did not alter the overall secondary metabolite profile. Physiological comparison between Δsol and wild-type S2753 demonstrated that growth dynamics and biofilm formation of both strains were similar; however, the Δsol mutant displayed reduced motility rings compared to the wild type. Reintroduction of sol restored solonamide production and motility to the mutant, indicating that solonamides influence the motility behavior of P. galatheae S2753. Proteomic analysis of the Δsol and wild-type strains found that eliminating solonamides influenced many cellular processes, including swimming-related proteins and proteins adjusting the cellular cyclic di-GMP concentration. In conclusion, our results revealed the biosynthetic pathway of solonamides and their ecological benefits to P. galatheae S2753 by enhancing motility, likely by altering the motile physiology.

IMPORTANCE The broad range of bioactive potentials of cyclodepsipeptides makes these compounds invaluable in the pharmaceutical industry. Recently, a few novel cyclodepsipeptides have been discovered in marine Proteobacteria; however, their biosynthetic pathways remain to be revealed. Here, we demonstrated the biosynthetic genetic basis and pathway of the antivirulence compounds known as solonamides in P. galatheae S2753. This can pave the way for the biological overproduction of solonamides on an industrial scale. Moreover, the comparison of a solonamide-deficient mutant and wild-type S2753 demonstrated that solonamides stimulate the swimming behavior of S2753 and also influence a few key physiological processes of the native producers. These results evidenced that, in addition to their importance as novel drug candidates, these compounds play a pivotal role in the physiology of the producing microorganisms and potentially provide the native producer competitive benefits for their survival in nature.

INTRODUCTION

Microorganisms produce a vast array of secondary metabolites, and many of these natural products are used in the clinic, in agriculture and horticulture, and in food processing (1–3). Most natural products have been derived from soil bacteria and fungi; however, the marine environment is emerging as a promising source of novel bioactive compounds, as revealed by a range of omics technologies, including genome mining of already-cultured microorganisms (4, 5). Bacteria of the marine bacterial family Vibrionaceae have predominantly been thought of as pathogens or symbionts (6) but have recently emerged as a group with natural product potential, producing a range of antibacterial compounds (4, 7–9).

One Vibrionaceae strain exhibiting potent antibacterial properties is Photobacterium galatheae S2753, isolated from a green mussel in the Solomon Sea (10). P. galatheae S2753 produces the antibiotic holomycin (11, 12) that is not only an antibacterial compound but also influences the biofilm formation of the producer as well as other marine bacteria (12). P. galatheae S2753 also produces two series of cyclodepsipeptides, solonamides and ngercheumicins, which reduce the virulence gene expression of Staphylococcus aureus (13, 14). Antivirulence drugs are of significant interest, since they do not target the viability of the pathogen but impede the pathogen without the high selective resistance pressure that is seen in antibiotic therapy (15), and that, according to the World Health Organization, is one of the three most significant global health threats (16).

Solonamides are non-ribosomally derived peptides (NRPs) consisting of four amino acids and a short 3-hydroxy fatty acid typically six to 10 carbons long. In the case of solonamides A and B, the amino acids are l-phenylalanine, d-leucine, d-alanine, and l-leucine (l-valine in solonamides C and D) (14). The cyclic feature of solonamides and their analogs is structurally similar to the quorum sensing (QS) autoinducing peptide (AIP) of S. aureus (14). The solonamides act as receptor antagonists by competitively binding to the AIP receptor, AgrC, without activating the latter QS signaling pathway. This results in the inhibition of the QS system that controls the virulence of S. aureus (14, 17–19). Cyclodepsipeptides produced by actinomycetes also inhibit cyclic-peptide-mediated QS in many Gram-positive bacteria (20). Staphylococci are not common in the marine environment and not associated with the niches typically occupied by Vibrionaceae, and it is thus not likely that AgrC interference is the natural function of these compounds. However, despite the considerable interest in cyclodepsipeptide bioactivity, neither the biosynthetic gene clusters nor the native function of solonamides in nature is known. Understanding the natural role of such compounds could lead to a targeted search to discover similar, potential antivirulence compounds.

The purpose of this study was to genetically identify the biosynthetic gene cluster of solonamides and to determine the function of solonamides in P. galatheae by analyzing the physiological differences between a solonamide-negative mutant and the wild-type (WT) strain.

RESULTS

The gene sol in BGC2 of P. galatheae S2753 is responsible for solonamide production.

AntiSMASH 5.0 (21) analysis of the P. galatheae S2753 genome predicted eight biosynthetic gene clusters (BGCs) on the larger chromosome and three on the smaller chromosome (12). On the larger chromosome, four were classified as nonribosomal peptide synthetase (NRPS) or NRPS-like BGCs, and the predicted NRPS of BGC2 consists of four modules, consistent with the number of amino acid residues in the solonamides (Fig. 1). The in silico prediction indicated the loading of a phenylalanine by the first A domain. Two epimerization (E) domains in modules 2 and 3 correspond to the two d-amino acid residues, d-Leu and d-Ala, respectively, common for solonamides A to D. Therefore, BGC2 is a putative solonamide BGC, and the core 16-kb NRPS gene was tentatively named sol.

FIG 1.

Schematic representation of biosynthetic gene cluster 2 (BGC2) as predicted by antiSMASH and the predicted biosynthetic pathway of the solonamides. (A) ORFs encoded in the putative solonamides BGC. Genes with conserved known functions were marked with gene names. The core NRPS gene, sol, was highlighted in red. (B) Domain organization and modules of the NRPS encoded by the sol gene and proposed biosynthetic pathway to solonamides. Domain annotation: FAS, fatty acid synthase; C, condensation; A, adenylation; T, thiolation and peptide carrier; E, epimerization; TE, termination with a thioesterase domain.

Within the MIBiG 2.0 database (22), the sol gene displayed high similarity to a 14-kb NRPS-encoding gene responsible for the production of xenematides, a group of cyclodepsipeptides produced by Xenorhabdus nematophila (MIBiG accession number: BGC0001825) (23). Based on the proposed biosynthetic pathway of xenematides and the annotation of the sol gene (see Table S1 in the supplemental material), a biosynthetic pathway for solonamide biosynthesis was proposed (Fig. 1). The starting unit is either 3-hydroxyhexanoic acid (in solonamide A), 3-hydroxyoctanoic acid (in solonamides B and C), or 3-hydroxydecanoic acid (in solonamide D), potentially derived from fatty acid biosynthesis. In the first module, an l-phenylalanine is added, followed by a leucine epimerized to the d configuration by the epimerization domain. Similarly, in module 3, an l-alanine is adenylated and epimerized to a d-alanine, followed by the final incorporation of either l-leucine (solonamides A and B) or l-valine (solonamides C and D). Finally, lactonization is likely catalyzed by the thioesterase (TE) domain between the 3-hydroxy-group on the fatty acid starter unit and the thioester-bound carbonyl group of the l-leucine/l-valine. Chemical structures of solonamides A and B have been provided, while those of C and D have not, because both solonamide A and B were previously confirmed by nuclear magnetic resonance (NMR) (14), while only suggested structures of solonamide C and D have been reported in the literature (24).

In addition to the sol genes, the products of four other genes in BGC2 were predicted as additional biosynthetic enzymes in secondary-metabolite biosynthesis, i.e., a pimeloyl-acyl carrier protein methyl ester esterase (BioH), a putative two-domain glycosyltransferase, and two conserved proteins of unknown function (Table S2). Proteins with regulatory function in signal transduction were located downstream of the sol gene (Fig. 1; Table S2), including a transcription elongation factor (GreB) and a two-component system that is responsible for osmoregulation encoded by the histidine kinase gene envZ and the response regulator gene ompR. Genes associated with stress response as well as DNA damage and repair were also arranged in BGC2; i.e., dinF encodes a DNA-damage inducible transporter and a ppGpp synthase and hydrolase gene spoT. The flagellar biosynthesis protein FliL was encoded by a gene in BGC2 located about 27 kb upstream of sol gene (Fig. 1; Table S2).

To confirm the secondary metabolites derived from BGC2, the core biosynthetic gene, sol, was in-frame deleted, constructing the Δsol mutant. The sol gene was cloned to plasmids and reintroduced into the Δsol mutant to generate a complemented Δsol::sol strain. Plasmid pBBR1MCS2 was transferred into the Δsol strain to construct the Δsol::NC mutant, which was used as a control for the Δsol::sol mutant (Table 1). These strains were grown in 3-(N-morpholino)propanesulfonic acid (MOPS)-buffered Sigma sea salt medium supplemented with glucose (SSBG medium), and culture extracts were analyzed by ultrahigh-performance liquid chromatography coupled to diode array detection and high-resolution mass spectrometry (UHPLC-DAD-HRMS) (Fig. 2A). Accurate m/z values for the [M+H]⁺ adducts of solonamides A and B (solonamide A, m/z = 559.3490; solonamide B, m/z = 587.3803) were used to detect the metabolites, and HRMS fragmentation data compared to previous literature (14) were used to verify the identities of the compounds (Fig. S1). The tentatively identified solonamides C and D (solonamide C, m/z = 573.3647; solonamide D, m/z = 601.3960) (24) were also included in the comparison of the wild type and Δsol mutant, but with no MS/MS data available to compare with. In the wild-type S2753 and Δsol::sol cultures, solonamides A and B were both detected, but in the Δsol and Δsol::NC cultures, no production was observed, demonstrating that sol is involved in biosynthesis of solonamides in P. galatheae S2753. The tentatively identified solonamides C and D were also produced only by the wild-type and Δsol::sol cultures, further suggesting that these compounds are solonamide analogues. As controls, both holomycin and ngercheumicins were detected at similar levels in the wild-type and Δsol mutant cultures. Holomycin ([M+H]⁺ = 214.9943) is included in the chromatograms in Fig. 2A.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Characteristic(s)	Source or reference
Escherichia coli
PIR1	F− Δlac169 rpoS(Am) robA1 creC510 hsdR514 endA recA1 uidA(ΔMluI)::pir-116	Invitrogen
WM3064	thrB1004 pro thi rpsL hsdS lacZΔM15 RP4-1360 Δ(araBAD)567 ΔdapA1341::[erm pir]	Strain developed by William Metcalf at UIUC (66)
Photobacterium galatheae S2753
Wild type		10
ΔhlmE strain	Deletion of core gene hlmE for holomycin biosynthesis	12
Δsol strain	Deletion of entire core gene sol for solonamide biosynthesis	This study
Δsol::sol strain	Δsol; pBBR1-MCS2-sol	This study
Δsol::NC strain	Δsol; pBBR1-MCS2	This study
ΔhlmE Δsol strain	Deletion of sol and hlmE genes	This study
Staphylococcus aureus 8325	Gram-positive human pathogen	53
Vibrio anguillarum 90-11-287	Gram-negative fish pathogen	52
Plasmids
pDM4	sacB cat; R6Kγ origin	54
pBBR1-MCS2	neor kanr lacZa	67
pDM4-d-sol	pDM4 backbone carrying recombinant arms targeting sol gene	This study
pBBR1-MCS2-sol	pBBR1-MCS2 backbone carring sol	This study

FIG 2.

Chemical analyses and antimicrobial activity of the culture extracts. (A) Combined extracted ion chromatograms of solonamides A to D and holomycin in the Photobacterium galatheae S2753 wild-type cultures versus the Δsol::sol complemented strain (top) and the Δsol mutant versus the Δsol::NC strain carrying the empty plasmid (bottom). (B) Mass spectrometry data were analyzed via GNPS. The figure shows a part of the molecular network and the distribution of an unknown group on compounds in wild-type (pink) and Δsol (yellow) strains. The proportion of each color shows the relative abundance of the analyzed feature in each sample group. (C) Well diffusion assay showing the antimicrobial activity of the culture extracts of WT, Δsol, ΔhlmE, and ΔhlmE Δsol strains against the Gram-negative fish pathogen Vibrio anguillarum 90-11-287 and the Gram-positive human pathogen Staphylococcus aureus 8325.

The overall secondary metabolite profile of Δsol was similar to that of wild-type cultures.

In addition to the targeted analysis of solonamides, we compared the global secondary metabolite profiles of wild-type and Δsol mutant cultures in both exponential and stationary growth phases using LC-MS analysis. The deletion mutant showed the expected absence of solonamides, and the production levels of solonamides in the wild-type were similar for both late-exponential-phase and stationary-phase cultures (10 h and 24 h, respectively) (Fig. S3).

Additionally, Global Natural Products Social (GNPS) molecular network analysis was used to generate an overview of the overall secondary metabolites of the wild type and the deletion mutant, also after both 10 and 24 h of growth (Fig. S2). The main difference was observed between the different sampling time points. The primary difference, other than solonamide production, between the metabolite profiles of the wild-type and Δsol mutant samples was seen in a group of compounds detected only in the 24-h samples. The compounds were late eluting in the m/z range 595 to 677 Da but did not result in any plausible hits when searched against GNPS and databases such as Antibase or Reaxys. From the network analysis, several of the compounds were detected in both the wild type and the Δsol mutant (Fig. 2B). For example, the compound with m/z [M+H]⁺ 625.5498 was detected in both the WT and mutant, but at higher levels in the mutant, whereas compounds with m/z [M+H]⁺ 611.5340 and 621.5190 were detected exclusively in the wild type and the Δsol mutant, respectively (Fig. 2B).

To test if any potentially unidentified compounds harbored antimicrobial activity, we performed an agar diffusion assay by using the extracts from cultures of wild-type S2753, ΔhlmE (12), Δsol, and Δhlm Δsol mutants, and a negative-control extract from medium (Fig. 2C). The antibiotic activity in the wild-type and Δsol culture extracts against both Vibrio anguillarum and S. aureus was monitored No inhibition zone was observed for extracts from the ΔhlmE and ΔhlmE Δsol mutants cultures, suggesting, as previously found (12), that holomycin confers the significant antimicrobial activity of S2753.

Neither the growth nor the biofilm formation of P. galatheae S2753 was influenced in the Δsol mutant.

We compared the growth dynamics and biofilm formation of the Δsol strain and wild-type S2753. Since Mansson et al. (14) isolated comparable yields of solonamides from S2753 cultures grown in a medium based on Sigma sea salt, the growth dynamics of the Δsol mutant and the wild-type P. galatheae S2753 were determined in SSBG medium (24). The doubling times of wild-type, Δsol, Δsol::sol and Δsol::NC strains were 45.16 ± 4.68, 42.99 ± 1.18, 46.15 ± 5.89, 45.61 ± 3.28 min, respectively, and all cultures reached a stationary phase with the maximum cell density of 1 × 10⁹ CFU/mL (Fig. 3A). Small cell aggregations and big clumps were observed in all samples, likely resulting in the slightly decreasing numbers of CFU per milliliter at the later time point (Fig. 3A).

FIG 3.

Growth dynamics (A) and biofilm formation (B) of Photobacterium galatheae S2753 WT, Δsol, Δsol::sol complemented, and Δsol::NC cultures in SSBG medium. (A) The beige and blue arrows indicate the proteomics sampling points in the late exponential and stationary phases, respectively, and the asterisk indicates the initial time point of cell aggregation. (B) The crystal violet assay was used as the biofilm assay, and cultures of the holomycin-deficient ΔhlmE mutant were included as a control for the biofilm formation assay. Significant differences between groups were determined by Tukey’s test with one-way analysis of variance (ANOVA). Data are means and standard deviations (n = 6). ***, P < 0.05.

P. galatheae S2753 forms biofilm in stagnant cultures, and this is enhanced by holomycin production (12); we wondered whether it could also be influenced by solonamide molecules. However, as determined by a crystal violet assay, biofilm formation did not differ between the wild type and the Δsol mutants (Fig. 3B).

The motility of the Δsol strain decreased compared to that of wild-type S2753.

Swimming assays in soft-agar (0.3% agar) SSBG medium plates were set up, and the Δsol mutant was significantly less motile than the wild-type S2753 in the SSBG swim plates (Fig. 4). The diameters of rings formed by the Δsol mutant in soft-agar plates were 58.5% ± 10.6% of those formed by the wild-type strain (Fig. 4A) when grown at 25°C and 22.9 ± 6.5% at 40°C (Fig. 4). The Δsol::sol cultures exhibited restored motility (Fig. 4). Further, chemical complementation by adding fractions containing solonamide B to the plate also restored the ring diameters of the Δsol mutant to the level of the wild-type strain but did not enlarge the ring size of the wild-type strain (data not shown).

FIG 4.

Motility assay in 0.3% agar-SSBG medium plate of Photobacterium galatheae S2753 WT, Δsol, Δsol::sol complemented, and Δsol::NC cultures. (A) Plates were incubated at 25°C for 15 h (left) or at 40°C for 60 h (right). (B) Analysis of the motility ring radius sizes. Significant differences between groups were determined by Tukey’s test with one-way ANOVA. Data are means and standard deviations (n = 4). ***, P < 0.001.

Global proteomic analysis of the wild type and Δsol mutant.

To uncover the underlying physiological changes leading to the change in swimming (motility) by deletion of production of solonamides, we performed a proteomic analysis of wild-type S2753 and the Δsol mutant. For each strain, two cultures grown in SSBG medium were sampled at the late exponential and late stationary growth points, respectively. Samples were separated into cell and culture supernatant components for proteomic profile analysis and comparison. The proteomic analysis detected 3,062 proteins in total (59% of the total encoded proteins). Among these, 140 proteins significantly (Student’s t test, P < 0.01) changed over 4-fold (log₂, >2 or <−2) between the Δsol and wild-type cultures (Fig. 5A; Table S3). Only six of the 140 proteins were from the supernatant samples, four from the exponential time point and two from the stationary time point. The other 134 proteins were from cellular samples (Student’s t test, P < 0.01; log₂, >2 or <−2). In samples from the exponential growth phase, 15 proteins decreased and 44 increased in the Δsol strain compared to the wild type. In the stationary-phase samples, 29 and 49 proteins decreased and increased, respectively, in the Δsol cultures compared to the wild-type ones.

FIG 5.

Volcano plots of statistical analysis applied to proteomic samples (A) and COG function classification of proteins with significant changed relative abundance (B). (A) The plots are based on the fold change (log₂) and the P value (−log₁₀) of all proteins identified in four groups of data sets. Δsol sol deletion mutant; WT, wild type; exp, exponential-phase sampling point; stat, stationary-phase sampling point; SN, culture supernatant. The threshold value for the cutoff was a combination of a P value of ≤0.05 and a log₂ fold change of ≥2. Green and red circles indicate the relative abundances of proteins in the Δsol cultures that were statistically significantly lower or higher than those of the wild-type samples. The full list of significantly changed proteins can be found in Table S3. (B) All proteins with significantly changed relative abundances from panel A were classified according to the COG database, and the protein counts assigned to each category were plotted. Lower and higher relative abundances in the Δsol than the wild-type sample sets are indicated by blue/negative numbers and red/positive numbers, respectively.

The proteins were classified into the Clusters of Orthologous Groups (COG) classification system (25). Comparison of expressed patterns between COG classes showed that three known biological processes involving 17 COG groups were influenced by knocking out the sol gene (Fig. 5B). Approximately 28% of the 140 proteins functioned in metabolic processes, including transport and metabolism of amino acids, carbohydrates, coenzymes, lipids and inorganic ions, energy production and conversion, and secondary metabolite biosynthesis, transport, and catabolism. Moreover, a quarter were assigned to functions in cellular processes and signaling pathways, i.e., cell cycle control, cell division, cell envelope biogenesis, cell motility, posttranslational modification, signal transduction, intracellular trafficking, secretion, and defense. In addition, 17% were involved in information storage and processing of transcription, translation, ribosomal structure and biogenesis, replication, recombination, and repair. The remaining influenced proteins were either poorly characterized or without a COG class number. Functional classification via the EGGNOG database gave similar results.

Notably, the relative abundances of five motility-related proteins were significantly different between the Δsol mutant and the wild type (Fig. 6). The gene PHGAL_v1_a0091 encodes a protein homologous to the cyclic di-GMP phosphodiesterase CdgJ, and its relative protein abundance in Δsol decreased to only 2.14% of the wild-type value (log₂ = −5.54; P = 7.29 × 10⁻⁵) in the exponential samples. The relative abundance of several proteins involved in flagellum assembly (encoded by PHGAL_v1_b1049) and pilus assembly (encoded by PHGAL_v1_a0319) also decreased in stationary-phase samples of the Δsol strain compared to the wild type. Consistently, a negative regulator of flagellin synthesis (FlgM homolog, encoded by PHGAL_v1_0801) in the Δsol strain increased to more than 16-fold that of the wild type. In contrast, a flagellar hook-associated protein, FlgL (encoded by PHGAL_v1_a0817), increased up to 22-fold in the Δsol strain compared to the wild type. The other flagellar-assembly-related proteins were detected at a similar relative abundance in all sample sets.

FIG 6.

Heat map showing the fold changes of five motility-associated proteins in different comparison sample sets. Blue and red indicate that the relative abundance of protein was lower or higher in Δsol samples, respectively. Statistically significantly different (log₂ <−2 or >2; P < 0.05) data are indicated by asterisks.

The relative abundances of core enzymes in each BGC of S2753 were also compared. In addition to proteins related to the solonamide BGC, core proteins of seven other BGCs were detected in all cell samples (Table 2). Comparison between sample sets of the wild type and the Δsol strain showed that the abundances of core enzymes in BGC1 and BGC4 increased in the Δsol strain compared to those in the wild type in stationary-phase samples. However, none of these changes was statistically significant when a 4-fold change was used as the cutoff. The differences between the wild-type and Δsol exponential-phase cells were insignificant, as shown in Table 2. Other detected BGC core proteins were similar in abundance in wild-type and Δsol mutant samples.

TABLE 2.

Relative abundance of the core biosynthetic enzyme(s) of biosynthetic gene clusters^a

BGC ID	Predicted product(s)	ID of the core gene(s)	Relative abundance in Δsol compared to WT samples
1	Unknown NRPs	a0050	NS (exp); NS (stat, 3.59-fold)
2	Solonamides	a0131	Not applied
3	Ectoine	a1233	Not detected
4	Siderophore	a1543-47	NS (exp); + (stat)
5	NRPs	a1580	NS
6	NRP-T1PKs	a2046-48	Not detected
7	Beta-lactone	a2907-08-12	NS
8	Bacteriocin	a3214	Not detected
9	NRPs-like	b0071	NS
10	T1PK-NRPs	b0515-16	NS
11	Holomycin	b1251-59	NS

^a+, significant increase of >4-fold; NS, change either not significant or less than 4-fold; exp, exponential phase; stat, stationary phase.

DISCUSSION

Cyclodepsipeptides display a broad range of bioactivities, including antivirulence, antimalarial, antitumor, antifungal, and insecticidal activities (26). As a consequence of their unique structural properties, many cyclodepsipeptides interact with proteins but are more resistant to hydrolyzing enzymes, leading to a valuable potential in pharmaceutical and clinical usage (27). Hence, as one of the promising sources of biologically active compounds for drug discovery, the cyclodepsipeptide family has been studied extensively with respect to bioactivity, isolation, and biosynthesis, mainly focusing on cyclodepsipeptides produced by fungi and bacteria and especially Actinomycetes (26–29). Recent studies have also discovered many potent bioactive cyclodepsipeptides in marine Proteobacteria, including didemnins that were isolated from cultures of Tistrella mobilis and several compounds produced by Vibrionaceae, including kahalalides, solonamides, ngercheumicins, and unnarmicins (13, 14, 30–32). However, the genetic basis, i.e., the biosynthetic gene clusters, for these have not been identified. As a basis for molecular studies, BGC2 of P. galatheae S2753 was predicted by antiSMASH as a candidate for solonamide production. The 16-kb NRPS gene of BGC2, designated sol, was knocked out in this study, and the Δsol mutant did not produce solonamides. Reintroduction of sol restored the solonamide production in the mutant. Therefore, the sol gene was genetically identified as the core gene for solonamide biosynthesis of P. galatheae S2753. In contrast, the biosynthesis of the different precursors of solonamides (14, 24) could not be bioinformatically assigned to any genes in BGC2, indicative of the involvement of more genes outside BGC2 in solonamide production. Buijs et al. found that andrimid impedes the production of solonamides (33). Since andrimid is a fatty acid biosynthesis inhibitor (34), the biosynthesis of solonamides is very likely to involve more genes from the general fatty acid-biosynthetic pathways, as predicted in the previous studies, in line with the incorporation of 3-hydroxy fatty acids into solonamides (14, 24). In addition to confirming the core biosynthetic gene, this study also investigated the potential roles of solonamides in the native marine producer by comparing a series of physiological, phenotypical, and proteomic differences between the generated Δsol mutant and the wild-type P. galatheae S2753.

Some bacterial secondary metabolites play a mediation role in regulating the general metabolites profile and physiological development of the native producers (35). One of those secondary metabolites is the cyclodepsipeptide hormaomycin, produced by Streptomyces griseoflavus (36, 37), which stimulates the production of antimicrobial secondary metabolites and the formation of aerial mycelium of various actinomycetes at nanomolar concentrations (37). Buijs et al. (33) and Giubergia et al. (24) found that the solonamide BGCs were transcribed in the early exponential growth phase, while most other BGCs were expressed in the late exponential to stationary phase (24, 33). The similar relative abundances of solonamides in extracts of late-exponential-phase and stationary-phase cultures also indicate that solonamides are produced primarily during the exponential phase. Although solonamides were produced in an earlier growth phase than the other secondary metabolites, the deficiency in the production of solonamides did not lead to any significant changes in other extracted secondary metabolites (Fig. S2) or the relative abundance of their related core biosynthetic enzymes (Table 2). In contrast, the proteomic analyses found that the production of solonamides led to general metabolic profile changes involving many cellular processes, including cell motility processes (Fig. 5), consistent with the finding that the motility of P. galatheae was significantly changed by abolishing the production of solonamides. This likely indicates that the P. galatheae solonamides, unlike the Streptomyces compounds, which have dual regulatory functions, serve as signal molecules influencing mainly the physiology but not the secondary metabolite profile of P. galatheae.

In the soft-agar motility plate assay, the chemotaxis response drives the bacteria to form a ring moving outward from the original point of inoculation, and the size of these rings is determined by growth as the cells multiply and increase their overall motility through the agar, including swarming and swimming behaviors (38). Since the sol deletion mutant grew similarly to the WT and formed swarming colonies with morphology similar to that of wild-type P. galatheae S2753, the difference in swimming behavior likely causes the different ring sizes displayed by the WT and sol deletion mutant in the soft agar. Bacterial secondary metabolites can influence the motility physiology of the producing organisms. For instance, some lipopeptides, such as surfactants, affect the biofilm formation and motility behavior (39, 40), while the same function was more recently found for several NRP compounds in marine bioactive Proteobacteria (12, 41). For example, the depsipeptides pseudovibriamides A and B promote motility and reduce the biofilm formation of Pseudovibrio (41). In P. galatheae, biofilm formation is influenced by its antibiotic, holomycin (12), and we demonstrate here that swimming is influenced by the solonamides.

We speculated that proteome analyses could provide indications of how solonamides impact the physiology of S2753. The putative type IV pilin protein was significantly decreased in the mutant, whereas there was a significant increase in the FlgM negative regulator and the hook-associated protein FlgL in the stationary growth phase of the Δsol mutant compared to the wild type. This was in accordance with the observations of the swimming assay in 0.3% soft agar, as an increase in the FlgM alters the biogenesis of flagella (42) and the flagellar hook length affects the biofunction of flagella (43). Both could lead to a defect in motility. There was a significant decrease of a bis-(3′-5′)-cyclic dimeric GMP (c-di-GMP) phosphodiesterase in the Δsol mutant (Fig. 6). As a second messenger, c-di-GMP functions in mediating flagellum-dependent motility in several bacterial species (44). For instance, in the marine bacterium Vibrio cholerae, a high c-di-GMP level maintains the elevated biofilm matrix production and eventually results in complete repression of the flagellar production and less swimming (45). Abolishing solonamide production in P. galatheae might alter the degradation of c-di-GMP and might consequently decrease the motility of the Δsol. Our data, however, cannot point to the exact molecular mechanism by which solonamides decrease the diesterase.

The concentration of nutrients varies enormously in oceanic environments (46). As a result, marine bacteria have developed many strategies to enable their survival, one of which is chemotactic swimming to seek favorable chemical niches (47–49). Marine bacteria are often fast swimmers, because swimming speed is the most critical parameter resulting in rapid spreading, colonization, and increased uptake of dissolved nutrients (50). Here, we found that the solonamide-producing P. galatheae S2753 strains expanded faster and occupied more areas in low-nutrient swimming plates than the Δsol mutant (Fig. 4), suggesting a competitive ecological benefit of solonamides for the native producer P. galatheae S273 in the marine environments.

In conclusion, the core biosynthetic gene of solonamide production was successfully identified in P. galatheae S2753, and a solonamide-deficient (Δsol) mutant was generated for investigating the effects of solonamides in the native producer. The reduced motility in soft agar of the Δsol cultures indicated that solonamides are directly involved in the physiology of S2753. Based on the proteomics, we speculate that these physiological influences of solonamides might occur via regulatory proteins but the molecular mechanism remains unclear. Our results pave a way to further investigate the ecological and adaptive benefits of solonamide production in P. galatheae S2753, thus facilitating the discovery of further antivirulence compounds from similar ecological environments.

MATERIALS AND METHODS

Bacterial strains, plasmids, primers, and growth conditions.

A list of the strains and plasmids used in this study can be found in Table 1. P. galatheae was cultured in Difco Luria-Bertani (LB) broth (BD 244620) or Difco LB agar (BD 244520) during conjugation. In the physiological experiments, P. galatheae was grown in Difco marine broth 2216 (BD 279110), Difco marine agar 2216 (MA; BD 212185), or Sigma sea salt medium (24) supplemented with glucose (SSBG) at a final concentration of 0.2%. Glucose was dissolved in Milli-Q water to 20% (wt/vol) stock solutions, sterilized with a 0.2-μm filter, and stored at 4°C. Sucrose (Merck 57-50-1) was dissolved in Milli-Q water to a 20% (wt/vol) solution and autoclaved at 100°C for 60 min.

Escherichia coli PIR1 used in cloning was grown in modified LB broth (tryptone, 10 g/L; yeast extract, 5 g/L; NaCl, 1 g/L [pH 7.0]) and modified LB agar (tryptone, 10 g/L; yeast extract, 5 g/L; NaCl, 1 g/L; agar, 15 g/L [pH 7.0]) (51). The donor strain for conjugation, E. coli WM3064, was grown in LB broth or LB agar supplemented with 2,6-diaminopimelic acid (DAP) to a final concentration of 0.3 μM. Vibrio anguillarum 90-11-287 (52) was cultured in MA or basal agar (1% [wt/vol] agar, 3% [wt/vol] Instant Ocean [https://www.aquariumsystems.eu], 0.33% [wt/vol] Casamino Acids [BD Biosciences, 223050]) and 0.4% (wt/vol) glucose. Staphylococcus aureus 8325 (53) was cultured in BHI broth (Thermo Scientific, CM1135) or in basal agar with 1% (wt/vol) peptone (BD Biosciences, 211677) and 0.4% (wt/vol) glucose.

Cultures and precultures from cryostocks were streaked on plates, followed by inoculation of single colonies in liquid medium. S2753 was grown at 25°C either with or without shaking, while V. anguillarum, and S. aureus were grown without shaking at 25°C. MA plates with S2753 for colony counting were incubated at 40°C. E. coli was grown at 37°C with shaking for liquid cultures. Other growth conditions are specified in the descriptions of the relevant methods.

PCR amplification and gel electrophoresis.

PCRs were prepared using either TEMPase Hot Start 2× master mix K (VWR Chemicals, 733-2413), PrimeSTAR Max DNA polymerase 2× master mix (TaKaRa, R045B) or Q5 high-fidelity 2× master mix (New England Biolabs, M0492S) with a 100 pM concentration of each primer (forward and reverse), template DNA, and distilled water (dH₂O) up to the desired volume. Primers were obtained from Integrated DNA Technologies (Leuven, Belgium), and a list of primers can be found in Table 3. PCR amplifications were carried out using a T100 thermal cycler (Bio-Rad, catalog no. 1861096). The PCRs were designed according to the manufacturer’s guidelines.

TABLE 3.

Primers used in this study

Primer	Sequence (5′–3′)	Use
CmR Fw	GGCATTTCAGTCAGTTGCTC	Amplification of the cat gene in pDM4 (12)
CmR Rv	CCATCACAAACGGCATGATG
Dsol-P1	GCTCTAGATTCATCTGTCGGGTCTCTGG	Amplification of the right recombineering arm of sol gene
Dsol-P2	CTGATCAGGTCATTCGATGTTTTCTTGAGTTGGGTG
Dsol-P3	ACTCAAGAAAACATCGAATGACCTGATCAGGACAGG	Amplification of the left recombineering arm of sol gene
Dsol-P4	CCGCTCGAGATCGACAGGCCATCTTCTCC
Sol-Fw	CCCTCGAGTCTGGTGTTGCTTTGTCACC	Amplification of the promoter region and 5′ segment of sol gene
Sol-link1	CCGTTAACTTCCCTGCCGCATAAATAC
Sol-link2	AAGTTAACGGAACGCCGTAACCTG	Amplification of the 3′ segment of sol gene
Sol-Rv	CGCTCTAGAATCTTGGATGAGGGAGCCTG
del-sol-P1	CAGGAGTGAACGGCAGTC	Diagnostic PCR and sequencing when confirming the in-frame deletion mutant
del-sol P2	CGAGTACTCAATGGAGTCAGC
del-sol P3	CCATCACAGCAACATGCTGG
del-sol P4	CAGCATCAGGATCGGCAAC
sol-seq-1	AGGGAACAAAAGCTG	Sequencing of sol
sol-seq-2	CGAGCCAGGCAATATC
sol-seq-2R	ACGGGAAAGTTCGC
sol-seq-3	CCGCCGGTTGTCCTG
sol-seq-3R	TGCCAGTAACTCAGC
sol-seq-4	GGTGCGGGAGCAACA
sol-seq-4R	CTTTCAGGGAGGTTC
sol-seq-5	GCGATAAGCTTGCCG
sol-seq-5R	CGAAGTCATGGCCGG
sol-seq-6	AGGGCAGCGTTGTCC
sol-seq-6R	CCGTCCGGAAACCGT
sol-seq-7	ACTCTGGTCGGGACC
sol-seq-7R	GGCCGGTTGAGGTAG
sol-seq-8	TGGTCTGGCCCGGGG
sol-seq-8R	CCTGTAACATCTCTT
sol-seq-9	CCATTATTCAGACCA
sol-seq-9R	ACCATAACTGAGTTT
sol-seq-10	CAGACTCTGGTGGTG
sol-seq-10R	AGCCGCCAGTACCTT
sol-seq-11	GACCGGGAAGGCAGC
sol-seq-11R	CTATAGGGCGAATTG

For the initial amplification of the fragments for direct cloning, a two-step PCR was used with an increased annealing temperature for the second step. Q5 high-fidelity 2× master mix was used for this PCR with the following program. 98°C for 30 s; 5 cycles of 98°C for 10 s, 56°C for 30 s, and 72°C for 20 s/kb; followed by 28 cycles of 98°C for 10 s, 68°C for 30 s, and 72°C for 20 s/kb; and a final extension step at 72°C for 5 min.

For detection of DNA segments, either GelRed nucleic acid stain (5×) (Biotium, 41003) or 6× purple gel loading dye (New England Biolabs, B7024S) with ethidium bromide (EtBr) staining followed by separation with electrophoresis on 1% (wt/vol) agarose (Invitrogen, 16500-500) gels that were prepared using 1× Tris-acetate-EDTA (TAE) buffer (diluted from 50× TAE buffer; Thermo Scientific, B49). Gels were visualized using the Bio-Rad GelDoc XR+ and Image Lab software. Gel band extractions were carried out using the Zymoclean gel DNA recovery kit (Zymo Research, D4007). PCR product purification was carried out using a QIAquick PCR purification kit (Qiagen, 28106). DNA concentrations were determined using a DS-11 spectrophotometer (DeNovix). DNA sequencing in this study was undertaken by Macrogen Europe (Amsterdam, Netherlands). Sequencing results were analyzed using the ApE plasmid editor V.2.0.53c.

Plasmid construction.

The suicide plasmid pDM4-del-sol was constructed by the restriction cloning method. Approximately 1.0-kb upstream and downstream regions flanking sol were amplified using primer pairs Dsol-P1/2 and Dsol-P3/4 (Table 3). Overlap PCR was used to merge the amplified DNA fragments, and the subsequent products were cloned into the suicide vector pDM4 (54) by using the XbaI and XhoI digestion sites to form pDM4-del-sol. The mutant allele generated by PCR amplification and cloned onto the pDM4 plasmid is an in-frame deletion of the sol gene. For complementation, the sol gene and its promoter region were amplified using the primer pairs Sol-Fw/Sol-link1 and Sol-link2/Sol-Rv (Table 3), cloned into plasmid pBBR1-MCS2 using XhoI and XbaI, and sequenced using the sol-seq primer set (Table 3). Genomic DNA of wild-type P. galatheae S2753 was used as the template for amplifying the various segments in this section. All plasmids were purified and sequenced by Macrogen Europe to ensure that no mutations were introduced into the desired deletion mutant.

Genomic DNA and plasmid extraction.

Genomic DNA (gDNA) was extracted using the Wizard Genomic DNA purification kit (Promega, A1120) according to the manufacturer’s protocol. Plasmids were extracted using the QIAprep Spin miniprep kit (Qiagen, 27106) according to the manufacturer’s protocol with the following changes. After the wash step, an additional centrifugation step (1 min, ~11,000 × g) was done to remove residual ethanol. The column was transferred to the new Eppendorf tube and incubated at room temperature for 1 to 2 min to evaporate any ethanol left on the column. Restriction digests were performed to verify the recombinant plasmids using restriction enzymes from New England Biolabs.

Construction and complementation of the deletion mutant.

A detailed protocol for conjugation and genetic manipulation of S2753 can be found in reference 12. Plasmids were transferred into E. coli WM3064 via electroporation, and the recombinant strains were used as the donor in the subsequent conjugation procedures. The donor and the recipients, i.e., wild-type S2753 or the solonamide-deficient (Δsol) mutant, were grown to the mid-exponential phase, mixed by centrifugation, and mated on 0.22-μm-pore-size membranes placed on agar plates containing LB with 0.3 μM DAP for 4 h at 37°C. After mating, cells were recovered in LB medium at 25°C with shaking for 30 min and spread onto LB agar with 15 μg/mL chloramphenicol (pDM4 backbone vectors) or 50 μg/mL kanamycin (pBBR1MCS2 backbone vectors) for incubation at 25°C until transconjugants appeared. For the second homologous crossover, precultures of the single-crossover mutants were inoculated with a 1,000-fold dilution and incubated with shaking at 25°C until reaching exponential phase. Cultures of 100 μL were then plated on LB agar with 10% (wt/vol) sucrose and incubated overnight. Allelic exchange between pDM4-del-sol and the chromosome was used to make the in-frame deletion on the genome of P. galatheae S2753. Colonies grown on the sucrose plates and sensitive to chloramphenicol were purified and confirmed by PCR with the primers del-sol-P1 and del-sol-P4 (Table 3) according to the protocol described by Zhang et al. (12). DNA sequencing of the PCR products was used to verify that an in-frame deletion had occurred.

Chemical extraction.

Cultures in Sigma sea salt medium supplemented with glucose were incubated at 25°C with shaking for 48 h. Chemical extraction was performed on 10 mL culture as described by Giubergia et al. (24). An equal volume of ethyl acetate was added to each culture and mixed by inversion. The mixture was mixed gently for 10 min until a clear division of layers appeared. The organic phase (upper layer) was transferred to a glass vial and then transferred to a heating block (35°C) to evaporate the liquid using N₂ gas. Extracts were resuspended in methanol (MeOH) and stored at −20°C before analysis. Since solonamide production was not altered significantly by culture with different carbon sources, culture extracts for the overall secondary metabolite profile analysis were prepared by the same procedure on cultures grown in MOPS-buffered Sigma sea salt medium with 0.2% (wt/vol) glucose (SSBG) medium, and the exponential- and stationary-phase cells were sampled at 10 and 24 h, respectively.

Secondary-metabolite profile analysis by UHPLC-HRMS.

Ultrahigh-performance liquid chromatography–high-resolution mass spectrometry (UHPLC-HRMS) was performed on an Agilent Infinity 1290 UHPLC system (Agilent Technologies, Santa Clara, CA) equipped with a diode array detector. Separation was obtained on an Agilent Poroshell 120 phenyl-hexyl column (2.1 by 150 mm; particle size, 1.9 μm) with a linear gradient consisting of water and acetonitrile, both buffered with 20 mM formic acid, starting at 10% acetonitrile and increasing to 100% in 10 min, at which point the concentration was held for 2 min, returned to 10% acetonitrile in 0.1 min, and left for 3 min (0.35 mL/min, 60°C). An injection volume of 1 μL was used. MS detection was performed as described by Giubergia et al. (24). The secondary metabolite profile was analyzed by using Agilent MassHunter qualitative analysis B.07.00. Detection of [M+H]⁺ ions of solonamides A, B, and C was used to evaluate the presence of these compounds in Δsol mutants along with fragmentation data compared to the literature.

For GNPS molecular networking (55), MGF files were exported using Agilent MassHunter qualitative analysis B.07.00, and triplicates of data generated from the wild type, Δsol mutants, and extracted sterile medium were uploaded as five groups. A molecular network was created using the online workflow (https://ccms-ucsd.github.io/GNPSDocumentation/) on the GNPS website (http://gnps.ucsd.edu). The data were filtered by removing all MS/MS fragment ions within ±17 Da of the precursor m/z. MS/MS spectra were window filtered by choosing only the top 6 fragment ions in the ±50-Da window throughout the spectrum. The precursor ion mass tolerance was set to 0.5 Da, and the MS/MS fragment ion tolerance was 0.02 Da. A network was then created where edges were filtered to have a cosine score above 0.7 and more than 6 matched peaks. Further, edges between two nodes were kept in the network if, and only if, each of the nodes appeared in the other’s respective top 10 most similar nodes. Finally, the maximum size of a molecular family was set to 100, and the lowest-scoring edges were removed from molecular families until the molecular family size was below this threshold. The spectra in the network were then searched against GNPS’ spectral libraries. The library spectra were filtered in the same manner as the input data. All matches kept between network spectra and library spectra were required to have a score above 0.7 and at least six matched peaks. GNPS data were analyzed using Cytoscape (56).

Well diffusion assay for testing antibacterial activity.

The protocols described by Gram et al. and Hjelm et al. (7, 57) were used. Well diffusion plates were prepared by melting basal agar (for 90-11-287) and basal agar with 1% (wt/vol) peptone (for 8325) and adjusting the temperature to 44°C in a water bath. Glucose was added to a final concentration of 0.4%. One microliter of culture was added per mL agar and mixed by swirling, and a 50-mL medium-culture mixture was poured into 15-cm-diameter petri dishes. The plates were left to solidify and dry in a flow bench. Wells 6 mm in diameter were punched, and 40 μL extract was added to each well. Plates were incubated bottom down at 25°C and inspected after 24 h.

Growth experiments.

Precultures were diluted to 10³ CFU/mL in 25 mL SSBG medium (2% Sigma sea salt, 0.3% Casamino Acids, 0.2% glucose, and 40 mM MOPS [pH 7.5]) in 250-mL flasks. The cultures were incubated at 25°C with shaking at 160 rpm for 48 h. Growth dynamics were measured by determining either the optical density at 600 nm (OD₆₀₀) or the number of CFU every 3 h. Data were plotted using Origin 9.0 (Origin Lab). When grown in the microtiter plates, the precultures were prepared in the same way and diluted to 10³ CFU/mL. A volume of 200 μL of each biological replicate was added to the well. The microtiter plates were incubated at 25°C with consistently linear shaking, and the OD₆₀₀ was measured with a Cytation 5 imaging reader (BioTek) every 10 min for 48 h.

Biofilm assay.

A modified protocol based on the work of Djordevic et al., Jensen et al., and O’Toole and Kolter (58–60) was used. Precultures were diluted to an OD₆₀₀ of 0.01 in 1 mL SSBG. One hundred microliters of culture was transferred to the wells of a microtiter plate in biological triplicates. SSBG medium was included as a negative control. Border wells were filled with 100 μL Milli-Q water to prevent desiccation. The plate was incubated in a humidity chamber with a wet paper towel in the bottom for 48 h at 25°C. After incubation, the OD₆₀₀ was measured in a SpectraMax i3 reader (Molecular Devices). Culture media and nonadhering bacteria were removed, and the wells were washed with Milli-Q water and dried for 5 min in a flow bench. One hundred twenty-five microliters of 1% crystal violet was added to each well, and staining proceeded for 15 min. Crystal violet was removed, and wells were washed three times with Milli-Q water. EtOH (96%) was added to each well and mixed with crystal violet. The plate was incubated for 30 min; then, the samples were transferred to a new microtiter plate, and crystal violet intensity was measured at OD₅₉₀ in a SpectraMax i3 reader.

Motility assay.

Motility plates were prepared by adding agar to a final concentration of 0.3% (wt/vol) into SSBG medium. Precultures were prepared as mentioned above. A sterile inoculation needle was dipped in the overnight preculture. The same inoculation needle was stabbed into the agar layer of the plate, but not to the base of the petri plate, as described by Ha et al. (61). Plates were then incubated right side up at 25 or 40°C until swimming colonies were observed. When needed, the motility rings on the soft-agar plates were recorded with a Canon camera. To observe the swimming cells, bacteria from either the motility rings or the edge of colonies on MA plates were placed in 1 μL of SSBG medium located on glass slides and observed with a Nikon microscope. To observe the elongated cells in a swarming colonies (62) on the agar plates, the plates were placed directly under the 40× objectives and observed with a Nikon microscope.

Solonamide B purification.

One liter of P. galatheae culture in SSBG medium, including all cells and culture liquid, was extracted with ethyl acetate (EtOAc) (twice, 1 L each) to afford 623 mg of extract. The extract was subjected to semipreparative HPLC using a C₁₈ column (Phenomenex Luna C18; 100 Å, 5 μm, 250 by 21.2 mm), using a solvent gradient system of 10 to 100% acetonitrile (ACN)-H₂O (20 mM formic acid [FA]) in 30 min. Fraction collection was performed using MS detection by targeting the solonamide B [M+H]⁺ ion (m/z 587.4). The enriched fraction (90 mg) containing solonamide B was further purified by reverse-phase HPLC (RP-HPLC) using a phenyl hexyl column (Phenomenex Luna C6-Ph; 100 Å, 5 μm, 250 by 10 mm) and a gradient of 50 to 85% ACN-H₂O (50 ppm trifluoroacetic acid [TFA]) to yield solonamide B (8.4 mg, >90% purity).

Proteomics sample preparation and analysis.

Proteomics sample preparation and analyses were conducted as described by Wang et al. (63). Cells were grown in buffered SSBG medium, and samples were collected at exponential- and stationary-phase points (Fig. 3). Cultures were harvested at 5,000 × g for 5 min, and the supernatant was collected and filtered through a 0.02-μm film to remove cells. The cell pellets were washed twice with 1 mL ice-cold phosphate-buffered saline (PBS) buffer, centrifuged at 7,000 × g for 1 min, and transferred to Eppendorf protein LoBind tubes. Lysis buffer (64) was added, and samples were incubated at 95°C for 5 min and subsequently cooled on ice. Ice-cold acetone was added to the supernatant to a final percentage of 80% acetone and stored at −20°C for overnight to precipitate all proteins. The supernatant samples were centrifuged at 5,000 × g for 20 min, and then the acetone was removed carefully without disturbing the protein pellets. The remaining acetone was evaporated in a fume hood. After that, the protein pellet was resuspended in lysis buffer and boiled at 95°C for 5 min. All samples were treated as described by Lechman et al. (65) with few modifications. Samples were sonicated for 150 s using BioRuptor and the protein concentrations were determined using BCA Gold. A sample of 10 μg was adjusted with lysis buffer to 20 μL and then diluted threefold with digestion buffer (10% ACN, 50 mM HEPES [pH 8.5]). LysC was added in a 1:50 ratio, and the sample was incubated for 4 h at 37°C. Samples were diluted 10-fold with digestion buffer relative to the starting volume, trypsin was added in a 1:100 ratio, and samples were incubated at 37°C overnight. Since P. galatheae S2753 forms biofilms, to increase the sample quality, an extract centrifugation step at 20,000 × g for 10 min was added for all samples before they were diluted with 2% TFA to a final percentage of 1% TFA. Pipette tips of 250 μL were loaded by adding four layers of C₁₈ disks after activating the stage tips with 20 μL 100% MeOH, 20 μL buffer B (80% ACN, 0.1% FA), and two 20-μL portions of buffer A′ (3% ACN, 1% TFA). Samples after the extract centrifuge step were loaded onto the stage tip. Filters were washed twice with buffer A (0.1% FA in Milli-Q water), and samples were subsequently eluted with buffer B′ (40% ACN, 0.1% FA). Buffer was then evaporated in a SpeedVac 5301 instrument (Eppendorf) for 90 min at 60°C, and proteins were resuspended in buffer A* (2% ACN, 1% TFA). Concentration was determined using fluorometers and a NanoDrop One spectrophotometer. Five hundred nanograms of each sample was injected for proteomics detection.

Proteome data acquisition and analysis.

The proteomic data were analyzed with the Q-Exactive HF-X Orbitrap MS coupled to nanoLC EASY1200 at the DTU Proteomics Core, and the data acquisition approach was as described by Wang et al. (63). MS performance was verified for consistency by running complex cell lysate quality control standards, and chromatography was monitored to check for reproducibility. The raw files were analyzed using Proteome Discoverer 2.4. Label-free quantitation (LFQ) was enabled in the processing and consensus steps, and spectra were matched against the P. galatheae S2753 database obtained from NCBI (GCA_000695255.1). The proteomic data were processed using the Perseus software (66) (v. 1.6.15; https://maxquant.net/perseus/) as described by Wang et al. (63). Proteins were filtered out if they were present in fewer than three of the five replicates of any sample group. The normalized relative abundance values were then log₂ transformed, and missing values were imputed from a downshifted normal distribution (width, 0.3; downshift, 1.8). The volcano plot function in Perseus was applied to visualize the differences between samples, and significant differences were set to a fold change of more than 2 (log₂, >2 or <−2) and a false discovery rate (FDR) with a significance threshold (q) of 0.01.

Data availability.

The mass spectrometry data have been deposited in the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org; accessed 2 February 2022) via the PRIDE partner repository with the data set identifier PXD031490.