Pseudomonas putida Biofilm Depends on the vWFa-Domain of LapA in Peptides-Containing Growth Medium

The biofilm of Pseudomonas putida is complexly regulated by several intercellular and extracellular factors. The cell surface adhesin LapA of this bacterium is a central factor for the biofilm and, consequently, the regulation of lapA expression, for example, by Fis. It has been recently shown that peptides in growth media enhance the formation of P. putida biofilm, but not as a source of carbon and nitrogen. Moreover, the peptide-dependent biofilm appeared especially clearly in the fis-overexpression strain, which also has increased LapA. Therefore, we investigate here whether there is a relationship between LapA and peptide-dependent biofilm. The P. putida strains with inducible lapA expression and LapA without the vWFa domain, which is described as a domain similar to von Willebrand factor domain A, were constructed. Thereafter, the biofilm of these strains was assessed in growth media containing extracellular peptides in the shape of tryptone and without it. We show that the vWFa domain in LapA is necessary for biofilm enhancement by the extracellular peptides in the growth medium. The importance of vWFa in LapA was particularly evident for the fis-overexpression strain F15. The absence of the vWFa domain diminished the positive effect of Fis on the F15 biofilm.


Introduction
The two adhesins of Pseudomonas putida, LapA and LapF, are responsible for attachment, initial biofilm formation, and maintaining the structural integrity of mature biofilm. LapA is presumed to be an essential adhesin for attachment and biofilm formation, while LapF is characterised as an adhesin for cell-cell interaction in mature biofilm. However, the role of LapA in biofilm appears to be more critical [1][2][3][4].
The transport and exposure of Pseudomonas fluorescens LapA onto the cell surface have been extensively studied. Due to the similarity of the system to P. putida, it can be considered as a model for both bacteria. The LapA of P. fluorescens is transported to the cell surface via a type I secretion system (T1SS) that consists of three proteins, LapB, LapC and LapE [5]. LapB is expected to recognise the C-terminal signal sequence of LapA. The C-terminus of the protein is transported out of the cell, while the N-terminus remains in the periplasm [6]. LapA remains tethered to the outer membrane via LapE, inducing attachment to biotic and abiotic surfaces [7][8][9][10]. The periplasmic protease LapG releases LapA from LapE after the proteolytic attack of the N-terminus of LapA, thereby reducing the biofilm. Depending on the presence of c-di-GMP, LapD, submerged to the inner membrane, can inhibit the LapG activity by binding to it [11]. Unlike P. fluorescens, the P. putida LapA transporter genes are not located consecutively; lapBC (PP_0167 and PP_0166) form an operon, whereas lapE (PP_4519) is located separately. It is important to note that lapA and lapB of P. putida share the same, approximately 1 kbp long, promoter region, with the genes encoded in the opposite direction [12].
The LapAs are large extracellular proteins; in P. fluorescens Pf0-1 and P. putida KT2440 the size of the protein is 519 kDa and 888 kDa, respectively [12]. Based on the similarity of protein sequence, several domains have been found in LapAs that could be involved in its transport and adhesion. While LapA N-terminus remains in the periplasm, the extracellular repeats make up the majority of LapA. Although these repeats are not conserved among LapA homologs, it is suggested that the functionally similar repetitions provide a tubular structure, ensuring the location of the C-terminus far away from the cell [13]. The C-terminus of LapA, the extracellular part of this protein, contains the T1SS signal sequence required to transport with the LapBCE complex and a repeat-in-toxin (RTX) domain, described as a domain that binds calcium [10,12,14]. One particular domain of the C-terminus, the von Willebrand factor A-like domain (vWFa), has our attention. In P. putida strain KT2440, amino acids 8159-8283 of LapA are predicted to form the vWFa-domain based on the Pfam database and P. fluorescens Pf0-1 in 4710-4830, respectively [12]. The vWFa of P. fluorescens LapA is associated with adhesion into hydrophilic surfaces [10], but so far, the function of P. putida vWFa has not been described.
The von Willebrand factor (vWF) is a eucaryotic multimeric glycoprotein found in blood plasma, platelets, endothelial cells, and subendothelial cells. Its primary function is to induce the adhesion of platelets and thrombus formation in the case of vascular injury [15]. Several eukaryotic proteins, such as type IV collagen, type VII collagen, cartilage matrix protein, undulin, leukocyte β2 class receptors, complement components C2, B, etc., contain a domain A of vWF (vWFa). These proteins are mostly cell surface proteins or extracellular proteins [16]. In addition to eucaryotic proteins, domain A of the von Willebrand factor is found in several bacterial proteins, especially among extracellular adhesins such as LapA homologs [11]. Hence, domain A is particularly characteristic of extracellular proteins involved in cell adhesion, migration, homing, pattern formation, and signal transduction [16].
Chronic infection of all pathogenic bacteria begins with their attachment to the surface. Notably, bacteria have been reported to bind to vWF or have surface proteins with vWFalike domains used for attachment. In any case, the protein-protein interaction plays a significant role in developing virulence. For example, Staphylococcus aureus uses Spa protein to attach to vWF or its proteolytic products. S. aureus can attach to several blood and matrix proteins; however, the vWF is a target in the early stage of infection [17]. It has been demonstrated that the vWFa-like domain plays a central role in triggering the virulence of pathogenic bacteria. For example, the activation of Pseudomonas aeruginosa virulence genes depends on two conditions: quorum sensing and the mechanical signal involved in recognising host cells. PilY1 is a P. aeruginosa surface protein with a vWFa domain. The deletion of this domain was accompanied by a loss of virulence of P. aeruginosa, and it was thought that the binding of the PilY1 vWFa domain to host cells was a mechanical signal that, in addition to the quorum-sensing signal, triggers the expression of virulence genes [18].
We have previously shown that the biofilm of P. putida depends on LapA and the proteinaceous component in the medium [1,19]. Specifically, according to the crystal violet absorption, P. putida forms more biofilm when tryptone or poly-lysine are present in the medium [19]. In addition, we have ascertained that Fis activates lapA transcription in P. putida [20]. We presumed that additional expression of fis from the P tac promoter by IPTG, i.e., an artificial increase in the amount of Fis in the cell (so-called fis-overexpression), has a positive epistatic effect on the P. putida biofilm (Fis-enhanced biofilm). However, we recently ascertained that the positive impact of Fis on the P. putida biofilm depends on the presence of peptides in the medium [19]. In other words, the overexpression of fis in cells can enhance biofilm by activating lapA transcription only when peptides are present in the medium [19].
In this study, we hypothesised that the extracellular peptide promotes P. putida biofilm via LapA. To test this, we constructed a strain of P. putida in which lapA transcription is under the control of the P tac promoter, and the addition of IPTG regulates its transcription. In other words, we subtracted the Fis-enhancing effect on the P. putida biofilm from the LapA effect. In addition, we deleted the vWFa-domain sequence from P. putida lapA in order to ascertain the involvement of the vWFa domain in peptide-dependency of biofilm and found that P. putida biofilm is amplified by extracellular peptides if the LapA contains the correct vWFa-domain.

LapA Is Essential for Peptide-Dependent Biofilm
Knowing the peptide-dependency of P. putida biofilm [19], we tested whether it is based on LapA or LapF. We previously reported that LapA is a crucial factor in forming P. putida biofilm in LB and that LapF is required for P. putida cell surface hydrophobicity [1,2,20]. Thus, we assessed the biofilm formation of PSm and F15 deletant strains ∆lapA, ∆lapF, and double deletant ∆lapAF in the following media: M9-0.2CAA; M9-0.2CAA+tryptone, and LB (Table S2 and Figure 1). Figure 1. Biofilm of P. putida strains PSm, F15 and deletion-mutants ∆lapA, ∆lapF, and ∆lapAF constructed from them. "-" indicates original strain without deletions. Cells were grown in (A) M9-0.2CAA, (B) M9-0.2CAA+tryptone, and (C) LB medium. Cells were grown with or without 1 mM IPTG. Arithmetic means of at least three independent sets of measurements are shown. The 95% confidence intervals of the arithmetic means and the homogeneous groups are shown above the columns by lower-case letters, tested by following Bonferroni's test for unequal n. Identical letters in homogenous groups denote nonsignificant differences (p ≥ 0.05) between averages of biofilm.
The two terms are used hereafter: Fis-enhanced biofilm and peptide-dependent biofilm. Fis-enhanced biofilm appears in P. putida F15 strain when the expression of fis is increased by IPTG compared to the biofilm of F15 without IPTG supplementation (compare biofilms of P. putida F15 in Figure 1B,C). Peptide-dependent biofilm appears in P. putida strains when the peptides are present in growth media compared to the biofilms formed in growth media without peptides (compare biofilms in Figure 1A,C) [19]. If the arithmetic means of the samples are significantly different (p < 0.05), it is considered a Fis-enhanced or peptidedependent biofilm. Otherwise, the increased amount of Fis in the cells or the peptides in the growth medium does not affect biofilm.
The F15 has been a helpful strain in evaluating the positive effects of peptides on P. putida biofilm. The Fis-enhanced biofilm was formed in the presence of the peptides [19], the lapA transcription was activated by fis-overexpression [20], and lapF transcription was repressed [21]. We hypothesised that if LapA or LapF are involved in the peptidedependence of biofilm, then the peptide-dependence of biofilm should be absent in strains where lapA or lapF genes are deleted, even in cells with an elevated amount of Fis.
We observed that the deletion of lapA decreased the biofilm of P. putida wild-type strain PSm, regardless of the medium (Figure 1). The lapA deletion strain PSm∆lapA had 1.3-1.9 times decreased biofilm than the wild-type strain PSm (p < 0.032), except for the lapA, lapF double deletion derivative PSm∆lapAF in M9-0.2CAA+tryptone, whose biofilm was similar to PSm ( Figure 1). The lapF-deletion strain PSm∆lapF formed biofilms similarly to the wild-type strain PSm in the presence of peptides in media ( Figure 1B,C).
In the case of fis-overexpression strain F15, the Fis-enhanced biofilm appeared in the tryptone-containing growth media with a 2.3-fold difference in M9-0.2CAA+tryptone (p < 0.0001) and a 2.0-fold difference in LB (p < 0.0001), but not in M9-0.2CAA (1.1-fold difference, p = 1; Figure 1). The strain with the deleted lapA gene F15∆lapA did not produce a Fis-enhanced biofilm ( Figure 1B,C). At the same time, the lapF deletion strain F15∆lapF retained the positive effect of fis-overexpression on the F15 biofilm in M9-0.2CAA+tryptone and LB (3.9 and 2.0-fold differences, respectively, p < 0.0001), indicating that lapA expression is needed for the enhanced biofilm by Fis in the presence of peptides in the medium ( Figure 1B,C).
We previously observed the positive effect of poly-L-lysine (pK) on P. putida F15 biofilm; it restored the Fis-enhancing effect for P. putida strain F15 grown in the M9-0.2CAA medium [19]. We supplemented M9-0.2CAA with 0.4 mg/mL of pK and assessed the biofilm formation of lapA and lapF deletants of P. putida ( Figure 2). The Fis-enhanced biofilm appeared only in the presence of functional LapA ( Figure 2). LapF did not impact the Fis-enhanced biofilm ( Figure 2). However, the Fis-enhancing effect was observed for the F15∆lapAF double mutant (1.5-fold difference, p < 0.0012), probably due to an unknown mechanism rather than polypeptides or adhesins ( Figure 2). Biofilm of P. putida strains PSm, F15, and deletion-mutants ∆lapA, ∆lapF, and ∆lapAF constructed from them in the M9-0.2CAA medium with poly-L-lysine (pK) or without it. "-" indicates original strain without deletions. Cells were grown with or without 1 mM IPTG. Arithmetic means of at least three independent sets of measurements are shown. The 95% confidence intervals of the arithmetic mean and the homogeneous groups are shown above the columns by lower-case letters, tested by following Bonferroni's test for unequal n. Identical letters in homogenous groups denote nonsignificant differences (p ≥ 0.05) between averages of biofilm.
The results from P. putida wild-type strain PSm were more complex. Specifically, the presence of pK in M9-0.2CAA increased the biofilm of PSm∆lapA but not the biofilm of PSm∆lapF, which had functional LapA and lacked LapF ( Figure 2). However, as the positive effect of pK appeared on PSm∆lapAF double mutant, similarly PSm∆lapA, it indicates that the impact of poly-lysine could be more complex, and besides the adhesins, more factors could influence the biofilm. Because the absence of LapA abolished the peptide-dependence effect on the P. putida biofilm (Figure 1), the question arose whether the peptide dependency on the P. putida biofilm could be caused via lapA expression. Accordingly, the lapA expression in P. putida was assessed at transcriptional and protein levels. The lapA transcription was evaluated by β-galactosidase assay, and the abundance of LapA protein was estimated in crude cell lysates ( Figure 3). LacZ activity was measured using plasmid p9_PlapA1-8, which carries the lapA promoter region in front of the reporter gene lacZ (Table 1). The amount of adhesine determined from the crude lysate of cells grown with IPTG was divided by the amount of LapA determined from the crude lysate of cells grown without IPTG. The averages of at least 6 biological parallels are given, with 95% confidence intervals. The homogenous groups are shown above columns by lower-case letters. Identical letters denote nonsignificant differences (p ≥ 0.05) between the averages of the ratios.
This study Plasmids p9_PlapA1-8 951 bp long promoter-area of lapA cloned in front of lacZ (Amp r ) [20] p9TTBlacZ A promoter probe vector containing lacZ reporter gene (Amp r ); RK2-based vector [26] pSW I-SceI-expressing plasmid (Amp r ) [27] pGP-FpFy-Km-lacItac-lapFSD R6K suicide vector pGP704-L containing P tac promoter with lacI repressor gene (Amp r , Km r ) [21] pEMG R6K suicide vector (Km r ) [28] pEMG-lapB-lapA Derivate of pEMG containing in the opposite direction of 5 ends of lapA and lapB genes (Km r ) This study pEMG-lapB-Pm-xylS-lacI-P tac -lapA Derivate of pEMG-lapB-lapA containing Pm promoter in front of lapB 5 end and P tac promoter in front of lapA 5 end, xylS and lacI genes (Km r ) This study pEMG-P tac A-natB Derivate of pEMG-lapB-lapA containing P tac promoter in front of lapA gene and native promoter region in front of lapB gene, lacI gene ( Our results revealed that transcription of lapA was dependent on the medium but did not correlate with the F15 biofilm (Figures 1 and 3A) [19]. F15 revealed the positive effect of fis-overexpression only in the presence of peptides ( Figure 1). In all media, LacZ activity in F15 in the presence of 1 mM IPTG was 1.26 (M9-0.2CAA), 1.3 (M9-0.2CAA+tryptone), and 1.48 (LB) times higher than in the non-IPTG variant (p < 0.0001), indicating that regardless of the growth medium, the lapA transcription was always increased in the presence of the higher amount of Fis ( Figure 3A). LacZ activity in PSm or F15 containing the empty vector p9TTBlacZ was less than 0.5 Miller units.
The crude cell lysates were prepared from the bacteria grown for 18 h in the appropriate medium to quantify the adhesins LapA and LapF. Because cell lysates prepared from different media were electrophoresed on PAA gels separately, the ratio of adhesins with and without IPTG was compared separately in every medium. In contrast to the level of lapA transcription, the amount of LapA protein in the crude cell lysate was affected by the content of the growth medium of bacteria ( Figure 3B-D).
Since the effect of Fis on the F15 biofilm was evident (Figures 1 and 2), we focused on comparing the amount of LapA in F15. Overexpression of fis in the tryptone-containing medium increased the amount of LapA 1.50 times (M9-0.2CAA) and 1.46 times (LB) in the F15 strain, which correlated with the previously observed increase in F15 biofilm formation in the same growth conditions (Figures 1 and 3C) [19]. However, the fis-overexpression did not elevate the amount of LapA in F15 grown in the M9-0.2CAA medium ( Figure 3C). Thus, the unchanged amount of LapA in fis-overexpressing cells might be the reason why F15 did not form a Fis-enhanced biofilm in the tryptone-free M9 medium ( Figure 1).
It is known that Fis represses lapF transcription [21]; therefore, we used LapF as a control for LapA measurements. Indeed, in contrast to LapA, the fis-overexpression decreased approximately 4 times the amount of LapF in F15 in every media compared, indicating no correlation with a media-dependent biofilm of P. putida F15 (Figures 1 and 3D).
In conclusion, the reduced amount of LapA in crude cell lysate prepared from cells grown in a medium without tryptone correlates with a reduced amount of P. putida biofilm.

Construction of a Strain with a Modified lapA Transcription and a LapA Strain without a vWFa Domain
Although we knew that Fis increases lapA transcription and LapA is required for biofilm formation (Figures 1 and 3), we had to split the Fis-effect from the LapA-effect on the biofilm. There are two reasons: (i) Fis can enhance biofilm through another factor than LapA, and (ii) P. putida forms the biofilm even in the absence of LapA in the cell [1,20]. Therefore, P. putida strains were constructed in which native lapA is under the P tac promoter, and IPTG controls its transcription. P. putida strain PaW85 (isogenic for KT2440), one of the standard strains of P. putida, was used for this purpose (Table 1). Additionally, the PSm derivative of the PaW85 was used to evaluate the biofilm in the context of PSm and F15 ( Table 1).
The construction of the lapA expression system was problematic for several reasons. Firstly, lapA is a very long gene (26 kbp) [12]; therefore, we could not use vector systems. Secondly, chromosomal lapA shares a promoter region with its transporter genes lapBC encoded opposite to lapA [12,20]. The disruption of the joint promoter region can affect the transcription of both genes. We used several expression systems to control the transcription of both lapA and lapBC genes. For example, an expression system was constructed that ensured the expression of lapA from the P tac promoter by IPTG and the transcription of lapBC from the Pm promoter by benzoate (data not shown). In this case, transcription in both directions was regulated by extracellularly added effectors. Unfortunately, the addition of phenolic compounds to the medium affected the biofilm formation per se; these strains were excluded from the study (data not shown). We had to use a construct that had the P tac promoter immediately upstream of the lapA gene. The reversely expressed lapBC retained the entire native promoter area without the possibility of external regulation by effectors. With this system, we constructed a P-PANB strain from PaW85 (Table 1). This variant is not ideal because, despite induced lapA transcription, the effect on the biofilm may remain modest, as the natively expressed transporter may begin to limit LapA exposure on the outer membrane.
The third part of the LapA transporter, LapE, is described in P. fluorescens as an outer membrane subunit where LapA N-terminus will trap. In this way, the extracellular LapA is anchored in the outer membrane [32]. The lapE deletion prevents the LapA transport to the outside of the cell, and the biofilm should weaken. Thus, this deletant strain was used as a negative control for P-PANB strains. The lapG deletion, on the other hand, was used to construct a positive control. LapG is a periplasmic protease that, when activated, cleaves the N-terminus of LapA and releases LapA into the environment [32], resulting in a reduction of biofilm. Thus, the deletion of lapG should produce a robust P. putida biofilm, as LapA is not removed from the outer membrane. Before this research, we had constructed lapE and lapG deletion variants from P. putida strain PSm, PSm-E, and PSm-G, respectively. Therefore, we created the IPTG-inducible lapA transcription strains PSm-E-PANB and PSm-G-PANB, based on PSm-E and PSm-G, respectively ( Table 1).
The sequence coding a von Willebrand factor type A (vWFa) domain in P. putida lapA was deleted (from amino acid S8167 to 8281P, designated Avwf − ). Recent studies have shown that the deletion of the vWFa domain in P. fluorescens LapA reduces biofilm formation to glass and hydrophilic plastic [10,32]. The von Willebrand factor itself, after which the bacterial adhesin domains were named, is a multimeric glycoprotein important for the protein-protein interaction that causes thrombus formation and bleeding arrest after injury of eukaryotic tissue [15]. Inspired by these works, we asked whether the vWFa domain of P. putida LapA may have a role in enhancing peptide-mediated biofilm. In other words, is the LapA vWFa domain required for peptide-dependent biofilm formation.
First, it was necessary to evaluate the dependence of lapA expression on IPTG in the constructed strains. For this purpose, the amount of LapA from the crude cell lysates was assessed by SDS-PAA gel-electrophoresis. In all strains containing the P tac promoter in front of the lapA gene, LapA was visibly increased in all media by 1 mM of IPTG ( Figures 4A, 5A and 6A). In addition to LapA, the unknown protein complexes were visible in these gels that could not be LapA. The unknown protein complexes (probably the complexes of ribosomes) migrate slower than LapA, appear irregularly, and appear in PSm∆lapA, PSm∆lapF ( Figures 4A, 5A, 6A, 7B and S5-S12), and PSm∆lapAF strains ( Figures S5-S7). However, we cannot exclude that lapA would or would not be expressed in P-PANB without IPTG, as unknown protein complexes migrated higher or at approximately the same height as LapA in SDS-PAA gels ( Figures 4A, 5A and 6A).

LapA Domain vWFa Is Involved in the Peptide-Dependent Biofilm of P. putida PaW85
We were interested in the behaviour of P. putida PaW85-based strains. These strains do not contain mini-Tn7 insertion after the glmS gene, and lapA transcription is directly regulated by IPTG without indirect effects on the fis-overexpression in the cell (Table 1). It is essential to consider two facts: (i) LapA exposure on the outer membrane is complexly regulated and depends on proteolysis and the presence of effectors, (ii) the expression of LapA transporters (lapBC and lapE) is not regulated in these strains.
The biofilm was evaluated by every medium separately. In the M9-0.2CAA medium, the biofilm of all PaW85-based strains was similar ( Figure 4B). Although P-PANB forms more biofilm in the presence of IPTG than in its absence (p = 0.025), P-PANB biofilm does not differ from P-PANB-Avwf − biofilm (p > 0.108); there is no significant biological difference among the biofilms of these strains.
In the LB medium, the P-PANB biofilm was increased 1.3-fold in the presence of IPTG (p < 0.0001), whereas deletion of the vWFa domain resulted in the independence of IPTG ( Figure 4D). The biofilm of the P-PANB-Avwf − strain was similar regardless of the presence or absence of IPTG ( Figure 4D). The biofilm of the original strain PaW85 did not depend on the vWFa domain ( Figure 4D).    Inclined arrows indicate an unknown protein complex. (C) Biofilm of P. putida F15-based strains. An arithmetical average of at least three biological parallels with 95% confidence intervals are given. Homogeneity groups are shown above the columns by lower-case letters. Identical letters denote nonsignificant differences (p ≥ 0.05) between averages of biofilm. Multifactorial ANOVA was used for statistical analysis, analysed separately by media.
In a polylysine-containing M9-0.2CAA medium, IPTG did not affect the biofilm of the P-PANB strain ( Figure 4E). Still, the biofilm of vWFa domain deletion strains (Avwf − ) was decreased compared to the original strains from which they were constructed ( Figure 4E). Independent of IPTG, the strain P-PANB-Avwf − had 1.5 times less biofilm than the P-PANB strain (p < 0.0001, Figure 4E). For PaW85 versus P-Avwf − , deletion of the vWFa domain reduced biofilm 1.2-fold (p = 0.063) and 1.4-fold (p < 0.001) in the absence and presence of IPTG, respectively ( Figure 4E).
Our previous study showed that adding water-soluble cellulose to a medium enhances P. putida biofilm [19]. Deleting the vWFa domain did not affect the biofilm in the M9-0.2CAA+cellulose medium; both vWFa-deletion strains behaved similarly to its origin strain, P-Avwf − similarly to PaW85 and P-PANB-Avwf − similarly to P-PANB ( Figure 4F). However, the addition of IPTG to the medium increased 1.2-times biofilms of P-PANB strains compared to cells grown without IPTG, P-PANB (p = 0.006) and the P-PANB-Avwf − strain (p < 0.0001; Figure 4F). Biofilms in the M9-0.2CAA+PGA medium were comparable and did not depend on the lapA expression with IPTG or the presence of the vWFa domain ( Figure 4G).

Biofilm of P. putida PSm∆lapE-and PSm∆lapG-Derived Strains
We constructed a lapE deletion mutant from PSm, the PSm-E, which lacked the third subunit of LapA transporter, LapE ( Table 1). The lapE deletion reduced 1.6 and 1.7-times the biofilm of PSm in the M9-0.2CAA medium with 1 mM IPTG (p < 0.0001) and without IPTG (p < 0.001) accordingly ( Figure 5B).
Similar results were obtained in the M9-0.2CAA+tryptone medium, where the biofilm of PSm-E was decreased 1.6 and 1.7 times with and without the IPTG in the medium, respectively, compared to PSm ( Figure 5C). However, in the LB medium lapE deletion did not affect the biofilm of P. putida wild-type strain PSm ( Figure 5D).
Since the transporter of LapA was incomplete in ∆lapE-based strains, we expected that manipulating the lapA expression by IPTG or deleting vWFa would not affect the biofilm in these strains. Indeed, the IPTG did not affect PSm-E-PANB and PSm-E-PANB-Avwf − biofilm in any of the used media ( Figure 5B-D). However, in the M9-0.2CAA medium, for some reason that we cannot explain, the vWFa deletion reduced 1.3 times PSm-E-PANB biofilm (p < 0.001; Figure 5B).
Deletion of lapG should trap all LapA transported on the outer membrane and enhance the biofilm. Indeed, all strains with lapG deletion ( Figure 6) have a more robust biofilm than the wild-type PSm in all media ( Figure 5). The presence of 1 mM IPTG in the growth medium increased PSm-G-PANB biofilm compared to the absence of IPTG increased biofilm 2.27 times in M9-0.2CAA (p < 0.0001), 1.6 times in M9-0.2CAA+tryptone (p < 0.0001) and 1.23 times in LB (p < 0.0001; Figure 6B-D). The deletion of the vWFa domain from LapA (P-PANB-Avwf − ) did not reduce the biofilm to the same level as that detected in the PSm-Avwf − , indicating that LapA in significant excess may affect biofilm without the vWFa domain (Figures 5 and 6). Additionally, the PSm-G-PANB strain without IPTG had a relatively robust biofilm compared to PSm in the presence of tryptone in media ( Figures 5C,D and 6C,D). These results suggest the P tac promoter may still leak. However, we cannot state that this would result in a relatively robust biofilm of PSm-G-PANB in media without IPTG.

The vWFa Domain Is Essential for F15 Biofilm
We hypothesised that the Fis-enhanced biofilm of P. putida F15 in LB medium is caused, at least in part, by the positive effect of Fis on lapA expression [1,20]. Although the effect of Fis on lapA transcription was modest, the lapA deletion in F15 vanished this effect, indicating that LapA could be involved in the biofilm enhancement by Fis (Figure 1) [1]. Therefore, we constructed an F15-Avwf − strain in which the fis was possible to overexpress by IPTG, and LapA lacked the vWFa-domain. We expected to see the reduced effect of fis-overexpression on the biofilm in this strain.
The fis-overexpression in F15 and F15-Avwf − strains by IPTG was assessed in M9-0.2CAA, M9-0.2CAA+tryptone and LB media ( Figure 7A). The expression of fis was similar in both strains in all media tested, and IPTG induced the expression of fis from the P tac promoter in these strains ( Figure 7A). After that, the LapA presence in the crude cell lysates of F15-Avwf − strains was assessed ( Figure 7B). The ratio of LapA in cells grown in the presence of 1 mM IPTG versus without IPTG was similar to F15; 1.21 (95% confidence interval of arithmetical mean 0.60) in M9-0.2CAA, 2.04 (1.48) in M9-0.2CAA+tryptone and 1.58 (1.22) in LB medium. Although the differences in LapA ratios of F15-Avwf − were statistically insignificant in the media assessed, LapA ratios increased similarly to F15 when IPTG and tryptone were added to the media ( Figure 3C).
F15 and F15-Avwf − strains formed biofilm similarly in M9-0.2CAA, without a Fisenhanced biofilm ( Figure 7C). Despite the similar amount of LapA and Fis in both strains, differently from the strain F15, the IPTG did not improve F15-Avwf − biofilm in M9-0.2CAA+tryptone or LB medium F15 ( Figure 7C). In contrast, the deletion of the vWFa domain did not abolish the Fis-enhanced biofilm in the M9-0.2CAA+poly-L-lysine medium ( Figure 7C). This could be explained by a strong positive charge effect of poly-L-lysine that binds negatively charged LapA, regardless of the vWFa-domain.

Discussion
The P. putida biofilm is a dynamic phenotype depending on the growth phase of bacteria and the environment where the bacterium lives [1,19]. Our previous research showed that peptides in growth media increase the P. putida biofilm [19]. In addition to the fact that the peptides may act as an additional source of C and N, the peptides may induce a more substantial biofilm by being a structural component of the biofilm matrix [19].
To study biofilm, we used P. putida strain F15, which carries the additional fis gene under the control of P tac (IPTG). We used the fis-overexpression strain F15 because the use of strains lacking or down-regulating fis was not possible due to the essentiality of the fis gene or instability of the constructed strain. Surprisingly, this artificial strain served as an excellent tool for identifying the factors influencing P. putida biofilm. Although it is an overexpression strain, it is possible to study the physiology of the bacterium. For example, Fis represses transcription of lapF through a single binding site in front of lapF, thereby reducing cell surface hydrophobicity [2,21]. Mutation of the Fis-binding site restored lapF transcription, amount of LapF in cells, and ultimately, surface hydrophobicity [2,21]. Hence, this direct effect of Fis on the bacterial phenotype through one binding site in front of another gene encouraged us to use F15 in LapA expression studies as well.
The main aim of this study was to assess the involvement of surface adhesins, LapA and LapF, in the peptide-dependent P. putida biofilm. We knew that the peptides increase the biofilm of the P. putida wild-type strain PSm, and the so-called Fis-enhancing effect on the F15 biofilm only occurs in the presence of the peptides [19]. Moreover, we assessed the Fis-enhancing effect on biofilm in LB and suggested that the searchable factor could be LapA, as Fis activates lapA transcription and increases the amount of LapA in F15 in LB [1,20]. Surprisingly, Fis activates lapA transcription in all studied growth media, but the increased quantity of LapA and Fis-enhanced biofilm appeared when growth media were supplemented with peptides (Figures 1-3). It indicates that in the absence of peptides in growth media, the post-transcriptional regulatory steps could affect the expression of lapA more intensively. For example, in the absence of peptides in the growth medium, the translation of lapA may be downregulated, LapA itself may be effectively degraded, or LapA may be released from the surface due to the activity of LapG. We can only speculate which factor might cause the peptide-dependent post-transcriptional regulation of lapA expression. For example, the RsmA, a component of the GacS/A signalling pathway, binds RNAs and can regulate translation initiation, RNA stability, and/or transcript elongation [33] and the involvement in the regulation of the GacS/A signalling pathway P. putida lap-genes expression has been previously shown [34,35]. One of three RsmA homologs in P. putida, RsmE, down-regulates lapA expression in LB medium [35]. On the other hand, LapA can be efficiently released from the outer membrane due to the increased amount of active LapG in the periplasm. The amount of active LapG depends not only on the expression of the lapGD operon, the amount of protease in the periplasm, but also on the effectors such as cdi-GMP and ppGpp; and the availability of calcium [4,36,37]. However, we cannot exclude the possibility that other regulators are involved in the peptide-dependent expression of lapA. Thus, the post-transcriptional steps in lapA expression, including translation and the amount of LapA on the cell surface, may depend on the components of the growth medium.
The importance of LapG proteolytical activity appears in comparing P-PANB, PSm-E-PANB, and PSm-G-PANB biofilms in the growth medium without tryptone (M9-0.2CAA). Namely, LapA quantity is elevated in all strains by IPTG supplementation (Figures 4A,  5A and 6A), while only PSm-G-PANB, the lapG deletion strain, had a robust biofilm in the M9-0.2CAA medium ( Figures 4B, 5B and 6B). Thus, proteolytic cleavage of LapA from the outer membrane probably affects peptide-dependent biofilm formation more than the transporters. However, we cannot exclude the possibility that P-PANB biofilm is a result of a lack of LapA transporters, indicating peptide dependent regulation of lapBC and/or lapE genes.
The P-PANB strain constructed from PaW85 had only a modest, albeit significant, effect on biofilm in peptide-containing media upon induction of lapA transcription by IPTG ( Figure 4). However, the Avwf − -strains of PaW85, without the vWFa domain in LapA, had a significantly decreased biofilm relative to the strains with the correct LapA ( Figure 4). The importance of the vWFa-domain to the biofilm was particularly evident when P-PANB and P-PANB-Avwf − were compared. Both strains had lapA transcription under the control of P tac promoter (IPTG). Still, the absence of the vWFa-domain in LapA eliminated the effect of the elevated amount of LapA by IPTG in the media containing the peptides (Figure 4). We decided to use previously constructed PSm∆lapE and PSm∆lapG strains as a negative and positive control of inducible lapA expression. In the first case, increasing the amount of LapA or the absence of the vWFa domain should not significantly affect the biofilm because in the absence of LapE, LapA does not reach the cell surface. Indeed, in general, the biofilm of lapE-minus strains did not depend on the induction of lapA transcription by IPTG or the absence of vWFa ( Figure 5). In addition, experiments on the lapG deletion strain confirm the importance of the vWFa domain for P. putida biofilm formation ( Figure 6). The LapA strain without the vWFa domain resulted in less biofilm than the strains with correct LapA, although the absence of lapG always produced a stronger biofilm than the wild-type strain (Figures 5 and 6). In sum, we can conclude that not only does the presence of LapA cause the peptide-dependent P. putida biofilm, but LapA must also have the correct vWFa domain.
Since the absence of the LapA vWFa domain negatively affected the biofilm in P. putida wild-type strains, we investigated the effect of the lack of this LapA domain on the F15 biofilm (Figure 7). So far, we hypothesised that, among other factors, Fis could increase the F15 biofilm via lapA because Fis activates lapA transcription to some extent [20]. The complete loss of the Fis-enhancing effect for biofilm in LB and M9-0.2CAA+tryptone medium was astonishing when we deleted the vWFa-domain sequence from the F15 lapA gene, F15-Avwf − (Figure 7). Therefore, the enhancing effect of the Fis to P. putida biofilm described so far must be entirely or predominantly dependent on LapA. Consequently, Fis enhances lapA transcription, at least to some extent, and the LapA vWFa domain increases biofilm when peptides are present in the medium.
However, the question arises why the lack of vWFa domain had a pronounced impact on the F15 biofilm, whereas wild-type PaW85 with elevated LapA had only a modest effect (Figures 4 and 7). It seems that Fis may more broadly control the lapA expression and its persistence on the cell surface. For example, Fis may indirectly regulate LapA persistence on the cell surface through other genes, whereas the wild-type strain undergoes native strong down-regulation. As a result, the wild-type bacterium may lose most of the expressed lapA to the environment. Indeed, according to our preliminary data, Fis represses transcription of the lapGD operon (data not shown), and it may explain at least partly a substantial biofilm enhancement with elevated Fis in P. putida F15. However, the transport and maintenance of LapA on the cell surface seems to be more critical for biofilm than transcriptional regulation. Consequently, the role of Fis in regulating the expression of other lap genes or the peptide-dependent regulation needs to be further investigated.
The LapA vWFa domain of P. fluorescens has been previously described as essential for hydrophilic surface binding [10,32]. Our research has shown the importance of poly-Llysine, a hydrophilic positively charged biomolecule for P. putida's LapA vWFa-domain ( Figure 7). Notably, hydrophilic Na-cellulose and PGA, which are negatively charged biopolymers, have no direct relationship to biofilm enhancement via vWFa ( Figure 4F,G). Referring to our previous study, neither uncharged poly-DL-alanine nor negatively charged poly-L-glutamate had a biofilm-enhancing effect on P. putida F15 [19]. However, the positively charged spermine enhanced P. putida F15 biofilm similarly to poly-L-lysine [19].
Moreover, the F15-Avwf − strain entirely or largely lost the positive effect of the peptides on biofilm formation ( Figure 7C). Therefore, the surface's hydrophilicity and positive charge appear to be necessary for the binding LapA by the vWFa domain to the surface. However, we cannot say whether vWFa is specific for a particular sequence or whether it binds to a surface only by its electrostatic forces, which is an issue for future research to explore.
However, how precisely LapA binds to the surface has not been elucidated. One explanation may be the importance of other LapA domains for surface binding. For example, the F15∆lapA and F15-Avwf − biofilms in the medium M9-0.2CAA+pK form biofilm differently (Figures 2 and 7C). The complete absence of LapA eliminates the Fisenhancing effect for biofilm, whereas LapA without the vWFa domain is significantly reduced but does not entirely abolish it. Thus, binding through another LapA domain may be necessary. However, it may simply be an electrostatic effect appearing in excess of adhesin and biopolymer. Due to the increased amount of Fis in P. putida F15, the cell surface remains hydrophilic, the amount of LapA increases, and the amount of LapF decreases drastically (Figure 3) [2]. The resulting biofilm formation may be promoted through non-specific binding between the positively charged poly-lysine and LapA.
In conclusion, this research was initiated because we had previously determined that LapA, extracellular peptides, and increased Fis in the cell promoted P. putida biofilm formation. The question was whether these factors are interrelated. The LapA domain of vWFa was found to be essential for biofilm enhancement when peptides were added to the medium. It was particularly evident in P. putida with an increased amount of Fis. This so-called biofilm-enhancing Fis effect wholly or partially depends on the LapA vWFa domain. In addition, it seems that the most crucial regulatory step in the lapA expression and maintenance of LapA on the cell surface is the post-transcriptional and not the transcriptional step.
The deletion strains and lapA expression strains were constructed according to a previously published protocol [28]. To this end, constructed pEMG vectors (Table 1) were electroporated [42] into the corresponding P. putida strain to obtain a cointegrate. Plasmid pSW was then introduced into the cells, which generated deletion variants from the strains by activating homing endonuclease SceI expression [27]. All vectors and modifications on the P. putida chromosome were checked by sequencing to exclude PCR and recombination errors (Table S1).

Assessment of Biofilm
Biofilm formation was monitored in flat-bottom microtiter plates (Greiner Bio-One GmbH, Kremsmünter, Austria; 96-well cell culture plate Cat-No 655180) by the Fletcher method [43]. A 100 µL measure of LB or M9-0.2CAA medium with or without biopolymers and 1 mM of IPTG was placed in the well and inoculated with 5 µL overnight-grown bacteria. Bacteria were grown for 24 h at 30 • C and stained with 25 µL of 1% crystal violet for 15 min. The wells were washed twice with 200 µL of water, and crystal violet was extracted twice with 180 µL of 96% ethanol. The absorption (A 540 ) of three times diluted crystal violet extracts in water was measured by a microtiter plate reader Tecan Sunrise-basic [25]. Antibiotics were added to the growth medium only for the pre-growing of inoculum; a biofilm assay was performed without antibiotics. At least three independent measurements were performed.

Assessment of β-Galactosidase Activity
An indirect method, β-galactosidase activity, was used to assess lapA transcription in different media in the presence and absence of 1 mM IPTG for 18 h at 30 • C. LacZ activity was measured in P. putida PSm and F15 containing the p9_PlapA1-8 or p9TTBlacZ plasmids (Table 1); therefore, the growth media was supplemented with 1.5 mg benzylpenicillin mL −1 [20]. The β-galactosidase measurements in cell suspension were performed according to Sambrook et al. (1989) [44]. At least three independent measurements were performed.

Preparation of Cell
Lysates, Protein Quantification, SDS-Polyacrylamide Gel Electrophoresis, and Western Immunoblot Assay P. putida strains were grown in 5 mL of the respective medium without antibiotics for 18 h in the presence and absence of 1 mM IPTG [1]. For the preparation of crude cell lysate, the cells were collected by centrifugation and suspended in RIPA buffer (50 mM Tris/HCl pH 7.4, 1% Triton X-100, 0.1% SDS, 0.5% sodium deoxycholate, 1% NP-40, 500 mM NaCl [45]), 100-fold diluted Halt protease and phosphatase single-use inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA, USA) and 5 mM EDTA, and incubated for 4 h at 37 • C [1]. The cells' debris was separated by centrifugation at 12,000× g for 15 min at 4 • C [1]. The total amount of protein in the cleared supernatant was measured spectrophotometrically by the content of tryptophan [46]. Then, 40 µg and 10 µg of total protein were diluted 1:1 in 2× Laemmli sample buffer for protein quantification and Western immunoblot assay, accordingly, and incubated for 30 min at 37 • C [1]. For quantification, the proteins were separated on a 4-8% gradient PAA gel at 4 • C overnight by electrophoresis at 20 V using the following running buffer: 37.5 mM Tris, 95 mM glycine, and 0.05% SDS. After that, proteins in SDS-PAA gels were silver strained [47] and the intensity of LapA and LapF bands were quantified by ImageQuantTM TL 8.2 software.

Statistical Analysis of the Results
Factorial ANOVA was used to assess the variability of data in experiments. The multiple comparisons of means were conducted following Bonferroni's test for unequal n.
For statistical tests, the significance level was set at p < 0.05. Homogenous groups with the same letter indicate similarity of the arithmetic means (p ≥ 0.05). The calculations were performed using Statistica 13 software.