Studies performed on alginate in some plant-associated
Pseudomonas have revealed that this polysaccharide plays minor structural roles in their biofilms, including the bacterial phytopathogen
P. syringae and the plant-beneficial bacteria
P. fluorescens,
P. chlororaphis, and
P. putida [
59,
70,
73,
81,
82]. The alginate-deficient derivative of the
P. syringae pv. glycinea PG4180.muc strain formed biofilms to the same extent as the wild-type strain in flow-cell chambers [
81]. However, the biofilm architecture of the PssUMAF0158
∆alg8 strain, which does not produce alginate, showed slightly but significantly lower surface coverage and volume than the wild-type strain [
59], as was previously described in
P. aeruginosa [
80]. Alginate is overproduced in some strains of
P. syringae upon exposure to copper bactericides, which are usually applied to reduce the disease incidence caused by some plant pathogens [
32]. This could be explained because exopolysaccharide production has been generally associated with a higher tolerance against toxic compounds [
2,
83]. Previous works have indicated that alginate polysaccharides are involved in the pathogenic interaction of
P. syringae with plants [
29,
84,
85]. For instance, the alginate mutant of the
P. syringae pv. syringae 3525 strain, the causal agent of bacterial brown spot on bean, is significantly impaired in the colonization of bean (host) and tomato (non-host) leaves, and although it retains the ability to generate symptoms, the symptoms are less severe than those induced by the wild-type [
29]. However, these results have not been observed in other
P. syringae strains [
59,
86,
87,
88]. For example, in
P. syringae pv. glycinea PG4180 strain, the causal agent of bacterial blight of soybean, the expression of the AlgT regulator protein, but not alginate production
per se, promotes survival and symptom development in plants [
88]. Similarly, the PssUMAF0158
∆alg8 mutant strain is not altered in the induction of symptoms in tomato compared to the wild-type strain [
59].
The structural functions displayed by alginate in the biofilms of plant-pathogenic
Pseudomonas are in line with those observed in the plant-beneficial
Pseudomonas. The alginate mutant of
P. fluorescens SBW25 still forms biofilms in flow-cell chamber experiments, but they are thinner than those formed by the wild-type strain [
82]. This result is consistent with the flow-cell chamber phenotypes of the PssUMAF0158 and PAO1 alginate mutant derivative strains [
59,
80]. However, the alginate mutant of the biocontrol agent
P. chlororaphis PCL1606 (PcPCL1606 ∆
alg8) forms biofilms to the same extent as the wild-type in flow-cell chamber experiments and is not impaired in initial surface attachment, showing nonsignificant differences in surface coverage and volume values with respect to the wild-type [
70]. Alginate has been described as the primary polysaccharide that promotes hydration under desiccating stress in
P. putida [
89,
90]. In fact, alginate slightly contributes to the biofilm architecture of
P. putida under water-limiting conditions [
90]. The functions performed by alginate polysaccharide in both
P. fluorescens and
P. putida strains in vivo seem to be more relevant than those in vitro. For instance, the CHA211 and CHA213M mucoid variants of the
P. fluorescens CHA0 strain, which overproduce alginate, enhance their biofilm formation abilities on carrot roots compared to the wild-type strain [
91]. The genomic region located upstream of the
algD gene of
P. putida KT2440 is active during the colonization of maize root, which suggests that this polysaccharide could be a fitness determinant for the rhizosphere colonization ability of this bacterium [
92]. Overall, these studies indicate that alginate is not a critical component for biofilm formation in vitro in plant-associated
Pseudomonas and that its role seems to be more prominent in vivo, facilitating colonization and providing protection against stressors.
3.2. Cellulose Exopolysaccharide
Cellulose is a polymer made of D-glucose residues joined by β-1,4 glycosidic linkages and is considered a relevant biofilm matrix molecule in many environmental
Pseudomonas species [
93,
94]. Several biosynthesis and regulation mechanisms have been described for bacterial cellulose, but a common role of this component is to facilitate the establishment of efficient host-bacteria interactions [
95]. Previous studies reported that several plant-associated
Pseudomonas species can produce cellulose, including the plant-associated pathogenic bacteria
P. syringae,
P. asplenii,
P. marginalis,
P. corrugate, and
P. savastanoi and beneficial bacteria, such as
P. fluorescens and
P. putida [
72,
93,
94]. Within the
Pseudomonas genus,
P. fluorescens SBW25 (SBW25) is traditionally used as the model strain for the study of bacterial cellulose. In SBW25, cellulose polysaccharide is encoded on a ten-gene operon (
wssA-J) that corresponds to the PFLU0300-PFLU0309 genomic region [
96]. This exopolysaccharide is involved in the formation of floating biofilms, also called pellicles, in many strains of the species mentioned above, including the SBW25 [
31,
93,
96,
97,
98,
99]. The
P. aeruginosa species does not contain the cellulose operon [
72]. In particular,
P. aeruginosa PAO1 and PA14 strains, which have been traditionally used as model strains for conducting biofilm studies within the
Pseudomonas genus, were tested for cellulose production, but in line with in silico observations [
72], they were not found to produce this exopolysaccharide [
94]. However, the PAO1 and PA14 strains contain a seven-gene operon that encodes Pel, which is a polymer composed of partially acetylated 1→4 glycosidic linkages of N-acetylgalactosamine and N-acetylglucosamine [
100]. The
pel operon is poorly conserved among environmental
Pseudomonas [
72,
100,
101]. Interestingly, Pel promotes the formation of pellicle biofilms, as has also been described for cellulose [
102].
Among all plant-pathogenic
Pseudomonas that have been reported to produce cellulose, studies regarding its structural roles within biofilms and biological significance have essentially been conducted on
P. syringae. The
P. syringae pv. syringae (Pss) UMAF0158 (PssUMAF0158) strain, the causal agent of bacterial apical necrosis (BAN) on mango trees, and
P. syringae pv. tomato DC3000 (PtoDC3000), responsible for bacterial speck disease on tomato plants, produce cellulose as the main exopolysaccharide of their biofilms [
31,
59,
103]. The biofilm structures formed by PssUMAF0158 and PtoDC3000 in micro-well plates are highly similar, consisting of pellicles with wrinkles on the surface that are weakly attached to the walls of the culture vessels [
31,
103]. Despite the structural similarities found in vitro, the biological performance of cellulose seems to differ in both strains. Cellulose allows PssUMAF0158 to adhere to mango leaves, and its production intimately affects the epiphytic and pathogenic stages of this strain over the plant surface [
31]. Hence, the incidence and severity of necrotic symptoms developed by PssUMAF0158 on tomato leaflets are lower in the wild-type than in cellulose mutants (∆
wssB and ∆
wssE mutants) and practically nonexistent in the cellulose-overproducing strain [
31]. These results, together with the fact that the highest BAN symptoms coincide with cool and wet periods [
104], support the proposed lifecycle of Pss strains over the mango tree, in which biofilm formation would be mainly needed during the epiphytic phase (spring/summer) when the bacteria are more exposed to the external environment, and protection against its challenging conditions becomes crucial for survival [
50]. Interestingly, the link observed between cellulose production and PssUMAF0158 transition through epiphytic and pathogenic stages over the mango plant surface has not been reported in PtoDC3000. The disease symptoms developed in tomato by the PtoDC3000 wild-type strain were not different from those of its
∆wssBC-derived mutant [
105]. Furthermore, in disagreement with what has been observed in PssUMAF0158, cellulose overproduction in PtoDC3000 does not lead to a significant impact on virulence [
103]. However, the PtoDC3000
armZ gene mutant, which does not produce alginate and does overproduce cellulose, has a reduced virulence compared to the wild-type strain [
105]. Although PssUMAF0158 and PtoDC3000 are categorized as
P. syringae species and belong to the
P. syringae complex, this complex is comprised of a hodgepodge that, in effect, includes many other taxonomically related species [
49]. A previous study revealed that the phylogenetic relationship between
P. syringae pv. syringae B728a strain, closely related to PssUMAF0158, and PtoDC3000, is not very proximate. In fact, PtoDC3000, together with other strains of the tomato pathovar, seems to form a new species
Pseudomonas tomato, pending a deeper taxonomic analysis [
49]. This evidence, together with the fact that the infection assays were performed using different tomato cultivars and inoculation approaches, could all eventually account for different results.
Regarding beneficial plant-interacting
Pseudomonas, studies on bacterial cellulose have been mainly conducted on
P. fluorescens and
P. putida species. Biofilm experiments on SBW25 determined that the gradients occurring within a static microcosm immediately select for the emergence of variants that occupy different niches [
106]. Among those variants, the air-liquid (A–L) interface is colonized by wrinkly spreader (WS) pellicles, an SBW25-derived mutant that overproduces cellulose compared to the wild-type equivalent [
107]. In
P. putida mt2 and its plasmid-free derivative KT2440 [
108] strains, cellulose plays minor roles in biofilm formation in vitro [
73,
89], while two additional exopolysaccharide gene clusters, putida exopolysaccharide A (Pea) and putida exopolysaccharide B (Peb), are essential for biofilm formation in this species [
73,
89]. Instead, the role of cellulose exopolysaccharide in
P. putida seems to be directed more towards conferring protection, as water-limiting conditions and increasing osmolarity highly induce cellulose expression of
P. putida mt2 [
89,
109]. In addition, the cellulose mutant of
P. putida mt2 strain accumulates significantly more reactive oxygen species (ROS) than the wild-type strain upon exposure to matric and copper stressors [
109]. During plant–bacteria interactions, the cellulose exopolysaccharide of SBW25 contributes to the ecological performance of this strain in the rhizosphere and phyllosphere of sugar beet [
61]. Thus, a cellulose-defective mutant of SBW25 (SM13) was compared against the wild-type in the rhizosphere, phyllosphere, and bulk soil surrounding the rhizosphere of sugar beet seedlings, and the results showed no significant differences between the fitness of SM13 relative to the wild-type in bulk soil, but significant differences were found in the rhizosphere and phyllosphere, especially in the phyllosphere [
61]. Something similar has been reported in the
P. putida mt2 strain in which the cellulose mutant is impaired in the colonization of the maize rhizosphere during competition with the wild-type equivalent [
73]. These studies indicate that, while the cellulose operon does not seem to be critical for biofilm formation under laboratory conditions in
P. fluorescens and
P. putida, their roles in these species seem to be more pronounced in vivo.
3.3. Psl Exopolysaccharide
The Psl polysaccharide was first described in
P. aeruginosa [
102,
110,
111], and its structural analysis determined that it consists of a repeating pentasaccharide subunit of D-mannose, D-glucose, and L-rhamnose in a 3:1:1 ratio [
112]. In PAO1, Psl was formerly described to be encoded by the 15-gene operon
psl (
pslA-O), which corresponds to the PA2231-PA2245 genomic region [
102,
110,
111]. However, later works revealed that the last three genes of the operon (
pslMNO) constitute an independent transcriptional unit [
113,
114,
115] and are not truly required to produce Psl [
112]. Except for the case of the
P. aeruginosa PA14 strain, which does not produce Psl due to the absence of
pslA-D genes [
102], the
psl gene cluster is present in multiple strains of
P. aeruginosa [
72,
80], where it plays key roles in their biofilm lifestyles [
80]. Several studies have proven the involvement of Psl in adhesion to biotic and abiotic surfaces, biofilm architecture, motility, and protection against stressors [
16,
19,
116,
117,
118,
119]. Although research on Psl polysaccharides has been mostly conducted in
P. aeruginosa, the existence of a
psl-like gene cluster has been reported in some environmental nonaeruginosa
Pseudomonas [
59,
70,
72,
101]. Generally, the
psl-like gene clusters found in nonaeruginosa
Pseudomonas either lack orthologues to
pslMNO genes or are found scattered in the genome outside the cluster. The bacterial phytopathogen PssUMAF0158 contains a
psl-like gene cluster that does not include orthologues to the
pslCLMNO genes and encodes a putative acetyltransferase between the
pslJ-and
pslK-like genes that might perform a related function to that of acyltransferase PslL [
59]. Interestingly, the
psl-like gene cluster of PssUMAF0158 seems to be highly conserved among the plant-associated phylogroups belonging to the
P. syringae complex [
59]. The biocontrol agent PcPCL1606 also contains a
psl-like gene cluster, which lacks the
pslLMNO genes and encodes a putative acetyltransferase between the
pslJ- and
pslK-like genes, similar to that of PssUMAF0158 [
70]. However, the
psl-like gene cluster of PcPCL1606 is not present in some phylogroups of the
P. fluorescens complex and is partially present in others, according to the strains included in a previous study [
70]. It is completely absent in the
corrugata,
jessenii, and
koreensis phylogroups; only present in
Pseudomonas GM21 strain of the
mandelii phylogroup; and is partially present within the
P. fluorescens phylogroup. Interestingly, a
psl-like gene cluster is found in all the strains of the
P. chlororaphis phylogroup that have been assessed [
70], which suggests that this polysaccharide could be relevant for biofilm formation in this species.
The first study regarding Psl composition in
P. aeruginosa PAO1 determined that this polysaccharide was a galactose- and mannose-rich exopolysaccharide [
120]. Support for this information came from three pieces of evidence. First, a chemical composition analysis of exopolysaccharide preparations of WFPA801, a PAO1-derived Psl-inducible strain, determined the presence of galactose, mannose, and glucose, as well as trace amounts of xylose, rhamnose, and N-acetylglucosamine. Second, staining of planktonically grown WFPA801 cells with FITC-HHA lectin, which binds to some mannosyl units, and FITC-MOA lectin, which binds to some galactosyl units, revealed green fluorescent signals on the WFPA801 surface. Ultimately, mutants of the
pslH gene, which encodes a putative galactosyltransferase, and the
pslI gene, which encodes a putative mannosyltransferase, were deficient in attachment, yielding a similar phenotype to that of the WFPA800 null Psl-producing strain [
120]. Two years later, the structural analysis of Psl was published, indicating that it likely consisted of a pentasaccharide repeating unit of mannose, glucose, and rhamnose in approximate ratios of 3:1:1 [
112]. Interestingly, galactose, which was reported as the major component of Psl in the first study [
120], was not detected as a component of Psl in the structural analysis [
112]. The authors stated that different growth conditions were used in both studies, which could account for some variations in composition, as described previously [
121]. Therefore, there is some thought that different forms of Psl might be produced even in the same strain depending on the growth conditions. Be that as it may, mannose seems to be a key component of the Psl structure in
P. aeruginosa. An analysis of the composition and structure of the putative Psl polysaccharide produced by PssUMAF0158 and PcPCL1606 has not yet been conducted, but some hints exist regarding the existence of a polysaccharide that resembles Psl in
P. syringae and
P. fluorescens. Thus, it was reported that, in addition to alginate and levan,
P. syringae PG4180 produced a third exopolysaccharide (EPS) that consisted of a fibrous structure in its biofilms and bound to
Naja mossambica lectin (NML) [
81]. Interestingly, the monosaccharide specificity of NML is mannose [
122]. Furthermore, two
P. fluorescens strains isolated from rotted bell pepper, PF-05-2 and PM-LB-1, produced a novel exopolysaccharide composed of mannose, rhamnose, and glucose (1:1:1 molar ratio) substituted with pyruvate and acetate [
123]. The biofilm formed by the PcPCL1606 wild-type strain but not its Psl-like-derived mutant, contains a polymer that binds to banana lectin, which also binds to mannose residues [
70,
124].
The ∆
pslAB mutant of PAO1 is severely attenuated in biofilm initiation and biofilm development in flow-cell chamber experiments [
110,
111]. Interestingly, similar results were observed in the biocontrol strain PcPCL1606, in which the ∆
pslE mutant was severely affected in early surface attachment and development of a mature biofilm architecture compared to the wild-type strain [
70]. The biofilms developed by PAO1 and PcPCL1606 wild-type strains showed an intricate architecture in flow-cell chambers, which consisted of a multilayer of cells that covered the chamber surfaces [
70,
110,
111]. However, the biofilm phenotype of the PAO1 ∆
pslAB and PcPCL1606 ∆
pslE mutant strains consisted of a monolayer of loosely aggregated cells, which suggests that this exopolysaccharide could also be important for cell-to-cell interactions [
70,
110,
111]. Similarly, the Psl-like polysaccharide of the phytopathogen PssUMAF0158 is also involved in biofilm architecture [
59]. Compared to the more developed biofilm of wild-type PssUMAF0158, the PssUMAF0158 ∆
pslE biofilms consisted of scattered cell aggregates across the flow-cell chamber surface [
59]. These cell aggregates were disrupted in the double mutant ∆
wssE,pslE strain, which did not produce both cellulose and Psl-like polysaccharides [
59]. Curiously, this phenotype was also observed in some
P. aeruginosa strains, where the cell aggregates formed by their derived ∆
psl mutants were disrupted in the ∆
psl∆
pel double mutants, affected in both Psl and Pel polysaccharide production [
116]. The fact that cellulose and Pel polysaccharides are both involved in the formation of pellicle biofilms [
101], and that the cell aggregates formed by these
Pseudomonas ∆
psl strains are disrupted when either cellulose or Pel is not produced, indicates that both polysaccharides could play redundant structural roles within biofilms, as has been previously suggested [
59,
101]. Indeed, it is not common to find both genomic regions encoding cellulose and Pel in the same
Pseudomonas strain. Thus, just 12 out of 600
Pseudomonas genomes (2%) that have been analyzed in a recent study [
72]—which belong to four different groups:
P. asplenii,
P. fluorescens,
P. fragi and
P. oryzihabitans—possess both clusters (), although whether they are functional remains unknown. With these recent data, the identity and coverage of both clusters have been analyzed in these 12
Pseudomonas spp. strains using the
wss operon of SBW25 and
pel operon of PAO1 as references.
Table 2. Pseudomonas spp. strains obtained from Blanco–Romero et al., (2020) that have been reported to contain the wss and pel clusters.
Strain |
wss Cluster 1 |
pel Cluster 1 |
|
Identity (%) |
Coverage (%) |
Identity (%) |
Coverage (%) |
P. agarici NCPPB 2472 |
71.03 |
82 |
69.68 |
93 |
P. azotoformans F77 |
82.58 |
100 |
70.63 |
89 |
P. azotoformans LMG_21611 |
83.72 |
99 |
70.50 |
91 |
P. extremorientalis LMG 19695 |
89.82 |
99 |
70.59 |
92 |
P. lundensis AU1044 |
71.31 |
112 |
72.32 |
91 |
P. lurida L228 |
83.17 |
100 |
71.37 |
85 |
P. lurida MYb11 |
82.99 |
100 |
71.48 |
85 |
P. oryzihabitans USDA-ARS-USMARC-56511 |
68.70 |
53 |
71.33 |
98 |
P. oryzihabitans FDAARGOS_657 |
70.45 |
57 |
71.50 |
98 |
P. psychrotolerans PRS08-11306 |
70.40 |
58 |
71.66 |
98 |
P. psychrotolerans CS51 |
70.11 |
50 |
74.27 |
94 |
P. trivialis IHBB745 |
91.38 |
99 |
73.28 |
91 |