Sensing preferences for prokaryotic solute binding protein families

Abstract

Solute binding proteins (SBPs) are of central physiological relevance for prokaryotes. These proteins present substrates to transporters, but they also stimulate different signal transduction receptors. SBPs form a superfamily of at least 33 protein Pfam families. To assess possible links between SBP sequence and the ligand recognized, we have inspected manually all SBP three‐dimensional structures deposited in the protein data bank and retrieved 748 prokaryotic structures that have been solved in complex with bound ligand. These structures were classified into 26 SBP Pfam families. The analysis of the ligands recognized revealed that most families possess a preference for a compound class. There were three families each that bind preferentially saccharides and amino acids. In addition, we identified families that bind preferentially purines, quaternary amines, iron and iron‐chelating compounds, oxoanions, bivalent metal ions or phosphates. Phylogenetic analyses suggest convergent evolutionary events that lead to families that bind the same ligand. The functional link between chemotaxis and compound uptake is reflected in similarities in the ligands recognized by SBPs and chemoreceptors. Associating Pfam families with ligand profiles will be of help to design experimental strategies aimed at the identification of ligands for uncharacterized SBPs.

Solute binding proteins (SBPs) are of central physiological relevance for bacteria and form a super‐family composed of at least 33 Pfam families. Based on the analysis of the totality of currently available 3D holo‐SBP structures, we show that most families preferentially bind members of a ligand family. This information will be helpful to identify the ligands recognized by uncharacterised SBPs.

INTRODUCTION

Solute binding proteins (SBPs) form a protein superfamily that is composed of at least 33 individual protein families in the Pfam database (Mistry et al., 2021). SBPs are found in all kingdoms of life (Scheepers et al., 2016). In Gram‐positive bacteria and archaea, they are tethered to the external face of the membrane, whereas in Gram‐negative bacteria they are present as diffusible proteins in the periplasm (Chu & Vogel, 2011). SBPs differ largely in size (20–65 kDa), but share the same topology consisting of two lobes connected by a hinge (Fukamizo et al., 2019). SBPs bind different ligands at a conserved binding site that is formed by both lobes of the protein.

The type of ligand recognized by SBPs is almost unlimited, including amino‐ and organic acids, peptides, sugars, polyamines, quaternary amines, purines, metal ions, oxoanions, vitamins, quorum sensing signals and different siderophores (Matilla et al., 2021). Bacteria typically contain many SBP genes. We have recently published the SBP repertoire of 49 bacterial model strains that possess on average 60 ± 49 SBP genes, which frequently represent several percent of their genomes (Ortega et al., 2022). In addition, a proteomics study has shown that the cellular level of many SBPs is very high (Matilla et al., 2023).

This abundance of SBPs is a reflection of the multiple functions of this protein superfamily. First, the main function of SBPs is to present different solutes to transmembrane transporters to initiate compound uptake (Gautom et al., 2021; Luo et al., 2021; Sakanaka et al., 2019). Second, solute‐loaded SBPs also bind to different transmembrane receptors such as chemoreceptors, sensor histidine kinases, protein kinases or diguanylate cyclases/phosphodiesterases to stimulate signal transduction cascades (Bhattacharyya et al., 2018; Carey et al., 2018; Lopes & Sourjik, 2018), permitting the coordination of transport with signal transduction. SBP‐mediated receptor stimulation appears to be a universal and widespread mechanism, and the elucidation of the role of SBPs in signal transduction is currently a major research effort (Matilla et al., 2021). Third, there is strong evidence that SBPs carry out additional functions that are currently poorly understood. One such example is the phosphate‐specific PstS SBP of Pseudomonas aeruginosa, which apart from presenting inorganic phosphate to the PstABC transporter (Peng et al., 2017) and stimulating the phosphate‐specific chemoreceptor CtpL (Rico‐Jimenez et al., 2016), also form an appendage‐like structure on the cell surface that contributes to multi‐drug resistance (Zaborina et al., 2008) and biofilm formation (Neznansky et al., 2014). SBPs are also of significant biotechnological relevance. They can serve as precursors for enzyme evolution (Clifton et al., 2018; Noda‐Garcia & Tawfik, 2020) and are used for biosensor construction (Smith et al., 2022; Tavares & van der Meer, 2021; Vongsouthi et al., 2021).

In a previous study, we have reported the ligands that are recognized by sensor domains of chemoreceptors, sensor kinases and transcriptional regulators belonging to 127 Pfam families, information that has been included into the Pfam database (Matilla, Velando, Martín‐Mora, et al., 2022). This study has been very well received by the scientific community and the primary objective of the present study is to define the ligand profiles of the individual SBP Pfam families.

Previous studies have shown that structural alignments of all available 3D SBP structures resulted in the identification of clusters that showed a preference for a ligand or family of ligands (Berntsson et al., 2010; Scheepers et al., 2016). Here, we have assessed whether the ligand specificity is reflected in sequence similarity, as defined by the classification of SBPs into different Pfam families (Mistry et al., 2021). In fact, the inspection of the information currently associated with the 33 SBP families in Pfam, provides either no or little information on the ligands recognized by family members. In the present study, we have manually inspected all of SBP three‐dimensional (3D) structures deposited in the protein data bank (pdb) (Burley et al., 2019), retrieved the ligands bound to all prokaryotic proteins and classified proteins according to their Pfam family. We show here that many SBP families possess a clear preference for a ligand family. This work will help to annotate the function of other members of this important protein superfamily.

EXPERIMENTAL PROCEDURES

Identification and classification of ligands bound to SBP 3D structures

The pdb ID codes of 3D structures for each of the SBP families deposited in the pdb (Burley et al., 2019) were retrieved from the Pfam database (Mistry et al., 2021). Each of the structures was inspected, and the ligand bound to the ligand‐binding site between the two lobes of the structure was identified. The publications (if available) associated with each structure in pdb were consulted to confirm that a given ligand is of functional relevance.

Ligands have then been classified into groups and subgroups (Table S1). To calculate the relative abundance of ligands for a given protein family, the totality of ligands observed in a given family was set as 100%. If a given ligand was bound to multiple structures, each of the structures was included for the total ligand count. The relative abundance of ligands as shown in Figures 1, 2, 3, 4 is the ratio of the number of individual ligands or members of ligand families relative to the total number of ligands per family.

FIGURE 1.

SBP families that preferentially recognize saccharides. The ligands that bind to the 3D structures of family members deposited in the protein data bank are shown. Data for the SBP_bac_1, Peripla_BP_4 and SBP_bac_8 families are based on 37, 44 and 30 different proteins respectively (Table S1).

FIGURE 2.

SBP families that preferentially bind different amines. The ligands that bind to the 3D structures of family members deposited in the protein data bank are shown. Data on the SBP_bac_3, Peripla_BP_6, Lipoprotein_9, Bmp and OpuAC families are based on 40, 21, 19, 7 and 14 different proteins respectively (Table S1).

FIGURE 3.

SBP families that preferentially recognize iron and iron‐containing compounds, oxoanions, bivalent metal anions and phosphates. The ligands that bind to the 3D structures of family members deposited in the protein data bank are shown. Data on the Peripla_BP_2, SBP_bac_6, SBP_bac_11, ZnuA, PBP_like_2 and Phosphonate_bd are based on 25, 6, 12, 20, 13 and 7 different proteins respectively (Table S1).

FIGURE 4.

The SBP families without clear ligand preferences. The ligands that bind to the 3D structures of family members deposited in the protein data bank are shown. Data on the DctP and SBP_bac_5 families are based on 58 and 28 proteins respectively (Table S1).

Construction of the phylogenetic tree

SBP sequences were extracted from UniprotKB (UniProt Consortium, 2023) with a custom Python script implementing the Bioservices (Cokelaer et al., 2013) and Biopython (Cock et al., 2009) packages. For each protein, a query from the main pdb code was made to Uniprot. FASTA sequence files, the Uniprot entry ID, gene locus and PFAM family were extracted for each protein. Sequences were then aligned using the MAFFT NS‐I algorithm (Katoh et al., 2019) and subsequently used for the construction of the phylogenetic tree using Fast Tree (JTT + CAT) with a Shimodaira‐Hasegawa test for phylogeny (Price et al., 2010). Both tools were obtained from the TREND pipeline (Gumerov & Zhulin, 2020). The phylogenetic tree was represented by iTOL (Letunic & Bork, 2021).

RESULTS AND DISCUSSION

The superfamily of SBPs is comprised of at least 33 individual Pfam protein families (Table S2). The individual pdb entries were inspected manually to extract structures of prokaryotic stand‐alone SBPs that have been solved in complex with bound ligand (holo‐structures). Protein families from eukaryotes (e.g. ANF_receptor, ABC_transp_aux) and families in which SBPs form sensor domains of receptors (e.g. Peripla_BP_3) were excluded (Table S2). For another four families, no holo‐structures were available (Table S2). In total, 748 prokaryotic holo‐SBP structures belonging to 26 SBP families were retrieved (Tables S1 and S2). A number of structures were of the same protein crystallized with different ligands (Table S1). In total, ligand‐bound 3D structures were obtained for 403 different proteins that belonged to 175 different organisms. Protein ligands were extracted manually (Table S1) and SBPs classified according to their protein family. The ligand range of populated SBP families is depicted in Figures 1, 2, 3, 4, whereas ligands that bind to less‐populated families are listed in Table 1.

TABLE 1.

List of less‐populated SBP families.

Pfam protein family		No. of proteins	Ligands bound
Code	Family name
PF00532	Peripla_BP_1	4	Monosaccharides, quaternary amines
PF03401	TctC	5	Malate, citrate, L‐Asp, L‐Glu, 2‐oxoadipic acid, hexanedionic acid
PF04392	ABC_sub_bind	3	L‐Trp, L‐Tyr, L‐Phe
PF05048	NosD	2	Difructose, L‐kestose
PF06646	CypI	2	Vitamin B1
PF09084	NMT1	6	Thiamine, aminopyrimidine
PF12727	PBP_like	2	Phosphate
PF13433	Peripla_BP_5	1	Amide
PF16868	NMT1_3	1	Glutamic acid
PF18610	Peripla_BP_7	1	D‐Glucose

Note: The ligands bound to the corresponding 3D structures are listed.

A significant number of populated SBP families showed a clear preference for a compound family. These families bind preferentially saccharides (Figure 1), amino acids, purines, quaternary amines (Figure 2), iron and iron‐chelating compounds, oxoanions, metal ions or phosphate (Figure 3). In contrast, the well‐populated families, DctP and SBP_bac_5, did not show a clear ligand preference (Figure 4).

Three SBP families have a ligand preference for mono‐, di‐ and oligosaccharides

Saccharides and amino acids appear to be the primary ligand families for SBPs (Table 2). Members of 10 different SBP families were found to bind saccharides (Table 2). There were three SBP families, SBP_bac_1, SBP_bac_8 and Peripla_BP_4, that showed a clear ligand preference for saccharides (Figure 1, Table S1). With 37, 30 and 44 proteins, respectively, these three families were well‐populated. The families SBP_bac_1 and SBP_bac_8 showed a wide ligand spectrum as they bind mono‐, di‐ and oligosaccharides, whereas family Perpla_BP_4 binds exclusively monosaccharides and the quorum sensing signal autoinducer‐2.

TABLE 2.

Link between SBP‐mediated compound uptake and chemotaxis.

SBP		Chemoreceptor (selection)
Ligand family	Protein families	Name	Species	Ligand recognized	Reference
Mono‐, di‐ and oligosaccharides	SBP_bac_1, SBP_bac_5, SBP_bac_6, SBP_bac_8, SBP_bac_11, Peripla_BP_1, Peripla_BP_4, Peripla_BP_7, DctP, NosD	Tlp2‐4	Campylobacter jejuni	Mannose, fucose, galactose	Taha et al. (2022)
		Tlp11	C. jejuni	Galactose	Feng et al. (2022)
		McpA	Bacillus velezensis	Ribose, ribitol, mannose, fucose	Day et al. (2016)
		Tar	Escherichia. coli	Maltose (via MBP SBP)	Zhang et al. (1999)
Monocarboxylic acids	DctP, Peripla_BP_6	McpV	Sinorhizobium meliloti	Propionate, acetate, glycolate	Compton et al. (2018)
Dicarboxylic acids	DctP, TctC	McpS	Pseudomonas putida	Malate, fumarate, oxaloacetate a	Lacal et al. (2010)
Amino acids (including non‐proteinogenic amino acids)	SBP_bac_3, SBP_bac_8, Peripla_BP_6, Lipoprotein_9, TctC, ABC_sub_bind, NMT1_3, DctP, Peripla_BP_2, SBP_bac_11, OpuAC	PctA/PctB	Pseudomonas aeruginosa	Multiple amino acids	Gavira et al. (2020)
		McpC	Bacillus subtilis	Multiple amino acids	Glekas et al. (2012)
		McpU	S. meliloti	Multiple amino acids	Webb, Compton, Del Campo, et al. (2017)
Phosphate and derivatives	PBP_like_2, Phosphonate_bd, PBP_like	CtpH	P. aeruginosa	Phosphate	Rico‐Jimenez et al. (2016)
		CtpL	P. aeruginosa	Phosphate (via PstS SBP)	Rico‐Jimenez et al. (2016)
Purines/pyrimidines	Bmp, NMT1	McpH	P. putida	Uric acid, purine, adenine a	Fernández et al. (2016)
		Tap	E. coli	Thymine, uracil (probably via SBP)	Liu and Parales (2008)
Oxoanions	SBP_bac_1, SBP_bac_11, Peripla_BP_2, PBP_like_2, Phosphonate_bd	AFM38850	Agrobacterium tumefaciens	Arsenite	Shi et al. (2017)
		SO_1056	Shewanella oneidensis	Chromate	Boyeldieu et al. (2021)
Quaternary amines	OpuAC, Peripla_BP_1, SBP_bac_1	McpX	S. meliloti	Choline, glycine betaine a	Webb, Compton, Castañeda Saldaña, et al. (2017)
		PctD	P. aeruginosa	Choline, acetylcholine, betaine	Matilla, Velando, Tajuelo, et al. (2022)
Polyamines	SBP_bac_8	TlpQ	P. aeruginosa	Spermidine, putrescine a	Corral‐Lugo et al. (2018)
Metal ions and metal chelating compounds	Peripla_BP_2, SBP_bac_1, SBP_bac_5, SBP_bac_6, SBP_bac_8, SBP_bac_11, ZnuA, DctP	SO_1056	S. oneidensis	Ni2+ (repellent)	Boyeldieu et al. (2022)
Peptides	SBP_bac_5, SBP_bac_3, DctP	Tap	E. coli	Various dipeptides (via DppA SBP)	Manson et al. (1986)
Autoinducer‐2	Peripla_BP_4	PctA/TlpQ	P. aeruginosa	Autoinducer‐2	Zhang et al. (2020)
		Tsr	E. coli	Autoinducer‐2 (via LsrB SBP)	Hegde et al. (2011)

Note: The major ligand families of SBPs and the corresponding SBP protein families are shown on the left. Representative examples for chemoreceptors that respond to SBP ligand families are listed on the right.

^a A selection is given, compounds with highest affinity.

With 44 different proteins, Peripla_BP_4 is one of the most populated families. Although it is specific for monosaccharides, family members bind 27 different sugars. Thus, a diversity of monosaccharides are recognized. The most abundant monosaccharide recognized by this family was glucose, with seven SBPs (Table S1). Around 15% of the Peripla_BP_4 structures had autoinducer‐2 bound (Figure 1 and Table S1). Autoinducer‐2 shows structural similarities with furanose, which may explain its recognition by this SBP family. Autoinducer‐2 is a central quorum sensing signal (Pereira et al., 2013), and SBPs that recognize this compound may interact with signal transduction receptors.

Three SBP families bind preferentially amino acids

Proteins belonging to 11 SBP families were found to bind amino acids (Table 2 and Table S1). Among them, families SBP_bac_3 and Peripla_BP_6 bind a range of different amino acids, whereas Lipoprotein_9 binds exclusively L‐Met and its close derivative selenomethionine (Figure 2 and Table S1). Clear amino acid preferences emerged for the SBP_bac_3 and Peripla_BP_6 families. Whereas SBP_bac_3 proteins bind preferentially amino acids with positive change (L‐Arg, L‐Lys and L‐His), Peripla_BP_6 binds preferentially the three amino acids with a branched side chain (L‐Leu, L‐Val and L‐Ile). Lipoprotein_9 was one of the families with the narrowest ligand spectrum (Figure 2 and Table S1). All 19 proteins of this family recognize exclusively L‐Met or selenomethionine.

The bacterial hosts for these proteins show a wide phylogenetic distribution, including different Gram‐positive and Gram‐negative strains (Table S1). Selenomethionine is a naturally occurring amino acid, and several bacteria metabolize this compound (Stolz et al., 2006). Selenomethionine mimics closely the properties of L‐Met, and selenomethionine‐containing proteins are frequently used to resolve the crystallographic phase problem using multi‐wavelength anomalous diffraction techniques (Wenzel et al., 2019).

Almost all bacterial species possess biosynthetic pathways for methionine (Ferla & Patrick, 2014). L‐Met is the only amino acid that is specifically recognized by a SBP protein family, suggesting a particular physiological relevance that may be related to the fact that this compound is also required for the synthesis of S‐adenosylmethionine, the main cellular carrier of methyl groups (Ferla & Patrick, 2014). L‐Met appears to be sensed almost exclusively by this family since there was only one other family, Peripla_BP_6, that recognizes L‐Met (Table S1). Based on the evidence that this is a methionine‐specific family, we propose to rename the Lipoprotein_9 family SBP_Met.

Many bacteria possess amino acid racemases that interconvert the L‐ and D‐ forms of amino acids (Miyamoto & Homma, 2021). Of the 61 amino acid‐responsive SBPs, there were only three proteins that bind a D‐isomer [pdb ID 4F3S & 4JB7, unpublished; pdb ID 4NAP, (Vetting et al., 2015)]. This preference for L‐amino acids agrees with data on amino acid‐responsive chemoreceptors that also recognize primarily L‐isomers (Matilla, Velando, Martín‐Mora, et al., 2022).

Families bmp and OpuAC bind preferentially purine/pyrimidine containing compounds or quaternary amines

Although the Bmp family comprises only seven members, there was a clear ligand preference for purines (adenine, guanine) and pyrimidines (uridine, thymidine) (Figure 2 and Table S1). In addition, there were several purine nucleosides (adenosine, guanosine and inosine) among the ligands (Table S1). Riboflavin (vitamin B2) was the only ligand that did not match this pattern. However, part of the riboflavin structure shares strong similarities with the purine skeleton. Two additional families, Peripla_BP_2 and NMT1, were also found to bind riboflavin and thymidine, respectively (Table S1).

Quaternary amines contain a positively charged nitrogen atom linked to four carbon atoms. Many physiologically relevant quaternary amines possess a trimethylated amino group. Family OpuAC showed a clear preference for quaternary amines (Figure 2). Most abundant are family members that recognize glycine betaine, followed by choline, acetylcholine, proline and carnitine (Table S1). These compounds are important osmo‐ and thermo‐protectants (Caldas et al., 1999; Chen & Beattie, 2008; Meadows & Wargo, 2015) or intermediates in their synthesis (Yang et al., 2022). Glycine betaine can reach a cellular concentration of up to 1 M in cells subjected to osmotic stress, and it has been shown to stabilize proteins (Singh et al., 2009). A single protein of the SBP_bac_1 family [pdb ID 6R1B, (Fenn et al., 2019)] binds quaternary amines (Table S1).

Two SBP families for iron acquisition

Iron is a critical nutrient for the growth and survival of most bacterial species (Frawley & Fang, 2014). There are two well‐populated SBP families for the acquisition of iron, Peripla_BP_2 and SBP_bac_6. Whereas most Peripla_BP_2 members bind a range of different iron‐chelating compounds, SBP_bac_6 members bind iron ions directly (Figure 3 and Table S1). The most abundant ligand (seven proteins) of the Peripla_BP_2 family is protoporphyrin IX, which is ubiquitously present in small amounts in all living cells as a precursor for haeme synthesis (Sachar et al., 2016). The other large group of ligands for this family are siderophores, which are iron‐scavenging molecules produced by many bacteria (Kramer et al., 2020). Examples of such compounds recognized by Peripla_BP_2 members are ferrichrome, gallichrome, staphyloferrin B, ferrioxamine B and E, staphyloferrin A and ferri‐Bacillibactin (Table S1).

Families SBP_bac_11 and ZnuA bind preferentially metal oxoanions or metal ions

The primary ligands for family SBP_bac_11 are different metal oxoanions like molybdate, tungstate, chromate and perrhenate (Figure 3). About half of the family members recognize molybdate and/or tungstate. Despite the close chemical and physical similarities of both of these oxoanions, there are SBPs that were able to recognize specifically one or the other, whereas other SBPs recognize both of them (Aguilar‐Barajas et al., 2011; Andreesen & Makdessi, 2008). Several family members bind the structurally related sulphate and phosphate ions. Furthermore, a number of phosphorylated compounds like 3‐phosphoglycerate, phosphoenolpyruvate and glucose‐6‐phosphate were identified as ligands (Table S1), suggesting that the recognition of the phosphate moiety is essential for binding. SBP_bac_1, DctP, PBP_like_2 and Phosphonate_bd are additional families that recognize oxoanions like tungstate, molybdate and arsenate (Figure 3 and Table S1).

Twenty different proteins of the ZnuA family were retrieved (Table S1). The ligand profile of this family is very well defined since all proteins recognize metal ions. Almost half of the family members bind Zn²⁺ ions. Whereas Zn²⁺, Mn²⁺ and Fe²⁺ had a similar molecular weight, Cd²⁺ ions are significantly heavier. Many ZnuA family members bind specifically a single metal ion. Some proteins of the families SBP_bac_5, Peripla_BP‐2, SBP‐bac_1, DctP, SBP_bac_6, SBP_bac_8 and SBP_bac_11 also bind different metal ions, such as Ni²⁺, Co²⁺, Zn²⁺, Fe³⁺ or Zr⁴⁺ (Table S1).

Two SBP families recognize phosphorous containing compounds

Phosphorous is an essential macroelement and is required for many different biological processes. Almost all PBP_like_2 family members recognize inorganic phosphate, phosphate, whereas two of them also bind arsenate and tungstate (Figure 3). As indicated above, both of these oxoanions are structurally similar to phosphate and the toxicity of arsenate is due to its binding to phosphate‐binding proteins (Dani, 2011). According to the Pfam database, the Phosphonate_bd family binds alkylphosphonates. Our analysis shows that the ligand range of this family is much larger and includes different phosphates, phosphites and pyrophosphates (Table S1, Figure 3). There were only two proteins, PhnD1 and PhnD, that recognize the alkylphosphonates ethylphosphonate and 2‐aminoethylphosphonate, respectively. One protein in the low‐populated family PBP_like also binds phosphate (Table S1).

DctP and SBP_bac_5 lack a clear ligand preference

Not all SBP families possess a clear ligand preference. This is exemplified by the DctP family, which with 58 members, is the most populated protein family. As shown in Figure 4, there was no clear ligand preference. In addition to different sugars and carboxylic acids, the DcpT family binds ligands as diverse as zinc and calcium ions, dipeptides, ectoine, the auxin indole‐3‐acetic acid, glycerol‐3‐phosphate and N‐acetyl‐neuraminic acid (Table S1). In total, there were three families that bind peptides (Table 2). However, the primary family for peptide recognition was SBP_bac_5, for which peptides represent almost half of the ligands recognized (Figure 4). SBP_bac_5 family members are rather permissive for peptide length, with ligands ranging from dipeptides to nonapeptides (Table S1). In addition, members of this family bind a number of structurally very diverse ligands including Ni²⁺, glutathione, different sugars, haemin or EDTA (Figure 4 and Table S1).

Evidence for convergent SBP evolution

We constructed a phylogenetic tree of the proteins analysed in this study (Figure 5). Among the three families that recognize preferentially saccharides, members of the SBP_bac_1 and SBP_bac_8 families were interwoven (Figure 5). In contrast, the third saccharide binding family, Peripla_BP_4, was phylogenetically distant from the former two families, suggesting convergent evolution for saccharide binding. Similar observations were made for the amino acid sensing SBPs, in which Peripla_BP_6 was distant to the two vicinal families, Lipoprotein_9 and SBP_bac_3. In addition, the phosphate‐binding families PBP_like_2 and Phosphonate_bd, were also phylogenetically distant. The tree shown in Figure 5 is based on sequence similarity. The inspection of a tree constructed on structural similarity (Scheepers et al., 2016) is also consistent with the notion of convergent evolutionary processes.

FIGURE 5.

Phylogenetic tree of SBPs analysed in this study. Branches have been coloured according to the corresponding PFAM family. Families that show a preference for ligands in the same compound family are boxed: red: saccharides, orange: amino acids, green: phosphate and related compounds. The phylogenetic tree was taken from iTOL.

Parallelism between SBP and chemoreceptor ligand profiles

The primary function of SBPs is to present solutes to transport proteins to initiate their uptake. Chemotaxis permits bacteria to access compounds of interest to promote their uptake. Data support that accessing nutrients and, more generally, locating niches that are optimal for growth are the major functions of bacterial chemotaxis (Colin et al., 2021). Typically, chemotaxis is initiated by the direct recognition of chemoeffectors by chemoreceptor sensor domains or the binding of signal‐loaded SBPs (Matilla et al., 2021). Because gaining access via chemotaxis and uptake of compounds are functionally linked, we compared SBP and chemoreceptor ligand profiles. In Table 2, we show that for every major SBP ligand family, there are chemoreceptors that respond to the same ligands. In most cases, chemoreceptors recognize chemoeffectors directly, indicating that there are sensor domains with similar ligand specificities as SBPs. Relatively few chemoreceptors rely on the recognition of ligand‐loaded SBPs (Table 2).

CONCLUSIONS

Previous studies have established a link between SBP structure and ligand range (Berntsson et al., 2010; Scheepers et al., 2016). In structural alignments of all available SBP 3D structures, clusters were defined that are specific for members of a ligand family. In the present study, we have investigated whether there is a link between SBP sequence and the type of ligand recognized. For the majority of SBP Pfam families, we report here the ligand profile. Currently, there is no or very scare information on the ligands recognized in the SBP family entries at the Pfam database (Mistry et al., 2021). In analogy to our previous study on ligand recognition by chemoreceptors, sensor kinases and transcriptional regulators (Matilla, Velando, Martín‐Mora, et al., 2022), the information on ligands recognized by SBPs will be introduced into the Pfam database, which will permit the scientific community to gain rapid first information on the ligands that may be recognized by an uncharacterized SBP. Such information will permit to rationally orient efforts to experimentally identify the ligand of an uncharacterized SBP. Chemotaxis permits bacteria to assess compounds of interest that are subsequently taken up by transporters. Chemotaxis and transport are thus functionally linked. Although the structural basis for ligand recognition at SBPs and chemoreceptors is different, we show here that for every major class of SBP ligands, there are characterized chemoreceptors that bind to the same ligands, revealing major environmental cues that have driven the co‐evolution of transport systems and chemoreceptors.

AUTHOR CONTRIBUTIONS

Jean Paul Cerna‐Vargas: Data curation (equal); formal analysis (equal); investigation (equal); methodology (equal); visualization (equal); writing – review and editing (equal). Beatriz Sánchez‐Romera: Data curation (equal); formal analysis (equal); writing – review and editing (equal). Miguel A. Matilla: Conceptualization; funding acquisition; project administration; writing – review and editing. Álvaro Ortega: Data curation (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); project administration (equal); writing – review and editing (equal). Tino Krell: Conceptualization (equal); funding acquisition (equal); project administration (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal).

FUNDING INFORMATION

This work was supported by the Spanish Ministry for Science and Innovation/Agencia Estatal de Investigación 10.13039/501100011033 (grants PID2020‐112612GB‐I00 to TK, PID2019‐103972GA‐I00 to MAM and PID2021‐122202OB‐I00 to AO) and the Junta de Andalucía (grant P18‐FR‐1621 to TK). JPCV was supported by the grant Unión Europea‐NextGeneration EU RD 289/2021 UPM‐Recualifica Margarita Salas.

CONFLICT OF INTEREST STATEMENT

The authors do not declare a conflict of interest.

Supporting information

Table S1

Table S2

Cerna‐Vargas, J.P. , Sánchez‐Romera, B. , Matilla, M.A. , Ortega, Á. & Krell, T. (2023) Sensing preferences for prokaryotic solute binding protein families. Microbial Biotechnology, 16, 1823–1833. Available from: 10.1111/1751-7915.14292