TtsI: Beyond Type III Secretion System Activation in Rhizobia

Abstract

The expression of the rhizobial symbiotic genes is controlled by various transcriptional regulators. After induction with appropriate plant flavonoids, NodD is responsible for the activation of the expression of genes related to Nod factor synthesis and secretion, but also, in most rhizobia harbouring a symbiotic type III secretion system (T3SS), the expression of ttsI. The ttsI gene encodes the positive regulator of the expression of T3SS-related genes, including those coding for structural components and for type III-secreted effector proteins. However, besides this general role among T3SS-harbouring rhizobia, different works have shown additional functions of TtsI in the regulation (positive or negative) of other bacterial traits such as the production of modified lipopolysaccharides or different types of motility (swimming or surface spreading). Interestingly, these additional functions appear to be rather specific than general among rhizobia. Moreover, in Sinorhizobium fredii HH103, TtsI affects the expression of various genes belonging to the nod regulon, including several transcriptional regulators. This review summarizes all the well-known bacterial traits affected by TtsI and describes other rhizobial genes that are regulated by TtsI but whose function remains to be established.

1. The Rhizobia–Legume Symbiotic Interaction

Rhizobia are soil proteobacteria able to establish a symbiotic nitrogen-fixing interaction with legumes [1]. For a successful interaction, rhizobia must infect host plant roots and colonize new organs, called nodules, formed by the plant [2,3,4]. Inside nodules, rhizobia are endocytosed into the so-called symbiotic cells and differentiate into bacteroids, which are the bacterial form able to fix N₂ into ammonia. Whereas rhizobia are fed with carbon and energy sources by the plant, most of the ammonia synthetized will be assimilated by the host, which allows it to grow in nitrogen-poor soils. Because of this, the rhizobia–legume symbiosis is important not only from an ecological perspective but also from an agricultural point of view because some of the most important crops are legumes (soybean, bean, etc.) [5].

The rhizobia–legume symbiosis relies on a complex signal interchange between both partners [2,6]. Legume root exudates contain various compounds, including flavonoids, that are supposed to physically interact with the bacterial protein NodD [7,8]. This interaction leads to the activation of NodD, which in turn binds to conserved promoter sequences, called nod boxes, located upstream of the symbiotic genes resulting in the induction of a set of rhizobial genes known as the nod regulon (Figure 1). The NodD proteins of different rhizobia vary in the spectrum, narrow or broad, of flavonoids they can perceive, which is related to the range of compatible legumes. Some of the genes belonging to the nod regulon are involved in the synthesis and secretion of a set of bacterial molecular signals termed Nod factors (NFs). These molecules are oligosaccharides of N-acetyl glucosamine (typically 3–5 residues) that harbour different decorations, such as sulphate, carbamoyl, or (methyl)fucose, among others, and including always the presence of a fatty acid linked to the N-acetyl glucosamine residue located in the non-reducing end of the NF. Each rhizobial species (or strain) produces a set of NFs that differ in their length and decorations, including variations in the fatty acid. In addition, different rhizobial species (or strains) produce distinct sets of NFs [9]. The NF molecules interact with specific LysM receptors located in the membrane of root hairs and, when these rhizobial molecules are appropriate, trigger the start of rhizobial root infection and nodule organogenesis [2,6]. Thus, the first specificity barrier of this symbiosis relies on two recognition events, NodD/flavonoids and LysM receptors/NFs. If both processes are successful, rhizobial infection is initiated and nodules are formed.

There are two main types of rhizobial root infection, intercellular and through infection threads (ITs) [10]. The first mode is considered more primitive than the second, and it takes place in approximately 25% of known legume species. The second mode involves root hair curling and the invagination of the plasma membrane of this cell, leading to the formation of a tube in which rhizobia will enter and move toward the nodule primordium. Different ways of intercellular infection can be found depending on whether NFs are required or not. On the other hand, the IT mode of infection depends on the participation of NFs, not only in the entry of rhizobia into the root but also in later steps of the interaction, such as the bacterial transit toward the nodule, the bacterial colonization of nodules, and the complete transition to nitrogen-fixing symbiosis [11]. In fact, in Mesorhizobium loti, there are two different NodD proteins, NodD1 and NodD2, that perceive plant compounds either in the rhizosphere and inside nodule cells (NodD2) or within infection threads (NodD1) [12].

In addition to NFs, other bacterial molecules are required for a successful symbiotic interaction, either by avoiding or counteracting plant defence responses, or by acting as signal molecules (Figure 1). Thus, some rhizobia possess a type III secretion system (T3SS) able to deliver effector proteins (T3Es) inside plant host cells [13,14]. On the other hand, different rhizobial surface polysaccharides (RSP) such as exopolysaccharide (EPS), lipopolysaccharide (LPS), K-antigen capsular polysaccharide (KPS), or cyclic glucans (CGs) perform important, sometimes essential, roles in symbiosis [8]. Although periplasmic cyclic glucans (CGs) have resulted to be absolutely required for a successful symbiosis in all the interacting couples studied so far, the symbiotic relevance of the other RSPs might vary depending on the specific rhizobium–legume couple [15,16].

The T3SS is the secretion system most extended among rhizobia, present in some strains belonging to the Mesorhizobium, Bradyrhizobium, Rhizobium, and Sinorhizobium genera and predominant among Bradyrhizobium spp. and S. fredii but rare in S. meliloti and not described in S. medicae (

Table 1. Reported secretion systems involved in rhizobia–host plant symbiosis.

Genus	Species	Demonstrated Presence of Functional TXSS 1			Comments	References
		T3SS	T4SS	T6SS
Bradyrhizobium					T3SS present in most strains.	[8,13,14]
	B. diazoefficiens	Yes				[23,24]
	B. elkanii	Yes				[25]
	B. japonicum	Yes				[26]
	B. vignae	Yes				[27]
	Bradyrhizobium sp.	Yes		Yes	T6SS described in LmicA16 strain.	[20,28]
Cupriavidus
	C. taiwanensis	Yes			Atypical T3SS.	[29]
Mesorhizobium
	M. amorphae	Yes				[30]
	M. loti	Yes	Yes		T4SS described in R7A strain.	[18,22]
Rhizobium
	R. etli			Yes	T6SS described in Mim1 strain. T3SS genes described in some strains.	[31,32]
	R. leguminosarum			Yes	T6SS described in RBL5787 strain. T3SS genes described in some strains.	[31,33]
Sinorhizobium
	S. fredii	Yes		Yes	T3SS present in most strains. T6SS described in USDA257 strain.	[8,13,14,34]
	S. medicae		Yes		T4SS present in most strains.	[17]
	S. meliloti		Yes		T4SS present in most strains. Low frequency of T3SS (functionality not demonstrated).	[8,13,17]

¹ While the presence of genes encoding different secretion systems has been described in various rhizobia species, this does not necessarily imply that these systems are functional or involved in symbiosis. In some cases, further research is needed to determine if these secretion systems are active and whether they play a role in the interaction between rhizobia and their host plants, especially for T6SS.

) [13,17]. Although previous reports mentioned the presence of a functional T3SS in several strains belonging to the genus Rhizobium, it is important to remark that these strains have been reclassified and assigned to other rhizobia genus such as Sinorhizobium (as is the case of many S. fredii strains, including NGR234) [13,14]. In addition, other secretions systems have been found to be present and symbiotically relevant in certain rhizobia, including the T4SS in M. loti R7A and many S. meliloti and S. medicae strains [17,18,19], or the T6SS in Rhizobium etli and Bradyrhizobium sp. LmicA16 [20,21]. Curiously, the T4 effector Msi059 of M. loti R7A is an orthologue of the T3e Mlr6316 of M. loti MAFF303099 [22]. In recent years, an increasing number of rhizobial species and strains possessing these secretion systems have been identified, a development enabled by advancements in computational technologies and bioinformatics tools. Importantly, despite the substantial amount of information derived from in silico studies, the functionality of these predicted systems needs to be experimentally validated.

2. TtsI, the Key Regulator of the Rhizobial Type III Secretion System

In rhizobia, genes coding for the structural components of the machinery of the T3SS and for some T3Es are clustered in the tts region, commonly located in the symbiotic plasmid (pSym). Together with this genomic region, genes coding for other T3Es can be found scattered in the genome, including the chromosome and plasmids other than the pSym. Generally, the transcription of the genes from the tts region and associated T3Es requires the presence of flavonoids and two transcriptional regulators, NodD and TtsI [35,36]. However, in the β-rhizobium Cupriavidus taiwanensis, the activation of the T3SS genes seems to be independent of both flavonoids and TtsI. Instead, it appears to be induced by glutamate, similarly to what is observed in plant pathogens, such as Ralstonia solanacearum and Pseudomonas aeruginosa [29]. For α- rhizobia, within the tts region, ttsI is the only gene preceded by a nod box. Thus, when NodD is activated by appropriate flavonoids, it induces the expression of ttsI. TtsI, in turns, binds to specific promoter sequences, called tts boxes, found upstream of genes coding for the structural components of the T3SS machinery and its T3Es, inducing their transcription (Figure 1). Interestingly, the expression of M. loti R7A VirA, which together with VirG constitutes the two-component regulatory system of the T4SS gene of this strain, is activated by NodD and flavonoids through a nod box located upstream of virA [19]. Thus, at least in this case, the master regulator of the R7A T4SS is subjected to the same symbiotic regulatory cascade as the T3SS of most rhizobia that harbour this secretion system.

The tts boxes are well conserved in rhizobia and the consensus sequences between Sino- and Bradyrhizobia are quite similar. Zehner et al. [37] identified three regions in the promoter. The modification of two or more nucleotides in any of these conserved regions significantly reduced gene expression, indicating their essential role in TtsI recognition.

TtsI is homologous to transcriptional regulators from two-component systems, which are composed of a sensor protein with histidine-kinase activity and a transcriptional regulator that modulates the response [38]. The sensor with kinase activity, after perceiving an external stimulus, would autophosphorylate in a histidine residue. Consequently, the transcriptional regulator is activated once phosphorylated by the sensor in an aspartic acid (Asp). The rhizobial TtsI possesses a glutamic acid (Glu) instead of the conserved Asp that acts as a primary acceptor of the phosphorylation in other transcriptional regulators, such as OmpR in Escherichia coli or NtrC in Salmonella typhimurium. It has been demonstrated that the substitution of Asp to Glu in these regulators leads to their constitutive activation, bypassing the need for the sensor kinase partner [36,39,40,41]. The fact that TtsI does not need to be activated by phosphorylation to regulate the transcription of tts genes is supported by other evidence, such as the absence of known sensors associated with the T3SS in rhizobia and the activation of tts genes in the absence of flavonoids when ttsI is overexpressed [42]. The transcription of these genes can hence be tightly regulated by flavonoids and other general transcriptional regulators that are components of the rhizobial nod regulon, closely associated with infection and nodule development.

The presence of rhizobial symbiotic secretion systems can entail a wide variety of phenotypes, including positive, neutral, or negative, depending on the specific bacteria–plant interacting couple [13,14,18,20,21]. Interestingly, some (but not all) rhizobial strains that infect intracellularly in a NFs-independent process use the T3SS to promote infection and nodulation [14,43,44,45]. For example, the T3E ErnA of Bradyrhizobium strain ORS3257 promotes nodulation in Aeschynomene most probably by directly altering plant gene expression [44].

Sinorhizobium fredii HH103 is a broad host range strain able to nodulate many legumes, including American and Asiatic varieties of soybean [46]. In HH103, the absence of TtsI or that of the T3SS component RhcJ results in a partial impairment of symbiosis with soybean [42,47], whereas in Lotus japonicus, a non-host legume of this strain, effective nodules are formed with both T3SS mutants. This suggests that HH103 T3Es can block nodulation in this plant upon their recognition. In fact, NopC is described as the main responsible for this blocking phenotype [48]. In soybean cultivars Williams and Peking, plants inoculated with the ttsI mutant exhibited a significant decrease in both the number and fresh weight of nodules formed, as well as a reduction in the plant shoot dry weight compared to plants inoculated with the parental strain [42]. Later studies have found that the mutation in ttsI also reduces nodule number and nodules dry weight in SN14 (cultivated soybean) and ZYD00006 (wild soybean) [49].

3. The nod Regulon of S. fredii HH103

The nod regulon of S. fredii HH103 is very complex: it comprises around 100 genes and involves various transcriptional regulators. The expression of 92 out of these 100 genes is dependent on NodD1, which is the main activator of this regulon [50,51]. However, at least four additional regulators participate in the modulation of the expression of these genes: NodD2, NolR, SyrM, and TtsI. NodD2 and SyrM have a dual role, repressing the expression of some genes and activating that of others, whereas NolR exerts a general repressor effect. In addition, TtsI, whose expression is driven by a nod box as mentioned above, is the activator of the T3SS assembly and T3E secretion.

Overall, following interaction with appropriate flavonoids, such as genistein, NodD1 triggers the production of NFs, the assembly and functioning of the T3SS, and the bacterial ability to spread on surfaces, whereas which the production of EPS [42,52,53,54] (Figure 2). SyrM is also required for the flavonoid-dependent repression of EPS production [55]. In addition, recent works [51,56] show that HH103 TtsI regulates the expression of several genes of the nod regulon apparently not related to the T3SS, as described below. Both NodD2 and NolR repress NF production and the T3SS, but the latter regulator is also required for surface motility [54,57].

4. In Some Rhizobia, TtsI Regulates Not Only the T3SS but Also Other Bacterial Traits

Although TtsI is considered the T3SS-related gene key regulator, at least in some rhizobia, this protein plays a role in controlling other symbiotic traits as well. Thus, in S. fredii NGR234, but not in strains HH103 and USDA257, TtsI is involved in the synthesis of the rhamnan component of a symbiotic specialized lipopolysaccharide (LPS), which contains a singular O-antigen polysaccharide and an altered core oligosaccharide-lipid A [36,58,59]. In fact, in NGR234, the expression of the rmlABCD genes, which are involved in the production of this Rhamnose-rich LPS, is driven by a tts box located in its promoter region. Rhamnose-rich LPS is believed to play a crucial role in facilitating bacterial release from infection threads and in protecting bacteria from plant defence responses [36,60]. To our knowledge, among rhizobia harbouring a T3SS, the presence of rmlABCD orthologues (mlr7550 to mlr755) have only been reported for M. loti MAFF303099 [36,61]. Interestingly, a putative tts box is located upstream of mlr7550, although the activation of these genes by TtsI, to our knowledge, remains to be elucidated.

On the other hand, TtsI appears to be crucial for biofilm formation on glass tubes in M. japonicum MAFF303099. Specifically, authors suggest that the biofilm formation capacity depends on a functional T3SS, or at least on pili formation [62]. In this strain, the second messenger cyclic di-GMP, which regulates the transition between motile and planktonic lifestyles in many bacteria, represses the expression of several T3SS genes, including ttsI. In addition, in the presence of flavonoids, MAFF303099 TtsI negatively regulates swimming motility [63]. In S. fredii HH103, the presence of inducing flavonoids does not affect swimming but triggers surface motility [54]. In this strain, surface motility is the result of flagella-dependent and flagella-independent mechanisms, and the T3SS apparatus (and therefore, TtsI) is required for both of them. Interestingly, TtsI positively regulates the expression of an HH103 flagellar gene, flgJ, which presents a functional tts box in its promoter region [56]. Interestingly, the tts box preceding flgJ does not perfectly follow the consensus sequence previously defined for HH103 [59], but it is functional, as demonstrated by its ability to drive gene expression in the presence of flavonoids in a NodD1- and TtsI-dependent manner 56]. At least two other HH103 flagellar genes (flgN and motF) are also induced by flavonoids, NodD1 and TtsI [50,51], although the mechanism of this induction remains to be elucidated since neither nod nor tts boxes have been identified upstream of flgN and motF. In any case, TtsI appears to activate HH103 genistein-induced surface spreading through its role as master regulator of the T3SS but also because it activates the expression of several flagellar genes. In addition, Fuentes-Romero et al. [64] demonstrated that flagellar-dependent surface motility in this bacterium is also induced by the presence of a non-ionic osmotic stress (400 mM mannitol) in a TtsI-dependent manner. It is important to remark that neither the presence of genistein nor the lack of NodD1, NolR, or TtsI affects the presence of flagella in HH103. Clearly, more research is required to shed light in how the T3SS participates in surface motility and to elucidate whether the expression changes in several flagellar genes promoted by genistein and the NodD1 and TtsI regulators have an effect on the HH103 flagella-dependent surface motility.

Unusually, in certain cases, TtsI also seems to act as a negative regulator. In this sense, as mentioned above, the M. loti MAFF303099 TtsI is involved in the negative regulation of bacterial swimming motility in semi-solid media. In this bacterium, induction with flavonoids and the presence of NodD negatively affects flagella synthesis and the expression of visN, a homolog of a described rhizobial positive regulator of flagellar genes [62,63]. Presumably, this effect may occur through the assembly of a T3SS complex. Consequently, it has been suggested that the presence of a T3SS complex at the membranes may act as a signal for a negative regulatory cascade affecting flagellar genes or may physically affect flagella assembly [63]. The flagellar switch-off mechanism has been previously identified in phytopathogens, pointing out that it facilitates bacteria to evade plant defence responses triggered by flagellins during plant infection [65]. However, the rhizobial flagellum has not been identified as a plant defence elicitor so far [66]. Thus, the general observation that rhizobial flagella are no longer necessary within the infection threads may support the maintenance of the TtsI–motility inverse regulation throughout evolution [63]. As an exception of the rule, in the particular case of the symbiosis between R. leguminosarum and Pisum sativum, most flagellar assembly and basal body apparatus proteins are required in bacteroids and nodule bacteria [67].

Transcriptomic studies carried out in S. fredii HH103 have revealed that TtsI, in addition to inducing all the genes related to the T3SS and the three flagellar genes mentioned above, affects the expression of other genes of the nod regulon located on the chromosome and megaplasmids e and d (

Table 2. Summary of HH103 genes whose expression is affected at least 1.5-fold in the ttsI mutant regarding the wild-type strain (WT) when both strains are cultured in the presence of the nod gene-inducer genistein [50,51].

Locus Tag/Gene Name	Putative Function	Fold Change Expression in the WT + Genistein a	Fold Change Expression in the ttsI Mutant + Genistein a	Ratio ttsI Mutant/WT b
Chromosome
SFHH103_00346/flgJ	Flagellar protein	7.57	2.07	0.27
SFHH103_00347/flgN	Flagellar protein	6.98	1.52	0.22
SFHH103_00348/motF	Flagellar protein	4.34	1.24	0.29
SFHH103_00844	Zinc-finger protein, putative RNA binding	3.23	1.20	0.37
SFHH103_01317	Siderophore synthetase component	0.31	0.99	3.16
SFHH103_01920	NAD(P)-dependent oxidoreductase	4.59	1.49	0.32
SFHH103_02192	Calcium-binding exoprotein	6.90	11.64	1.69
SFHH103_02323/hmuU	Hemin ABC transporter	0.30	1.26	4.13
SFHH103_03749	TetR family transcriptional regulator	3.88	1.53	0.39
pSfHH103e
SFHH103_05319	TetR family transcriptional regulator	0.29	0.47	1.62
SFHH103_05320	Multidrug efflux transporter protein	0.20	0.42	2.08
SFHH103_05321	Multidrug efflux transporter protein	0.18	0.40	2.23
SFHH103_06013/pdh	Putative pyruvate dehydrogenase E1 component	3.58	2.41	0.67
pSfHH103d
psfHH103d_208	ABC-type transport system, periplasmic component	17.43	39.14	2.25
psfHH103d_275	Hypothetical protein	3.45	1.43	0.41
psfHH103d_274	Hypothetical protein	3.36	1.24	0.37
psfHH103d_367	SyrM LysR family transcriptional regulator	5.81	9.29	1.60

^a Fold change values regarding the wild-type strain in the absence of flavonoids [33,34]. ^b In bold: genes up or downregulated in the mutant with a ratio >2 or <0.5, respectively.

) [50,51]. On the one hand, TtsI appears to slightly induce (more than two-fold) the expression of genes coding for a putative RNA-binding protein containing a Zinc-finger domain, a transcriptional regulator of the TetR family (SFHH103_03749), an NAD(P)-dependent oxidoreductase, and two hypothetical proteins. On the other hand, TtsI appears to slightly downregulate the expression of two genes related to iron uptake, two genes coding for the components of a multidrug efflux pump, one gene coding for the periplasmic component of an ABC-type transport system, and the coding gene of another TetR transcriptional regulator (SFHH103_05319). Navarro-Gómez et al. [51] also reported a negative effect of TtsI on the expression of another symbiotic regulator, SyrM, which was further confirmed by qPCR and β-galactosidase assays of syrM::lacZ mutants carrying either a mutated version of ttsI or extra copies of the wild-type ttsI gene. In addition, these authors also showed a negative autoregulation of ttsI in this strain. Thus, TtsI plays a role in the fine-tuning of the complex regulatory network that controls symbiotic genes in this bacterium and affects the expression of different regulatory proteins. All these results suggest that the TtsI protein of S. fredii HH103 performs additional roles, beyond the T3SS, in symbiosis.

5. Conclusions and Perspectives

Since the discovery of the role of TtsI in the regulation of the T3SS genes in rhizobia, its function has been considered to be solely related to this process. In fact, when studying the role of rhizobial T3Es in the symbiosis of various rhizobia species with their host plants, mutants affected in ttsI have been widely used. However, recent advances in the understanding of the rhizobial T3SS and its regulation have revealed an unexpected complexity in the function of TtsI and also showed that this regulator might act not only as an activator but also as a repressor. While the role of TtsI as the activator of the symbiotic T3SS is common in all the rhizobial species that harbour this secretion system, the additional roles mentioned in this review seem to be rather specific for the rhizobial strain in which they have been described. However, is this specificity always real or are these observations due to the absence of the appropriate studies in other rhizobia? In our opinion, future research is required to elucidate the different bacterial traits that are regulated by TtsI and how widespread they are. The fact that Type III secretion (T3S) is common to the bacterial flagellum and to the type 3 secretion apparatus [68] leads to another important question that should be addressed in the future: is this fact related to the involvement of the T3SS in the motility of different rhizobia? In addition, when studying the relevance of the symbiotic T3SS, it is becoming increasingly important to include mutants affected in genes related to the structural components of this system, such as rhcJ or rhcV, rather than relying solely on the use of ttsI mutant strains. However, conducting a comparative analysis of the symbiotic phenotypes presented by ttsI mutants and those affected in the structural genes of the T3SS would be a promising area for future research, since it could provide information about other putative roles of TtsI in the rhizobium–legume symbiosis. Thus, this approach would provide important crucial information for the scientific community.