The Two-Component System YesMN Promotes Pneumococcal Host-to-Host Transmission and Regulates Genes Involved in Zinc Homeostasis

ABSTRACT

The ability to sense and respond rapidly to the dynamic environment of the upper respiratory tract (URT) makes Streptococcus pneumoniae (Spn) a highly successful human pathogen. Two-component systems (TCSs) of Spn sense and respond to multiple signals it encounters allowing Spn to adapt and thrive in various host sites. Spn TCS have been implicated in their ability to promote pneumococcal colonization of the URT and virulence. As the disease state can be a dead-end for a pathogen, we considered whether TCS would contribute to pneumococcal transmission. Herein, we determined the role of YesMN, an understudied TCS of Spn, and observe that YesMN contributes toward pneumococcal shedding and transmission but is not essential for colonization. The YesMN regulon includes genes involved in zinc homeostasis and glycan metabolism, which are upregulated during reduced zinc availability in a YesMN-dependent fashion. Thus, we identified the YesMN regulon and a potential molecular signal it senses that lead to the activation of genes involved in zinc homeostasis and glycan metabolism. Furthermore, in contrast to Spn monoinfection, we demonstrate that YesMN is critical for high pneumococcal density in the URT during influenza A virus (IAV) coinfection. We attribute reduced colonization of the yesMN mutant possibly due to increased association with and clearance by the mucus covering the URT epithelial surface. Thus, our results highlight the dynamic interactions that occur between Spn and IAV in the URT, and the role that TCSs play in modulation of these interactions.

INTRODUCTION

The nasopharynx is the initial site of acquisition and the primary reservoir of the human-adapted extracellular respiratory pathogen Streptococcus pneumoniae (Spn; the pneumococcus). Spn resides in the mucosal surface of the upper respiratory tract (URT), and carriage is typically asymptomatic. From this niche in the URT, pneumococcus can spread to another host or cause disease within the same host. Disease states associated with Spn include pneumonia, septicemia, otitis media, and meningitis. Invasive pneumococcal disease is associated with over a million deaths worldwide (1–3).

The ability to colonize and thrive in different host environments is a hallmark of Spn pathogenesis. Multiple studies have shown that pneumococcus modifies its global gene expression depending on the colonization site (4–7). Adapting and responding to the dynamic environments that bacterial pathogens encounter requires rapid responses, some of which are mediated through two-component signal transduction systems (TCSs). Bacterial TCSs generally consist of two proteins, a histidine kinase (HK) that senses external stimuli, and a response regulator (RR). Upon autophosphorylation of the HK in response to external stimuli, the phosphate group is transferred to the RR, which undergoes a conformation change allowing it to act as a regulator (8, 9).

The Spn genome encodes 13 TCSs, each consisting of an HK and a cognate RR, and one orphan response regulator (8, 10). As with other bacterial pathogens, several of these TCSs contribute to pneumococcal colonization and virulence. They sense a variety of signals, including, but not limited to, reactive oxygen species, oxidative stress, antibiotics, antimicrobial peptides, pH, and temperature. In response, they modulate the expression of genes involved in reducing oxidative stress, uptake of nutrients, carbon metabolism, modulating competence and fratricide, releasing bacteriocins, and other factors that affect host-pathogen interactions (8, 10, 11). Several TCSs of Spn have been extensively characterized, including BlpRH, ComDE, and CiaRH. BlpRH is involved in quorum sensing and bacteriocin production; ComDE, through quorum sensing, regulates Spn transformation machinery under specific environmental conditions; CiaRH regulates expression of the serine protease HtrA and the stress response machinery, and negatively influences competence (8, 12). Furthermore, there is cross talk signaling between pneumococcal TCSs, where they respond to the same environmental signal (ComDE and CiaRH both respond to acid stress) and regulate the expression of each other (12–14).

Specific pneumococcal factors, such as pneumolysin (15), dlt locus-dependent d-alanylation of Spn teichoic acid (16), and the type and amount of Spn capsular polysaccharide (17), are essential for Spn host-to-host transmission. These factors modulate the host inflammatory response or the dynamic interactions with mucus protecting the URT epithelial layer. Recent high-throughput transposon mutagenesis (Tn-seq) screens have identified Spn factors important for colonization, virulence, and host-to-host transmission (16, 18, 19). As TCSs can regulate hundreds of genes, it was not surprising that they would be identified in these screens to contribute toward pneumococcal pathogenesis. However, one of the least characterized TCSs, YesMN (also called TCS07), was found to contribute toward pneumococcal host-to-host transmission in two dissimilar Tn-seq approaches with different animal models (16, 19), indicating that YesMN regulates factors that are important determinants of pneumococcal spread.

The putative regulon of YesMN was initially identified by overexpressing YesN (RR) and performing next-generation RNA sequencing (20). Genes involved in host-derived glycan metabolism and transport were among the YesMN regulon. Specifically, genes encoding enzymes for N-glycan metabolism and transport were identified to be upregulated in a YesMN-dependent manner (20). These results were unsurprising, as free carbon is scarce in the URT (21). Spn devotes considerable resources to acquiring carbon by cleaving host glycans, such as heavily glycosylated mucin proteins that form the mucus layer (22, 23). Nevertheless, the signal and the biological relevance in the nasopharynx of YesMN remain poorly defined.

Herein we aimed to validate, using our infant mouse pneumococcal challenge model, the in vivo role of YesMN suggested by Tn-seq screens (16, 18) with a yesMN deletion mutant tested outside the context of a mixed infection. In addition, we sought to better define the YesMN regulon using a yesMN deletion strain rather than overexpression. Our results indicate that YesMN does indeed play a role in transmission but not in URT colonization. Transcriptomic analysis revealed the YesMN regulon includes zinc (Zn²⁺) homeostasis genes and validated the published role in glycan metabolism (20). Importantly, we found that YesMN impacts its regulon in response to changes in zinc levels, suggesting that zinc may be the signal that is sensed by YesMN.

RESULTS

Effect of yesMN deletion on pneumococcal colonization, shedding, and host-to-host transmission.

A transposon mutagenesis screen with a serotype 4 isolate (TIGR4) revealed putative Spn factors that contribute toward pneumococcal nasopharyngeal colonization and shedding (16, 18). A candidate of particular interest was the HK YesM that forms part of the two-component system (TCS) YesMN (TCS07) (Fig. 1A). Overexpression of the response regulator (RR) YesN in D39 (type 2 isolate) leads to an increase in the expression of genes that code for N-linked glycan-metabolizing enzymes. Also, compared to the parental D39, ΔyesMN had a slightly reduced virulence in a lung infection model (20).

FIG 1.

The two-component system YesMN of pneumococcus contributes toward its shedding and host-to-host transmission. A. Schematic representation of the genetic organization of the TCS YesMN of S. pneumoniae with the TIGR4 designation. Adjusted P values for the genes from the Tn-seq screen that identified YesMN as being important for pneumococcal shedding are listed (16), with the solid line representing the gene deletions constructed for the current study. B. The 4-day-old pups were challenged i.n. with the indicated strain, and nasal secretions were collected, and bacterial load was enumerated for 5 days postinfection. Median CFU values for each day are indicated, and each symbol represents the number of CFUs measured from a single pup (n ≥ 5 for each strain tested). Shedding differences between the two strains on a given day were calculated using the Mann-Whitney U test. C. Pups were challenged i.n. at day 4 of age with the indicated construct, and bacterial shedding from nasal secretions were quantified for 5 days postinfection. Violin plots of pooled 5 days of shedding are indicated with median shown in red. Data are from 6 to 10 pups per group. Statistical differences were calculated using the Kruskal-Wallis analysis of variance with Dunn’s posttest. D. Colonization density of the WT, isogenic mutants, and the chromosomally complemented strain in cultures of URT lavage fluids obtained from pups at day 9 of age. Median values are shown. Dotted line represents the limit of detection (LOD). E. Summary of transmission efficiency of the WT, ΔyesMN mutant, and the corrected strain (yesMN⁺) from colonized index pups at the age of 4 days to naive contact pups in the same litter. The transmission efficiency was determined by the number of contact pups (at a 1:1 ration to index mice) colonized by S. pneumoniae at the age of 14 days. Fisher’s exact test was used to determine the statistical difference between the WT and the ΔyesMN mutant and yesMN⁺. *, P ≤ 0.05; **, P ≤ 0.01; ****, P ≤ 0.0001; n.s, not significant.

We were particularly intrigued by identifying the yesMN TCS in our shedding screen, as of the 13 pneumococcal TCSs, YesMN regulation is poorly understood (8). Moreover, another in vivo transposon screen to identify putative transmission factors of Spn using a ferret model identified the RR YesN as being necessary for host-to-host transmission (19). To confirm that the TCS YesMN impacts pneumococcal shedding, we constructed mutants with unmarked, in-frame deletion of HK yesM (ΔyesM), the RR yesN (ΔyesN), or both yesMN (ΔyesMN). Deletion of yesMN was not associated with impaired growth (20) and all the constructed mutants in the study grew as wild-type (WT) type 4 isolate (data not shown).

To determine if YesMN contributes toward Spn shedding, we first assessed ΔyesM, which was identified in our shedding screen (16). Shedding was measured daily (days 1 to 5 postinoculation) from the inoculated pups. Compared to the WT type 4 strain, the isogenic mutant ΔyesM shed poorly over 5 days (Fig. 1B). Because of established day-to-day variation (24), daily shedding values were also pooled for comparison (Fig. 1C). The HK (ΔyesM), RR (ΔyesN) mutants, and the complete TCS knockout (ΔyesMN) all showed reduced median shedding compared to either the WT type 4 strain or the corrected mutant (yesMN⁺ mutant). These results included an intralitter comparison of strains inoculated into pups from multiple litters to control for host and environmental effects on shedding. The contribution of yesMN to shedding was not due to differences in colonization density, as all the strains colonized equally well at the conclusion of the shedding experiment at the age of 9 days (Fig. 1D).

Next, we determined whether reduced shedding attributed to the yesMN mutants manifests as a defect in host-to-host transmission. Using a 1:1 ratio of Spn-inoculated pups (index) to uninoculated littermates (contact pups), ~40% transmission was observed with the WT type 4 strain. In contrast, no transmission events were detected with the ΔyesMN strain. Transmission to WT levels was restored with the corrected strain (yesMN⁺) (Fig. 1E). Furthermore, all constructs colonized the pups at a high density at the conclusion of the transmission study at the age of 14 days (see Fig. S1 in the supplemental material). Thus, we validated our and others' observations, confirming that under Spn monoinfection, YesMN does not affect colonization but is required for sufficient pneumococcal shedding to allow for host-to-host transmission.

Comparative transcription analysis of YesMN mutant.

Recently, RR yesN overexpression led to the identification of the YesMN regulon (20). As pneumococcal TCSs exist in a balanced cross talk, overexpression of a single RR can potentially affect other signaling pathways. To provide more physiologically relevant molecular testing conditions, we determined the global gene expression patterns in a ΔyesMN strain to identify the putative YesMN regulon. The WT type 4 and the ΔyesMN strains were grown in tryptic soy yeast (TSY) medium, and total RNA transcript abundance was measured using RNA-seq. A total of 137 differentially regulated genes were identified in the ΔyesMN compared to the WT strain (Log2FC>±1, Padj < 0.05), of which 88 were downregulated and 49 were upregulated (Fig. 2A to C) (Table S1). Besides genes of unknown function, most of the genes identified were involved in the uptake of nutrients and their metabolism (Fig. 2B to C). The top downregulated genes encode proteins involved in zinc homeostasis, host-glycan and carbohydrate transport and metabolism, and the locus associated with purine metabolism (Table 1). The putative YesMN regulon included genes previously shown to be necessary for pneumococcal colonization and/or shedding (16, 18). Notably, genes that encode enzymes involved in host-glycan metabolism (Table 1) identified in our RNA-seq analysis were correspondingly identified when YesN was ectopically overexpressed in the D39 background (20). Thus, our RNA-seq results align with the previously described study and identified new members of the YesMN regulon.

FIG 2.

Identification of the genes regulated by the pneumococcal TCS YesMN. The transcriptome of wild-type TIGR4 and the isogenic ΔyesMN mutant was compared through RNA-seq analysis. Strains were grown in TSY to OD₆₂₀ of ~0.4, RNA isolated, and prepared for analysis as described in Materials and Methods. A. Volcano plot depicting genes whose RNA transcripts were differentially regulated (Log2FC > ±1; P < 0.05). Genes involved in zinc homeostasis are highlighted. B and C. Functional annotation of genes appearing clockwise whose transcripts in the ΔyesMN mutant compared to the wild type were either downregulated (B) or upregulated (C).

TABLE 1.

Major functional classes of genes whose transcripts were downregulated in a ΔyesMN mutant^a

TIGR4 designation	Gene name and description	Function	Log2FC
Manganese and zinc homeostasis
SP_2136	pcpA; choline binding protein (CbpN)	Adherence and virulence (79, 80)	−2.87
SP_1002	adcAII; zinc transport system substrate-binding protein	Colonization and virulence (26, 81)	−2.59
SP_1174	phtB; histidine triad protein	Adherence and immune evasion (79, 82)	−2.44
SP_1003	phtD; histidine triad protein	Adherence and immune evasion (79, 82)	−2.08
SP_1004	phtE; histidine triad protein	Adherence and immune evasion (82, 83)	−1.49
SP_1175	phtA; histidine triad protein	Adherence and immune evasion (79, 82)	−1.13
SP_1650	psaA; ABC transporter	Manganese/zinc transport, virulence (84, 85)	−1.23
SP_1648	psaB; ABC transporter	Manganese/zinc transport, virulence (84)	−1.18
SP_1649	psaC; ABC transporter	Manganese/zinc transport, virulence (84)	−1.21
Host glycan metabolism
SP_2143	GH38; putative exo-α-(1, 3)-mannosidase	Glycan metabolism, colonization, and virulence (18, 31)	−2.07
SP_2144	GH125; exo-α-(1, 6)-mannosidase	Glycan metabolism and virulence (31, 86)	−1.75
SP_0091	ngtP1; ABC transporter	Glycan transport (31)	−1.67
SP_0090	ngtP2; ABC transporter	Glycan transport (31)	−1.63
SP_0092	ngtS; ABC transporter	Glycan transport and virulence (44)	−1.61
SP_2142	ROK; putative sugar kinase	Glycan metabolism and virulence (31, 44, 86)	−1.55
SP_2141	GH20C; exo-β-hexosaminidase	Glycan metabolism (31)	−1.43
SP_0498	endoD; (GH85); N-glycan endo-β-N-acetylglucosaminidase	Glycan metabolism and virulence (31, 44, 87)	−1.14
Carbohydrate metabolism
SP_2021	6-Phospho-β-glucosidase	Putative cellobiase (25)	−1.16
SP_2022	Phosphotransferase system	Cellobiose-specific IIC component and virulence (25, 44)	−1.15
SP_0308	Phosphotransferase system	Cellobiose-specific component IIA (25)	−1.12
SP_2023	Phosphotransferase system	Cellobiose-specific IIB component (25)	−1.09
SP_2024	Phosphotransferase system	Cellobiose-specific IIA (25)	−1.09
SP_0307	Hypothetical protein		−1.08
Purine biosynthesis
SP_0044	purC	Purine biosynthesis (88, 89)	−2.71
SP_0052	Hypothetical protein		−2.37
SP_0045	purL; putative phosphor-ribo-sylformyl-glycinamidine synthase	Purine biosynthesis and Virulence (44)	−2.33
SP_0051	purD	Purine biosynthesis	−2.32
SP_0053	purE; 5-(carboxyamino)imidazole ribonucleotide mutase	Purine biosynthesis	−2.17
SP_0046	purF; amidophosphoribosyltransferase	Purine biosynthesis	−2.13
SP_0047	purM; phosphoribosylformylglycinamide cyclo-ligase	Purine biosynthesis	−2.11
SP_0048	purN; phosphoribosylglycinamide formyltransferase	Purine biosynthesis	−2.04
SP_0054	purK; 5-(carboxyamino)imidazole ribonucleotide synthase	Purine biosynthesis	−2.02
SP_0049	vanZ; hypothetical protein	Virulence (44)	−1.95
SP_0050	purH; phosphoribosylaminoimidazolecarboxamide formyltransferase	Purine biosynthesis and virulence (44)	−1.32

^aGene name, description, and function were based on information retrieved from KEGG and UniProt databases and literature searches.

Examination of the upstream region of the genes that form the YesMN regulon did not yield a consensus binding site for the RR07, suggesting further analysis is required to identify a putative consensus binding site. Furthermore, CiaR that forms the part of the CiaRH TCS and several of its regulon members were in the list of genes whose transcripts were upregulated (~2-fold) in the ΔyesMN strain (Table S1), indicating cross-regulation of pneumococcal TCS. In addition, the highly conserved fruRBA operon (25) involved in fructose uptake and metabolism was also upregulated. Taken together, our RNA-seq analysis provides a global view of the YesMN regulon. Herein, we focused on genes involved in N-glycan metabolism and zinc homeostasis as these were part of the most underrepresented genes/pathways identified in our screen.

To validate the results from the RNA-seq analysis, quantitative real-time PCR (qRT-PCR) was performed targeting 11 specific genes implicated in either zinc homeostasis or host-glycan metabolism (Table 1; Fig. 3). Neuraminidase A (nanA) and B (nanB) were selected as controls, as they were not identified by the RNA-seq screen to be modulated by YesMN. The YesMN-regulated genes to be validated included adcAII, which encodes an adhesion lipoprotein with specificity for Zn²⁺ (26); and phtD and phtE, known as pneumococcal histidine triad (Pht) proteins that form an operon with adcaII. Pht family proteins are also involved in Zn²⁺ acquisition, and upregulated under low Zn²⁺ (27, 28). pcpA, which encodes choline-binding protein N that acts as an adhesin (29), was also observed in the RNA-seq screen to be regulated by YesMN. However, in the presence of Mn²⁺, PsaR, a metal-dependent regulator, negatively regulates the expression of several genes, including pcpA (30). Besides genes involved in Zn²⁺ homeostasis, we tested endoD, GH20C, ROK, GH38, and GH125, which are involved in N-glycan metabolism (20, 31). We observed a statistically significant ~2-fold reduction in the transcripts of the selected genes when comparing ΔyesMN to the WT strain (Fig. 3A and B). No reduction was observed in the transcripts of control genes (nanA and nanB) that do not belong to the YesMN regulon (Fig. 3A). Taken together, our results strongly suggest that YesMN regulates genes involved in zinc homeostasis and glycan metabolism.

FIG 3.

YesMN regulates genes involved in zinc homeostasis and glycan metabolism. Differential gene expression in ΔyesMN of selected genes by quantitative real-time PCR (qRT-PCR). WT and the ΔyesMN isogenic mutant were grown to OD₆₂₀ of ~0.4 in TSY broth at 37°C, RNA isolated, and prepared for qRT-PCR analysis. Shown is fold change in transcription of genes between WT and ΔyesMN involved either in (A) zinc homeostasis or (B) glycan metabolism. nanA and nanB that are not part of the YesMN regulon were tested as controls (gray bars). TIGR4 16s (rrsA, SP_0775) specific primers were used as the housekeeping gene for 2^–ΔΔCt analysis. For each biological replicate (n ≥ 3 for WT and ΔyesMN), qRT-PCR was conducted in triplicate. Bars represent mean ± SEM. Statistical differences between the WT and the yesMN mutant were calculated using the Mann-Whitney U test. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

YesMN appears to respond to Zn²⁺ levels to modulate its regulon.

Based on our RNA-seq analysis that identified genes involved in Zn²⁺ homeostasis belonging to the YesMN regulon, we hypothesized that YesMN responds to Zn²⁺ levels in the environment. Previously, adcaII and pht gene transcripts were found to be upregulated when grown in N, N, N′, N′-tetrakis-(2-pyridylmethyl)-ethylene-diamine (TPEN), a membrane-permeable zinc chelator (32). Correspondingly, the TSY-growing WT strain, when shifted to TSY containing 30 μM TPEN, led to an upregulation in the transcript levels of the adcaII and pht genes, as well as pcpA (~15- to 30-fold) (Fig. 4A). In addition, transcripts of glycan metabolism genes were upregulated (Fig. 4B), albeit less (~2- to 4-fold) than the zinc homeostasis genes. In contrast, in the ΔyesMN strain, TPEN treatment did not increase the transcripts of genes involved in zinc homeostasis or glycan metabolism to the same level as WT (Fig. 4A and B). Zinc homeostasis genes were still upregulated (~3- to 4-fold) in the ΔyesMN strain under TPEN, suggesting that other factors also contribute to the increase in expression of these genes under reduced levels of Zn²⁺. Collectively, our data suggest that the Zn²⁺ homeostasis and glycan metabolism genes form part of the YesMN regulon and respond to reduced Zn²⁺ levels.

FIG 4.

YesMN responds to reduced availability of zinc. Quantitative real-time PCR analysis was used to examine gene expression in the WT and the ΔyesMN isogenic mutant in response to zinc chelation. Strains were grown to an OD₆₂₀ of ~0.4 in TSY broth at 37°C. Half of the culture was transferred to a clean tube and treated with TPEN and the other half remained untreated; these were grown for an additional 30 min at 37°C. RNA was isolated and prepared for qRT-PCR analysis as described in the Materials and Methods. TIGR4 16s (rrsA, SP_0775)-specific primers were used as the housekeeping gene for 2^–ΔΔCt analysis. Bars (mean ± SEM) in red represent WT whereas gray bars represent ΔyesMN and indicate the fold change in transcript abundance between treated (TPEN) or untreated zinc homeostasis genes (A) or glycan metabolism genes (B). For each biological replicate (n ≥ 3 for WT and ΔyesMN), qRT-PCR was conducted in triplicate. Statistical differences between the WT and the ΔyesMN mutant were calculated using the Mann-Whitney U test. **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

YesMN is required for robust colonization during concurrent influenza A infection.

Epidemiological data show that bacterial-influenza A virus (IAV) coinfections occur in more than 30% of all IAV cases, predominantly with pneumococcus (33, 34). These coinfections are associated with increased morbidity and mortality (35, 36). Spn-IAV coinfection alters the nasopharyngeal landscape of the infected host, with an increase in inflammation and mucus production, allowing Spn to proliferate further (37–39). Recently, a transposon mutagenesis screen using a type 3 isolate identified Spn factors important for nasopharyngeal colonization with a concurrent IAV infection. The list of genes identified included several that form part of the YesMN regulon and are involved in N-glycan metabolism (40). We hypothesized that even though the ΔyesMN strain colonizes as WT under monoinfection, its regulon members are potentially critical for colonization during IAV coinfection.

To test our hypothesis, we inoculated pups with either the ΔyesMN strain or the WT type 4 isolate at 4 days of life and, on day 9, gave them the mouse-adapted influenza A virus/Puerto Rico/8/34(H1N1), here referred to as “PR8,” and followed the modulation of Spn shedding from the pups over 5 days post-IAV infection. Analogous to our previous studies, the WT strain shed at a high level during IAV coinfection (24). Like the results for monoinfection (Fig. 1C), the ΔyesMN strain also had a defect in shedding during IAV coinfection (Fig. 5A). However, unlike monoinfection, where the ΔyesMN strain colonized as WT (Fig. S1), with a concurrent IAV infection, the mutant colonized poorly (Fig. 5B), suggesting that the YesMN regulon promotes robust colonization in the shifting landscape of the nasopharynx. Furthermore, the shedding and colonization defect was YesMN specific as the yesMN⁺ strain shed and colonized as WT. Thus, our data show that even though YesMN is dispensable during monoinfection, it is critical for pneumococcal colonization during an IAV coinfection and provides further evidence that Spn-IAV interactions are not necessarily synergistic in the upper respiratory tract.

FIG 5.

YesMN is required for robust pneumococcal shedding and colonization in the setting of influenza A virus coinfection. A. Four-day old pups were challenged i.n. with either the WT, the isogenic yesMN mutant, or the corrected yesMN⁺ strain. Five days later, pups were challenged with influenza A virus, and pneumococcal shedding in nasal secretions was quantified daily for 5 days. Pooled median shedding values of 5 days are shown, where each symbol represents CFU values obtained from a single pup. B. Colonization density of each pneumococcal construct in cultures of URT lavage fluids obtained from pups at day 14 of age. Median values are shown. Dotted line represents the LOD. Data are from 8 to 11 pups per group. Statistical differences were calculated using the Kruskal-Wallis analysis of variance with Dunn’s posttest. *, P ≤ 0.05; **, P ≤ 0.01; n.s, not significant.

YesMN impacts the interaction of pneumococcus with human nasal fluid.

The YesMN regulon includes genes that encode proteins that act as adhesins, promote virulence, and contribute toward N-linked glycan metabolism (Table 1). Previous studies demonstrated that high pneumococcal shedding is dependent upon (i) ability of Spn to cause a robust inflammatory response in the URT that results in an increase in nasal secretions, and (ii) its ability to escape mucus present on the URT epithelial surface (15, 17, 37, 41, 42). We therefore examined whether the reduced shedding from the URT of the ΔyesMN strain during mono-infection is due to reduced inflammation. The chemokines (CXCL-1, CXCL-2) and cytokine interleukin-1β (IL-1β) are considered hallmarks for Spn-induced inflammation in the URT (15), and their levels were measured following inoculation with WT or the yesMN mutant. At 2 days postinoculation, we did not observe upregulation of these inflammatory markers in mice inoculated with either strain. (Fig. S2). Thus, our qRT-PCR data likely precludes Spn-mediated inflammatory response in the URT as the basis for reduced shedding associated with the ΔyesMN mutant.

Based on our observation, we postulated that an increase in nasal secretions leads to a reduced load of the ΔyesMN mutant in URT during coinfection. To determine whether YesMN contributes to interactions with mucus, we performed a previously described in vitro solid-phase human mucus binding assay using pooled human nasal fluid (hNF) (43). Adherence of the ΔyesMN strain to mucus was significantly higher than the WT type 4 isolate, suggesting that YesMN regulates pneumococcal factors that promote mucus evasion (Fig. 6). This enhanced adherence was not because of reduced capsular polysaccharide (CPS), as both the mutant and the WT strain made similar levels of CPS (Fig. S3). Moreover, this enhanced interaction with hNF was specific to the ΔyesMN mutant, as the yesMN⁺ strain restored reduced interaction with hNF. Overall, our results demonstrate that YesMN-regulated factors influence interactions with the components that constitute hNF.

FIG 6.

YesMN affects human nasal fluid association. Comparison of binding level of WT, the ΔyesMN, and the yesMN⁺ strains to immobilized human nasal fluid (hNF). A total of 10 μg of hNF in 100 μL was immobilized in a 96-well micotiter plate overnight and then blocked with 0.1% BSA. After gentle washing, bacteria (2× 10⁴ CFU/100 μL DMEM) were added, and the plate was incubated for 2 h at 30°C. After 19 washes, adherent bacteria were determined by resuspending with 0.001% Triton X-100 following plating on TSY agar plates supplemented with 200 μg/mL streptomycin. Values are means of three independent determinations in triplicate ± SEM. Statistical differences were calculated using the Kruskal-Wallis analysis of variance with Dunn’s posttest. **, P < 0.01; ***, P < 0.001; n.s, not significant.

DISCUSSION

In the current study, we sought to increase our understanding of nasopharyngeal colonization and host-to-host transmission of the human-adapted pathogen Streptococcus pneumoniae using an infant mouse model (24). Two-component systems respond to various overlapping signals that allow for quick adaptation and establishment by Spn within the nutrient-poor but dynamic environment of the nasopharynx (8). Two independent Tn-seq screens identified one of the least characterized TCSs (YesMN) as contributing to the spread of Spn from one host to another without impacting its colonization of the URT (16, 19). Thus, to provide a molecular understanding of how YesMN promotes pneumococcal spread, we identified its regulon members and the signal it uses to modulate their expression.

Previous studies with adult mice demonstrated that YesMN impacted pneumococcal virulence (20, 44, 45). Our tractable infant mouse model allowed us to examine the contribution of YesMN in pneumococcal colonization, shedding, and host-to-host transmission. YesMN knockout strains had reduced shedding without impacting colonization (Fig. 1). Consequently, we determined whether YesMN impacts Spn host-to-host transmission, as our previous work had shown that reduced shedding correlates with poor transmission dynamics (15). Our transmission studies demonstrated that a ΔyesMN strain is unable to transmit, providing further evidence that reduced shedding impacts transmissibility, and highlights the importance of YesMN to pneumococcal spread to another host. Hence, our data established that YesMN-coordinated regulation of the genes in the nasopharynx is likely critical for robust Spn shedding and transmission.

To gain insight into the genes within the YesMN regulon that could influence the virulence and transmission phenotypes, we performed RNA-seq with a serotype 4 WT (TIGR4) and isogenic ΔyesMN strains. We observed that transcripts of genes involved in Zn²⁺ homeostasis, N-linked glycan metabolism, carbohydrate metabolism, and purine metabolism were downregulated in the ΔyesMN strain (Table 1). This data aligns well with the transcript changes published by Andreassen et al. (20) using the genetically distinct D39 strain (serotype 2). Although Spn isolates are genetically diverse with over 90 serotypes, the genes within the YesMN regulon are conserved across Spn isolates (11, 20). Even with this genetic regulon conservation, it is possible that there are small genetic differences between the two isolates tested. An important caveat to this study is that the RNA-seq was performed on bacteria grown in TSY, which is likely to be nutritionally dissimilar to the nasopharyngeal environment. However, several of the genes identified in this in vitro screen have been previously identified as essential for either colonization or shedding (16, 18, 44), suggesting that the TSY growth medium is not inappropriate for such analyses.

In our RNA-seq analysis, we identified genes that are involved in zinc homeostasis, which we verified through subsequent qRT-PCR and illustrated that YesMN contributes under Zn²⁺ chelation conditions to the upregulation of zinc homeostasis and glycan metabolism genes. Trace metals, including zinc, are critical for the enzymatic function of both eukaryotic and prokaryotic cells (46). Zinc deficiency, modeled in murine and human studies, is associated with the reduced phagocytic activity of polymorphonuclear (PMN) cells and degradation of the mucus layer, which leads to increased disease severity and bacterial burden (47–52). Likewise, an increase in dietary zinc is concomitant with a reduction in invasive pneumococcal disease (53–55).

In keeping with the critical role of zinc, the host modulates Zn²⁺ levels in the blood and respiratory tract in response to pneumococcal infection (nutritional immunity) (50, 56). In this dynamic microniche environment where Zn²⁺ concentrations can fluctuate, Spn must strike a delicate balance between Zn²⁺ sufficiency and toxicity, maintained by Zn²⁺ homeostasis genes (56–59). The URT is a nutritionally poor environment with low levels of Zn²⁺ where the host mucin proteins, which contain labile Zn²⁺, potentially serve as a source (60, 61). Correspondingly, Ogunniyi et al. (62) observed an upregulation in the URT, relative to in vitro, of the transcripts of the histidine triad proteins (pht). These proteins, together with AdcA and AdcaII, are required for zinc uptake (26, 28).

The HK YesM is predicted to contain a large extracellular loop allowing it to detect small molecules and ions and providing it potentially with the ability to respond to a diverse number of signals (8). Interestingly, when we used TPEN, a high-affinity chelator of Zn²⁺, to mimic the zinc starvation that Spn likely encounters in the URT, we observed a YesMN-dependent regulation of the pht, adcaII, and glycan metabolism genes (Fig. 4A and B). Thus, our data suggest that YesMN coordinates the expression of these genes, resulting in enhanced shedding and subsequent transmission of Spn. Surprisingly, we identified pcpA to be a member of the YesMN regulon. In contrast to the adcaII and the pht genes, whose transcripts are upregulated under reduced Zn²⁺ levels in the environment, pcpA expression is regulated by CodY (63) and PsaR (30). PsaR regulatory roles are dependent on divalent ion concentration. High levels of Mn²⁺ cause downregulation, whereas high levels of Zn²⁺ lead to upregulation of pcpA (64). Generally, there are higher levels of Mn²⁺ in the nasopharynx compared to Zn²⁺ (30, 64); therefore, PsaR would likely downregulate the expression of pcpA. However, pcpA expression in the URT is higher than the blood (4), and functional antibodies have been observed in children without local or systemic infection (65), suggesting that pcpA is expressed in the URT. The opposing role of PsaR and YesMN would suggest the essentiality of pcpA expression under specific conditions, and that YesMN under reduced levels of Zn²⁺ might override the downregulation of pcpA by PsaR.

As Spn-IAV coinfections occur frequently and can result in higher morbidity and mortality (35, 36), we investigated the role of YesMN during Spn-IAV coinfection of the nasopharynx. Similar to Spn monoinfection, we observed reduced Spn shedding from pups colonized with the ΔyesMN mutant (Fig. 5A). However, in contrast to monoinfection, pups colonized with the ΔyesMN mutant coinfected with IAV had a defect in URT colonization (Fig. 5B). During Spn-IAV coinfection, Spn shedding and URT colonization density is known to increase over time, which has been attributed to an increase in nasal secretions and reduced mucociliary action (37, 38, 66). Moreover, increased Spn shedding leads to an increase in host-to-host transmission (38). These data suggest that Spn-IAV coinfection is beneficial for Spn dissemination. However, Spn impacts IAV titers and host-to-host transmission negatively (67, 68). Furthermore, IAV infection leads to the dispersal of Spn biofilms and modulation of its transcriptome (69). Collectively, these data suggest a complex, possibly antagonistic relationship, where IAV infection changes the nasopharyngeal landscape, to which Spn has to respond dynamically. Our results further build on these data, where the YesMN regulon provides Spn with an advantage in the dynamic environment of the URT during Spn-IAV coinfection, where N-glycan metabolism and other sugars cleaved from glycoconjugates for metabolism possibly allows for an increase in Spn density in the nasopharynx. Additionally, our hNF binding assay suggests that YesMN-regulated factors contribute toward the ability to escape nasal secretion and not be cleared by the host. Thus, it is likely that YesMN-dependent changes in gene expression allow Spn to navigate the inflamed URT, where N-linked glycan metabolism and other factors allow Spn to thrive and escape in high numbers to transmit to another host, although we cannot rule out other effects of YesMN-dependent factors involved in N-linked glycan metabolism, such as providing a carbon source as a nutrient for growth.

In conclusion, our studies have enhanced our understanding of one of the least characterized TCSs of S. pneumoniae. We demonstrate that YesMN contributes toward Spn shedding and host-to-host transmission during pneumococcal monoinfection. We attribute reduced shedding of the ΔyesMN mutant to an inability to escape the mucus that lines the epithelial layer of the URT, suggesting that YesMN encodes factors critical for Spn-mucus interactions. These YesMN-dependent factors include surface proteins required for zinc homeostasis and glycan metabolism. Furthermore, we provide evidence for antagonistic interaction during Spn-IAV coinfection, where YesMN promotes the ability of Spn to thrive in the inflamed URT cocolonized with IAV. Elucidation of specific YesMN-dependent factors that impact the ability of Spn to persist and thrive in the inflamed URT would provide further mechanistic insights into bacterial colonization and dissemination.

MATERIALS AND METHODS

Ethics statement.

This study was conducted according to the guidelines outlined by National Science Foundation animal welfare requirements and the Public Health Service Policy on Humane Care and Use of Laboratory Animals (70). All animal work was done according to the guidelines provided by the American Association for Laboratory Animal Science (AALAS) (71) and with the approval of the Wake Forest Baptist Medical Center Institutional Animal Care and Use Committee (IACUC). The approved protocol number for this project is A20-057.

Growth conditions and strain construction.

Strains used in the current study are listed in Table 2. Pneumococcal strains were grown at 37°C in tryptic soy broth (Becton, Dickinson [BD]) supplemented with 0.3% yeast extract (TSY). The strains in this study are derivatives of AZ57, which is a streptomycin-resistant derivative of the type 4 strain TIGR4 (24). Once Spn cultures reached the desired optical density at 620 nm (OD₆₂₀), they were washed, and diluted in sterile phosphate-buffer saline (PBS) for intranasal inoculation and other assays. To quantitate S. pneumoniae, serial dilutions were routinely plated on TSY agar-streptomycin (200 μg/mL), supplemented with either catalase (6,300 U/plate; Worthington Biochemical Corporation) or 5% defibrinated sheep’s blood, and incubated overnight at 37°C with 5% CO₂. The yesM, yesMN, and yesN deletion mutants were constructed using the Janus cassette (72) in a two-step process. Briefly, the Janus cassette was amplified from genomic DNA isolated from strain AZ59 (15) using the MasterPure DNA purification kit (Lucigen), with flanking regions (~1 kb) upstream and downstream of the gene of interest using isothermal assembly. Strain AZ57 was then transformed with the PCR product and the transformants selected on TSY agar with kanamycin (125 μg/mL). The resulting mutant was confirmed through PCR and selected for making an in-frame clean gene deletion. The mutant was transformed with PCR product using the primers listed in Table S2 in the supplemental material. All in-frame clean gene deletions contain the first and last 5 amino acid coding sequence of the gene of interest. The corrected mutant yesMN⁺ was constructed using primers (YesMN-US-1/YesMN-DS-6) on genomic DNA from AZ57, and the resultant PCR product was transformed into strain AZ83; transformants selected on TSY agar-streptomycin (200 μg/mL). We performe qRT-PCR (described below) to rule out potential downstream polar effects of the gene deletions. Comparing transcript levels of the genes immediately downstream of yesMN did not identify any differences between the WT and the ΔyesMN strain, suggesting the deletion of yesMN did not impact transcription (Fig. S4).

TABLE 2.

Strains used in the study

Strain	Descriptiona	Reference or source
AZ57	P2406; (TIGR4SmR)	24
AZ59	P2423; TIGR4 ply::kan-rpsL+; KmR, SmS	15
AZ69	AZ57; yesM::kan-rpsL+; KmR, SmS	This study
AZ81	AZ69; ΔyesM; SmR	This study
AZ83	AZ57; yesMN::kan-rpsL+; KmR, SmS	This study
AZ84	AZ83; ΔyesMN; SmR	This study
AZ85	AZ57; yesN::kan-rpsL+; KmR, SmS	This study
AZ86	AZ86; ΔyesN; SmS	This study
AZ88	AZ83; yesMN+, SmR	This study

^aKm, kanamycin; Sm, streptomycin.

Infant mouse infection model for shedding, colonization, and transmission.

All mouse studies were performed with C57BL/6J specific pathogen-free (SPF) mice that were initially obtained from the Jackson Laboratory (Bar Harbor, ME), then bred and maintained in the animal facility at Biotech Place, Wake Forest Baptist Medical Center. The pups were housed with their dam (mother) for the duration of the study and gained weight like uninfected animals.

Pups were infected as described previously (24). Briefly, at day 4 of life, the pups were inoculated intranasally (i.n.) with ~2,000 CFU of S. pneumoniae suspended in 3 μL of PBS. Postinoculation, shedding was determined by gently tapping the nares of pups (20 taps per pup) on to a TSY agar plate supplemented with either streptomycin (200 μg/mL) or kanamycin (125 μg/mL) to only allow for the growth of Spn. The nasal secretions were spread using a sterile cotton-tipped swab, and the plates were incubated as described above. Because of inherent variability in shedding from day-to-day, the CFU values obtained from individual pups were pooled for the 5-day duration of the study. To control for environmental and microbiome variability, shedding of different strains was compared for pups within the same litter. The mouse-adapted influenza A virus PR8 was used for coinfection studies. Five days postinoculation with Spn, pups were challenged i.n with 2× 10⁴ PFU of PR8 suspended in 3 μL of PBS. Daily shedding was quantified as above, but because of higher shedding values, only 10 taps per pup were carried out.

To determine colonization density, pups were euthanized at the designated time points by CO₂ asphyxiation followed by cardiac puncture. The upper respiratory tract (URT) was exposed and lavaged with 200 μL of sterile PBS from a needle inserted into the trachea, and fluid collected from the nares. The limit of detection for Spn in lavage fluid was 33 CFU/mL.

Pneumococcal monoinfection transmission studies were performed as previously described (24). Briefly, at day 4 of life, half of a cohort of pups were randomly selected to be inoculated with the indicated Spn strain. These index mice were then returned to their uninfected littermates (contact pups) and cohoused for 10 days postinoculation. To determine transmission from the index pups, all pups were euthanized at the age of 14 days, and nasal lavage fluid was collected and serially diluted on TSY agar plates supplemented with appropriate antibiotic.

RNA-seq sample preparation and analysis.

Spn cultures were grown as described above to an OD₆₂₀ of ~0.4, and RNA isolated using the Qiagen AllPrep bacterial DNA/RNA/protein kit following the manufacturer’s protocol. Purified RNA samples were analyzed for quality by electrophoretic tracing (Agilent Bioanalyzer). The Illumina TruSeq Stranded Total RNA LT kit with bacterial Ribo depletion probes (Illumina) were used to construct cDNA libraries from samples with RNA integrity (RIN) values equal to 10. Briefly, 500 ng of total RNA was rRNA depleted, fragmented, reverse-transcribed to cDNA, then purified via AMPure XP magnetic beads. Sequencing adaptors were then ligated to end-repaired fragments, and the libraries were preamplified by PCR and quantified on a Qubit 3.0 (Thermo Fisher, USA). Dual indexed libraries were pooled and sequenced to a target read depth of 50 M reads on an Illumina NextSeq 500 using a 75-bp single-end sequencing and a high-output flow cell. For all samples, >90% of sequences achieved >Q30 Phred quality scores (FASTQC analysis, Babraham Bioinformatics). Contaminating adapters were removed with Trimmomatic (73). Reads were aligned to the genomic reference Streptococcus pneumoniae TIGR4 using Bowtie 2 (74). Gene counts were determined using featureCounts software (75). Differentially expressed genes were identified by negative binomial modeling using DESeq2 (76) with false discovery correction (Benjamini-Hochberg). This work was performed by the Cancer Genomics Shared Resource of the Wake Forest Baptist Comprehensive Cancer Center.

RNA isolation and qRT-PCR.

Host RNA was isolated by lavaging the URT of pups inoculated with Spn or mock inoculated (PBS) at the designated time point using RLT lysis buffer (Qiagen) to obtain RNA from the epithelium. RNA was purified and cDNA generated as per manufacturer’s (Bio-Rad) instructions. qRT-PCR was performed using Bio-Rad SYBR green master mix, ~2.5 ng cDNA, and 0.5 mM primers per reaction mix. Primers directed toward glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used as an internal control. Samples were run on CFX384 Touch real-time PCR detection system (Bio-Rad) in duplicate, and each run was repeated twice. RNA expression was quantified using the ΔΔC_T threshold cycle (C_T) method. Primers used in the current study were previously described (77, 78).

qRT-PCR was performed on TSY broth-grown Spn cultures to confirm the genes that form the YesMN regulon. The WT and ΔyesMN strains were grown in TSY broth statically at 37°C until they reached an OD₆₂₀ of ~0.4 and (2 mL) was spun down at room-temperature at 15,500 × g for 2 min. Samples were treated with RNAprotect bacteria reagent (Qiagen), and RNA was isolated using the AllPrep bacterial DNA/RNA/protein kit (Qiagen) following the manufacturer’s protocol. cDNA and qRT-PCR were carried out as described above with the primer sets listed in Table S3.

To determine the effect of Zn²⁺ on the expression of YesMN-dependent genes, WT and ΔyesMN were grown statically in TSY broth at 37°C until they reached an OD₆₂₀ of ~0.3. At that point, samples were divided and either treated with Zn²⁺-chelating agent TPEN (30 μM) or untreated for 30 min at 37°C. Afterwards, RNA isolation, cDNA synthesis, and qRT-PCR were performed as described above.

Human nasal fluid binding assay.

Adherence of different strains to hNF was assessed using a solid-phase binding assay, as previously described (41, 43). Pooled nasal secretion samples from human volunteers were purchased from LeeBio (Maryland Heights, MO, USA; 991-13-S), sonicated on ice to homogenize samples, and protein amount was determined using the Coomassie Plus protein assay reagent (Thermo Scientific). Briefly, 10 μg total hNF protein (in 100 μL PBS) was immobilized on a 96-well polystyrene plate (Sarstedt REF:82.1581.001) by centrifugation (250 × g) for 3 min at room temperature and incubated overnight at 37°C. Plates were gently washed 3 times with 100 μL DMEM, blocked with 100 μL 0.1% bovine serum albumin (BSA) in DMEM for 2 h, then washed again 3 times with 100 μL DMEM. Spn strains to be tested were grown to an OD₆₂₀ of ~0.4 in TSY broth and diluted into DMEM to 2× 10⁴ CFU/100 μL, and 100 μL aliquoted into each well. The plate was centrifuged (250 × g for 3 min) and incubated at 30°C with 5% CO₂ for 2 h. The wells were then washed 19 times with DMEM, and the adherent bacteria were removed by adding 100 μL of 0.001% Triton X-100-PBS and incubating for 15 min at room temperature followed by vigorous mixing. The samples were then serial diluted and plated on TSY agar plates with antibiotics and incubated overnight at 37°C with 5% CO₂. Each strain was tested in triplicate on three different days.

ELISA for quantifying capsular polysaccharide.

A capture ELISA to determine the CPS amount of pneumococcal isolates was performed as described previously (17). Briefly, Immulon 2HB plates (Thermo Scientific) were coated with serotype 4-specific rabbit antiserum (Staten Institut, Denmark) at a dilution of 1:5,000, and incubated overnight at room temperature. The next day, strains to be tested were grown to OD₆₂₀ of 0.5 and resuspended in PBS. Samples were sonicated on ice for 30 s in total (5 s on and 15 s off) using the Branson Sonifier 250 (Duty cycle 40%, output control 2). The antibody-coated plate was washed 5 times with wash buffer (10 mM Tris, pH 7.4, 0.02% NaN₃, 150 mM NaCl, 0.05% Brij), and 100 μL of 2-fold serially diluted bacterial lysate was aliquoted and incubated for 2 h at RT. The cell lysate was removed, wells washed 3 times with wash buffer, incubated with 1% BSA in PBS for 1 h, and washed again 3 times. Then, 100 μL of CPS antibody (1:3,000 in 1% BSA/PBS; monoclonal against serotype 4 CPS, kindly provided by Moon Nahm, UAB) was applied and incubated for 24 h with gentle agitation at 4°C. The plates were then washed 5 times with wash buffer and incubated with goat antimouse antibody conjugated to alkaline phosphatase (Sigma-Aldrich; A3688-1 MI; 1:10,000 1% BSA/PBS) for 2 h at room temperature. The plate was washed 5 times before being developed with phosphatase substrate p-nitrophenyl phosphate (PNPP) (Thermo Scientific, product number 34047) for 20 min and read at OD₄₀₅ on a Biotek H1 synergy plate reader.

Statistical analysis.

All statistical analyses were performed using GraphPad Prism (version 9.0) software (GraphPad Software, Inc., San Diego, CA). Unless otherwise specified, statistical differences were calculated using the Mann-Whitney U test (comparing two groups) or the Kruskal-Wallis test with Dunn’s postanalysis (comparing multiple groups).

Data availability.

The RNAseq data generated and statistical analysis performed are available in Table S1 in the supplemental material.