StkP- and PhpP-Mediated Posttranslational Modifications Modulate the S. pneumoniae Metabolism, Polysaccharide Capsule, and Virulence

ABSTRACT Pneumococcal Ser/Thr kinase (StkP) and its cognate phosphatase (PhpP) play a crucial role in bacterial cytokinesis. However, their individual and reciprocal metabolic and virulence regulation-related functions have yet to be adequately investigated in encapsulated pneumococci. Here, we demonstrate that the encapsulated pneumococcal strain D39-derived D39ΔPhpP and D39ΔStkP mutants displayed differential cell division defects and growth patterns when grown in chemically defined media supplemented with glucose or nonglucose sugars as the sole carbon source. Microscopic and biochemical analyses supported by RNA-seq-based global transcriptomic analyses of these mutants revealed significantly down- and upregulated polysaccharide capsule formation and cps2 genes in D39ΔPhpP and D39ΔStkP mutants, respectively. While StkP and PhpP individually regulated several unique genes, they also participated in sharing the regulation of the same set of differentially regulated genes. Cps2 genes were reciprocally regulated in part by the StkP/PhpP-mediated reversible phosphorylation but independent of the MapZ-regulated cell division process. StkP-mediated dose-dependent phosphorylation of CcpA proportionately inhibited CcpA-binding to Pcps2A, supporting increased cps2 gene expression and capsule formation in D39ΔStkP. While the attenuation of the D39ΔPhpP mutant in two mouse infection models corroborated with several downregulated capsules-, virulence-, and phosphotransferase systems (PTS)-related genes, the D39ΔStkP mutant with increased amounts of polysaccharide capsules displayed significantly decreased virulence in mice compared to the D39 wild-type, but more virulence compared to D39ΔPhpP. NanoString technology-based inflammation-related gene expression and Meso Scale Discovery-based multiplex chemokine analysis of human lung cells cocultured with these mutants confirmed their distinct virulence phenotypes. StkP and PhpP may, therefore, serve as critical therapeutic targets.

candidate regulators control transcription of the cps locus genes. These include CcpA, a GntR family transcriptional regulator SPD_0064, a MarR family transcriptional regulator SPD_0379, DNA-binding protein HU, CodY, GlnR, and RitR (51). Although CcpA ensures optimal metabolic fitness in S. pneumoniae (53), its role in modulating capsule production, metabolic fitness, and ultimately virulence of S. pneumoniae through StkP activity is presently unknown.
In the present investigation, we have created and characterized the type-2 encapsulated pneumococcus D39 strain-derived mutants lacking StkP or PhpP. We have demonstrated that StkP and PhpP-mediated reversible phosphorylation reciprocally regulate genes encoding the entire capsule operon and corresponding expression of the capsule. Further, this regulation is mediated via the reciprocal StkP/PhpP-mediated reversible phosphorylation status of catabolite control protein (CcpA), resulting in differential regulation of the transport and metabolism of nonglucose sugars. The latter plays a vital role in forming building block sugar components of the pneumococcal polysaccharide capsule and critical sugars required for colonization and infection (53)(54)(55)(56). Biochemical, proteomic, and global transcriptome analyses of D39DStkP and D39DPhpP mutants have revealed that StkP and PhpP independently, together, and reciprocally modulate the expression of several bacterial genes. Further, NanoStringand Meso Scale Discovery Electrochemiluminescence (MSD)-based inflammatory genes transcriptome analyses and multiplex cytokine/chemokine analyses of human lung cells cocultured with the pneumococcus mutants establish that StkP and PhpP modulate host inflammatory responses, and thus together, they play a pivotal role in pneumococcus virulence and pathogenesis.

RESULTS
Phenotypic characteristics of D39DPhpP and D39DStkP mutants and their corresponding complemented strains. In the present study, we used the markerless D39DPhpP mutant, as previously reported (36). D39DStkP was created with aad9 (spectinomycin resistance) as the differential resistance marker. Genome sequence analysis of the D39DStkP mutant versus D39 wild type did not show any changes in the sequence except the absence of stkP. On the other hand, the D39DPhpP mutant grown in the THY medium showed the deletion of one nucleotide, G, between residues number 1995571 and 199572 (in the 4th passage of the mutant but not in the 1st passage). This mutation was found to be localized in the cbpA/spd_2017 gene spanning between nucleotides 1995045 and 1997150, translating into truncation of the last 56 amino acids and a loss of the last repeat region of the CbpA protein (Fig. S1). Both mutants grew slower in C1Y chemically defined medium (CDM). The wild-type gene complemented D39DStkP and D39DPhpP mutant strains regained growth patterns akin to the wild-type strain (Fig. 1A). Western blot analysis of the whole-cell lysates of D39DPhpP and D39DStkP mutants revealed the absence of the expression of PhpP and StkP and the corresponding presence of StkP and PhpP, respectively (Fig. 1B). In a similar assay, the phpP-and stkP-complemented mutant strains (D39DPhpP::phpP and D39DStkP:: stkP) showed the restoration of the expression of the corresponding encoded proteins confirming the mutant integrity (Fig. 1B). These mutants and complemented strains were then subjected to various biological and biochemical assays to understand the contribution of the PhpP and StkP proteins to different metabolic, regulatory, and virulence properties of the encapsulated strain of S. pneumoniae D39, as described in subsequent sections.
Phenotypic characteristics of pneumococcal strains lacking StkP have been extensively studied (28,31). Characteristic features of D39DPhpP mutants were elongated shapes with significant variations and irregularity with multiple parallel septa resulting in a long-chain formation. Similarly, the D39DStkP mutant strain was also found to be defective cell division, showing irregular septum with the appearance of ovoid morphotypes as revealed by scanning and transmission electron microscopy ( Fig. 1C and E). The length and width sizes of D39 mutants were significantly larger (P , 0.01) compared to the D39 wild-type strain. The morphometric analysis of complemented strains showed an overall trend of canonical shape and size of the wild-type strain (Fig. 1E). However, in this first report on PhpP and StkP mutants derived from the parent encapsulated strain D39, we were able to demonstrate additional essential features. As shown in Fig. 1C, transmission and scanning electron microscopy (TEM and SEM) of the wild-type D39 strain showed several distinct, irregular appendages signifying the surface-decorated capsule. However, the cell surface of D39DPhpP showed a relatively much smoother surface, indicating the reduced formation of the capsular polysaccharide. The phpP-complemented strain demonstrated the restoration of this lost function and mimicked the wild-type cell surface structure. The D39DStkP mutant, on the other hand, showed notably increased amounts of the capsule, suggesting the role of StkP/PhpP in reciprocal regulation of the capsule formation. (Fig. 1C). The differential putative capsule formation feature was further examined by immunofluorescence microscopy using an antipolysaccharide type-2 antibody that was preadsorbed with heat-killed D39-derived nonencapsulated R6 strain. The increased and decreased antibody binding to the D39DStkP and D39DPhpP mutant strains, respecticely, (Fig. 1D) strongly indicated that pneumococcal StkP and PhpP have a role in the modulation of capsule production.
PhpP and StkP reciprocally regulate capsule formation in S. pneumoniae. We estimated the amount of polysaccharide capsules extracted from an equal optical density (OD) volume-equilibrated pneumococcal culture suspensions to confirm the microscopic findings. Qualitative analysis of the polysaccharide capsule in Western immunoblots assays using the anti-type-2 capsule antibody revealed significantly decreased capsule-specific reactivity in the D39DPhpP mutant relative to the wildtype D39 strain and the phpP-complemented DPhpP strain ( Fig. 2A). The D39DStkP mutant, on the other hand, showed a substantially increased amount of polysaccharide capsule, which was restored in the stkP-complemented DStkP mutant to the wild-type level (Fig. 2B).
We further substantiated these findings by examining rhamnose-sugar contents quantitatively in each capsule extract ( Fig. 2C and D). Compared to R6 (negative capsule derivative of D39 strain and negative control), the wild-type strain showed the presence of rhamnose content in the range of 2 to 2.5 mg/mL. Furthermore, the results obtained from the mutants showed significantly decreased (P , 0.0001) and increased (P = 0.039) rhamnose contents in the capsule extracted from D39DPhpP ( Fig. 2C) and D39DStkP (Fig. 2D) mutants, respectively, compared with the wild-type D39 and their respective complemented strains. Together, these data indicated that the StkP-and PhpP-mediated reversible phosphorylation contributed to decreasing and increasing capsule production in pneumococci through a specific mechanism.
StkP/PhpP-modulated capsule formation is independent of the MapZ-regulated cell division in S. pneumoniae. MapZ or LocZ protein, as a crucial cell division protein, is involved in proper septum placement in S. pneumoniae (30,31) by regulating its binding to peptidoglycan and other accessory proteins (57,58). Similarly, peptidoglycan and capsule biosynthesis are coordinately regulated in Gram-positive bacteria (59). StkP/PhpP-mediated reversible phosphorylation of MapZ plays a specific role in the recruitment of FtsZ at the septum (25,31,58). This event initiates the assembly of the Immunofluorescence microscopy was performed using type-2 specific antibody preadsorbed with heat-killed S. pneumoniae nonencapsulated D39-derived R6 strain and corresponding Cy3-labeled antirabbit IgG conjugated antibody. (E) Quantitative morphometric analysis determining lengths, widths, and number of septa present in different pneumococcal strains. N = number of fields, each representing an average of randomly selected 4 to 6 bacteria. Statistical analysis as indicated in the text. P value ,0.05 was treated as a significant difference. mutants, respectively, using type-2 polysaccharide capsule-specific antibody as described above for IF microscopy. The respective complemented strains show the wild-type capsule formation in each. Quantitative estimation of rhamnose contents in the polysaccharide capsule extracted from the (C) D39DPhpP mutant and (D) D39DStkP mutant strains and their comparison to the parent wild-type strain and their corresponding complemented strains. Each error bar represents an average (6 SD) quantity of rhamnose detected in the polysaccharide extracted from three individually grown cultures. P , 0.05 was treated as a significant difference and was calculated using GraphPad Prism 6. S. pneumoniae R6 strain (nonencapsulated) was used as a negative control. (E) The left panel is Coomassiestained SDS-PAGE gel and serves as a loading control for the duplicate right panel representing Western immunoblot-based qualitative analysis of polysaccharide capsule in D39-derived single DStkP, DPhpP, DMapZ, and DPhpP mutants and DMapZ-DStkP, DMapZ-DPhpP, and DStkP-DPhpP double mutants, using type-2 polysaccharide capsule-specific antibody as described above for IF microscopy.
other division proteins at the septum of nonencapsulated pneumococcus (31). Hence, we investigated whether the MapZ-regulated cell division process indirectly influenced the altered capsule formation in the D39DPhpP and D39DStkP mutant in the encapsulated D39 pneumococcus strain. To address this, we created a D39DMapZ mutant strain by replacing the mapZ gene with the promoterless kanR gene employing the allelic replacement method. The genetically confirmed D39DMapZ mutant showed expected cell division defects (Fig. S2), as previously demonstrated in the nonencapsulated pneumococcus strain (28)(29)(30). However, this mutant with intact StkP and PhpP did not reveal any difference in the capsule formation. We, therefore, surmised that the absence of the peptidoglycan-associated MapZ did not influence differential capsule formation by cell wall-associated StkP and cytoplasmic PhpP and, thus, did not interfere in the capsule biosynthesis in pneumococcus.
Since StkP and PhpP are also involved in cell division, we investigated further to determine to what extent MapZ influences the modulatory activity of StkP and PhpP in capsule formation. To that end, we created double mutants, D39DMapZ-DStkP (kan R , aad9 R ), D39DMapZ-DPhpP (kan R ), and D39DStkP-DPhpP (aad9 R ), as described in the Materials and Methods. All mutants, including D39DStkP-DPhpP, showed slow growth (maximum OD 0.3 to 0.4) associated with cell division defects in the initial passage. However, the growth patterns of all mutants except D39DStkP-DPhpP improved (OD 0.8 to 0.9) when the starting inoculum for each of these strains was used from the consecutively multiple-passaged culture, as described in the later section. Recently, similar observations have been reported (27,40) for StkP/PhpP mutants explaining that multiple passages of StkP and PhpP mutants accumulate suppressor mutants, which do not influence cell division phenotype but are responsible for enhanced growth. Our limited genome analysis of the D39DPhpP mutant after four passages concurs with this notion. Most mutants showed loss of chains and wild-type ovo-coccoid forms and predominance of diverse sizes of aggregated forms of round or deformed shape of coccoid structures resulting from the unevenly arranged septa formation. (Fig. S2D).
To determine the modulation of polysaccharide capsule formation in these mutants, whole-cell lysates obtained from each mutant at their normalized total protein concentration (1 mg/mL) were subjected to Western blot analysis using the antitype-2 polysaccharide antibody. The result showed a significant decrease in capsule formation in the mutant lacking PhpP. Thus, D39DPhpP, D39DMapZ-PhpP, and D39DStkP-PhpP showed negligible amounts or the absence of polysaccharide capsules. At the same time, capsule contents in mutants lacking StkP and/or MapZ were more than or equal to the wild-type D39 strain, confirming that the polysaccharide capsule formation is independent of MapZ-associated cell division defect (Fig. 2E, Fig.  S2C).
PhpP and StkP differentially influence the expression of a variety of genes and reciprocally regulate the transcription of capsule-related genes. Since the relative activity of these enzymes changes based on the intracellular metabolic environment and cell wall stress (27,60,61), the primary focus of this study was to dissect the molecular and cellular bases of regulatory functions of PhpP/StkP in encapsulated pneumococcus besides its role in cytokinesis. Thus, to better understand the impact of StkP/PhpP-mediated reversible phosphorylation in the transcriptional regulation of genes in pneumococcus, we first examined global mRNA expression levels of genes in D39 wild-type, and isogenic D39DPhpP and D39DStkP mutant strains employing Illumina-based RNA-seq transcriptome analysis. This analysis was conducted using high-quality DNA-free total RNA extracted from cultures grown to the mid-log-phase. The data of cDNA libraries of an average of 160 bp inserts and their paired sequence data derived from nine samples (three samples each of the wild type and two mutants) were then analyzed and deposited in the GEO database (GEO accession no. GSE113337). At least 1.5 million reads with more than 90% of the entire genome were resolved for each sample using SOAP aligner/SOAP2 alignment software tools (62) and BOWTIE (63). Based on our previously used statistical analysis tools (64), we were able to obtain a robust analysis of all data with SOAP2 tools, which detected transcripts of more than 98% of genes out of a total of 2,076 genes of the D39 genome (NC_008533.1/CP000410.1) (65). The stringent criteria employed ($1.95-fold) in the present study (66) enabled us to identify with high confidence a total of 516 differentially expressed genes (DEGs) (25.6% of total 2,076 genes) in D39DPhpP (366 genes) ( Fig.  3A and B) and D39DStkP (333 genes) mutant strains ( Fig. 3A and C, Table S1). Among these DEGs, a group of 183 (50%) and 150 DEGs (45.04%) were unique to D39DPhpP (Table S2A) and D39DStkP mutants (Table S2B), respectively. PhpP and StkP independently regulated these genes. Both mutants displayed similar patterns of differential regulations for 150 genes (24 upregulated and 126 downregulated genes). (Table S3). The reciprocal regulation was observed for only 33 genes (6.39% of the total DEGs in these mutants) ( Table 1). In addition to the phpP gene (SPD_1543), the genome of S. pneumoniae D39 also consists of two additional genes encoding the metal-binding Ser/Thr phosphatase family, SPD_0539 (834 bp/277 aa) and SPD_1061 (729 bp/242 aa), which harbor conserved catalytic residues reported in PhpP. Irrespective of their sparse protein sequence identities and similarities with PhpP, the expression levels of these genes were found to be unaltered (1.15-and 1.2-fold increased transcript abundance). Further, the expression level of immediate upstream and downstream flanking genes of phpP and stkP genes also remained unaltered. Hence, the observed changes in D39DPhpP and D39DStkP are effectively due to the deletion of the phpP or stkP (spd_1543/1542) gene. The overall gene expression profiles supported by ClueGo analysis (Fig. 4) in these two mutants indicated that both PhpP and StkP participated independently, reciprocally, and together in regulating several carbohydrate transport-metabolism-, nucleotide-purine/  pyridine metabolism-, protein synthesis-, and iron transport-related genes. This differential global gene expression repertoire resulted in the observed phenotypes of S. pneumoniae described above (Figs. 1 and 2). Further, 6 of 13 gene pairs encoding two-component systems showed differential regulation in individual or both D39DStkP and D39DPhpP mutants (Table S4A). For example, the transcript abundance of rr08 (spd_0081/saeR) was upregulated only in D39DStkP (2.28-fold) but moderately in D39DPhpP (1.32-fold). Similarly, the expression of rr10 (vncS/R, spd_0525/ spd_0524) was downregulated (;3.0-fold) only in D39DPhpP but not in D39DStkP (;1.0 to 1.2-fold). The remaining four two-component systems (rr03/liaSR-spd_0351/0352, rr13/blpHR-spd_0469/0468, rr05/ciaRH-spd_0702/0701, and rr12/comDE-spd_2064/2063) were severely downregulated (2-to 31-fold) in both mutants (Table S4A).
Concerning reciprocally regulated genes in these two mutants (Table 1), the differential regulation was essentially associated with the genes responsible for capsule production, carbohydrate metabolism, and iron/ferrichrome transport (Table 1). Particularly, while the transcript abundance of several genes of the capsule (cps) operon (spd_0315 to spd_0331) was significantly increased in D39DStkP, the entire 17 cps genes displayed decreased transcript abundance by ;9 to18-fold (Table 1) in the D39DPhpP mutant. (see also Fig. 5D). Furthermore, the qRT-PCR analysis of capsule genes (cps2A, cps2E, and cps2L) also revealed decreased and increased mRNA transcript abundance in D39DPhpP and D39DStkP mutants, respectively, confirming that PhpP and StkP reciprocally regulate the transcription of capsule-related genes. Similarly, a matching transcript abundance profile for an additional eight genes by qRT-PCR validated RNA-seq analysis in general ( Table 2).
PhpP/StkP positively influences pneumococcal carbohydrate metabolism and contributes to metabolic fitness. Pneumococci use glucose as the primary carbon source for growth and energy. In addition, S. aureus Stk1-mediated phosphorylation of CcpA   Table S1 to S4 for detail).
StkP-and PhpP-Modulated Pneumococcal Capsule and Virulence Infection and Immunity modulates CcpA-regulated carbohydrate metabolism (67), a master regulator of carbohydrate metabolism (68). Since the pneumococcus CcpA is essential to ensure optimal metabolic fitness in S. pneumoniae (53), we hypothesized that the mutant strains lacking StkP or PhpP would display altered growth patterns when grown in different sugars. The reduced growth of D39DPhpP mutant in the presence of nonglucose sugars in the initial passage indicated that the uncontrolled StkP-mediated phosphorylation in the absence of PhpP might adversely affect the bacterial ability to grow in the presence of nonglucose sugar (Fig. 5A). Noticeably and as reported previously for nonencapsulated pneumococcal strain grown in the glucose-containing medium (69), different pneumococcal mutants, including double mutants lacking MapZ-StkP, MapZ-PhpP, and StkP-PhpP showed improved growth after 4 to 5 sequential passages (Fig. 5B). However, the growth pattern of D39DStkP-DPhpP in the presence of glucose and different nonglucose sugars did not improve significantly ( Fig. 5A and B). Transcriptome analysis of D39DStkP and D39DPhpP mutants for "carbohydrate metabolism and transport" genes revealed downregulation of most genes belonging to the lactose, galactose, fructose, cellobiose, maltodextrin, mannose, mannitol, and fucose metabolism and phosphotransferase system (PTS) (Fig. 5C, Table S4B), supporting the observed growth retardation of the mutants in nonglucose sugars. Noticeably, the expression of the gene encoding CcpA (catabolite control protein), the master regulator of sugar metabolism (53,68), and its coregulator, HPr, and HPr kinase/phosphorylase did not show any significant changes in D39DPhpP and D39DStkP mutants (ccpA/spd_1797, 21.25 DPhpP /-1.46 DStkP -fold; ptsH/hpr/ spd_1040, 21.13 DPhpP /21.67 DStkP -fold; and hprK/P/spd_1244, 21.17 DPhpP /1.04 DStkP -fold). Based on these results showing modulated functional expression of several genes, including those encoding the proteins involved in the carbohydrate metabolism of D39DStkP and D39DPhpP mutant strains, we hypothesized that posttranslationally modified CcpA could be responsible for the observed changes.
StkP and PhpP reversibly phosphorylate CcpA. The above results and published reports showing the possible direct and indirect role of CcpA in influencing the expression of pneumococcal capsule genes (51,53,70), prompted us to examine whether StkP/ PhpP-mediated reversible phosphorylation exerts its effects via posttranslational modification of CcpA. In the radioactivity-based in vitro phosphorylation assays, StkP and PhpP reversibly (phosphorylation and dephosphorylation, respectively) phosphorylated recombinant CcpA (Fig. 6A). Additionally, to determine the in vivo phosphorylation of CcpA by StkP, the kinase domain of StkP (StkkP) and His-tagged CcpA were expressed in the pCDFDuet vector and D39DCcpA::his-ccpA complemented strain. The in vivo phosphorylated purified, and SDS-PAGE resolved CcpA protein bands were subjected to mass-spectrometry analysis and compared with nonphosphorylated and in vitro StkP-phosphorylated (using cold ATP) CcpA. The comparative analysis of in vitro (using cold -ATP) and in vivo phosphorylated CcpA proteins revealed Ser19 (In vivo and in vitro), Thr 22 (in vitro and in vivo), and S238 (in vitro, with low probability) as three major phosphosites in phosphorylated CcpA (Fig. 6B). Subsequently, three variant forms of the recombinant CcpA protein, CcpAS19A, CcpAT22A, and CcpAS238A, were subjected to in vitro phosphorylation by StkP. Each of these variant proteins showed substantially decreased phosphorylation ( Fig. 6A), indicating that all three residues contribute to the StkP-mediated CcpA phosphorylation. StkP-phosphorylated CcpA does not efficiently bind to the cps2A promoter. The reciprocal expression levels of the polysaccharide capsule, along with the corresponding transcript abundance of cps2 genes in D39DPhpP and D39DStkP mutants and the ability of CcpA to bind to the cre locus within the cps2A promoter (51,70,71) suggested that the StkP-mediated phosphorylation of CcpA might adversely affect this binding. Upon examining the CcpA binding to the 32 P radiolabeled cps2A promoter by EMSA, we observed that CcpA interacted with the cps2A promoter and shifted its migration dose-dependently (1 to 3 mM), as reported previously (70). However, the in vitro phosphorylated CcpA in the presence of increasing amounts of StkkP (0.5 to 3.0 mM) displayed a dose-dependent reduced binding to the Pcps2A probe, and proportionately increased amounts of the nonmigrated labeled promoter probe (Fig. 7). These results thus indicated that the binding of CcpA to the Pcps2A promoter and the observed differential transcript abundance of capsule operon genes and corresponding capsule production in D39DStkP and D39DPhpP mutants were inversely proportional to the StkP-mediated CcpA phosphorylation.
PhpP and StkP play an important role in the modulation of S. pneumoniae virulence. While the pneumococcus capsule is known to be the major virulence factor, several other virulence factors also play a crucial role in pneumococcal adherence to and invasion of nasopharyngeal cells and hence in the virulence (12,14,72). Along with differential transcript abundance of capsule biosynthesis-related genes, seven genes (blpU, strH, pspA, endoD, nanA2, nanA, and spd0335) of the 16 differentially regulated potential virulence genes in the D39DStkP and D39DPhpP and virulence-regulating two-component systems were downregulated (Table 1, Table S4A). These results prompted us to investigate whether differentially regulated virulence genes participated in pneumococcal virulence.
To that end, we examined the virulence potential of D39DPhpP and D39DStkP mutants compared to their parental wild-type and the complemented strains employing mouse infection models for septicemia and nasopharyngeal colonization/lung infection. The results of the septicemia model revealed the complete attenuation of virulence in the D39DPhpP mutant strain as all mice survived for 10 observation days (Fig. 8A) and even when observed after 3 weeks postinfection (data not shown). In contrast, mice infected similarly with D39DStkP displayed attenuation with only 50% mortality, possibly due to the increased production of the capsule (Fig. 7B). Mice infected similarly with the wild-type and the phpP-or stkP-complemented strains died within 2 The diagram shows that the cps2 operon contains 17 genes and an upstream promoter region located between dexB and aliA genes. The functional nature of promoter region encompasses transcription terminator (TT), insertional element (ISE), repeat region RUP, spacing sequence (SS), and core promoter (CP). (B) Electrophoretic mobility shift assay (EMSA) for StkP-phosphorylated and nonphosphorylated CcpA was conducted using cps2A as the binding probe. Left panel, EMSA was performed using a 32 P-labeled fragment (-250/135) with an increasing (1 to 3 mM) amount of CcpA. Right panel, EMSA was performed using a 32 P-labeled probe, the promoter probe. A sequence of the 59 upstream (-250/135) region of cps2A was used with a constant amount of CcpA and increasing amounts of StkP (0.5 to 3 mM, lanes 4 to 9). Lanes 1 and 2, EMSA was performed using a 32 P-labeled probe (10,000 dpm) in the absence (lane 1) or presence (lane 2) of 100Â excess amount of cold probe as a competitor to confirm the specific DNA/protein complex. Electrophoresis of the 32 P-labeled probe alone is shown in the first lane of each panel. Results represent three similarly conducted independent experiments. B* represents bound form with retarded migration of the labeled probe. F* represents unbound, fast-migrated, free probe. (C) Impact of the repressor activity of StkP-mediated phosphorylation on the transcription of cps2 genes and the observed regulation of the polysaccharide capsule formation. to 5 days postinfection ( Fig. 8A and B). A similar pattern of virulence attenuation was also observed when the mice were infected by the intranasal route using 2.82 log 10 higher inoculum (5 Â 10 7 I/N versus 7.5 Â 10 4 I/V CFU/animal) of D39DPhpP (Fig. 8C) and D39DStkP mutants (Fig. 8D). However, ;47% of infected mice (7 [D39DStkP] and 8 [D39DPhpP] out of 17 mice) belonging to the wild-type and complemented groups still survived. Irrespective of the route of infection, the inclusion of two different infection models consistently demonstrated that the D39DPhpP mutant, which showed significant decreases in the capsule formation without displaying significant growth defects in the enriched media, was attenuated for virulence. The D39DStkP mutant, on the other hand, despite its growth defect, showed relatively more virulence during systemic infection.
Consistent with the virulence pattern described above, the median pneumococcal CFU counts measured at 24 h, and 48 h in the blood of the mice retroorbitally infected with the D39DStkP/PhpP mutant strains were found to be significantly low in comparison to those in the wild-type and corresponding complemented strain-infected mice ( Fig. 8E and F). These results indicated that the mutant strains could not survive in the blood and were phagocytosed and killed readily by innate host immune responses. In contrast, the wild-type and the complemented strains survived and multiplied ( Fig. 8E and F). We also observed a similar trend at 24 h and 48 h when we counted CFU in the blood of mice infected intranasally with these strains. However, the total number of CFU was relatively low in the mutant-infected mice, and 5 of 17 survived at the end of the observation period ( Fig. 8G and H).
Along with these findings, the Gram-staining of the tissue sections obtained from the lungs of the mice infected with the D39DPhpP mutant strain revealed the absence  of any discernible bacterial colonization of the lung tissues (Fig. S3). In contrast, the lumen and the interstitial spaces of the lung alveoli of the wild-type-and the phpPcomplemented strain-infected mice showed a profuse number of bacteria in the lumen (Fig. S3). Furthermore, the histopathology (hematoxylin and eosin/H and E staining) of the sections of lung tissues obtained from the wild-type and complemented straininfected mice also revealed thick inflamed alveolar walls compared to the lungs of D39DPhpP mutant-infected mice exhibiting thin-walled alveoli with much larger lumen space (Fig. S3). The histopathological pattern of D39DStkP and corresponding complemented strains mimicked similarly (data not shown). D39DPhpP mutant evokes distinct protective inflammatory responses in the lung epithelial cell culture. To further dissect whether in addition to the pneumococcal division process, its growth defect, and the presence or absence of the polysaccharide capsule per se, other host factors are responsible for pneumococcus virulence, we cocultured D39-WT and isogenic mutant strains D39DStkP, D39DPhpP, and D39DMapZ with A549 human lung cells for 4 h and determined the impact of bacterial-cell interactions on the expression of 255 inflammation-related genes employing NanoString technology. We examined unamplified mRNA copy numbers of 255 inflammatory genes in one sample to obtain multigene differential expression profiling. Out of 255 genes, only 48 genes showed significant difference (P , 0.05) in A549 cells while in contact with these mutants (19 with D39DStkP, 32 with D39DPhpP, and 11 with D39DMapZ) (Table S5A and B) compared with the wild-type D39-Wt strain (Fig. 9A). Upon further analysis of the results of these mutants' interaction with host cells, the heat map of these differentiated genes in A549 cells showed distinct virulence phenotypes. (Fig. 9B, Table S5B). Particularly, A549 cells cocultured with D39DPhpP in comparison to D39DStkP showed upregulated mRNA expression (1.5-to 3fold) of 9 inflammatory genes (FOS-1.5, MAFK-1.55, IL1A-1.59, IL-8-1.67, CCL2-1.74, MAFF-1.55, CCL20-2.18, IL-6-2.8, and PTGS2-3.1) (Fig. 9B, Table S5B sheet 1).
In a separate similar experiment, tissue culture supernants from cocultured A549 cells with different D39-derived mutants were collected at 0, 2, and 4 h and examined quantitatively for the secretion of IFN-g , TNF-a, and 8 interleukins by multiplex electrochemiluminescence (ECL)-based meso scale discovery (MSD) cyotkine/chemokine analysis (Fig. 9C). These results revealed significantly less amounts (P , 0.05, P , 0.01) of most chemokines in the supernatants of A549 cells cocultutred with D39DPhpP and D39DStkP versus D39-WT and D39DMapZ collected after 2 h of incubation. The double mutants lacking MapZ and StkP or PhpP, and lacking StkP and PhpP showed similar low expression levels at 2 h. The significantly dampened critical inflammatory chemokine responses against D39DPhpP and D39DStkP in comparison to D39DMapZ and D39-WT supports the observed virulence attenuation of D39DPhpP mutant and reduced virulence of D39DStkP in comparison to D39-WT (Fig. 8).
PhpP and StkP together, thus, served as a critical cognate enzyme pair to control pneumococcal cell division, virulence, and metabolic fitness by direct or indirect modifications of transcription factors such as CcpA and subsequently transcription of genes required for polysaccharide capsule formation and other metabolic/transport genes. In addition, the mutants' differential cell surface structures and ability to grow in the glucose or nonglucose sugarenriched host environment also modulate the host-inflammatory genes during infection. Thus, StkP-and PhpP-mediated posttranslational modifications play a crucial role in the modulation of pneumococcal virulence and disease pathogenesis. Given this multimodal functionality, both StkP and PhpP can serve as promising therapeutic targets to counteract encapsulated and nonencapsulated pneumococcal diseases.

DISCUSSION
Pneumococcal colonization of the nasopharynx is a prerequisite for the subsequent steps involving bacterial adherence and persistence for the disease commencement. Although the polysaccharide capsule is the major virulence factor, other noncapsule factors, including several surface proteins, also have been incriminated in pneumococcal adherence, invasion, and dissemination (6). Pathogenic mechanisms underlying pneumococcal diseases, like many other Gram-positive pathogens, are thus multifactorial (6, 73, 74). Among many virulence regulators, the role of Ser/Thr kinase (StkP) also has been incriminated in pneumococcal virulence primarily as a regulator of the fundamental growth process, i.e., cytokinesis (28,39). The PhpP serves as a counteracting or balancing factor in the homeostasis of cytokinesis. However, the role of StkP and PhpP, unlike similar homolog enzymes of other Gram-positive pathogens (36,43,75,76) in other cellular activities, has not been well recognized. Published microarray analyses of noncapsulated pneumococcus-derived mutant lacking StkP has revealed differential expression of genes that encode proteins involved in cell wall metabolism, pyrimidine biosynthesis, DNA repair, iron uptake, and oxidative stress (77). In the present study, we provide an essential and first insight into the global regulatory role of PhpP and StkP in the encapsulated pneumococcus pathophysiology. Considering several advantages of RNA-seq-based transcriptome analysis (78) over the limitations of microarray analysis (79), we opted to carry out RNA-seq-based analysis of the DPhpP and DStkP mutants derived from the encapsulated pneumococcus strain D39. The major highlights of the present investigation are the revelation of the mechanism underlying the modulation of the polysaccharide capsule formation and the metabolism genes involved in the building blocks of pneumococcal capsules. (Table 1, Table S2). The additional highlights of the comparative global transcriptome analysis of D39DStkP and D39DPhpP mutants included 45 to 50% of the DEGs unique to be regulated by either StkP or PhpP, and 29% DEGs were are under the control by both enzymes. The observed direct and indirect regulations indicate that although StkP and PhpP have reciprocal biochemical functions, their physiological existence seems to be mutually exclusive as a monocomponent regulatory system and may interact with other corresponding cross-reacting counteracting kinase or phosphatase enzymes. Results obtained from the present transcriptomic study showing modulated expression of many genes could also be the direct and indirect of the reversible phosphorylation of the gene products as described recently in two phosphoproteomic analyses of DPhpP and DStkP mutants derived from the nonencapsulated S. pneumoniae strains (26,27). Although these analyses did not show the implications of the observed in vivo phosphorylation in a mechanistic form, our study complements the published phosphoproteomic analyses to quite an extent.
In the present study, transcriptome analysis of the D39DPhpP mutant versus D39-WT revealed the precisely targeted functional category of genes responsible for metabolic fitness and virulence. Pneumococcus lacks the respiratory chain or tricarboxylic acid cycle for energy production, leaving glucose as the primary source of energy (65). The ability to survive and thrive in various host niches endowed with physiologically diverse microenvironments and typically deficient in glucose and other essential cation contents reflects that the pneumococcus possesses a regulatory mechanism to utilize other sugars as an alternative carbon source (80). In the upper and lower respiratory tracts, mucins are major components of the mucus that cover the epithelial surfaces (81). Mucins are heavily O-glycosylated glycoproteins and composed of N-acetylglucosamine (GlcNAc), N-acetylgalactosamine (GalNAc), N-acetylneuraminic acid/sialic acid (neuNac), galactose, fucose, and sulfated-sugars linked to the core protein (82)(83)(84). Sialic acid also plays a dynamic role in pneumococcal adhesion and invasion and as a molecular signal for forming biofilms, nasopharyngeal colonization, and invasion (85,86). The major content of human saliva is also sialic acid (84). Galactose is a critical sugar in pneumococcal colonization and infection (56). Pneumococcus relies and thrives on these glycan-derived sugars. Pneumococcus is endowed with more than 20 phosphotransferase systems (PTS) to properly transport various complex sugars (80). Besides some sugar-specific PTS, many PTS are responsible for transporting more than one sugar, and some sugars, such as galactose, may use more than one PTS (80). Given such complexity, the ability of PhpP and StkP to influence the expression of several PTSs (Table S4B) in a similar manner indicates that StkP-PhpP, individually and as a pair, play a decisive role in this complex metabolic regulation, and ultimately in pneumococcal colonization (56). Since the key genes involved in galactose metabolism (spd_1634/galK, spd_1633/galT, and spd_1635/galR, and spd_1648/ galE) were downregulated (2 to 4-fold) (Table S1, Fig. 5A to D) in both D39DStkP and D39DPhpP mutants, the growth of these mutants in the presence of galactose could not be sustained beyond an OD of 0.2 to 0.4. The recent proteomic studies of nonencapsulated D39 strain-derived PhpP and StkP mutants also have observed differential expression and increased phosphorylation of PTS proteins, especially in galactose operon (26,27). The mechanism underlying protein phosphorylation affecting PTS gene expression is presently unknown. These results, however, support our observation that none of the mutants lacking StkP, PhpP, and/or MapZ could grow efficiently in the culture media containing nonglucose sugars, especially in the presence of galactose. Although presently unknown, the mechanism of this downregulation is an important field of future research as galactose metabolism is physiologically linked to polysaccharide capsule production (Fig. 5D).
Catabolite control protein (CcpA) enables pneumococcus (53) and many other Gram-positive pathogens (67,68,(87)(88)(89)(90) to utilize nonglucose carbohydrate sources and survive in the glucose-depleted host microenvironment. CcpA thus plays an essential role in maintaining pneumococcal metabolic fitness and virulence. It also plays an indirect role in polysaccharide capsule production in certain Gram-positive bacteria, including pneumococci (53,55,91), and partially regulates galactose metabolism (53,91). In the present study, however, we observed the downregulation of genes controlling several nonglucose sugar metabolism and transport but not that encoding CcpA. These results indicated that the functionality of CcpA, i.e., its binding ability to the "cre"-region containing promoters, might have been altered, resulting in compromised metabolic fitness in the absence of PhpP and StkP. (Fig. 7). Based on the differential expression of proteins involved in sugar metabolism, a recent proteomic analysis study of pneumococcus mutants lacking StkP and PhpP made a similar prediction (27). However, StkP in this study did not seem to target CcpA. A report showing the abrogation of the binding ability of the S. aureus Stk1-phosphorylated CcpA to the "cre" promoters, such as ccpA, citZ, tst, pckA, ald, and malR indicated that the Stk1-mediated phosphorylation in S. aureus potentially modulates the CcpA repressor activity toward the target genes (67). To understand how capsule assembly is coordinated with cell wall biosynthesis in S. aureus, a recent report has shown that S. aureus Ser/Thr kinase, PknB senses cellular lipid level II levels and negatively controls capsule synthesis based on proteomic analysis and showed increased capsule production in the absence of PknB (59). The study on LytR-CpsA-Psr (LCP) glycopolymer transferases, especially CpsA, has shown that they play an essential role in the attachment of capsule assembly to the cell wall (92). While the study on the S. aureus capsule has focused on the capsule biosynthesis purely on proteomic and biochemical levels (59), our study has highlighted that StkP and PhpP modulate pneumococcal capsule biosynthesis by reciprocal expression levels of genes responsible for the production of building blocks of the polysaccharide capsule. At present, it is not known whether CpsA or other LCP motif-containing enzymes serve as the substrate for StkP in pneumococcus or if there is cross phosphorylation/dephosphorylation between StkP/PhpP and Cps2D (BY-tyrosine kinase)/CpsB (Tyrosine phosphatase) or non-BY (bacterial Tyrosine) kinase (Ubk/ubiquitous bacterial kinase) (93) or PTPs (protein tyrosine phosphatases). This knowledge gap can be attributed to the fact that previous studies on StkP and PhpP/UbK are carried out using nonencapsulated pneumococcus strains to focus primarily on their cell division, formation of the ovo-coccoid morphology and mechanisms of cytokinesis. Such cross-phosphorylation phenomena are reported in Gram-positive bacteria such as Bacillus subtilis (94-96) and S. pyogenes (64,97). Recently, pneumococcal VncR has been reported to regulate capsule polysaccharide synthesis, although in a strain-specific manner (52). VncR is the response regulator of the vancomycin resistance locus (vncRS operon or rr10/spd_0524/0525). In the absence of PhpP, but not in the absence of StkP, both vncRS genes were 3-fold downregulated, supporting the downregulated capsule formation in D39DPhpP. Whether the regulation of polysaccharide capsule formation by CcpA through reversible phosphorylation StkP/PhpP also occurs through VncRS/RR10 TCS is presently unknown. ComE, an essential response regulator negatively regulates the expression of capsular polysaccharide locus (98). Both comED/rr12/spd2063-2064 expression is severely downregulated in D39DStKP and D39DPhpP mutants (20 to 30-fold) (Table 4SA) and thus downregulation of comE in D39DStKP may relate with increased production of polysaccharide, a similar correlation with D39DPhpP could not be made.
The complete transcription of the cps genes depends on the core promoter (CP) immediately upstream cps2A and additional elements upstream of CP extending up to the dexB gene (51). These other elements constitute (i) the insertional elements (IE), (ii) repeat sequence of pneumococcus (RUP), and (iii) the spacing sequence (SS) (Fig. 7A). Their individual roles vary in the transcriptional regulation of the cps2A gene (71). Interestingly, while the deletion of the SS element and core promoter significantly reduces the expression of the cps2A gene and the capsule formation, the deletion of the RUP region increases the expression of the Cps2A gene (71). Based on the location of the RUP and the homology between IE and RUP, it is presumed that the RUP may influence gene expression and contribute to the biology of pneumococcus (71). The RUP region thus may participate in such regulation by fine-tuning the transcription of the cps locus and CPS production by recruiting a regulatory factor (71), although the precise nature of such a regulatory factor has not been characterized. More recently, several putative regulators, including CcpA, a GntR family transcriptional regulator SPD_0064/cpsR, a MarR family transcriptional regulator SPD_0379, DNA binding protein HU, CodY, GlnR, and RitR, have been identified with the potential binding site within the promoter region of cps2 (51). The RUP region contains the "cre" binding site, and based on which, we predicted the role of the StkP-phosphorylated CcpA in the repression of the genes. As described above, VncR protein also binds the 218-bp Pcps, but only in the presence of serum in a type-specific manner (52), and positively regulates the expression of cps2 genes. It is unknown whether StkP phosphorylates VncR and modulates the VncR-mediated cps2 gene expression.
Unlike Stk1-mediated phosphorylation of S. aureus CcpA at Thr 18 and Thr 33 (67), the pneumococcus StkP in vitro phosphorylated CcpA at its S19, Thr22, and S238 residues. Irrespective of these differences, the in vitro reversible phosphorylation of CcpA by StkP and PhpP, and EMSA of the StkP-phosphorylated CcpA reveal the compromised binding ability of CcpA to the cps2A promoter and strongly support our hypothesis providing a novel regulatory mechanism of the pneumococcus capsule formation. In addition to the CcpA-mediated regulation of the capsule via binding to the cre region in the RUP segment of the core promoter, a recently identified CpsR (SPD_0064) also binds to the RUP region and suppresses capsule expression (51). The CpsR binding region is 15 bp downstream of the cre locus, i.e., 2146 to 2114 bp relative to the transcription start site (TSS) of the cps locus. In the absence of PhpP, i.e., in the background of unchecked phosphorylation activity, the transcript abundance of the spd_0064/cpsR is upregulated by ;9 folds (Table S2). Thus, the increased binding of CpsR may result in repression of the cps locus and capsule formation. However, it is unknown how StkPmediated phosphorylation modulates the expression of cpsR. A defined biochemical link between CcpA and CpsR binding to the cps2 promoter is presently unknown. The RUP forms a stem-loop structure (99). It is likely that the recruitment of phosphorylated CcpA at the cre site of RUP or the possible phosphorylation of CpsR by StkP may affect the folding of its stem-loop structure and alter the binding of CpsR to RUP and, in turn, may result in modulated transcription of the cps locus genes and capsule production.
Given this information, we surmised that the increased and decreased expression levels of the polysaccharide capsules genes and capsule contents in D39DStkP and D39DPhpP mutants is likely due to the altered and reversible phosphorylation-modulated CcpA binding to the region upstream of the cps2A gene within its promoter. However, understanding the direct impact of StkP and PhpP on translational and phosphorylation levels of proteins involved in Cps biosynthesis and their interactions with CcpA and other mediators involved in polysaccharide capsule gene transcription via Pcps in pneumococcus need further investigation.
StkP/PhpP-mediated posttranslational modifications and their regulated metabolic transport, virulence factors, and the capsule formation required for pneumococcal in vivo fitness at the mucosal level directly impact pneumococcal adherence, invasion, and evasion processes. However, the modulation of virulence manifestation is both bacteria and host-dependent (100), and like many other Gram-positive pathogens, the virulence of pneumococcus is also multifactorial (6). Thus, besides the mutants' ability to multiply and survive within the nutritionally variable and hostile host environments, the nature of the virulence phenotype of pneumococcus mutants also depends on host responses to the modulated bacterial cell surfaces. In this study, we have demonstrated that the capsule-depleted D39DPhpP mutant displays the wild-type growth pattern but remains attenuated and does not survive in both mouse infection models. D39DStkP, on the other hand, although defective in cell division and unable to divide efficiently, show relatively more virulence compared to D39DPhpP. Thus, while cell division defects may be crucial for bacterial survival in the host, the bacterium-specific innate immune responses have an equal role in virulence. To that end, we have demonstrated that division/growth defective D39DStkP, D39DPhpP, and D39DMapZ are phenotypically distinct in terms of their ability to evoke host-innate immune responses. D39DPhpP, in comparison to D39-WT, D39DMapZ, and D39DStkP, is unable to elicit important proinflammatory chemokines (Fig. 9C). Similarly, as revealed by NanoString gene expression analysis, the interaction of D39DPhpP with the host cells also increases the expression of specific protective chemokine genes (Fig. 9B, Table S5B). The increased expression of CCL2 and CCL20 prevents pneumococcus-mediated sepsis in humans and mice (101)(102)(103). Although IL-6 and IL-8 have a protective role by recruiting neutrophils and clearing pneumococcus burden during lung infection (104,105), overwhelming responses as in the case of D39-WT and comparable responses in the presence of D39DMapZ can also cause imbalance resulting in sepsis. With an increased capsule expression level per se or by curtailing the exposure of other bacterial components, the highly encapsulated D39DStkP mutant may evade this host defense mechanism by dampening the protective proinflammatory responses allowing bacterial survival through other invasive or evasive events (106). Thus, the role of StkP, apart from morphogenesis, is also in allowing bacteria to induce better surfaceexposed protein-mediated protective cytokine responses by limiting polysaccharide capsule formation. The PhpP activity, on the other hand, may increase bacterial virulence by promoting capsule formation and invoking inflammatory responses or evading protective innate immune responses.
This study thus elucidates that while StkP is essential for the cell division process in pneumococcus, its kinase activity-mediated posttranslational modification on CcpA serves as a repressor of capsule gene transcription and subsequent capsule biosynthesis. PhpP, however, seems to act as an essential counter enzyme reversing this repressor activity. StkP and PhpP, thus, play crucial roles in fine-tuning the vital cellular functions involved in pneumococcal pathogenesis and maintaining the homeostasis of capsule formation required for the pneumococcal innate immune evasion, survival, proliferation, and dissemination to distant vital organs to cause severe and often fatal diseases. PhpP and StkP can serve as important targets for developing novel therapeutics against pneumococcal infections.

MATERIALS AND METHODS
Ethics statement. This study was carried out per the guidelines outlined in the "Guide for the Care, and Use of Laboratory Animals" of the National Institutes of Health and the NC3R S recommended ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines (https://www.nc3rs.org.uk/arrive -guidelines). In addition, all animal experiments, anesthesia procedures, and early removal criteria were observed and performed following the protocol (no. 2011A00000051) approved by the Ohio State University Institutional Animal Care and Use Committee (IACUC).
Bacterial strains, cell lines, and growth conditions. The wild-type Streptococcus pneumoniae D39 (type 2) (65), D39-derived mutants, and corresponding phpP/stkP complemented strains were grown routinely at 37°C in Todd-Hewitt broth (Diffco) supplemented with 0.5% (wt/vol) yeast extract (THY) with or without antibiotics, in a CO 2 incubator as described previously (36) for maintenance. D39-derived R6 strain, which lacks the polysaccharide capsule, was used as a capsule-negative control strain (42). For pneumococcus transformation experiments and growth curve determination, bacterial cells were grown in a chemically defined medium containing 0.5% yeast extract (C1Y) (36). The antibiotics streptomycin (150 mg/mL), spectinomycin (200 mg/mL), kanamycin (200 mg/mL), and chloramphenicol (5 mg/mL) were used in various experiments during the generation of the mutant and complemented strains. Human lung adenoma epithelial cell lines A549 (CCL-185, ATCC) were cultured and maintained in DMEM medium supplemented with 10% fetal bovine serum and streptomycin-penicillin under the 5% CO 2 environment using a CO 2 incubator as described previously (36,42,64).
Generation of recombinant pneumococcal wild-type StkP, PhpP, CcpA, and variant CcpA proteins. Recombinant StkP and PhpP were obtained as described previously (36). Recombinant His-tag CcpA was produced by cloning the PCR-amplified ccpA gene between NdeI and BamHI sites within the multiple cloning sites of the pET14b plasmid (see Table S6 for primers). Four clones of pET14b plasmids containing the entire synthetic wild-type ccpA genes and their variant counterparts ccpAS19A, ccpAT22A, and ccpAS238A were custom synthesized (GeneScript, Piscataway, NJ). The plasmids were then used to transform E. coli DH5a and BL21(DE3-PlysS) strain. Recombinant wild-type and variant CcpA proteins were individually expressed. The His-tag-proteins were then affinity purified using Ni-NTA (Nickel-nitrilotriacetic acid) agarose (Qiagen) as described previously (36,64).
Generation of PhpP and StkP knockout mutants and complementation. The markerless phpPknockout mutant, previously derived from an encapsulated S. pneumoniae type 2 strain, D39 (36), was used in the present study. Briefly, the markerless pneumococcus mutant lacking PhpP was created using the Janus-cassette that contained the kanamycin (kan) resistance gene, followed by the recessive rpsL gene (Sm S ), and by employing a two-step negative selection strategy (36,107). In the first step, the Sm S -Kan R phenotype of a D39DPhpP-Janus strain containing Janus-cassette was obtained after transformation of the laboratory-generated D39-Sm R strain with a DNA construct containing the Janus cassette flanked on either side by an upstream and a downstream region of the phpP gene (36). This strain was then subjected to the second step of negative selection, and the Janus cassette was replaced with a PCR product containing the upstream and downstream regions of the phpP gene spliced together by spliced-overlap extension (SOE). The resulting DphpP mutant (phenotype Kan S /Sm R ) was selected on streptomycin-containing blood agar plates. The genetic integrity of the D39DPhpP mutant was confirmed as described previously by PCR and DNA sequencing using appropriate screening primer pairs (36).
Unlike the markerless D39DPhpP mutant, the isogenic mutant lacking StkP was created by the allelic replacement method by using the pFW6 vector, which has the spectinomycin resistance gene (aad9) flanked by two multiple cloning sites (MCS)-I and II and lacks a transcription terminator after the stop codon of aad9 (108). Briefly, the PCR-amplified 503-bp upstream sequence of stkP (primer pairs, StkP-up-F and StkP-up-R) was cloned between the SalI and restriction sites within MCS-I. Similarly, the PCR-amplified 500-bp downstream sequence (primer pairs StkP-Dwn-F and StkP-Dwn-R) was cloned at PstI restriction sites within MCS-II. The PCR-amplified product was obtained using a primer pair, StkP-up-F and StkP-Dwn-R and template pFW (stkP) to transform the competent D39-WT cells to replace the stkP gene with aad9 as described previously (36). The transformants were selected on TSA-blood agar plates containing 200 mg/mL spectinomycin. The genetic integrity of D39DStkP mutants was confirmed by PCR, and DNA sequencing using appropriate primers and custom-genome sequencing (BGI Americas). In addition, the absence of the expression of PhpP and StkP proteins was confirmed by Western immunoblotting using the affinity-purified rabbit anti-StkP and anti-PhpP antibodies (36).
To restore the PhpP and StkP expression in the mutants, the genetically confirmed mutants, D39DPhpP and D39DStkP, were respectively complemented with pDC123 plasmids containing either the wild-type phpP or stkP gene along with its native RBS (109). Briefly, the phpP gene (741 bp), along with its 18 bp RBS, was cloned between KpnI and EcoRI sites of the pDC123 vector, using a specific primer pair (pDC-phpP-F and pDC-phpP-R). Similarly, the wild-type stkP gene was cloned between the same restriction sites using the primers pairs (pDC-StkP-F and pDC-StkP-R) and cloned into the pDC123 (Chl R ) plasmid. PhpP and StkP were then expressed under the control of the cat gene promoter. The mutant strains (D39DPhpP and D39DStkP) were made competent in the presence of CPS-1 peptide, transformed with their respective pDC123-phpP and pDC123-stkP plasmids.
The resultant phpP/stkP-complemented mutant strains (D39DPhpP::phpP/D39DStkP::stkP) were then selected on blood agar plates containing chloramphenicol (5 mg/mL). The absence of PhpP and StkP proteins in the D39DPhpP and D39DStkP mutants and their restoration in the complemented strains were confirmed by Western immunoblot analysis and qRT-PCR.
Generation of MapZ knockout mutant from the D39 S. pneumoniae strain. D39DMapZ mutant was created using a 3,133-bp synthetic gene construct (GenScript, NJ) encompassing upstream and downstream regions of the mapZ gene (spd_0342) linked in-between with a promoterless kanR gene cassette containing its own RBS (ribosomal-binding site). The synthetic gene, therefore, included in tandem (i) the 847-bp upstream region of the mapZ gene (spd_0342) (i.e., the last 787 residues of the upstream region of spd_0341, 12 residues of the intergenic region, and first 48 residues of the mapZ gene), followed by (ii) 1,443 residues of the promoterless kanamycin gene (kanR) along with its ribosomal binding site (RBS-GGAGGTAAATAA) derived from the pFW13 vector (108), and (iii) the 843-bp downstream region of the mapZ gene (i.e., the last 42 residues of mapZ, followed by 75 residues of the intergenic region and first 726 bp residues of the downstream gene, gndA [spd_0343]) (Fig. S2A). This synthetic DNA was cloned between BamHI and SphI sites of the pUC57 plasmid (DmapZ-pUC57). A 3152 bp PCR-amplified product obtained with a primer pair DmapZ-up-F/DmapZ-Dn-R (Table S6) was used to transform the D39 wild-type strain. Colonies of D39DMapZ mutant were selected on kanamycin (200 mg/mL)-containing blood agar plates. The genetic integrity of the mutant was confirmed by PCR and sequencing using appropriate flanking primers. (Table S6).
Generation of double mutants. Using D39DMapZ and D39DPhpP mutants, three additional double mutants were derived. To create the D39DMapZ-DStkP mutant strain, D39DMapZ mutant was transformed with the PCR product used for creating D39DStkP mutant wherein stkP was replaced with the aad9. To create D39DMapZ-DPhpP, D39DPhpP was transformed with the PCR product used to create D39DMapZ wherein mapZ was replaced with kanR. Similarly, D39DPhpP was transformed using a PCR product with an upstream flanking region of the phpP gene and downstream region of the stkP gene amplified from D39DStkP mutant to obtain D39DStkP-DPhpP (see Table S6). The double knockout mutants were selected on blood agar plates containing kanamycin and/or spectinomycin described above. D39DMapZ-DStkP mutant was selected on kanamycin-spectinomycin plates. D39DMapZ-DPhpP and D39DStkP-DPhpP mutants were selected on kanamycin, and spectinomycin blood agar plates, respectively. The genetic integrity of these mutants was confirmed by PCR using appropriate primers (Table S6).
RNA extraction. For RNA-Seq and quantitative real-time PCR analyses, total RNA was extracted from three independent pneumococcus cultures grown in THY broth at 37°C grown at a mid-log-phase in the CO 2 incubator. Briefly, the bacterial cultures were pelleted, washed, and adjusted to OD 620 of 0.6. Next, 10 mL of these normalized cultures were pelleted and resuspended in 200 mL of the lysing buffer (PBS containing 25 mg/mL pneumo-phage lysin) (110) and further incubated for 1 h at 37 C for total lysis. The total RNA was extracted using the Norgen total RNA-extraction kit (Norgen Biotek Corp., Thorold, ON, Canada) per the manufacturer's instructions. The DNA-free total RNA was obtained by treatment with the RNase-free DNase (Millipore). Qualitative and quantitative analyses of the total RNA were determined by the Agilent RNA6000 bioanalyzer (Agilent Technology, Santa Clara, CA). High-quality total RNA was determined based on RNA integrity (RIN) .7.
RNA-Seq-based transcriptome analysis. The DNase-treated total RNA samples were subjected to RNA-Seq analysis in the commercial facility of BGI Americas (Boston, MA, USA). High-quality total RNA (RNA integrity [RIN] .7, 23S/16S .1.0) was determined by the Agilent RNA6000 bioanalyzer (Agilent Technologies), and subsequently, rRNA was removed from each preparation and subjected to corresponding cDNA preparation. Short fragments were made via heat treatment, purified, and resolved using EB (elution buffer) buffer for end repair and the addition of single nucleotide A (adenine). Short fragments were then connected to adapters. Following agarose gel electrophoresis, the suitable fragments were selected for PCR amplification. After confirming the quality and quantity of the sample library, the libraries were sequenced using an Illumina HiSeq 2000 with 100 bp sequencing. Three biological repeats each of D39-WT, D39DPhpP, and D39DStkP were subjected to RNA-seq transcriptome analysis using the high-quality total RNA (20 mg) preparations, The primary sequencing data (raw as well as filtered clean reads) produced by the Illumina HiSeq 2000 were essentially analyzed as recently described (64) and aligned with the reference sequence of S. pneumoniae D39 using the SOAP aligner/SOAP2 v2.20 software (62). The alignment of data were used to calculate the distribution of the reads on the reference genes and perform coverage analysis and quantification analysis of gene expression. In addition, Gene ontology enrichment analysis was used for pathway enrichment analysis. The published algorithm assessed the gene expression profile's significance (111). Threshold values for the P value and false detection rate (FDR, adjusted P value, or q value) , 0.01 and the absolute value Log 2 ratio $1 were used as the cutoff criteria to designate significantly differentiated genes (66). The RNA-seq-based nine transcriptome data files (three each corresponding to the D39-WT, D39DStkP, and D39DPhpP strains) submitted to the GEO database entry were approved as GSE113337.
Differentially Quantitative real-time reverse transcriptase-PCR. For the quantitative real-time reverse transcriptase-PCR (qRT-PCR) analysis, total RNA was extracted from three independently grown cultures of the wild-type D39 strain and the corresponding isogenic D39DPhpP and D39DStkP mutants as described above. The first-strand cDNA and subsequent determination of mRNA expression levels of genes were carried out using SYBR green qRT-PCR master mix (Roche), specific primers (Table S6), and a LightCycler 480 (Roche) real-time PCR instrument, as described previously (64). The absence of any PCR-amplified band other than the one expected in a PCR using the D39 genomic DNA as a template confirmed the integrity of primers. The copy numbers of all the genes were normalized to the housekeeping gene, 16S rRNA. The linear fold change in the mRNA expression levels was analyzed using Exor4 software (Roche), and $2-fold decreased/increased transcript abundance was considered a significant change.
Electron microscopy. Transmission and scanning electron microscopy (TEM and SEM) of the wildtype, mutant, and stkP/phpP-complemented pneumococcal strains were carried out as described previously (64). Briefly, pneumococcus strains were grown to their mid-log-phase, and the bacterial pellets obtained after centrifugation were washed 3 times with 0.1 M cacodylate buffer, pH 7.4, and resuspended in freshly made 0.1 M cacodylate buffer containing 2.5% glutaraldehyde and 4% paraformaldehyde for 30 min on ice followed 4°C for overnight fixation. After fixation, the samples were processed for TEM as well as SEM at the Ohio State University Central Microscope Imaging-Core Facility (CMIF). TEM was performed using a cryo-capable digital TE microscope (Technai G2 Spirit; EFI), and SEM using a field emission gun-equipped SE microscope (Nova SEM400, EFI), as described previously (64). Quantitative analysis of the different morphotypes (lengths and widths) in each strain was conducted manually based on multiple fields of SEM. Similarly, the number of septa per bacterium was counted based on multiple TEM fields. Statistical analysis was conducted based on data obtained from ;20 to 50 SEM/TEM fields, each representing an average length, width, and number of septa from 4 to 6 bacteria by a nonparametric t test with Welch's correction using GraphPad Prism 6.0 software. A P value less than 0.05 was treated as a significant difference.
Immunofluorescence microscopy. Immunofluorescence microscopy of D39 wild-type, D39DStkP, D39DPhpP, and corresponding complemented strains was performed to detect the expression level of the type-2 polysaccharide capsule using rabbit type-2 polysaccharide antibody (Statens Serum Institute, Copenhagen). The R6 pneumococcal strain grown to an OD 600 of 0.5 to 0.6 in 10 mL of THY broth was heat-killed by incubating the culture at 65°C for 2 h, followed by cooling at room temperature (RT). Heat-killed bacteria were then washed, pelleted by centrifugation, and suspended in 1 mL of PBS. The latter was mixed with an equal volume of pneumococcus type-2 rabbit antibodies, incubated overnight at 4°C under constant rotation, and centrifuged. The supernatant containing nonadsorbed antibody specific to type-2 polysaccharide (1:50 dilution) was incubated with the pneumococcal wild-type, mutant, and complemented strains for 2 h, followed by incubation with Cy3-conjugated antirabbit IgG (1:500 dilution) antibody for 1 h to detect the presence of the capsule. The bacterial pellets were washed and stained with 49,6-diamidino-2-phenylindole (DAPI) dye. The bacterial suspension was then thinly spread on the polylysine-glass slides and air-dried. The stained slides were then observed under a fluorescence microscope (Nikon Eclipse E600) interfaced with a Nikon CCD camera (DSFi1c). Images of the same field were separately captured using red and blue filters and merged using NIS-element (version 4.13) image analysis software.
Growth curves in the presence of different carbon sources. Initially, the growth curve patterns of the D39 wild-type, D39DStkP, D39DPhpP, and stkP and phpP-complemented mutant strains were determined using a chemically defined C1Y glucose medium (36). Subsequently, D39-WT, and D39DStkP, D39DPhpP, D39DMapZ, D39DMapZ-DStkP, D39DMapZ-DPhpP, and D39DStkP-DPhpP mutant strains were examined in the C1Y medium supplemented with different carbon sources (1% [wt/vol] glucose, maltose, lactose, galactose, and maltodextrin). The fresh culture grown in the chemically defined C1Y medium to mid-log-phase and adjusted to an OD 620 of 0.6 was diluted to 1:100 in the C1Y medium with different carbon sources. Individual wells of sterile 96-well round-bottom microtiter plates were seeded with 200 mL of the D39 wild-type and various mutant strains diluted cultures. The spectrophotometer was programmed (FluorStar Galaxy software, BMG) to analyze the growth kinetics of culture in each well of the microtiter plate every hour for 15 h at 37°C with an adjustment of horizontal shaking for 10 s before every OD 620 reading. Growth curves were obtained from three independently grown cultures, each grown in triplicate wells. The initial background values in different groups obtained at the time 0 were subtracted from values obtained at other time points in each group. The corrected values from three individual cultures were used to plot the growth curves using GraphPad Prism-6.
Extraction of capsule. The capsular polysaccharide was extracted from late log phase-grown strains of S. pneumoniae. The cultures were pelleted, washed once with PBS by centrifugation, and adjusted to an OD 620 of 1.0. Each of these cultures was then centrifuged, and the resulting pellet was resuspended in 1 mL of lysis buffer (PBS with 0.2% sodium deoxycholate, 25 mg/mL pneumolysin, 5 mg/mL DNase, 10 mg/mL RNase, and 10 mM MgCl 2 ) for 1 h at 37°C. The culture lysates were centrifuged for 15 min at 17,000 Â g at 4°C. The capsule material was extracted from the supernatant with chloroform (1:1 vol/vol). The upper aqueous phase containing capsule material was removed after centrifugation (17,000 Â g for 10 min at 4°C) and adjusted to the original starting volume.
Estimation of capsular contents. The presence or absence of capsules extracted from the wild-type, single and double mutants and/or complemented pneumococcal strains in whole-cell lysate was qualitatively measured by Western immunoblot analysis. In addition, the presence of polysaccharide capsules was visualized using the capsule type 2-specific antibody described above for immunofluorescence microscopy and corresponding conjugate antibody by the chemiluminescent or chromogenic method, as described previously (64).
The quantitative analysis of the extracted polysaccharide was carried out by measuring the rhamnose sugar content in the extracted capsule as described previously (116). Briefly, 1 mL of diluted samples was mixed with 4.5 mL of 15.86 M sulfuric acid and incubated for 20 min at room temperature. The reaction mixture was then boiled for 10 min and cooled on ice for 20 min. Subsequently, 100 mL of 3% cysteine was added and incubated for 2 h at room temperature. The development of the colored furfural precipitates was spectrophotometrically measured at 430 nm. PBS-containing wells were treated as controls. The rhamnose content in the samples was determined based on the standard curve obtained using different rhamnose concentrations (10 to 100 mg/mL). Results received from three independent experiments, each with three technical replicates, were statistically analyzed by the nonparametric Student's t test with Welch correction using GraphPad Prism 6. A P value less than 0.05 was treated as a significant change.
In vitro phosphorylation. In vitro autophosphorylation assays were performed using 2 mg of StkP in the presence of 1 mCi g 32 P-ATP (specific activity 3,000 Ci/mMol, PerkinElmer) at 30°C for 45 min in a final volume of 30 mL phosphorylation buffer (50 mM Tris/HCl, pH 7.5, 1 mM dithiothreitol, 5 mM MgCl 2, and/ or 5 mM MnCl 2 ). The dephosphorylation of StkP was carried out in the presence or absence of an equimolar concentration of PhpP. Similarly, reversible phosphorylation of CcpA was carried out in the presence or absence of StkP, and PhpP, described above. Protein phosphorylation status was determined by SDS-PAGE followed by Coomassie stain and autoradiography.
In vivo phosphorylation of CcpA by StkkP. In vivo phosphorylation of CcpA was performed by expressing the stkkP and ccpA genes in pCDFDuet-1 vector (Novagen). The stkkP gene was cloned between NdeI and KpnI sites within the multiple cloning site-2 (MCS2) using a PCR product obtained with a primer pair, StkkP-F/StkkP-R. The ccpA gene was cloned between BamHI and HindIII within the MCS-1 region of the plasmid flanked by 6ÂHis tag-encoding region in the upstream region of the plasmid, using a primer pair, His-CcpA-F/HisCcpA-R. The genetically verified plasmid pCDF-His-CcpA/ STKK was used to transform DH5a E. coli and pLysS BL-21 E. coli. CcpA was overexpressed in the BL21 strain under the induction of 1 mM IPTG. The in vivo StkkP-phosphorylated CcpA protein was purified by the Ni-NTA affinity column chromatography described above.
Similarly, His-CcpA was expressed in D39DPhpP complemented with the wild-type his-ccpA gene using the pDC123 complementation plasmid (Chl R ) as described above. In vivo phosphorylated His-CcpA was then purified using Ni-NTA chromatography as described above.
Mass spectrometry analysis. Purified in vitro and in vivo phosphorylated His-tagged CcpA proteins bands resolved by SDS-PAGE and Coomassie stained were excised and subjected to in-gel trypsin digestion followed by phosphopeptide enrichment and LC-MS/MS mass spectrometry analysis using the Orbitrap Fusion Tribrid mass spectrometer at the OSU Campus Chemical Instrument Center (CCIC) Mass Spectrometry and Proteomics (MSP) Facility. Briefly, gel pieces were washed in 100 mL of 50% methanol/ 5% acetic acid for 30 min and then dried and suspended in 75 mL 50 mM ammonium bicarbonate (with 5 mg/mL dithiothreitol [DTT]) for 30 min at RT. The DTT-containing ammonium bicarbonate was aspirated, and the gel pieces were suspended in 50 mM ammonium bicarbonate containing 15 mg/mL iodoacetamide and incubated for 30 min in the dark at RT. The gel pieces were then alternately incubated with 50 mM ammonium bicarbonate and acetonitrile, respectively, for 10 min and then dried. Dried gel slices were rehydrated with 25 mL of 50 mM ammonium bicarbonate and digested with 75 mL of sequencing grade-modified trypsin (10 mg/mL in 50 mM ammonium bicarbonate; Promega, Madison WI) for 6 h at 37°C. The peptides were repeatedly extracted from the polyacrylamide gel with 50% acetonitrile and 5% formic acid. The subsequent phosphopeptide enrichment approach was based on the well-established low pH TiO 2 enrichment method (117). Enriched phosphopeptides were then subjected to LC/MS-MS analysis using a Thermo-Scientific Orbitrap Fusion mass spectrometer equipped with an EASY-Spray Sources was operated in the positive ion mode. Peptides were separated using mobile phase A (0.1% formic acid in water) and mobile phase B (acetonitrile with 0.1% formic acid) with a flow rate set at 300 nl/min. Typically, mobile phase B was increased from 2% to 35% in 30 min and then increased from 35 to 55% in 5 min and again from 55% to 90% in 5 min and then kept at 90% for another 2 min before being brought back quickly to 2% in 1 min. MS/MS ion scan spectra recorded between m/z 400 and 1,600 were generated for the most abundant peaks to determine the amino acid sequence. The full scan was performed at FT mode, and the resolution was set at 120,000 to achieve high mass accuracy MS determination. Sequence information from the MS/MS data were processed by converting the raw files into a merged file (.mgf) using an in-house program, RAW2MZXML_n_MGF_batch (merge.pl, a Perl script). The resulting mgf files were searched using Mascot Daemon by Matrix Science version 2.5.1 (Boston, MA), and the database searched against the most recent Swiss-Prot or NCBI databases (NC_008533.2/CP000410.2, dated 22 Feb 2022). The mass accuracy of the precursor ions was set to 10 ppm. The fragment mass tolerance was set to 0.5 Da. Considered variable modifications were oxidation (Met), deamidation (Asn and Gln) and carbamidomethylation (Cys) and phosphorylation (Ser/Thr/Tyr). Proteins identified with at least 2 unique peptides were considered for reliable identification. A decoy database was also searched to determine the false discovery rate (FDR), and peptides were filtered according to the FDR. Phosphorylated peptides were manually validated, and proteins with a Mascot score of 50 or higher and a P value ,0.01 with a minimum of two unique peptides from one protein having a -b or -y ion sequence tag of five residues or better were accepted.
Electrophoretic mobility shift assays. Electrophoretic mobility shift assay (EMSA) experiments were performed using a 32 P end-labeled 285-bp probe encompassing the 250-bp region of the promoter element upstream and 37 bp downstream (-250 to 135 nt) of the cps2A gene spd_0315. These probes were PCR-amplified using a primer pair, PCps2A-F/PCps2A-R. Briefly, the binding of the different concentrations of purified nonphosphorylated CcpA (1 to 3 mM) with the labeled cps2A promoter was carried out in the final 40 ml of reaction mixer containing the labeled probe (20,000 CPM/mL) in EMSA buffer (2 mg/mL poly dI-dC [Sigma], 10 mM Tris pH 7.5, 35 mM KCl, 1 mM EDTA pH 7.5, 1 mM DTT, 6% glycerol and 1 mM MgCl 2 ). Subsequently, a constant concentration of nonphosphorylated CcpA protein was incubated with different concentrations of StkkP (0.5 to 3.0 mM), followed by incubation with poly di-dC containing EMSA buffer for 5 min at room temperature. The reaction mixture was further incubated at 37°C for 25 min after adding the probe. The CcpA-bound and free probes in the reaction mixtures were resolved by 4.5% nondenaturing polyacrylamide gel electrophoresis (200V Â 30 min) using 0.5Â Trisborate-EDTA (TBE) buffer and visualized by autoradiography. In addition, the 100Â cold probe was mixed with the labeled probe in the EMSA reaction buffer to determine the specificity of CcpA binding to the Pcps2A promoter.
In vivo bacterial virulence assay. The virulence potential of the D39 wild-type and corresponding mutants (D39DPhpP and D39DStkP) and phpP/stkP-complemented strains (D39DPhpP::phpP and D39DStkP::stkP) was assessed by employing two mouse infection models. For the septicemia infection model, 17 CD-1 mice (5 weeks old, 20 to 22 g, Charles River Laboratories) were anesthetized with isoflurane and injected with 100 mL of the diluted culture in sterile PBS (5 Â 10 3 CFU) via the retroorbital route. The ability of these pneumococcal strains to cause systemic infection followed by colonization of the nasopharynx was determined by employing the intranasal infection model. As described above, a group of briefly anesthetized 17 mice was infected intranasally with pneumococcal strains (5 Â 10 7 CFU/ 20 mL). A group of 8 mice that received PBS alone (retroorbitally or intranasally) served as a shaminfected control group. Intranasal infection experiments were performed in two instances (10 and 7 mice per group and housed as 3 to 5 mice/cage). At the end of the experiments, the survival/mortality data were combined for individual mutants and corresponding complemented strains.
The morbidity/mortality for all these groups was monitored 3 to 4 times daily in the first 3 days postinfection and twice daily for 10 days. Experimental animals showing signs meeting the criteria for early removal were euthanized by using compressed CO 2 gas (CO 2 pressure 10 to 30% of the chamber volume/min) followed by cervical dislocation to confirm death. The Log-Rank test was applied to statistically analyze the percent survival for each group, and the survival curve was plotted using GraphPad Prism 6 software. The systemic bacterial burden in these animals was determined by counting CFU in an aliquot of 5 mL of blood collected at 24 h and 48h postinfection. Median CFU values obtained from each group (n = 10) were statistically analyzed by Mann-Whitney nonparametric unpaired test. The lungs were excised from three animals in each group on day three postinfection (72 h), then fixed in 10% formalin and further processed by the OSU Histopathology Core Facility for tissue sectioning and staining for histopathology. Histopathological slides were examined after Grams and hematoxylin and eosin (H&E) staining to determine the bacterial load and extent of lung pathology. Slides were scanned using Leica ScanScope XT2 high resolution (0.5 mm/pixel) scanner and viewed at Â20 and Â40 magnification using Aperio ScanScope Image viewer. NanoString technology-based inflammatory genes expression analysis. The confluent cultures of A549 cells (CCL-185, ATCC), maintained in 6-well tissue culture plates, were cocultured with D39 wildtype and isogenic mutant D39DStkP, D39DPhpP, and D39 DMapZ pneumococcus strains (multiplicity of infection [MOI] ;100:1 bacteria:cell) for 4 h and total DNase-free RNA was isolated from triplicate tissueculture samples for each mutant using Norgen total RNA isolation kit as described above. The quantity and quality of RNA samples were assessed by NanoDrop 2000 (Thermo Fisher Scientific, USA). RNA samples were then subjected to multiple inflammatory gene expression analyses using NanoString nCounter inflammation profiling panels (NanoString, USA). Data analysis was carried out with the nSolver Analysis Software 4.0 for the profiling Panel (NanoString, USA) according to the manufacturer's recommendations. Briefly, samples were examined using 50 ng of total RNA loaded for each sample. Probes for each gene in the panel were allowed to bind and detect the mRNA molecules of that target gene in the sample and copy numbers of mRNA molecules were counted by NanoString technology. The gene-expression data were normalized in two steps; positive-control normalization (with internal positive-control sequences) and housekeeping normalization, differential expression plots, fold change values for each gene, and other analyses were obtained using the nSolver Analysis Software version 4.0 (NanoString, USA). The list of genes used for analysis is given in Table S8 in the supplemental material, with corresponding annotations.
Meso Scale Discovery analysis of cytokines/chemokines. The concentrations of IFN-g , TNF-a, and 8 different interleukins (IL-1-b, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13) in culture supernatants of A549 cells cocultured with D39-WT and other derived single and double mutants collected at 0, 2, and 4 h time intervals were analyzed using V-plex Plus proinflammation panel 1 (human) kit (C4049-1) (Meso Scale Diagnostic, Rockville, MD) based on an electro chemiluminescent detection method according to the manufacturer's recommendations. All collected samples were stored at 280°C, thawed, and diluted 1:3 before analysis. Data were collected and analyzed using MSD Meso 1300/SQ 120 Quickplex Microplate Reader equipped with Discovery workbench data analysis software (MSD, Rockville, MD) and located in the OSU CCTS-supported Analytical & Development Laboratory. Briefly, 96-well microtiter plates were coated with linker-coupled capture antibodies (specific to analyte listed above and provided by the manufacturer) for 1 h and then aspirated and washed with washing buffer (PBS/0.05%Tween 20) 3 times. Standards and supernatants (25 mL) were added to appropriate wells, incubated for 1 h with shaking, and washed again 3 times with washing buffer. Detection antibodies were added to each well and incubated for 1 h at room temperature and washed, as described above. Finally, 150 mL of a reading buffer was added to each well. The plate was analyzed on the MDS instrument. Standard curves were formed by fitting electrochemiluminescence signals from calibrators, and analyte concentrations were extrapolated from a standard curve calculated using a four-parameter logistic fit using MSD Workbench 3.0 software. Data were then plotted and statistically analyzed using GraphPad Prism-6 software.
Data availability. RNA-seq analysis data are deposited in GEO database (GEO Accession no. GSE113337).

ACKNOWLEDGMENTS
This study was supported by OSU Department of Pathology internal fund no. 50271 to V.P. The authors acknowledge the help of Daniel Nelson (University of Maryland) for providing pneumococcus specific phage lysin, to Liwen Zhang and Arpad Somogyi (OSU Comprehensive Cancer Center [CCC] supported proteomic facility) for mass spectrometry analysis. Transmission and electron microscopy images presented in this report were generated using the instruments and services at the OSU Campus Microscopy and Imaging Facility (CMIF) and the Department of Pathology Clinical Core facility. The proteomic and CMIF core facilities are supported by NIH award number grant P30 CA016058. The Fusion Orbitrap instrument is supported by NIH award number grant S10 OD018056. Histopathology and slide scanning were performed using the instruments and services at the OSU Department of Pathology Core Facility, a division of OSU Human Tissue Resource Network (HTRN). MSD analysis was funded and supported by award number UL1TR002733 from the National Center for Advancing Translational Sciences to OSU Center for Clinical and Translational Science (CCTS). The content is solely the responsibility of the authors. It does not necessarily represent the official views of the National Center for Advancing Translational Sciences or the National Institutes of Health.
We declare no conflicts of interest with the contents of this article.