S.pn serotype 2 strain D39 was provided by Zunyi Medical College. QFY was provided by the Affiliated Hospital of Changchun University of Traditional Chinese Medicine (Jilin, China). Acetate (#695092) and butyrate (#B103500) were purchased from MilliporeSigma. Mouse interleukin-lβ (IL-1β, #MLB00C), interleukin-6 (IL-6, #M6000B), and tumor necrosis factor-alpha (TNF-α, #MTA00B) enzyme-linked immunosorbent assay (ELISA) kits were obtained commercially (R&D Systems, Minneapolis, MN, USA). Rabbit polyclonal antibodies against NF-κB P65 (#ab16502), NF-κB p-P65 (#ab194726), inhibitor of NF-κB (IκBα) (#ab7217), glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (#ab8245) and β-tubulin (#ab18207) were purchased from Abcam (Cambridge, MA, USA). Rabbit polyclonal antibodies against NOD-like receptor protein 3 (NLRP3) (#15101S) and cleaved caspase-1 (cleaved-Casp1) (#89332S) were purchased from Cell Signaling Technology (Beverly, MA, USA).

QFY was provided by the Affiliated Hospital of Changchun University of Traditional Chinese Medicine (Jilin, China). Each component powder was accurately weighed and formulated to generate QFY based on the following ratio (by weight): Scutellaria baicalensis Georgi, Forsythia suspensa (Thunb.), Belamcanda chinensis (L.) DC, and Fritillaria cirrhosa D. Don of 8:8:8:3. According to the standard procedure (National Pharmacopoeia Committee, 2005), QFY powder was decocted 3 times with 200 mL of water at 100 °C to obtain the aqueous extract. All decocted solutions were combined and centrifuged then the supernatant was dried under vacuum to produce a brown powder. The extract was stored in the laboratory at −80°C until use. In previous studies, 19 main chemical compounds comprising QFY were detected using a method based on high performance liquid chromatography with tandem mass spectrometry (HPLC-MS/MS). Comparisons of results obtained from analysis of ten batches of QFY recipe indicated batch-to-batch consistency within the range of 95% to 98% of all major chemical constituents. Thus, different batches of QFY retained principal active components of the four Chinese herbal ingredients with acceptable batch-to-batch variability ( Figure S1 ; Tables S1, 2 ).

The S.pn strain D39 (Berry et al., 1989) was cultured at 37°C in Todd-Hewitt broth then the culture was inoculated onto tryptic soy broth (TSB) agar plates and incubated under the same conditions overnight. Next, bacterial colonies were collected from the plates and cultured in Todd-Hewitt broth supplemented with 5% yeast extract until growth reached mid-log phase (as determined from the optical density of the culture at 600 nm).

After female BALB/c mice (approved by the Animal Ethics Committee of Changchun University of Chinese Medicine-20190116) were received from the Experimental Animal Center of Changchun University of Traditional Chinese Medicine, they were housed and maintained for 6 weeks until they weighed 20 ± 2 g. Next, mice were allowed to rest for 5 d to acclimate before being subjected to experimental manipulations. For experimentation, mice were randomly assigned to nine groups: Control, Model, QFY (L) (0.21 g/kg), QFY (H) (0.42 g/kg), gut microbiota-depleted (DEPL), S.pn+DEPL, S.pn+DEPL+QFY, S.pn+DEPL+Acetate, and S.pn+DEPL+Butyrate (n = 7 per group). In experiments with DEPL mice, mice received high-dose QFY.

To deplete the gut microbiota, (designated as DEPL in group names above), mice were treated with a broad-spectrum antibiotic in drinking water as previously reported (Schuijt et al., 2016). Next, mice assigned to groups Model, QFY (L), QFY (H), S.pn+DEPL, S.pn+DEPL+QFY, S.pn+DEPL+Acetate, and S.pn+DEPL+Butyrate were lightly anesthetized via inhalation of isoflurane then were infected via left nasal inoculation of a 25-μL volume of nose drops containing 2.5 × 10⁸ CFU/mL of S.pn to establish the in vivo pneumonia model (Dessing et al., 2010). Beginning at the time of S.pn infection, microbiota-deficient mouse groups received drinking water containing a broad-spectrum antibiotic. Meanwhile, mice in QFY treatment groups QFY (L), QFY (H), and S.pn+DEPL+QFY were intragastrically administered QFY twice daily, while mice in the S.pn+DEPL+Acetate group were administered acetate (200 mM) and mice in the S.pn+DEPL+Butyrate group received butyrate (100 mM) in drinking water. All mice were sacrificed at 48 h post-infection.

Body weight, lung tissue wet weight to dry weight (W/D) ratio, and immune system-associated organ indices were determined in order to assess general characteristics of mice during experiments. Determinations of mouse organ immune index values (for thymus and spleen) were conducted using the following formulas: thymus index = (thymus weight/mouse weight) × 10, spleen index = (spleen weight/mouse weight) × 10.

As report previously (van Deel et al., 2016; Ruscitti et al., 2017), mice were anaesthetized with isoflurane then lung imaging was performed using a Quantum FX Micro-CT scanner (PerkinElmer, Inc., Waltham, MA). The X-ray system of the scanner uses a microfocus tube with a focal spot size of 5 μm at 4 W and produces X-rays delivered as a cone-shaped beam. Images were acquired using X-ray tube settings of 90 kVp and 160 μA. Projection radiographs were taken throughout the 360-degree gantry rotation cycle during a total scan time of 14 min.

Lung tissues of mice were prepared for histology and analyzed as described previously using the following steps: immersion of tissues in 4% paraformaldehyde followed by paraffin embedding, sectioning to generate slices, attachment of slices to glass slides, dewaxing of slices with xylene, immersion of slides in eosin staining solution, dewaxing of slices until transparent, and sealing of stained slices on slides. Total lung inflammation scores (TLISs) were determined as previously reported (Dessing et al., 2007). The entire lung surface was analyzed with respect to the following parameters: interstitial damage, vasculitis, peri-bronchitis, edema, thrombus formation, and pleuritis.

The trachea was exposed through a midline incision and bronchoalveolar lavage fluid (BALF) was harvested by instilling 1 mL of sterile isotonic saline into the lungs. BALF fluid was then tested for levels of IL-6, IL-1β, and TNF-α by ELISA.

Total microbial DNA was extracted using the OMEGA Soil DNA Kit (D5625-01) (Omega Bio-Tek, Norcross, GA, USA) from mouse cecal samples (n = 7). Next, PCR amplification of 16s rRNA gene V3-V4 was performed using forward primer 799F (5’-AACMGGATTAGATACCCKG-3’) and reverse primer 1193R (5’-ACGTCATCCCCACCTTCC-3’). PCR amplicons were quantified then were sequenced using the Illlumina MiSeq platform with MiSeq Reagent Kit v3 at Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China). For microbiome community profiling evaluations, sequence data were analyzed using QIIME2 and R software packages (v3.2.0) (Bolyen et al., 2019).

30 µg of lung tissue lysate was loaded per lane then samples were electrophoretically separated using 10% SDS-PAGE gels. Primary antibodies specific for the following antigens were used at 1:1000 dilution: β-tubulin, GAPDH, NLRP3, cleaved-Casp1, NF-κB p65, p-NF-κB p65, and IκB. After blots were incubated with primary antibodies overnight at 4°C, they were washed then incubated with horseradish peroxidase-conjugated IgG secondary antibody (either goat anti-mouse or goat anti-rabbit) diluted 1:5000. Protein bands were visualized and analyzed using a chemiluminescent imaging system (FluorChem, ProteinSimple, San Jose, CA, USA).

All data are presented as means ± SD as calculated using Student’s t-test Comparisons between groups were conducted using one-way or two-way ANOVA followed by Bonferroni or Tukey’s post-tests as indicated. Differences were considered significant vs Control group, ^# P < 0.05, ^## P < 0.01; vs Model group, ^* P < 0.05, ^** P < 0.01; n.s is used to denote results that are not significant.

As shown in Table 1 , as compared with the Control group, body weights and spleen and thymus indices were decreased in untreated infected mice by 48 h after S.pn infection. However, the lung W/D ratio was reduced after QFY treatment (P < 0.05), suggesting that QFY treatment of infected mice significantly reduced pulmonary edema and preserved host immune system balance during S.pn infection.

Next, micro-CT scans and hematoxylin and eosin (H&E) staining were used to investigate lung inflammation induced by S.pn infection, as well as protective effects of QFY against lung tissue injury. In one micro-CT scan image, patchy, non-homogeneous high-density shadows and marginal haziness were observed in the lower lobe of the lung. Meanwhile, representative histological findings of tissues from intreated S.pn-infected mice showed extensive signs of pneumonia manifesting as deepened lung tissue color accompanied by signs of interstitial inflammation, vasculitis, bronchitis, and edema. In addition, lobes exhibited signs of congestion, swollen tissues, and evidence of neutrophil infiltration. Scans of lungs of infected animals receiving QFY treatment also exhibited patchy, non-homogeneous shadows, but these shadows were less dense than patches appearing in lung scan images of untreated infected animals. These imaging results indicated that lung tissues of QFY-treated infected mice exhibited less bleeding and swelling and more closely resembled lung tissues of healthy mice than did lung tissues of untreated infected mice ( Figures 1A, B ). Moreover, based on the TLISs semi-quantitative scoring system described in the Methods section, histopathological scores were much lower in the QFY-treated infected mouse group as compared with the untreated model group, indicating that QFY treatment of mice prevented inflammatory lung damage induced by S.pn infection ( Figure 1C ). Antibiotic treated also reduced the TLISs, but we observed that a considerable number of mice had diarrhea and loose feces which not observed in QFY group, regrettably ( Figure S2 ).

In addition, we measured levels of inflammatory cytokines in mice BALF specimens. As shown in Figures 1D–F , S.pn infection led to significantly increased inflammatory cytokine levels of TNF-α, IL-1β, and IL-6 in BALF (p < 0.01), while QFY treatment of infected mice led to significantly decreased BALF levels of these cytokines. Thus, these results suggest that S.pn infection induced lung inflammation that was reversed by QFY treatment.

Levels of NF-κB-NLRP3 pathway proteins that play key roles in lung immune responses were evaluated by Western blot analysis. As shown in Figures 1G–I , the NLRP3 signaling pathway was activated by S.pn invasion, as indicated by increased levels of NLRP3 and cleaved-Casp1 proteins. In addition, NF-κB, a transcription factor associated with NLRP3 pathway regulation, was also activated by S.pn infection, as confirmed by both an observed 43.32% decrease in IκBα level and a 3.79-fold increase in NF-κB p65 phosphorylation level as compared with corresponding Control group levels. These relative differences indicate that S.pn infection triggered a domino effect by activating an intense inflammatory cascade accompanied by induction of a mature cytokine response that induced high-level release of cytokines such as IL-1β. Strikingly, QFY treatment markedly maintained expression of these proteins at normal levels, especially in the QFY(H) group. Thus, hereafter data associated with QFY treatment groups was obtained using a high dose of QFY (0.42 g/kg) if not otherwise mentioned. Furthermore, after S.pn infection of mice, measurements of bacterial loads revealed no significant differences between QFY-treated and untreated groups ( Figure S3 ).

Taken together, the abovementioned data revealed that S.pn infection-induced lung inflammation was reversed by QFY treatment and that this QFY therapeutic effect may be linked to inhibition of NLRP3 and NF-κB p65 pathways.

To explore the relationship between the inflammatory response and intestinal microecology, we analyzed the microbiota within mice cecal digesta. Community richness and diversity were evaluated using Chao 1 and Shannon indices ( Figures 2A, B ), with results showing that the Model group had markedly lower Shannon (P = 0.0011) and Chao 1 index (P = 0.0011) values as compared to corresponding Control group values. Importantly, QFY treatment restored infection-induced alterations of microbial community richness and diversity levels to nearly normal levels.

Principal coordinates analysis (PCoA) revealed overall differences in microbiota classification profiles among groups ( Figures 2C, D ), especially between Model and Control groups, while QFY treatment group results overlapped with those of the Control group. At the phylum level, the Venn diagram illustrating overlap of Operational taxonomic units (OTUs) profiling ( Figure 2E ) and hierarchical cluster analysis results ( Figure 2F ) revealed statistically significant differences in microbiota profiles between Control and Model groups, whereas the gut microbiota profile of the QFY-treated infected group more closely resembled that of the Control group. Taken together, the abovementioned results indicated that administration of QFY treatment shifted the microbiota profile from one of disease status towards one of healthy status.

Next, in order to further compare species composition differences, heatmap were used to visualize significant bacterial taxonomic profile differences among groups at the phylum level ( Figure 3A ) and genus level ( Figure 3B ). In addition, LEfSe (LDA, linear discriminant effect size) was used to construct a hierarchical tree based on sample composition for use in difference analysis of results for all classification levels ( Figure 3D ).

S.pn infection caused significant perturbations of the microbiota at the phylum level ( Figures 2F , 3A ). Phyla Firmicutes and Bacteroidetes dominated Control and QFY group profiles, while the phylum Proteobacteria was most prominent in the Model group profile. At the genus level, bacterial community profiles differed markedly between Control/QFY and Model groups ( Figure 3B ).

Notably, abundances of genera Ralstonia, Sphingomonas, Ochrobactrum, and Streptococcus were significantly increased after S.pn infection as compared to their respective abundances in Control and QFY groups ( Figure 3C ). Interestingly, SCFAs-producing bacterial genera mainly included Coprococcus, Prevotella, Butyricicoccus, [Ruminococcus], Rikenella, Dehalobacterium, and Parabacteroides, the levels of which were all decreased in untreated S.pn-infected mice and restored to normal levels in QFY-treated infected mice.

Observed differences in microbiota profiles among mouse groups in this study prompted us to test SCFAs levels in feces of each group ( Figure 3E ). As expected, S.pn infection decreased fecal levels of acetate and butyrate, while QFY treatment led to increases of both SCFAs to normal levels. Spearman correlation analysis revealed an interrelationship between SCFAs and corresponding gut microbiota among Control, S.pn-infected and QFY-treated S.pn-infected mice ( Figure 3F ). Intriguingly, highly significant negative correlations were observed between SCFAs and TLISs and between SCFAs and levels of cytokines IL-1β, IL-6, and TNF-α. Conversely, acetate and butyrate levels were each significantly positively correlated with abundances of genera Coprococcus, Prevotella, Butyricicoccus, [Ruminococcus], Rikenella, Dehalobacterium, and Parabacteroides, were significantly negatively correlated with abundances of Ralstonia and Streptococcus.

Taken together, alterations of gut microbiota composition and function may be attributed to effects of S.pn infection, which subsequently resulted in dysregulation of innate host immune responses. Conversely, QFY treatment led to marked reversal of infection-induced microbial imbalances to normalize the gut microbiota. This effect was positively correlated with increased levels of SCFAs acetate and butyrate and increased abundance of gut probiotic organisms that produce beneficial SCFAs.

Mice were depleted by oral administration of broad-spectrum antibiotics in drinking water [as described previously (Schuijt et al., 2016)], we conducted all experiments as mentioned below in gut microbiota-depleted mice (termed DEPL mice). As shown in Figure 4 , microbiota depletion was confirmed by observations of markedly lower gut microbial community richness and diversity in DEPL mice ( Figures 4A–D ). In addition, relative abundances of beneficial gut bacteria in DEPL mice were significantly decreased ( Figures 4E–G ), fecal SCFAs levels were decreased ( Figure 4H ), and DEPL mice exhibited more severe pneumonia symptoms ( Figure 5 ). These results align with results reported recently showing that gut microbiota plays a protective role in the host defense against pneumococcal pneumonia (Schuijt et al., 2016). Taken together, the abovementioned results indicate that the gut microbiota-depleted mouse model was successfully established in this work. Moreover, QFY treatment of gut microbiota-depleted mice afflicted with S.pn-induced pneumonia exhibited normalization of microbiota profiles and associated metabolite levels, especially with regard to SCFAs. Thus, effects of SCFAs may play pivotal roles in observed QFY therapeutic effects associated with alleviation of pulmonary infectious diseases.

To investigate whether gut flora and SCFAs were potential targets of QFY pneumonia therapy, DEPL mice were treated with QFY after S.pn infection. Figure 5 displays results demonstrating that QFY exhibited no therapeutic effect in mice with prior gut microbiota depletion induced by administration of a broad-spectrum antibiotic. Specifically, results of H&E staining and inflammatory cytokine level determinations revealed no differences between QFY-treated and untreated S.pn-infected DEPL mice, indicating that gut microbiota depletion counteracted the beneficial QFY effect. Based on the widely accepted fact that SCFAs fermented by gut flora are converted into metabolites with anti-inflammatory properties, we next administrated two SCFAs, acetate and butyrate, to S.pn-infected DEPL mice and observed marked changes associated with improved health status after SCFAs treatment. As shown in Figure 6 , although QFY did not exert a protective effect in DEPL mice against pneumonia, a protective effect was achieved by administration of acetate and butyrate. More specifically, administration of these SCFAs improved the health of S.pn-infected DEPL mice as compared to untreated DEPL mice, as evidenced by decreased inflammatory cytokine levels and greater alleviation of pathology, as indicated by lung imaging findings.

It has been reported that SCFAs acting as NF-κB inhibitors block the inflammatory cascade (Chi et al., 2020; Wu et al., 2020). Therefore, we next evaluated the relative expression levels of proteins among experimental groups using Western blot analysis. As shown in Figures 6F–H , levels of NF-κB pathway proteins that were dramatically elevated after S.pn infection were decreased after acetate and butyrate treatments. However, in untreated infected mice, changes of levels of NLRP3, a target protein of NF-κB that plays a key role in immune responses to infectious diseases, paralleled changes of phosphorylated NF-κB levels, leading to cleavage of pro-caspase-1 and triggering of the inflammatory cascade.

In conclusion, we demonstrated that there is a close relationship between lung infection and disruption of intestinal flora. QFY therapeutic effects appeared to be associated with both repair of disordered intestinal flora through restoration of basal SCFAs levels to normal levels (in uninfected mice) and inhibition of NF-κB-NLRP3 pathway activity, which together alleviated lung injury induced by bacterial pneumonia.