3.1. The mouse pneumonia model was successfully constructed
In this study, bronchial instillation of LPS was used to establish a mouse pneumonia model. According to the pre-experiments, a single bronchial instillation of LPS can only cause a mild pneumonia state that lasts for approximately two days, so the method of instilling LPS every other day is used to maintain the mice in a pneumonia state during the dosing cycle (Fig. 1A). To verify whether the mice remained in a state of pneumonia during the 7-day experimental period, we randomly selected 3 mice from the GP or XP groups (model group) every day, sacrificed them, and collected BALF to detect inflammatory factors. We used ELISA to measure the levels of IL-6 and TNF-α in BALF collected from three randomly sacrificed mice daily. The results are shown in Fig. 1C and 1D, and it is clear that the content of IL-6 and TNF-α in BALF was significantly increased from the first day of modeling compared with that in the BALF of the control group without modeling, and the content of IL-6 and TNF-α also increased with the increase in the number of models, which means that the pneumonia state of the mice was also aggravated. Inflammatory cytokine testing alone did not fully confirm that the mice were in a state of pneumonia, so we also performed H&E staining on the mouse lungs (Fig. 1B). As shown in Fig. 1B, H&E staining revealed obvious inflammatory cell infiltration in the GP and XP groups compared with the nonmodeled GC and XC groups. From the morphological photographs of the lungs, it can also be seen that compared with those of the control groups (GC and XC groups), the lung tissues of the GP and XP groups were slightly enlarged and slightly larger, and some parts were dark red. Based on the above results, a 7-day mouse model of pneumonia was successfully established in our study.
3.2. Mice in the pneumonia group had less kidney damage and less dA-ALI production
The 7-day body weight curve of each group showed a downward trend (Fig. 2A), which may be caused by the stimulation of the mice by daily gavage. There were significant differences in the body-to-kidney weight ratio (Fig. 2B) between the GC group and the control group, and there were also significant differences between the GC group and the GP group (pneumonia CAM group). There were no significant differences among the XP, XC, and control groups. Based on these results, we can say that the GC group may have some kidney injury compared with the GP group and the control group, but the degree of injury needs to be further tested and confirmed. Therefore, we not only measured the serum creatinine (CR) and urea nitrogen (BUN) concentrations (Fig. 2D&2E) in the mice but also performed H&E staining (Fig. 2C) on the kidneys of the mice to further confirm the degree of kidney damage in each group of mice. As shown in Figs. 2D and 2E, the serum CR and BUN levels in the mice in each group exhibited the same high and low trends; the levels in the GP group were slightly lower than those in the GC group, and there was a significant difference. The levels in the XP group were significantly lower than those in the XC group, there was a significant difference, and the levels of GC, GP and XC were significantly greater than those in the control group. A comparison of the XC vs. XP and GC vs. GP groups revealed that the CR and BUN levels of the XP and GP groups in the pneumonia model group were lower than those in the corresponding nonmodel XC and XP groups.
According to the H&E staining results (Fig. 2C), the glomerular structure in the field of vision of the GP, XP and XC groups was intact, the outline was clear, the cyst cavity was not expanded, the number of mesangial cells was clear, the number was normal, no necrosis was found, and there was no obvious inflammatory cell infiltration in the tissue. In addition, only the GP and XC groups exhibited brush border injury in a small number of renal tubules. In the GC group, individual glomeruli exhibited atrophy, mesangial cells were significantly reduced, the cyst cavity was slightly dilated, a small number of brush border injuries of renal tubules occurred, a small amount of protein mucus was visible in the renal tubular lumen, and no obvious inflammatory cell infiltration was found in the tissues. In summary, we preliminarily concluded that the ARR and CAM cause less kidney damage in pneumonic mice than in healthy mice. The XP group were even at the same level as those in the control group.
Finally, we extracted DNA from mouse kidney tissue, hydrolyzed it into a single nucleotide, and measured the concentration of the resulting DNA adduct (dA-ALI). The results are shown in Fig. 2F, within our detection limits, with dA-ALI detected only in the GC and GP groups. The dA-ALI content in the GP group was significantly lower than that in the GC group. Since pneumonia was the only variable between the GP and GC groups, we believe that pneumonia may have protected the body, resulting in less drug toxicity, less dA-ALI production, and less kidney damage.
3.3. NQO1 protein in the healthy group was significantly greater than that in the pneumonia group
Both CYP1A1/2 and NQO1 are enzymes involved in the metabolic process of AAI. WB analysis revealed no difference in CYP1A1/2 expression between the pneumonia group and the healthy group (Fig. 3B&3C), while NQO1 expression was significantly greater in the healthy group than in the pneumonia group (Fig. 3D). This result suggested that the reduction in kidney injury and adducts may be related to NQO1.
3.4. Metabolomic analysis of pneumonic mice and healthy mice
Serum metabolomics analysis of pneumonia mice and healthy mice was performed by multivariate analysis of the serum metabolic profiles of mice in the healthy (GC, XC group) and pneumonia (GP, XP group) groups. The PCA score plot showed that healthy mice and pneumonia mice were completely different (Fig. 4A&4B). As shown in the orthogonal projections to latent structures-discriminant analysis (OPLS-DA) score plot (Fig. 4C&4D), the metabolite profile distributions for the healthy (XC, GC) and pneumonia groups (XP, GP) were clearly separated, with both \({R}^{2}Y\) and \({Q}^{2}\) greater than 0.5. Each point in the volcano plot (Fig. 4E&4F) represents a peak, and the graph contains all the materials measured. The abscissa represents the fold change of the group of contrasted substances (taking the logarithm with 2 as the base), the ordinate represents the P value of Student's t test (taking the negative number with 10 as the base logarithm), and the scatter size represents the VIP value of the OPLS-DA model; the larger the scatter is, the greater the VIP value. Metabolites that are significantly upregulated are indicated in red, metabolites that are significantly downregulated are shown in blue, and metabolites that are not significantly differentiated are shown in gray. Figure 4E shows the metabolite levels in the healthy group administered CAM compared to those in the pneumonia group (GC vs. GP), and Fig. 4G shows the metabolite levels in the healthy group administered ARR compared to those in the pneumonia group (XC vs. XP).
According to the criteria OPLS-DA VIP (variable importance in the projection) > 1.0 and P value \(<\) 0.05, we identified 430 differentially abundant metabolites (123 upregulated and 307 downregulated) and 286 differentially abundant metabolites (83 upregulated and 203 downregulated) in mice administered CAM and ARR, respectively, that were significantly affected by pneumonia. The differentially abundant metabolites were mainly lipids and lipid-like molecules, organic acids and derivatives, and organic oxygen compounds, such as glucosylceramide, trans-octadecanoic acid, myristoleic acid, methyl dihydrojasmonate, and sebacic acid. Most metabolites differed between CAM-treated healthy mice and pneumonic mice, suggesting that these specifically altered metabolites may be related to the alleviation of the side effects of the drug caused by pneumonia.
We analyzed the metabolic pathways of mice that were administered CAM or ARR. The results are shown in Fig. 4I&5J; the abscissa represents the RichFactor corresponding to each pathway, and the ordinate is the name of the KEGG metabolic pathway. The size of the dot indicates the number of differentially abundant metabolites enriched in the pathway. The color indicates the size of the p value, and the smaller the p value is, the more reddish the color, and the more significant the enrichment. Because CAM and ARR are both herbs that contain AAI, we believe that there is a certain similarity in the mechanism by which pneumonia alleviates toxic side effects, so we directly examined the metabolic pathways in which they were enriched. As shown in Fig. 4I&4J, glycerophospholipid metabolism, thiamine metabolism and sphingolipid metabolism are all pathways that may be involved in alleviating the toxic side effects of pneumonia.
3.5. Transcriptomic analysis based on pneumonic mice and healthy mice
To explore the transcriptomic changes underlying the differences in the effects of CAM and ARR on healthy mice and pneumonia mice, RNA-seq data were obtained from healthy and pneumonia samples. The PCA score plot (Fig. 5A&5B) showed that there was a large difference between the pneumonia group and the healthy group, and the reproducibility within the group was acceptable. The heatmap was obtained by taking the DEGs of all comparison groups and taking them as the differential gene sets and then performing cluster analysis on the differential gene sets. In the figure, the abscissa is the sample name, and the ordinate is the normalized value of the differential gene FPKM; the redder the color is, the greater the expression level, and the greener the color is, the lower the expression level (Fig. 5C). The volcano plot clearly reflects the genes that are significantly expressed in both groups and is indicated in red and blue. The results showed that the expression of 236 genes in the CAM group (GP vs. GC) significantly changed in the pneumonia group, of which 212 genes were upregulated and 24 genes were downregulated (\(p\le 0.05\), Fig. 5D). The expression of 779 genes in the ARR group (XP vs. XP) significantly changed during pneumonia, of which 562 genes were upregulated and 779 genes were downregulated (\(p\le 0.05\), Fig. 5F). These results indicated that pneumonia status had a significant effect on transcription in both the CAM and ARR groups. Since CAM and ARR are both herbs that contain AAI, we believe that there is a certain similarity in the mechanism of reducing their toxic side effects, so we directly searched for the same pathway through KEGG pathway enrichment analysis. As shown in Fig. 5E and 5G, chemical carcinogenesis—reactive oxygen species—was the only pathway enriched in the CAM-administered group and the ARR-administered group.
The reactive oxygen species pathway in chemical carcinogenesis refers to a carcinogenic mechanism mediated by reactive oxygen species molecules. Reactive oxygen species, including oxygen radicals and peroxides, are highly reactive and can oxidize biomolecules, leading to cell damage and mutations, thereby increasing the risk of cancer. According to the KEGG differential gene enrichment analysis of GP vs. GC and XP vs. XC, the chemical carcinogenesis-reactive oxygen species pathway showed a downward trend. It is likely that pneumonia can alleviate the toxic side effects of AAI-containing herbal medicines by inhibiting this pathway.
3.6. Determination of oxidative stress markers in the serum and kidney tissues of mice
We measured the following oxidative stress-related indicators [15, 16]in the serum and kidney tissues of the mice: reactive oxygen species (ROS), malondialdehyde (MDA), superoxide dismutase (SOD), oxidized glutathione (GSSH), reduced glutathione (R-GSH), and total glutathione/oxidized glutathione (T-GSH/GSSH). We first measured ROS in kidney tissue, as shown in Fig. 6A. The ROS in the healthy group (GC, XC group) were significantly greater than those in the pneumonia group (GP, XP group), which was also consistent with our transcriptomic enrichment results. Figure 6B shows that the SOD concentration in kidney tissue was significantly greater than that in serum in the pneumonia group, while there was no significant difference in the healthy group. This may indicate that the kidneys have a strong antioxidant capacity in the pneumonic state. Figure 6D shows that the MDA concentration in kidney tissue was lower than that in serum in all groups, suggesting that the degree of oxidative damage in kidney tissue may be lower than that in other organs. Figure 6C shows that the R-GSH concentration in the kidney tissue of the pneumonia group was significantly greater than that in the healthy group, suggesting that the kidneys of the pneumonia group may have experienced greater antioxidative stress. Figure 6E shows that the T-GSH/GSH ratio in the renal tissue of the pneumonia group was significantly greater than that in the renal tissue of the healthy group, while the opposite was true for the serum. These results indicated that the reduction in renal tissue in the pneumonia group was greater than that in the control group, which was conducive to normal metabolism and functional maintenance of cells, while the oxidative state of renal tissue in the healthy group was greater, which may have led to cell damage and abnormal function.
3.7. Pneumonitis can attenuate mouse kidney injury and dA-ALI adduct production through the Keap1-Nrf2 pathway
Based on the results of transcriptome KEGG enrichment and the results of oxidative stress indicators in section 3.5, we deduced that pneumonia may reduce kidney injury and dA-ALI production through the Keap1-Nrf2 pathway. Therefore, we performed WB analysis of Keap1 and Nrf2 proteins in mouse kidney tissue. WB results are shown in Fig. 6G. Grayscale analysis revealed that Keap1 in the pneumonia group (GP, XP) was significantly lower than that in the healthy group (GC, XC), while Nrf2 was significantly greater than that in the healthy group (Fig. 6F). This finding is consistent with our hypothesis that pneumonitis may attenuate kidney injury and adduct production via the ROS-Keap1-Nrf2 pathway.