Introduction

Space is a complex environment characterized by a variety of extreme conditions, such as low temperatures, intense radiation, high vacuum, and microgravity. Although space stations are shielded from most extreme conditions, microgravity is still a serious mutagenic factor1. Microgravity is a state of apparent weightlessness that occurs due to the significant reduction in physical forces exerted by gravity; it is typically characterized by a gravity value ranging from ~10−3 to 10−6 g and is accompanied by low-shear and low-turbulence conditions2. Studies have shown that spaceflight can impact astronauts’ lung tissue, cardiovascular system, and immune function3,4,5. Additionally, research has indicated that certain microbes may be introduced into the space environment by humans intentionally or inadvertently. Within enclosed spacecraft, the risk of microbial infection increases the longer astronauts spend in orbit. A variety of opportunistic pathogens, such as Klebsiella pneumoniae (K. pneumoniae), have been detected by post-spaceflight tests both in the astronauts and on the spacecraft6.

K. pneumoniae, a gram-negative bacterium, is commonly found on animal mucosal surfaces and in everyday environments such as soil and water. In humans, this bacterium predominantly resides in the gastrointestinal tract but is sometimes present in the nasal and pharyngeal areas. K. pneumoniae can invade the bloodstream or other tissues through the nasal and pharyngeal areas, leading to systemic infections. Consequently, K. pneumoniae is a significant risk factor for severe community-acquired infections in clinical settings7. A prevalent opportunistic pathogen, K. pneumoniae is non-pathogenic under typical conditions; its pathogenicity is conditional upon certain factors, and pathogenic K. pneumoniae is particularly common in individuals with compromised immune systems, often presenting as invasive infections. In the space environment, the immune systems of astronauts are compromised5, increasing the risk of K. pneumoniae infection.

In recent years, interest in the genetic traits and mutation mechanisms of microorganisms under microgravity conditions has increased. However, because changes for microorganisms to experience authentic space environments are limited, ground-based simulated microgravity (SMG) systems have been developed8. The Rotary Cell Culture System (RCCS), a widely utilized tool, employs a specialized culture dish—the High-Aspect Rotating Vessel (HARV)—to simulate microgravity accurately on Earth. Properly filled with liquid culture medium and rotated at precise speeds and angles, the HARV effectively mimics microgravity conditions9.

Studies have shown that 1 month of spaceflight can alter the biofilm formation ability of Acinetobacter baumannii and the growth rate and drug resistance of Salmonella enteritidis10,11. Furthermore, after SMG induction, genes and proteins involved in diverse metabolic processes and biological pathways were differentially regulated in Stenotrophomonas maltophilia12. However, the study periods have been limited predominantly to 14–30 days, a timeframe that differs markedly from the typical 3–6 month mission duration for astronauts. Certain phenotypic traits of K. pneumoniae, including morphology, growth rate, and biofilm formation capacity, are reportedly altered by short-term SMG exposure13, with SMG-treated strains exhibiting the upregulation of genes associated with type 3 fimbriae14. However, it remains unclear whether these alterations persist under long term SMG conditions. Moreover, RNA sequencing (RNA-seq) has been used to identify differential expressed genes (DEGs) in pathogens under SMG versus control conditions, yet differential expressed metabolites (DEMs) have seldom been characterized13,14,15. Consequently, we cultured a clinical strain of K. pneumoniae under SMG for 56 days and employed multiple experimental techniques to investigate its phenotypic changes. RNA-seq and metabolome sequencing analyses were conducted to elucidate the potential regulatory mechanisms of DEGs and DEMs in K. pneumoniae under SMG.

Based on existing research, we aimed to investigate whether the phenotypic characteristics of K. pneumoniae would continue to change under long term SMG induction, and search for genes and metabolites involved in abnormal regulation of K. pneumoniae under SMG. We aimed for our findings to serve as a valuable reference for future research and the development of safer space missions.

Results

Changes in the phenotypic characteristics of K. pneumoniae under SMG

First, we conducted growth tests on the strains cultured under SMG and NG conditions for the 24-h period of the 56th day (Fig. 1A). The OD600 values revealed that there was a difference in the growth rate of K. pneumoniae between the two environments at 1 h, and the differences in the measured values between the two environments remained non-significant until ~12 h. Notably, between 12 and 18 h, K. pneumoniae experienced a brief growth spurt, during which the differences between the two groups were significant (p < 0.05). This period could mark the inflection point of an S-shaped (sigmoidal) growth curve. After 18 h, the bacteria entered a plateau phase, indicating a stable growth stage. After incubation at 37 °C for 24 h, the final OD600 values were significantly different between the NG group and the SMG group, suggesting that K. pneumoniae exhibited slower growth under SMG than under NG.

Fig. 1: Phenotypic changes in K. pneumoniae under SMG.
figure 1

A Growth curves of K. pneumoniae in SMG and NG. OD600 readings were measured for each strain. B Morphology of K. pneumoniae in SMG and NG. C Biofilm quantification of K. pneumoniae in SMG and NG. OD590 readings were measured for each strain to determine the amount of biofilm formed. The data represent specific values from three independent experiments.

Full size image

Second, TEM was used to observe the single-cell morphology of the strains cultured under the two conditions (Fig. 1B). Compared with the strains cultured in NG, the strains cultured in SMG presented more hypertrophy, lighter intracellular staining, clearer intracellular ultramicro substances, and visibly distinguishable capsules. Notably, the cells of the strain cultured in NG still maintained a spindle-shaped cell morphology, whereas the cells of the SMG-cultured strain tended to be more circular.

K. pneumoniae is well-known for its robust capacity to form biofilms16. To investigate whether this characteristic would exhibit short-term phenotypic response changes under SMG, we evaluated the biofilm formation ability of K. pneumoniae under SMG using crystal violet staining (Fig. 1C). The results indicated SMG group and NG group exhibited darker staining, suggesting the formation of biofilms in each group. However, the degree of staining was difficult to distinguish with the naked eye. Further quantitative analysis showed the OD590 values of the two groups were also very similar. Upon conducting a t-test, we found that there was indeed a statistically significant difference between the two groups (p = 0.0069). This finding indicated that after 56 days of cultivation, the thickness of K. pneumoniae biofilms in the SMG group slightly increased compared with that in the NG group.

The DEGs between the SMG and NG groups of K. pneumoniae were involved mainly in metabolic function

After Pacbio sequencing and assembly, the reference genome of K. pneumoniae was obtained and 5388 coding sequences were detected (Supplementary Data 1). RNA-seq was subsequently conducted on both the SMG and NG populations. Following the removal of adapters, low-quality sequences, and potential contaminants, the clean rate of all samples was greater than 99%, and the quality scores of Q20 and Q30 were greater than 98% and 96%, respectively (Fig. 2A). Meanwhile, the reads of NG and SMG groups were 21,546,100 ± 12,546 and 20,582,228 ± 33,529, respectively, with a stable GC content of about 54% (Supplementary Data 1). This indicated that the sequencing data was of high quality and could truly reflect the genetic information of the samples. To elucidate the regulatory network of K. pneumoniae under SMG and NG conditions, we identified DEGs between the two groups according to fragments per kilobase of transcript per million mapped reads (FPKM) values (Supplementary Data 1).

Fig. 2: RNA-seq analysis of K. pneumoniae between the NG and SMG.
figure 2

A RNA-seq quality control information. B Volcano plot of the DEGs. The red, blue and gray dots represent upregulated, downregulated and not significantly changed genes, respectively. C Bar plot of DEGs mapped with COG.

Full size image

A total of 486 DEGs were identified, among which only four genes were upregulated (Fig. 2B); these genes were annotated as hypothetical protein KP1_1104, bifunctional enzymes with N-succinyldiaminopimelate-aminotransferase/acetylornithine transaminase activities, arginine N-succinyltransferase and hypothetical protein in NR. Further analysis identified the four upregulated DEGs were annotated as csiD, astC, and astA (with the last gene remaining unannotated) in the KEGG database, which were involved in amino acid metabolism pathways, particularly arginine and proline metabolism. As for downregulated genes, 476 genes could be annotated by NR, further KEGG results showed that these genes were mainly involved in transport, metabolism and bacterial secretion (Supplementary Data 1).

In addition, 302 DEGs were mapped to COG categories (Fig. 2C). In the majority of cases, the DEGs were categorized according to the following functions: carbohydrate transport and metabolism, amino acid transport and metabolism, inorganic ion transport and metabolism, energy production and conversion, intracellular trafficking, secretion, and vesicular transport, cell wall/membrane/envelope biogenesis. And these categories were classified in the COG database as related to substance metabolism, cellular processes, and signal transduction. Consequently, after 56 days of cultivation, the DEGs between the SMG group and the control group were associated mainly with various metabolic pathways, with a small portion participating in cellular processes and signaling.

The pathways in which DEGs were enriched under SMG were involved mainly in catabolic metabolism, growth-related processes and biofilm formation

To explore the mechanism of differential regulation of K. pneumoniae under SMG, we further analyzed these DEGs (Supplementary Data 2), and found carbohydrate transport (p = 4.33E-04), catabolic process (p = 5.92E-04) and single-organism catabolic process (p = 7.32E-04) were meaningful (Fig. 3A). Clearly all of these biological processes were related to the metabolism of substances in cells, including the production of energy and the transformation of substances. Meanwhile, catabolic activity and lyase activity enriched many DEGs, including hycCDE, eutBDG, astAC, and ulaBCE, which involved in energy metabolism and amino acid degradation (Fig. 3B).

Fig. 3: Enrichment analysis of DEGs in K. pneumoniae between the NG and SMG groups.
figure 3

A GO enrichment results of DEGs. B Network diagram of DEGs in GO terms. C KEGG enrichment results of DEGs. D Relationship diagram of growth-related pathways in KEGG. The blue lines in (B) and (D) represent downregulated genes, whereas the red lines represent upregulated genes.

Full size image

KEGG analysis revealed majority of DEGs were enriched in fructose and mannose metabolism, phenylalanine metabolism, and the degradation of aromatic macromolecular compounds (Fig. 3C), indicating K. pneumoniae modulated metabolic pathways under SMG. Additionally, we found that the functions of the DEGs in pathways were related to the metabolism of carbohydrate and amino acid transport, and they could facilitate energy and signal transduction, and act as carbon source to support growth (Table 1). Studies had found metabolism, signal transduction and genetics in bacteria may lead to slow growth17,18. Our results showed fructose, ascorbate and phenylalanine had produced many downregulated genes, which was consistent with our observation that K. pneumoniae grew slowly under SMG (Fig. 3D).

Table 1 Key metabolic and biofilm formation pathways associated with DEGs between the SMG and NG groups
Full size table

On the other hand, the bacterial secretion system and biofilm formation were highly enriched. We observed that genes related to the secretion system, extracellular polysaccharides, membrane, and pilus assembly, such as gsp operon, imp/vas secret system core components and fim gene, were down regulated, and these genes are components of bacterial biofilm formation19,20. Their abnormal expression may cause changes in the biofilm thickness of K. pneumoniae under SMG19,20,21.

Furthermore, Gene set enrichment analysis (GSEA) also revealed significant differences in some gene sets of K. pneumoniae grown under the two conditions (Supplementary Fig. 1). The phosphotransferase system (PTS), which was widely present in bacteria and is involved in carbon/nitrogen metabolism, plays an important role in the overall metabolism of bacteria22. The expression of PTS in K. pneumoniae was significantly reduced under SMG, which affected the metabolism of K. pneumoniae. Moreover, the gene sets related to the bacterial secretion system and biofilm formation, specifically in Pseudomonas aeruginosa, were enriched and presented decreased expression under SMG. The DEMs of K. pneumoniae under SMG conditions were significantly enriched in phenylalanine metabolism.

In total, there were 37 intracellular and 55 extracellular DEMs were identified, and it was found that both intracellular and extracellular DEMs were mainly classified as organoheterocyclic compounds, organic acids and derivatives, nucleosides, nucleotides, and analogs, lipids and lipid like molecules in Human Metabolome Database (HMDB) (Fig. 4A, B). There were 6 upregulated and 31 downregulated metabolites in intracellular, and 47 upregulated and 8 downregulated metabolites in extracellular. There were 6 upregulated and 31 downregulated metabolites in intracellular, and 47 upregulated and 8 downregulated metabolites in extracellular. Further KEGG analysis results showed that (Supplementary Data 3), phenylalanine metabolism was the most significantly enriched pathway (p = 0.0315) (Fig. 4C), including trans-cinnamate, phenylethylamine, phenylacetaldehyde, and phenylacetylglutamine. Moreover, pathways associated with peptidoglycan and lipopolysaccharide biosynthesis, which are integral to the outer membrane and cell wall, were enriched. Notably, glutathione metabolism demonstrated the greatest enrichment, although it was not statistically significant (p = 0.0563) (Fig. 4D).

Fig. 4: Metabolomic analysis of K. pneumoniae between the NG and SMG groups.
figure 4

A, B Clustering of the expression patterns of intracellular and extracellular DEMs. C, D KEGG enrichment analysis of intracellular and extracellular DEMs. E Schematic diagram of phenylalanine degradation pathways involving DEGs. The blue and brown dashed lines represent trans-cinnamate degradation and phenylacetate degradation, respectively. The expression patterns of each DEG are listed according to their respective groups.

Full size image

Phenylalanine, an essential amino acid in organisms, is commonly converted to phenylacetate by bacteria23, and the phenylalanine metabolism pathway includes both synthesis and degradation. Degradation can be divided into phenylacetate degradation and trans-cinnamate degradation. The mhp gene cluster, which is responsible for the 3-(3-hydroxyphenyl)propionate catabolic pathway, is crucial for trans-cinnamate degradation23. Under SMG, trans-cinnamate expression was reduced, likely suppressed by the mhp gene cluster, which facilitates its breakdown and integration into the citrate cycle (Fig. 4E). Additionally, both the mhp gene cluster and its metabolic operon hcaD presented decreased expression in the transcriptome under SMG, suggesting a downregulation in the trans-cinnamate degradation pathway. This result also indicated a corresponding decrease in energy production. For phenylacetate degradation, only three DEMs were identified.

Correlation network of DEGs with DEMs under SMG, especially those involved in both intracellular and extracellular arginine and proline metabolism

RNA-seq and metabolomics analyses revealed that DEGs and DEMs were enriched in multiple metabolic pathways in K. pneumoniae under SMG. Co-expression network analysis revealed that DEGs regulated specific DEMs involved in the physiological activities of bacteria (Supplementary Data 4). And some key DEGs regulated multiple intracellular compounds (Fig. 5A), for example, astA simultaneously regulated thymine, lysophosphatidylethanolamine, and two forms of phosphatidylcholine, while astC indirectly affected three metabolites and linked to the biosynthesis of aspartate family amino acids. N8-acetylspermidine, a product of spermidine metabolism in the nucleus, can be converted back to spermidine by deacetylase in the cytoplasm24. Our findings showed N8-acetylspermidine was downregulated under SMG and regulated by a transport-related genes (tabA), carbon and nitrogen source utilization genes (eutMP, glcG), and other genes (madB, selA, parB, yhaV). The expression of extracellular compounds was similarly regulated by numerous DEGs (Fig. 5B). A variety of extracellular DEMs centered on guanosine were interconnected through DEGs, including carbon source utilization proteins (rhaP, pduD, srlB), phosphorylation genes (celA, selA), and membrane-related genes (omph, mpB, gspH, hcp, papC), creating a regulatory network of genes that target metabolism and the transmembrane transport of carbohydrates. Furthermore, the genes astAC, along with puuB and csiD, were involved in the regulation of 11 extracellular compounds. These findings showed that DEGs actively regulated the arginine and proline metabolism pathway and formed a close and complex network with DEMs, suggesting that under long term SMG, some DEGs in K. pneumoniae cooperated with DEMs to participate in the regulation of arginine and proline.

Fig. 5: Co-expression analysis of DEGs and DEMs in K. pneumoniae via STRING.
figure 5

A Intracellular. B Extracellular. The orange circles represent compounds, the outer red circles represent DEMs, blue icons represent pathways that may be affected, and green icons represent DEGs. The green triangles and green diamonds are related to metabolism and the membrane, respectively.

Full size image

Discussion

K. pneumoniae is a prevalent opportunistic pathogen capable of causing pneumonia and other infections7. SMG is a state where gravity in the ground environment is reduced to an almost negligible level using relevant techniques. Our findings indicate that under SMG, K. pneumoniae exhibited phenotypic changes, including slower growth, enlarged and more circular cell morphology, and increased biofilm thickness. Concurrently, our study utilized multiomics analysis to elucidate the impact of SMG on K. pneumoniae from different perspectives; for example, RNA-seq analysis revealed that the DEGs of K. pneumoniae under SMG compared to NG were mostly downregulated, and multiple metabolic pathways and biofilm-related pathways were significantly enriched. The integration of metabolomics further elucidated the interplay between DEGs and DEMs associated with amino acid (phenylalanine, arginine, proline) and pyrimidine metabolism, suggesting a potential causal link to the observed phenotypic alterations in K. pneumoniae under SMG. This approach expanded upon a single node representing the sole characteristic of the pathogen and revealed some unique patterns that differed from previous findings.

Bacteria can take up carbon/nitrogen sources and convert them into energy to maintain normal growth25, and astAC played a crucial role in the conversion of arginine to glutamate18. Our results showed the expression of astAC and glutamate in K. pneumoniae were upregulated under SMG, and glutamate was easily converted to stable glutamine, but this process required energy. This energy may be derived from EutMP, glcG, and pduF, which utilize ethanolamine, glycolate, and propanediol as carbon sources to maintain normal physiological functions26. Under SMG, these genes were downregulated in K. pneumoniae, which may result in reduced carbon source utilization and affect normal growth26. Meanwhile, the decrease in intracellular N8-acetylspermidine expression may also affect normal cell proliferation and differentiation27. Additionally, K. pneumoniae could fully utilize phenylalanine as a nitrogen source to support growth28, and our results suggested the low expression of propanediol utilization proteins may indicate that K. pneumoniae can directly utilize phenylalanine to support growth, reducing energy requirements29,30,31. The expression levels of mhp gene clusters, trans-cinnamate, and phenylacetate derivatives involved in phenylalanine metabolism were downregulated under SMG, which may also to the natural entry of this pathway into the citrate cycle, resulting in reduced energy production by K. pneumoniae23,28. In summary, we found the downregulated expression of carbon and nitrogen source utilization proteins, intermediate metabolites of arginine and proline and phenylalanine degradation and their regulatory genes in K. pneumoniae, resulting in a decrease in overall energy in the later stage, which affected the energy that K. pneumoniae could use to support growth and normal growth. This decrease was likely one of the reasons for the slower growth of K. pneumoniae under SMG32.

A biofilm is a surface-attached microbial community consisting of bacteria and their secreted extracellular polymeric substances33, and its formation is a multistage process that can be divided into five distinct phases33. In the attachment phase, bacteria reversibly adsorb onto biological or non-biological surfaces34, and during the colonization phase, bacteria reinforce this adsorption through pilus structures and extracellular polysaccharides35. Our results revealed that pilus assembly proteins, their chaperone proteins and the extracellular polysaccharides of K. pneumoniae under SMG were significantly regulated during this process. Type I fimbriae are encoded by the fim gene cluster, of which fimA is the main subunit, which interacts with another key gene to promote biofilm formation36. Polysaccharides are among the main contributors to the structural stability of biofilms37, and poly-beta-1,6-N-acetyl-D-glucosamine (PGA) was one of these polysaccharides. pgaD was an essential gene that encodes proteins involved in PGA synthesis38, and its differential expression may affect the content of PGA, thereby affecting the normal biofilm formation process during the colonization period. During the development and maturation stages, bacteria proliferated to form multiple layers of cells and began to produce and secreted various products such as extracellular polymeric substances and enzymes39. We found that membrane-related genes (ompH) and multiple secretion system genes, especially gspCHIN in type II secretion system and hcp and VgrG in type VI secretion system, were significantly downregulated. The gsp operon gene controlled the general pathway of extracellular matrix secretion, and hcp and VgrG formed a pilus-like structure, assisting in the transfer of effector proteins to the extracellular40 and jointly promoting the formation of bacterial biofilm24. At the active dispersal phase, bacteria separate under various conditions to prepare for the subsequent biofilm formation process41. Our results revealed that many membrane-related genes and metabolites were expressed at low levels in SMG conditions, which may result in an inability of K. pneumoniae to complete the normal dissipation process, which is one of the reasons for the increase in biofilm thickness in the SMG group.

Bacteria can survive in almost all environments due to their strong characteristics, and changes in the environment may lead to different phenotypic responses42. Our results revealed that under SMG, the single-cell morphology of K. pneumoniae changed from short rods to circular, indicating K. pneumoniae was undergoing short-term phenotypic response under SMG. Unfortunately, our omics analyses did not directly screen for DEGs and DEMs related to this phenomenon, and we speculated it may be caused by bacteria being in the SMG, leading to the dispersion of cytoplasm in all directions, which will require further validation and research in the future.

Acinetobacter baumannii, Salmonella enteritidis and Stenotrophomonas maltophilia10,11,12 also explored their mechanism changes under SMG by observing their phenotypic changes and gene expression, and these studies also introduced genomic analysis to identify the mutated genes. In our study, we proposed a multi-layered model connecting SMG-induced phenotypic changes with transcriptomic/metabolite differential expression and gene-metabolite covariation. Notably, the observed transcriptomic-metabolite covariation under SMG conditions demonstrated how metabolomics bridged molecular dynamics with phenotypic manifestations, highlighting its critical role in functionally connecting these biological layers. However, as our current study lacked genomic data, whether these changes involved genomic-level variations remains to be investigated.

However, there was no doubt that our experiment still had shortcomings. The conclusions drawn currently required further direct evidence to support the observed phenotypic changes. To strengthen our findings, future work should include more rigorous induction cultures, detailed phenotypic assessments, comprehensive omics analyses, and robust validation studies. It was especially crucial to conduct thorough experiments with pathogenic bacteria under actual microgravity conditions for a prolonged duration. Furthermore, it was essential to delve deeper into the effects of microgravity on the pathogenicity and virulence of highly virulent strains, including K. pneumoniae. This additional research will help to elucidate the potential changes in their behavior and characteristics when exposed to the unique conditions of a microgravity environment. Understanding these alterations is crucial for assessing the risks and implications for human health and safety in space-related activities.

In the current study, we explored changes in phenotypic, transcriptomic, and metabolomic characteristics of K. pneumoniae after culturing on the RCCS for 56 days on the ground. Expression of genes or metabolites associated with bacterial metabolism and biofilm formation ability was altered after SMG. Research on SMG on the ground will help to understand the potential changes in growth and metabolism of pathogens such as K. pneumoniae in the microgravity environment. It can also simulate the physiological and psychological problems that astronauts may face in space, thereby providing health protection for astronauts during spaceflight and offering new ideas for the monitoring of microbial behavior.

Methods

Bacterial strains and growth conditions

This study did not involve human or animal subjects. The original K. pneumoniae strain was isolated from sputum samples of a patient suffering from alcoholic liver cirrhosis at the Second Center of Hepatobiliary Surgery/Liver Transplantation Research Center of the Fifth Medical Center of the People’s Liberation Army General Hospital (formerly the 302nd Hospital of the People’s Liberation Army); the strain was preserved at −80 °C and utilized as a reference standard. A schematic of the experimental setup from the establishment of the experimental groups, including the SMG group and the normal gravity (NG) control group, through the culture of bacterial cells in HARV bioreactors within an RCCS system (Synthecon, Texas, USA) is depicted in Supplementary Fig. 29. The culture bottle’s axis was aligned parallel to the ground in the SMG group and perpendicular in the NG group, and the other culture conditions were the same for both groups. The experiment commenced with an overnight culture of the bacterial strain at 37 °C, which was agitated at 200 revolutions per minute (rpm). Serial dilutions of the bacteria were prepared in glass tubes for further use. Then, 10 mL of Luria–Bertani (LB) broth was added to the HARV carefully to eliminate air bubbles. Next, 100 μL of the bacterial culture was inoculated into the HARV for both groups. After thoroughly mixing the contents and removing any bubbles, the HARVs were positioned in the RCCS under conditions of constant temperature (37 °C) and humidity. The rotational velocity was set as 25 rpm9. Subculturing was performed every 24 h, following this specific procedure: 1) removed HARVs corresponding to SMG and NG, and gently tapped to mix the internal bacterial suspension. 2) used a 0.5 ml syringe to draw 0.1 ml (1%) of bacterial suspension from the sampling port of HARV and inoculated it into a new HARV containing 10 ml LB liquid medium. 3) tapped HARV again to mix the bacterial suspension and expelled any bubbles. 4) labeled the HARV and placed it in the corresponding position to continue cultivation. The entire process was operated under sterile conditions, and the tools used were all clean disposable items. Lastly, the 1 ml of the bacterial mixture being cryopreserved in solution at a final glycerol concentration of 30%. Cultivation was maintained for a total of 56 days.

Growth rate

The original strain was inoculated into HARVs containing 10 ml of LB medium under both SMG and NG conditions and incubated at 37 °C for 24 h. The optical density at 600 nm was monitored hourly for 24 h in 200 μl aliquots of the culture using ultraviolet–visible spectrophotometer. Each group consisted of 3 independent samples for this experiment.

Transmission electron microscopy (TEM)

The strains were cultured in LB medium under two distinct conditions and then washed with sterile phosphate-buffered saline (PBS; pH 7.4). To prepare samples for observation via electron microscopy (FEI, USA), a copper mesh with its carbon surface facing upward was placed on a sealed film that was cooled using an icebox. Next, 20 μL of the sample, ultrapure water, and PTA dye solution were separately applied to the copper mesh. The mesh was allowed to absorb the liquids for 1 min and then transferred to ultrapure water and immediately dried with filter paper. The carbon surface of the copper mesh was subsequently dyed in PTA dye solution for 0.5 to 1 min and dried again with filter paper. Finally, the copper mesh was left to air dry naturally and observed via electron microscopy (FEI, USA).

Crystal violet staining

Two hundred microliters of bacterial suspension culture from the 56th day were added to each well of a 24-well flat-bottom microtiter plate with an equal volume of the culture medium (LB = 1:1) and then incubated at 37 °C for 48 h. The wells were subsequently washed three times with PBS, and the cells were stained with 200 μl of 0.1% crystal violet (Sigma, St. Louis, MO, USA) for half an hour. After the plates containing the three strains were washed with PBS and allowed to dry, the stained biomass samples were dissolved in 33% glacial acetic acid. The OD590 for each well was measured with an ultraviolet spectrophotometer (Beckman, GER). All the experiments were conducted in triplicate to ensure reliability.

Original strain genome sequencing and data analysis

The original strain was revived and grown on agar plates in fluid medium at 37 °C with agitation. Genomic DNA sample was isolated from the cell pellets with a Bacteria DNA Kit (OMEGA), and quality control was subsequently carried out on the purified DNA samples. Genomic DNA was quantified by using TBS-380 fluorometer (Turner BioSystems Inc., Sunnyvale, CA). High qualified DNA sample (OD260/280 = 1.8–2.0, >6 μg) was utilized to construct fragment library and sequenced by PacBio.

Then, after quality control of the raw reads using Trimmomatic43, ABySS44, canu45, and GapCloser46 were used successively to obtain the final assembly results. And we used ab initio prediction method to get gene models for original strain. Gene models were identified using GeneMark47. Then all gene models were blastp against Non-Redundant Protein Sequence Database (NR), Cluster of Orthologous Groups (COG), Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) to do functional annotation by blastp module. And these annotations serving as references for further analysis.

RNA sequencing

Total RNA was extracted from the tissue using TRIzol® Reagent according to the manufacturer’s instructions (Invitrogen), and genomic DNA was removed using DNase I (Takara). The RNA quality was assessed with a 2100 Bioanalyzer (Agilent), and the RNA concentration was measured with an ND-2000 (NanoDrop Technologies). Only high-quality RNA samples were used to construct the sequencing library. For the RNA-seq strand-specific libraries, the TruSeq RNA Sample Preparation Kit from Illumina (San Diego, CA) was utilized, starting with 5 μg of total RNA. rRNA was removed using the RiboZero rRNA Removal Kit (Epicenter), and the RNA was subjected to fragmentation. cDNA synthesis, end repair, A-base addition, and ligation of Illumina-indexed adapters were subsequently performed following Illumina’s protocol. The libraries were then size-selected for cDNA fragments ranging from 200 to 300 bp using a 2% low-range ultra agarose gel, followed by 15 cycles of PCR amplification with Phusion DNA polymerase (NEB). After quantification via TBS380, the paired-end libraries were sequenced on an Illumina NovaSeq 6000 system (150 bp*2, Biozeron, Shanghai, China). Each group had used three samples for sequencing and subsequent transcriptome analysis.

Comparative transcriptomic data analysis

The raw paired-end reads were trimmed and quality controlled by Fastp48, and clean reads were aligned against the reference genome using Bowtie 249. The gene-wise read count was determined using HTSeq-count50. DEGs were identified by calculating the expression levels of transcripts with the FPKM method. The edgeR package from R/Bioconductor51 was used to normalize the read counts and analyze differential gene expression. The DEGs between two groups were selected using the following criteria: the logarithmic of fold change (FC) was greater than 2 and the false discovery rate (FDR) should be less than 0.05. To understand the functions of the differential expressed genes, GO functional enrichment and KEGG pathway analysis were carried out by Goatools52 and KOBAS53 respectively. DEGs were significantly enriched in GO terms and metabolic pathways when their Bonferroni-corrected P-value was less than 0.05.

GSEA has been extensively applied to identify underlying pathways involved in various conditions54. The samples were classified as high or low according to the expression of K. pneumoniae. NOM P-values < 0.05, |NES | >1 and FDR q < 0.25 were considered to indicate statistical significance.

Metabolite detection and data analysis

After centrifugation, the samples cultured for 56 days were subjected to non-targeted metabolic analysis. In this analysis, bacterial cell samples were used to study intracellular metabolism, whereas bacterial liquid samples were used to study extracellular metabolism. Tissues (100 mg) were individually ground with liquid nitrogen, and the homogenate was resuspended in prechilled 80% methanol and 0.1% formic acid by vortexing. The samples were incubated on ice for 5 min and then centrifuged at 15,000 rpm and 4 °C for 5 min. Some of the supernatant was diluted to a final concentration of 53% methanol with liquid chromatography–mass spectrometry (LC–MS)-grade water. The samples were subsequently transferred to a fresh Eppendorf tube and then centrifuged at 15,000 × g and 4 °C for 10 min. Finally, the supernatant was injected into a Vanquish UHPLC system (Thermo Fisher, Germany) coupled with an Orbitrap Q ExactiveTM HF mass spectrometer (Thermo Fisher, Germany) at Biozeron Co., Ltd. (Shanghai, China)55.

The raw data files generated by UHPLC-MS/MS were processed using the Compound Discoverer 3.1 (CD3.1, Thermo Fisher) to perform peak alignment, peak picking, and quantitation for each metabolite. The main parameters were set as follows: retention time tolerance, 0.2 min; actual mass tolerance, 5 ppm; signal intensity tolerance, 30%; signal/noise ratio, 3; and minimum intensity, 100,000. After that, peak intensities were normalized to the total spectral intensity. The normalized data were used to predict the molecular formula based on additive ions, molecular ion peaks and fragment ions. And then peaks were matched with the mzCloud, mzVault and MassList database to obtain accurate and relative quantitative results. Statistical analyses were performed using the statistical software R (R version R-3.4.3), When data were not normally distributed, normal transformations were attempted using of area normalization method.

Detected metabolites were annotated using the KEGG database and HMDB. Principal component analysis and partial least squares discriminant analysis were performed with metaX (a flexible and comprehensive software for processing metabolomics data). We performed univariate analysis (t test) to calculate the statistical significance (P-value). The metabolites with variable importance in projection >1, P-value < 0.05 and FC ≥ 2 or FC ≤ 0.5 were identified as DEMs. The functions of these metabolites and their metabolic pathways were studied using the KEGG database. The metabolic pathway enrichment of DEMs was calculated: when the criterion x/n > y/N was satisfied, the metabolic pathway was considered enriched; when the P value of the metabolic pathway was <0.05, the metabolic pathway was considered significantly enriched.

Co-expression network analysis of RNA-seq and metabolome sequencing data

To investigate the associations between genes and metabolites, the Pearson correlation coefficient was computed for the expression levels of each gene and metabolite. Genes and metabolites with correlation coefficients ≥0.8 and ≤−0.8 were selected. Cytoscape software (version 3.9.1) was used to visualize the relationship between the metabolome and transcriptome.

Statistical analysis

The data are presented as the means ± standard deviations. Statistical analysis was conducted with a two-tailed unpaired Student’s t test or analysis of variance for multiple comparisons. A p value of less than 0.05 was considered to indicate statistical significance, whereas a p value of less than 0.01 was considered to indicate high significance.