Characterization of a small non-coding RNA S612 in Bacillus subtilis

Abstract

Small regulatory RNAs (sRNAs) are non-coding RNA molecules that fine-tune various cellular processes and respond to various environmental stimuli. In Bacillus subtilis, the regulatory mechanisms and specific targets of several sRNAs remain largely unknown. In this study, we identified and characterized S612 as a self-terminating sRNA in B. subtilis. The expression of S612 is regulated by external signals, including nutrient availability and salt concentration. Overexpression of S612 induced filamentous cells with extensive cellular elongation and complete inhibition of sporulation, indicating its potential to control cell morphology and spore formation. S612 directly targets and downregulates genes through post-transcriptional base pairing with mRNAs, including ylmD, trpE, ycxC, yycS, rapH, and amyE, some of which are involved in cell membrane integrity, cell wall synthesis, and sporulation initiation. Therefore, we propose that S612 is an important post-transcriptional regulator of cell morphology and sporulation.

Graphical abstract

1. Introduction

Small regulatory RNAs (sRNAs), typically 50–300 nucleotides in length, are key players in bacterial gene regulation. By modulating metabolic pathways and stress responses, sRNAs ensure cellular adaptability to dynamic environments [1,2]. The regulatory mechanisms of sRNAs span from the transcriptional to translational levels, with most sRNAs regulating post-transcriptionally [3,4]. In addition, some sRNAs directly interact with proteins [[5], [6], [7], [8]].

Previous research on sRNAs has primarily focused on Gram-negative bacteria such as Escherichia coli and Salmonella enterica [9,10]. However, in Bacillus subtilis, the mechanisms and targets of nearly 100 sRNAs remain unknown [11,12]. In B. subtilis, the identified sRNAs are involved in the tricarboxylic acid (TCA) cycle [[13], [14], [15]], transmembrane transportation [14], toxin-antitoxin (TA) system [16,17] and other processes. Most of the identified sRNAs act post-transcriptionally by base pairing with target mRNAs. The stability or translational efficiency of mRNAs varies after binding to sRNAs [13,18] and is sometimes mediated by RNase [19] and RNA-binding proteins (RBPs) [9,20,21]. The types of encoded RNase and RBP differ between Gram-positive and Gram-negative bacteria [9,19,20,22]. In B. subtilis, RNase Y and J1 play major roles in controlling the abundance of mRNA and sRNA [[23], [24], [25]]; Hfq [[26], [27], [28], [29], [30]] and CsrA [31,32] facilitate base pairing and affect the abundance of mRNA and sRNA. Understanding the mechanisms of sRNA is essential for advancing our knowledge of microorganisms and developing regulatory tools [[33], [34], [35]].

In this study, we characterized an sRNA from B. subtilis, named S612. S612 is expressed in response to environmental stresses such as high concentrations of glucose, nitrogen, or salt. Cells of B. subtilis overexpressing S612 were over 10 times longer than normal cells and failed to produce spores after 48 h of culture. S612 targets several genes critical for cell membrane function, cell wall formation, and sporulation initiation. This study provides new insights into the regulatory roles of sRNAs in B. subtilis and highlights S612 as a potential modulator of filamentation and sporulation.

2. Experimental procedures

2.1. Strain and plasmid construction

B. subtilis wild type was applied in this study. The endogenous sRNA S612 was knocked out to construct the B. subtilis Δs612-surX via CRISPR/Cpf1 [36]. The potential S612 target gene ylmD was knocked out to construct the B. subtilis ΔylmD using CRISPR/Cpf1. Donor sequences and sgRNAs were incorporated into the plasmid pcrF19NM. The sgRNAs and donor sequences used in this study are listed in Table S4. E. coli strains JM109 and Top 10 were used for plasmid construction and amplification, respectively. All plasmids and primers used in this study are listed in Tables S1 and S2, respectively. sRNA expression plasmids were constructed using p43nmk (kanamycin resistance) as the template. The p43 promoter was used to overexpress S612 and its mutant, whose activity peaked at approximately 8–12 h [[37], [38], [39]]. gfp expression plasmids were constructed using pHT01 (ampicillin-resistant in E. coli and chloramphenicol-resistant in B. subtilis) as the template. The pSpoVT promoter was utilized to regulate the transcription of GFP, which was fused with the designated 5′ UTRs of predicted S612 target genes [40,41]. The alteration in fluorescence serves as an indicator of the 5′ UTR regulated by S612. DNA fragments were conjugated using a one-step cloning kit (YEASEN). The linearized plasmids were amplified via polymerase chain reaction (PCR) using PrimeSTAR® Max DNA Polymerase (Takara). The primers used for the PCR are listed in Table S3. PCR products were self-cyclized after transformed into E. coli JM109 or Top10 using endogenous DNA recombinase. The sequences of the recombinant plasmids were confirmed by DNA sequencing.

2.2. Medium and cultivation

Luria-Bertani (LB) medium (10 g/L tryptone, 10 g/L NaCl, and 5 g/L yeast extract) was used for routine cell cultivation. Difco sporulation medium (DSM) [42] consisted of 5 g/L peptone, 3 g/L meat extract, 1 g/L KCl, 1 mM MgSO₄, 1 × 10⁻² mM MnCl₄, 1 mM CaCl₂, and 1 × 10⁻² mM FeSO₄. The culture media shown in Fig. 3B, C, are variants of DSM obtained by adding designated concentrations of NaCl, glucose, or NH₄Cl. The following antibiotics were added when necessary: chloromycetin, 5 μg/mL; kanamycin, 50 μg/mL; ampicillin, 50 μg/mL. To measure the growth of the strain overexpressing S612, cells were pre-cultured in 96-well plates containing 200 µL of DSM medium at 37 °C for 12 h. Subsequently, the cells were transferred to fresh 96-well plates with 200 µL of DSM medium, and the initial OD₆₀₀ was adjusted to 0.1. In the shake flask experiments, cells were pre-cultivated in 50 mL tubes containing 5 mL LB medium at 37 °C for 12 h and then transferred to a designated medium (25 mL) with an initial OD₆₀₀ of 0.1. Where appropriate, media were supplemented with antibiotics at the following concentrations: 50 µg/mL ampicillin in E. coli; 5 µg/mL chloramphenicol in B. subtilis; 50 µg/mL kanamycin in E. coli or in B. subtilis.

Fig. 3.

Characterization of S612 promoter. The details of the medium used are shown in Section 2.2. A. Activity of pS612 in the LB medium. B. Activity of pS612 in DSM containing 0 M NaCl and 0.85 M NaCl. C. Activity of pS612 in DSM containing 0 % glucose and 2 % glucose. Both cultures contain 0.085 M NaCl. D. Activity of pS612 in DSM containing 0 % NH₄Cl and 1 % NH₄Cl. Both cultures contain 0.085 M NaCl.

2.3. Florescence measurement

The strains were grown in 50 mL tubes containing 5 mL LB medium at 37 °C for 12 h and then transferred to 500 mL flasks containing 25 mL LB medium with a starting OD₆₀₀ of 0.1. After 12 h of cultivation in flasks, 1 mL of the cultivated medium was removed as a sample from every flask in 1.5 mL tubes. After centrifugation at 4500 rpm for 5 min, the cells were collected, washed, resuspended in distilled water and diluted 10-fold. The diluted cells in each tube were added to a 96-well plate. The fluorescence intensities of GFP (excitation, 490 nm; emission, 530 nm), mKate (excitation, 588 nm; emission, 635 nm), and YFP (excitation, 490 nm; emission, 527 nm) were measured using an Infinite 200 PRO [43,44].

2.4. Plasmid electro-transformation into B. subtilis

The overnight culture of B. subtilis was diluted 16-fold in a growth medium (LB medium containing 0.5 mol/L sorbitol) and grown at 37 °C to an OD₆₀₀ of 0.85–0.95. The cells were cooled on ice-water for 10 min and harvested by centrifugation at 4 °C and 5000 rpm for 5 min. The cells were washed four times with ice-cold washing solution (0.5 mol/L sorbitol, 0.5 mol/L mannitol, and 10 % glycerol). Then, they were suspended in 1/40 of the culture volume using washing solution. The competent cells were stored at –80 °C until used with some decrease in transformation efficiency. We added 5 µL (200–1000 µg/µL) plasmid DNA to 90 µL of competent cells. The plasmid and cell mixture was homogenized by gently mixing several times with a pipette. The mixture was then transferred to a pre-chilled cuvette. The cuvette was incubated for 1–1.5 min. The moisture was wiped from the cuvette and inserted into the device (Gene Pulser Xcell™ Total System). We immediately added 1 mL outgrowth medium (LB medium containing 0.5 mol/L sorbitol and 0.38 mol/L mannitol) and incubated the mixture for 3 h at 37 °C. The outgrowth medium was centrifuged at 5000 rpm for 5 min. The precipitation was plated onto selective LB agar plate and incubated overnight at 37 °C.

2.5. Statistical analysis

Statistical analyses were performed using GraphPad Prism, version 9 (GraphPad Software). All data were measured with 3–6 biological replicates. The experimental parameters are listed in the corresponding figure legends.

2.6. Sporulation efficiency measurement

The cells were cultured in 25 mL DSM at 37 °C, 220 rpm for 48 h. One milliliter of the culture was incubated at 80 °C for 20 min. The incubated culture was serially diluted 8–10 times. We spread 100 μL of the diluted culture on a plate and cultured it at 37 °C overnight. The number of colonies was counted using CountThings Software.

2.7. Total RNA extraction and cDNA synthesis

Total RNA and RNA derivatives were extracted from B. subtilis after 12 h of culture. Detailed methods followed the protocols provided with the DNAaway RNA Mini-Preps Kit EZ-10 (Sangon Biotech) and the miRNA 1st Strand cDNA Synthesis Kit MR101–01 (Vazyme).

2.8. RT-qPCR detection of gene expression levels

qPCR primers were designed to amplify cDNA fragments of 80–100 nucleotides in length using Primer-Blast (https://www.ncbi.nlm.nih.gov/tools/primer-blast). The detailed methods are described in the product manual of the Taq Pro Universal SYBR qPCR Master Mix Q712 (Vazyme). The reference genes selected were clsA and rpoB [45].

3. Results

3.1. S612 is a self-terminated small non-coding RNA affecting cell morphology and sporulation in B. subtilis

To investigate sporulation-related sRNA in B. subtilis, we overexpressed six uncharacterized endogenous sRNA candidates based on their base-pairing genes or reported functions (Table S5). The sporulation efficiencies of the six sRNA overexpression strains were measured (Fig. S1). Among these strains, overexpression of S612 completely inhibited sporulation in B. subtilis.

Northern blotting analysis conducted by Irnov et al. identified S612 as a 150-nucleotide transcript [12]. The secondary structure of S612, predicted by RNAfold [[46], [47], [48]], revealed a strong Rho-independent terminator (a genomic sequence encoded at the 3′ end of the transcription unit, characterized by a GC-rich repeat followed by a T-tract [49]) (Fig. 1A). To verify the termination efficiency (TE) of S612, we inserted s612 between mKate and yfp into a plasmid and calculated the TE value by measuring the fluorescence intensity of mKate and YFP (Fig. S2) [50]. The termination efficiency of S612 was approximately 90 % (Fig. 1B), which corroborates that S612 can be terminated by the intrinsic terminator encoded at its 3′ terminal.

Fig. 1.

Structural analysis of S612. A. Predicted secondary structure and self-termination of S612. The secondary structure was predicted based on NUPACK and RNAfold based on minimum free energy (MFE) and partition function. The transcription starting site and length of transcript were referred to northern blot conducted by Irnov et al. [12]. B. Self-termination efficiency of S612. The termination efficiency of S612 was tested by measuring the fluorescence of mKate and YFP, after adding sRNA between the two fluorescent proteins on plasmid. There was no sRNA added between mkate and yfp in the control group. The formula used to calculate termination efficiency is provided in Supplementary Fig. 2.

To further investigate the function of S612 in B. subtilis, we measured the growth curve of the S612 overexpression strain and observed a slight growth retardation in the S612 overexpression strain (Fig. 2A). Microscopic observations revealed that the S612 overexpression strain exhibited extreme elongation, with almost no spores detected after culturing for 48 h (Fig. 2B, C). Electron microscopy confirmed the presence of filamentous cells (Fig. 2C). Overall, these results led us to conclude that S612 is a self-terminated small non-coding RNA that affects cell morphology and sporulation in B. subtilis. However, the mechanism through which S612 overexpression induces these physiological changes remains unclear.

Fig. 2.

S612 modulates the growth and shape of B. subtilis.A. Growth curve of S612 overexpression strain (Bs168/p43nmk-s612) compared with that of the control strain (Bs168/p43nmk). Cells were cultured in 96-well plates and monitored dynamically using a plate reader. B. Sporulation measurement of S612 overexpression strain and contrast after a 48-h growth. The overexpression of S612 completely inhibited sporulation. C. Microscopy of S612 overexpression strain and control after a 48-h growth. The differences between normal and filamentous cells are shown in both images.

3.2. The S612 expression is subject to the regulation of environmental signals

To investigate the regulation of S612 transcription, we fused the S612 promoter, located 500 bp upstream of the transcription start site, to gfp, and measured bacterial growth and fluorescence properties in the designated media (Fig. S3). In the LB medium, the S612 promoter exhibited almost no transcriptional activity (Fig. 3A). However, the promoter was activated in the defined medium, DSM. This indicates that the S612 promoter is subject to nutrient supply regulation. According to predictions from the Database of Transcriptional Start Sites (DBTBS) [44,45], S612 transcription is likely regulated by the sigma factor SigA and the transcriptional regulator DegU (Fig. S4). Previous studies have shown that DegU can modulate gene expression in response to environmental stresses, including high concentrations of glucose [46], salt [47,48], and nitrogen [49]. To further assess whether the S612 promoter is responsive to additional environmental signals, we measured GFP fluorescence under DSM conditions supplemented with 0.85 M NaCl, 2 % glucose, or 1 % NH4Cl (Fig. 3B, C, and D). We observed that the addition of a high concentration of glucose suppressed S612 promoter activity and that high concentrations of NaCl suppressed S612 promoter activity during the exponential growth phase.

Overall, our analysis suggests that the expression of S612 is regulated by environmental signals, particularly the availability of nutrients and the presence of stress-inducing factors such as high glucose and salt concentrations.

3.3. S612 is nonessential and its function is independent of SurX

Analysis of the genomic location of s612 revealed another independently expressed sRNA (Fig. 4A), SurX, located in the base-pairing region of s612 [51]. This finding led us to consider whether the observed physiological changes could be attributed to SurX rather than S612. To rule out the potential confounding effect of SurX, we knocked out the s612-surX in the genome using CRISPR/Cpf1 and constructed B. subtilis Δs612-surX [36]. Subsequently, S612 and SurX were overexpressed separately in B. subtilis Δs612-surX. RT-qPCR results showed that S612 expression was higher than that of SurX in the wild-type strain, whereas SurX expression was negligible in the S612-overexpression strain (Fig. 4B). Microscopic observations revealed that the overexpression of S612 resulted in filamentous cells of B. subtilis, whereas the overexpression of SurX had no effect on cell morphology (Fig. 4C). This result excludes SurX as a contributing factor to the observed physiological changes. Upstream of surX is ymzD, and the predicted SurX promoter resides within the ymzD coding sequence. This raised the question of whether SurX functions as the 3′ UTR of ymzD and whether S612 affects ymzD expression. Although predictions indicated only a very slight transcriptional anti-correlation between ymzD and S612 [11], qPCR analysis revealed significantly lower ymzD expression in the S612-overexpression strain. Based on these results, we conclude that the function of S612 is not influenced by SurX and ymzD might be one of the targets of S612. The S612 genome is shown in Fig. 4D.

Fig. 4.

Analysis of S612 genomic context. A. Genomic context of S612. The antisense strand locates identified non-coding RNA surX. Upstream of surX is putative integral inner membrane protein ymzD.B. Relative expression level measurement of S612, SurX, and ymzD mRNA by RT-qPCR. The relative expression levels were measured both in Bs168 and Bs168/p43-s612.C. Microscopy of Bs168 Δs612-surX/p43nmk, Bs168 Δs612-surX/p43nmk-s612, and Bs168 Δs612-surX/p43nmk-surX. The strains were cultured for 12 h. D. Sequence of s612 and ymzD on genome. The transcription starting site of S612, the bidirectional terminator of S612, and the start and termination codon of ymzD are marked out.

3.4. S612 regulates multiple genes

Target prediction using CopraRNA (http://rna.informatik.uni-freiburg.de/CopraRNA/Input.jsp) [52,53] identified numerous potential target candidates. Seven putative targets were confirmed in a previous proteomic study (Table 1) [54]. Interaction analysis between sRNA and target mRNA using IntaRNA (http://rna.informatik.uni-freiburg.de/IntaRNA/Input.jsp) [55] predicted that all these base-pairing regions are located in the 5′ untranslated region (UTR) and the translation initiation region (TIR) (Fig. 5). We constructed a GFP-based reporter plasmid containing the 5′ UTR and TIR sequences of seven potential S612 target genes [[56], [57], [58]] (Fig. 6A). These plasmids were transformed into B. subtilis wild type and B. subtilis Δs612-surX. The results revealed that the GFP reporter with the 5′ UTR sequences of ycxC, yycS, and ylmD were obviously repressed in the wild type strain (Fig. 6B).

Table 1.

Genes regulated by S612.

Gene name	Annotation	Hybridization energy (kcal/mol)
trpE	Biosynthesis of tryptophan	−26.03
ylmD	Peptidoglycan editing factor	−29.57
rapH	Response regulator aspartate phosphatase H. Control of sporulation initiation and ComA activity	−10.2
ycxC	Putative permease	−8.68
yycS	Putative lipoprotein	−10.14
amyE	Alpha-amylase-Starch degradation	−12.34

Fig. 5.

Prediction of RNA duplex formation between selected target mRNAs and S612. Arrows indicate the S612 mutations tested in Fig. 7. Numbers indicate the base-pairing region. For example, the base-pairing region between S612 and yycS is located 8 to 1 bp upstream of the yycS translation starting site and downstream of the S612 transcription starting site (91 to 100 bp).

Fig. 6.

Screening of S612 targets. A. The schematic diagram of the GFP-based reporter plasmid and mechanism of how S612 inhibits gfp expression. The GFP reporter plasmids contain 5′ UTR of S612 target sequences, including pHT01-amyE 5′UTR-gfp, pHT01-ylmD 5′UTR-gfp, pHT01-rapH 5′UTR-gfp, pHT01-ycxC 5′UTR-gfp, pHT01-yycS 5′UTR-gfp, pHT01-ycbP 5′UTR-gfp, and pHT01-trpE 5′UTR-gfp. B. The measurement of relative fluorescence/OD₆₀₀ in Bs168 Δs612-surX and Bs168 wild type.

We also observed that the intensities of GFP fluorescence with some other 5′ UTRs were slightly repressed. We hypothesized that an increased abundance of S612 transcripts would lead to significant inhibition of GFP fluorescence. To test this, gfp plasmids containing 5′ UTR of predicted target genes were co-transformed into B. subtilis Δs612-surX with either an S612 overexpression plasmid or a control plasmid. Six specific targets, namely, trpE, ylmD, rapH, ycxC, yycS, and amyE, were repressed (Fig. 7). Using compensatory base-pair exchange experiments, the S612 mutants (M1, M2, M3, and M4) were generated (Fig. 5) [56]. S612 mutants showed significantly reduced GFP inhibition in vivo compared to wild-type S612 (Fig. 7), which further validated binding at the predicted positions. Together, these results demonstrate that S612 is a trans-acting sRNA that inhibits the expression of several genes.

Fig. 7.

Compensatory base-pairing exchange verification. A. The plasmids applied in this experiment. B. The measurement of fluorescence/OD₆₀₀ in Bs168 Δs612-surX/pSpoVT-5′ UTR-gfp containing p43nmk, p43nmk-s612, and p43nmk-s612 mutants.

3.5. S612 acts post-transcriptionally

To determine whether the targets were downregulated at the translational or post-transcriptional level, reverse transcription quantitative polymerase chain reaction (RT-qPCR) was performed in the S612 overexpression strain, compared with the wild-type strain. Analysis of the relative expression levels showed that the mRNA levels of trpE, ylmD, ycxC, yycS, and amyE were downregulated by two-fold or more in the S612 overexpression strain (Fig. 8). In contrast, the mRNA abundance of rapH and ycbP showed no significant changes. These findings indicate that S612 primarily regulates these predicted targets at the post-transcriptional level.

Fig. 8.

Relative expression measurement by RT-qPCR. The relative expression levels of predicted S612 target genes were measured in Bs168 WT and Bs168/p43-s612 using RT-qPCR.

3.6. The physiological function of S612 in B. subtilis

Based on the annotation of the S612 target genes, we hypothesized that YcxC and YycS are involved in cell membrane formation. The tertiary structures predicted by AlphaFold and the transmembrane helices predicted by TMHMM [59] indicated that YcxC was located on the cell membrane (Fig. S6). Thus, YcxC may play a critical role in maintaining cell membrane integrity.

A previous study revealed that ylmD lies in the Spo0A operon in B. subtilis, and its expression is repressed under conditions that trigger sporulation [60]. It has also been shown that ylmD encodes a peptidoglycan-editing factor located in the division and cell wall clusters of many Gram-positive organisms [61]. We knocked out ylmD in B. subtilis and observed longer cells than the wild-type strain but not as long as S612 overexpression strain (Fig. S5). Knockout of ylmD homologs in E. coli (yfiH) also results in longer cells, but only in the early exponential growth phase [61]. The role of ylmD in sporulation-specific cell division has previously been investigated in Streptomyces [62]. However, in B. subtilis, the role of ylmD in endospore formation remains unclear [63]. Based on our results and those of previous studies, we hypothesized that the inhibition of ylmD is a crucial factor in filamentation (Fig. 8).

The overexpression of S612 led to the absence of sporulation after culturing for 48 h in DSM (Fig. 2B and S1), and no spores were observed under the microscope (Fig. 2B). However, the mechanism through which S612 inhibits sporulation remains unclear. Although rapH has been shown to control sporulation initiation, it plays a negative role by dephosphorylating Spo0F, causing phosphoryl groups to flow away from Spo0A, and inhibiting sporulation (sporulation is triggered when Spo0A∼P reaches a certain level) [64,65]. However, this does not explain the mechanism by which S612 inhibits sporulation. We hypothesized that the regulation of sporulation by S612 could be a coordinated effect of downregulation of several targets. However, the relationship between S612 and sporulation requires further research. Fig. 9

Fig. 9.

Summary of speculated regulatory function of S612 in B. subtilis. The red arrows indicate inhibition and the black arrows indicate activation.

4. Discussion

In this study, we characterized S612 as a trans-acting sRNA that influences sporulation and cell morphology. S612 was independently expressed, with a self-termination efficiency greater than 90 %. The base-pairing region is another independently expressed sRNA, SurX, which has no direct correlation with S612 in its physiological functions. However, the mRNA expression levels of SurX and ymzD mRNA were lower in the S612 overexpression strain, suggesting that ymzD may be an S612 target. As the SurX transcript was not been verified by northern blotting, the specific base-pairing region between SurX and S612, as well as whether SurX is on the same transcript as ymzD requires further investigation.

The S612 promoter was predicted to contain SigA and DegU binding sites. The S612 promoter showed low activity in LB medium, but high activity in DSM, indicating that it was induced by external signals. The S612 promoter was weaker under environmental stressors such as glucose, nitrogen, and salt, which trigger DegU expression. In conclusion, certain nutrients are important regulators of S612 levels. However, further experimental validation through an Electrophoretic Mobility Shift Assay (EMSA) is needed to investigate the interaction between the S612 promoter and its protein regulator.

Six targets were regulated by S612 post-transcriptionally. The known functions of these targets are related to cell membrane composition, cell wall synthesis, and sporulation initiation, which could correlate with the mechanisms of sporulation and filamentation after S612 overexpression. However, the specific functions of ycxC, yycS, and ylmD in B. subtilis remained to be elucidated. Additionally, the mechanism through which S612 affects sporulation remains unclear.

Overall, by studying endogenous S612 sRNAs, we established essential foundations for the investigation of S612. Confirmation of sRNA-target interactions is crucial for understanding sRNA functions. Although >100 sRNAs have been identified in B. subtilis, only a few target interactions have been validated. Furthermore, the potential of S612 to regulate sporulation in B. subtilis is promising for industrial production, as sporulation reduces production in the late stages of fermentation.

Data Availability Statement

All data generated or analyzed during this study are included in this published article and its supplementary information files or are available upon request.

CRediT authorship contribution statement

Anqi Peng: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Weijiao Zhang: Writing – review & editing, Validation, Supervision. Haibo Xiong: Investigation. Luyao Zhang: Investigation. Jian Cheng: Investigation. Yang Wang: Writing – review & editing, Supervision, Data curation, Conceptualization. Zhen Kang: Writing – review & editing, Supervision, Funding acquisition, Conceptualization.

Declaration of Competing Interest

This manuscript has not been published or presented elsewhere in part or in entirety and is not under consideration by another journal. We have read and understood your journal's policies, and we believe that neither the manuscript nor the study violates any of these. There are no conflicts of interest to declare.