Abstract

   Cryptotanshinone (CT) is an extract from the traditional Chinese
   medicine Salvia miltiorrhiza, which inhibits the growth of
   methicillin-resistant Staphylococcus aureus (MRSA) in vitro. This study
   aims to determine the antibacterial mechanisms of CT by integrating
   bioinformatics analysis and microbiology assay. The microarray data of
   [30]GSE13203 was retrieved from the Gene Expression Omnibus (GEO)
   database to screen the differentially expressed genes (DEGs) of S.
   aureus strains that were treated with CT treatment. Gene ontology (GO)
   and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway
   enrichment analyses were used to identify the potential target of CT.
   Data mining on the microarray dataset indicated that pyruvate kinase
   (PK) might be involved in the antimicrobial activities of CT. The
   minimum inhibition concentrations (MICs) of CT or vancomycin against
   the MRSA strain ATCC43300 and seven other clinical strains were
   determined using the broth dilution method. The effects of CT on the
   activity of PK were further measured. In vitro tests verified that CT
   inhibited the growth of an MRSA reference strain and seven other
   clinical strains. CT hampered the activity of the PK of ATCC43300 and
   five clinical MRSA strains. CT might hinder bacterial energy metabolism
   by inhibiting the activity of PK.

   Subject terms: Computational biology and bioinformatics, Microbiology

Introduction

   One of the largest concerns in public health is the continual emergence
   of multidrug-resistant bacterial pathogens, which severely limits
   treatment options. Staphylococcus aureus is a particularly problematic
   pathogen that is prevalent in human and animal populations. This
   organism commonly causes infections of the superficial skin, soft
   tissue, surgical wounds, and sometimes the bloodstream and lungs.
   Methicillin-resistant S. aureus (MRSA) and specific strains with
   reduced susceptibility to vancomycin could cause infections and
   diseases that are difficult to treat or resistant to the empirical
   antibiotics^[31]1. The global supply of antibiotics that are available
   for the treatment of infections that are associated with this
   microorganism is decreasing.

   Many studies have reported that medicinal herbs from different
   countries exhibited anti-MRSA activities, which was due to their
   phytochemical contents^[32]2. These plants could be employed as
   alternatives for drug development to stop, or control, or both MRSA
   infections.

   Cryptotanshinone (CT) is a fat-soluble extract from the traditional
   Chinese medicine Salvia miltiorrhiza (Danshen), which dilates blood
   vessels, and has antitumor and antiinflammatory activity^[33]3. In
   addition, CT inhibits the growth of S. aureus and MRSA in vitro^[34]4;
   however, the mechanism of action is unknown, which limits its further
   applications.

   Microarray platforms were recently employed to understand gene
   expression in bacteria that were treated by various Chinese medicines
   to identify some pathogen genes that are associated with antibacterial
   mechanisms^[35]5. In addition, the Gene Expression Omnibus (GEO)
   database offers methods for the downstream bioinformatics mining of
   gene expression profiles in a variety of bacteria.

   In this study, the differentially expressed genes (DEGs) were
   identified between strains of S. aureus that were treated with normal
   saline and CT by mining microarray datasets from the GEO database,
   which aimed to identify and confirm the mechanism associated with the
   antibacterial effects of CT.

Results

DEGs

   The volcano and heatmaps plots were generated to show the down and
   upregulated genes in the GEO datasets ([36]GSE13203). A total of 64
   overlapping DEGs (absolute log[2]FC > 1 and FDR < 0.05) were
   identified, which included 33 downregulated and 31 upregulated genes,
   respectively (Fig. [37]1). The top ten up and downregulated DEGs are
   listed in Table [38]2.

Figure 1.

   [39]Figure 1
   [40]Open in a new tab

   Screening of DEGs in MRSA induced by CT treatment: (a) volcano plots;
   and (b) hierarchical cluster analysis (heatmaps) of the common DEGs
   between CT treated samples and normal controls in GSE 13203.
   Statistically significant DEGs were defined with p < 0.05 and absolute
   log[2]FC > 1.0 as the cutoff threshold.

Table 2.

   Top10 up and downregulated genes in DEGs.
   Category Gene symbol Description log[2]FC Adjusted p-value
   Upregulated psuG Pseudouridine-5′-phosphate glycosidase 2.332 0.077
   cydB Cytochrome bd-I ubiquinol oxidase subunit II 2.029 0.011
   hrtB Hemin transport system permease protein 1.784 0.022
   cap5C Capsular polysaccharide biosynthesis protein Cap5C 1.780 0.063
   pflA Pyruvate formate-lyase activating enzyme 1.737 0.142
   cydA Cytochrome bd-I ubiquinol oxidase subunit I 1.653 0.018
   crtQ Zeta-carotene desaturase 1.647 0.018
   clfA MSCRAMM family adhesin clumping factor 1.640 0.018
   opuD2 Glycine betaine transporter 2 1.557 0.039
   crtM Dehydrosqualene synthase 1.534 0.022
   psuG Pseudouridine-5′-phosphate glycosidase 2.332 0.077
   Downregulated esaB Type VII secretion protein –2.607 0.011
   pyk Pyruvate kinase –2.285 0.164
   hutH Histidine ammonia-lyase –2.268 0.022
   isdB Iron-regulated heme–iron binding protein –2.106 0.025
   isdG Staphylobilin-forming heme oxygenase –2.006 0.020
   ribD Bifunctional diaminohydroxyphosphoribosylaminopyrimidine
   deaminase/5-amino-6-(5-phosphoribosylamino) uracil reductase –1.981
   0.574
   gltS Sodium/glutamate symporter − 1.975 0.015
   nuc Thermonuclease − 1.960 0.022
   modA Molybdate ABC transporter substrate-binding protein − 1.860 0.057
   [41]Open in a new tab

   FC fold change.

GO and KEGG pathway of DEGs enrichment analysis (Fig. [42]2)

Figure 2.

   [43]Figure 2
   [44]Open in a new tab

   Functional and signaling pathway analysis of the overlapped DEGs
   according to BP, CC, MF, and KEGG pathways.

   Enriched biological process (BP), cellular component (CC), and
   molecular function (MF) analyses were used to better understand the
   biological functions of overlapping DEGs. The results indicated that
   the significantly enriched GO term for BP was pathogenesis, the
   phosphoenolpyruvate (PEP)-dependent sugar phosphotransferase system
   (PTS) and translation. The significantly enriched GO terms for CC were
   the integral components of the membrane, plasma membrane, and
   cytoplasm. In addition, the significantly enriched GO terms for MF were
   metal ion binding, transmembrane transporter activity, and
   oxidoreductase activity.

   In addition, the results of the KEGG pathway analysis revealed that
   these DEGs were primarily enriched in metabolic pathways, biosynthesis
   of secondary metabolite ABC transporters, and two-component systems.

   Based on the results from the KEGG analysis, the effect of CT on S.
   aureus was concentrated on the metabolic pathways, moreover, the
   results of the GO-BP analysis indicated that the effect of CT on S.
   aureus concentrated on pathogenesis and the PEP-dependent sugar (PTS).
   PTS is a distinct system that is used by bacteria for sugar uptake when
   the energy source is phosphoenolpyruvate (PEP). In addition, PTS acts
   as a complex protein kinase system that regulates a wide variety of
   transport, metabolic, and mutagenic processes and the expression of
   numerous genes^[45]6. Among the DEGs, the pyk gene, which encodes PK
   was significantly downregulated. PK catalyzes the irreversible
   conversion of adenosine diphosphate (ADP) and PEP into adenosine
   triphosphate (ATP) and pyruvic acid, which are crucial for cellular
   metabolism. Therefore, PK plays a key role in controlling metabolic
   flux and ATP production. Meanwhile, as a highly conserved enzyme from
   animals to humans, PK has recently been identified as an essential gene
   for the survival of bacteria such as Haemophilus influenzae,
   Streptococcus pneumoniae, and Mycobacterium tuberculosis^[46]7–[47]10.
   In addition, it has been reported that MRSA is inhibited in vitro by
   hampering the activity of PK^[48]11–[49]13. Following CT action on S.
   aureus, PTS could detect changes in the surrounding environment that
   might further affect S. aureus metabolism via PK.

MICs of vancomycin or CT against the MRSA strains

   The MICs of vancomycin or CT against eight MRSA strains are given in
   Table [50]1 and Fig. [51]3. Among the seven clinical strains, five were
   derived from sputum and two were derived from blood, all of which were
   positive after a cefoxitin screening test. The MIC of vancomycin
   against ATCC43300 strain was 0.9 μg/mL, the MICs of vancomycin against
   seven other clinically isolated strains were between 0.45 and
   1.9 μg/mL. This reflects the status of vancomycin as the first-line
   treatment for MRSA since the 1950s. The MIC of CT against a standard
   strain ATCC43300 was 1.9 μg/mL, the MICs of CT against seven other
   clinically isolated strains were between 0.9 and 3.9 μg/mL. This result
   showed that CT displays bacteriostatic action against MRSA, which
   agreed with previous reports^[52]14, which suggested that CT could be
   used to fight MRSA infection.

Table 1.

   Specimen and tests information on MRSA strains.
   Strains Source Date of isolation (month and year) Antibiotic
   resistances Antibiotic sensitivity V MIC (μg/mL) CT MIC (μg/mL)
   MRSA01 Sputum 04/2017 P, L, E, O, CIP, M, T, CL G, TI, LINE, RI, V,
   TMP/SXT 0.45 0.9
   MRSA02 Blood 04/2017 P, L, E, O, G, CIP, M, T, CL TMP/SXT, TI, RI,
   LINE, V 0.9 0.9
   MRSA03 Sputum 05/2017 P, G, L, E, O, CIP, M, T, CL V, TMP/SXT, TI, RI,
   LINE 0.9 0.9
   MRSA04 Sputum 07/2017 P, L, E, O, CIP, M, T, CL TMP/SXT, TI, RI, V,
   LINE 1.9 1.9
   MRSA05 Blood 07/2017 P, L, E, O, CIP, M, T, CL TMP/SXT, TI, RI, LINE, V
   1.9 1.9
   MRSA06 Sputum 08/2017 P, L, E, G, O, CIP, M, T, CL TMP/SXT, TI, RI,
   LINE, V 1.9 1.9
   MRSA07 Sputum 08/2017 P, G, L, E, O, CIP, M, T, CL TMP/SXT, TI, RI,
   LINE, V 0.9 3.9
   ATCC43300 BJZY – – – 0.9 1.9
   [53]Open in a new tab

   P penicillin, L levofloxacin, E erythromycin, O oxacillin, CIP
   ciprofloxacin, M moxifloxacin, T tetracycline, CI clindamycin, G
   gentamicin, TI tetracycline, LINE linezolid, RI rifampin, V vancomycin,
   TMP/SXT trimethoprim–sulfamethoxazole, BJZY Beijing Zhongyuan Company,
   V MIC MIC of vancomycin to MRSA, CT MIC MIC of CT to MRSA.

Figure 3.

   [54]Figure 3
   [55]Open in a new tab

   MIC assay of MRSA strains: (a) MICs of CT to seven MRSA strains
   isolated from clinical specimens and standard MRSA strain ATCC 43300;
   and (b) MICs of vancomycin to seven MRSA strains isolated from clinical
   specimens and standard MRSA strain ATCC 43300.

The effect of CT on the activity of MRSA PK

   In this study, the impact of CT was examined on the activities of PK in
   a standard MRSA strain ATCC43300 and seven other clinical strains. As
   shown in Fig. [56]4, in ATCC43300, MRSA3, and MRSA6, CT at a
   concentration of double the MIC reduced the activity of PK (p < 0.05).
   However, CT at the concentration of the MIC did not show an effect on
   the activity of PK. In MRSA1 and MRSA4, CT at the concentration of MIC
   and double the MIC reduced the activity of PK significantly (p < 0.05),
   with no difference between both concentrations. In MRSA7, CT at the
   same concentration and double the MIC decreased the PK activity with a
   difference between both concentrations (p < 0.05). In MRSA2 and MRSA5,
   CT did not affect PK activity at either concentration.

Figure 4.

   [57]Figure 4
   [58]Open in a new tab

   Effect of CT on the activity of PK of eight MRSA strains. ^#Control
   group versus 1 × MIC group p < 0.05, *control group versus 2 × MIC
   group p < 0.05, **:1 × MIC group versus 2 × MIC groups p < 0.05.

Discussion

   The evolution of MRSA demonstrates its genetic adaptation into a
   first-class multidrug-resistant pathogen. Following the introduction of
   penicillin and methicillin, S. aureus rapidly developed resistance to
   these β-lactam compounds^[59]15. In the 1950s, vancomycin was used to
   treat MRSA infections, but recently the emergence of MRSA strains that
   are less sensitive to vancomycin (VISA) or even resistant to vancomycin
   (VRSA) means that clinicians are less confident when dealing with MRSA.
   Therefore, there is an urgent requirement to explore effective drugs
   for MRSA^[60]16. Plants that grow in the natural environment are
   affected by pathogenic microorganisms, such as bacteria and viruses.
   Some plants have developed two major strategies to defend against
   pathogenic microorganisms: (1) plants have a lot of epidermal hairs, a
   thick waxy stratum corneum, and small stomata to block the invasion of
   pathogenic microorganisms; and (2) plants produce antibiotic
   substances, such as tannins and alkaloids to inhibit or kill invading
   pathogenic microorganisms^[61]17. Therefore, potential antimicrobial
   chemicals could be found in plants to treat pathogenic microorganisms,
   such as bacteria.

   Salvia miltiorrhiza (Danshen) has been widely used in traditional
   Chinese medicine to treat a variety of diseases, which include coronary
   artery disease, acute ischemic stroke, hyperlipidemia, chronic renal
   failure, chronic hepatitis, and Alzheimer’s disease. In addition,
   Danshen has no serious adverse effects. The main bioactive constituents
   of S. miltiorrhiza include water-soluble phenolic acids and lipophilic
   diterpenoid tanshinones^[62]18. CT is one of the major tanshinones that
   were isolated from the roots of the Danshen plant. Recent studies have
   shown that CT has the potential to treat and prevent the previously
   mentioned diseases and it is a potent antibiotic agent^[63]19.

   In this study, the antibacterial effect of CT on MRSA was tested in
   vitro, and the MICs of CT on MRSA reference strain ATCC43300 and seven
   clinical strains were detected using a two-fold serial dilution method.
   The results showed that CT displayed bacteriostatic action against
   MRSA, which was consistent with previous reports^[64]14, which suggests
   that CT could be used to fight MRSA infection.

   Although thousands of herbal compounds have been listed as
   antimicrobial phytochemicals^[65]20, limited understanding of the
   mechanisms limits the application of these substances. To determine the
   mechanism of CT inhibition on the growth of MRSA, bioinformatics tools
   were used. In this study, the data from was retrieved from [66]GSE13203
   and systematically analyzed the differential gene expression of S.
   aureus were systematically analyzed between the CT treatment and
   control. Therefore, 64 DEGs were identified, which included 33
   downregulated and 31 upregulated genes.

   Then, GO and KEGG analysis of these DEGs revealed that CT had a greater
   impact on the metabolism of S. aureus. Because MRSA is a special
   species of S. aureus that carries a multidrug resistance gene, the
   metabolic pathways of MRSA and S. aureus are similar. As a facultative
   anaerobe, S. aureus uptakes a variety of nutrients that include
   glucose, mannose, mannitol, glucosamine, N-acetylglucosamine, sucrose,
   lactose, galactose, and beta-glucosides. The central pathways for
   glucose metabolism are the Embden–Meyerhof–Parnas pathway and the
   pentose phosphate cycle. Lactate is the end product of anaerobic
   glucose metabolism and acetate, and CO[2] are the products of aerobic
   growth conditions^[67]21. In the glucose metabolic pathway, PK is the
   rate-limiting enzyme of energy metabolism, which plays a central role
   in the carbohydrate metabolism of MRSA. It catalyzes the final
   rate-limiting step of glycolysis. During this irreversible process, the
   high-energy phosphate bond in the PEP molecule is transferred to ADP to
   generate ATP. The metabolites PEP and pyruvate are related to other
   biosynthetic pathways, and therefore, the potential inhibition of PK
   could obstruct MRSA energy metabolism. The potential inhibition of PK
   might result in the decreased metabolism of MRSA^[68]13,[69]22–[70]24.

   A PK test kit was used to detect the effects of CT on the activity of
   PK of MRSA in vitro. The results showed that except for in the MRSA2
   and MRSA5 groups, CT hampered the PK activity of MRSA. PK was
   identified as a highly interconnected essential hub protein in MRSA,
   with structural features distinct from human homologs. Currently, the
   majority of antibiotics in use are directed at critical proteins that
   are unique to the bacteria and without human homologs to avoid
   mechanism based toxicity. In addition, based on the supposition that
   hub proteins are critical for bacterial survival and they should be
   very sensitive to mutations^[71]25, and therefore, targeting them
   should reduce the potential to develop resistant strains and species.
   Therefore, CT could be a valuable antibacterial candidate because of
   its effect on the PK of MRSA.

   In this study, CT did not affect PK activity in MRSA2 and MRSA5, which
   were collected from a blood specimen and the five other strains were
   isolated from a sputum specimen. In different environments, S. aureus
   will utilize a variety of metabolic pathways, which depend on the
   oxygen content, source of carbohydrates, and other factors^[72]26.
   However, the MIC of CT was 0.9 μg/mL for MRSA2, and 1.9 μg/mL for
   MRSA5, which indicated that CT might have additional mechanisms that
   affect the growth of MRSA in addition to affecting the PK activity of
   both strains. As shown in Table [73]2, in addition to the pyk gene, a
   significant downregulation of isdB and isdG were observed, both of
   which belong to the iron-regulated surface determinant family. The
   GO-MF analyses revealed that CT affected the metal ion binding of S.
   aureus. Iron is vital for the growth and proliferation of nearly all
   organisms, including MRSA. In addition, iron is required for the
   colonization of host tissues by MRSA and subsequent
   pathogenesis^[74]27. Therefore, in addition to affecting PK activity,
   CT might inhibit MRSA growth by interfering with its iron metabolism,
   however, this requires further investigation.

   The most significant limitation of this study was that only changes in
   the activities of PK MRSA were examined after treatment with CT. Future
   research should include measuring the expression of PK of MRSA after
   treatment with CT. In addition, only seven MRSA strains were isolated
   from a clinical environment; therefore, a larger number of samples is
   required in future research to verify the antimicrobial mechanism of CT
   on MRSA.

   In summary, CT inhibited the growth of MRSA in vitro. To determine the
   mechanism of activity, multiple bioinformatics tools were used combined
   with a comprehensive analysis of gene expression profiles to identify
   the key signaling pathways. CT interfered with the activities of PK,
   which is the key rate-limiting enzyme during glycolysis in MRSA. The
   results of this study might provide new insights into the antibacterial
   mechanisms of CT. However, further research and studies with larger
   sample sizes are required to confirm these findings.

Methods

Acquisition of gene expression profiling data

   The microarray data for [75]GSE13203 that was deposited by Feng et
   al.^[76]28 was retrieved from the GEO
   database([77]http://www.ncbi.nlm.nih.gov/geo/). The gene expression
   profile was generated using the Microarray Analysis Suite 5.0
   (Affymetrix, Santa Clara, CA, USA). [78]GSM333383, [79]GSM333384, and
   [80]GSM333385 were used for S. aureus specimens that were treated with
   CT for 45 m (intervention group). [81]GSM333386, [82]GSM333387, and
   [83]GSM333388 were used are for the S. aureus specimens that were
   treated with saline for 45 m (control group).

Raw data preprocessing and screening of DEGs

   First, the chip data files were downloaded and the gene name
   information, sample number, and value were copied into a Microsoft
   Office Excel 2019 table (the Excel table cell format selected the text
   format) to sort the data and delete incomplete data, followed by saving
   all data as "input.txt". Then, gene expression differences between the
   CT treated samples and the control group were identified using the
   limma R package with the Empirical Bayes method^[84]29. The
   Benjamini–Hochberg false discovery rate (FDR) was used to correct for
   p-value. In this study, adjusted p-value < 0.05 and absolute Log[2]fold
   change (FC) > 2.0 were set as the cutoff criteria to screen DEGs.
   Furthermore, a volcano map was plotted with the ggplot2 package
   according to the adjusted p-value and log[2]FC. In addition,
   hierarchical clustering analysis of DEGs was performed and visualized
   using the pheatmap package in R Language.

Functional and pathway enrichment analysis

   To explore the biological functions and the pathways involved in the
   significant up and downregulated DEGs, Gene Ontology (GO) and Kyoto
   Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were
   performed using the clusterProfiler package in R language^[85]30. GO is
   a comprehensive database that describes gene functions in three parts:
   BP, CC, and MF. GO function enrichment uses p-values < 0.05 as the
   threshold for significant enrichment. KEGG is a comprehensive database
   that integrates genomic, chemical, and system function
   information^[86]31. KEGG pathway enrichment uses p-values < 0.05 as the
   cutoff criteria for significant enrichment.

Bacteria strains and materials

   The MRSA reference strain ATCC43300 was purchased from Zhongyuan
   Company (Beijing, China). The seven clinical MRSA strains used in this
   study were isolated from the Microbiology Laboratory, Jiangsu
   Provincial Hospital of Traditional Chinese Medicine (Nanjing, Jiangsu
   Province, China) (Table [87]1). Mueller–Hinton Broth II (MHB II) and
   Mueller–Hinton agar were purchased from BD Biosciences (Sparks Glencoe,
   MD, USA). Stock solutions of varying concentrations were dissolved in
   dimethylsulfoxide (DMSO), which were purchased from Sigma-Aldrich
   (Shanghai, China).

MICs

   The minimum inhibitory concentrations (MICs) of CT or vancomycin
   against MRSA strain ATCC43300 and seven other clinical strains were
   determined in triplicate by broth microdilution or broth macrodilution
   using two-fold serial dilutions in MHB II, according to CLSI/NCCLS
   M100-S28^[88]32. The MICs were defined as the lowest concentration at
   which no visible growth was observed.

The effect of CT on the activity of MRSA PK

   PK activity was determined using a continuous assay coupled with
   lactate dehydrogenase (LDH), and the change in absorbance at 340 nm
   that was caused by NADH oxidation was measured using a SmartSpec Plus
   spectrophotometer (Bio-Rad, Laboratories (Shanghai), Hercules, CA,
   USA). The CT was diluted according to the MIC concentration of each
   MRSA strain and double the MIC concentration, with a control group of
   no CT treatment samples. In a 96-well plate, 50 μL of diluted CT was
   added in triplicate for each concentration. MRSA cells were harvested
   in the logarithmic growth phase and were diluted to a 0.5 turbidity
   standard bacterial suspension (1.5 × 10^8 cfu/mL). A 96-well plate that
   contained 50 μL of bacteria suspension diluted in MHB II 1:500 in each
   well were mixed gently and placed in a 37 °C constant temperature
   incubator for 12 h. After the 96-well plate was removed from the
   incubator, 1 × 10^5 cfu bacteria respectively were collected and
   centrifuged at 3000 rpm, at 4 °C for 5 m. After removing the
   supernatant, 100 μL of extract fluid (BC0540, Solarbio Science &
   Technology, Beijing, China) was added to the pellets and the bacterial
   cells were broken in an ultrasonic cell disruptor (Xinzhi JY92-IIN,
   Saide Electic, Hangzhou, Zhejiang Province, China). After centrifuging
   at 8000g, at 4 °C for 10 m, 30 μL of the supernatant was incubated with
   reagents (BC0540, Solarbio Science & Technology, Beijing, China) to
   initiate the reactions. PK activity proportional to the rate of change
   at 340 nm was expressed as specific activity (μmol/min/mg), which is
   defined as the amount of PK that catalyzes the formation of 1 μmol of
   either product per minute.

Author contributions

   J.W.Z. and Q.S. designed the experiments and wrote the manuscript.
   J.W.Z. performed the bioinformatic analysis. H.D.W. performed the in
   vitro experiments; Y.Z. collected and analyzed the data. All authors
   reviewed the manuscript.

Data availability

   The datasets generated during and/or analyzed during the current study
   are available from the corresponding author on reasonable request.

Competing interests

   The authors declare no competing interests.

Footnotes

   Publisher's note

   Springer Nature remains neutral with regard to jurisdictional claims in
   published maps and institutional affiliations.

References