Target-based screening for natural products against Staphylococcus aureus biofilms

Abstract

As a notorious food-borne pathogen, Staphylococcus aureus can readily cause diseases in humans via contaminated food. Biofilm formation on various surfaces can increase the capacity of viable S. aureus cells for self-protection due to the stubborn structure of the biofilm matrix. Increased disease risk and economic losses caused by biofilm contamination in the food industry necessitate the urgent development of effective strategies for the inhibition and removal of S. aureus biofilms. Natural products have been extensively used as important sources of “eco-friendly” antibiofilm agents to avoid the side effects of conventional strategies on human health and the environment. This review discusses biofilm formation of S. aureus in food industries and focuses on providing an overview of potential promising target-oriented natural products and their mechanisms of S. aureus biofilm inhibition or removal. Hoping to provide valuable information of attractive research targets or potential undeveloped targets to screen potent natural anti-biofilm agents in food industries.

Keywords:

• anti-biofilm agent
• food
• natural products
• Staphylococcus aureus
• target–oriented screening

Introduction

In recent years, food safety has become a worldwide focus of public health. As one of the most common food safety issues, microbial contamination causes food spoilage and foodborne illness, not only seriously affecting people’s health but also incurring huge economic losses. Bacterial biofilms are the main culprits of sanitary problems and economic losses in the food industry. Bacteria can remain in food and attach to various surfaces by building three-dimensional rigid biofilm structures. Biofilm formation can enhance bacterial resistance to environmental stresses, which leads to increased cleaning difficulty and poses challenges for food hygiene (Avila-Novoa et al. 2018). In the food industry, the formation of biofilms in piping, equipment or heat exchangers can lead to the potential for metal corrosion, a decrease in operational efficiencies of equipment and an increase in maintenance costs (Xu et al. 2018). On the global scale, the annual impact of biofilms is incalculable, resulting in the loss of millions of dollars or more in food industries (Alonso, Harada, and Kabuki 2020; Cattò and Cappitelli 2019). Furthermore, biofilm formation occurs not only on environmental processing surfaces but also on food itself, leading to potential for cross-contamination and foodborne illness. Staphylococcus aureus is one of the most common foodborne pathogens related to foodborne disease outbreaks, especially in the meat and dairy industries (Guo et al. 2020). Thus, removal or inhibition strategies for S. aureus biofilms are an area of growing worldwide concern for public health.

In recent years, significant research efforts have been devoted to discovering biofilm inhibition and removal strategies, which have yielded several promising achievements. Traditional cleaning methods, including heat, detergents, surfactants, and disinfectants (Simões, Simões, and Vieira 2010). To avoid the negative effects of conventional chemical and physical strategies on humans and the environment, such as high energy consumption, high toxicity, the emergence of resistance, etc., efforts are focused on natural products to develop “green” anti-biofilm agents (Elmasri et al. 2015; Xu et al. 2017; Wu et al. 2019a).

Much attention has been directed toward target-based screening for natural antibiofilm agents over the past two decades. For example, quorum sensing (QS) systems are typical research target related to biofilms (Truchado et al. 2015; Chen et al. 2018). Although some natural products have been found which can inhibit S. aureus biofilm by disturbing QS system, further studies are still necessary because the mechanisms are still not fully understood. Additionally, an increasing number of new antibiofilm targets have been studied for S. aureus via various methods (e.g. computer virtual screening, molecular biology methods, etc.) (Yan et al. 2019; Sayed et al. 2020; Liang et al. 2019). These findings provide the basis for screening effective and “green” antibiofilm agents.

This review discusses common food environments where S. aureus biofilms easily occur; subsequently, we focus on the promising natural products of target-based screening and discuss their antibiofilm mechanisms. In addition, research limitations and future prospects for natural antibiofilm product target-oriented screening are provided.

S. aureus biofilms in food industries

In food processing operations and storage, in addition to existing in raw materials, most bacteria are associated with food contact surfaces and easily accumulate to form biofilms, which induce cross-contamination for transmitting pathogens to food. S. aureus can develop biofilms on different surfaces with nutrient-rich food residues, especially in the dairy and meat industries (Figure 1).

Figure 1. Potential S. aureus biofilm formation contact surfaces along the whole food supply chain.

In dairy industry, numerous instances of contamination have been reported. MRSA or MSSA biofilm-forming strains were isolated from raw milk or milk products such as cheese (Kang et al. 2020; Dai et al. 2019; Basanisi et al. 2017). Throughout the whole dairy supply chain, production involves milk and cheese in contact with the walls of the equipment. Because biofilms easily attach to stainless steel surfaces of equipment once contaminated, they can secrete exotoxin, which is a challenge for food safety in the dairy industry (Dutra et al. 2018; Avila-Novoa et al. 2018). More importantly, in dairy equipment, the development of biofilms is very rapid: 4–12 h (Meesilp and Mesil 2019). Furthermore, biofilm formation and dispersion behavior of foodborne pathogens are different in different dairy substrates due to varied composition and chemical alterations, such as pH. Alonso et al observed a greater dispersion of biofilm cells in skimmed milk (Alonso and Kabuki 2019). In addition, Carvalho et al. found that S. aureus has the capacity to form biofilms in low-density polyethylene packages when stored at 5 °C in the presence of Minas Frescal cheese whey, which indicates that packaging may be a potential source for dairy contamination (Carvalho et al. 2021).

In the meat industry, contamination caused by S. aureus biofilms mainly involves three aspects: (i) raw materials: as one of the most common pathogens originating from animals consumed as food, S. aureus has been discovered and isolated from pork, chickens, raw fish (Ali et al. 2021; Carvalho et al. 2020; Wu and Su 2014; Ou et al. 2020; Sullivan et al. 2020; Lin et al. 2019). Under appropriate conditions, initial attachment of biofilms on the surface may occur and induce cross-contamination through cutting boards, knives, and working surfaces (Llorens et al. 2012); (ii) food processing environment and equipment: in addition to slaughterhouses, operator bodies and clothes, which are often ignored, provide the perfect places for bacterial adhesion. Once bacteria are present on these surfaces due to incomplete cleaning, biofilm formation easily occurs (Al-Shabib et al. 2017a; de Souza et al. 2014); (iii) transportation and storage occur throughout the food chain, from transportation to sale, and bacterial biofilm formation easily occurs due to unsanitary environments or nonstandard worker operations (Miao et al. 2017).

Target-based screening for natural products

Biofilm formation is manipulated by various regulators involved in a complex regulatory network. Thus, to screen effective natural antibiofilm agents, it is necessary to research and determine one or multiple molecular targets. The biofilm formation process contains 4 stages: (1) initial biofilm attachment, (2) accumulation of layers, (3) biofilm maturation production of extracellular polymeric substrate (EPS), and (4) detachment of cells, as shown in Figure 2. Natural products can control biofilms by targeting these stages of the process. In the following discussion, we focus on promising natural products based on different molecular targets involved in the process.

Figure 2. Scheme of the network pathways governing biofilm formation and disruption in S. aureus. PIA, Polysaccharide intercellular adhesin; eDNA, Extracellular DNA; QS, Quorum sensing; LuxS, S-ribosylhomocysteine lyase; AI-2, Interspecies autoinducer; Agr, Accessory gene regulator protein locus; Spx, Global transcriptional regulator Spx; IcaR, Biofilm operon icaADBC HTH-type negative transcriptional regulator; SarA, Transcriptional regulator; SarR, HTH-type transcriptional regulator; σB, RNA polymerase sigma factor; PSMs, phenol-soluble modulins; CidA, Holin-like protein.

Target adhesion and attachment stage

At the beginning of biofilm formation, adhesion to various surfaces is a major strategy by S. aureus cells to remain in food or invade host cells. As adhesive matrix molecules, a vast array of Staphylococcal surface proteins are mobilized in this process, such as fibronectin-binding proteins FnbpA, FnbpB and ClfA. These proteins can be regarded as research targets for anti-biofilm agents discovery. As approved edible spice by national standard in China (GB1886.270-2016), tea tree oil showed high biofilm removal activity and eradicate pre-formed biofilm by 75% at minimum inhibitory concentration (MIC) concentration for 48 h, which targets surface proteins FnbB, as shown in Table 1 (Zhao et al. 2018). Thus, it can be considered as a potential promising natural product for development of anti-biofilm agents.

Table 1. Natural Products screening for inhibiting biofilm formation or removing pre-formed biofilm targeted adhesion proteins.

Class	Source	Compounds	Structure	MIC	Concentration, contact time (Biofilm formation inhibition rate)	Concentration, contact time (Pre-formed biofilm removal activity)	Mechanisms of action (Targets)	References
	Essential oils from Alaska yellow cedar trees, some herbs, and grapefruit	(+)-nootkatone		200 μg/mL	100 μg/mL, 24 h (99.8%); 50 μg/mL, 24 h (>90%) 25 μg/mL, 24 h (47%)	200 μg/mL, 24 h (50%)	spa	(Farha et al. 2020)
Flavanol	Tea, broccoli	Kaempferol		>1024 µg/mL	64 µg/mL, 512 µg/mL, 12 h (80%, 90%)	NS	Srt A, fnbpA, fnbpB, clfA, clfB, sarA	(Ming et al. 2017)
Flavonoid	Licorice	Licochalcone A		4 μg/ mL	NS	64 μg/ mL, 24 h, 48 h 5000–6000 mm thickness reduction	fnbB, clfA	(Shen et al. 2015)
	Tea, berries and red wine	Myricetin		>200 μM >1024 μg/ml	100 μM, 24 h (≈75%) 64 µg/mL, 18–24 h (≈60–70%)	NS	Srt A, Bap	(Silva et al. 2017)(Lopes et al. 2017)
	Geranium phaeum, Fagopyrum esculentum	Isovitexin		>1024 μg/mL	256 μg/mL, 24 h (≥ 61% inhibition)	NS	Srt A	(Mu et al. 2018)
	Ashitaba	Chalcone		>76 μM	76 μM, 18 h (≈60%)	NS	Srt A	(B. Zhang et al. 2017)
Essential oil	Melaleuca alternifolia	Tea tree oil	NS	1–2 mg/mL	NS	2 mg/mL, 48 h (75%)	fnbB	(Zhao et al. 2018)
Phenolic	Pine needles of Cedrus deodara	3-p-trans-coumaroyl-2-hydroxyquinic acid		5 mg/mL	0.3125–2.5 mg/mL, 24 h (53–88%)	No disruption effect	Srt A, ClfB, FnbpB, etc.	(Wu et al. 2019b) (X. Liu et al. 2021)
Phenolic acid	Salvia miltiorrhiza Bunge	Salvianolic acid A		>1024 μg/ml	256 μg/mL, 18 h (33.14%)	NS	Srt A	(Mu et al. 2020)
Phenolic	Oak	Tannic acid		40–160 µg/mL	20–80 µg/mL, 24 h (34%–100%)	NS	Peptidoglycan	(Dong et al. 2018)
Phenols	Loquat leaves	Resveratrol		350 μg/mL	100 μg/mL, 18 h (39.85%±0.18)	150 μg/mL, 18 h (23.42%±0.15 removal	Surface proteins and capsular polysaccharides. (cap)	(Qin et al. 2014)
Anthraquinone	P. cuspidatum and R. palmatum	Emodin		4–8 μg/ mL	4 μg/ mL, 24 h (≈60% inhibition)	NS	Srt A	(Yan et al. 2017)
Bibenzyl	Dendrobium chrysotoxum	Erianin		512 μg/mL	64 μg/mL, 24 h (≈50%)	NS	Srt A	(Ouyang et al. 2018)
Terpenoid	Loquat leaves	Ursolic acid		37 μg/mL	30 μg/mL, 18 h (66.30%±0.18)	NS	Reduce adhesins (isdB, srtB, ebh, and sdrC), amino acids metabolism (arcA, aur)	(Qin et al. 2014)

NS = Not specified

Some of these surface proteins carry a C-terminal LPXTG motif, which can be recognized by the membrane cysteine transpeptidase sortase A (SrtA). Briefly, the transpeptidation reaction mediated by SrtA is indispensable for anchoring these surface proteins on the cell wall (Figure 3). As important drug targets for controlling some attached proteins that mediate biofilm accumulation and adhesion, SrtA is not essential for bacterial growth or viability (Nitulescu et al. 2019). Thus, its inhibition should produce mild evolutionary pressure that will not favor the emergence of resistance. Inhibitors targeting this process are divided into two categories: (i) blocking SrtA activity directly by interacting with the catalytic active site. Chlorogenic acid has been found to bind to C184 and G192 in SrtA (Wang et al. 2015). Maresso et al. found that small molecules aryl (β-amino) ethyl ketones inhibit SrtA through irreversible, covalent modification of their active site cysteine (Maresso et al. 2007). Promising natural products, such as kaempferol, have been found to inhibit S. aureus biofilm formation (>80% inhibition rate at a sub-MIC 64 µg/mL, 18 h) by decreasing SrtA activity and the expression of the fnbpA, fnbpB, clfA, clfB and sarA genes (Table 1). Some other natural products, such as myricetin, esculein and palmatine chlorides, showed potent inhibitory activity on SrtA by virtual screening studies (Nitulescu et al. 2017; Nitulescu et al. 2016). A subsequent study proved that myricetin has biofilm inhibitory activity (64 µg/mL, 24 h, ≈70% inhibition rate) (Table 1), which indicates the reliability of a structure-based virtual screening strategy for target-oriented inhibitor discovery. (ii) SrtA can also recognize and couple synthetic molecules containing LPXTG motifs onto the peptidoglycan layer of S. aureus (Nelson et al. 2010). Therefore, molecules containing the LPXTG motif may be potential competitive inhibitors that interfere with the recognition of the LPXTG motif on surface proteins.

Figure 3. Mechanism of surface protein anchoring to the cell wall mediated by sortase A. Surface proteins precursors P1 with a C-terminal LPXTG motif are produced in the cytoplasm. SrtA is responsible for cleaving the amide bond between the threonine and glycine of the motif and catalyzing the formation of amide bond to form a Gly5 cross-bridge between the surface protein fragment and the peptidoglycan biosynthesis intermediate lipid II (Gao et al. 2016). Key molecule can be seen as antibiofilm and antibacterial targets, encircled in red. P1, P2, P3 represent Surface proteins precursors in different stages. MSCRAMMs: microbial surface components recognizing adhesive matrix molecule.

Target maturation stage

In S. aureus mature biofilms, the matrix accounts for more than 90% of the dry mass, whereas the microorganisms account for less than 10%. The EPS matrix comprises various types of polymeric substrates, including polysaccharides, nucleic acids (such as extracellular DNA), proteins, lipids, etc., and is indispensable for the construction of a three-dimensional biofilm architecture scaffold (Figure 2). Many natural products were found to reduce biofilms by destroying or decreasing biofilm matrix EPS directly, as presented in Table 2.

Table 2. Natural Products screening for inhibiting biofilm formation or removing pre-formed biofilm targeted EPS.

Class	Source	Compounds	Structure	MIC	Concentration, contact time (Biofilm formation inhibition rate)	Concentration, contact time (Pre-formed biofilm removal activity)	Mechanisms of action (Targets)	References
Terpenoid	Essential oils from Alaska yellow cedar trees, some herbs, and grapefruit	(+)-nootkatone		200 μg/mL	100 μg/mL, 24h (99.8%); 50 μg/mL, 24h (>90%) 25 μg/mL, 24h (47%)	200 μg/mL, 24h (50%)	IcaA, sarA,	(Farha et al. 2020)
	B. monnieri	Bacoside A		400 μg/mL	NS	200 μg/mL, 24h (≈90%)	EPS	(Parai et al. 2017)
	Litsea cubeba	Citral		500 μL/mL	NS	2000 μL/L, 45 °C, 60 min (> 5 log CFU reduction)	EPS	(Espina et al. 2017)
	Cabreuva oil	Farnesol		≈2 mg/mL	44 μg/mL, 24h (reduce significantly)	1–2 mg/mL, 20 min, (18–78%)	PIA	(Jabra-Rizk et al. 2006) (Unlu et al. 2018)
Carbohydrate	Marine biowaste	Chitosan		200–600 μg/mL	100 μg/mL, 24 h, (95.44%–56%)	100 μg/mL, overnight (60.69%)	EPS, slime synthesis	(Rubini et al. 2018a)
	Geranium	1,2,3,4,6-Penta-O-Galloyl-D-Glucopyranose		>50 μM	6.25, 12.5, 25 and 50 μM ,6 h (75%–99%)	NS	EPS	(Lin et al. 2011)
Essential oil	Melaleuca alternifolia	Tea tree oil	NS	0.25 mg/mL	0.25 mg/mL, 24 h, 51.23%	2 mg/mL, 24 h, 100%	EPS	(Zhao et al. 2018) (T. Liu et al. 2020)
	Peppermint	Menthol, menthone, (+)-isomenthone, and (+)-isomenthol	NS	5 mg/mL 3.1 μL/mL	0.25 mg/mL-2 mg/mL, 24 h, (≈1–6 log CFU/mL inhibition) 3.1 μL/mL, 24 h, 74.7%	3.1 μL/mL, 24 h, 67.5%	EPS	(Kang et al. 2019) (Bazargani and Rohloff 2016)
	Pogostemon heyneanus	(E)-nerolidol, acetophenone, Linalool and (E)-cinnamaldehyde	NS	2–6%v/v	0.5–3%v/v, 24 h (60–80%)	1–3% v/v, 24 h (70–80%)	Reduce EPS and slime synthesis	(Rubini et al. 2018b)
	Cinnamomum tamala		NS	2–6%v/v	1–3%v/v, 24 h (55–65%)	1–3% v/v, 24 h (60–65%)	Reduce EPS and slime synthesis	(Rubini et al. 2018b)
Acyclic sesquiterpene	Pogostemon heyneanus	Nerolidol		0.025%v/v	0.0125%v/v, 24 h (88%)	0.0125%v/v, 24 h (85%)	Reduce EPS and slime synthesis	(Rubini et al. 2018b)
Flavonoids	Licorice	Licochalcone A		1–8 μg/mL	MIBC: 8–64 μg/ mL 16–128 μg/ mL, thinner	NS	sarA, cidA, icaA	(Shen et al. 2015)
	Ruta leaves, Tobacco leaves, et al.	Rutin		800–1600 μg/ mL	1/16MIC-1/2MIC, 48 h (19–88%)	NS	Exopolysaccharide, PIA	(Al-Shabib et al. 2017b)
	Tea, berries and red wine	Myricetin		>200 μM >1024 μg/mL	100 μM, 24 h (≈75%) 64 µg, 18–24 h (≈60–70%)	NS	EPS	(Silva et al. 2017)(Lopes et al. 2017)
	Scutellaria spp.	Baicalein		256 μg/mL	32 μg/mL, 3 days (≈60%–80% inhibition)	64 μg/mL, 3 days, 7 days (≈1.5log CFU/mL reduction)	icaA, sarA	(Chen et al. 2016)
Pyranonaphthoquinone	Mitracarpus frigidus	Psychorubrin		5–80 μg/mL	2.5–80 μg/mL, 24 h (12%–52%)	2.5–80 μg/mL, 24 h (46%–85%)	EPS	(Lemos et al. 2018)
Polyhydroxylphenolic	Gall, tea, et al.	Gallic acid		2–8 mg/mL	2 mg/mL,24 h (31.7%)	4 mg/mL, 24 h (5.55 log CFU/ mL reduction)	Ica operon	(M. Liu et al. 2017)
Trihydroxyanthraquinone	P. cuspidatum and R. palmatum	Emodin		4–8 μg/ mL	4 μg/ mL, 24 h (≈60%)	NS	eDNA, cidA, icaA, dltB, and sarA	(Yan et al. 2017)
Phenols	Magnolia Bark	Honokiol		4–16 μg/ mL	NS	MBIC: 32–128 μg/ mL	eDNA, PIA, icaA, sarA, cidA	(Li et al. 2016)
	Caesalpinia sappan L	Brazilin		32 µg/mL	8–32 µg/mL, 24 h, 48 h (≈26–52%, 41%–60%)	64 µg/mL, 48 h (0.35 log CFU/mL reduction)	icaA	(Peng et al. 2018)
	Oregano oil	Carvacrol		>1 mM	0.5–1 mM, 24 h (≈25%–80%)	500μL /L, 45 °C, 60 min (> 5 log CFU reduction)	EPS	(Burt et al. 2014) (Espina et al. 2017)
Lipopeptides	Bacillus subtilis	Surfactin		32 μg/mL	8–16 μg/mL, 24 h (40–70%);	500 μg/mL, 1000 μg/mL, 2000 μg/mL,1–4 h (50%–90%)	icaA, icaD	(J. Liu et al. 2019)
Diterpene alkaloid	Totarol tree, Coptidis rhizoma	Totarol, Berberine chloride		0.5–2 μg/mL, 32–128 μg/mL	NS	4 μg/mL ≈ 37% 128 μg/mL (≈45%) Two compounds Combination, 24 h (84.5%)	icaA sarA, cidA	(Guo et al. 2015)
Aldehydes	Camphor leaves	Cinnamaldehyde		0.06–2 mg/mL	0.24 mg/mL, 6 h (57.1% ± 2.0% reduction in metabolic activity)	1–2 mg/mL, 20 min, (22–78%)	eno, ebps, fib, icaA, icaD SarA	(Kot et al. 2020) (Unlu et al. 2018)

NS = Not specified

Strategy against polysaccharide intercellular adhesin (PIA)

As a major fraction of the EPS matrix, exopolysaccharides are essential for biofilm formation. The biosynthesis of PIA is mediated by the intercellular adhesion (IcaADBC) locus (Archer et al. 2011). The natural product (+)-nootkatone (edible spices, GB 2760-2014) from grapefruit significantly prevented the biofilm formation (inhibition rate 99.8%, 100 μg/mL) of S. aureus by inhibiting the IcaA operon, as listed in Table 2. Another natural product gallic acid used as antioxidant which is regarded as safe in the food industry also exhibited anti-biofilm activity and regulated ica operon expression (Oliveira et al. 2020; M. Liu et al. 2017). Furthermore, the transcriptional regulator SarA can upregulate the production of biofilms by modulating ica transcription and the expression of poly-N-acetylglucosamine (PIA precursor) (Arciola et al. 2012). The biofilm inhibition activity of licochalcone A and emodin has been identified by decreasing the expression of sarA (Yan et al. 2017). Additionally, Ma et al. proposed that the S-ribosylhomocysteine lyase LuxS can reduce cell-to-cell adhesion by repressing rbf expression and inducing the downregulation of PIA. Thus, rbf might be considered as a potential research target (Ma et al. 2017).

Strategy against extracellular DNA (eDNA)

Another component in S. aureus biofilms, eDNA, requires the presence of matrix proteins to adhere to the cells, playing a role as an electrostatic net to hold cells together in large clumps (DeFrancesco et al. 2017). Indeed, biofilms can be reduced by adding DNase I exogenously, indicating the importance of eDNA in biofilm formation. Additionally, the CidA was indicated to positively increase the release of DNA during the development of a biofilm (Mann et al. 2009). Thus, eDNA release, or CidA, can be seen as a target for anti-biofilm agent screening (DeFrancesco et al. 2017). Some promising plant-derived compounds, including licochalcone A and emodin, were discovered to inhibit or disturb preformed biofilms by suppressing cidA, as shown in Table 2 (Shen et al. 2015; Yan et al. 2017). Although they have not been approved for use as food additives, their high inhibitory activity at low doses shows their potential for development as antibiofilm agents in the food industry.

Strategy against extracellular proteins

Considerable amounts of extracellular proteins within the biofilm matrix, including enzymes and nonenzymatic proteins, can far exceed the polysaccharide content. Numerous extracellular enzymes in biofilms can break down biopolymers to low-molecular-mass products and promote the detachment of bacteria from biofilms. The serine proteases SspA and SplA-F, cysteine proteases SspB and ScpA and metalloprotease Aur promote biofilm dispersal in S. aureus mediated by the global regulators SarA and SigB (Lister and Horswill 2014).

Nonenzymatic proteins, consisting of cell surface proteins and extracellular carbohydrate-binding proteins (biofilm-associated surface protein Bap, surface protein G SasG, fibronectin binding proteins FnBPs, clumping factor ClfA, enolase, etc.), are involved in the formation and stabilization of the biofilm matrix network and contribute to connecting the bacterial surface with extracellular EPS related to biofilm formation contribute to the attachment of bacterial cells on different surfaces. Natural products, such as emodin from P. cuspidatum, suppress biofilm formation by inhibiting these surface adhesion proteins, as shown in Table 2.

Target quorum sensing involved in dispersal stage

The quorum sensing (QS) system participates in bacterial cell-to-cell communication behavior, including biofilm formation. Thus, it is regarded as an important target for the development of antibiofilm agents. Interference with the QS system in S. aureus was achieved in two ways: disturbance of the Agr/AIP system and of the LuxS/AI-2 system, which are involved in intraspecific and interspecies communication, respectively.

Disrupting the agr/AIP system

In response to autoinducer peptide (AIP) signaling, the agr system regulates cell communication and virulence secretion via a series of cascading reactions. Briefly, AIP is first processed and secreted by the precursor peptide AgrD and integral membrane endopeptidase AgrB. Subsequently, extracellular AIP binds to the cognate transmembrane-bound receptor histidine kinase AgrC and promotes the phosphorylation of AgrA (Figure 4). Agr quorum-sensing activation leads to the production of enzyme proteins such as proteases, lipases, and nucleases that participate in the biofilm dispersal process. AgrA also directly activates the transcription of the psmα and psmβ genes, which encode phenol-soluble modulins (PSMs) that participate in biofilm dispersal (Horswill and Gordon 2020; Kim et al. 2017; Boles and Horswill 2008). From this perspective, activation of the agr system might be considered a more attractive strategy for biofilm inhibition (Horswill and Gordon 2020). However, most efforts to date have concentrated on the development of quorum inhibitors to reduce virulence because agr can positively increase virulence factors, such as enterotoxin and hemolysin secretion (Wang and Muir 2016). It is a challenge to precisely manipulate the Agr quorum sensing system to inhibit biofilm formation while not increasing virulence. This factor hinders the development of biofilm inhibitors targeting the agr quorum for S. aureus.

Figure 4. Mechanism of quorum sensing system (QS) Agr/AIP for biofilm regulation in S. aureus. The precursor of autoinducer peptide (AIP) is produced in the cytoplasm mediated by AgrD, Subsequently, AgrB is involved in the externalization of mature AIP. AgrC-AgrA two component system is activated at an AIP threshold concentration. Phosphorylated AgrA activates transcription of RNAII and RNAIII. The RNAIII inhibits biofilm by down- regulating adhesion surface proteins. AgrA increases transcription of the psmα and psmβ operons, encoding PSM peptides which improve biofilm dispersion. Key molecule can be seen as antibiofilm and antibacterial targets, encircled in red (Xie et al. 2019; Srivastava et al. 2014; Nicod et al. 2014).

Currently, the activation of agr has been proposed as a biofilm dispersal promoting strategy (Wang and Muir 2016). The first identified agr activation agents are the AIPs themselves. Activation of agr through AIP addition in established biofilms can trigger biofilm dispersal (Boles and Horswill 2008). The AIPs produced by one strain can cross-inhibit a strain producing a different agr class, which are competing AIPs (Cech and Horswill 2013). AIPs may also be regarded as potential targets; however, their roles in biofilm formation still need to be determined. Furthermore, Xie et al. reported that the activation of AgrC involves disruption of an intrasteric inhibitory docking interaction in the AgrC dimer, which provided a potential mechanistic strategy for inhibiting biofilms via the activation of AgrC (Xie et al. 2019). Qin et al proposed that resveratrol can enhance MRSA agr function at the RNA level to inhibit biofilm formation (Qin et al. 2014). Brazilin isolated from Caesalpinia sappan L inhibits S. aureus biofilm formation (32 µg/mL, 60% inhibition rate) by upregulating AgrA mRNA expression, as shown in Table 3. In contrast, some natural products have been found to have antibiofilm activity by downregulating the quorum-sensing system regulator AgrA in S. aureus (Sharifi et al. 2018; Chen et al. 2016). The results contradict the mechanism discussed above. This is possibly due to the complicated regulatory network involved in the multifactor response to biofilm removal or inhibition.

Table 3. Natural products screening for inhibiting biofilm formation or removing pre-formed biofilm targeted QS system.

Class	Source	Compounds	Structure	MIC	Concentration, contact time (Biofilm Inhibition rate)	Concentration, contact time (Pre-formed biofilm removal activity)	Mechanisms of action (Targets)	References
Essential oils	Thymus daenensis	Carvacol, γ-terpinene, p-cymene, etc.	NS	0.0625 μL/ mL;	0.0312 μL/mL), 24 h (≈80%–90%)	0.0625 μL/mL, 24 h (≈30%–40%)	QS/Agr	(Sharifi et al. 2018)
	Satureja hortensis	Thymol, γ-terpinene and p-cymene, etc.	NS	0.125 μl/ mL	0.0625 μl/mL, 24 h (≈60%)	0.0625 μL/mL, 24 h (≈40%–45%)	QS/Agr	(Sharifi et al. 2018)
Monoterpenoid	Essential oils of Eucalyptus globulus	1,8-cineole		0.048–3.125 mg/mL	0.095-0.003 mg/mL, 24 h (≈70%–97% );	MIC-4MIC, 24 h (77.46 ± 1.91–90.81 ± 4.05%)	QS/Agr	(Merghni et al. 2018)
Lipopeptides	Bacillus subtilis	Surfactin		32 μg/mL	8–16 μg/mL, 24 h (40–70%);	500 μg/mL, 1000 μg/mL, 2000 μg/mL, 1–4 h (50%–90%)	QS (LuxS/AI-2)	(J. Liu et al. 2019)
Extracellular protease	Actinomycetes culture supernatants	Proteinase K	NS	<0.1 μg /mL	0.1 μg /mL, > 80%	NS	Detach biofilm and enhance biofilm dispersal	(Park et al. 2012)
Polyphenols	Caesalpinia sappan L	Brazilin		32 µg/mL	8–32 µg/mL, 24 h, 48 h (≈26–52%, 41%–60%)	64 µg/mL, 48 h (0.35 log CFU/mL)	QS (agrA)	(Peng et al. 2018)
Phenols	Oregano oil	Carvacrol		>1mM	0.5–1 mM, 24 h (≈25%–80% inhibition)	500 μL/L, 45 °C, 60 min (> 5 log CFU reduction)	QS	(Burt et al. 2014) (Espina et al. 2017)
	Loquat leaves	Resveratrol		350 μg/mL	100 μg/mL, 18 h (39.85%±0.18);	150 μg/mL, 18 h (23.42%±0.15)	Disturb Agr-QS	(Qin et al. 2014)
	Tea	Tea polyphenols	NS	300 μg/mL	300 μg/mL, 5 days, (≈40%)	NS	QS/AI-2	(H. Zhang et al. 2014)
Aldehydes	Black garlic	5-hydroxymethylfurfural		>125 μg/mL	125 μg/mL, 24 h (82% )	NS	QS	(Vijayakumar and Ramanathan 2018)
Flavonoids	Scutellaria spp.	Baicalein		256 μg/mL	32 μg/mL, 3 days (≈60%–80%)	64 μg/mL, 3 days, 7 days (≈1.5log CFU/ml reduction)	QS	(Chen et al. 2016)

NS = Not specified

Interfering with LuxS/AI-2

The interspecies autoinducer AI-2 synthesized by the LuxS enzyme, which is involved in QS sensing, participates in interspecies cell communication, including biofilm formation. In Vibrio harveyi, Escherichia coli and Salmonella typhimurium, the pathways involved in AI-2 receptors and the transduction process have been well researched (Rajamani and Sayre 2011; Chen et al. 2002), but the mechanism of AI-2-mediated signal transduction in S. aureus remains inadequate. The inactivation of LuxS has been identified in S. aureus to result in increased biofilm formation and PIA production. The LuxS/AI-2 system regulates biofilm formation by modulating the transcriptional regulation of the ica locus in S. aureus (Ma et al. 2017; Rowe et al. 2016) (Figure 5). Cluzel et al. first demonstrated an AI-2-producing enzyme regulated by Ser/Thr phosphorylation (Cluzel et al. 2010). This has been seen as a potential strategy for perturbing AI-2 signaling for anti-biofilm and anti-virulence agent development (M. Guo et al. 2013). Natural products, such as surfactin of microbial origin, were identified as antibiofilm compounds targeted LuxS/AI-2 QS system with increased AI-2 activity. The 70% inhibition was achieved at 1/2 MIC (16 μg/mL), and 90% removal was achieved for the established biofilm after just 4 hours at 2 mg/mL (Table 3). Although the mechanism has not been clarified, the LuxS/AI-2 QS system can be a hypothetical competitive target due to its important role in intercellular communication.

Figure 5. Mechanism of quorum sensing system (QS) LuxS/AI-2 for biofilm regulation in S. aureus. Biosynthesis of Autoinducer 2 (AI-2) is catalyzed by LuxS and repressed the expression of rbf. Rbf could bind to the SarX and rbf promoters to upregulate their expression. Rbf can improves PIA-dependent biofilm formation via the repression of icaR expression. Polysaccharide intercellular adhesin (PIA) is synthesized regulated via proteins IcaA, IcaD, IcaB, and IcaC, encoded within the intercellular adhesin (ica) operon. IcaA and IcaD synergistically synthesize UDP-N-acetylglucosamine and exported via IcaC. Subsequently, IcaB regulates the partial deacetylation of PIA, improving adhesion via increased positive charge. Key molecule can be seen as antibiofilm and antibacterial targets, encircled in red.

The QS system is the key regulator of dispersal in biofilms by S. aureus. Biofilm dispersion, which is the last step in the biofilm formation process, is no less critical than the previous steps. At this stage, planktonic cells detach from biofilms and dissolve in neighboring spaces, determining the beginning of a new biofilm and becoming a new source for food contamination (Alonso and Kabuki 2019). Increased antibiotic susceptibility has been observed with the occurrence of the dissociation of the three-dimensional structure of the biofilm. Thus, several studies have proposed that the induction of biofilm dispersal may be a promising strategy (Landini et al. 2010; Lister and Horswill 2014). However, there are opposite voices. If biofilms accumulate on the contact surfaces of food packaging or equipment, induced dispersal could result in cross infections if the antibacterial agent fails to eradicate the released cells, and bacterial cells can detach from the surfaces, increasing the microbial load on the food product (Rossi et al. 2020). Although targeting QS system is a promising strategy, studies will need to address these challenges.

As depicted in Tables 1 and 2, a number of natural products ((+)-nootkatone, tea tree oil, emodin and myricetin) target SrtA, SarA and EPS substrates simultaneously. In addition, natural products carvacrol and surfactin extracted from Bacillus subtilis (Tables 2, 3) disturb QS system while reducing EPS production. Indeed, it is a coherent process for biofilm formation from the initial adhesion stage to the dissociation stage which involves complex pathway crosstalk. Multi-target compounds screening is an effective strategy for potent anti-biofilm agents discovery. Furthermore, according to Table 2, most natural compounds that target EPS (e.g. tea tree oil, essential oils of Pogostemon heyneanus, Baicalein, Carvacrol, etc.) are found to exhibit not only high biofilm formation inhibitory/prevention activity at sub-MIC concentrations but also well pre-formed biofilm removal activity at higher concentrations. However, few natural products target adhesin proteins show biofilm removal activity, as shown in Table1. Possibly because targeting EPS substrates is more conducive to break up the stable matrix facilitating the removal of mature biofilm. Moreover, essential oils from plants (e.g. tea tree oil, essential oils of Pogostemon heyneanus, Thymus daenensis, Eucalyptus globulus, etc.) (Tables 2, 3) all show biofilm inhibitory and scavenging activity which may be due to the synergy result from interactions between various components. Obviously, these products can serve as potential candidates of promising anti-biofilm agents.

Conclusions

In this review, most natural products have the ability to prevent biofilm formation at sub-MIC concentrations. This fact highlights that the capability of natural products to inhibit biofilms is linked to mechanisms other than growth inhibition. Indeed, an ideal antibiofilm inhibitor is expected to exert no pressure on bacterial growth to develop resistance. In addition, appropriate dosages and contact times are important for avoiding the overuse of antibiofilm agents that could be toxic to the end consumer. Thus, the present review summarizes useful information on the mechanisms, targets and dosage of natural compounds against biofilms.

The development of natural products urgently requires further exploration of their antibiofilm mechanism. Alternatively, to improve the potency and avoid negative effects, structural optimization of natural products is required. In view of drug resistance alleviation and increasing efficiency, the combination of different natural products or multitarget screening are promising strategy to inhibit or eradicate S. aureus biofilms. Within this review, many natural products against S. aureus biofilms have been discovered by targeting various substrates which involve in biofilm regulation. Overall, this review provides valuable information on promising targets for screening anti-biofilm agents. In addition, these natural products provide attractive candidates for the application and development of anti-biofilm agents against S. aureus in the food industry.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the National Nature Science Foundation of China (No. 31772082), the Natural Science Foundation of Jilin Province (No. 20180101249JC) and Jilin University Excellent Young Teacher Training Program.