**Insights into the Synthesis, Secretion and Curing of Barnacle Cyprid Adhesive via Transcriptomic and Proteomic Analyses of the Cement Gland**

#### **Guoyong Yan 1,2, Jin Sun 3, Zishuai Wang 4, Pei-Yuan Qian <sup>3</sup> and Lisheng He 1,\***


Received: 4 March 2020; Accepted: 29 March 2020; Published: 31 March 2020

**Abstract:** Barnacles represent one of the model organisms used for antifouling research, however, knowledge regarding the molecular mechanisms underlying barnacle cyprid cementation is relatively scarce. Here, RNA-seq was used to obtain the transcriptomes of the cement glands where adhesive is generated and the remaining carcasses of *Megabalanus volcano* cyprids. Comparative transcriptomic analysis identified 9060 differentially expressed genes, with 4383 upregulated in the cement glands. Four cement proteins, named Mvcp113k, Mvcp130k, Mvcp52k and Mvlcp1-122k, were detected in the cement glands. The salivary secretion pathway was significantly enriched in the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the differentially expressed genes, implying that the secretion of cyprid adhesive might be analogous to that of saliva. Lysyl oxidase had a higher expression level in the cement glands and was speculated to function in the curing of cyprid adhesive. Furthermore, the KEGG enrichment analysis of the 352 proteins identified in the cement gland proteome partially confirmed the comparative transcriptomic results. These results present insights into the molecular mechanisms underlying the synthesis, secretion and curing of barnacle cyprid adhesive and provide potential molecular targets for the development of environmentally friendly antifouling compounds.

**Keywords:** barnacle; cement gland; cyprid adhesive; transcriptome; cement protein

#### **1. Introduction**

Barnacles are major marine fouling organisms that can secrete adhesives to attach themselves permanently to underwater substrates on which they live [1]. The adhesive generated by barnacles has been termed barnacle adhesive or barnacle cement [2], usually, the adhesive is a thin layer only a few microns thick, but it is nevertheless capable of adhering barnacles tightly to different foreign materials throughout their entire lifespan without failure, even under conditions of strong wave action [3]. Due to its robust strength and durability underwater, barnacle adhesive has attracted the attention of scientists from different fields. Some scientists have attempted to develop biomimetic underwater glues [4], and others have engaged in developing antifouling compounds to prevent the

adhesive attraction [5]; however, both lines of research require a good understanding of the molecular mechanisms underlying the regulation of barnacle cyprid cementation.

The life cycle of barnacles includes the nauplius I-VI, cyprid, juvenile and adult stages. The cyprid stage is the last planktonic stage during which barnacle cyprids search for suitable substrates on which to attach and metamorphose. The adhesive that cyprid larvae secrete from the attachment discs on their antennules is named cyprid adhesive, which is synthesized in a pair of cement glands. The cement glands are composed of two cell types, α cells and β cells, both of which are secretory; these cells synthesize proteins and lipids, respectively [6,7]. After settlement, the cement glands disappear gradually, and only cement cells spread over the connective tissue on the base plates of adult barnacles [8,9]. An adult adhesive is generated in the cement cells and is secreted into the adhesive interface through a well-developed duct system [8,10], and a recent study showed that acorn barnacles secrete phase-separating fluid, which is rich in lipids and reactive oxygen species to clean the surface before cement deposition [11].

Although both cyprid and adult barnacles can generate adhesive, it is almost impossible to collect enough cyprid adhesive for common research use; thus, almost all related research has been based on adult barnacle adhesive. Previous research has found that barnacle adhesive consists of proteins (90%), lipids (1%), carbohydrates (1%) and inorganic ash (4%) [2], indicating that the majority of the secretion is proteinaceous. Detailed analysis of the components of barnacle adhesive was hindered by its inherent insolubility [12] until Kamino and his colleagues developed a nonhydrolytic method of dissolving most of the barnacle adhesive from adult *Megabalanus rosa*, leading to the identification of the barnacle cement proteins Mrcp100k, Mrcp68k, Mrcp52k, Mrcp20k and Mrcp19k [12–15] and their homologues in different species [16–19]. In the past few years, high-throughput transcriptomic and proteomic approaches have facilitated the discovery of novel cement proteins, such as cp114k and cp43k; and many other genes and proteins that might be related to cementation, such as peroxidases and lysyl oxidases, have also been identified [20–22].

In contrast to the well-studied proteins in mussel and tube worm adhesives, barnacle cement proteins have no 3,4-dihydroxyphenylalanine (DOPA) [3,23]. There are no phosphorylation modifications in any characterized cement protein, although phosphoproteins have been reported in the cement gland and adhesive interface [3,7,24]. Quinone-type cross-linking has been thought to be responsible for the adhesion of barnacle adhesive [25,26], but sufficient evidence is lacking [23]. Dickinson and colleagues hypothesized that the barnacle cement polymerization process was similar to blood clotting caused by transglutaminase cross-linking [27], but this hypothesis was repudiated by Kamino [28]. Recently, growing evidence has demonstrated that hydrophobic interactions among cement proteins and amyloid-like conformations might play important roles in the self-assembly and curing of barnacle adhesive [13,29]. It is highly possible that barnacle adhesive functions through multiple mechanisms.

Here, to the best of our knowledge, for the first time, we obtained the transcriptome of cement glands dissected from *Megabalanus volcano* cyprids through RNA-seq. Comparative analysis with the carcass transcriptome was performed to screen for Differentially Expressed Genes (DEGs). Cement proteins produced in the cyprid cement gland were identified and characterized. Further analysis of the DEGs was performed to identify genes and pathways that might be involved in the synthesis, secretion and curing of barnacle cyprid adhesive. The proteome of the cement gland was also obtained by LC–MS/MS analysis, and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was performed to verify and complement the comparative transcriptomic analysis results.

#### **2. Results**

#### *2.1. Transcriptome Sequencing, Assembly and Annotation*

The pair of cement glands was isolated from the whole body of *M. volcano* cyprids (Figure 1), and the remaining parts were collected together and defined as the carcass. The cement glands and carcasses were separately subjected to RNA extraction, cDNA library construction and Illumina high-throughput sequencing. In total, 65.60 M raw reads each were obtained for the cement glands and carcasses, and 65.36 M (99.63%) and 65.44 M (99.76%) clean reads remained, respectively, after poor-quality reads were filtered out. All clean reads for the cement glands and carcasses were subjected to *de novo* assembly, and a total of 38,538 and 55,537 unigenes were obtained, with mean lengths of 679 bp and 565 bp and N50 values of 1092 bp and 779 bp, respectively. Redundancy was further removed for the unigenes from the cement glands and carcasses to obtain a global dataset (all-unigenes). Finally, 67,299 unigenes remained, with a mean length of 613 bp and an N50 value of 928 bp (Table 1). Benchmarking Universal Single-Copy Orthologs (BUSCO) was used to assess the completeness of the transcriptome assembly, and the results showed that 76.8% of the BUSCOs were complete, 12.5% were fragmented and 10.7% were missing.

**Figure 1.** Dissected cement gland of a *Megabalanus volcano* cyprid.


**Table 1.** Summary of sequencing, assembly and annotation results.

To determine the gene functions of the assembled unigenes, six public databases were searched for annotations. In total, 22,139 (57.45%), 25,099 (45.16%) and 32,505 (48.30%) unigenes were annotated for the cement gland, carcass and global datasets, respectively. The detailed annotation results are summarized in Table 1. The coding regions of unannotated unigenes were predicted with ESTScan, and 5360 were predicted to have coding regions. Ultimately, 33,882 protein-coding unigenes were obtained from the transcriptome. To gain an overview of the functions of all the unigenes, the Gene Ontology (GO), Cluster of Orthologous Groups (COG) and KEGG databases were used for unigene classification. The GO classification results showed that 6830 nr-annotated unigenes were assigned to the three major functional categories. In detail, the unigenes were annotated in 19, 25 and 19 subcategories in the GO Cellular Component (GOCC), GO Biological Process (GOBP) and GO Molecular Function (GOMF) categories, respectively (Figure S1A). With regards to the COG database, 12,584 unigenes were annotated and further classified into 25 categories (Figure S1B). Moreover, the KEGG analysis showed that 21,600 unigenes were mapped to 235 pathways.

#### *2.2. Comparative Transcriptomic Analysis and Characterization of DEGs*

Genes that have different expression levels in different organs or tissues are thought to have specific functions consistent with the roles that the organs or tissues play in the whole body. Hence, comparative transcriptomic analysis was performed to screen for DEGs between cement glands and carcasses based on Fragments Per Kilobase of Transcript Per Million Mapped Reads (FPKM) values. As a result, a total of 9060 DEGs were identified, among which 4383 DEGs were upregulated and 4677 were downregulated in the cement glands compared with carcasses (Figure 2A). The numerous DEGs between cement glands and carcasses reflect the great differences between these samples at the transcriptional level, which are caused by tissue/organ-specific gene expression.

**Figure 2.** Comparative analysis of unigene expression levels between cement glands and carcasses. **A**. Volcano plot of the Differentially Expressed Genes (DEGs) between cement glands and carcasses. The blue spots indicate unigenes that were upregulated in cement glands, the red spots indicate downregulated unigenes, and the black spots indicate unigenes that were not differentially expressed. **B**. Gene Ontology (GO) enrichment analysis of the upregulated DEGs in cement glands, terms related to protein synthesis and protein modification are in red. **C**. KEGG enrichment analysis of all the DEGs between cement glands and carcasses, salivary secretion pathway (ko04970) is in red.

DEGs that were upregulated in cement glands were subjected to GO enrichment analysis to determine the functions of the overrepresented gene set. In the GOCC category, 10 terms were significantly enriched (*p* < 0.05; Figure 2B). Among them, four terms (GO: 0005840, ribosome; GO: 0044391, ribosomal subunit; GO: 0015935, small ribosomal subunit and GO: 0031985, Golgi cisterna) were related to protein synthesis and further protein modification, which is consistent with the role of the cement gland in generating cement proteins [6,7]. In total, 508 of the 4383 DEGs upregulated in cement glands were predicted to be transcription factors based on AnimalTFDB 2.0 (Table S1). In addition, 622 of the 4383 had no hits in any public databases, but 148 of them were predicted to have coding regions by ESTScan. The SignalP and TMHMM servers were used to analyze the sequence characteristics of the 148 DEGs [30]. Seven of the DEGs with predicted signal peptides but without transmembrane domains were suggested to be putative cement gland-secreted proteins (Table S2). These putative secreted proteins that are highly expressed in cement glands also have the potential to be novel cement proteins.

Further KEGG enrichment analysis was performed on all the DEGs to identify the predominant pathways that might be involved in the specific function of the cement gland. The results showed that eight pathways were significantly enriched (*p* < 0.001; Figure 2C). These pathways included oxidative phosphorylation (ko00190), the citrate cycle (TCA cycle; ko00020), 2-oxocarboxylic acid metabolism (ko01210) and cysteine and methionine metabolism (ko00270), which are involved in fundamental material and energy metabolism; the cytosolic DNA-sensing pathway (ko04623), which plays important roles in the innate immune response; and salivary secretion (ko04970), collecting duct acid secretion (ko04966) and cardiac muscle contraction (ko04260), which function in the secretion and release of secretory substances. These pathways are closely related to the role of the cement gland as a synthetic and secretory organ.

#### *2.3. Characterization of Cement Proteins in the Cement Gland*

According to the BLASTx results, four barnacle cement proteins were identified in the cement gland transcriptome, and they were named Mvcp113k, Mvcp130k, Mvcp52k and Mvlcp1-122k. Mvcp113k had the highest similarity (81%) with Mrcp100k from *M. rosa*, and Mvcp130k had the highest similarity (65%) with Aacp100k from *Amphibalanus amphitrite*. The sequence similarity between Mvcp113k and Mvcp130k was 51.7% (Figure 3A). Sequence alignment of the six cp100k homologues from four species of acorn barnacle revealed that they were conserved in different species, especially at the *N*-terminus (Figure S2). Comparative analysis of the amino acid composition of the six cp100k homologues revealed that they also had similar amino acid composition and theoretical pI values, which ranged from 9.63 to 9.99 (Table S3). Mvcp52k had the highest similarity (83%) with Mrcp52k from *M. rosa*. Sequence analysis showed that Mvcp52k also contained four repeat sequences with lengths of 129, 124, 120 and 113 amino acids, respectively, and that each repeat sequence contained a Cys residue located in nearly the same region (Figure 4A,B). Mvlcp1-122k had the highest similarity (89.6%) with the newly reported Mr-lcp1-122k, a cement gland-specific protein from *M. rosa* [31]. Sequence analysis found that Mvlcp1-122k and Mr-lcp1-122k were predicted to have one *N*-glycosylation site and seven mucin type GalNAc O-glycosylation sites (Figure 5A). The transcriptome quantitative results showed that Mvcp113k, Mvcp130k and Mvcp52k were almost exclusively expressed in the cement glands (Figures 3B and 4C), and Mvlcp1-122k was expressed in both cement glands and carcasses and had especially high expression level in the cement glands, which ranked the 91st of all the unigenes (Figure 5B).

**Figure 3.** Cement protein-100 kDa homologues expressed in cement glands. **A**. Sequence alignment of Mvcp113k (MK336236) and Mvcp130k (MK336237) of *M. volcano*, the homology level of the sequences = 100% are shaded in black. **B**. Expression levels (Fragments Per Kilobase of Transcript Per Million Mapped Reads (FPKM) values) of Mvcp113k and Mvcp130k in cement glands and carcasses.

**Figure 4.** Cement protein-52 kDa homologue expressed in cement glands. **A**. Sequence alignment of Mvcp52k (MK336235), Mrcp52k (BAL22342.1) from *M. rosa* and Aacp52k (AKZ20820.1) from *A. amphitrite*, the homology level of the sequences = 100% and ≥ 50% are shaded in black and blue; 4 repeat sequences are labeled as Rep-1, -2, -3 and -4, and boxed in yellow, purple, green and orange rectangles respectively. **B**. Sequence alignment of the four repeat sequences of Mvcp52k, the homology level of the sequences = 100%, ≥ 75% and ≥ 50% are shaded in black, pink and blue, locations of a Cys residues are boxed in red rectangle. **C**. Expression levels (FPKM values) of Mvcp52k in cement glands and carcasses.

**Figure 5.** Cement gland-specific protein 122 kDa homologues expressed in cement glands. **A**. Sequence alignment of Mvlcp1-122k (MT024661) and Mrlcp1-122k (MK490677) of *M. rosa*, the homology level of the sequences = 100% are shaded in black. Asn-Xaa-Ser/Thr sequon is boxed in red rectangle, and the Asparagine predicted to be *N*-glycosylation site is marked with blue asterisk, predicted mucin type GalNAc O-glycosylation sites are marked with inverted purple triangle. **B**. Expression levels (FPKM values) of Mvlcp1-122k in cement glands and carcasses.

#### *2.4. Characterization of Enzymes in the Cement Gland*

All the enzyme-coding unigenes in the cement gland transcriptome were summarized according to their KEGG annotation results, and a total of 5913 enzyme-coding unigenes were ultimately identified. These unigenes were classified into six groups according to the enzyme commission scheme, with 946 classified as oxidoreductases, 2204 classified as transferases, 2091 classified as hydrolases, 207 classified as lyases, 214 classified as isomerases and 251 classified as ligases; furthermore, 1026 of the 5913 enzyme-coding unigenes were upregulated in the cement glands (Table S4). Lysyl oxidase (MvLOX), a homologue of AaLOX-1 in *A. amphitrite*, which has been detected in the adult barnacle adhesive interface [20,21], was also identified in the cement gland transcriptome of *M. volcano*. The putative protein sequence of MvLOX contained 510 amino acids and had the highest similarity (75%) to AaLOX-1, it had a conserved lysyl oxidase domain located at the C-terminus, and a predicted signal peptide indicating that it could be secreted out of cells (Figure 6A). MvLOX had higher expression level in the cement gland than the carcass (Figure 6B), suggesting that it might be involved in cyprid cementation. In addition, enzymes that involve in chitin synthesis (chitin synthase) and degradation (chitinase) were expressed in the cement gland as well as in the carcass (Table S5).

**Figure 6.** Characterization of *M. volcano* lysyl oxidase. **A**. Sequence alignment of MvLOX and AaLOX-1 (AQY78507.1) from *A. amphitrite*, the homology level of the sequences = 100% are shaded in black, predicted signal peptide is boxed in yellow rectangle and conserved lysyl oxidase domain is boxed in red rectangle. **B**. Expression levels (FPKM values) of MvLOX in the cement glands and carcasses.

#### *2.5. Proteomic Analysis of the Cement Gland*

LC–MS/MS analysis was performed to obtain the proteome of the cement glands, and a total of 352 proteins were identified based on 1504 peptides. All of the identified proteins were mapped to 77 pathways. KEGG enrichment analysis showed that pathways related to protein synthesis (ribosome (ko03010)) and energy metabolism (oxidative phosphorylation (ko00190), fructose and mannose metabolism (ko00051) and sulfur metabolism (ko00920)) were highly enriched, confirming that the production and secretion of the adhesive requires adequate energy supply. The salivary secretion pathway (ko04970) was also significantly enriched (Figure 7), which is consistent with the transcriptomic analysis results. The proteomic analysis partially validated the results from the comparative transcriptomic analysis.

**Figure 7.** KEGG enrichment analysis of all the proteins identified in the cement gland proteome. The salivary secretion pathway (ko04970) is in red, and the PPAR signaling pathway (ko03320) is boxed in blue rectangle.

Furthermore, the PPAR signaling pathway (ko03320), which is mainly involved in lipid metabolism, was also significantly enriched (Figure 7). Lipid-binding proteins are important components of the PPAR signaling pathway (ko03320), and two lipid-binding proteins were identified in the cement gland proteome (Table S6). Conserved domain database (CDD) analysis showed that these two proteins had hits in the lipocalin (pfam00061) and lipocalin\_7 (pfam14651) domain families, respectively (Figure S3), which belong to the lipocalin superfamily (cl21528). These findings indicate the basic function of these lipid-binding proteins as lipocalins that transport small hydrophobic molecules such as lipids. Mv-FABP1 (Unigene13631\_All) was found to be upregulated in the cement glands according to the comparative transcriptomic results, suggesting that it is likely involved in the regulation of lipid accumulation and metabolism in the cement glands. Lipids have been found to be integral components in mussel adhesion [32], and lipid-binding proteins have also been identified in the adhesive glands of marine tube-building polychaetes [33], implying their universal importance for marine adhesives.

#### **3. Discussion**

The cement gland is the primary organ responsible for the synthesis and secretion of barnacle cyprid adhesive, and it plays an indispensable role in the larval settlement of barnacles. However, because cyprids are very small and have a pair of carapaces [34], it is very difficult to harvest the pair of cement glands (Figure 1). Therefore, research on barnacle cyprid adhesive system is relatively rare, especially research at the molecular level. In this study, three-day-old *M. volcano* cyprids were chosen for cement glands dissection, when temporary settlement behavior began to appear, suggesting that most of the cyprids were in the preparation for settlement. Sufficient cement glands for RNA-seq were isolated successfully, and the transcriptomes of cement glands and carcasses were obtained for the first time. The high occupation of clean reads, complete BUSCO and high N50 value indicated that the quality of the transcriptome was sufficient for further analyses [18,35]. Comparative transcriptomic analysis is an efficient and reliable method for screening DEGs involved in specific regulatory functions. Here, a false discovery rate (FDR) ≤ 0.001 rather than the common FDR ≤ 0.01 was used to improve the confidence level of the differential analysis results because of no biological replicates. In total, 9060 DEGs were identified, 4383 of which were upregulated in the cement glands. Unigenes that might function in cyprid cementation were further characterized with the aim to decipher the underlying molecular mechanisms.

Transcription is the first step in gene expression, while transcription factors control whether genes are transcribed and the rate of transcription [36]. Those transcription factors that had a higher expression level in the cement glands might play transcriptional regulatory roles during the initial expression of cement proteins and other related proteins in the cement glands [37]. However, due to the lack of whole genome information of barnacles, we are currently incapable of obtaining the promoter sequences of the cement protein-coding genes; otherwise, we could identify the upstream transcription factors that regulate the transcription of these genes based on predicted transcription factor binding sites in their promoter sequences [38].

Barnacle larval settlement is a highly energy-consuming process [39,40]. Here, several energy metabolism pathways were enriched in both transcriptomic and proteomic analyses (Figures 2C and 6), suggesting that a portion of energy might be used for adhesive synthesis and secretion. This finding is consistent with the findings of studies on organs that are analogous or homologous to cement glands, such as silkworm silk glands [41], planthopper salivary glands [42] and scorpion venom glands [43], in which abundant protein synthesis and high energy metabolism are demanded. In addition, in the marine environment, biofilms are crucial mediators of barnacle larval settlement [44], which means that larvae are exposed to active and dense microbial environments that include potential pathogens. Cement glands are attractive targets for pathogens because they connect to the external environment directly via the cement ducts in the antennules; however, it is possible that cyprids protect themselves from invading pathogens through the cytosolic DNA-sensing pathway, which can detect foreign DNA and induce further immune responses [45]; acid secretion regulated by the collecting duct acid secretion pathway might also be involved in maintaining a sterile interior and surface cleaning.

Until now, studies on adhesive secretion have focused only on morphological and physiological characteristics, and exocytosis was found to be the major mode of adhesive secretion, but detailed molecular mechanisms underlying this process are poorly understood [46,47]. The significant enrichment of the salivary secretion pathway found in this study implies that the molecular mechanisms underlying cyprid adhesive secretion by cement glands might be analogous to that of saliva secretion by salivary glands [48,49]. In secretory cells, suitable environment cues for settlement are transformed into neural signals. On the one hand, these signals cause adenylyl cyclase activation and intracellular cAMP accumulation, and elevated cAMP induces the secretion of proteins such as cement protein. On the other hand, these signals also activate phospholipase C, causing increases in intracellular Ca2<sup>+</sup> that lead to ion and water secretion. In cyprids, the secretion of adhesive from the secretory cells and the release of these adhesive from the cement gland are two separate processes. It is presumed that secreted adhesive is accumulated by the median collecting duct and is temporarily stored in the cement

duct and/or muscular sac; once a suitable substrate for permanent attachment has been found, the adhesive is released explosively through the adhesive disc with the pumping action of the muscular sac [46,47,50]. The muscular sac is composed of a layer of circular muscle (Figure 1) [50]. As the cardiac muscle contraction pathway was also found to be enriched in the current study, we speculated that the contraction of the muscular sac might be the same as that of cardiac muscle, which is initiated by electrical excitation of myocytes and is mediated by calcium cycling and signaling [51].

Protein is the dominant component of both cyprid and adult adhesive [6,7,12], but whether the two types of adhesive possess the same kinds of cement protein remains unclear. Cyprid and adult adhesive have been thought to be different but share a portion of cement proteins [3]. According to the transcriptome of adult *M. volcano* [34], Mvcp130k, Mvcp113k and Mvcp52k that identified in the cement glands of cyprids are also expressed at the adult stage. The place where larval adhesive generated is quite different from adults and they have different ways to be released to the surface [9,10], suggesting that there might be potential larval-specific cement proteins compared to well-studied adult cement proteins [21]. Recently, Aldred et al. (2020) reported the existence of a cement gland-specific protein (lcp1-122k) in barnacle *A. amphitrite* and *M. rosa*, and speculated that it functioned through glycosylation with chitin [31]. The homologue of lcp1-122k was also identified in the cyprid of *M. volcano* in this study, while it was not found in the transcriptome of adult *M. volcano* [34]. Chitin metabolism related enzymes express in the cement gland, implying that chitin could be generated in the cement gland. Further identification of potential glycosylation sites in the lcp1-122k homologues provide increased evidence that cement gland-specific protein lcp1-122k might be a glycoprotein (Figure 5A). Cheung and Nigrelli observed that the cement gland at the nauplius VI stage presented considerable biosynthetic activity while the activity in the cement gland at the cyprid stage was relatively low [8], no direct evidence has shown that cement proteins are produced at the nauplius VI stage up to now. Here, a high level of cement protein-coding gene mRNAs (Mvcp130k, Mvcp113k, Mvcp52k and Mvlcp1-122k) was detected in the cyprid cement gland transcriptome, indicating that the cement proteins used for settlement are *de novo* synthesized in the cement gland at the cyprid stage. In the present study, we reported the existence of two cp100k homologues in *Megabalanus* barnacles and detected the expression of cp52k in barnacle cyprids for the first time. Two cp100k homologues of the barnacle species *M. volcano* were identified in this study, and three cp20k homologues and two cp100k homologues were also found in previous studies [52], but the different roles that these homologous cement proteins might play have not been clarified based on their sequence characteristics and predicted physical and chemical properties. We speculated that the diversity of cement protein homologues might be a mechanism for guaranteeing the successful settlement of barnacles under different circumstances. Notably, gene duplication has played vital roles in spider silk gland evolution [53]; however, whether similar gene duplication has occurred in the evolution of the barnacle cement gland will be difficult to determine until the whole barnacle genome is available.

Enzymes have been reported to play important roles in secretory glands, such as snail salivary glands [30] and tube worm adhesive glands [33]. What is more, many enzymes have been proposed as specific targets for antifouling compounds [5]. In adult barnacle adhesive, lysyl oxidase has been predicted to oxidize the lysine in cement proteins to reactive allysine and further form durable lysine protein cross-links that involve three proximate allysines and a lysine side chain [20]. The elevated expression level of MvLOX in the cement gland suggests that MvLOX has the potential to play the same role in cyprid adhesive, indicating that allysine-mediated cross-links might be at least one of the factors involved in cyprid adhesive curing. Inexplicably, neither phenoloxidase nor catecholoxidase was identified in the cement gland transcriptome, although phenoloxidase and catecholoxidase activity have been reported in barnacle cyprid cement glands [2,6]. In well-studied marine tube worm cement proteins and mussel foot proteins, post-translational modification of tyrosine residues into DOPA is of great importance for adhesion [54,55]. Seven types of enzymes have been reported to contribute to the L-DOPA metabolic pathway, including phenylalanine dehydrogenase, phenylalanine hydroxylase, tyrosine aminotransferase, aspartate aminotransferase, histidinol-phosphate aminotransferase, l-amino-acid

oxidase, tyrosinase and peroxidase [33]; however, most of these enzymes were absent in the cement gland transcriptome except aspartate aminotransferase and peroxidase, suggesting that the functional mechanism of barnacle cyprid adhesive differs from that of marine adhesives that rely on DOPA.

In total, 352 proteins were identified in the cement glands by LC–MS/MS. Notably, no cement proteins were identified, which is probably due to the technical limitations of mass spectrometry, because some abundant proteins might increase the difficulty of identifying cement proteins expressed at relatively low levels [22]. Protein database based on a whole genome might be more helpful for protein identification, considering the spatial and temporal expression difference of transcriptomes. The significant enrichment of lipid metabolism-related pathways implies that lipids play important roles in cementation, which is consistent with the discovery that cyprid adhesive is a biphasic system consisting of phosphoproteins and lipids and that lipids are secreted first to create a conducive environment for phosphoproteins and modulate the protein phase simultaneously [7]. Moreover, lipids have been reported to augment amyloid-beta (Aβ) peptide generation, and various lipids and their assemblies can interact with amphiphilic Aβ peptide to change Aβ aggregation [56]. Considering previous findings that amyloid-like nanofibrils are the main components of adhesive plaques from the barnacle *Balanus amphitrite* [57] and that certain peptides from a bulk 52 kDa cement protein [29] and a full-length 19 kDa cement protein [58,59] can self-assemble into amyloid fibrils, we hypothesize that lipids might also participate in the curing of barnacle adhesive by affecting cement protein amyloid fibril aggregation.

In the present study, comparative transcriptomic analysis and proteomic analysis were used to screen for genes, proteins and pathways that were involved in barnacle cyprid cementation, which is of great importance to decipher the molecular mechanisms underlying the synthesis, secretion and curing of barnacle cyprid adhesive. As adhesive production/release has been reported to be the most common targets for antifouling compounds development, and antifouling compounds that have specific molecular targets are considered more likely to be non-toxic rather than function through toxic killing [5]; thus, these genes, proteins and pathways have the potential to be molecular targets for novel non-toxic antifouling compounds. In recent years, many antifouling biocides with potential environmental risk, such as tributyltin (TBT), were restricted to use, and non-toxic antifouling compounds with specific targets are urgently need; we hope that our findings in this research to be helpful for the development of environmentally friendly antifoulants.

#### **4. Materials and Methods**

#### *4.1. Larval Culture*

Larval culture was performed as described in our previous study [34]. Briefly, adult *M. volcano* barnacles were collected from the rocky shore, cleaned thoroughly in the laboratory. After drying in the air for 24 h, they were transferred into 0.22 μm-filtered sea water (FSW) to release embryos. The released embryos were collected and hatched in FSW at 28 ◦C to obtain swimming nauplii; then, all the nauplii were transferred into another tank and cultured at a density of 1 larva mL−<sup>1</sup> in autoclaved FSW at 25 ◦C with the light:dark cycle of 12 h:12 h, and were fed with *Chaetoceros gracilis* at about <sup>1</sup> <sup>×</sup> 106 cells/mL every day until they transformed into cyprids.

#### *4.2. Cement Gland Dissection*

Dissection of the cement gland from *M. volcano* cyprids was performed following a protocol described for *M. rosa* [46]. Three-day-old cyprids were placed into modified barnacle saline [60] containing 462 mM NaCl, 8 mM KCl, 32 mM MgCl2 and 10 mM HEPES (pH = 7.5) for relaxation. Relaxed cyprids were dissected under a stereoscope with a pair of finely etched tungsten needles. The swimming appendages were removed first, and then the bivalved carapaces were separated to expose the internal cavity; the pair of cement glands was carefully pulled away from adjacent tissues with the needles without touching them directly to avoid being damaged. The isolated cement glands were

sucked out with a pipette, and the remaining parts were all collected as the carcass. The cement glands and carcasses were separately transferred to RNAlater (Invitrogen, Carlsbad, CA, USA) and stored at −80 ◦C until use.

#### *4.3. RNA Extraction, cDNA Library Construction and Sequencing*

Thirty cement glands and 10 carcasses of *M. volcano* cyprids were pooled together, respectively. Total RNA from the cement gland and carcass samples was extracted with TRIzol Reagent (Invitrogen) following the manufacturer's instructions. The quality and quantity of the RNA samples were measured with an Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA). RNA-seq for the cement glands and carcasses was performed by BGI China using equal amount of RNA. Briefly, sequencing libraries were constructed using a Nextera XT DNA Library Preparation Kit (Illumina Inc., San Diego, CA, USA) according to the manufacturer's instructions, and high-throughput sequencing was performed on an Illumina HiSeq 4000 platform (Illumina Inc.) with the model of paired-end (PE) 101 bp.

#### *4.4. De Novo Assembly and Annotation*

The raw sequence reads were filtered to obtain high-quality clean reads by removing reads with adaptor sequences, more than 5% unknown nucleotides, or more than 20% low-quality bases with SOAPnuke version 1.5.2 [61]. The read data have been submitted to the SRA of NCBI under the accession numbers SRR6516776 and SRR6516777. Clean reads were used for all the following analyses. The clean reads for the cement glands and carcasses were separately subjected to de novo assembly with the Trinity algorithm version 2.0.6 [62]; they were further assembled and redundancy was removed with TGICL version 2.1 [63], using the parameters of "repeat\_stringency = 0.95, minmatch = 35, minscore = 35". The completeness of transcriptome assembly was assessed by BUSCO version 3.0 with the arthropoda\_odb9 database [64]. The assembled unigenes were annotated by searching against public databases including the NCBI nonredundant (nr) database, the nucleotide (Nt) database, Swiss-Prot, the COG database and the KEGG database (e-value < 0.00001) using BLASTx version 2.2.23 [65]; GO classification was performed with Blast2GO [66]. ESTScan version 3.0.2 was used to predict the coding regions of the unigenes that had no hits in any databases [67].

#### *4.5. Comparative Transcriptomic Analysis*

The clean reads for the cement glands and carcasses were mapped to the assembled whole transcriptome (all-unigenes) with software Bowtie 2 version 2.2.5 [68]. The number of reads mapped to every unigene was counted with SAMtools [69]. The expression levels of the unigenes were quantified as FPKM [70]. DEGs were analyzed with the in-house software PossionDis [71,72] based on the Poisson distribution according to Audic and Claverie (1997), which can provide a quantitative assessment of differential expression without replicates [73]. *p*-values were then corrected by the FDR [74] in multiple testing. The threshold of FDR ≤ 0.001 and |log2 Fold Change| ≥ 1 was used to determine the DEGs. GO and KEGG enrichment analyses were performed based on the cumulative hypergeometric distribution method using the online OmicShare tools (version 1.0, http://www.omicshare.com/tools).

#### *4.6. Protein Extraction and in-Solution Digestion*

The same cement gland sample that was used for RNA extraction with TRIzol Reagent (Invitrogen) was also subjected to protein extraction according to the manufacturer's protocol. Briefly, after phase separation, the interphase and organic phenol-chloroform phase was subjected to a protein isolation procedure; in the last protein resuspension step, 100 μL of 10 M urea with 50 mM DTT (Sigma, St. Louis, MO, USA) was added to the protein pellet. The mixture was vortexed thoroughly and centrifuged at 16,000× *g* for 10 min, and the supernatant was transferred into a new tube. The protein concentration was quantified with an RC-DC kit (Bio-Rad, Hercules, CA, USA) following the manufacturer's instructions. For in-solution digestion, protein was alkylated by incubation in 40 mM

iodoacetamide (Sigma) for 20 min at room temperature in the dark and then diluted 10-fold with 25 mM tetraethylammonium bromide (TEAB; Sigma). Trypsin (Promega, Madison, WI, USA) was added at an enzyme-to-substrate ratio of 1:50 (*w*/*w*), and the mixture was incubated for 16 h at 37 ◦C. After tryptic digestion, the peptide solution was desalted with Sep-Pak C18 cartridges (Waters, Milford, MA, USA) and dried using a SpeedVac (Thermo Electron, Waltham, MA, USA).

#### *4.7. LC–MS*/*MS Analysis*

Dried peptide fractions were reconstituted with 0.1% formic acid and analyzed two times using an LTQ-Orbitrap Elite coupled to an Easy-nLC system (Thermo Fisher, Bremen, Germany) as described previously [75]. Raw data obtained from LC–MS/MS analysis were converted into MGF format files with the software Proteome Discovery version 1.3.0.339 (Thermo Finnigan, San Jose, CA, USA) and then searched against the protein database deduced from the *M. volcano* cyprid transcriptome with 67,764 sequences including both 'target' and 'decoy' sequences with Mascot version 2.3.02 (Matrix Sciences, London, UK). The search parameters used were identical to those described by Mu and colleagues [75], except that the ion score cut-off was set to 25 to achieve 95% confidence in identification. Proteins with at least one matched unique peptide were retained, and the threshold of 1% FDR was used for the final protein identification. Proteomic data are available via ProteomeXchange with identifier PXD012779.

#### *4.8. Sequence Analysis*

Transcription factor identification was performed with the software DIAMOND version 0.8.23 [76] based on AnimalTFDB version 2.0 [77]. The SignalP 4.1 server was used to predict the presence and location of signal peptide cleavage sites [78], and the TMHMM Server 2.0 was used to predict transmembrane helices in proteins [79]. The NetNGlyc 1.0 Server and NetOGlyc 4.0 Server were used to predict *N*-Glycosylation sites and mucin type GalNAc O-glycosylation sites in proteins [80]. Sequence alignment was performed with ClustalX 2.1 (EMBL, Heidelberg, Germany) with the deduced protein sequences as input. Sequence alignment results were shaded with DNAMAN 8.0 (Lynnon Biosoft, San Ramon, CA, USA).

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1660-3397/18/4/186/s1, Figure S1: All-unigenes classification, Figure S2: Sequence alignment of cp100k homologues from different species, Figure S3: CDD analysis of the lipid-binding proteins, Table S1: List of predicted transcription factor-coding unigenes that were upregulated in the cement gland transcriptome, Table S2: List of potential novel cement proteins, Table S3: Amino acid composition of cp100k homologues from different species, Table S4: Classification of all the enzyme-coding unigenes in the cement gland transcriptome, Table S5: List of enzymes involve in chitin synthesis and degradation, Table S6: List of lipid-binding proteins identified in the cement gland proteome.

**Author Contributions:** L.H., P.-Y.Q. and G.Y. conceived and designed the project. G.Y. prepared the samples, performed the wet experiments and drafted the manuscript. G.Y. and Z.W. performed the bioinformatics analysis. J.S. performed the protein search with Mascot. All authors reviewed the manuscript. L.H. supervised the work. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by grants from the National Key Research and Development Program of China (2016YFC0302504), the National Natural Science Foundation of China (31460092), the National Key Research and Development Program of China (2016YFC0304905) and the Hong Kong Branch of Southern Marine Science and Engineering Guangdong Laboratory (SMSEGL20Sc01).

**Acknowledgments:** Special thanks to Ling Fang from the Instrumental Analysis and Research Center of Sun Yat-Sen University for performing the LC–MS/MS analysis.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **Marine Microbial-Derived Antibiotics and Biosurfactants as Potential New Agents against Catheter-Associated Urinary Tract Infections**

**Shuai Zhang 1, Xinjin Liang 2,3, Geoffrey Michael Gadd <sup>3</sup> and Qi Zhao 4,\***


**Abstract:** Catheter-associated urinary tract infections (CAUTIs) are among the leading nosocomial infections in the world and have led to the extensive study of various strategies to prevent infection. However, despite an abundance of anti-infection materials having been studied over the last forty-five years, only a few types have come into clinical use, providing an insignificant reduction in CAUTIs. In recent decades, marine resources have emerged as an unexplored area of opportunity offering huge potential in discovering novel bioactive materials to combat human diseases. Some of these materials, such as antimicrobial compounds and biosurfactants synthesized by marine microorganisms, exhibit potent antimicrobial, antiadhesive and antibiofilm activity against a broad spectrum of uropathogens (including multidrug-resistant pathogens) that could be potentially used in urinary catheters to eradicate CAUTIs. This paper summarizes information on the most relevant materials that have been obtained from marine-derived microorganisms over the last decade and discusses their potential as new agents against CAUTIs, providing a prospective proposal for researchers.

**Keywords:** marine microorganisms; urinary catheter; antibiofilm; antifouling; coating

#### **1. Introduction**

Urinary catheters are hollow, partially flexible tubes that are designed to drain urine from the bladder. The earliest use of urinary catheters can be traced back to the third century B.C., but the modern indwelling catheter, called the 'Foley' catheter, was designed by Frederick B. Foley in the mid-1930s [1]. To date, over 100 million urinary catheters are used worldwide per year since catheterization rates remain high at 20% in non-intensive care units and 61% in intensive care units (ICUs) [2]. Despite the care taken to avoid contamination, catheters are still susceptible to infections as they provide direct access for uropathogens from the outside environment into the urinary tract, impairing local host defence mechanisms of the bladder [3,4]. Opportunistic uropathogens (Table 1) are mainly faecal or skin microbiota from the patients that can enter the bladder through the catheter lumen (34%) or along the catheter–urethral interface (66%) causing infections and complications, such as encrustation, bladder stones, bacteriuria, pyelonephritis, septicaemia and endotoxic shock (Figure 1) [3,5–7]. According to the European Centre for Disease Prevention and Control (ECDC), catheter-associated urinary tract infections (CAUTIs) account for 27% of all hospital-acquired infections in developed countries, and over 1 million cases occur in the USA and Europe [1,8]. In the UK, CAUTIs cost the NHS GBP 1–2.5 billion and account for approximately 2100 deaths annually [9].

**Citation:** Zhang, S.; Liang, X.; Gadd, G.M.; Zhao, Q. Marine Microbial-Derived Antibiotics and Biosurfactants as Potential New Agents against Catheter-Associated Urinary Tract Infections. *Mar. Drugs* **2021**, *19*, 255. https://doi.org/ 10.3390/md19050255

Academic Editor: Tom Turk

Received: 1 April 2021 Accepted: 27 April 2021 Published: 29 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Figure 1.** Anatomical cross-section of the renal system in a male showing CAUTIs.

Depending on clinical indications, the duration of catheterization may be short- (<7 days) or long-term (>28 days). Early work showed that 10–50% of patients undergoing short-term catheterization developed bacteriuria, and all patients undergoing long-term catheterization became infected, regardless of whether the catheter system was open or closed [10]. As a basic survival strategy, bacteria that encounter a catheter surface submerged in urine become attached within minutes [11]. These attached bacteria begin to phenotypically change, producing extracellular polymeric substances (EPS) (mainly exopolysaccharides and proteins) that allow the emerging biofilm community to develop a complex, three-dimensional structure within hours. Once established, the biofilms are very difficult to eradicate and exhibit a high tolerance to antibiotics and other biocidal treatments, making them a continuous focus for infections that can only be eliminated by the constant removal of the catheters [3,7,12–15]. Furthermore, in the presence of ureasepositive bacteria (e.g., *Proteus mirabilis*), urease catalyses the hydrolysis of urea into carbon dioxide and ammonia, which increases the urine pH, leading to the precipitation of calcium and magnesium phosphate crystals (encrustation) and the formation of crystalline biofilms on the catheter, which eventually results in the complete blockage of the catheter [7,16,17]. Previous attempts to prevent CAUTIs include improving sterile techniques to inhibit the access of microbes into the urinary tract and limiting microbial accumulation on the catheter surface by intermittent catheterization. However, clinical evidence shows that these efforts do not lead to a noticeable reduction in CAUTIs [15]. Therefore, developing novel urinary catheters with antibiofilm and antiencrustation properties remains the most direct and promising strategy for this significant clinical problem.


GN: Gram-negative; GP: Gram-positive.

#### **2. Current Anti-Infection Strategies against CAUTIs and Challenges**

Current commercial urinary catheters can be generally classified as standard and antimicrobial catheters (Figure 2). Owing to their superior malleability, several materials used for making standard catheters include polyvinyl chloride (PVC), polyurethane (PU), silicone and latex [20]. Of these, silicone has emerged as the material of choice for urinary catheters due to distinct advantages, including excellent biocompatibility, no allergic reactions, superior chemical and thermal stability and good mechanical strength [2,26]. Morris et al. [27] compared the antiencrustation performance of 18 types of catheter and found that all-silicone urinary catheters took the longest time to encrust and block. The main reason for this lies in that the all-silicone catheter has a wider lumen, allowing a faster urine flow, which prevents the accumulation of crystalline deposits. However, recent studies demonstrated that there was no significant difference between the development of infections or bacterial adherence on silicone catheters as compared to other types of catheters [20,26,28,29].

**Figure 2.** Classification of urinary catheters in clinical use and recently reported antimicrobial and antifouling materials for the prevention of CAUTIs.

Given that bacterial adhesion is the critical step in the progression of biofilm formation, numerous attempts have been made to endow the catheters with antiadhesive or antimicrobial properties, or both. Antiadhesive coatings are designed to prevent microbial adhesion through mechanisms of steric repulsion, electrostatic repulsion or low surface energy instead of killing the microbes [15]. To date, hydrogel- and polytetrafluoroethylene (PTFE)-coated catheters are commercially available, but clinical studies demonstrate that their efficacies against CAUTI are insignificant when compared with standard catheters [2,4,26,30]. Despite their lubricating features that may help improve patient comfort, hydrogel- or PTFE-coated catheters are only suitable for short- or medium-term catheterization (<28 days). Antimicrobial coatings are characterized by bactericidal or bacteriostatic activities that protect the catheters from microbial adhesion and migration. Recent reviews of commercial antimicrobial catheters identified two main strategies used: silver-based coatings and antibiotic impregnation [1,7,31]. However, all these catheters have only been reported to yield positive results for short-term application [7,20,32].

Silver is among the few FDA-approved antimicrobial materials for urinary catheter coatings, and its antimicrobial activity is associated with the release of silver ions. To date, silver has been applied in catheter coatings in the forms of bulk silver, silver alloy (silver/gold or palladium) and silver-hydrogels. Silver is very prone to oxidation in aqueous conditions, and the release of silver ions from these coatings often undergoes an initial burst-release phase followed by a slow-release phase. In clinical trials, the long-term antimicrobial efficacy of these silver-based coatings has proven limited [1,28,33]. In this scenario, attempts have been made to introduce silver nanoparticles into catheter coatings to attain enhanced antimicrobial efficacy. However, concerns have also been raised about the potential toxicity towards patients due to the fast and excessive release of silver ions [33,34]. In comparison, despite studies suggesting that the overuse of antibiotics may result in the development of antibiotic resistance, certain types of antibiotic-impregnated catheters have been proven to be more effective than silver-based antimicrobial catheters in preventing CAUTIs [3,20,35–37]. For example, nitrofurazone-impregnated catheters are commercially available, and studies comparing silver-alloy coated catheters, silver-hydrogel coated catheters, and nitrofurazone-impregnated catheters have found that nitrofurazone could effectively reduce the risk of symptomatic CAUTI and bacteriuria in short-term catheterization by impairing bacterial adherence and planktonic growth, while silver-based catheters have only demonstrated minimal effects [28,38]. Pickard et al. [33] compared the ability of silver-alloy-coated catheters and nitrofural-impregnated catheters for the reduction in incidence of symptomatic CAUTIs in adults requiring short-term catheterization via a clinical model, and the results demonstrated that nitrofural-impregnated catheters were more effective than the silver-alloy-coated catheters. In addition, impregnating antibiotics into catheters provides a cost-effective strategy to manufacture antimicrobial catheters, as this can be achieved by simply submerging swollen catheters in antibiotic-containing solutions for antibiotic encapsulation [31]. However, there are still certain concerns about using nitrofurazone in urinary catheters, such as patient discomfort [38] and the potential risks of developing tumours [39] and resistance in bacteria. This has hindered research in this field, and there is a growing demand for developing new antibiotics instead.

Apart from silver and antibiotics, recent research has also focused on a variety of antifouling materials (e.g., poly(ethylene glycol) (PEG) and polyzwitterions) [2,40] and biocidal materials (e.g., nitric oxide, antimicrobial peptides, enzymes and bacteriophages) [41–44] for urinary catheter coatings and has reported with varying levels of success, which offers potential for complete protection against CAUTIs. However, the promising performance of anti-infection coatings under laboratory conditions has not been translated into clinical success to date. Therefore, exploring novel anti-infection materials for urinary catheters will remain a current and broad interest in the upcoming decade.

#### **3. Marine Microbiota as a Source of Novel Anti-Infection Materials**

For short-term urinary catheters, the use of antibiotics has proven to be a cost-efficient strategy to prevent CAUTIs in clinical trials, but a major concern is the development of antibiotic resistance, particularly the increasing emergence of new forms of multidrug resistance among uropathogens that can render these antibiotics useless after repeated applications. Therefore, the identification of new resources for novel antibiotics with new modes of action has become the research focus over the past two decades. Currently, over 75% of the antibiotics currently available on the market are derived from terrestrial organisms, and the discovery of new antibiotics has declined considerably (only two new classes of antibiotics have been commercialized since 1962) due to difficulties in identifying novel and effective compounds [45–47].

Oceans cover 71% of the Earth's surface, but less than 5% have been explored to date, presenting an unexplored area of opportunity [48,49]. In recent decades, significant progress in the clinical development of marine-derived drugs has been achieved, and the discovery of novel antimicrobial substances from marine organisms has indicated their potential to combat existing medical device-related infections [50–54]. However, there are concerns that the overexploration of marine organisms may affect the balance between ecological constraints and economic activity. In this scenario, marine microbiota are emerging as a viable source of bioactive materials [55]. Diverse marine habitats provide unique conditions for marine microorganisms to develop into complex and diverse assemblages, and the isolation and extraction of bioactive secondary metabolites from such organisms are relevant to the discovery of novel anti-infection agents [56–59]. Therefore, microbes isolated from previously unexplored marine habitats may lead to the discovery of novel structures with potent antibiotic activity, and marine microbial-derived antibiotics are believed to be a promising alternative to overcome existing problems [60–62].

For long-term catheters, microbes are more prone to colonize and build biofilms on the surfaces with the attenuation of antimicrobial efficacy. Catheters with only biocidal activity are considered insufficient to eradicate CATUIs, as recent studies demonstrated that a 'foundation layer' composed of both dead and live bacteria can form on the catheter surface in long-term catheterization, which protects bacteria from contact with underlying biocidal substances [4,63]. To solve this problem, more research has focused on endowing catheter surfaces with both biocidal and antiadhesion properties [4,9,15,64]. Biosurfactants (BSs) are amphipathic secondary metabolites produced by microorganisms and have become a promising anti-infection material for medical applications [65–67]. Recent studies of the anti-infection functions of biosurfactants have identified three main routes, including killing microbes or inhibiting microbial growth, resisting microbial adhesion and disrupting biofilm formation [68–70]. For example, several BSs isolated from terrestrial-derived microbes, such as surfactin [71] and iturin [72], display potent antimicrobial activities, which make them relevant molecules in combating infections. Van Hoogmoed et al. [73] reported that biosurfactants released by *Streptococcus thermophilus* significantly inhibited *Candida* spp. adhesion to silicone rubber, which indicates their potential for use as a defence against colonizing strains on urinary catheters. These biosurfactants are mostly obtained from terrestrial-derived microorganisms, while biosurfactants produced by marine microorganisms have been less explored. Over the last decade, marine microbial-derived biosurfactants have attracted increasing interest, as marine microbes may exhibit unique metabolic and physiological capabilities producing novel metabolites with potent biological activities. To date, a considerable number of marine microbial-derived biosurfactants with antimicrobial, antiadhesive and antibiofilm activities have been obtained, and several have been proven to be effective against a broad spectrum of uropathogens, including Grampositive and Gram-negative bacteria, as well as the yeast *Candida albicans*, which could potentially offer a new solution to the problems associated with long-term catheterization.

In this review, we summarize the diversity of marine microbial-derived antibiotics and biosurfactants discovered over the past 10 years and discuss their potential for use in combatting CAUTIs in short-term and long-term urinary catheters.

#### **4. Marine Microbial-Derived Antibiotics with Broad-Spectrum Antimicrobial Activity**

Antibiotics are low-molecular-weight compounds produced by microorganisms that kill or inhibit the growth of other microorganisms at low concentrations [74]. The current understanding of the biological activity of antibiotics is mainly cantered on their primary cellular targets. The mechanisms of antibiotic action can be classified into 5 categories: (1) inhibition of cell wall synthesis (the most common mechanism), (2) inhibition of protein synthesis, (3) alteration of cell membranes, (4) inhibition of nucleic acid synthesis and (5) antimetabolite activity [75,76]. However, not all antibiotics have broad-spectrum antimicrobial activity, and their efficiency decays along with catheterization due to a limited shelf-life. Some microbes may be inherently resistant to certain antibiotics, while they may also mutate and/or exchange genetic material with other microbes, leading to the development of multidrug resistance [77,78]. Furthermore, some antibiotics may exhibit toxicity at or close to their therapeutic dose [47]. Considering the diversity of uropathogen species, the ideal antibiotic for urinary catheters should exhibit broad-spectrum antimicrobial activity and avoid developing resistance without inducing cytotoxicity in patients.

Generally, antibiotic compounds based on their structures and/or biosynthetic origins can be classified as alkaloids [79], quinones [80], phenols [81], polyketides [82], terpenes [83], polyketides [84,85] and peptides [86]. Table 2 lists the recent discoveries of marine microbial-derived compounds with broad-spectrum antimicrobial activity. Although very few of them have been developed into clinical trial phases, several have demonstrated to be effective against a broad spectrum of uropathogens, including Grampositive and Gram-negative bacteria, fungi, as well as methicillin-resistant *Staphylococcus aureus* (MRSA). Therefore, they could be an alternative to conventional antibiotics for short-term catheters against CAUTIs caused by those pathogens.

Over the last two decades, the most studied group of microbes producing antimicrobial substances is the Firmicutes phylum (particularly the genus *Bacillus*) [87]. Tareq et al. [88–91] reported the discovery of four types of bioactive molecules with broad-spectrum antimicrobial activity against both Gram-positive and Gram-negative bacteria and fungi. Gageostatins A–C (Figure 3a) and gageopeptides A–D (Figure 3b) are two types of non-cytotoxic linear lipopeptides isolated from a marine *Bacillus subtilis* 109GGC020. The authors proposed a biosynthetic pathway based on nonribosomal peptide synthetases (NRPS) for the production of gageotetrins. Comparative studies showed that gageostatins A–C were more active against fungi than bacteria with minimum inhibitory concentration (MIC) values of 0.02–0.04 μM. Gageopeptides A–D exhibited potent antifungal and moderately broad antibacterial activity, while not showing cytotoxicity to human myeloid leukaemia K-562 and mouse leukemic macrophage RAW 264.7 cell lines. Gageomacrolactins A–C (Figure 3c) are three macrolactin derivatives obtained from the secondary metabolites of marine *Bacillus subtilis* 109GGC020 and displayed strong broad-spectrum activity against Gram-negative and Gram-positive bacteria and fungi. By comparing the structure-function of the gageomacrolactins A–C with that of known macrolactins 4–7, the authors demonstrated that antibacterial activity was not affected by the position of the epoxide group but highly dependent on the hydroxyl group at C-15 of the macrolactone ring. Ieodoglucomides 1 and 2 (Figure 3d) are two unique glycolipopeptides produced by a marine *Bacillus licheniformis* 09IDYM23 which act as moderate antimicrobial molecules. Podilapu et al. [92] reported the successful synthesis of ieodoglucomides A and B through a high-yielding route using per-O-TMS glucosyl iodide, making them promising candidates for industrial application.

Apart from *Bacillus* spp., Uzair et al. [93] reported the isolation of 4-[(Z)-2 phenyl ethenyl] benzoic acid (kocumarin) from the marine *Kocuria marina* CMG S2, which exhibited pronounced and rapid growth inhibition against fungi and pathogenic bacteria, including methicillin-resistant *Staphylococcus aureus* (MRSA). Although its antimicrobial mechanism has not been clearly elucidated, research on the functional groups involved in the molecular mechanism would reveal further insights into this atypical class of antibiotics. Schumacher et al. [84] described the first antimicrobial polyketide (bonactin) isolated from the liquid culture of a *Streptomyces* sp. BD21-2, a compound displaying antifungal and broad-spectrum antibacterial activity against microbes, including *Bacillus megaterium*, *Micrococcus luteus*, *Klebsiella pneumoniae*, *Staphylococcus aureus*, *Alcaligenes faecalis*, *Escherichia coli* and *Saccharomyces cerevisiae*. However, such marine microbial-derived compounds with activity against both bacteria and fungi are still very rare to date.

Over the past 10 years, numerous molecules possessing broad-spectrum antibacterial or antifungal activities have been isolated from marine microorganisms. Glycosylated macrolactins A1 and B1 (Figure 3e) were isolated from a marine *Streptomyces* sp. KJ371985, which inhibited *Bacillus subtilis*, *Escherichia coli*, *Pseudomonas aeruginosa* and *Staphylococcus aureus* with MICs of 0.03–0.22 μM. These compounds inhibit the peptidyl transferase activity and binding of the acceptor substrate to bacterial ribosomes. The higher solubility in the polar solvents of sugar-containing macrolactins could be an advantage in clinical applications [94]. Bacteriocins are natural peptides synthesized by bacteria for the purpose of killing/inhibiting other bacterial strains whilst not harming the producing bacteria through specific immunity proteins [95]. Elayaraja et al. [96] purified a bacteriocin from marine *Lactobacillus murinus* AU06, which exhibited a broad inhibitory spectrum against both Gram-positive and -negative

bacteria. Marinocine is a broad-spectrum antibacterial protein synthesized by the melanogenic marine bacterium *Marinomonas mediterranea*, which generates hydrogen peroxide that kills bacteria [97]. However, recent studies demonstrated that the molecular basis of the antibacterial activity was L-lysine dependent, and activity was inhibited under anaerobic conditions. Mollemycin A (Figure 3f) is an antibacterial glyco-hexadepsipeptide-polyketide isolated from an Australian marine-derived *Streptomyces* sp. (CMB-M0244), which exhibited exceptionally potent and selective growth inhibitory activity against Gram-positive and Gram-negative bacteria (IC50 10–50 nM) but did not show any antifungal activity against *Candida albicans*. The cytotoxicity test also showed that mollemycin A was proportionately less cytotoxic toward human neonatal foreskin fibroblast cells [98]. Thiomarinols A–G (Figure 3g) were discovered as a class of polyketide antibiotics from marine *Alteromonas rava* SANK 73390, which displayed broad-spectrum activity against Gram-positive and Gram-negative bacterial species [99]. These thiomarinols act through inhibiting bacterial isoleucyl-transfer RNA synthetase and have pronounced activity against MRSA, with MICs ≤ 0.01 μg/mL [60]. Similar antibacterial compounds were also obtained from actinobacteria, such as *Pseudonocardia carboxydivorans* M-227 [100] and *Streptomyces* sp. JRG-04 [101]. Fungi (primarily *Candida species*) account for 20–30% of CAUTI cases, and *Candida albicans* is the most prevalent pathogen found in CAUTI biofilms [1,37]. Antifungal (particularly anti-*Candida*) compounds derived from marine microbes have been reported for metabolites produced by *Streptomyces* sp. ZZ338 [102], *Streptomyces* sp. SNM55 [103], *Bacillus subtilis* KC433737 [104], *Janthinobacterium* spp. ZZ145 and ZZ148 [105], *Trichoderma* sp. MF106 [106] and *Stagonosporopsis cucurbitacearumstrain* G019 [107]. These microbial strains, respectively produced actinomycins D, V and X0<sup>β</sup> (Figure 4a) (MIC, 9.83–9.96 μM); mohangamides A and B (Figure 4b) (IC50, 4.4 and 20.5 μM); 5-hydroxymethyl-2-furaldehyde (5HM2F) (Figure 4c) (MBIC, 400 μg/mL); janthinopolyenemycin A and B (Figure 4d) (MIC, 15.6 μg/ML, MBC, 31.25 μg/mL); trichodin A (Figure 4e) (IC50, 25.38 ± 0.41 μM); and didymellamide A (Figure 4f) (MIC, 3.1 μg/mL).

**Table 2.** Marine microbial-derived compounds with broad-spectrum antimicrobial activity.


B: Gram-positive and Gram-negative bacteria; F: fungi; UN: unnamed material.

**Figure 3.** Structures of (**a**) gageosatins A–C [90], (**b**) gageopeptides A–D [91], (**c**) gageomacrolactins A–C [89], (**d**) ieodoglucomides 1 and 2 [88], (**e**) glycosylated macrolactins A1 and B1 [94], (**f**) mollemycin A [98], and (**g**) thiomarinols A–G [99].

**Figure 4.** Structures of (**a**) actinomycins D, V and X0<sup>β</sup> [102], (**b**) mohangamides A and B [103], (**c**) 5-hydroxymethyl-2 furaldehyde (5HM2F) [104], (**d**) janthinopolyenemycin A and B [105], (**e**) trichodin A [106], and (**f**) didymellamide A [107].

#### **5. Marine Microbial-Derived Biosurfactants: New Agents against CAUTIs**

Biosurfactants (BSs) are amphipathic secondary metabolites produced by microorganisms that consist of both hydrophilic and hydrophobic moieties [45]. Compared to conventional chemical surfactants, these biomolecules have many advantages, such as low toxicity; high biodegradability; biocompatibility; low critical micelle concentrations (CMC); an ability to function over wide ranges of pH, temperature and salinity; as well as greater selectivity, and can be produced from renewable, cheaper substrates [110,111]. Such characteristics allow BSs to play a key role in multidisciplinary applications in industrial and environmental fields and make them a green alternative to their chemical counterparts. Furthermore, some BSs have been reported for their specific bioactivities, which play an essential role in the survival of microbial producers against other competing microbes [69,112]. In nature, BSs can be secreted extracellularly or remain attached to cell surfaces, reducing surface tension at the interface, thereby reducing surface contamination and aiding microbial motility in potentially hostile environments [113]. BSs may also have a range of therapeutic and biomedical benefits and could be used instead of conventional antibiotics to combat infections [114]. Current research regarding BSs has mostly focused on microbes from terrestrial sources (particularly soil-isolated microbes such as species of *Bacillus*, *Pseudomonas* and yeasts) [45,112], while practical applications of BSs in healthcare are still limited [115]. For this reason, marine habitats have again emerged as a potential source for isolating new BSs, and the discovery of new BS-producing microorganisms is attracting increasing interest.

BSs according to their molecular weight can be grouped into two categories: (a) high-molecular-mass (HMW) molecules (also called bioemulsifiers), such as polysaccharides, proteins, lipopolysaccharides, lipoproteins and lipoheteropolysaccharides, and (b) low-molecular-mass (LWM) molecules (generally 500 to 1500 Da), which can be further subdivided into glycolipids, lipopeptides, phospholipids, polymeric compounds and neutral lipids based on their chemical composition and microbial origin [45,65,116]. Recent

research has mainly focused on the LWM BSs, and several have been reported to display antimicrobial, antiadhesive and antibiofilm activities against a broad spectrum of uropathogens (including multidrug-resistant pathogens), indicating their use as promising anti-infection materials for urinary catheters (Table 3).

#### *5.1. Antimicrobial Activity*

Biosurfactants have been reported to exhibit antimicrobial activity via different mechanisms of action, which primarily destroy the cell wall or plasma membrane by disrupting their integrity and permeability [65,68,117,118]. More specifically, their amphiphilic characteristics and affinity for lipid bilayers allow BSs to interact with cell membranes, leading to cell lysis and metabolite leakage, which ultimately results in cell death [116]. To date, various marine microbes have been reported to produce antimicrobial BSs, of which glycolipids and lipopeptides are the two primary isolated families that display broad-spectrum antimicrobial activity.

Glycolipids are a class of carbohydrate molecules made of mono-, di-, tri- and tetrasaccharides in combination with long-chain aliphatic acids or hydroxyaliphatic acids [119], of which trehalolipids, rhamnolipids and sophorolipids are of the most interest [120]. Reported marine microbial-derived glycolipid BSs with broad-spectrum antimicrobial activity against both bacteria and fungi are listed in Table 3. A glycolipid BS isolated from a marine *Staphylococcus saprophyticus* SBPS 15 exhibited dual functions: (1) excellent antibacterial and antifungal activities against a broad spectrum of clinical human pathogens, including Gram-positive bacteria (*Bacillus subtilis* and *Staphylococcus aureus*), Gram-negative bacteria (*Escherichia coli*, *Pseudomonas aeruginosa*, *Klebsiella pneumoniae* and *Salmonella paratyphi*) and fungi (*Aspergillus niger*, *Candida albicans* and *Cryptococcus neoformans*), and (2) potent surface tension-reducing activity (32 mN/m). This BS also showed superior stability over a broad range of pH (3–9) and temperature (up to 80 ◦C) [121]. Another glycolipid BS purified from extracts of a tropical marine strain of *Serratia marcescens* demonstrated similar dual functions, inhibiting the growth of *Candida albicans* and *Pseudomonas aeruginosa* with MIC values of >25.0 μg/mL and preventing adhesion by up to 99% [122]. The glycolipid BS also displayed biofilm disrupting activities against the test strains, which were believed to result from synergistic antimicrobial and surfactant activity. Glycolipid BSs were also isolated from the marine actinobacteria *Brevibacterium casei* MSA19, *Streptomyces* sp. MAB36 and *Brachybacterium paraconglomeratum* MSA21. The MSA19 glycolipid was bacteriostatic and could inhibit microbial attachment and disrupt both fungal and bacterial biofilms in both individual strains and mixed cultures at 30 μg/mL [123]. The MAB36 glycolipid also possessed strong antimicrobial activity against pathogenic bacteria and fungi and demonstrated excellent stability over a wide range of pH, temperature and ion concentrations [124]. The BS produced by the sponge-associated *Brachybacterium paraconglomeratum* MSA21 displayed broad antibiotic activity and it was suggested that the discovery of polyketide synthase (pks II) genes might pave the way for making new BSs by exploiting the microbial genes and enzymes, making the green production of BSs possible in the future [125]. Another glycolipid BS was isolated from a marine halotolerant bacterium (*Buttiauxella* sp. M44) that exhibited significant stability over a broad range of pH (7–8), temperature (20–60 ◦C) and salinity (0–3%) and potent antimicrobial activity against both bacteria and fungi [126]. Moreover, the combination of a cheap energy and carbon source (e.g., molasses) as well as a response surface method (RSM) could favor production and expand possible uses in the future.

Lipopeptides (LPs) are composed of peptide chains (short linear or cyclic structures) with lipid moieties and are the most widely reported class of biosurfactants with antimicrobial activity. *Bacillus* species are the most studied LP-producing strains that have been reported to produce several antimicrobial lipopeptide biosurfactants (LPBs), such as surfactin, iturin and fengycin, although most of these strains are from terrestrial ecosystems [70,72]. Liu et al. [71] reported the first surfactin (mainly composed of the nC14 and anteiso-C15-surfactin) isolated from marine *Bacillus velezensis* H3, which exhibited

antimicrobial activity against a broad range of pathogens, including *Staphyloccocus aureus*, *Klebsiella peneumoniae*, *Pseudomonas aeruginosa* and *Candida albicans*. The surfactin demonstrated a lower inhibitory effect than the antibiotic polymixin B but greater antifungal activity against *Candida albicans* than the antibiotic vancomycin. Similar antibacterial and antifungal activities were also found for the BSs produced by *Halobacterium salinarum* [127]. Another surfactin produced by the marine actinobacterium *Nocardiopsis alba* MSA10 was shown to effectively inhibit the growth of *Enterococcus faecalis*, *Klebsiella pneumoniae*, *Micrococcus luteus*, *Proteus mirabilis*, *Staphylococcus aureus*, *Staphylococcus epidermidis* and *Candida albicans*, however, without effects on *Escherichia coli* and *Pseudomonas aeruginosa* [128]. Apart from surfactin, Balan et al. [129] first reported the discovery of aneurinifactin from the marine *Aneurinibacillus aneurinilyticus* SBP-11, which showed broad-spectrum antibacterial activity against *Klebsiella pneumoniae*, *Escherichia coli*, *Staphylococcus aureus*, *Pseudomonas aeruginosa*, *Bacillus subtilis* and *Vibrio cholerae*. The antibacterial mechanism was proposed to be derived from the anchoring of the BS on the bacterial cell membrane that disrupted its integrity, resulting in the production of hydroxyl radicals, which in turn caused lipid peroxidation and pore formation in the membrane [129]. However, antifungal activity was not investigated. It was also reported that a LPB (pontifactin) produced by marine *Pontibacter korlensis* SBK-47 displayed high antibacterial activity against *Streptococcus mutans*, *Micrococcus luteus*, *Salmonella typhi* and *Klebsiella oxytoca* and moderate activity against *Klebsiella pneumonia* and *Vibrio cholerae*. The molecule also showed antiadhesion potential against *Bacillus subtilis*, *Staphylococcus aureus*, *Salmonella typhi* and *Vibrio cholerae* [130]. Lawrance et al. [131] discovered an antibacterial LPB from the marine sponge-associated bacterium *Bacillus licheniformis* NIOT-AMKV06, which displayed significant bacteriostatic activity against a range of human pathogens, including *Enterococcus faecalis*, *Klebsiella pneumoniae*, *Micrococcus luteus*, *Proteus mirabilis*, *Salmonella typhi*, *Shigella flexneri*, *Staphylococcus aureus* and *Vibrio cholerae*. The production of these biosurfactants in heterologous host strains was achieved by cloning three gene clusters (sfp, sfpO and srfA) in *Escherichia coli*, with production being increased three-fold over the original strain. In addition, a lipopeptide BS produced by the marine *Bacillus circulans* is the only one found to be effective against multidrug-resistant (MDR) clinical strains while not having any haemolytic activity, indicating its potential to combat infections caused by such pathogens [132].

#### *5.2. Antiadhesive and Antibiofilm Activities*

Biosurfactants have also been reported to inhibit microbial adhesion and biofilm formation without exerting antimicrobial activity [69,133,134]. Despite the precise mechanisms of such activity not being fully understood, biosurfactants may affect interactions between microbes and surfaces through several modes of action: (1) modification of the physico-chemical properties (e.g., surface charge, hydrophobicity and surface energy) of the surface, which reduces microbial adhesion [111]; (2) suppression of the expression of biofilm-related genes, which inhibits microbial adhesion [135]; (3) promotion of the solubilization of biofilms encouraging bacterial detachment [136]; and (4) the interference of quorum sensing leading to decreased biofilm formation [70]. Numerous studies have shown that the prior adhesion of BSs to catheter surfaces (surface conditioning) reduced microbial adhesion and colonization [137–139].

In general, bacterial adhesion to a surface is regulated by diverse factors (e.g., growth medium, substrate and cell surface). The most frequently cited theory is the two-step adhesion model in which the adhesion process is divided into two distinct phases: reversible adhesion and irreversible adhesion [11,12,140]. Reversible adhesion describes a dynamic process in which bacteria can easily attach to or detach from a surface due to a combination of surface–cell interactions including van der Waals, electrostatic double-layer, Lewis acid-base, Brownian motion and hydrophobic interactions [64,141]. Walencka et al. [142] reported that BSs can affect cell-to-surface interactions by altering surface tension and charge to overcome the initial electrostatic repulsion barrier. For instance, Meylheuc et al. [143] reported that a stainless-steel surface coated with the BS produced by *Pseudomonas fluorescens*

significantly reduced microbial adhesion. The presence of BS in the conditioned surface remarkably reduced the surface energy and enhanced surface hydrophilicity, leading to a decrease in attraction due to a reduction in the van der Waals forces and an increase in electron–donor/electron–acceptor characteristics. Jemil et al. [144] also reported that electrostatic repulsion between negatively charged surfaces (coated with anionic lipopeptides) and the negatively charged microbial surface could aid in inhibiting microbial adhesion. However, following reversible adhesion, microbial attachment gradually becomes stronger with time through a range of interfacial rearrangements (e.g., removal of interfacial water, protein conformational changes and an increase in hydrophobic interactions) and the production of adhesins, eventually form biofilms [145]. Despite numerous studies highlighting the difficulties in eliminating biofilms, BSs have been shown to disrupt or remove biofilms by penetrating and absorbing at the interface between the solid substrate and the biofilm, thereby reducing interfacial tension [134,146].

Antiadhesive and antibiofilm activities have also been reported for marine microbialderived BSs. Hamza et al. [134] reported a non-toxic glycolipid BS (SLSZ2) derived from a marine epizootic bacterium *Staphylococcus lentus* SZ2 that effectively prevented the adhesion of *Vibrio harveyi* and *Pseudomonas aeruginosa* without showing bactericidal activity. In addition, this BS efficiently inhibited biofilm formation and disrupted the pre-formed biofilms of both strains by ~80%. Another glycolipid BS produced from a marine *Symphylia* sp. also exhibited antiadhesive activity against a range of pathogens (*Candida albicans*, *Pseudomonas aeruginosa* and *Bacillus pumilus*) and could disrupt pre-formed biofilms of these cultures in a concentration-dependent manner [122]. Similarly, a glycolipid BS isolated from the marine actinobacterium *Brevibacterium casei* MSA19 disrupted biofilm formation in *Escherichia coli*, *Pseudomonas aeruginosa* and *Vibrio* spp. under dynamic conditions. Moreover, biofilm disruption activity was consistent against mixed and individual biofilm bacteria at ~30 μg/mL. The lipopeptide BS produced by marine *Bacillus circulans* exhibited promising antiadhesive activity against several potential pathogenic strains, and the BScoated surface effectively reduced microbial adhesion by up to 89% at a concentration of 0.1 g/L. Moreover, the pre-formed biofilms were removed with efficiencies between 59 and 94% for all the strains tested, demonstrating its potential in biomedical applications [147]. Song et al. [148] reported a lipopeptide BS produced by marine *Bacillus amyloliquefaciens* anti-CA that inhibited biofilm formation and dispersed pre-formed biofilms of *Pseudomonas aeruginosa* PAO1 and *Bacillus cereus*. The obtained data indicated that the BS could suppress the expression of the PslC gene which is associated with exopolysaccharide production in *Pseudomonas aeruginosa* PAO1. Pradhan et al. [149] also reported a lipopeptide BS produced by a marine *Bacillus tequilensis* CH which effectively inhibited pathogenic biofilms (*Escherichia coli* and *Streptococcus mutans*) on both hydrophilic and hydrophobic surfaces at a concentration of 50 μg/mL.




#### **Table 3.** *Cont.*

#### **6. Opportunities, Challenges and Future Perspectives**

Currently, research on anti-infection materials for urinary catheters is still increasing, and there is a growing demand for catheters with stronger antibiofilm and antiencrustation capabilities. Compared to the conservative strategy of optimizing existing products (e.g., widening the internal lumen [1], replacing bulk silver with silver nanoparticles [4], and redesign of micro/nano-scale surface topography [153]), the exploration of novel antiinfection materials from marine resources has clearly become an exciting potential solution to address some current limitations. Owing to the vast diversity of marine microorganisms, a number of new bioactive molecules with antimicrobial, antiadhesive and antibiofilm activities have been discovered that could be potentially combined with urinary catheters to combat CAUTIs.

The antimicrobial strategy aims to endow urinary catheters with microbicidal or microbiostatic properties to prevent the contact of microbes with the catheter surface or inhibit microbial migration along the catheter, thus preventing biofilm formation and encrustation. For short-term catheters, the use of antibiotics has proven to be the most efficient and cost-effective strategy, whereas the application of conventional antibiotics (e.g., nitrofurazone) has been questioned due to the safety concerns and the presence of antibiotic resistance. Despite attempts having been made to develop new antibiotics to solve this problem, the difficulty in identifying novel and effective compounds has led to a slowdown of research in this field. In recent decades, the isolation of microbes from previously unexplored marine habitats has led to the discovery of a number of novel antimicrobial compounds with new modes of action to combat the current antibiotic resistance threat. Several of these have been proven to possess broad-spectrum antimicrobial activities against a wide range of uropathogens, which could be potentially applied to urinary catheters to eradicate CAUTIs. Technically, these antimicrobial molecules could be directly deposited on the catheter surfaces or applied in coatings. Considering that CAUTIs may be derived from both extraluminal and intraluminal routes, it would be ideal to endow both surfaces with antimicrobial properties, despite extraluminal infections being clinically more common. For example, commercial catheters have been conventionally impregnated with antimicrobials, such as antibiotics, and function via a release model. However, this is only a temporary solution for short-term catheterization, and microbial contamination can resume once the antibiotics are removed. Given that the efficacy of antimicrobial materials is often concentration dependent, a challenge in this type of catheter modification strategy is to achieve and preserve adequate local delivery of the antimicrobials during catheterization in vivo without inducing resistance and patient cytotoxicity. This could be achieved by developing a novel release system with controlled release kinetics instead of directly altering the drug loading. For example, hydrogel coatings, due to their lubricating properties, have been widely applied in urinary catheters. Milo et al. [154] reported a smart infection-responsive hydrogel coating for urinary catheters. The coating is a duallayered polymeric system consisting of a lower 'reservoir' layer of poly(vinyl alcohol) (PVA) hydrogel (containing bacteriophage), capped by an upper layer of the pH-responsive polymer. The upper layer swells when urinary pH is elevated due to *P. mirabilis* infection, exposing the PVA reservoir layer to urine, resulting in the release of bacteriophage from the coating to prevent the encrustation and blockage of urinary catheters. In this way, the release of antimicrobials could be controlled in a smart way to prevent/retard CAUTIs and associated complications. Researchers will need to further conduct thorough in vitro and in vivo leaching and cytotoxicity studies to determine the optimum operational conditions and identify any problematic side effects.

In long-term catheterization, extensive biofilms containing >5 × <sup>10</sup><sup>9</sup> viable cells/cm<sup>2</sup> can be found on catheters and can be responsible for the persistence of the infections [155]. Moreover, the biofilms are often composed of multispecies consortia, which have been reported to be more virulent and have a higher tolerance to antibiotics [7]. Therefore, in addition to killing the uropathogens, attempts have also been made to combine catheter surfaces with antiadhesive materials to retard the occurrence of CAUTIs. To date, only

PTFE and hydrogel have entered clinical use, while their efficacy in resisting CAUTI and encrustation is still controversial [4,17,33]. Other antiadhesive materials, such as polyzwitterions, have recently emerged as promising candidates, although their instability in long-term catheterization hinders practical application [1]. In this scenario, research on marine microbes has led to the discovery of novel antiadhesive molecules with greater activity and stability. For instance, marine microbial-derived antiadhesive materials, such as biosurfactants and cyanobacterial extracellular polymers [59,156], can inhibit microbial adhesion via multimechanisms. Currently, several biosurfactants have been reported to influence both cell-to-cell and cell-to-surface interactions that can inhibit microbial adhesion as well as disrupt biofilm formation [67,70,112]. Technically, biosurfactants could be doped in coatings and act on microbial cells via a contact mode. These features may aid in preventing biofilm formation and encrustation as well as extending the lifetime of indwelling urinary catheters. Furthermore, they could be employed in combination with other strategies (e.g., use of antimicrobial materials and micro/nano patterning) to achieve synergistic effects.

In fact, although these marine microbial-derived molecules have shown potential in combating CAUTI-related pathogens and/or biofilms, it should be noted that their clinical effectiveness still needs to be verified in future. Currently, there has been a lack of accomplished research in this area, and this paper provides a prospective proposal for researchers. Currently, the most significant obstacles hindering the development of marine microbial-derived products are supply issues. This is mainly due to difficulties associated with the isolation and growth of producing microorganisms. In addition, for marine microbial-derived biosurfactants, most of these materials are unidentified mixtures of compounds, and the further pharmaceutical development of such mixtures is very challenging. To date, the development of new chemical and physicochemical approaches and tools has led to great advances in the isolation and structure elucidation of novel minor marine secondary metabolites, which could not be isolated/detected in the past [54]. On the other hand, urinary catheters are still considered as low-cost clinical care products, although commercial 'anti-infective' catheters are usually priced two to five times higher than standard catheters. Therefore, in addition to the cost of employing new materials, a cost-effective coating technology should also be addressed.

#### **7. Conclusions**

Marine microorganisms can produce metabolites of varied chemical structures displaying antimicrobial, antiadhesive and antibiofilm activities. These compounds present enormous potential for the discovery of new agents to overcome the current challenges associated with CAUTI. Owing to technical limitations in the past, bioactive compounds derived from marine microorganisms have not yet progressed into clinical trials. However, recent advances in oceanographic science and metabolome screening have led to a boost in the discovery of new species and metabolic profiles as well as new molecules with potent anti-infection activities against CAUTI-associated pathogens. Research efforts should be applied to the exploration of marine microbes in the search for new bioactive compounds for the development of novel anti-infection agents.

**Author Contributions:** S.Z. and X.L. are first co-authors. Conceptualized and writing, S.Z. and X.L.; Revision and editing, G.M.G. and Q.Z.; Supervision and resources, Q.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the UK Engineering and Physical Sciences Research Council (EP/P00301X/1).

**Acknowledgments:** The authors acknowledge the financial supports from the UK Engineering and Physical Sciences Research Council (EP/P00301X/1).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **Research Strategies to Develop Environmentally Friendly Marine Antifouling Coatings**

**Yunqing Gu 1, Lingzhi Yu 1, Jiegang Mou 1,\*, Denghao Wu 1, Maosen Xu 1, Peijian Zhou <sup>1</sup> and Yun Ren <sup>2</sup>**


Received: 22 June 2020; Accepted: 16 July 2020; Published: 18 July 2020

**Abstract:** There are a large number of fouling organisms in the ocean, which easily attach to the surface of ships, oil platforms and breeding facilities, corrode the surface of equipment, accelerate the aging of equipment, affect the stability and safety of marine facilities and cause serious economic losses. Antifouling coating is an effective method to prevent marine biological fouling. Traditional organic tin and copper oxide coatings are toxic and will contaminate seawater and destroy marine ecology and have been banned or restricted. Environmentally friendly antifouling coatings have become a research hotspot. Among them, the use of natural biological products with antifouling activity as antifouling agents is an important research direction. In addition, some fouling release coatings without antifoulants, biomimetic coatings, photocatalytic coatings and other novel antifouling coatings have also developed rapidly. On the basis of revealing the mechanism of marine biofouling, this paper reviews the latest research strategies to develop environmentally friendly marine antifouling coatings. The composition, antifouling characteristics, antifouling mechanism and effects of various coatings were analyzed emphatically. Finally, the development prospects and future development directions of marine antifouling coatings are forecasted.

**Keywords:** antifouling mechanism; antifouling coating; antifoulant; environmentally friendly; polymer

#### **1. Introduction**

Marine fouling organisms refer to the general term for all kinds of marine organisms attached to the surface of marine facilities and causing damage to human marine economy [1]. Marine fouling organisms not only endanger the development of the marine economy, but also hinder human exploitation of the ocean. At present, more than 4000 marine fouling organisms have been discovered [2], such as microorganisms mainly including bacteria, diatoms and Ulva spores, etc.; large fouling organisms mainly include barnacles, bryozoans, mussels and algae.

Various fouling organisms easily attach to ships, oil platforms and breeding facilities in the ocean [3]. Barnacles and other invertebrates are firmly attached to the hull surface by secreting a biological adhesive. As the barnacle grows, its edges will destroy the anticorrosion layer on the hull surface, thereby accelerating the hull corrosion [4]. As shown in Figure 1 [5], when the hull surface is attached, fouling organisms will corrode the hull surface and increase the roughness of the hull surface, thereby reducing the speed of the ship [6–8], increasing fuel consumption and greenhouse gas emissions [9,10]. In recent years, with the rapid development of offshore industries such as submarine oil and marine power generation, the damage of marine fouling organisms to man-made facilities has become more severe. For example, when an offshore oil platform is attached, it will increase the weight of the facility and weaken its ability to resist tsunami and storm risks. In addition, fouling organisms can block drainage pipes, jeopardizing the safety and service life of marine facilities. Some marine fouling organisms enter the non-native land by attaching to the bottom of the ship, thereby causing biological invasion and causing great harm to the global marine ecology [11]. Every year, a lot of money is invested in ship surface cleaning and maintenance of marine facilities. With the continuous development of the human marine economy, the economic losses caused by marine fouling organisms are getting larger and larger, thus, effective and economic prevention methods are getting more and more attention.

**Figure 1.** Hull attached to fouling organisms [5].

The key to managing biological attachment is to block biological attachment from the source. Applying antifouling coating is currently the easiest and most widely used antifouling method. Traditional antifouling coating often use poisonous drugs to poison fouling organisms, the main components of which are organic tin [12] and cuprous oxide. However, the accumulation of metal elements in metal compounds in fish and shellfish will cause biological variation and death and pollute seawater, and thus harm the marine ecological environment. Taking into account the greater harm of organic tin to the marine environment [13], organic tin coatings have been banned worldwide. Although cuprous oxide is less toxic, as it continues to accumulate, its impact on marine ecology will become greater and greater. Therefore, the development of efficient and environmentally friendly marine antifouling coatings has become a research hotspot [14].

There are two main research directions for new antifouling coatings. One direction is to develop non-toxic and environmentally friendly antifoulants that replace metal compounds, that is, to find antifouling active substances from the ocean. An antifoulant developed by marine natural products can effectively reduce the harm to the environment, which is also one of the hot issues in current research [2,15]. In addition, some terrestrial biological products have also attracted attention. The purification or modification of natural products through chemical methods can obtain antifoulants with extremely strong antifouling capabilities. The second direction focuses on fouling release coatings, which will modify the surface characteristics of the material, making it difficult for fouling organisms to adhere or make them easier to remove after attachment. Due to the complex diversity of the marine environment, it is becoming more and more difficult for a single mechanism of antifouling coating to meet the requirements of use, and some composite antifouling coatings that combine multiple antifouling principles have also appeared. With the development and application of nanotechnology and polymer materials, the antifouling ability of antifouling coatings has been further improved [16].

This article first reveals the fouling mechanism of marine fouling organisms. On this basis, it mainly summarizes the development status of environmentally friendly antifouling coatings in three aspects from antifoulant type antifouling coatings, fouling release antifouling coatings and other important antifouling coatings. Our analysis reveals the antifouling mechanism of various antifouling coatings, and reviews the advantages and disadvantages of the new coating. Finally, the development trend of marine antifouling coatings is forecasted.

#### **2. Formation Mechanism of Fouling**

In order to better solve the problem of marine biological fouling, it is necessary to understand how the attachment of marine fouling organisms to the surface of objects occurs. Marine biological fouling is a process where biological communities accumulate on the surface of materials. The process is complex. As shown in Figure 2, biological fouling can be divided into three stages: conditioned film, micro-fouling and macro-fouling.


However, this model cannot apply to all marine fouling organisms. The actual fouling of the ocean is sometimes not in this order. For example, the cyprids of barnacle *Amphibalanus amphitrite* can adhere to the surface of materials without biofilm.

**Figure 2.** Hull surface fouling process and main fouling organisms [5].

#### **3. Marine Antifouling Coating**

Thus far, the application of antifouling coating is the most effective means of antifouling. Self-polishing copolymer-based coatings containing TBT had high efficiency in preventing the settlement and growth of marine fouling organisms, thus they were once used widely. However, they were globally banned in 2008 due to their persistent toxicity to non-target organisms. It is time to develop environmentally friendly systems to prevent marine biofouling. Table 1 summarizes the current popular research directions of environmentally friendly marine antifouling coating.


**Table 1.** Main research directions of marine antifouling coatings.

#### *3.1. Natural Product Antifoulant*

Cuprous oxide antifouling coatings have been used widely in today's marine ship. However, it also has potential environmental risks. Because the copper element in the cuprous oxide coating is enriched in the ocean, it causes a lot of death of seaweed and destroys the ecological balance, and will eventually be restricted or banned. In order to prevent the antifouling coating from harming the ecological environment, the researchers tried to find useful substances from nature to replace copper-based coatings. At present, the research of new antifoulants mainly focuses on the extraction and optimization of natural products. Many natural substances extracted from animals and plants have a good antifouling effect [21]. Compounds such as terpenoids, steroids, carotenoids, phenolics, furanones, alkaloids, peptides and lactones extracted from the ocean all have antifouling activity [22,23]. Irritation substances extracted from terrestrial plants such as oleander and pepper are also important sources of antifoulants.

#### 3.1.1. Marine Products with Antifouling Activity

Some marine organisms produce certain metabolites that inhibit fouling biological activity, which helps to develop pollution-free antifoulants composed of natural products [24]. In the past few years, some marine biological products with anti-pollution function have been discovered [25,26]. For example, Zhang et al. [27] discovered subergorgic acid (SA) from *Subergorgia suberosa*, which proves that it is non-toxic and has a strong inhibitory effect on attachments. In addition, Zhang et al. [28] conducted an in-depth study on the structure-activity relationship of SA and found that both ketone carbonyl and the double bond are necessary elements with antifouling activity. The experimental scheme is shown in Figure 3. Firstly, the esterification product 1 was readily obtained from SA using CH3I and K2CO3; secondly, NaBH4 or LiAIH4 was used to reduce ketone carbonyl to hydroxyl group; thirdly, LiOH(3eq)/TFH/H2O was used, trying to remove the methyl to give 3; finally, a strong oxidant 30%H2O2/TFA was used to modify the double bond of SA into two hydroxyl groups to obtain product 4. Products 1 and 2 still have antifouling activity, while products 3 and 4 have no antifouling activity, thus, ketone carbonyl and double bond groups are necessary elements for antifouling effect. On this basis, benzyl esters and methylene chains with proven antifouling activity were added to SA to form new SA derivatives. According to the methods depicted in Figure 3b, with dry DMF as solvent and

K2CO3 as base, SA can smoothly react with various benzyl halides and quantitatively obtain the desired benzyl ester. As shown in Figure 3c, using SA and dibromoalkanes of various lengths as starting materials, compounds 15–20 were easily synthesized in good yields. The antifouling test results show that the antifouling effect of all benzyl esters of SA is basically the same or stronger than that of SA containing unsubstituted benzyl rings. The results also show that the antifouling effect of SA derivatives containing methylene chains is not as good as that of SA, and the influence of the length of methylene chains on the antifouling effect of SA derivatives depends on the functional group at the opposite position of the methylene chain. The impact of this type of antifouling agent on non-target marine life needs further testing. However, the antifouling active functional groups proved in its experiments have certain value for the research of antifouling compounds.

**Figure 3.** Research and preparation of subergorgic acid (SA) compounds. (**a**) Identification of bioactive functional groups of SA. (**b**) Synthesis of benzyl esters of SA. (**c**) Synthesis of SA derivatives containing methylene chain of various lengths [28].

In nature, marine organisms have evolved many biological strategies to interact with microorganisms to protect themselves from pathogens or from being parasitized. [29]. In particular, sponges and their associated microbiota produce compounds that interfere with quorum sensing (QS) mechanisms [30,31]. Quorum sensing is a synchronization mechanism within a bacterial population. The bacteria use QS to communicate, regulate their behavior and assess their population density. The use of marine biological secretions to interfere with bacterial QS can effectively inhibit the formation of bacterial biofilms, making it difficult for other organisms to attach [32]. Tintillier et al. [33] studied the

marine sponge *Pseudoceratina sp. 2081* and isolated four new tetrabromotyrosine derivatives exhibiting antifouling and quorum sensing inhibition (QSi) properties. Structures of the isolated bromotyrosine metabolites is showed in Figure 4. They tested the antifouling activity of 6 kinds of extracts. Among the six compounds tested, the most active were 1, 3 and 5. The mode of action of those compounds are not based on toxicity. They operate through a targeted mode of action on bacteria and microalgae, as they only affect adherence and not growth.

**Figure 4.** Structures of the isolated bromotyrosine metabolites [33].

Indole derivatives found in bryozoans and ascidians have been shown to have antifouling effects [34]. However, due to the uncontrollable release rate of indole derivatives in antifouling coatings, the effect of antifouling coatings with added indole derivatives is poor. Feng et al. [35] prepared acrylate resins suspending the indole derivative structure in their side chain by the free-radical polymerization. Its preparation method and antifouling principle are shown in Figure 5. The prepared indole derivatives have good antibacterial and algae inhibiting effects. The self-polishing rate of the acrylate resin polymerized with indole derivatives is reduced, thereby obtaining a long-term self-polishing effect. The dynamic simulation measurements and static immersion measurements of the coating were performed to study in the marine environment. The results show that acrylate resin containing indole derivatives has long-lasting efficacy in biological control and the acrylate resins suspending the indole derivative structure in their side chain can make fouling organisms difficult to grow. In addition, indole derivatives inhibit the marine algae growth by interfering with the equilibrium of calcium ions in algal cells and decreasing the abundance of cellular Ca2<sup>+</sup> [36]. Under the action of seawater, the acrylic resin continuously renews its surface through the hydrolysis of ester groups. This self-polishing property makes attached organisms easily fall off the surface, exposing new indole derivative structures. Therefore, the combination of these two functions of the resins can significantly improve their antifouling properties. Since indole derivatives are easily degraded by microorganisms, they will not cause a great impact on the ecological environment.

Most marine fouling organisms secrete bio-gum to enhance surface adhesion. Bio-glue is a protein, and many biological enzymes in nature can degrade proteins. The essence of biological enzyme is protein, which is easy to decompose, causes no pollution and will not damage the marine ecology, thus, the biological enzyme with antifouling function is an ideal antifoulant. Wang et al. [37] isolated a marine proteolytic bacterial strain of *Bacillus velezensis* from sea mud and found that the protease produced by it had obvious inhibitory effects on barnacles, diatoms and mussels. This method uses bacteria to obtain proteases, which has the possibility of mass production. However, because marine fouling organisms secrete complex binders, polysaccharides and even lipids, it is difficult to control marine fouling by a single protease. Therefore, in the application research of enzyme-based environmental protection antifouling materials, it is necessary to use a combination of several enzymes to control marine antifouling.

**Figure 5.** Preparation process and working mechanism of acrylate resin with indole derivatives [35].

Currently, these marine biological antifoulant usually cannot obtain a sufficient amount from the ocean, and are difficult to chemically synthesize at low cost, thus, to a certain extent, it is still difficult to commercially mass-produce. However, it is important to develop a more perfect antifoulant by studying its antifouling mechanism through chemical methods and exploring the functions of various groups.

#### 3.1.2. Terrestrial Products with Antifouling Activity

Terrestrial organisms are also a valuable source of natural antifoulants. Compared with marine life, many terrestrial plants are easy to produce commercially because of their wide distribution or large-scale cultivation [38]. Liu et al. [39] isolated four cardenolides from *Nerium oleander L*. and evaluated their antifouling activity and toxicity to non-target organisms. The result showed that all of the tested compounds showed a strong inhibitory activity against barnacle settlement, with very low or moderate toxicity to non-target organisms. Their antifouling effect far exceeds that of TBT, Irgarol and copper. However, before it is widely used, its degradability and toxicity to non-target organisms need further study.

Chalcone is a compound commonly found in natural products. Celery was once the main source of chalcone, and now it is mainly synthesized by chemical means. Almeida et al. [40] studied the synthesis methods of 16 kinds of chalcone derivatives and studied the antifouling properties and ecotoxicity of chalcone through biological experiments. The results show that chalcone can effectively prevent the settlement of mussel larvae and inhibit the accumulation of other fouling microorganisms. In addition, they proved that these compounds have low ecotoxicity and clarified their great potential in the field of marine antifouling. Sathicq et al. [41] evaluated the synthetic furan-based chalcone antifouling coatings. They used furan rings instead of B rings to make a variety of different furylchalcones antifouling coatings, and conducted field experiments. Experimental results show that: using furylchalcones as an antifouling coating is an effective means of preventing marine fouling organisms; the synthesis of furylchalcone is simple and efficient, which is beneficial to reduce economic costs; furylchalcone is degradable, having little impact on the marine environment. Despite this, the mechanism of action of chalcone derivatives as antifoulant is not yet clear, thus, it is necessary to thoroughly understand the toxicity of furanyl chalcone in non-target marine organisms before applying it to the marine environment.

Camphor extracted from the trunk of camphor tree is a terpene natural organic compound, which can also be synthesized in large quantities by chemical means [42]. It is often used to repel insects and mosquitoes in daily life. In theory, it is a potential compound with antifouling capabilities [43]. Borneol extracted from herbs such as lavender and chamomile is a crystalline cyclic alcohol. Borneol-based polymer, isobornyl methacrylate (IBOMA) has been proven to have a strong activity against bacterial infections and can be used to prepare antibacterial coatings [44]. Hu et al. [45] synthesized an environmentally friendly antifouling coating based on antibacterial polymer (IBOMA) and natural antifouling agent (camphor). The prepared coating can slowly degrade in sea water and release borneol. Borneol itself has antibacterial properties [46], and at the same time, it provides a self-renewing surface to prevent dirt from sticking. In addition, camphor was released, providing a special antibiosis and antifouling surface with sterilized polymers to play a synergistic antifouling effect. Due to the hydrolysis of acrylic silicone resin and the increase in carboxylate content after immersion, the hydrophobicity of the coating surface quickly becomes dehydrated. The experiment proves that the produced coating can be controlled and slowly released, and has a continuous effect. Marine testing has shown the great potential of the coating, but the coating still needs a longer time to evaluate.

Capsaicin extracted from chili is a spicy vanillin amide alkaloid with antibacterial effects and is also regarded as a natural non-toxic antifoulant [47]. Wang et al. [48] conducted extensive research on capsaicin and its derivatives and reported six capsaicin derivatives for the first time, proving that capsaicin and its derivatives have excellent antifouling properties. Liu et al. [49] used capsaicin as an antifoulant and used high-density polyethylene (HDPE) as a substrate to make an antifouling coating by flame spray. Capsaicin, HDPE and coating preparation methods are shown in Figure 6. Compared with the raw materials, the morphology and grain size of capsaicin in the coating have changed significantly, and the coating has significant antibacterial and antifouling capabilities. In addition, capsaicin can be extracted from capsicum relatively easily, thus, this antifouling technology has low cost and high feasibility, and is suitable for widespread promotion.

**Figure 6.** Capsaicin coating raw materials and preparation method. (**a**) FESEM image of the starting capsaicin powder; (**b**) FESEM image of the starting high-density polyethylene (HDPE) powder; (**c**) the fabrication route for the HDPE–capsaicin antifouling coatings [49].

Biological antifoulants come from a wide range of sources, and the effect is quite excellent. However, the method of extracting it from organisms is still difficult to use on a large scale. Even though some biologically active substances can be artificially synthesized, it is far less convenient than metal compounds. In addition, natural product antifouling agents may have ecotoxicity, that is, they might be toxic to non-target organisms. Therefore, for the research of natural antifouling agents, we should not only pay attention to their antifouling properties, but also explore their ecotoxicity.

#### *3.2. Fouling Release Coatings*

The fouling release coatings (FRC) itself does not contain antifoulants, mainly to use the dual characteristics of fouling organisms that are difficult to attach to the surface of low surface energy materials or easy to desorb on the surface to prevent or reduce the attachment of marine fouling organism. Compared with traditional self-polishing antifouling coatings (SPC), FRC uses its own physical properties to hinder the attachment of fouling organisms, and will not release antifouling agents into the ocean, thereby not causing pollution to the marine environment.

Biological attachment behavior has a direct relationship with the surface characteristics of materials. Different surfaces have different or even opposite effects on biological adhesion. Low surface energy coatings mainly inhibit the attachment of fouling organisms by physical means, that is, the use of the characteristics that fouling organisms are difficult to attach to the surface of low surface energy materials to prevent or reduce the attachment of marine fouling organisms. Research shows that when the surface energy is less than 25 mJ/m2, that is, the contact angle between the coating and the liquid is greater than 98◦, and it can effectively reduce the adhesion of fouling organism [50]. At present, this type of antifouling coating mainly includes two major categories of silicone and organic fluorine.

#### 3.2.1. Organic Fluorine

In fluorine-containing polymers, due to the strong electronegativity of fluorine atoms, low polarizability and high C-F bond energy (460 kJ/mol), these materials are given high chemical stability and oleophobic and hydrophobic properties [51]. Arukalam et al. [52] prepared a perfluorodecyltrichlorosilane-based poly (dimethylsiloxane)-ZnO (FDTS-based PDMS-ZnO) nanocomposite coatings with surface energies within 20–30 mN/m for possible antifouling and anti-corrosion applications. ZnO is that the coating has antibacterial ability and changes the roughness of the coating surface, thereby improving its anti-adhesion ability. FDTS can modify the surface energy of the coating to keep its surface energy within 20–30 mN/m. Electrochemical impedance spectroscopy (EIS) test shows that the coating has excellent corrosion resistance. These characteristics mean that this coating has a good application prospect in the field of marine antifouling. Xu et al. [53] prepared a polymer with pendant branched poly (ethylene glycol) (PEO) and poly (2,2,2-trifluoroethyl methacrylate) (PFMA) structural units. The structure is shown in Figure 7. The b-PFMA-PEO asymmetric molecular brush with its side chains densely distributed with the same repeat unit is used to prepare a fluorine-containing synergistic nonfouling/fouling-release surface. A spin-cast thin film of the b-PFMA-PEO asymmetric molecular brush exhibits a synergistic antifouling property, in which PEO side chains endow the surface with a nonfouling characteristic, whereas PFMA side chains display the fouling-release functionality because of their low surface energy. Compared with the bare surface, the protein adsorption of the surface of the coating containing asymmetric molecular brushes is reduced by 45–75%, and the cell adhesion is reduced by 70–90%, showing considerable antifouling performance.

**Figure 7.** The structure of synergistic antifouling surface [53].

3.2.2. Silicone

Due to the high price and difficulty in preparation, fluoropolymers currently have few commercial products, and silicone polymer antifouling coatings have become a research and development hotspot. The silicone polymer has good desorption ability, and can easily make marine fouling organisms fall off by brushing or flowing. For example, PDMS has a combination of low surface energy and low elastic modulus, and it has gradually become the base of most antifouling coatings. Selim et al. [54] prepared a super-hydrophobic PDMS-Ag@SiO2 core-shell nanocomposite antifouling coating using the modified Stöber methods. The formation process of the coating is shown in Figure 8. The Ag@SiO2 core-shell nanofiller was inserted into the surface of the PDMS material by the solution casting method, and a strong coating was formed according to the hydrazination curing mechanism. The water contact angle (WCA) of the coating was determined to be 156◦, and the surface free energy was 11.15 mJ/m2. Biological tests prove that the coating has significant inhibitory effects on different bacterial strains, yeasts and fungi.

**Figure 8.** Schematic diagram of Ag@SiO2 core-shell nanosphere structure [54].

Low surface energy coatings do not require the use of antifoulant and they generally have a smooth surface, which is of great significance for reducing the navigation resistance of the vessels and reducing fuel consumption. The obvious disadvantage of low surface energy coatings is the poor adhesion on the hull surface. In practical applications, intermediate coatings must be used to enhance the adhesion of the coatings, which complicates the coating application process and increases the cost of use. Compared with other types of coatings, low surface energy coatings are more easily destroyed. In addition, most low surface energy coatings have poor desorption ability to strong adherent organisms such as diatoms.

#### *3.3. Biomimetic Antifouling Coating*

Biomimetic antifouling coating, also known as microstructured surface antifouling coating, its mechanism of action is to destroy the physical attachment of marine organic matter by preparing a surface similar to the microstructure contained in the biological epidermis. For example, the skin or

surface of sharks, shells, etc. [55,56] all have different structures of microstructures. This microstructured surface plays an important role in preventing or inhibiting the pollution of barnacles, algae and bacteria. The current main methods of making microstructured surfaces include laser etching [57], photolithography [58], three-dimensional (3D) printing [59], etc.

The shark skin surface structure has been extensively studied due to its drag reduction [60,61] and antifouling properties, and its various microstructures have been proven [62,63]. For example, in order to further study the antifouling effect of bionic shark skin, Pu et al. [64] prepared the biomimetic shark skin using the polydimethylsiloxane (PDMS)-embedded elastomeric stamping (PEES) method. PDMS itself is a low surface energy antifouling material, which has a certain antifouling ability, but it is prone to bioaccumulation in static or low flow state. The surface microstructure of shark skin is constructed on PDMS material, and the coating still has a low surface energy and exhibits extremely strong hydrophobicity. Air is hidden in the surface microstructure to form a thin layer of air, which can effectively reduce the interaction between the surface and protein molecules. Therefore, it is difficult for the protein to adhere to the surface and inhibit the attachment of some fouling organisms. The structure of shark skin and biomimetic shark skin is shown in Figure 9. Marine field tests show that the hydrophobicity of the PDMS material itself and the surface microstructure can effectively prevent the attachment of marine fouling organisms.

**Figure 9.** SEM images of shark skin and biomimetic shark skin: (**a**) the riblet structures of shark surface; (**b**) the surface of biomimetic shark skin prepared [64].

In marine environment, *Laminaria japonica* still has excellent antifouling ability in a relatively static state compared to those parade creatures. Zhao et al. [65] reported the synergistic effect between surface topography and chemical modification, and used sodium alginate and guanidine-hexamethylenediamine-PEI to replicate the surface microstructure on the surface of PDMS replicas by a layer-by-layer assembly method. The production process is shown in Figure 10. The water contact angle of PDMS with a microstructured surface is reduced to 35.3◦, which is more hydrophilic. After the algae test and the bacteria test, the surface has extremely strong antifouling performance, and the antibacterial ability is up to 96.2 ± 1.3%.

Lotus leaf surface is one of the most famous superhydrophobic surfaces and displays excellent anti-biofouling performances [66,67]. Jiang et al. [68] investigate the super-repellency of the lotus leaf towards the bacterial medium, together with its mechanical bactericidal activity against the attached bacteria, and for the first time reveal the synergistic antibacterial effects of its bacterial repellency and physical rupture by nanotubes. They designed and developed a hierarchically structured superhydrophobic surface integrated with regularly spaced micro-pillar arrays and packed nanoneedles. This surface can remarkably prolong its efficiency under much harsh conditions with respect to the conventional superhydrophobic surfaces, without causing any potential antimicrobial resistance. Schematic illustration of fabrication procedure of lotus leaf-like structures is shown in Figure 11, and the microscale structure of the surface is showed in Figure 12. The surface static water contact angle exceeds 170◦, and the rolling angle is less than 1◦. The surface's anti-adhesion efficiency for *E. coli* can reach more than 99%. A small amount of adhered bacteria can be completely killed, effectively overcoming the disadvantages of bacteria adhesion accumulation on the surface of a single structure sterilization.

**Figure 10.** Schematic diagram of material preparation process [65].

**Figure 11.** Schematic illustration of fabrication procedure of lotus leaf-like structures [68].

**Figure 12.** Microscale structure of different scales on the patterned surface [68].

The biomimetic antifouling coating contains no antifoulant and will not release compounds into the sea water. It is an eco-friendly antifouling coating. However, the construction of surface microstructures is more complicated, it is difficult to repair after damage and it is less effective in real marine environments. With the rapid development of Micro-Electro-Mechanical System (MEMS) and laser repair technologies, bionic antifouling coatings are still a promising development direction.

#### *3.4. Photocatalytic Antifouling Coating*

The use of photocatalytic to enhance the antifouling ability of the hull has attracted the attention of many scholars [69]. When some nanocomposites are exposed to sunlight, their surfaces exhibit strong oxidizing and reducing properties. Considering both economic and ecological aspects, this method has no pollution and low cost. TiO2 nanocrystals are widely used materials that modify coating surfaces to provide considerable mechanical reinforcement and surface wettability. Selim et al. [70] conducted an in-depth study on photocatalytic antifouling coatings, and prepared a PDMS/TiO2 hybrid nanocomposite. Its preparation method and action mechanism are shown in Figure 13. When sunlight illuminates TiO2 nanoparticles, it will excite carriers such as electrons and holes. The photogenerated electrons are transferred from the valence band to the conduction band, while the holes remain in the valence band. The electrons in the conduction band can be reduced to superoxide anion radicals, and the holes with strong oxidizing ability can oxidize water to produce hydroxyl groups. Hydroxyl groups still have strong oxidizing properties, which can attack the unsaturated bonds of organic substances or extract H atoms of organic substances to generate new free radicals. The chain reaction is excited and the bacteria are decomposed. In addition, the modified material exhibits strong hydrophilicity, which will produce a thin water film on the surface to block the adhesion of fouling organisms.

**Figure 13.** Preparation and catalytic principle of silicone/TiO2 nanocomposite [70].

However, because the band gap of TiO2 nanoparticles is 4% of sunlight, they can only absorb ultraviolet rays. Therefore, it is necessary to further improve the visible light absorption effect of the material. Zhang et al. [71] synthesized a new visible light-sensitive InVO4/AgVO3 photocatalyst with a PN node structure through ion exchange and in-situ growth process. Its mechanism of action is shown in Figure 14. When the bacteria come into contact with the catalyst, the H<sup>+</sup> and O2- ions catalyzed by the surface of InVO4/AgVO3 will destroy these cell walls and cell membranes, resulting in severe cell rupture, cell effluent and DNA molecule destruction. About 99.9999% of *Pseudomonas aeruginosa* (*P. aeruginosa*), *Escherichia coli* (*E. coli*) and *Staphylococcus aureus* (*S. aureus*) were killed over 0.5 InVO4/AgVO3 at 30 min.

**Figure 14.** Mechanism of InVO4/AgVO3 under light [71].

#### *3.5. Nano-Composite Antifouling Coating*

Due to the diversity of the working environment and the increased use requirements, it is difficult to achieve a better antifouling effect with a single performance antifouling coating. Composite antifouling coatings have become the mainstream of current antifouling coatings research [72]. Tian et al. [73] prepared a series of composite antifouling coatings consisting of silicone elastomer and nanocomposite hydrogel. They use AgNPs as cross-linking agents to improve the compatibility of hydrogels and PDMS, combine different antifouling mechanisms together, and improve the antifouling performance of hybrid coatings. The preparation principle of the coating is shown in Figure 15. Silicone and hydrogel work together to improve the desorption ability of the coating. Nanoparticles AgNPs are not only used to enhance the compatibility of silicone and hydrogel, but also make the coating have a bactericidal effect. After soaking in seawater for two years, the surface of the coating has no barnacles nor other scaling phenomena, further verifying its antifouling ability.

**Figure 15.** Schematic illustration of the formation of the pure silicone film and the hybrid coatings [73].

Selim et al. [74] prepared a superhydrophobic coating of silicone/β-MnO2 nanorod composite for marine antifouling. They studied how the self-cleaning and antifouling features were affected by controlling the β-MnO2 nanorod preparation and distribution in the silicone matrix. The preparation method of β-MnO2 and the construction method of the coating are shown in Figure 16. PDMS, as the substrate of the coating, is itself a low surface energy coating. By using β-MnO2 nanorods to construct a rough structure on the surface of PDMS, the coating is made superhydrophobic. The composite coating prevents fouling organisms from adhering through super-hydrophobicity and low surface energy. In addition, β-MnO2 nanomaterials also enhance the stability of the coating to temperature and PH.

**Figure 16.** The preparation method of β-MnO2 and the construction method of the coating [74].

#### *3.6. Other Antifouling Coatings*

In addition to the several coatings introduced above, some novel research results have emerged in recent years. These novel coatings have novel ideas and are expected to become new hot spots in antifouling coatings.

#### 3.6.1. Microcapsules Coating

Li et al. [75] prepared a form of environment-friendly microcapsules through mini-emulsion polymerization as show in Figure 17. They used zinc acrylate resin to wrap the synthetic microcapsules into a coating, and investigated the slow release efficiency and antifouling effect of the coating. The microcapsules had a poly(urea-formaldehyde) (PUF) shell and a mixed core of silicone oil and capsaicin. Dimethyl silicone oil has low surface tension, good water repellency, good lubricity and has the characteristics of pollution resistance, antifouling, anti-adhesion and so on. Capsaicin extracted from natural peppers does not destroy the biological chain, and can be used as a marine antifouling agent to kill plant spores and animal larvae attached to the outer surface of ships. It has broad application prospects in the field of antifouling materials. PUF microcapsules can combine multiple ingredients to delay the release rate of silicone oil and capsaicin, and extend the service life of the coating. Since the surface of the microcapsule antifouling paint has a micro-nano raised structure and the internal material can be slowly released, it has good hydrophobicity and antifouling properties. The release rate of the antifouling agents from the coating is determined by their slow release from the microcapsules to the surrounding coating matrix, and the encapsulated biocide is protected from degradation. This research has certain guiding significance for the development of long-term antifouling coatings.

**Figure 17.** Mechanism of urea-formaldehyde microcapsule formation [75].

#### 3.6.2. Dynamic Surface Antifouling

Xie et al. [76] first proposed the concept of Dynamic Surface Antifouling (DSA). The dynamic surface refers to a changing surface that continuously renews itself in seawater and thus decreases the adhesion of biofouling. Based on this strategy, they developed a series of biodegradable and reproducible polymer dynamic surface coatings, which have excellent antifouling properties and mechanical properties. The dynamic surface can shorten the contact time between the organisms and the surface, and make the interaction between them close to non-adhesion, thereby reducing the interaction force between the surface and the fouling organisms, making the fouling organisms difficult to attach. For degradable polymers, the surface renewal is a spontaneous process where water flow is not necessary. They developed degradable self-polishing copolymers (DSPC), made of poly (ester-co-silyl methacrylate) [77]. They inserted ester bonds into the backbone of a silyl acrylate copolymer for the first time by copolymerizing 2-methylene-1, 3-dioxepane (MDO), tributylsilyl methacrylate (TBSM) and methyl methacrylate (MMA) via radical ring opening polymerization (RROP). The degradable main chains can significantly improve the erosion of the TBSM based copolymer and avoid swelling. Consequently, the coating has excellent antifouling properties during static immersion in seawater. In addition, this type of coating can be used as a carrier and release control system for antifouling agents.

#### 3.6.3. Oil-Infused Polymers

Oil-infused 'slippery' polymer surfaces and engineered surface textures have been separately shown to reduce settlement or adhesion strength of marine biofouling organisms. Kommeren et al. [78] created liquid-infused surfaces embossed with surface structures. They used a photo-embossing process to create perfluorinated oil-infused fluorinated meth (acrylate) coatings with tuneable surface topography. Figure 18 showed the changed in coating surface morphology with the addition of oil-infused. Oil-infused polymers are made by exposing bulk polymeric materials such as fluoropolymer or PDMS to an excess of a chemically-matched oil, such as silicone or perfluorinated oils. The polymers absorb the oil, leaving a thin liquid layer on the material surface and a reservoir of oil in the bulk polymer. This allows oil to diffuse to the interface and replenish the surface liquid layer as it becomes depleted. The antifouling coating of this strategy can effectively resist bacterial adhesion under static and flowing conditions. The performance of the coating can be adjusted by controlling the surface fluctuations, thus, the coating can adapt to a variety of use environments.

**Figure 18.** Microstructure of the coating surface. (**a**) A three-dimensional (3D) height profile of the photo-embossed features in a polymerized coating without the addition of perfluorinated oil. (**b**) Measured feature height. (**c**) Micrograph of coating surface. [78].

#### 3.6.4. Sol-Gel Coating

Sol-gel derived functional coatings are commercially available for many practical applications and are emerging as suitable non-toxic alternatives to biocidal antifouling coatings. The sol-gel process provides the capacity to include inorganic and organic components at the nanometric scale. The surface energy of sol-gel is low and through modification can change the hydrophobicity and hydrophilicity of its surface. Richards et al. [79] studied several novel transparent sol-gel materials and determined their effectiveness as antifouling coatings for potential application to marine deployed sensors, camera lenses, solar panels or other related technologies. They developed a range of sol-gel coating with increasing water surface wettability and roughness. The surface energy of antifouling coatings was altered by modification of surface chemistry, while surface topology was roughened by the incorporation of amorphous fumed silica within the sol-gel. A future application of this work would be to incorporate sol gel coatings on the optical windows of sensors to reduce early stage fouling.

#### 3.6.5. Coating Based on a Synergistic Strategy

Xie et al. [80] designed a new type of antifouling costing with a longer period of validity based on a synergistic antifouling strategy. As shown in Figure 19, the antifouling paint is made of polyacrylates, tert-butyldimethylsilyl methacrylate (TBSM), eugenol methacrylate (EM) and poly(vinylpyrrolidone) (PVP). EM can add eugenol groups to the coating covalently, so that the coating has the function of contact inhibition. PVP can enhance the hydrophilicity of the coating to block biological adhesion. Eugenol can be gradually released into the sea water to expel marine life. Methyl methacrylate (MMA) ethyl acrylate (EA) and *n*-butyl acrylate (BA) are used in the coating to enhance the mechanical properties and adhesion of the coating, thereby improving the durability of the coating. TBSM can make the coating have a stable hydrolysis rate, so that eugenol can be released stably. Antifouling tests show that the coating can effectively prevent protein adsorption, bacterial adhesion and diatom adhesion. The actual marine environment test shows that the validity period of the coating is at least 8 months. This research has created conditions for the development of environmentally friendly, efficient and long-lasting antifouling paints.

**Figure 19.** Preparation of synergistic antifouling coating [80]. (**a**) Schematic diagram of coating preparation. (**b**) Synthesis of key copolymer PAPS-EM.

#### **4. Concluding Remarks**

Every year, marine fouling organisms cause huge economic losses, and antifouling coating is currently the most direct and effective means of prevention. It is important to develop stable, durable and environmentally friendly antifouling coatings. The current antifouling coatings are difficult to cover in terms of stability, toxicity and cost of use, which seriously hinders its popularization and use. In the future, the research and use of marine antifouling coating will surely develop in the direction of being highly efficient, non-toxic, non-polluting and degradable.

Antifouling coating containing antifoulant mostly outperform coatings without antifoulant. It is imperative to develop new and highly effective antifoulant to replace cuprous oxide, which is currently widely used. Obtaining natural antifouling substances from nature, especially extracting antifouling active substances from marine organisms, is the most ideal source of antifoulants. Through in-depth research on the antifouling mechanism of natural antifouling active substances, the development of highly effective and non-polluting antifouling agents is an important direction of the development of antifoulants. In addition, releasable coatings are environmentally friendly and have potential development prospects. The fouling release coating is mostly suitable for ships with high speed and fast travel, and it is less effective for ships that are moored for a long time or traveling at low speed.

Coatings that rely on a single antifouling mechanism cannot achieve the desired antifouling effect, and future trend will need to integrate multiple methods. Composite coatings that integrate multiple antifouling mechanisms will surely become a research hotspot, especially some nano-composite and modified coatings. With the use of nanotechnology, marine antifouling technology has been greatly improved. When the material reaches the nanometer level, its performance will change qualitatively, and the surface performance of the coating or the release efficiency of the antifoulant will be significantly improved. In addition, combining antifoulant with fouling release coatings is also an important development direction.

**Author Contributions:** Conceptualization and design, Y.G., L.Y. and J.M.; Writing—manuscript preparation, Y.G., L.Y. and D.W.; Writing—review and editing, L.Y., M.X., P.Z. and Y.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Zhejiang Provincial Natural Science Foundation of China (LY19E050003) and by the National Natural Science Foundation of China (51779226).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Marine Drugs* Editorial Office E-mail: marinedrugs@mdpi.com www.mdpi.com/journal/marinedrugs

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com

ISBN 978-3-0365-2146-6