Bioactive Marine Xanthones: A Review

The marine environment is an important source of specialized metabolites with valuable biological activities. Xanthones are a relevant chemical class of specialized metabolites found in this environment due to their structural variety and their biological activities. In this work, a comprehensive literature review of marine xanthones reported up to now was performed. A large number of bioactive xanthone derivatives (169) were identified, and their structures, biological activities, and natural sources were described. To characterize the chemical space occupied by marine-derived xanthones, molecular descriptors were calculated. For the analysis of the molecular descriptors, the xanthone derivatives were grouped into five structural categories (simple, prenylated, O-heterocyclic, complex, and hydroxanthones) and six biological activities (antitumor, antibacterial, antidiabetic, antifungal, antiviral, and miscellaneous). Moreover, the natural product-likeness and the drug-likeness of marine xanthones were also assessed. Marine xanthone derivatives are rewarding bioactive compounds and constitute a promising starting point for the design of other novel bioactive molecules.


Introduction
The marine environment occupies more than half of the Earth's surface and harbors the largest pool of biodiversity. Among others, marine microorganisms produce specialized metabolites used mostly in interspecies competition and defense from predators. The harsh conditions found in the sea spurs the development of specific biosynthetic pathways that produce metabolites bearing novel scaffolds, quite different from those found in terrestrial sources [1]. Specialized metabolites were optimized by evolution to establish flawless interactions with biological targets [2]. The sum of these factors makes the marine environment a prolific source of structurally diverse bioactive molecules that have a pharmacological interest [3].
Among the most relevant chemical classes of specialized metabolites isolated from the marine biodiversity, xanthones are a class of oxygen-heterocycles containing a heterocycle containing a dibenzo γ-pyrone moiety [4][5][6]. Depending on the nature and position of substituents, xanthone derivatives show a wide variety of biological activities making the xanthone scaffold a "privileged structure" in Medicinal Chemistry with a good potential for the discovery of novel hits, leads, and drugs [7].
As specialized metabolites, xanthones are often used as chemical defense agents. The most prevalent described activities were antitumor (43.5%) and antimicrobial (antibacterial (31.7%), antifungal (12.4%), and antiviral (10.6%) which provides some sort of protection to other competitive or predator marine organisms (Figure 2c). Interestingly, 22% of the identified marine xanthones presented more than one biological activity (sum of all activities > 100% in Figure 2c).  The bioactive marine xanthones were mostly isolated from marine fungi, namely from fungi belonging to the Aspergillus genus (41%, Figure 2b). Only a few examples were isolated from bacteria, and among them, the Streptomyces genus was the source that provided more bioactive xanthones (4.5%). Half of the reported bioactive marine xanthones (53%) were isolated from endophytic microorganisms associated with macroorganisms, like mangroves (39 marine xanthones), sponges (38 marine xanthones), algae (16 marine xanthones), corals (12 marine xanthones), jellyfish (1 marine xanthone), and seaweed (1 marine xanthone).
As specialized metabolites, xanthones are often used as chemical defense agents. The most prevalent described activities were antitumor (43.5%) and antimicrobial (antibacterial (31.7%), antifungal (12.4%), and antiviral (10.6%) which provides some sort of protection to other competitive or predator marine organisms (Figure 2c). Interestingly, 22% of the identified marine xanthones presented more than one biological activity (sum of all activities >100% in Figure 2c).

Chemical Space of Bioactive Marine Xanthones
Bioactive marine xanthones are produced by and act in living organisms, and to fulfill their specific biological task, they are structurally optimized by nature. Therefore, defining their chemical space is important for designing new molecules with desirable properties and pharmacological potential.
To describe the chemical space occupied by bioactive marine xanthones, several molecular descriptors, embracing different molecular, physico-chemical, and topological properties, were calculated using the RDkit (release 2021_03_5 Q1 2021) and SwissADME [17]. For each marine xanthone, the following molecular descriptors were calculated: molecular weight (MW), fraction of sp 3 carbons (Fsp 3 ), number of rotatable bonds (RB), lipophilicity (Log P), topological polar surface area (TPSA), and solubility (Log S) ( Table S1). The molecular descriptors were analyzed accordingly to the structural type of xanthones ( Figure 3).
Molecular size can be expressed in terms of MW, which is a predictor for pharmacokinetics behavior because bioavailability usually decreases as the molecular size increases [18]. In terms of size, the majority of the marine xanthones presented MWs within the range of 300 to 600 g·mol −1 . "Complex" xanthones presented the highest mean value (621.7

Chemical Space of Bioactive Marine Xanthones
Bioactive marine xanthones are produced by and act in living organisms, and to fulfill their specific biological task, they are structurally optimized by nature. Therefore, defining their chemical space is important for designing new molecules with desirable properties and pharmacological potential.
To describe the chemical space occupied by bioactive marine xanthones, several molecular descriptors, embracing different molecular, physico-chemical, and topological properties, were calculated using the RDkit (release 2021_03_5 Q1 2021) and SwissADME [17]. For each marine xanthone, the following molecular descriptors were calculated: molecular weight (MW), fraction of sp3 carbons (Fsp3), number of rotatable bonds (RB), lipophilicity (Log P), topological polar surface area (TPSA), and solubility (Log S) ( Table S1). The molecular descriptors were analyzed accordingly to the structural type of xanthones ( Figure 3).
Molecular size can be expressed in terms of MW, which is a predictor for pharmacokinetics behavior because bioavailability usually decreases as the molecular size increases [18]. In terms of size, the majority of the marine xanthones presented MWs within the range of 300 to 600 g·mol −1 . "Complex" xanthones presented the highest mean value (621.7 g·mol −1 ) and the highest value dispersion (standard deviation of 85) due to the inclu-sion of several dimeric and glycosylated structures. On the other side, "simple" xanthones and "hydroxanthones" have the lowest MW mean values (306.9 and 311.2 g·mol −1 , respectively) as these xanthones are composed by the xanthonic nucleus with simple substituents (such as hydroxyl or methyl or methoxy group).
Mar. Drugs 2022, 20, 58 11 of 36 droxanthones" were the most soluble group (mean value of −2.74), while "complex" xanthones were the most poorly soluble group (mean value of −5.84) (Figure 3f). "Simple" and "hydroxanthones" are above the log S value of −4, which is considered an acceptable value for a drug [23]. This trend is related to lipophilicity ( Figure 3j) and size ( Figure 3k). The poor solubility of complex xanthones is ascribed to their high molecular weight and high lipophilicity, while the smaller and/or more hydrophilic hydroxanthones have good water solubility (Figure 3f). Solubility of marine xanthones seems not to be affected by polarity as log S and TPSA were not correlated ( Figure 3l).  Molecular flexibility depends primarily on carbon saturation (Fsp 3 ) and on the number of rotatable bonds (RB). A high number of RB and/or Fsp3 means that the molecule has conformation flexibility which results in a less planar, less rigid, and more com-plex three-dimensional shape. Both RB and Fsp3 are important for determining oral bioavailability [10,19]. "Simple" xanthones have the lowest Fsp 3 values (mean value of 0.11) as saturated bonds are only present in substituent groups (Figure 3b). "Prenylated" and "O-heterocyclic" xanthones have equal mean Fsp 3 values (mean values of 0.31), but the latter showed higher value dispersion (interquartile range of 0.03 for "prenylated" and 0.17 for "O-heterocyclic" xanthones, Figure 3b), which is a consequence of higher 3D complexity rather than larger size (Figure 3g). "Complex" xanthones have a high value dispersion (standard deviation of 0.07, Figure 3b) due to their wide range of sizes (Figure 3g). Fsp 3 values greater than 0.42 are considered to be suitable values for a drug [19], and half of the "hydroxanthones" obey this criterion. In agreement with Fsp 3 analysis, "O-heterocyclic", "complex", and "prenylated" xanthones have a higher number of freely rotating bonds (median values of 3, 3, and 5 for "O-heterocyclic", "complex", and "prenylated", respectively). "Hydroxanthones" have the highest Fsp 3 values (median value of 0.38), but they have the lowest number of RB (median value of 1.50), meaning that saturated bonds belong to the cyclic system of the hydroxanthonic nucleus ( Figure 3c). Good oral absorption is associated with a number of RBs < 10 [10], and the vast majority of the marine xanthones fulfill this criterion.
Lipophilicity, assessed by log P, is a key parameter that affects both pharmacodynamics and pharmacokinetics [20]. "Hydroxanthone" (mean value of 0.68) and "simple" xanthones (mean value of 2.48) have the lowest log P values as they are frequently substituted with hydrophilic groups (hydroxyl and carboxylic). "O-heterocyclic" xanthones presented log P values similar (median value of 3.32) to the xanthone itself (calculated log P of 2.95), meaning that the additional O-heterocyclic moiety does not contribute significantly to lipophilicity. "Prenylated" xanthones have the highest log P value (mean value of 4.68), significantly higher than the "O-heterocyclic". "Complex" xanthones (median of 2.57) present the highest dispersion of log P values (standard deviation of 1.2), putting in evidence their structural diversity. The increase or decrease lipophilicity of marine xanthones is dependent on the substitution pattern, namely on the presence of hydrophilic or lipophilic substitutions. The size of the marine xanthone was not correlated with increasing lipophilicity as different sized molecules have quite similar log P values (Figure 3h), such as compound 10 (314 g·mol −1 , log P 2.14) and compound 157 (638 g·mol −1 , log P 2.16).
Molecular polarity, evaluated as the sum of surfaces of polar atoms in a molecule (TPSA), has been used to predict the permeability of drugs [21]. Different xanthone types have quite similar mean TPSA values (ranged from 74.8 Å 2 for "prenylated" up to 110.7 Å 2 for "hydroxanthones"), with the exception of complex xanthones that have significantly higher values (mean value of 203.9 Å 2 ) (Figure 3e). "Complex" xanthones are the only type that violates the preconized 140 Å 2 limit value [10]. In marine xanthones, the polarity is mostly related to MW, as TPSA values increase almost linearly with the MW (Figure 3i). In the case of marine xanthones, this is attributed to the increased number of polar atoms, such as oxygen or nitrogen, with increasing MW.
The water solubility, expressed as log S, is an important parameter for drug bioavailability. Compounds with poor water solubility have poor absorption and oral bioavailability, the evaluation of their bioactivity might be erratic, and the formulation development will be challenging [22]. The solubility trend observed with marine xanthones was: "hydroxanthones" > "simple" > "O-heterocyclic" > "prenylated" > "complex". "Hydroxanthones" were the most soluble group (mean value of −2.74), while "complex" xanthones were the most poorly soluble group (mean value of −5.84) (Figure 3f). "Simple" and "hydroxanthones" are above the log S value of −4, which is considered an acceptable value for a drug [23]. This trend is related to lipophilicity ( Figure 3j) and size (Figure 3k). The poor solubility of complex xanthones is ascribed to their high molecular weight and high lipophilicity, while the smaller and/or more hydrophilic hydroxanthones have good water solubility (Figure 3f). Solubility of marine xanthones seems not to be affected by polarity as log S and TPSA were not correlated (Figure 3l). NP-likeness allows measuring the similarity of a molecule to natural products [15]. The NP-likeness score quantifies this similarity; the higher the score, the higher the resemblance of that molecule to an NP [15]. NP-likeness scores of marine xanthones were calculated using the web service NaPles [24]. Figure 4a depicts the probability density function, based on kernel density estimation (KDE), of NP-likeness score for all NPs and synthetic molecules (SMs), as well as the score of marine xanthones. As expected, the NP-likeness score of marine xanthones falls within the range of NP. NP-likeness score of marine xanthones was analyzed considering the different types of xanthones (Figure 4b). Xanthone itself is similar to SMs (NP-likeness score of 0.19), while hexahydroxanthone presents a high score (NP-likeness score of 1.57). Within marine xanthones, "simple" xanthones have the lowest similarity with NP molecules (mean score of 1.09). The extension of the degree of substitution, from the simple hydroxyl groups in "simple" xanthones, up to an additional xanthonic nucleus present in "complex" xanthones, leads to the increase in the similarity to NP. This is in agreement with the fact that usually, NP are structurally more complex than synthetic molecules [25].  Considering the reported biological activities of marine xanthones, the relationship between these and the molecular descriptors was established. Figure 5a-d displays the relationship of the most relevant molecular descriptors (MW, log S, log P, and Fsp 3 ) with the most representative biological activities (antitumor, antibacterial, antifungal, antiviral, and antidiabetic). As the number of xanthones reported for each biological activity is different, the results should only be compared when the number of reported molecules is similar (antitumor vs. antibacterial and antifungal vs. antiviral vs. antidiabetic).
Antitumor marine xanthones have a bimodal distribution of MW with 2 subsets of Drug-likeness allows estimating the probability of a molecule to become a drug administered orally [17]. The classical approach to drug-likeness is normally based on a set of criteria to which the compounds under study should obey. This approach provides a binary "yes or no" assessment, depending on if the compound obeys or not the preconized limit values. Drug-likeness of marine xanthones were evaluated considering the classical Lipinski [9], Veber [10], Ghoose [11], Egan [12], and Gleeson rules [13]. In this study, a compliance value, defined as 0 when a compound does not obey any of the preconized criteria of that rule and 1 when a compound fulfills all criteria, was calculated for each marine xanthones (Table S2). Figure 4c displays the obtained mean compliance values of each type of marine xanthones for each rule. A lighter color in the heatmap plotted in Figure 4c means higher compliance, while a darker color means less compliance. "Hydroxanthone", "O-heterocyclic", "simple", and "prenylated" xanthones meet most of the criteria defined by classical rules, except for the Gleeson rules. On the contrary, "complex" xanthones violate at least one criterion in all the considered rules ( Figure 4c). Among the classical rules, the Gleeson rules [13] were the best to discriminate the different types of xanthones. "Simple" and "hydroxanthones" obey most of the criteria, "prenylated" obey just some, and "complex" xanthones do not obey the generality of Gleeson's proposed criteria.
Classical rules have many exceptions, and there are many examples, namely among NP or NP-inspired, of successful drugs that violate them [26]. Quantitative estimate of drug-likeness (QED) is an alternative way for assessing drug-likeness. QED index is generated considering eight properties, namely MW, log P, TPSA, RB, number of hydrogen donors and acceptors, number of aromatic rings, and number of alerts for undesirable substructures [14]. Compared with classical drug-likeness rules, the QED method is more flexible because it does not use cutoffs but a continuous score index of drug-likeness. When all properties are unfavorable, the QED index is 0, and when all properties are favorable, the score is 1 [14]. The obtained QED indexes for marine xanthones clearly differentiate the distinct types (Table S2, Figure 4c,d), enabling sorting the marine xanthones in the following ascending order of drug-likeness: "complex", "prenylated", "simple", "O-heterocyclic", and "hydroxanthones". The low drug-likeness of "complex" marine xanthones is related to their high MW ( Figure 3a) and low solubility (Figure 3f), while the drug-likeness of "hydroxanthones" is ascribed to their low lipophilicity ( Figure 3d) and good water solubility ( Figure 3f).
Considering the reported biological activities of marine xanthones, the relationship between these and the molecular descriptors was established. Figure 5a-d displays the relationship of the most relevant molecular descriptors (MW, log S, log P, and Fsp 3 ) with the most representative biological activities (antitumor, antibacterial, antifungal, antiviral, and antidiabetic). As the number of xanthones reported for each biological activity is different, the results should only be compared when the number of reported molecules is similar (antitumor vs. antibacterial and antifungal vs. antiviral vs. antidiabetic).
Antitumor marine xanthones have a bimodal distribution of MW with 2 subsets of different sized compounds (one mode of 314.3 g mol −1 and another mode of 636.6 g mol −1 ) ( Figure 5a). The presence of a bimodal distribution is also observed for the solubility of antitumor marine xanthones (modes of −3.45 and −5.66) (Figure 5b), which is not surprising considering that log S and MW are strictly correlated ( Figure 3k). However, log P have a unimodal distribution with a mode of 2.33 ( Figure 5c). Similarly, antibacterial marine xanthones also showed a bimodal distribution of MW (modes of 336.3 and 628.6 314.3 g mol −1 ) and log S values (modes of −4.23 and −5.84) (Figure 5a,b) and a unimodal distribution of log P values (mode of 2.37) (Figure 5c). The features of the large-sized subset of antitumor/antibacterial marine xanthone, i.e., "obese" molecules that apparently violate drug-likeness but with a suitable log P value, is a trait of NPs molecules. Despite being often cited as exceptions to classical drug-likeness rules, NP molecules largely comply in terms of log P [27]. This is attributed to the way in which natural evolution took place, producing bioactive compounds that retain low hydrophobicity, even for molecules with high MW [27]. The major difference between the physicochemical properties of antitumor and antibacterial xanthones was in terms of carbon saturation. Antibacterial marine xanthones have lower and more dispersed Fsp 3 values than antitumor xanthones (median value of 0.38 and 0.32 for antibacterial and antitumor, respectively), raising the hypothesis that more rigidity might be an important aspect for the antibacterial activity ( Figure 5d).   Figure 5d). Antiviral, antifungal, and antidiabetic marine fulfill the limited preconized by the drug-likeness filters independently of the descriptor.

Biological Activities of Marine Xanthones
A total of 74 marine xanthones described in literature were evaluated for antitumor activity, measuring their growth inhibitory activity in different tumor cell lines ( Table 1). The cervical carcinoma cell line (HeLA), human lung carcinoma (A549), human breast adenocarcinoma cell line (MCF-7), and human leukemia cell line (HL-60) were the most used tumor cell lines in the biological assays ( Figure 6). The number of xanthones with a half maximum inhibitory concentration (IC50) lower than 10 µ M varied depending on the tested cell lines. For instance, the number of xanthones screened against HL-60, MCF-7, and Huh7 cells was almost the same, but the number of most potent xanthones was higher identified against HL-60 cells. None of the marine xanthones assayed against C4-2B, 22RV1, and RWPE-1 presented an IC50 lower than 10 µ M, while for MGC-803, the most part presented an IC50 lower than 10 µ M.   (IC50 ≥ 300 µM) 2,6-Dihydroxy-3-methyl-9-oxoxanthene-8-carboxylic acid methyl ester (4) HEp-2 (IC50 = 8 µ g mL −1 ); HepG2 (IC50 = 9 µ g mL −1 ) Endophytic fungus (SK7RN3G1) isolated from mangrove collected in the South China Sea  A total of 54 marine xanthones described in literature were evaluated for antibacterial activity, measuring the growth inhibitory activity of different bacteria ( Table 2). Grampositive bacteria were more exploited (144 assessments) than Gram-negative (78 assessments) (Figure 7a). The number of xanthones with a minimum inhibitory concentration (MIC) lower than 4 µg mL −1 varied depending on the tested bacteria. Marine xanthones revealed a great selectivity for growth inhibition of Gram-positive bacteria as the percentage of marine xanthones with MIC lower than 4 is significantly higher in these types of bacteria (28% Gram-positive vs. 8% Gram-negative). This could be ascribed to a target potentially involving the peptidoglycan layer that is present in Gram-positive bacteria and absent in Gram-negative bacteria. The majority of marine xanthones were evaluated against S. aureus and E. coli (Figure 7b). Marine xanthones were particularly active against S. aureus, B. subtilis, E. faecalis, which are all Gram-positive bacteria.        xanthones revealed a great selectivity for growth inhibition of Gram-positive bacteria as the percentage of marine xanthones with MIC lower than 4 is significantly higher in these types of bacteria (28% Gram-positive vs. 8% Gram-negative). This could be ascribed to a target potentially involving the peptidoglycan layer that is present in Gram-positive bacteria and absent in Gram-negative bacteria. The majority of marine xanthones were evaluated against S. aureus and E. coli (Figure 7b). Marine xanthones were particularly active against S. aureus, B. subtilis, E. faecalis, which are all Gram-positive bacteria.    Paeciloxanthone (68) C. lunata (zone of inhibition 6 mm); C. albicans (zone of inhibition 10 mm)
A total of 9 marine xanthones were evaluated for anti-oxidant activity through the DPPH assay and ABTS or trolox equivalent antioxidant capacity (TEAC) assay (Table 6).
A total of 8 marine xanthones were evaluated for anti-inflammatory activity by measuring the inhibitory activity against cyclooxygenase (COX), by measuring the inhibition of inflammatory response induced by nitric oxide (NO) and NF-κB (factor nuclear kappa B), and by measuring the decrease in IL-6 cytokine production on LPS-stimulated macrophages ( Table 7).    The remaining biological activities were classified as miscellaneous (Table 8). Seven marine xanthones were evaluated for their immunosuppressive activity through the assessment of the inhibition of proliferation of mouse splenic lymphocytes stimulated with Con-A and LPS. One xanthone was evaluated for its anti-Alzheimer activity through the assessment of acetylcholinesterase inhibition. Three marine xanthones were evaluated for their antiprotozoal activity against Trypanosoma brucei, Trypanosoma cruzi, Leshnmania donovani, and Plamodium falciparum. Nine marine xanthones were evaluated for their aquatic pathogens biocide activity against Vibrio sp. Paeciloxanthone (68) Anti-Alzheimer: acetylcholinesterase inhibition (IC 50 = 2.25 µg mL −1 ) Paecilomyces sp. isolated from a mangrove collected in the Taiwan Strait [40] Chaetoxanthone A (78) Antiprotozoal: T. brucei rhodesiense (strain STIB 900) (IC 50 = 4.7 µg mL −1 ); T. cruzi (strain Tulahuen C4) (IC 50 ≥ 10 µg mL −1 ); L. donovani (strain MHOM-ET-67/L82) (IC 50 = 5.3 µg mL −1 ); P. falciparum (IC 50

Conclusions
As far as we know, 169 bioactive marine xanthone derivatives were reported in the literature up to 2021. They were isolated from microorganisms, mainly from Aspergillus sp., which normally live in an endophytic relationship with microorganisms (e.g., algae, sponge, mangrove, among others).
The chemical space occupied by bioactive marine xanthones was described through molecular descriptors. For each structural category, the distribution of the MW, Fsp3, number of RB, Log P, TPSA, and Log S values were described and analyzed. The descriptors were framed accordingly to the NP and drug-likeness concepts. Among the different structural categories of xanthones, "hydroxanthones" and "O-heterocyclic" xanthones are those that better resemble NPs and the ones that better fulfill the drug-likeness criteria. Therefore, hydroxanthones" and "O-heterocyclic" xanthones represent the most promising starting point for a hit-to-lead expansion.
Xanthones isolated from marine and terrestrial organisms share a similar biosynthetic pathway. However, marine-derived xanthones have not been as exploited as terrestrialderived xanthones in traditional drug discovery campaigns. The numerous and relevant bioactive xanthones isolated from the marine environment could inspire the development of new drugs. The data concerning this review allow us to go deeper in understanding molecular properties, at different levels, of such an important family of marine NP.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10 .3390/md20010058/s1, Table S1: Molecular descriptors of the bioactive marine xanthones, Table S2: NP-likeness and drug-likeness scores of the bioactive marine xanthones, Table S3: Chemical functional groups present in bioactive marine xanthones.