Bioactive Compounds Produced by Strains of Penicillium and Talaromyces of Marine Origin

In recent years, the search for novel natural compounds with bioactive properties has received a remarkable boost in view of their possible pharmaceutical exploitation. In this respect the sea is entitled to hold a prominent place, considering the potential of the manifold animals and plants interacting in this ecological context, which becomes even greater when their associated microbes are considered for bioprospecting. This is the case particularly of fungi, which have only recently started to be considered for their fundamental contribution to the biosynthetic potential of other more valued marine organisms. Also in this regard, strains of species which were previously considered typical terrestrial fungi, such as Penicillium and Talaromyces, disclose foreground relevance. This paper offers an overview of data published over the past 25 years concerning the production and biological activities of secondary metabolites of marine strains belonging to these genera, and their relevance as prospective drugs.


Introduction
For a long time fungi have been considered as a fundamentally terrestrial form of life. In the past few decades, this concept has started to be revised based on the emerging evidence that these microorganisms are also widespread in the marine habitat. New species recovered from marine substrates are reported repeatedly, which makes a reliable estimate of their actual number quite problematic [1]. Attention of researchers in the field often tends to be focused on the obligate marine species, defined for their ability to grow and sporulate exclusively in a marine habitat [2]. However, it is a matter of fact that many species found at sea are already known from terrestrial contexts, which makes their placement in a "facultative" category more appropriate. It is obvious that the mere isolation from a marine substrate does not imply a real adaptation of a fungus to develop in those particular conditions. Nevertheless, this aspect becomes secondary when considering discovery and exploitation of bioactive compounds, and in view of this objective the ecological versatility of the facultative marine fungi introduces them as being among the most valuable natural resources, deserving to be better characterized through more detailed genetic and biochemical analyses [3].
The issue of bioactive compound production is fundamental in understanding the complex ecological relationships established among and between sea-inhabiting organisms and microorganisms, and presents human nutritional implications due to the possibility that such fungal strains contaminate sea food, and their metabolites eventually act as mycotoxins [4,5]. However, the pharmaceutical industry can be regarded as the application field where products from marine fungi have the most substantial impact, since many of them have entered the clinical pipeline in view of being exploited as novel drugs. Quantitative considerations about fruitfulness in the discovery of new metabolites show that the number of compounds obtained from marine-derived fungi is increasing at a high rate. From a total of about 270 known before 2002, investigations in the field have added more than 800 such products up to 2010 [6], as a result of the availability of bioassay-guided fractionation systems, the accessibility to higher field NMR and mass spectrometers, and the development of the so-called hyphenated spectroscopy technologies (HPLC-MS, HPLC-NMR, etc.) [7]. The recent combining of natural product chemistry and metabolomic approaches in drug discovery can certainly contribute to the development of new leads from marine derived fungi [8].
Within the facultative marine fungi, species of Penicillium and Talaromyces are particularly known for their ability to produce important bioactive compounds. This paper offers an overview of the literature issued in the past 25 years concerning production and biological activities of secondary metabolites of marine strains belonging to the above genera, and their relevance as prospective drugs. Our review basically considers strains obtained from marine sources in a topographic sense, thus possibly including strains/species whose occurrence at sea is merely incidental. Conversely, we did not treat isolates from mangrove plants and their rhizosphere, whose connection with the sea is more remote, and probably deserve a dedicated review. As for the compounds, this overview does not consider primary metabolites mentioned in the cited references, including ergosterol and structurally related compounds [9,10]. Additional exclusions concern other common compounds which often represent intermediates in the synthesis of more complex secondary metabolites, such as orsellinic acid [11,12], and products obtained from mutant strains [13][14][15], or through co-cultivation of two or more strains [16].

Penicillium and Talaromyces: An Extraordinary Source of Bioactive Compounds
The Ascomycetous genus Talaromyces (Eurotiomycetes, Trichocomaceae) was initially designated to comprise the teleomorphs of a number of biverticillate Penicillium species. However, following the principle "one fungus-one name" recently affirmed in fungal taxonomy, by which a single holomorphic denomination is to be adopted for species presenting two alternating stages in their life cycle [17], the concept of Talaromyces has been recently extended to include all species in the Penicillium subgenus Biverticillium [18], while the name Penicillium is conserved sensu stricto for species belonging to the subgenera Aspergilloides, Furcatum, and Penicillium, for their associated Eupenicillium teleomorphs, and for species previously classified in a few related genera [19]. Information concerning production of secondary metabolites also supports the separation of the two genera in distinct monophyletic groups based on DNA sequencing [20], and the accumulation of novel data provides a remarkable contribution under the taxonomic viewpoint, particularly in view of a correct species ascription of the many strains which are provisionally reported as Penicillium/Talaromyces sp. [18,21].
Traditionally, species in the genus Penicillium and Talaromyces, which are fundamentally saprophytic and ubiquitary, have been regarded as a fruitful investigational ground for the finding of novel bioactive compounds, leading to the discovery of blockbuster drugs, such as penicillin [22] and the anticholesterolemic agent compactin [23], miscellaneous antitumor products [24], and mycotoxins contaminating food [25]. Most of these fundamental studies were carried out on strains from soil and food commodities. Thus, in a way it is not surprising that the number of bioactive compounds is continuously increasing when a so far inadequately explored context such as the sea has become the subject of systematic investigations. This inference is particularly valid for the species treated in this paper, considering that in a recent review Penicillium is reported as the second most common genus of marine fungi [26]. In our overview we recorded over 550 compounds, or compound families, from a total of about 150 strains belonging to 39 species of Penicillium and five species of Talaromyces (Table 1). Unclassified strains, referred to as Penicillium sp. or Talaromyces sp., represent a remarkable share (ca. 38%), which implies that the number of marine species within these genera is destined to increase when and if more work is carried out on some of these strains, eventually leading to their correct species ascription. With reference to this taxonomic aspect, the characterization of two novel species, Penicillium marinum [25] and Penicillium dravuni [27], deserves to be particularly mentioned. Cyclo(L-prolyl-L-tyrosyl), Cyclo(L-phenalanyl-L-prolyl), Cyclo(L-prolyl-L-valyl), Cis-bis(methylthio)silvatin, Bilain A-C, Pistillarin, Citromycin, 2,3-Dihydrocitromycin, Citromycetin, 2,3-Dihydrocitromycetin [    Penicillium sp.   Pinodiketopiperazine A, 6,7-Dihydroxy-3-methoxy-3-methyl phthalide, Cyclo(D-prolyl-D-valyl), Cyclo(trans-4-OH-D-prolyl-D-phenylalanyl), N-methylphenyldehydroalanyl-L-prolin-anhydrid, L-5-Oxoproline methyl ester, Rubralide C, Alternariol 2,4-dimethyl ether, Altenuene, 5 1 -Epialtenuene [132] Penicillium sp.    1 Current species name is specified in parentheses if different from the one reported in the original reference; 2 Information concerning the kind of organism is indicated as follows: B = brown alga; C = coral, soft coral; F = fish; G = green alga; M = mollusc; P = Angiosperm plant; R = red alga; S = sponge; T = tunicate; U = urchin; Z = bryozoan; 3 Products originally characterized from the corresponding strain are underlined; 4 Talaromycins A-C have been reported as new products. However, the same name was previously used for compounds with a different structure isolated from terrestrial Talaromyces strains [174,175].
Concerning sources, 49 strains were recovered from inanimate substrates, mainly sediment and water samples. As for living organisms, sponges appear to be the most widely reported hosts with 33 strains, confirming recent evidence of their significant interaction with fungi [176,177], while the other sources are represented by a disparate set of animals and plants including shellfish, gorgonians and corals, a few tunicate, urchin and fish species, brown, red and green algae, and a single Angiosperm plant (Zostera marina).
About half of the compounds listed in Table 1 (underlined names) were first characterized in strains from marine sources. This remark not only indicates, once again, that sea is a fruitful context for drug discovery, but also introduces a point of view that the ecological relationships established with marine organisms by species which are ordinarily reported from terrestrial environments may somehow address the biochemical pathways toward the synthesis of some peculiar compounds. In this sense, it must be emphasized that a number of unusual molecular structures have been first elucidated from this biological material (Figures 1-8).
The rest of the compounds itemized in Table 1 were first extracted and characterized for their bioactivity from terrestrial fungal strains, and a few of them are already known as drugs, or drug prospects. Particularly, mycophenolic acid is famous as the first known fungal antibiotic, discovered as a product of a strain of Penicillium brevicompactum even before the start of the 20th century [178], although its real structure was elucidated only after a few decades [179]. This compound displayed notable antibiotic, antiviral, and cytostatic properties, and has found consistent medical application as an immunosuppressive drug in the derivate form of mycophenolate mofetil [180,181]. First extracted from a strain of Penicillium griseofulvum [182], but later detected in many congeneric species, griseofulvin gained notoriety as an antimycotic drug, and more recently is being considered for its antitumor properties [183]. Again with reference to their antibiotic/cytotoxic properties, gliotoxin and the chaetoglobosins were first characterized from unrelated fungi, respectively Gliocladium fibriatum [184], and Chaetomium globosum [185]. However, both these compounds were later detected in a few Penicillium species, and considered for a series of interesting effects on human tumor cells [24]. Finally, 3-O-methylfunicone was first identified in connection with the antagonistic/mycoparasitic aptitude of the producing strains of Penicillium pinophilum (=Talaromyces pinophilus) [186,187], and later thoroughly characterized for its cytostatic properties on a number of human tumor cell lines, based on effects on cytoskeletal organization, cell cycle progression, the expression of pro-apoptotic genes, the inhibition of markers of tumor progression, and other mechanisms suppressing cell proliferation/migration [188][189][190][191][192][193][194][195][196]. Moreover, remarkable activity as a DNA polymerase inhibitor makes it one of the few known natural compounds displaying this particular effect [83]. Taken together, these valuable biological properties introduce 3-O-methylfunicone as a candidate molecule for more accurate clinical investigations in view of its development as an antitumor drug [197].
With a complex structure based on highly oxygenated, bicyclic and tricyclic frameworks, sorbicillinoids are a class of compounds which include over 50 members [198]. Their name derives from sorbicillin, which was first extracted from a terrestrial strain of Penicillium notatum (=Penicillium chrysogenum) [199]. However, a significant number of analog compounds showing peculiar structural models and consistent bioactive properties have been reported from marine fungi. Producing strains cited in this review were ascribed to a few unrelated species, such as Penicillium citrinum, Penicillium crustosum and Penicillium commune (Table 1). However, authorities in Penicillium taxonomy consider these strains to have been probably misidentified, by reason of strict evidence that these products are typical of P. chrysogenum and allied species [200].
Other products listed in Table 1 are best known for their noxious effects as mycotoxins contaminating foodstuffs. This is the case of cyclopiazonic acid, verrucosidin, fumitremorgin, and a few related tremorgenic toxins, secalonic acids, and particularly of citrinin and patulin [25]. Actually, the concern for dietary intake of mycotoxins produced by Penicillium strains has recently reached seafood, and specific investigations are being carried out in order to better assess the associated risk for consumers [54,128,201].

Bioactivities of Novel Compounds
By reason of the quite short time having elapsed after their discovery, most of the novel compounds obtained from marine strains of Penicillium/Talaromyces have been characterized for their biological properties and mechanisms of action only at a preliminary stage. In this regard, the largest category of bioactivity is undoubtedly represented by the cytotoxic/antiproliferative products (Table 2, Figure 1). In fact, assays on human or mammalian cell lines have become widespread following the recent trend to identify new natural antitumor compounds [202], and in view of pursuing this general aim there is a tendency to inappropriately consider these terms as synonyms [203]. Even if such a frequent semantic impropriety does not affect the significance of preliminary bioactivity screenings, the possible relevance of these molecules as antitumor prospects can be introduced only when a further characterization of their cytological effects is accomplished, which quite notably reduces the number of compounds deserving to be further examined in this review.

Bioactivities of Novel Compounds
By reason of the quite short time having elapsed after their discovery, most of the novel compounds obtained from marine strains of Penicillium/Talaromyces have been characterized for their biological properties and mechanisms of action only at a preliminary stage. In this regard, the largest category of bioactivity is undoubtedly represented by the cytotoxic/antiproliferative products (Table 2, Figure 1). In fact, assays on human or mammalian cell lines have become widespread following the recent trend to identify new natural antitumor compounds [202], and in view of pursuing this general aim there is a tendency to inappropriately consider these terms as synonyms [203]. Even if such a frequent semantic impropriety does not affect the significance of preliminary bioactivity screenings, the possible relevance of these molecules as antitumor prospects can be introduced only when a further characterization of their cytological effects is accomplished, which quite notably reduces the number of compounds deserving to be further examined in this review.

Bioactivities of Novel Compounds
By reason of the quite short time having elapsed after their discovery, most of the novel compounds obtained from marine strains of Penicillium/Talaromyces have been characterized for their biological properties and mechanisms of action only at a preliminary stage. In this regard, the largest category of bioactivity is undoubtedly represented by the cytotoxic/antiproliferative products (Table 2, Figure 1). In fact, assays on human or mammalian cell lines have become widespread following the recent trend to identify new natural antitumor compounds [202], and in view of pursuing this general aim there is a tendency to inappropriately consider these terms as synonyms [203]. Even if such a frequent semantic impropriety does not affect the significance of preliminary bioactivity screenings, the possible relevance of these molecules as antitumor prospects can be introduced only when a further characterization of their cytological effects is accomplished, which quite notably reduces the number of compounds deserving to be further examined in this review.
Another alkaloid with a unique spiro skeleton, penicitrinine A, was found to induce some typical modifications in melanoma cells undergoing apoptosis, such as shrinkage, fragmentation, and chromatin condensation. Assays based on annexin-V/PI double staining showed that apoptosis occurred at a higher rate than control cells treated with the chemotherapeutic drug 5-fluorouracile. Apoptosis followed the mitochondrial pathway, as indicated by down-regulation of the anti-apoptotic gene Bcl-2 and concomitant up-regulation of the pro-apoptotic gene Bax, and the ratio of Bcl-2/Bax expression, which decreased with increasing concentrations of the compound. Anti-metastatic dose-dependent effects were also observed as a result of suppression of invasiveness and inhibition of cell migration, which is an ill-fated tendency of melanoma cells. These latter effects are related to a down-regulation of matrix metalloproteinase (MMP-9) expression along with up-regulation of its inhibitor glycoprotein TIMP-1 [111].
Besides selectively suppressing cell growth and proliferation in five human cancer cell lines, pinophilins displayed a strong inhibitory activity on DNA polymerases of the A-, B-, and Y-families, particularly against DNA pol-α and -κ. The inhibitory effect was selective, since it was not observed on normal human cells (dermal fibroblasts and umbilical vein endothelial cells), possibly because their DNA replication rates are significantly slower than those of cancer cells [122]. Bioactivity as DNA polymerase inhibitors had been previously showed for two phenalenone compounds, the sculezonones A-B. Particularly, both compounds inhibited bovine DNA pol-α and -γ, and moderately affected the activity of DNA pol-ε. Moreover, DNA pol-β was inhibited by sculezonone A, and just weakly influenced by sculezonone B [205]. Another DNA polymerase inhibitor, the γ-lactam compound epolactaene [36], not only was effective on bovine DNA pol-α and rat DNA pol-β, but also disclosed inhibitory properties against human DNA topoisomerase II [206], which is a very important biomolecular mechanism considered in prospecting for antitumor drugs [207]. Consistent neuritogenic effects [36] make this compound even more valuable in view of possible medical applications.
Some degree of neural stimulation was also pointed out for the fellutamides. In fact these tripeptides were characterized as potent enhancers of the release of the nerve growth factor (NGF) from fibroblasts and glial-derived cells [208]. This effect results from the inhibition of proteasome catalytic activity, which leads to increased NGF gene transcription [209]. For different respects, neuroprotective properties have been reported for sorbicillactone A, which impaired the negative effects of important neurotransmitters such as L-glutamic acid and serotonin [62], and brevicompanines E and H (Figure 2), which have been characterized as neuroinflammation modulators [87]. More in detail, in BV2 mouse microglial cells brevicompanine E was found to inhibit production of tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), inducible nitric oxides (iNOS), and cyclooxygenase-2 (COX-2), and to reduce the DNA binding activity of the oncogenic nuclear factors AP-1 and NF-κB. Nuclear translocation of the latter was also inhibited, together with IκBα degradation, and Akt and c-Jun NH 2 -terminal kinase phosphorylation [210]. Similar anti-inflammatory effects were also evidenced in murine peritoneal macrophages for novel ester derivatives of hexylitaconic acid [115], and for penicillinolide A [138] (Figure 2). Weak NF-κB inhibitory properties were again reported from penilactone A [125], while sargassopenilline C has been found to inhibit the transcriptional activity of AP-1 [160] (Figure 2). Finally, and again in BV2 cells, 2E,4Z-tanzawaic acid D was found to inhibit the production of iNOS [158].
Penstyrylpyrone ( Figure 2) is another product reported for considerable anti-inflammatory activity deriving from inhibition of the expression of iNOS and COX-2, reduction of TNF-α and IL-1β production, suppression of phosphorylation and degradation of IκB-α, and of NF-κB nuclear translocation and DNA binding activity. These effects were found to be associated with the expression of heme oxygenase 1 (HO-1), an enzyme releasing anti-inflammatory degradation products during heme catabolism. Ultimately, these anti-inflammatory effects lead to a competitive inhibition of the protein tyrosine phosphatase 1B (PTP1B), which is known to play a major role in the negative regulation of insulin signalling. Therefore, this compound was introduced as a prospect therapeutic drug for the treatment of type II diabetes [140]. Inhibitory properties towards PTP1B were also disclosed for penostatins A-C [211], while verruculides A and B respectively displayed a strong and a moderate effect against this enzyme [173]. Finally, a moderate effect as an inhibitor of tyrosine kinases was reported for terrestrol G (Figure 2) [64].
Another target in the search for antitumor products is represented by the reactive oxygen and nitrogen species (ROS), whose excessive production results in oxidative stresses, DNA damage, and inflammation, as well as contributing to tumor initiation and promotion. Consequently, scavenging of the physiologically relevant ROS represents an effective strategy in preventing tumor initiation and promotion. Chromanone A (Figure 3) was characterized as a strong OH scavenger, which also dramatically inhibits the degree of DNA fragmentation. Moreover, it was able to act, in a dose-dependent manner, as a potent inhibitor of cytochrome P4501A, and as an inducer of GSH (cytosolic thiol) and GST enzymes, which both help in the destruction of peroxides, free radicals, and other xenobiotics [86].
Other enzyme modulatory activities are relevant in human medicine for the treatment of a number of complex diseases. This is the case for Alzheimer's disease, where compounds performing acetylcholinesterase inhibition can be considered in view of possible therapeutic use [212]. In this regard, moderate activity has been reported for products such as sorbiterrin A [65], and penicilliumine [149], while more potent effects have been evidenced for talaromycesone A and talaroxanthenone [161] (Figure 4). The latter compound also showed activity as an inhibitor of phosphodiesterase, which is a target in the treatment of inflammatory processes involved in pulmonary diseases [161]. Besides general activity as inhibitors of proteases, such as papain, ficin, and bromelain, the cathestatins, particularly cathestatin B (Figure 4), were introduced for possible useful effects in the treatment of osteoporosis deriving from the inhibition of bone collagen degradation, and the suppression of calcium release [35]. Finally, along the lines of their more famous analog mycophenolic acid, the penicacids ( Figure 4) were investigated for their immunosuppressive properties, and found to possess appreciable inhibitory effects towards inosine-monophosphate dehydrogenase [124].
Antibiotic properties have been assessed for a number of novel compounds ( Figure 5) against the bacterial and fungal species indicated in Table 3. Besides these general inhibitory effects, some peculiar mechanisms of action were evidenced for the dipeptide cis-cyclo(leucyl-tyrosyl) (Figure 6), which inhibits biofilm formation by Staphylococcus epidermidis without interfering with bacterial growth [144], and herqueidiketal, which was characterized for inhibitory properties against sortase A of Staphylococcus aureus. Since sortases are absent in mammals, this biochemical effect may be relevant for the development of novel antibiotics [130]. Moreover, the above-mentioned proteasome inhibitory effects by fellutamide B were again observed on Mycobacterium tubercolosis, introducing this peptide as a prospect drug to be more thoroughly investigated against such a deadly pathogen [213]. Terretrione D Antifungal C. albicans [167] Xestodecalactone B Antifungal C. albicans [47] m.r.: Methicillin resistant.
A few compounds displayed consistent effects against important viruses, such as the influenza virus A (H1N1) and HIV-1. Particularly, the cytopathic effects induced by the former in MDCK cells were found to be inhibited by sorbicathecols A-B (Figure 7) [153], while the latter was impaired by brevione F, which inhibited its replication in C8166 cells [92], and sorbicillactone A, which inhibited the expression of viral proteins and protected H9 cells (human T lymphocytes) against cytopathic effects [62].
Finally, some miscellaneous bioactive effects can be mentioned for a few compounds (Figure 8).
Widely considered as a model organism used to test the toxicity of chemicals, brine shrimp (Artemia salina) has been employed for demonstrating the toxic effects of products such as 6,7-dihydroxy-3-methoxy-3-methylphthalide [132], 13-O-prenyl-26-hydroxyverruculogen [148], adametizine A [155], the peniciadametizines [156], and the communesins [51]. The latter represent a numerically expanding series of cytochalasan alkaloids which have been also introduced to some extent for cytotoxic/antiproliferative properties (Table 2), and insecticidal effects resulting after oral administration to silkworms [215]. Insect neurotoxicity was also observed in assays carried out on larvae of the bluebottle fly (Calliphora vomitoria) [94]. Moreover, in a study employing a zebra-fish model, communesin I and two more novel compounds, fumiquinazoline Q and protuboxepin E, were reported for cardiotonic effects, as well as some extent of vasculogenetic properties assessed with reference to both number and length of vessels [168].

Conclusions
Data summarized in this review highlight the widespread occurrence at sea of Penicillium/Talaromyces strains, and their extraordinary potential as a source of novel bioactive compounds and drugs. As new data accumulate more and more, the awareness is increasing within the scientific community that these microorganisms represent a trove of unexplored biodiversity, and that more exhaustive investigations should be carried out. In this regard, a comprehensive work was recently published concerning diversity and antifungal properties of a group of 184 marine strains belonging to 36 different Penicillium species from Korea [216]. Also a comparison shows that as many as 18 of these species were not even mentioned in our review, which makes it very likely that our proposed record series will be considerably expanded if this biological material is further characterized in order to detect the biochemical determinants of the fungitoxic effects.
In the meantime, much work is to be done with reference to more complete characterization of the biological activities of the material accumulated so far, in view of the increasing number of products which can evolve to the drug level. A good example is represented by sorbicillactone A, whose notable antileukemic effects have stimulated studies for improving its laboratory yields in view of large scale production [217]. However, progress towards this ultimate step is largely dependent on the extent to which the pharmaceutical industry will prove to be prepared to grasp such a great opportunity. It is desirable that the recent policy adopted by most governments worldwide aimed at involving the manufacturing industry in funding for basic research, turns into a more decisive effort, with ensuing results, in this direction.