Trichoderma: A Treasure House of Structurally Diverse Secondary Metabolites With Medicinal Importance

Fungi play an irreplaceable role in drug discovery in the course of human history, as they possess unique abilities to synthesize diverse specialized metabolites with significant medicinal potential. Trichoderma are well-studied filamentous fungi generally observed in nature, which are widely marketed as biocontrol agents. The secondary metabolites produced by Trichoderma have gained extensive attention since they possess attractive chemical structures with remarkable biological activities. A large number of metabolites have been isolated from Trichoderma species in recent years. A previous review by Reino et al. summarized 186 compounds isolated from Trichoderma as well as their biological activities up to 2008. To update the relevant list of reviews of secondary metabolites produced from Trichoderma sp., we provide a comprehensive overview in regard to the newly described metabolites of Trichoderma from the beginning of 2009 to the end of 2020, with emphasis on their chemistry and various bioactivities. A total of 203 compounds with considerable bioactivities are included in this review, which is worth expecting for the discovery of new drug leads and agrochemicals in the foreseeable future. Moreover, new strategies for discovering secondary metabolites of Trichoderma in recent years are also discussed herein.


INTRODUCTION
Trichoderma is a fungal genus that was first described in 1794 (Persoon, 1794). This genus is well adapted to various ecological niches and is ubiquitous in most types of soils, roots, and foliar environments. Trichoderma species are beneficial for their commercial enzymes, plant growth-accelerating abilities, and biocontrol of plant diseases, indicating their promising industrial, agricultural, and medicinal potential (Cai et al., 2013;Bhardwaj and Kumar, 2017;McMullin et al., 2017). Globally, Trichoderma has proved to achieve great success as effective biological control drugs (Keswani et al., 2014). Many Trichoderma species, such as T. harzianum, T. hamatum, T. asperellum, T. atroviride, T. koningii, T. virens, and T. viride, are lucratively used as potent biocontrol agents worldwide (Bhardwaj and Kumar, 2017). These fungal species exhibit outstanding biocontrol capability against pathogenic microorganisms either through indirect (scrambling for nutrients, changing the ambient conditions, stimulating plant growth and defense responses) or direct (mycoparasitism) mechanisms (Bhardwaj and Kumar, 2017). Moreover, in addition to ecological effects, it is well known that Trichoderma can produce secondary metabolites that not only participate in signal transduction but also go through communications with various organisms (Keswani et al., 2014;Zeilinger et al., 2016). It is also believed that the successes of Trichoderma as biocontrol drugs are, at least partially, due to their capacity to secrete abundant secondary metabolites (Zeilinger et al., 2016).
Fungi play an irreplaceable role in the drug discovery in the course of human history, as they possess unique abilities to synthesize diverse secondary metabolites with significant medical potential (Li X.Q. et al., 2020). The discovery of penicillin from the filamentous fungal species Penicillium was a milestone in pharmaceutical research (Fleming, 1929). Since then, chemical studies regarding fungal secondary metabolites have become a research hotspot . A large number of fungal secondary metabolites have been discovered, many of which have potential as drug leads (Newman and Cragg, 2016). The fungal species belonging to Penicillium and Talaromyces are representative flora, with many secondary metabolites possessing intriguing chemical skeletons and bioactivities characterized from these species (Frisvad, 2014). For the genus Trichoderma, more than 1000 metabolites have been isolated from Trichoderma in recent years (Zeilinger et al., 2016). Accordingly, many reviews on various aspects of Trichoderma, not only for the chemical diversity of metabolites but also for the various bioactivities and their potential applications, have been published. Zeilinger et al. (2016) reviewed the selected Trichoderma-derived secondary metabolites and gave an all-round summary of genomic analysis and putative gene clusters involved in biosynthesis. Keswani et al. (2014) listed targeted metabolites of Trichoderma and pointed out the utilization potentiality in multifarious areas, especially in agriculture. As mentioned above, Trichoderma is a well-known biocontrol agent that is used globally. Since many Trichoderma species are some of the most prominent producers of anti-phytopathogenic secondary metabolites, Khan et al. (2020) exhibited 45 fungicidal secondary metabolites of Trichoderma sp. along with the structural overview, biosynthesis pathway, and action mechanism. Moreover, Reino et al. (2008) systematically summarized the metabolites of Trichoderma and their bioactivities. As of 2008, a total of 186 compounds and 269 references were cited, including a detailed study of the activities of biocontrol mechanisms (Reino et al., 2008). Herein, in order to update the relevant list of reviews of secondary metabolites of Trichoderma sp., we provide a comprehensive overview in regard to the newly described metabolites of Trichoderma from the beginning of 2009 to the end of 2020, with emphasis on their chemistry and various bioactivities. Moreover, new strategies for discovering secondary metabolites of Trichoderma in recent years have also been discussed.

LITERATURE SEARCH
To retrieve literature published up to 2020, an in-depth inspection was performed . The key words of "Trichoderma" and "secondary metabolites" were used to search for related literatures in Web of Science, with the times pan from 2009 to 2020 (Supplementary Figure 1). Additionally, other platforms such as Crossref, Google Scholar, Elsevier, and Springer Link were also searched at the same time. The retrieved articles were categorized according to natural product chemistry. In all, 63 records in the context of natural product research were retained and assessed to present this review. A total of 203 compounds were found from Trichoderma from 2009 to 2020. It should be pointed out that some omissions were inevitable during the literature search. However, with the greatest effort, we have demonstrated almost all of the relevant research herein.

Other Compounds
Two new sulfur compounds, designated thioporidiols A (198) and B (199), were produced by a culture broth of T. polypori FKI-7382 (Figure 10) . Both of them were determined to be C13 lipid structures with an N-acetylcysteine moiety. A chemical investigation of the endophytic fungus T. polyalthiae offered two diphenyl ethers, Violaceol I (200) and II (201). Notably, both of them were characterized from Trichoderma for the first time (Nuankeaw et al., 2020). Trichodenols A (202) and B (203), two new compounds with 4-(2-hydroxyethyl) phenol moieties, were isolated from an endophyte T. gamsii (Ding et al., 2015).

NEW STRATEGIES FOR DISCOVERING SECONDARY METABOLITES OF TRICHODERMA
Recent fungal genome sequencing indicated that the majority of biosynthetic gene clusters (BGCs) associated with secondary metabolites are cryptic (transcriptionally silent) or expressed at very low levels under general laboratory conditions (Ren et al., 2017). Therefore, despite a large number of secondary metabolites being characterized from Trichoderma, genome sequencing revealed that there were more BGCs than we discovered, especially in filamentous fungi. These findings suggested that those silent metabolic pathways urgently need to be stimulated, which may lead to the discovery of novel metabolites with attractive functions. To activate cryptic biosynthetic pathways, many innovative approaches, such as cultivation-based approaches, metabolomic profiling, and genome mining-based molecular approaches, have been developed in recent years. These new approaches were accomplished with various degrees of success. The following are typical examples of searching for secondary metabolites of Trichoderma induced by cultivation regulation, cocultivation, chemical epigenetic manipulation, and transcript regulation, as shown in Figure 11.
The marine-derived fungus Trichoderma sp. TPU199 was found to produce a series of diketopiperazines under different conditions (Figure 12) (Yamazaki et al., 2020a). Chemical investigations of this fungal strain under ordinary culture conditions led to the discovery of compounds 169-177. Then, this fungus produced the halogenated gliovirin-type ETPs 120 (Cl derivative of 117), 122 (Br derivative of 117), and 123 (I derivative of 177) when added with NaCl, NaBr, and NaI in culture medium, respectively. Moreover, TPU199 supplemented  with DMSO yielded 124, a new trithio derivative of 120. A continuous study indicated that, with the long time cultivation, an undescribed modified dipeptide 125 was obtained. Finally, two undescribed ETPs 126 and 127 were characterized under NaI-containing culture conditions. It is undoubtedly proven that changing the culture conditions can activate cryptic metabolic pathways.
Microorganism coculture based on microbial interspecies competition is an efficient path to stimulate cryptic BGCs. Cocultivation of Trichoderma sp. 307 and pathogenic bacterium A. johnsonii B2 yielded two undescribed sesquiterpenes (66 and 67) and de-O-methyllasiodiplodin (194 and 195)  . HPLC analysis indicated that they were derived from Trichoderma sp. rather than by the coculture. Chemical epigenetic manipulation has also proven to be an effective method to activate cryptic BGCs. Therefore, it was applied to the marine fungal strain T. harzianum XS-20090075 to mine its potential to synthesize secondary metabolites (Shi T. et al., 2020). A histone deacetylase inhibitor, sodium butyrate at 10 µM, significantly changed its metabolic profile and gave rise to three undescribed terpenoids, including a cleistanthane 105, a harziane diterpenoid 91, and a cyclonerane sesquiterpenoid 44. Interestingly, harziane diterpenoids were the dominant metabolites from this fungal strain under ordinary culture conditions, indicating the production of harzianes as the dominant metabolic pathway. In this study, the production of harzianes was hampered due to chemical epigenetic manipulation. In contrast, the biosynthetic pathways of cleistanthanes and cycloneranes were successfully activated, leading to the isolation of the new cleistanthane diterpenoid 105 and a series of cyclonerane sesquiterpenoids (44 and other known cycloneranes). This is the first report of cleistanthane diterpenoids isolated from Trichoderma species. This study provided solid example to show that it is efficient to activate the silent genes of Trichoderma species by chemical epigenetic manipulation.
The transcriptional control has also proven to be an effective approach. To activate the chemical potential of the endophytic fungus T. afroharzianum, a laeA-like gene overexpression transformant was built (Ding et al., 2020). Further chemical investigation of this transformant successfully yielded two new antifungal polyketides (189 and 190). This study indicated that transcriptional control could be a considerable strategy in activating more secondary metabolites and enhancing the silent potential metabolism of Trichoderma species.

BIOLOGICAL ACTIVITIES
The producing fungus, environmental source, and bioactivities of compounds 1-203 are listed in Table 1. As shown in Table 1, most compounds possess various moderate to potent biological activities. Among them, antimicrobial, antimicroalgal, and anticancer activities represent dominant bioactivities to assess the pharmacological potential of these natural products.
Detailed descriptions of these metabolites with promising biological activities are described as follows.
lagenarium with an MIC value of 16 µg/mL, which was stronger than that of the positive control carbendazim (MIC, 32 µg/mL) (Du et al., 2020). Furthermore, both of them showed potency against carbendazim-resistant B. cinerea. In contrast, compared to those of 22 and 23, trichothecene congener 24 only showed weak effects, indicating that the hydroxyl substituted in 23 may enhance its antifungal activity. Trichothecenes are reported to possess promising antifungal, phytotoxic and cytotoxic activities. Trichoderma-derived trichothecenes were mainly focused on their antifungal activity in the literature above, which highlighted their potential as biocontrol agents. Drimane sesquiterpenes 53-55 were active against C. miyabeanus by inducing hyphal branching at 1.0 and 10 µg/mL (Hirose et al., 2014). The novel norsesquiterpene 71 showed potent ability against C. lagrnarium with an MIC of 8 µg/mL (Du et al., 2020).

Phytotoxic Activities
In phytotoxicity assays, harzianums A (20) and B (21) decreased the shoot and root lengths of the dicot species Brassica chinensis and induced inhibitory effect of seed germination at 2 µg/mL . Moreover, 20 and 21 showed phytotoxicity against monocots, Oryza sativa and Echinochloa crusgalli, compared with the positive control 2,4dichlorophenoxyacetic acid (a chlorophenoxy herbicide most commonly used worldwide). The results indicated that 20 and 21 possess potent herbicidal potential for dicotyledon and/or monocotyledon weeds. All of the isolated harzianes 95-101 exhibited potent phytotoxicity at 200 ppm .

Other Activities
Cyclonerane sesquiterpenes 51 and 52 exhibited certain nematicidal activity against Meloidogyne incognita, with secondstage juvenile (J2s) lethal rates of 38.2 and 42.7% at 200 µg/mL (Du et al., 2020). Cyclodepsipeptides 112-114 also showed nematicidal activity against M. incognita (Du et al., 2020). New sesquiterpenes 72 and 73 showed potent NO scavenging effects, with IC 50 s of 15.3 and 9.1 µM, respectively (Zheng et al., 2011). Harzianoic acids 80 and 81 exerted potency to decrease the HCV RNA with EC 50 s of 24.5 and 20.4 µM, respectively (Li et al., 2019). Trichothioneic acid (134) showed OH radical-scavenging and singlet oxygen-quenching ability in a dose-dependent manner, which was equivalent to those of positive controls . The activity of harzianic acid (140) as a plant growth promoter was evaluated (Vinale et al., 2013). Treatment with 100 or 10 µM 140 significantly affected seed germination at 4 and 5 times stronger than that of the blank control. Naphthalene derivatives 145 and 147 showed moderate antifouling potency with EC 50 s of 29.8 and 35.6 µg/mL, respectively (Yu et al., 2021). The new de-O-methyllasiodiplodin 194 and 195 showed strong α-glucosidase inhibitory activity with IC 50 s of 25.8 and 54.6 µM, respectively, which were higher than acarbose (703.8 µM) .  (198)(199)(200)(201)(202)(203) according to their putative biogenetic sources. As shown in Figure 13A, 39.9% of the metabolites reported were sesquiterpenes, followed by polyketides with 26.6%. Taking diterpenes into account, terpenoids accounted for 51.72% of the obtained compounds, which indicated that species belonging to Trichoderma are considerable producing strains of novel terpenoids. It should be pointed out that some terpenoids, such as harzianes, are isolated exclusively from Trichoderma species. This review described 21 harziane diterpenes produced by Trichoderma. Considering their intriguing structures and bioactivities, much more attention should be devoted to this type of terpenoid in subsequent chemical studies.

Producing Strains
The genus Trichoderma comprises more than 340 species. Some of them are used as biocontrol agents, while some of them are promising producers of enzymes for industrial purposes. On the other hand, some Trichoderma species possess the unique capacity to synthesize various secondary metabolites with potent biological activities. In this review, a total of 17 identified species, including T. harzianum, T. brevicompactum, T. virens, T. gamsii, T. atroviride, T. longibrachiatum, T. asperellum, T. koningiopsis, T. koningii, T. citrinoviride, T. neokongii, T. spirale, T. afroharzianum, T. polypore, T. polyalthiae, T. erinaceum, and T. cremeum, are reported as the producing strains of the described metabolites. Among them, T. harzianum and T. brevicompactum were the most prolific strains, with 48 (23.76%) and 33 (16.34%) metabolites identified, respectively ( Figure 13B). The fungus T. harzianum is famous for widely used biocontrol agents, and it is also considered to be a promising producer of bioactive metabolic products. T. brevicompactum can synthesize trichothecene-type sesquiterpenoids with potent antifungal activity and high biotechnological value. Twenty-one novel trichothecenes (1-21) have been characterized from T. brevicompactum.

Environment Sources
The genus Trichoderma is widely distributed and has been isolated in soils, decaying wood, and endophytes in the inner tissue of host plants. Previous studies have mainly focused on terrestrial species of Trichoderma. However, Trichoderma from the marine environment are unexploited. It would be useful to examine marine-derived Trichoderma species since they may be induced to produce specific metabolites in hyperhaline environments. Accordingly, in recent years, increasing attention has been devoted to marine Trichoderma. As shown in Figure 13C, a total of 54.7% producing fungus were obtained from marine environments, including algae (32.02%), sponges (4.43%), soft corals (8.87), mangroves (3.94%), and other marine environments (5.42%, seawater, sediments), with 111 compounds characterized. Moreover, some fungi are obtained as endophytes from medicinal plants. Endophytic fungi, which harmoniously live in the inner tissues of their hosts without causing apparent diseases, are considered to be prolific sources of novel metabolites with remarkable pharmacological activities. It is estimated that 37.97% of these compounds were isolated from endophytic Trichoderma. From the above analysis, it can be concluded that marine environment and endophytes are more abundant sources of those productive strains.

Biological Activities
As discussed above, most of the presented compounds possess considerable biological activities, such as antimicrobial, antimicroalgal, anticancer, enzyme inhibitory, herbicidal, and nematicidal activities. Among them, antimicrobial (25.88%), anticancer (20.61%), and antimicroalgal (17.98%) activities were dominant in assessing the pharmacological potential of these metabolites (Figure 13D). It should be pointed out that a high proportion (73.40%) of the presented metabolites showed moderate to potent bioactivities. Even more importantly, a large number of them exhibit potent activities, which are higher than those of positive controls. For example, trichothecenes 22 and 23 showed higher antifungal effect on C. lagrnarium (MIC, 16 µg/mL) than the synthetic fungicide carbendazim (MIC, 32 µg/mL) (Du et al., 2020). Cyclopentenone 183 displayed potent cytotoxicities, whereas it was inactive toward the normal lung cell line (You et al., 2010). The selectivity index was even more remarkable than that of cisplatin, indicating 183 has high selective toxicity to cancer cell lines. Sesquiterpenes 72 and 73 showed potent NO scavenging effects (Zheng et al., 2011). These impressive bioactivities indicate that many of these compounds could be used as potential candidates for new drug discovery.

CONCLUSION
In the present review, we offer a detailed summary of recently isolated metabolites from Trichoderma from the beginning of 2009 to the end of 2020. As a result, a total of 203 metabolites are described herein, including their structural diversity and biological activities. Moreover, new strategies for discovering secondary metabolites of Trichoderma in recent years have also been discussed. Trichoderma has proven to be a treasure house of interesting secondary metabolites with medicinal importance. The biochemical studies of Trichoderma are untapped. Although a mass of metabolites have been isolated from Trichoderma species, the further excavation of those metabolites is worth expecting. By using new approaches to activate their silent gene clusters, including cultivation-based approaches, metabolomic profiling, and genome mining-based molecular approaches, an ever-increasing number of bioactive compounds will be obtained, which will be beneficial for the new drug discovery in the near future.

AUTHOR CONTRIBUTIONS
J-LZ and W-LT wrote this manuscript. Q-RH, Y-ZL, M-LW, L-LJ, CL, XY, H-WZ, and G-ZC collected and reorganized the literature data. X-XZ supervised the research work and revised the manuscript. All authors reviewed the manuscript.