Endophytic Fungi—Alternative Sources of Cytotoxic Compounds: A Review

Cancer is a major cause of death worldwide, with an increasing number of cases being reported annually. The elevated rate of mortality necessitates a global challenge to explore newer sources of anticancer drugs. Recent advancements in cancer treatment involve the discovery and development of new and improved chemotherapeutics derived from natural or synthetic sources. Natural sources offer the potential of finding new structural classes with unique bioactivities for cancer therapy. Endophytic fungi represent a rich source of bioactive metabolites that can be manipulated to produce desirable novel analogs for chemotherapy. This review offers a current and integrative account of clinically used anticancer drugs such as taxol, podophyllotoxin, camptothecin, and vinca alkaloids in terms of their mechanism of action, isolation from endophytic fungi and their characterization, yield obtained, and fungal strain improvement strategies. It also covers recent literature on endophytic fungal metabolites from terrestrial, mangrove, and marine sources as potential anticancer agents and emphasizes the findings for cytotoxic bioactive compounds tested against specific cancer cell lines.


INTRODUCTION
Human health is at constant risk due to the occurrence of different types of cancers. Cancer cells are characterized by the enormous replicative potential, apoptotic resistance, and invasive ability. The dislodging of cancer cells from the primary site may lead to the formation of cancer at the secondary site by a process of metastasis which eventually results in the death of an individual. Cancer is a major cause of death globally. An estimated 1,688,780 new cancer cases and 600,920 cancer-related deaths projected to occur in the US in 2017 (Siegel et al., 2017). In 2013, 14.9 million cancer incidences, 8.2 million cancer-related deaths were reported (Global Burden of Disease Cancer, 2015). The frequency of cancer-related mortality is increasing in developing countries at an alarming rate. Therefore, several research groups have been actively involved in the development of novel anticancer drugs throughout the globe. It is estimated that the total cost for the development of a new drug is ∼$2.6 billion. The global annual expenditure on anticancer drugs is ∼$100 billion and is estimated to increase to $150 billion by 2020 (Prasad et al., 2017). Many available anticancer drugs exhibit toxicity to proliferating normal cells, possess adverse effects, and are less effective against several types of cancer, which results in a need for bioactive compounds from natural products (Remesh, 2017).
Scientists have begun to comprehend that plants may be reservoirs for an infinite number of microorganisms, commonly referred as endophytes (Schulz et al., 2002;Strobel et al., 2004). Endophytes inhabit at internal plant tissue in a symbiotic association and can spend most of their life cycle within host plants. Approximately one million endophytic species are present in the plant kingdom (Fouda et al., 2015). It is noteworthy to mention that, a small number of endophytic fungi were reported to produce plant growth stimulating hormones such as gibberellic acid and indole acetic acid (Rai et al., 2014). Metabolites produced by endophytes can be influenced by the chemistry of their host plants . During the period of co-evolution, some of the endophytes have developed the ability to produce biologically active compounds that are similar or identical to their host plants Jia et al., 2016). Several studies have indicated the possible likelihood of medicinal plants that host endophytic fungi capable of producing pharmacologically important natural products. It is reasonable to postulate that the medicinal properties of these plants could be due to the endophytes residing within them. This has led to an investigation of beneficial compounds related to plant-endophyte interactions (Hardoim et al., 2008). The interaction between plants and endophytes is regulated by genes of both the organisms and controlled by the environmental factors (Moricca and Ragazzi, 2008). In other words, the endophyte may encounter metabolically unfavorable condition due to host defense chemicals (Schulz et al., 1999;Easton et al., 2009). The plant and endophyte association may also modulate the production of secondary metabolites in the host plant. The secondary metabolites isolated from endophytes displayed a broad spectrum of pharmacological properties including anticancer, antiviral, antibacterial, and antifungal activity (Jalgaonwala et al., 2017). In the subsequent sections, we discuss about the anticancer compounds extracted from endophytic fungus from diverse habitats.

NATURAL PRODUCTS AS SOURCES OF NOVEL ANTICANCER COMPOUNDS
Natural products obtained from endophytic fungi have been identified as continuing and prolific sources of anticancer agents that could have a profound impact on modern medicine for the advancement of anticancer drugs (Schulz et al., 2002;Newman and Cragg, 2007). In the modern era, the discovery of potent medicines with minimal side effects acts as an alternative to conventionally used methods of disease control and treatment (Kusari et al., 2009b;. In 2000, it was estimated that ∼57% of compounds in clinical trials for cancer therapy are natural products and their derivatives (Demain and Vaishnav, 2011). Early reports suggest that natural bioactive compounds isolated from endophytes may serve as alternate sources for the discovery of new anticancer drugs (Joseph and Priya, 2011;Alvin et al., 2014;Xie and Zhou, 2017).
Most of the reports dealing with the cytotoxicity of metabolites from endophytic fungi utilize antiproliferative assays in the cancer cell lines and high-throughput screening that essentially allows the evaluation of several compounds together for their potential anticancer activity (Aly et al., 2008;Lei et al., 2013). Isolation of natural bioactive compounds and screening them for the pharmacological properties offer a route for the discovery of drug candidates (Salvador-Reyes and Luesch, 2015). Several endophytic fungal strains have been identified and reported to produce new compounds that are effective in anticancer assays (Stierle and Stierle, 2015). Taxol, vincristine, etoposide, irinotecan, topotecan, and vinblastine are some of that plantderived anticancer drugs that are in clinical use for the treatment of several human cancers (Balunas and Kinghorn, 2005). The review describes the isolation of some of these compounds from endophytes and reports their cytotoxic effect in various cell lines.

CLINICALLY USED ANTICANCER DRUGS AND THEIR PRECURSORS FROM ENDOPHYTIC FUNGI
Although several natural compounds have been shown to possess good pharmacological and anticancer activities, they often face many challenges. The major challenges in developing natural compounds as drug candidates are identification of the right source; difficulty in extraction due to the excessive degradation by metabolic enzymes; low abundance of active principles; hindrance in large-scale production because of steric complexity high chiral centers, hydrogen bond donors/acceptors, bulky compounds, and diverse aromaticity (Singh et al., 2016). Identification of the good source of natural compounds can be achieved by random selection, folklore, codified systems of medicine, ethnopharmacology, ayurvedic classical texts, or zoopharmacognosy (Singh et al., 2016). Often, plant-derived bioactive natural compounds are of low abundance. For an instance, the total taxol content of the bark of Taxus baccata tree ranges between 0.064 to 8.032 g/tree, and it was also reported that a tree aged about 100 years can produce the dry bark yield of 5.74 kg (Nadeem et al., 2002). Therefore, it is essential to develop alternative strategies for the production of bioactive compounds using tissue culture, synthetic/semisynthetic approaches, biotransformation, and use of microbes that can produce desired products in large scale. This review presents bioactive compounds isolated from plant-associated fungal strains obtained from terrestrial, mangrove, and marine habitats, which are capable of inducing cytotoxicity/apoptosis in cancer cells and thereby possess efficient anticancer activity.

Taxol
Taxol (1) is the world's first billion-dollar anticancer drug and it is a highly functionalized polycyclic diterpenoid that belongs to a class of taxanes. In 1962, researchers from National Cancer Institute supported project, collected inner bark (phloemcambial tissue) of the Pacific yew tree (Taxus brevifolia) and analyzed for the presence of natural bioactive compounds. Initial screening of crude extract on cancer cells revealed good cytotoxic activity. It took several years to identify and isolate paclitaxel (trade name is taxol) in its pure form from the extract. Thereafter, paclitaxel was identified as a potent antitumor agent and made its way into clinical trials. One of the biggest hurdles faced during the initial days of taxol production is the requirement of six yew trees of 100 years old to treat one cancer patient (Demain and Vaishnav, 2011). In other words, 0.01 to 0.03% is the taxol content in dry weight of phloem of the yew tree. The constraints in the availability, isolation, and synthesis of taxol made the researchers to think of alternate sources for its production. The efforts resulted in the isolation of 10-deacetylbaccatin III (2) (a precursor for the synthesis of taxol) from T. baccata (European yew). The tree is abundant and bears high amount of 10-deacetyl-baccatin III in its needles and nowadays it is used as a precursor for the synthesis of taxol by semi-synthetic approach (Tulp and Bohlin, 2002). Eventually, FDA approved taxol for the treatment of several types of tumors including breast, ovary, and Kaposi's sarcoma. It is also claimed that taxol is the best-selling cancer drug ever manufactured (Gordon, 2011) with a market size of $1.6 billion in 2005 and its structural analog, docetaxel presented the sales of $3 billion in 2009 (Demain and Vaishnav, 2011). The efficacy and increased demand for taxol resulted in developing biotechnological approaches to prepare the drug (Kusari et al., 2014). In the present day, taxol is produced by semisynthetic approaches using 10-deacetyl-baccatin III, plant cell culture, and endophytic fungi. In a breakthrough, the T. brevifolia associated endophytic fungus Taxomyces andreanae was reported to produce taxol and related compounds (Stierle et al., 1993). This extraordinary feat led to the discovery of several new taxol-producing endophytic fungi from different host plants (Strobel et al., 1996;Strobel, 2003;Zaiyou et al., 2015). The production of paclitaxel was also identified in an angiosperm named Corylus avellena L which belongs to the family Betulaceae (Qaderi et al., 2012). In the next section, we have comprehensively discussed the mode of action of taxol in cancer cells, its endophytic fungal sources and cytotoxic ability.

Mode of Action
Paclitaxel represents a new class of antineoplastic agents and has a unique mode of action. It promotes and stabilizes the polymerization of microtubules and resists their depolymerization. In the presence of taxol, polymerized microtubule is resistant to depolymerization by cold (4 • C) and calcium chloride (4 mM) (Manfredi et al., 1982). This unusual stability of microtubules interfere with the mitotic spindle assembly, chromosome segregation which leads to mitotic arrest and eventually cell death (Schiff et al., 1979;Horwitz et al., 1986;Weaver, 2014).

Endophytic Fungi as Producers of Taxol
Several research groups have been focusing on the search for new sources of paclitaxel. Several Taxus species, such as Taxus baccata, Taxus chinensis, Taxus cuspidata, Taxus x media, Taxus floridana, Taxus canadensis, Taxus yunnanensis, Taxus mairei, Taxus sumatrana, and Taxus wallichiana, have been reported to produce taxol, albeit with significant variation in the taxane content within and among the different species (Tabata, 2006). Non-Taxus species, such as Cardiospermum halicacabum, Citrus medica, Cupressus sp., Ginkgo biloba, Hibiscus rosa-sinensis, Taxodium distichum, Podocarpus sp., Torreya grandifolia, Terminalia arjuna, and Wollemia nobilis, are also taxol producers (Flores-Bustamante et al., 2010). As discussed earlier, low taxol yields in the plants prompted researchers to discover alternate ways of taxol production which led to the identification of endophytic fungi as sources of taxol (Heinig and Jennewein, 2009). The discovery of taxol-producing endophytes from both Taxus and non-Taxus plants led to screening of array of plants for the identification of endophytes with taxol-producing abilities. Endophytic fungi belonging to the genera Alternaria, Aspergillus, Botryodiplodia, Botrytis, Cladosporium, Ectostroma, Fusarium, Metarhizium, Monochaetia, Mucor, Ozonium, Papulaspora, Periconia, Pestalotia, Pestalotiopsis, Phyllosticta, Pithomyces, and Taxomyces have been tested and reported for the production of paclitaxel and its analogs ( Table 1).
Paclitaxel production from endophytes is reported often in submerged culture (Wang et al., 2001;Zaiyou et al., 2017). Paclitaxel in the fungal extract can be detected by various analytical (Thin layer chromatography, High-performance liquid chromatography), spectroscopic (Matrix-assisted laser desorption/ionization-Time of flight [MALDI-TOF], Nuclear magnetic resonance [NMR], Fast atom bombardment [FAB], Electron spray ionization [ESI]), and immunological techniques (using monoclonal antibodies specific for paclitaxel) (Flores-Bustamante et al., 2010). Polymerase chain reaction-based methods can be used for the screening and identification of taxol-producing fungi. Previous studies have revealed that presence of genes encoding for the enzymes taxadiene synthase, 10-deacetylbaccatin III-10-O-acetyltransferase, and baccatin III 13-O-(3-amino-3-phenylpropanoyl) transferase could serve as molecular markers for the identification of fungi with taxol-producing abilities (Zhang et al., 2008;Vasundhara et al., 2016). Most taxol production has been reported using a liquid media such as potato dextrose broth (PDB), modified liquid  Kasaei et al., 2017 medium (MlD), and S7. However, the highest taxol yield of 846.1 µg/L was reported from the endophytic fungus, Metarhizium anisopliae of T. chinensis (Liu K. et al., 2009). The endophyte Cladosporium cladosporioides from Taxus media was also shown to produce 800 µg/L of taxol (Zhang P. et al., 2009). Sun et al. isolated 155 endophytic fungi from the tissue of Podocarpus and identified strain A2 as a taxol producer. Subsequently, A2 was classified as Aspergillus fumigates and the taxol yield obtained was 0.56 mg/L when grown in liquid potato dextrose medium. The isolated taxol inhibited the growth of Vero cells as similar to commercially available taxol (Sun et al., 2008). In another report, 34 endophytic fungi were isolated from the medicinal plant Salacia oblonga at Kigga village, Karnataka, India, and screened for their potential to produce taxol or taxanes. The authors used genomic mining approach to identify the taxolproducing fungus. Among the isolates, seven fungi were found to possess 10-deacetylbaccatin III-10-O-acetyltransferase gene and one revealed the presence of C-13 phenylpropanoid side chain-CoA acyltransferase gene (Roopa et al., 2015). Recently, 18 fungal isolates were screened for their ability to produce taxol. Cladosporium oxysporum isolated from Moringa oleifera was identified as taxol producing fungus and the taxol yield was found to be 550 µg/L. The fungal taxol suppressed the growth of HCT15 with an IC 50 value of 3.5 µM. The presence of gene encoding 10-deacetylbaccatin III-10-O-acetyltransferase confirmed the ability of an endophyte to produce taxol (Gokul Raj et al., 2015). Zaiyou et al. isolated 528 fungal strains from the bark of T. wallichiana var. mairei and only one strain was found to produce paclitaxel. The fungal strain was identified as Phoma medicaginis. The amount of paclitaxel produced when grown in whole potato dextrose broth culture, spent culture medium and dry mycelium was found to be 1.215, 0.936 mg/L, and 20 mg/kg, respectively (Zaiyou et al., 2017).

Cytotoxic Ability of Taxol and Its Precursors
Several cytotoxicity tests have been performed to determine the anticancer abilities of taxol and its precursors. One such interesting investigation highlighted that baccatin III (biosynthetic precursor of taxol) functions via a similar mechanism to taxol. Taxol and baccatin III were extracted from Fusarium solani and tested against HeLa, HepG2, Jurkat-JR4, OVCAR-3, T47D, Jurkat-JR16, and caspase-8-deficient Jurkat cells. Both compounds were able to inhibit cell proliferation of the above-mentioned tumor-derived cell lines with IC 50 values between 0.005 to 0.2 µM for taxol and 2 to 5 µM for baccatin III. These results indicate that although taxol induces apoptosis in the tested cells, there were difference in sensitivity between the tumor cells toward fungal taxol and baccatin III treatment. Among these two compounds, the potent inducer of apoptosis is debatable; because baccatin III is the more active molecule inside the cells during the growth period and is less active than taxol in vitro studies. Conversely, taxol has benefit over baccatin III in cellular uptake, microtubule binding kinetics and interaction with other proteins (Chakravarthi et al., 2013). The endophyte Diaporthe phaseolorum isolated from T. wallichiana var. mairei synthesizes baccatin III. The baccatin III content in PDB and spent culture medium was found to be 0.219 and 0.193 mg/L, respectively (Zaiyou et al., 2013). A detailed description of the taxoid biosynthesis pathway is the solution to improve the supply of taxol and other important taxoids. Several genes and enzymes of the Taxus pathway for the taxoid biosynthesis have now been established (Heinig and Jennewein, 2009). Future studies should focus on the isolation and identification of endophytes that are capable of producing a high amount of taxol, as well as optimization of the production conditions.

Podophyllotoxin
Podophyllotoxin (3), an aryl tetralin lignan derived from both gymnosperms and angiosperms, has also been investigated as an anticancer drug (Majumder and Jha, 2009); It is widely distributed in the genera Diphylleia, Dysosma, Juniperus, and Sinopodophyllum (also called Podophyllum) . Podophyllotoxin and its analogs are pharmacologically important due to their cytotoxic and antiviral activities (Gordaliza et al., 1994(Gordaliza et al., , 2000Abd-Elsalam and Hashim, 2013). At present, one of the major natural sources of podophyllotoxin is Sinopodophyllum plants. Etoposide (4) and teniposide (5) are the semisynthetic derivatives of podophyllotoxin approved for the treatment of cancer of lungs and testicles, other solid tumors and a variety of leukemias (Kusari et al., 2009a). Etoposide was developed as an alternative to podophyllotoxin in 1966 and received FDA approval in 1983. Subsequently, enzymeassisted asymmetric synthesis of (-)-podophyllotoxin and its C2epimer, (-)-picropodophyllin was described (Berkowitz et al., 2000). However, they are not economically feasible due to the low yield (Chandra, 2012). The supply of podophyllotoxin derivatives from traditional sources is limited due to their low abundance in the plants. A large amount of effort in the past several years has focused on improving production from different podophyllotoxin-producing plant species. Various alternative strategies have been implemented to enhance the production of podophyllotoxin-related compounds. Plant tissue culture is one of the alternate, reliable, and sustainable strategies to produce natural compounds of the plant origin (Ochoa-Villarreal et al., 2016).

Mode of Action
Etoposide and teniposide were reported to induce excellent antitumor activity in various cancer models. Their antitumor potential is due to their interaction with the enzyme topoisomerase II (Botta et al., 2001;Cortés and Pastor, 2003). Topoisomerase inhibitors impart their action by two mechanisms-either by eliminating the catalytic activity of topoisomerase II or by increasing the levels of topoisomerase II:DNA covalent complexes (often designated as topoisomerase II poisons) (Nitiss, 2009). Etoposide does not induce cytotoxicity by blocking the catalytic activity of topoisomerase II, but they poison mammalian topoisomerase II by dramatically increasing the covalent DNA cleavage complexes leading to the permanent double-stranded breaks. The increase in enzyme-associated DNA breaks serve as targets for genetic alterations such as recombination, exchange of sister chromatid, insertions, deletions, and translocation (Froelich-Ammon and Osheroff, 1995). The accumulation of permanent DNA breaks induces a series of events and drive the cell to death. During this process, topoisomerases serve as a physiological toxins (Van Maanen et al., 1988;Kaufmann, 1989;Hande, 1998;Pendleton et al., 2014).

Podophyllotoxins From Endophytic Fungi
The two strains of Phialocephala fortinii were isolated from the rhizomes of Podophyllum peltatum which produced podophyllotoxin and the yield ranged between 0.5 and 189 µg/L. The fungal extract was evaluated for cytotoxicity using brine shrimp lethality assay and the LD 50 values were in the range of 2-3 µg/mL (Eyberger et al., 2006). Similarly, Fusarium oxysporum, an endophytic fungus isolated from Juniperus recurva in Gulmarg region of South Kashmir, India. The fungus was found to produce podophyllotoxin and the yield was found to be 28 µg/g of dry mass . In another study, the endophytic fungus Trametes hirsute was isolated from the dried rhizomes of Podophyllum hexandrum which produced podophyllotoxin, podophyllotoxin glycoside and demethoxypodophyllotoxin. The isolated metabolites exhibited cytotoxicity in U-87 cell line . In another study, Aspergillus fumigatus Fresenius was isolated from Juniperus communis L. Horstmann collected from botanical gardens of Rombergpark, Dortmund, Germany. The fungus was found to produce deoxypodophyllotoxin and maximum yield of 100 µg/g of dry weight of mycelia was observed (Kusari et al., 2009a). In the next study, six fungi were collected from the rhizomes of Sinopodophyllum hexandrum (Royle) Ying plants at Taibai Mountains of China. Among the fungi, Mucor fragilis Fresen. produced podophyllotoxin and kaempferol. The podophyllotoxin yield was found to be 49.3 µg/g of mycelial dry weight (Huang et al., 2014). Recently, first report was published on production of podophyllotoxin from endophyte Alternaria tenuissima. The fungal endophyte was isolated from fresh roots of Sinopodophyllum emodi (Wall.) Ying at Xinglong Mountains, Gansu Province, China. The secondary metabolite analysis revealed the presence of podophyllotoxin in the fungal biomass (Liang et al., 2016). The list of podophyllotoxin producing endophytic fungi and their host plants has been tabulated and presented as Table 2.
Camptothecin Camptothecin (6), an anticancer drug that was first isolated from the bark of the Camptotheca acuminata. It is a pentacyclic pyrroloquinoline alkaloid and a parent compound for clinically used anticancer drugs such as irinotecan and topotecan. It is present in nature in 20-S-camptothecin form and 20-R-camptothecin enantiomer form is biologically inactive. Camptothecin is produced mainly by C. acuminata and Nothapodytes foetida. In National Cancer Institute screening studies, anticancer activity of wood extract of the C. acuminata was identified in 1958. In 1968, Wani and Wall isolated the cytotoxic agent from the extract of C. acuminata and identified the alkaloid as camptothecin (Takimoto, 2002). The purified compound exhibited good cytotoxic activity against several types of tumors. Conversely, clinical development of camptothecin was hampered in clinical trials due to poor water solubility and severe gastrointestinal toxicities (Takimoto, 2002). Eventually, water soluble and biologically active novel camptothecin derivatives were synthesized (Miyasaka et al., 1986). After four decades from the identification of antitumor activity of C. acuminata extract, two camptothecins, topotecan (7) and irinotecan (8) were approved by FDA for the treatment of colorectal cancer, small-cell lung cancer, and ovarian cancer (Blagosklonny, 2004). Meanwhile, camptothecin was identified to target topoisomerase-I in cells to impart its anticancer activity. Several camptothecin derivatives such as IDEC-132 (9-aminocamptothecin), rubitecan (9-nitrocamptothecin), and 10,11-methylenedioxy camptothecin have been identified as good anticancer agents (Ulukan and Swaan, 2002). In the next subsequent sections, we have discussed about the mechanism of action of camptothecin and endophytic fungi as its source.

Mode of Action
At the initial days, it was believed that camptothecins induce cytotoxicity in cancer cells by inhibiting DNA and RNA synthesis.
In the late 1980s, topoisomerase I was identified as the cellular target of camptothecin. Topoisomerase-I is involved in the relaxation of DNA supercoiling that is generated as a result of replication. Topoisomerase-I creates a nick in the single strand of DNA to release supercoiling. In parallel, topoisomerase-I forms an ester linkage with the 3' end of nicked DNA through its catalytic tyrosine to form topoisomerase-I cleavage complexes. Immediately 5'hydroxyl group of the nicked DNA strand nucleophilically attacks the tyrosyl-DNA-phosphodiester bond to bring the DNA to normal topology. In general, topoisomerase-I cleavage complexes are short-lived intermediates and hard to detect in the cellular environment (Pommier, 2006). Camptothecin and its derivatives interact with topoisomerase-I cleavage complexes and stabilize them. This results in the initiation of series of apoptotic events, finally leading to cell death. Studies using mutant yeasts (Saccharomyces cerevisiae and Schizosaccharomyces pombe) that lacks topoisomerase I activity were completely resistant to camptothecin indicating the absence of off-targets (Eng et al., 1988). It is now clearly demonstrated that the antitumor activity of camptothecin is due to its ability to inhibit DNA topoisomerase-I, which is involved in the swiveling and relaxation of DNA during replication and transcription.

Camptothecin From Endophytic Fungi
Camptothecin obtained from plant sources does not meet the requirement from the global market and consistent attempts have been underway to identify novel sources of camptothecin. Entrophospora infrequens is the endophytic fungus isolated from the inner bark of N. foetida in Konkan ghats, West coast of India. The yield of camptothecin was found to be 18 µg/mg of the chloroform extract. The isolated fungal camptothecin showed good cytotoxicity against A549, HEp2, and OVCAR-5 (Puri et al., 2005). In another study, the same fungus produced 4.96 mg of camptothecin per 100 g of dry mass in 48 h in a bioreactor . The two endophytic fungal strains of F. solani were isolated from Apodytes dimidiate in the Western Ghats, India. In broth culture after 4 days of incubation, the yield of camptothecin produced by two strains was found to be 37 and 53 µg/100 g (Shweta et al., 2010). In another study, Shweta et al. isolated three endophytic fungi (A. alternate, Fomitopsis sp., Phomopsis sp.) from fruit and seeds of Miquelia dentata and found to produce camptothecin, 9-methoxycamptothecin, and 10-hydroxycamptothecin. The yield of camptothecin obtained from A. alternata, Fomitopsis sp., and Phomopsis sp. was 73.9, 55.49, and 42.06 µg/g dry weight (Shweta et al., 2013). Three camptothecin-producing fungi, Aspergillus sp. LY341, Aspergillus sp. LY355, and Trichoderma atroviride LY357 were isolated from C. acuminata. The corresponding camptothecin yields were 7. 93, 42.92, and 197.82 µg/L, respectively. Unfortunately, LY341 and LY355 strains lost the camptothecin-producing capability with repetitive subculturing. On the other hand, consistent production of camptothecin was seen in LY357 from second to eighth generation (Pu et al., 2013;Kai et al., 2015).
In the next study, 161 fungi were isolated from C. acuminata. Among the isolates, Botryosphaeria dothidea X-4 fungus was reported to produce 9-methoxycamptothecin (Ding et al., 2013). The endophytic fungi Neurospora sp. was isolated from the seed of N. foetida and found to produce camptothecin. The isolated fungal camptothecin was tested against the A549 and OVCAR-5 cells and it showed good cytotoxicity in both the cell lines . In another study, the fungus (XZ-01) that belongs to Phomopsis sp. was isolated from twigs of C. acuminata collected from the Jiangshi Natural Reserve, Fujian, China. Details of the camptothecin-producing fungi and their corresponding host plants are provided in Table 3.

Vinca Alkaloids
The success story of isolation of vinca alkaloids from Madagascar periwinkle (botanical name is Catharanthus roseus G. Don and commonly called Vinca rosea) marks back to late 1950s. In Jamaican folklore, the extracts of periwinkle were used as an oral hypoglycemic agent in the absence of insulin treatment (Noble, 1990). Eventually, two groups (Noble, Beer, and Cutts at the Collip Laboratories; Svoboda, Johnson, Neuss, and Gonman at the Lilly Research Laboratories) investigated the antidiabetic property of periwinkle extracts but did not find substantial hypoglycemic activity. The intense research on the extracts of periwinkle led to the isolation of an alkaloid, which reduced the WBCs and depleted bone marrow in rats. The new alkaloid was named as vincaleukoblastine because of its origin and effect on WBCs; subsequently named as vinblastine. Thorough phytochemical investigation of the extract over a period of a decade led to the discovery of anticancer alkaloids including vinblastine (9), vincristine (10), vinleunosine, and vinrosidine (Johnson et al., 1963;Noble, 1990Noble, , 2016. The vinca alkaloids are terpenoid indoles obtained from the coupling of catharanthine monomer and vindoline and are mainly used in chemotherapy regimens due to their ability to reduce the number of white blood cells in acute lymphoblastic leukemia and nephroblastoma. Vinca alkaloids represent the second most used class of anticancer drugs in the treatment of various malignancies (Moudi et al., 2013).

Mode of Action
Vinca alkaloids target cell cycle progression by interfering with activities of the microtubule cytoskeleton. Microtubules are the cytoskeletal elements made up of tubulin heterodimers which are associated with transportation of cargo within the cell, motility of the cell itself, and progression of the cell cycle (Downing, 2000). In the structural organization of microtubules, α and β tubulins arrange themselves alternatively to form a protofilament and α subunits are projected toward minus end and β subunits toward plus end (confers polarity). Further, protofilaments arrange in a parallel fashion to form a hollow cylindrical structure called microtubule (Kirschner, 1980;Dumontet and Jordan, 2010). The microtubules grow and shrink rapidly by addition and removal of tubulin subunits by a phenomenon referred as dynamic instability of microtubules. The free subunits are added up at plus end with the subsequent release of subunits at minus end which results in the steady-state maintenance of microtubule length (Alberts, 2017). The shortening and lengthening of a microtubule are essential to pull sister chromosomes during mitosis and push out membranes respectively (Wilson et al., 1999). Impeding the dynamics and stability of microtubules can block cell division in multiple ways. Vinca alkaloids are found to interact with the vinca domain of β tubulin in order to destabilize the microtubule structure. The binding site is present toward the plus end of microtubule and vinblastine suppresses dynamic instability at plus ends (Wilson et al., 1999). Previous findings suggested that vinca derivatives inhibit cell cycle at metaphase of mitosis (Jordan et al., 1991). They bind to tubulin intracellularly, resulting in the subsequent dissolution of microtubules, and prevent cells from assembling the spindles needed for the division, thus arresting cells in mitosis, which is necessary to mediate their cytotoxic effects (Jordan et al., 1991;Newman and Cragg, 2007). Primarily vinca alkaloids interfere with the formation of microtubule and mitotic spindle dynamics coupled with the disruption of intracellular transport. Overall, vinca alkaloids impart their antineoplastic effect by serving as microtubuledestabilizing agents. Vincristine and vinblastine are approved for the treatment of Hodgkin lymphoma and structural analogs of vinca alkaloids (Vinflunine, Vinorelbine, Anhydrovinblastine) targeting tubulin polymerization are in phase-II/III trials for the treatment of breast cancer and carcinoma (Kaur et al., 2014).

Vinca Alkaloids From Endophytic Fungi
An endophytic fungus, Alternaria sp., isolated from the phloem of Catharanthus roseus was reported to produce vinblastine (Guo et al., 1998). Later, the endophyte, F. oxysporum isolated from the phloem of C. roseus was found to produce vincristine (Zhang et al., 2000). An endophytic fungus from the leaves of C. roseus was also reported to produce vincristine (Yang et al., 2004). The endophytic fungus, F. solani from C. roseus was screened for vinca alkaloids using thin layer chromatography and electron spray ionization-mass spectroscopic analysis. The fungus was identified to produce vincristine and vinblastine . The amount of vinblastine and vincristine in the culture filtrate was found to be 76 and 67 µg/L, respectively. The endophytic fungus Talaromyces radicus from C. roseus produced 670 µg/L of vincristine in modified M2 medium and 70 µg/L of vinblastine in PDB medium. Vincristine was partially purified and evaluated for cytotoxicity in HeLa, MCF7, A549, U251, and A431 cells. The treatment of vincristine resulted in the dose-dependent growth inhibition in HeLa, MCF7, A549, U251, and A431 with IC 50 values of 4.2, 4.5, 5.5, 5.5, and 5.8 µg/mL, respectively. However, the normal cells (HEK293) were not significantly affected (Palem et al., 2015).

CYTOTOXIC COMPOUNDS FROM ENDOPHYTIC FUNGI ASSOCIATED WITH TERRESTRIAL PLANTS
The secondary metabolites from endophytic fungi have been widely recognized as potent antitumor agents. The secondary metabolite and bioactive compounds from endophytic fungi are screened for their antiproliferative activity against different cancer cell lines (Please refer Tables 4-6 for examples). Preliminary screening of extracts or compounds using cytotoxicity assays allows the rapid identification of compounds with anticancer activity. It is also essential to identify compounds that discriminate between cancer and normal cells, and specifically induce cytotoxicity in cancer cells. Specific cytotoxicity is imperative in cancer therapy, as the standard chemotherapy often affects actively dividing normal cells. However, few anticancer agents may be carcinogenic (Blagosklonny, 2005). Hence, toxicity screening is one of the major steps employed to identify anticancer agents (Parasuraman, 2011). In the next section, compounds isolated from endophytic fungi associated with terrestrial plants and their anticancer potential have been discussed.
The endophytic fungus Massrison sp. was isolated from the roots of wild Rehmannia glutinosa collected from Wushe County, Henan Province, China. Extract from the culture broth revealed the presence of massarigenin D (30), spiromassaritone (31), and paecilospirone (32). These compounds contain a rare spiro-5,6-lactone ring skeleton. The cytotoxic effects of these three compounds were tested against LO2, HepG2, MCF7, and A-2 cells. The compounds showed good cytotoxicity against the tested cells. Spiromassaritone displayed potent cytotoxicity against all the tested cells with an IC 50 value ranging between 5.6 and 9.8 µg/mL (Sun et al., 2011).  Among the isolated compounds, only potent compounds are included in the column "isolated compound/s". NA, Not active; NR, Not reported; NT, Not tested.
The endophytic fungus was collected from stems of Salvia officinalis from the mountain of Beni-Mellal, Morocco. Chemical analysis of secondary metabolites revealed the presence of two known compounds namely, cochliodinol (33) and isocochliodinol (34). The bioactive compounds were tested for cytotoxicity against L5178Y cells. Cochliodinol was potent than isocochliodinol, with an EC 50 of 7.0 µg/mL compared with 71.5 µg/mL for isocochliodinol. The authors interpreted that, the difference in cytotoxicity of structurally related compounds was due to the position of prenyl substituents at the indole rings (Debbab et al., 2009b).  .
The endophyte Stemphylium globuliferum was isolated from the stem tissues of Egyptian medicinal plant Mentha pulegium collected in Morocco. Chemical investigation of extract revealed the presence of five novel compounds (alterporriol G, atropisomer alterporriol H, altersolanol K, altersolanol L, and stemphypyrone) and eight known compounds macrosporin,altersolanol A,alterporriol E,alterporriol D,alterporriol A,alterporriol B,and altersolanol J). The isolated compounds were tested for cytotoxic activity against L5178Y cells. An unresolved mixture of alterporriol G (44) and its atropisomer alterporriol H (45) exhibited the most potent cytotoxicity with an EC 50 value of 2.7 µg/mL. The known compound 6-O-methylalaternin (46) also exhibited good cytotoxicity with an EC 50 value of 4.2 µg/mL. The positive control (kahalalide F) exhibited an EC 50 value of 6.3 µg/mL (Debbab et al., 2009a).
In another study, a new dihydrobenzofuran-2,4-dione derivative named as annulosquamulin (49) along with 10 known compounds were obtained from the endophytic fungus Annulohypoxylon squamulosum. The endophyte was isolated from the stem bark of the medicinal plant Cinnamomum sp. at Fu-Shan Botanical Garden, I-lan County, Taiwan. Annulosquamulin was evaluated for its cytotoxicity against MCF7, NCI-H460, and SF-268 cell lines. The IC 50 values of annulosquamulin were 3.19, 3.38, and 2.46 µg/mL, respectively (Cheng et al., 2012).
The endophytic fungal strain Allantophomopsis lycopodina was isolated from a tree branch on the Gassan stock farm, Yamagata, Japan, yielding a new natural product, allantopyrone A (50), and a previously reported islandic acid-II methyl ester (51). Both compounds exhibited good cytotoxic activity against HL60 cells with IC 50 values of 0.32 and 6.55 µM, respectively. Cleavage of the genomic DNA is a hallmark event in the cells committed to apoptosis (Sebastian et al., 2016) which leads to the formation of cells with lesser DNA content called hypodiploid cells Roopashree et al., 2015). Treatment of HL60 cells with these compounds resulted in fragmentation of genomic DNA and N-acetyl-L-cysteine prevented this effect. Allantopyrone A demonstrated stronger cytotoxicity and apoptosis-inducing activity than the islandic acid-II methyl ester (Shiono et al., 2010).
An endophytic fungus, Cochliobolus kusanoi, isolated from the stem of Nerium oleander L. produced the known compound oosporein (52). The cytotoxicity of oosporein was evaluated toward A549 and presented the IC 50 value of 21 µM (Alurappa et al., 2014). The cytotoxic potential of oosporein in the MDCK cell line and its role in oxidative stress in vivo and in vitro have been previously reported by our research group (Ramesha et al., 2015).
Ginsenocin and penicillic acid also showed effective cytotoxicity with IC 50 values ranging from 0.49 to 7.46 µg/mL against tested cancer cells .
The novel alkaloids chaetoglobosin V (97) and W (98), together with six known congeners were obtained from C. globosum isolated from Artemisia annua at Nanjing, Jiangsu Province, China. Chaetoglobosin V and W displayed moderate cytotoxic activity against MCF7 and HepG2 cell lines with IC 50 values of 52.76 and 51.04 µM, respectively. The cytotoxicity of chaetoglobosin V and W was also assessed against KB, K562, MCF7, and HepG2 cells. Chaetoglobosin V displayed the IC 50 value of 23.53 and 27.86 µg/mL in KB and MCF7 cells whereas chaetoglobosin W showed 21.17 and 27.87 µg/mL against the same cells, respectively. However, both the compounds did not show cytotoxicity in K562 cells up to 30 µg/mL ).

CYTOTOXIC METABOLITES OF ENDOPHYTIC FUNGI FROM MANGROVE PLANTS
Mangrove endophytic fungi are the second largest ecological group of marine fungi and are potential sources for isolation of new bioactive compounds (Kalaiselvam, 2015). They have attracted researchers due to their importance in ecology (Calcul et al., 2013). Presently, only a small number of mangrove fungi have been studied. Fungal endophytic associations with mangrove allow them to successfully compete with saprobiotic fungi for decomposition of their senescent parts (Elavarasi et al., 2012). Mangrove-associated fungi provide a wide array of bioactive secondary metabolites with unique structures including alkaloids, benzopyranones, chinones, flavonoids, phenolic acids, quinones, steroids, terpenoids, tetralones, and xanthones (Huang et al., 2011). In the following section, natural compounds isolated from endophytic fungi from mangrove plants have been discussed.
Three new bianthraquinone derivatives, alterporriol K (124), L (125), and M (126), together with six known compounds (physcion, marcrospin, dactylariol, tetrahydroaltersolanol B, alternariol, and alternariol methyl ether) were isolated from Alternaria sp. ZJ9-6B from Aegiceras corniculatum collected from the South China Sea. In preliminary studies, Alterporriol K and L imparted moderate growth inhibitory activity against MDA-MB-435 and MCF7 cells with IC 50 values ranging from 13.1 to 29.1 µM (Huang et al., 2011). An ergochrome class mycotoxin, secalonic acid D (127) was isolated from the endophytic fungus No. ZSU44 and evaluated for its anticancer activity against HL60 and K562 cells. The results showed strong cytotoxicity towards HL60 and K562 cells with IC 50 values of 0.38 and 0.43 µM, respectively. Furthermore, apoptosis inducing effect of secalonic acid D was confirmed by annexin V-FITC/PI staining. Activation of executioner caspase and PARP cleavage are the crucial events in the cells undergoing apoptosis (Ashwini et al., 2015;Baburajeev et al., 2015).
Secalonic acid D induced the activation of caspase-3 and cleavage of PARP in HL60 and K562 cells . The endophytic fungus Phomopsis sp. was isolated from Excoecaria agallocha (Euphorbiaceae) in Dong Zai, Hainan, China. The secondary metabolite analysis revealed the presence of three unique compounds namely phomopsis-H76 A, B, and C (128-130). These metabolites did not show considerable cytotoxic activity against KB, KBv200, and MCF7 cell lines and IC 50 values were over 50 µM .
Three new sesquiterpenes, merulin A (131), B and C (132) were produced by the endophytic fungus XG8D (Basidiomycete) isolated from the mangrove plant Xylocarpus granatum (Meliaceae). Merulin A and C exhibited good cytotoxic activity against human breast and colon cancer cell lines with IC 50 values of 4.98 and 1.57 µg/mL for BT474; 4.84 and 4.11 µg/mL for SW620, respectively. Doxorubicin was used as vehicle control with IC 50 values of 0.09 and 0.53 µg/mL against SW-620 and BT-474 cell lines, respectively. (Chokpaiboon et al., 2010). A new pyrrolyl 4-quinolinone alkaloid, penicinoline (133), was isolated from Penicillium sp., from the bark of the mangrove Acanthus ilicifolius (Acanthaceae) collected from the South China Sea. Intramolecular dehydration of the compound yielded an unexpected lactam derivative (134), which was confirmed by single-crystal X-ray analysis. Penicinoline showed potent cytotoxicity with IC 50 values of 0.57 and 6.5 µg/mL toward the 95-D and HepG2 cell lines, respectively. However, penicinoline was inactive against HeLa, KB, KBv200, and HEp2 cell lines at the concentration of 100 µg/mL (Shao et al., 2010).
The endophytic Pestalotiopsis sp. was obtained from the leaves of Rhizophora mucronata from Dong Zhai Gang, Mangrove Garden on Hainan Island, China. The endophyte produced five novel cytosporones J -N (135-139), five new pestalasins A-E (140-144), and a new alkaloid pestalotiopsoid A (145). All the isolated compounds were investigated for their cytotoxicity against L5178Y, HeLa, and PC12 cells. All the compounds were found to be inactive at a concentration of 10 µg/mL (Xu et al., 2009b). The Pestalotiopsis sp. produced six new pestalotiopsones A-F (chromones) (146-151) and a new compound 7-hydroxy-2-(2-hydroxypropyl)-5-methylchromone. The fungus was isolated from the leaves of Rhizophora mucronata. Among the isolates, only pestalotiopsone F exhibited moderate cytotoxicity against the murine cancer cell line (L5178Y) with an EC 50 value of 8.93 µg/mL (Xu et al., 2009a).
Five highly oxygenated chromones named rhytidchromones A-E (152-156) were isolated from the endophytic fungus Rhytidhysteron rufulum BG2-Y. The fungal strain was isolated from the healthy leaves of Bruguiera gymnorrhiza collected from Pak Nam Pran, Prachuab Kiri Khan Province, Thailand.
Isolated compounds were tested for their cytotoxicity on a panel of cancer cell lines (MCF7, HepG2, Kato-3, and CaSki). All the compounds excluding rhytidchromone D, displayed cytotoxicity toward Kato-3 cell lines with IC 50 values ranging from 16.0 to 23.3 µM, while rhytidchromones A and C were active against MCF7 cells with IC 50 values of 19.3 and 17.7 µM, respectively (Chokpaiboon et al., 2016).

ENDOPHYTIC FUNGAL METABOLITES FROM MARINE HABITATS
Marine fungi comprise of obligate and facultative types; and classically they are defined as a form of ecological, but not a taxonomic group of fungi (Elsebai et al., 2014). The growth and sporulation of obligate marine fungi occur exclusively in marine environment. In case of facultative marine fungi, the growth, and sporulation in the marine habitat are dependent on the physiological adaptations (Raghukumar, 2008). All marine habitats such as marine plants (mangrove plants, algae, seagrasses, and driftwood), marine vertebrates and invertebrates (corals, bivalves, sponges, and crustaceans), and abiotic factors (soil) can host fungal strains. Many reviews have not focused significantly on bioactive natural compounds from endophytic fungi of marine origin. The fungal diversity in marine habitats has been underestimated because of their propensity to form aggregates (Rateb and Ebel, 2011). Sponges and algae contribute majorly as sources of fungi in marine habitats. A thorough investigation of algae and sponges is essential to explore the fungal communities associated with them. The marine environment has extreme conditions like high salinity, more amount of metals in hydrothermal vents, elevated hydrostatic pressure, and low temperatures in the deep sea (Raghukumar, 2008). The organisms that live and thrive in the adverse and extreme conditions can be expected to produce metabolites that are unique and might be of therapeutic interest. Therefore, it is essential to search for exotic and unique metabolite-producing organisms in these unusual environments. Marine-derived fungi are a prolific source of new bioactive compounds with a high degree of structural diversity and equally interesting pharmacological activities, namely, anticancer, antibacterial, antifungal, and antiviral activities, as well as involvement in specific enzyme inhibition activities (Bugni et al., 2004). The bioactive natural compounds from marine fungi have been investigated to smaller extents relatively with their terrestrial counterparts (Vita-Marques et al., 2008;Gamal-Eldeen et al., 2009). Several research groups have been involved in the isolation of marine microbes, analysis of their secondary metabolites and evaluation of possible pharmacological properties.
The endophytic fungus Chaetomium globosum (QEN-14) was isolated from the marine green alga Ulva pertusa (Ulvaceae) collected from Qingdao coastline, China. The fungus yielded seven new cytochalasan derivatives (cytoglobosins A-G) together with two known compounds (isochaetoglobosin D and chaetoglobosin F ex ). New isolates were evaluated for their cytotoxicity against P388, A549, and KB cancer cell lines. The results indicated that cytoglobosins C (167) and D (168) as potent anticancer agents towards A549 cells with IC 50 values of 2.26 and 2.55 µM, respectively (Cui et al., 2010b). The other compounds did not show significant cytotoxic effect up to 10 µM.

FUTURE CHALLENGES WITH ENDOPHYTES
Endophytic fungi offer new hope and possibilities for the production of novel bioactive compounds, which is challenged due to low yield, lack of information on precise fungiplant biochemical interaction, scale-up issues, and growth in the axenic state. There is a constant need to manifest these natural secondary metabolites in vitro, which despite being accomplished, has shown limited success in commercial production. There is a weak understanding of interactions of endophytic fungal organisms with other associated microbes. The use of plant tissue culture techniques to obtain higher yields has not been successful due to genetic instability in culture, slow growth of fungi, the formation of cell aggregates, and sensitivity to shearing (Charlwood and Rhodes, 1990). Additionally, the maintenance of tissue cultures is comparatively more expensive and time-consuming than fermentation technologies. Nevertheless, fungal fermentation provides the platform for the potent and sustainable production of bioactive compounds. Microbial fermentation offers certain advantages such as a simple fungal cell medium, low cost, faster growth rate of fungi with minimal risk of contamination, and in vitro optimized fermentation conditions that facilitate the production of secondary metabolites but fails to meet the requirements for sustained production, as observed in vivo in the host apoplast region. Studies examining genetic aspects to trigger specific genes for overproduction of target metabolites could also be beneficial. Another important hurdle is the difficulties associated with using a combinatorial chemical synthesis approach to generate complex metabolites of the bioactive compounds. These shortcomings may overcome by the use of inducers/elicitors of the specific biochemical pathway that can ultimately increase the yield. The complex structure of endophytic metabolites is a hurdle for use of the combinatorial chemical synthesis approach. Alternatively, the limitations may overcome by the use of signaling molecules and inexpensive precursor feeding in growth medium, which may act as efficient triggers of the biochemical pathway to achieve an elevated yield. Moreover, the presence of several unnecessary metabolites can interfere with the enzymatic activity in the growth medium. The negative feedback mechanism is a major aspect that must be critically investigated to achieve increase in yield with the fungal fermentation strategy (Zhao et al., 2010a). Endophytic studies employing genetic engineering techniques, transformation studies, gene cluster amplification for metabolite overproduction, mutagenesis, medium manipulation, culture condition optimization, and elicitor addition should help to increase the yield of the desired secondary metabolites Venugopalan and Srivastava, 2015).

CONCLUSION
Endophytes have been emerged as privileged sources of bioactive compounds, since after the discovery of taxol from T. andreanae. The secondary metabolites of endophytes have been used substantially for the sustainable production of therapeutically important compounds. The limited availability of bioactive principles in plant sources could be surpassed by exploiting the chemical entities in the endophytes. With the unfolding knowledge of the anticancer properties of endophytes and current technological advancements in "omics" and fermentation, endophytic fungi may play a lead role in providing a constant supply of chemotherapeutics with high target specificity and minimal side effects at an affordable cost. The attempts in genetic manipulations, overexpressing of the genes that are essential for the production of the specific bioactive metabolite, and other optimization strategies may contribute to the strain improvement and make endophytes as potential drug sources for the future. Therefore, exploring and exploiting of metabolites from endophytes in terrestrial, mangrove, and marine habitats may provide an excellent avenue for the discovery of drug candidates against deadly human diseases.

AUTHOR CONTRIBUTIONS
All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.