Approximately, 30 nHSTs secondary metabolites are known and characterized in which alternariol (AOH), alternuene (ALT), zinniol, tenuazonic acid (TeA), alteriol monomethyl ether (AME), brefeldin A (dehydro-), tentoxin (TEN), curvularin, and alterotoxin (ATX) I, II, III, are some of the known toxins produced by different Alternaria species (Andersen et al., 2015; Lee et al., 2015). Fujiwara et al. (1988) suggested that brefeldin A actions on Golgi-complex cause disassembly and act as an inhibitor of its secretion, while curvularin inhibits cell division by microtubule assembly disturbance (Robeson and Strobel, 1981) and tenuazonic acid inhibits process of protein synthesis (Meronuck et al., 1972). Zinniol disturbed membrane permeabilization (Thuleau et al., 1988), and tentoxin inhibits chloroplast photo-phosphorylation through binding to ATP synthase resulting in the ATP hydrolysis and ATP synthesis inhibition (Steele et al., 1978). Since, these toxins often target basic cellular processes, so these are regarded as potent mycotoxins. By UV/Vis and MS/MS spectra comparison, alternarionic acid, dehydrocurvularin, pyrenochaetic acid, and altechromone A were identified from species of Alternaria and other fungal SMs. Most of these are derivatives of AOH and TEN (Andersen et al., 2015).

On the basis of scientific observation, there are seven HSTs have been identified from A. alternata. These HSTs are families of closely related diverse group low molecular weight natural compounds (Tsuge et al., 2013). These families are: (1) Epoxy-decatrienoic acid (EDA)—AK-, AF-, and ACT- are the known toxic compound in this family, (2) Cyclic depsipeptide (cyclic tetrapeptide)—AM toxin also known as alternariolide, (3) Amino pento/polyketide (these are sphinganine analog compound)—AAL-toxin, and (4) Polyketide—ACR-toxin (syn. ACRL-toxin).

Interestingly, another HSTs is HC-toxin (a cyclic tetrapeptide), a well-known produced toxin by the plant pathogenic fungus Cochliobolus carbonum, has also been discovered in the species of Alternaria jesenskae, pathotype of Fumana procumbens, and the gene responsible for this toxin is AjTox2 identified by genomic sequencing. Culture filtrate confirmed HC-toxin production through reverse phase HPLC and TLC assay (Wight et al., 2013), while its taxonomic identity confirmed by sequencing of ITS regions (Labuda et al., 2008).

These toxins are involved in the development of few destructive diseases. Alternaria pathotypes produce HSTs which are diverse ranged chemical compounds ranging from low molecular weighted peptides to cyclic peptides. Wolpert et al. (2002) reported that HSTs were biologically produced on their diverse specific plants species, so that it determines the host range of toxin-producing pathogens. Often, not all genotypes of a host plant species are sensitive to the toxin, and similarly, not all isolates of a pathogen species produce the toxin. Table 1 showing host-specific toxins (HSTs) produced by different species of Alternaria.

These HSTs produced several effects on a narrow species that serve as host to the fungi and are necessary for disease. In contrast to the HSTs, the dominant properties of less virulent small-spored species of Alternaria producing host-specific toxins are mainly due to strong selection pressure corresponding from modern monocrop agriculture and newly developed susceptible genotypes (Chou and Wu, 2002). Till now, 13 HSTs are identified by the different variants of Alternaria species and most of these variants are considered as A. alternata pathotypes (Table 1). In HSTs producing and nHSTs producing Alternaria species, all pathogenic species have some extra chromosomes, while these extra chromosomes are not carried by non-pathogenic species (Akamatsu et al., 1999; Akamatsu, 2004). The extra chromosomes are not required for normal growth but have an adaptive advantage in some habitats to the individual species. These are referred as conditionally dispensable chromosomes (CDCs).

The plant pathogen fungi undergo several complex and crucial processing steps of pathogenesis viz. attachment to the plant surface, germination and formation of infectious structures, penetration, and colonization into host cell, which are crucially important steps to cause disease. Generally, plasma membrane, chloroplast, mitochondria, and some important enzymes are the inhibitory sites for the action of HSTs of A. alternata but the other target sites are ER, nucleus, vacuole, and Golgi bodies (Tsuge et al., 2013). All target sites are identified on the basis of different scientific studies (Tsuge et al., 2013; Meena et al., 2016a; Zhang et al., 2016). HSTs produce their effects on different targeted cell organelles and induce cell death in hosts (Figure 2). HSTs generally cause suppression of defense responses of genotype and also disturb other metabolic signaling pathway. HSTs have both properties such as necrosis and suppression of defense process of susceptible hosts (Wolpert et al., 2002). Kohmoto et al. (1989) observed the differences between pathogenicity and host cell death when interactions occur between AM-toxin producing isolates with apple and AK-toxin producing isolates with Japanese pear.

Alternaria HSTs play a vital role as an effector which determines the pathogenicity. Many scientists have reported that simultaneous treatment with HSTs and conidia of non-pathogenic A. alternata strains lead to the initiation of infection (Akimitsu et al., 1989; Yamagishi et al., 2006). The controversial finding of many scientists states that the gene silencing causing loss of HSTs production which leads to pathogenicity disappearance (Harimoto et al., 2008; Miyamoto et al., 2008). A hypothetical idea based on HST-receptor mediated model of pathogenesis was given by Pringle and Scheffer (1964). Before penetration fungal spp. produces a signal molecule to recognize host, and after that release fungal HSTs which specifically binds to host receptor site, and finally, these suppress the host defense against fungus and leads to cell death (Scheffer and Livingston, 1984; Kohmoto and Otani, 1991).

AAL-toxins are chemically propane 1,2,3-tricarboxylic acid (PTCA) which is esterified to 1-amino-11,15-dimethylheptadeca-2,4,5,13,14-pentol. These are structurally sphingosine and sphinganine analogs (Bottini and Gilchrist, 1981; Brandwagt et al., 2001). The scientist identified five types of AAL-toxin related molecules viz. sphingosine (TA), phytosphingosine (TB), sphinganine (TC), tetra-acetyl-phytosphingosine N- Lignoceroyl-d (TD), and L- sphinganine, each consisting with two isomers (Caldas et al., 1994). TA and TB analogs of AAL-toxins showing related specific action and showing 30–400 times higher activity than the other form analogs TC, TD, and TE (Caldas et al., 1994). So, these two toxins are taken into consideration and referred as AAL-toxins. These toxins are produced by Alternaria alternata f. sp. lycopersici, a pathogen causing stem canker in tomato (Lycopersicon esculentum) which exhibits high degree of host specificity and plays a major role in pathogenesis by causing leaf necrosis (Prasad and Upadhyay, 2010). TA type of toxin is most active and highly produced which has molecular mass of 522 kb. AAL-toxin sensitivity and insensitivity to plant tissue have been revealed (Spassieva et al., 2002). The necrotic symptoms of AAL-toxin can be detected from a susceptible line of detached leaflets (Brandwagt et al., 2001). Gilchrist (1997) assumed that AAL-toxin inhibits the activity of aspartate carbamoyl transferase (ACTase) on disruption of pyrimidine metabolism but it could not be confirmed experimentally. AAL-toxins isolated from Alternaria alternata f. sp. lycopersici inhibited ceramide synthase and induced programmed cell death (PCD; Michaelson et al., 2016).

Kawaguchi et al. (1991) observed that ethanolamine (EA), phosphoethanolamine (PEA) and five other related chemicals accumulation occur in AAL-toxin treated leaves. It was also found that cell free ACTase preparation of homozygous susceptible and homozygous resistant genotypes of host and non-host specific sources express differential AAL-toxin sensitivity. Generally, AAL-toxins have its effect on mitochondria but their exact target site is still uncertain. The possible AAL-toxin biosynthetic pathway was in under investigation because EA and PEA are primary and secondary intermediate metabolites of biosynthetic pathways.

Orolaza et al. (1992) suggested that ¹⁴C labeling of ethanolamine to susceptible AAL-toxin treated leaf discs have shown strong inhibition of EA incorporated into phosphatidyl ethanolamine (PtdEA). Therefore, phospholipid pathway involved enzymes were suggested as potential biochemical targets for AAL-toxins. AAL-toxins have structural similarities to fumonisins because AAL-toxins have one PTCA and fumonisins with two PTCA side chains esterified to aminopentol back bones (Brandwagt et al., 2000). Gilchrist (1998) referred both AAL-toxin and fumonisin collectively as sphinganine analog mycotoxins (SAMS) due to their structural and functional similarities, and toxicity to plant and mammalian cells. They also show inhibitory action to sphingolipid biosynthesis and induced PCD in both plant and mammalian cells (Abbas et al., 1994; Spassieva et al., 2002, 2006; Tsuge et al., 2013). The sensitivity and disease resistance in tomato to AAL-toxin are regulated by Asc1 (Alternaria stem canker resistance gene 1) gene locus which encodes a host recognition factor (Egusa et al., 2009). These plant-pathogen interactions come under direct or indirect interaction between AAL-toxin and Asc locus products in which Asc action is linked to sucrose transport, ethylene biosynthesis, pyrimidine metabolism, and cell death (Moussatos et al., 1994).

According to Zhang et al. (2011), AAL-toxin induced PCD in tomato leaves is promoted by both jasmonic acid and ethylene by disrupting sphingolipid metabolism. While, the experimental finding of Akamatsu et al. (1997) states that AAL-toxin deficient REMI mutants are non-pathogenic to sensitive tomato plants. AAL-toxins also induce apoptotic-like responses. The induced PCD was taken place by DNA laddering, TUNEL-positive cells and the formation of apoptotic like bodies (Tsuge et al., 2013). AAL-toxin induced PCD involves ceramide signaling and cell cycle disruption (Wolpert et al., 2002). The physiological effects of AAL-toxins represent development of necrotic lesions on fruits and leaves, inhibition of in-vitro development of calli, pollen, roots and shoots, and also reduce the viability of protoplasts and suspension cells (Ismaiel and Papenbrock, 2015).

AM-toxin is other type of HSTs, which is responsible for causing Alternaria blotch on apple, a worldwide distributed disease. It is cyclic depsipeptide of alternariolide (Ueno et al., 1975, 1977). This type of cyclic depsipeptide chemical structural compound is also found in other toxins of plant pathogens i.e., HC-toxin from Cochliobolus carbonum race1 (Gross et al., 1982) and tentoxin from Alternaria tenuis (Mayer et al., 2001). Cyclic peptides are synthesized by non-ribosomal pathways by large multifunctional enzymes called non-ribosomal peptide synthetase (NRPS), and also by polymerase chain reaction (PCR) based cloning with primers having highly conserved domains of fungal NRPS genes (Keller et al., 2005; Tsuge et al., 2013). AMT1 and AMT2 are two biosynthetic genes for AM-toxin (Johnson et al., 2000; Harimoto et al., 2007). AMT1 encodes non-ribosomal peptide synthetase attaching with four catalytic domains, which is responsible for activation of each residue in AM-toxin whereas AMT2 encodes an aldo-keto reductase enzyme required for biosynthesis of 2-hydroxy-isovaleric acid i.e., one of the AM-toxin residues (Harimoto et al., 2007, 2008). Chloroplasts are the important cellular organelles which serve as the primary site for AM-toxins. On the basis of AM-toxin acting site and known pathogenesis, it is identified that Alternaria alternata f. sp. mali pathogen may secrete AM-toxins that act on the susceptible leaf cells causing tissues damage (Zhang et al., 2015).

Alternaria alternata f. sp. mali (apple pathotype) has small chromosomes having <1.9 Mb size which is not found in non-pathogenic strains of A. alternata. It suggests strongly that AMT1 resides on small chromosome of 1.1–1.8 Mb of apple pathotype strains (Johnson et al., 2001). The apple pathotype mutant strains have 1.1 Mb CDCs encoding AMT genes which are responsible for AM-toxin biosynthesis and its pathogenicity to susceptible apple strains (Johnson et al., 2001). According to Covert (1998), fungal supernumerary chromosomes are not required for growth but it has advantages for colonizing certain ecological niches (Harimoto et al., 2007).

EST (expressed sequence tag) analysis of AMT genes encoded on 1.4 Mb chromosome in the apple pathotype strain IFO 8984 was performed to identify the structure, function, and origin of Alternaria alternata, in which CDCs responsible for HSTs biosynthesis (Harimoto et al., 2007). Harimoto et al. (2007) generated a cDNA library from AM-toxin producing culture and identified 80 unigenes from 40,980 clones with 1.4 Mb chromosome probe from 196 ESTs. The sequence analysis of these genes showed that most of the small chromosomes encoding these genes with unknown function and also most of the genes expressed at remarkably low level under testing condition. Comparison of the transcription levels of the genes in toxin-producing and non-producing cultures identified 21 genes, including AMT1 and AMT2, that were up-regulated (>10 fold) in toxin producing cultures. Sequence analysis suggested that the up-regulated genes include candidates for novel AM-toxin biosynthetic genes. Disruption of three genes, AMT2, AMT3, and AMT4 also upregulated in toxin-producing cultures IFO8984 having multiple copies of the genes in the genome, and all showed similarities in structures (Tsuge et al., 2013).

AF-toxins are other types of A. alternata produced HSTs toxins having <2 Mb sized CDCs encoding AFT genes (Hatta et al., 2002). AF-toxin produced by A. alternata has three related molecular species types viz. AF-toxin I, II, and III. AF-toxin I is highly toxic to both strawberry and pear (Nishimura and Nakatsuka, 1989; Tsuge et al., 2013). Another, AF-toxin II is toxic to only pear while toxin III is highly toxic to strawberry and slightly to pear (Maekawa et al., 1984). AF-toxin I and III are valine derivatives of 2,3-dyhydroxy-isovaleric acid and 2-hydroxy isovaleric acid respectively, while AF II is an isoleucine derivative of 2-hydroxy valeric acid (Tsuge et al., 2013).

Hatta et al. (2006) concluded structure of AF-toxin, on the basis of 1.0 Mb chromosomal strain of NAF8 and found 2–7 copies of 20 AFT regions. They also found many transposon-like sequences and most of which were inactive transposon fossils. On cellular level, AF-toxin affects plasma membrane of susceptible cells and causes a sudden increase in loss of K⁺ after a few minutes of toxin treatment (Park and Ikeda, 2008). These toxin-induced dysfunctions of plasma membrane which were confirmed by microscopic study with no other cell organelles (Tsuge et al., 2013). Electrophysiological studies of AF-toxin showed that plasma membrane becomes irreversibly depolarised (Namiki et al., 1986; Otani et al., 1989). A polarization occurs mostly in the respiration-dependent component of membrane potential which is sustained by H⁺ pump. AF-toxin possibly affects the plasma membrane H⁺ATPase (Tsuge et al., 2013). However, no direct effect of AF-toxin was observed on isolated susceptible host cell plasma membrane ATPase activity (Akimitsu et al., 2013).

AK-toxins are the esters of 9,10-epoxy 8-hydroxy 9-methyldecatrienoic acid (EDA) produced by Japanese pear pathotype of A. alternata and was first reported in the Japanese pear black spot disease (Nakashima et al., 1985; Nakatsuka et al., 1986; Tsuge et al., 2013). EDA is an intermediate for toxin biosynthetic pathways. H³-labeled EDA when added in a growing liquid culture of the Japanese pear pathotype strain, converted to AK-toxin. Tanaka et al. (1999) used restriction mediated-integration transformation (REMI) to isolate AK-toxin-minus (or AK-toxin lack) mutant (Akimitsu et al., 2013). They observed that mutants affected essential AK-toxin biosynthetic genes and structural and functional analysis of clones containing tagged site represented six AK-toxin biosynthetic genes i.e., AKT1, AKT2, AKT3, AKT4, AKTR, and AKTS1 (Tanaka and Tsuge, 2000). AKTR encodes transcription regulator having a zinc binuclear cluster DNA binding domain, a protein type of fungal Zn(II)2 Cys6 family (Tanaka et al., 1999; Tanaka and Tsuge, 2000) and other fungal genes encoding proteins showed similarity with many enzymes (Tsuge et al., 2013). So, it acts as regulatory genes for AK-toxin biosynthetic enzymes.

Japanese pear pathotype strains also have multiple copies of functional or non-functional homologous of the AKT genes (Tanaka et al., 1999; Tanaka and Tsuge, 2000). AKTS1 also found unique in this pathotype in DNA blot analysis when these techniques used to assess AKT homologous distribution in A. alternata pathogen (Tanaka et al., 1999). These AKTS1 genes were also present in the tangerine and strawberry pathotype and all three pathotype share common genes required for EDA biosynthesis.

REMI techniques used for AKT protein and AKT1 gene tagging to mutagenize which is required for biosynthesis of AK-toxin and pathogenicity of the Japanese pear pathotype (Tanaka et al., 1999; Wolpert et al., 2002). Another experimental observation of Masunaka et al. (2000) stated that AKT1 and AKT2 were also present in tangerine and strawberry pathotype. The two other AKT3 and AKTR genes are also required for AK-toxin biosynthesis. AKT1, AKT2, AKT3, and AKTR and all their homologous are present on a single chromosome (Tanaka and Tsuge, 2000). Imazaki et al. (2010) reported that enzymes of AK-toxin biosynthesis are localized in peroxisomes and peroxisomal targeting signal type1 (PS1)- like tripeptides at their C-terminal ends. Mutation in AaPEX6, which encodes a peroxin protein required for peroxisome biogenesis from the Japanese pear pathotype. Lack of this peroxisome function, AK-toxin production and pathogenicity become loss completely.

ACR-toxin, a polyketide of long fatty acid (Gardner et al., 1985a,b) causes rough lemon leaf spot disease. The target site of this toxin is mitochondria which makes it dysfunctional (Kohmoto et al., 1984; Akimitsu et al., 1989). ACR-toxin causes uncoupling of oxidative phosphorylation and causes leakage of NAD⁺ cofactor from tricarboxylic acid cycle (TCA) and loss of membrane potential (Akimitsu et al., 1989).

The clustered biosynthetic genes of ACR-toxin have generally occurred on <2 Mb small sized chromosomes (Ito et al., 2004; Miyamoto et al., 2008, 2009, 2010; Ajiro et al., 2010; Izumi et al., 2012). Masunaka et al. (2005) observed that lemon pathotype strains also carried a small chromosome of ~1.5 Mb which is correlated with the toxin production and pathogenicity to rough lemon.

To investigate the relationship of ACRS to sensitivity to ACR-toxin and hence susceptibility to A. alternata rough lemon pathotype, Ohtani et al. (2002) sequenced this region of the mitochondrial genome from resistant cultivars and 13 species of citrus and found that the regions in the resistant citrus are identical to that of rough lemon (Ohtani et al., 2002). However, examination of ACRS transcripts demonstrated that sensitivity to the toxin is not controlled by the presence or absence of ACRS but rather by post-transcriptional modification of the ACRS transcripts (Figure 3). The peptide encoded by ACRS was detected by immunoblotting only in rough lemon mitochondria, but not in toxin-insensitive citrus mitochondria, and the peptide appeared to consist of sodium dodecyl sulfate (SDS)-resistant oligomers that have been reported for many pore-forming transmembrane proteins (Figure 3; Ohtani et al., 2002).

ACRTS1 and ACRTS2 are the two genes required for ACR-toxin production. Hutchinson and Fujii (1995) classified ACRTS2 as types I enzyme, is a single large and multifunctional polypeptide having all necessary enzymatic domains for multiple cycles of condensation and β-keto processing. ACRTS1 has more than three copies in the rough lemon pathotype genome. However, combination of both homologous recombination-mediated gene disruption and RNA silencing functionally suppressed all paralogs. Therefore, this gene is required for ACR-toxin biosynthesis and pathogenicity (Izumi et al., 2012), but there are some limitations of homologous recombination-mediated gene disruption. It is difficult to disrupt entire duplicated and multiple paralog functional copies of genes. RNA silencing solves this problems by knockdown all transcripts from all functional copies in HST- producing A. alternata (Miyamoto et al., 2008).

RNA silencing method was first applied to tangerine pathotype HST biosynthesis gene ACTT2, a polyketide synthase (PKS) which is similar to rough lemon pathotypes. This is silenced by transforming the pathotype strain with a plasmid construct expressing hairpin ACRT2 RNA. The ACRT2-silenced transformants in which ACRT2 transcripts were not detectable, ACR-toxins were lost the production and pathogenicity, indicating that this gene encodes a PKS essential for ACR-toxin biosynthesis and hence pathogenicity (Izumi et al., 2012).

Polyketides are the largest families of SMs synthesized by fungi, microbes and plants (Anand et al., 2010). ACT-toxin is one of the types of polyketides HSTs of tangerine pathotype. A. alternata causes brown spot disease (Timmer et al., 2000; Miyamoto et al., 2009). ACT-toxin action appears to be complex in the cell and the primary site of action for this toxin is plasma membrane (Kohmoto et al., 1993; Miyamoto et al., 2008). Kohmoto et al. (1993) detected ACT-toxins ACTT1 and ACTT2 from purified germinating conidia fluids of tangerine pathotype and expected that it participates in the biosynthesis of ACT-, AK- and AF-toxins due to their related common chemical structure (Tanaka et al., 1999; Masunaka et al., 2000, 2005).

Molecular analysis by Masunaka et al. (2000, 2005) suggests that both ACT-toxins are present on a single chromosome with a size of 1.1–1.9 and ACTT2 toxins exists in the genome as multiple copies form. Harimoto et al. (2008) used repeated round of homologous recombination-mediated gene disruption which is essential to knockout all copies of a particular target gene but it is difficult to disrupt entire copies of ACTT1 genes. Homologous gene disruption has some limitations like lack of selectable marker for multiple transformants, low efficiency etc. Harimoto et al. (2008) also overcome these limitations by RNA silencing and beneficial to knock down all functional ACTT1 genes. Since, RNA silencing induced double stranded RNA (dsRNA) post-transcriptional gene silencing phenomenon. It is cleaved by Dicer (a nuclease of the RNA III family) into small interfering RNAs (siRNA) to a ribonucleo-protein complex [RNA-induced-silencing-complex (RISC)] (Bernstein et al., 2001; Hammond et al., 2001). RISC recognizes and degrades homologous mRNAs by complementary base pairing (Elbashir et al., 2001). The RNA silencing used vectors generate hairpin structure of RNA. This structure have sense and antisense sequenced target gene with an intron sequence-based hairpin head spacer which are effective and relatively stable for sequencing, which will be useful for HSTs genes function in the other pathotype of A. alternata as well as other filamentous fungi producing HSTs (Liu et al., 2002; Kadotani et al., 2003; Miyamoto et al., 2008).

RNA silencing has never been demonstrated in Alternaria spp., so Isshiki et al. (2003) had constructed a vector for generating hairpin RNA of the green fluorescent protein (GFP) gene in A. alternata. The silencing genes were over-expressed with green fluorescence protein for determining the gene function. In addition, they analyzed the function of ACTT2 gene in ACT biosynthesis. The ACTT2 gene having multiple copies and high sequence identity were silenced by introducing a vector generated ACTT2 hairpin RNA. The sequence identity of ACTT2 and ACTT1 is 98.8% while ACTT2 homologs identity from dual toxin producing A. alternata strain also expressed similarity of 99% for ACTT2 and 98.9% ACTT3 (Miyamoto et al., 2008).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer AMR and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.